2010 Programme - Esope PDF
2010 Programme - Esope PDF
2010 Programme - Esope PDF
ESOPE 2010 Livre des Resumés _Mise en page 1 02/09/10 12:55 Page1
ESOPE
2010
Palais des Congrès
PARIS
PROGRAMME
SOMMAIRE
www.afiap.org
Organisé par /
Organised by
A.F.I.A.P.
2010
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:27 Page1
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
ESOPE 2010
Journées d’Etudes Européennes
Equipements sous Pression
European Symposium
on Pressure Equipment
Sommaire du recueil / Summary of the booklet
INDEX çç
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:27 Page2
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:27 Page3
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:27 Page4
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:27 Page5
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:27 Page6
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page7
Le planning et les présentations pourront subir de légères modifications / The plan and the presentations can be sligthly modified
Les présentations sont en traduction simultanée français-anglais / The official languages are French and English with simultaneous translation
INDEX
INDEX çç Sommaire
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page9
PLAN
SALON
352B
SALON
352A
VESTIAIRE
SALON
351
ACCU
CONG EIL
RÈS
AMPHITHÉÂTRE
HAVANE
EXPOSITION
ZONE
POSTERS
INDEX çç Sommaire
ANAIS 1 P Q_Mise en page 1 08/09/10 08:39 Page1
Fiabilité de l'image
ANAÏS répond à la norme NF_EN_14096-1, "Essais Non
Destructifs - Qualification des systèmes de numérisation
des films radiographiques". La qualité de numérisation
permet une visualisation supérieure à celle sur négatoscope.
Pérennité de l'image
ANAÏS permet d'obtenir un fichier numérique inaltérable
et copiable à l'infini.
Analyse électronique
ANAÏS intègre un grand choix de fonctions de
traitement et d'analyse de l'image, inaccessibles
avec un négatoscope : modification du contraste
et de la luminosité, mise en relief, zoom, applica-
tion de filtre.
Encombrement réduit
Un DVD ou un disque dur permet de stocker jusqu’à plusieurs
dizaines de milliers de radiographies.
Tél : 04 92 78 21 60
http:// www.anais-images.fr - e-mail : [email protected]
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page11
équipements sous pression et modélisation de l’effet de compensation par surcontrainte «, par M. Fethi DAHMENE,
Å « Fluage à 500°C d’un joint soudé d’un acier 9Cr1Mo Modifié. Evolution de la microstructure & Comportement
Å « Développements par chaos polynomial creux et adaptatif pour la propagation d’incertitudes et les analyses de
According to article 2 of its statutes, the AFIAP missions include the promotion of the education on pressure equipment technology
and the promotion of research and its applications. In this context, the AFIAP attributes traditionally every year scholarships to
students working on projects concerning pressure equipment.
On the occasion of ESOPE 2010 conference, and in coherence with the main orientation of these days, the AFIAP wished to
encourage a research work selected by a jury representing the various actors of the profession on criteria of innovation and
interest for the competitiveness and the safety of pressure equipment.
The works selected for the final award attribution are the following:
Å “Assessment of a gouge in a pipeline for transportation of natural gas and hydrogen mixture», by Mr. Julien CAPELLE,
Å “Flaw assessments in pipes under internal pressure», by Mr Hassane MOUSTABCHIR,
Å “Acoustic Emission applied to high temperature examination of pressure equipment materials and modelization of
Mr Florian VIVIER,
Å « Minimum cost assessment of reliability of structures under cyclic loading» by Mr Alban NOTIN,
Å « Adaptive sparse polynomial chaos expansions for uncertainty propagation and sensitivity analysis » by
Mr Géraud BLATMAN.
The AFIAP Award is sponsored by TOTAL Company, and will be handed on Wednesday, September 29th during the lunch to Mr.
Fethi DAHMENE, for his PhD dissertation dedicated to the Acoustic Emission applied to high temperature examination of
pressure equipment materials and modelization of compensation by overpressure.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page12
Groupe ETAI/IDEXPO
46/50 rue Benoît Malon
94250 GENTILLY
Tél. 01 41 98 40 26 – Fax. 01 41 98 40 70
*************************************************************************************
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page13
SESSIONS
Résumés / Abstracts
P - ESOPE 2010 Pr
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page14
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page15
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page16
RÉSUMÉ / ABSTRACT
103SP
ÉQUIPEMENTS SOUS PRESSION EN EUROPE : ÉVOLUTIONS
RÉGLEMENTAIRES ET PROGRAMME DE R&D
ABSTRACT
"Pressure equipment in Europe : evolution of legal framework and programmes for R&D.
The presentation relates mainly to the regulatory framework for stationary pressure equipment in the
European Union and more in particular the Pressure Equipment Directive (97/23/EC) and the Simple
Pressure Vessels Directive (SVPD, 2009/105/EC).
Both directives are in the progress of being aligned to the new legislative framework for the internal
market for goods. The presentation will provide an overview of the main changes and will highlight
the implications of this process for the pressure equipment sector."
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page17
RÉSUMÉ / ABSTRACT
104SP
INTERNATIONALISATION ET OUVERTURE : LES NOUVEAUX ENJEUX
POUR LA RÉDACTION DES RÈGLES TECHNIQUES APPLICABLES AUX
ÉQUIPEMENTS SOUS PRESSION DES RÉACTEURS NUCLÉAIRES
RÉSUMÉ
Le travail de codification des règles applicables aux équipements importants pour la sûreté a été engagé
en France dès 1978 pour accompagner la construction du parc nucléaire français. Les besoins des
industriels étaient triples: traduire les règles de conception du licencieur américain, répondre par la
codification des pratiques industrielles aux objectifs de sécurité exprimés dans la réglementation
française récemment publiée à l’époque (arrêté du 26 février 1974) et stabiliser le référentiel de travail
entre l’exploitant – prescripteur – et le constructeur, chargé d’appliquer les prescriptions techniques.
Un important travail a alors été réalisé au sein de l’association créée à cet effet, l’AFCEN (Association
Française pour les Règles de Conception, de Construction et de surveillance en exploitation des
matériels des chaudières électronucléaires), conduisant à la publication de recueils de règles dans le
domaine des matériels mécaniques des réacteurs à eau sous pression, le RCC-M et le RSE-M, mais
également dans plusieurs autres domaines techniques: matériels mécaniques des réacteurs à neutrons
rapides, RCC-MR, électricité (RCC-E), combustible (RCC-C) notamment.
Depuis, un retour d’expérience significatif a été obtenu, les contextes réglementaires européen et
français ont évolué de manière importante dans le domaine des équipements nucléaires et non
nucléaires, et le marché des équipements est devenu de plus en plus international, tant au niveau des
projets qu’à celui des approvisionnements. Du statut de «recueils» d’exigences l’apparentant à une
spécification générale, l’ensemble des exigences doit devenir davantage un code, traduction d’un état
de l’art industriel et donc ouvert à une participation plus directe des industriels et des organismes
indépendants.
Par ailleurs, le champ couvert par l’AFCEN a été étendu aux règles de protection contre l’incendie
(RCC-I) et au Génie-Civil (RCC-G et ETC-C).
Le code doit également être un outil au service des utilisateurs pour faciliter la démonstration de la
conformité des équipements à la réglementation du pays destinataire du produit. Ceci a conduit
l’AFCEN à identifier les prescriptions de base applicables quelque soit le contexte et celles à intégrer
dans des annexes non obligatoires, dédiées à un contexte réglementaire spécifique.
La communication s’attachera à présenter l’évolution de l’organisation et des processus de l’AFCEN
et illustrera l’évolution des règles techniques sur quelques exemples.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page18
RÉSUMÉ / ABSTRACT
104SP
ABSTRACT
Work on standards applicable to large safety related equipment was undertaken in France as of 1978
to accompany the construction of the French nuclear power plants. The needs of manufacturers were
threefold: translate the design rules from the American licensor, meet the safety objectives expressed
in French regulations published at that time (order of February 26, 1974) through the codification of
industrial practices, and stabilize the reference system between the operator and the contractor
Mardi
INDEX çç Sommaire
INTERNATIONALIZATION AND ENLARGEMENT: THE NEW STAKES
FOR SETTING THE TECHNICAL RULES APPLICABLE TO
THE PRESSURE EQUIPMENTS OF NUCLEAR POWER PLANTS
ABSTRACT
Work on standards applicable to large safety related equipment was undertaken in France as
of 1978 to accompany the construction of the French nuclear power plants. The needs of
manufacturers were threefold: translate the design rules from the American licensor, meet the
safety objectives expressed in French regulations published at that time (order of February 26,
1974) through the codification of industrial practices, and stabilize the reference system
between the operator and the contractor responsible for applying technical specifications.
A significant work was carried out by AFCEN (the French Association for Design,
Construction and in-service inspection rules for nuclear island components), an association
created for this purpose, leading to the publication of a collection of rules related to
mechanical equipment for pressurized water reactors, RCC-M and RSE-M, and also in several
other technical fields: particularly mechanical equipment for fast neutron reactors, RCC-MR,
electrical components (RCC-E), and fuel (RCC-C).
Since AFCEN creation, an experience of code application was gained, the regulatory context
significantly evolved in Europe and in France in the nuclear and non nuclear fields, and the
equipment market became more and more international, both for projects and products
procurement. From the status of "sets" of requirements similar to general specifications, the
rules have to become more like Codes, being the translation of an industrial state of the art
and thus opened to a more direct participation of the industrial companies and independent
bodies.
In addition, the AFCEN scope was extended to Fire Protection (RCC-I and ETC-F) and Civil
Works (RCC-G and ETC-C).
The codes shall also be tools helping users in assessing the compliance of equipments to the
regulations of the country where the product will be used, leading to identify basic mandatory
requirements applicable to every application and non-mandatory appendices dedicated to
specific regulatory contexts.
The paper will attempt to present the evolution of the organization and processes of AFCEN
and will illustrate the evolution of the technical rules on the example of pressure equipments
of nuclear power plants.
INDEX çç Sommaire
RESUME
Le travail de codification des règles applicables aux équipements importants pour la sûreté a
été engagé en France dès 1978 pour accompagner la construction du parc nucléaire français.
Les besoins des industriels étaient triples: traduire les règles de conception du licencieur
américain, répondre par la codification des pratiques industrielles aux objectifs de sécurité
exprimés dans la réglementation française récemment publiée à l’époque (arrêté du 26
février 1974) et stabiliser le référentiel de travail entre l’exploitant – prescripteur – et le
constructeur, chargé d’appliquer les prescriptions techniques.
Un important travail a alors été réalisé au sein de l’association créée à cet effet, l’AFCEN
(Association Française pour les Règles de Conception, de Construction et de surveillance en
exploitation des matériels des chaudières électronucléaires), conduisant à la publication de
recueils de règles dans le domaine des matériels mécaniques des réacteurs à eau sous
pression, le RCC-M et le RSE-M, mais également dans plusieurs autres domaines techniques:
matériels mécaniques des réacteurs à neutrons rapides, RCC-MR, électricité (RCC-E),
combustible (RCC-C) notamment.
Depuis, un retour d'expérience significatif a été obtenu, les contextes réglementaires
européen et français ont évolué de manière importante dans le domaine des équipements
nucléaires et non nucléaires, et le marché des équipements est devenu de plus en plus
international, tant au niveau des projets qu'à celui des approvisionnements. Du statut de
"recueils" d'exigences l'apparentant à une spécification générale, l'ensemble des exigences
doit devenir davantage un code, traduction d'un état de l'art industriel et donc ouvert à une
participation plus directe des industriels et des organismes indépendants.
Par ailleurs, le champ couvert par l'AFCEN a été étendu aux règles de protection contre
l'incendie (RCC-I et ETC-F) et au Génie-Civil (RCC-G et ETC-C).
Le code doit également être un outil au service des utilisateurs pour faciliter la démonstration
de la conformité des équipements à la réglementation du pays destinataire du produit. Ceci a
conduit l'AFCEN à identifier les prescriptions de base applicables quelque soit le contexte et
celles à intégrer dans des annexes non obligatoires, dédiées à un contexte réglementaire
spécifique.
La communication s'attachera à présenter l'évolution de l'organisation et des processus de
l'AFCEN et illustrera l'évolution des règles techniques sur l'exemple des équipements sous
pression des réacteurs nucléaires.
INTRODUCTION
In the context of the recent changes in AFCEN organization and working methods, it is
important to remind why the previous organizations and methods were defined as they were,
and how the evolutions are consistent with context evolution. This constitutes the first part of
this paper.
As the ESOPE conference is more particularly dedicated to pressure equipment, the second
part will deal with the particular example of the design and construction rules applicable to
mechanical equipments, covered by the RCC-M, highlighting the most recent evolutions as
well as the orientations retained on future evolutions.
HISTORICAL REMINDERS
When the French government decided to launch a large nuclear program in the context of the
first petroleum shocks, the decision was taken to follow an American licence on design
INDEX çç Sommaire
aspects and to meet French (and later European) practices as far as procurement,
manufacturing and examination were concerned.
The first three loops and four loops (P4) plants were built according to this approach, with a
lot of practical adaptations and discussions with the Safety Authority and French Utility EDF
to agree on a consistent practice. This practice was then proposed for abroad applications, in
particular in the context of the South African Koeberg offer. It appeared that such approach
was difficult to justify in front of clients and foreign authorities, and that the "photography" of
this practice should be integrated in a consistent set of codes for further contracts.
Two conditions were identified in 1976. First, EDF should accept not to refer to its set of
requirements, called "CPFC" (Cahier des Prescriptions de Fabrication et de Contrôle, or "Set
of Manufacturing and Examination Requirements") after the issuing of such codes. Second,
the initiative should be supported by the French Authority. The first condition was obtained in
1977, after several exercises and comparisons were conducted, in particular taking the ASME
code as a guide in the case of pressure equipment. The second one took the form of a request
by the Authority [1] to issue a set of codes under the following condition: after agreement
with the Safety Authority, meeting the code would imply satisfying the applicable regulation.
The letter [1] did address process, mechanical and electrical equipments and civil works.
The final approvals issued by the safety authority by Fundamental Safety Rules were limited
to the acceptability of the proposed rules in front of the applicable regulation. These safety
rules did address civil works [2], mechanical [3] and electrical [4] equipments, fuel [5] and
fire protection [6]. Process rules were program specific (900 MWe, 1300 MWe, N4) and were
evaluated through specific Safety Authority letters.
The condition provided in [1] nevertheless was the guideline of the working groups, leading
to include in the code all what was necessary to meet the applicable French regulation, and
consequently to codes integrating national and context specific provisions. The pragmatic
approach followed could be summarized as follows for nuclear pressure equipments:
RCC-M
WORKING METHOD
The orientations retained led to working groups taking the "photography" of actual practices
plus anticipated progresses with the objective to obtain rules which could be applied without
derogations. The rules should be sufficiently precise so as to present no risks of diverging
interpretations and should permit to reach a high quality at the right cost.
Groups chaired by a contractor and a utility representative were asked to reach a technical
consensus on these grounds. In order to solve difficulties, "Resolution" committees were
established. After AFCEN creation on October 19, 1980, the resolution committees became
the RCC sub-committees and a "Writing (or Editorial) committee" was established to cover
the various codes issued by AFCEN with the objective to harmonize codes evolutions.
Modification requests may be issued by every user or by the AFCEN members themselves
depending on industrial needs. In preparing the code evolutions, the AFCEN committees rely
on development results and integrate plant experience feedback, taking advantage of the
centralized organization of the French nuclear industry [9].
Requests are first analyzed by the dedicated working group depending on the concerned sub-
committee and topic (design, materials, manufacturing, examination...), then discussed and
approved by the dedicated RCC subcommittee. In case an agreement is not reached on the
working group proposal, a revised instruction sheet is issued and analyzed again by the
working group. When an agreement is reached in the RCC subcommittee, a modification
sheet to the code is prepared by the Technical Secretary of the Sub-committee on the basis of
this agreement. This Sheet is verified by the working group leaders and the sub-committee
secretary, then approved by the President of the Editorial committee, and issued by the
General Secretary by delegation of the AFCEN Board of Directors. Discussion by the Board
is only needed in cases of significant evolutions, for example the issuing of a new Section.
INDEX çç Sommaire
All these activities are conducted according to rules described in the AFCEN Quality Manual.
Periodically, the set of issued modification sheets are gathered in an Addendum, which is
published. The addendum lists the modification sheets which are included, as well as the
concerned chapters. There cannot be a single modification – even on details – included in
Code addenda, unless it has been subjected to the complete process of modification request –
modification sheet instruction. When a new edition of the code is issued, it is strictly the sum
of the previous one as modified by the successive addenda (which means by the successive
modification sheets).
The above process is followed in French language (except when discussing ETCs, which were
initially issued in English). In this case, the only official AFCEN version is the French
edition. The translation of the RCC editions and of every subsequent modification sheets is
done according to the AFCEN procedure. In case of doubts, the French edition shall be
considered, and in case of translation error, the appropriate approach is to correct the code
following the "erratum" process.
These evolutions are presented periodically to the French Safety Authority. There is no formal
approval of codes, which are never mandatory, and the French Fundamental Safety Rule
process [2]-[6] is no more applied in the new regulatory context. Nevertheless, this code
assessment facilitates the conformance assessment of equipments on topics covered by the
code.
Interpretation requests are discussed by the working groups, which issue directly an answer
transmitted to the requestor. Instructions are periodically presented to the code subcommittee.
Interpretations were not edited, although this will evolve according to user needs. There was
no "Code cases" necessity, as far as a specific agreement may be reached through the
Equipment specification in case of new materials or processes. Nevertheless, the concept of
"transitory modification sheet" was integrated in the AFCEN procedures in order to anticipate
the need for a code case approach for abroad applications.
Some modification requests may necessitate an evaluation of their consequences. This is in
particular the case when a reference to a new standard is proposed, or for the development of
a new rule. In such case, a specific enquiry may be addressed to potential suppliers and,
where appropriate, an AFCEN development study may be decided.
The various "instruction sheets" used during the request analysis process are signed by two
AFCEN founding members. These signatures testify that a consensus was obtained within the
group on the chosen decision. A consensus is considered as being reached if no participant
expresses an explicit disagreement with the proposal.
As a general rule, the process of RCC-M preparation and updating remains consistent with the
initial orientations, although it evolves to reflect the evolution of the AFCEN status and
organisation which will be presented later, in particular concerning the participation of
external experts and new members representatives.
On standards evolution, one has to remind that the codes refer to many standards. More and
more of these standards become European or ISO standards, and it is the policy of AFCEN to
refer as far as possible to international standards. Standards concern materials, examination
methods, qualification of processes or personnel, etc. In case of evolution, one has to evaluate
the potential consequences and to adapt the code. In some cases, the new standard has not the
same scope as the old one, leading to prepare self-contained specifications in case a practice is
needed, which is no more covered by the new standard. In case of examination method
evolution, one has to adapt the associated criteria so as to maintain an equivalent level of
quality.
INDEX çç Sommaire
EVOLUTION OF CONTEXT
Form the industrial point of view, as mentioned previously, the RCCs were initially
developed in the 70s and 80s on the basis of actual applications. Since 1989, in the context of
the French-German agreement which did lead to the EPRTM development, common rules were
developed from the French RCC and German KTA provisions, leading to ETC (EPR
Technical Codes) [11].
This work later led to modification requests in order to integrate in the RCC practices which
are proposed for the EPRTM project. In other cases, ETCs were kept as the basis for EPR
development, and are the support of the current AFCEN work on civil works and fire
protection.
From the regulatory point of view, discussions were conducted in Europe during several years
to harmonize regulatory requirements with the objective to suppress trade barriers within the
European Union. It appeared necessary to avoid trying to harmonize all detailed technical
regulations in Europe and to focus, following a "New Approach" defined from a Resolution of
May 7, 1985 of the European Council, on "Essential requirements" formulated in general
terms, on which agreements could more easily be obtained. As a result, such an approach did
lead to two important consequences:
- A more important weight given to the standards which can provide rules consistent with
these essential requirements. Among these standards, "harmonized standards" issued
through the European Committee for Standardization (CEN) provide a "presumption of
conformity" to the essential legal requirements in the directives;
- An important role given to independent bodies, approved through a common process by
the National states and "notified" to the European Commission, which assess the
conformity with the essential requirements in the directives.
In the particular case of pressure equipment, this process did lead to the Pressure Equipment
Directive (PED) dated May 29, 1997 [12]. Nevertheless, being issued in a context where
nuclear safety aspects remain under the responsibility of National states, nuclear pressure
equipments were considered outside the scope of the PED, allowing every national body to
issue specific nuclear safety regulations. Generally, PED essential requirements remain
nevertheless the minimum grounds for nuclear safety related equipment, safety rules taking
precedence in case of conflict.
In France, the same approach was followed under the form of a specific Order dated
December 12, 2005 [13], which refers to the December 13, 1999 Decree transposing the PED
[14] in the French regulation, and supplement this Decree on topics covering safety aspects
and potential radioactive releases.
In order to satisfy these requirements, specific industrial codes and standards can be applied,
subject to a justification demonstrating conformance with the essential requirements in the
regulation, or a reference to harmonized standards may be chosen, completed where necessary
in the nuclear field to fulfil specific rules in the nuclear regulation.
The choice between the applicable approaches depends on the following aspects:
- the practices which are well mastered by the manufacturers, and the available operation
experience,
- the standard character of the application,
- the amount of complementary provisions to be added on a given standard to fulfil the
needs,
INDEX çç Sommaire
- the amount of justifications needed to guarantee conformance with essential requirements.
This leads to maintaining existing industrial codes such as RCC in cases rules are well
adapted to the applications under consideration, and to retain a more open attitude toward
using other nuclear codes or harmonized standards appropriately supplemented, where more
standard applications are concerned.
For example, according to Order [13], nuclear pressure equipments are classified in three
groups, N1 to N3 depending on the radiological consequences of a potential failure:
- N1 group includes the Main Primary and Secondary Systems of light water reactors, except
for small pipes, and equipment the failure of which is not considered in the safety reports.
The RCC-M class 1 rules are well adapted to these components,
- N2 group includes components the failure of which would lead to an important activity
release. Small N1 primary and secondary piping are subjected to the same technical
requirements. RCC-M class 2 rules are well adapted to these components, as well as the
corresponding KTA and ASME III NC provisions,
- N3 group includes nuclear equipment not classified N1 or N2, which could lead to a
significant radioactivity release. For these equipments, the pressure risk is the most
significant and harmonized standards may be used, provided they are supplemented to
cover radioactivity prevention aspects.
The applicable technical prescriptions are the essential requirements of the PED
supplemented by specific 2005 Order demands, the additional requirements being significant
for N1 equipment and more limited for N2 and N3 equipment. Specific provisions cover:
- The hazards analyses, which have to be consistent with the Safety Report,
- The design, which shall cover the various damage risks, including irradiation effects,
- The manufacturing, which shall meet stringent requirements for N1 equipments, in
particular part technical qualifications, to cover consequences of heterogeneities on
required material properties,
- Surface and volumetric examinations, which shall cover 100% of surface and volume of
N1 and N2 equipment,
- Material properties, which shall meet minimum (elongation at fracture, toughness) or
maximum (ultimate tensile strength) values given for various grade families in appendices
specific to each equipment group.
Obviously, if one intends to issue a code consistent with the applicable regulation, the above
regulatory provisions shall lead to code updating. This was one of the important aspects of the
2007 edition of the RCC-M Code.
The evolution of the industrial and regulatory contexts did lead to enlarging the AFCEN
scope by covering civil works and fire protection, including RCC-G and ETC-C and RCC-I
and ETC-F, respectively.
Codes on mechanical equipments were also concerned by this scope evolution, as the rules on
experimental reactors, previously covered by a private CEA code, the RCC-MX, are being
integrated in the RCC-MR which is being transformed into the RCC-MRx covering new
generation reactors (including the ITER project needs) as well as experimental reactors.
The current scopes of AFCEN codes resulting from this enlargement are given in figure 2,
and the AFCEN organization is given in figure 3.
INDEX çç Sommaire
Code Scope Current edition
RCC-M Mechanical Equipments of PWR nuclear islands 2007, Addendum 2009
RCC-MR Mechanical Equipments of new generation and 2007, RCC-MRx under
experimental reactors preparation
RCC-E Electrical Equipments 2005
RCC-C Fuel 2005
ETC-F Fire protection Under preparation
RCC-G Civil works 1988, ETC-C under preparation
RSE-M In-service inspection of mechanical equipments of 1997, Addendum 2005
PWR reactors
ETC-C Sub ETC-F Sub RCC-C Sub RCC-E Sub RCC-M Sub RCC-MR Sub RSE -M Sub
Committee Committee Committee Committee Committee Committee Committee
WG WG General Design
WG General Inspection
Installations
Examination
The second orientation resulting from the evolution of the industrial and regulatory context is
the opening of AFCEN general assembly and board of directors, as well as of their code
committees to representatives of industrial partners and independent bodies.
Members from industry and third parties were previously participating as "invited experts"
depending on the scope covered. These companies are now invited to joint AFCEN with full
rights associated to such membership.
As mentioned above, the result is that code modification instruction sheets which were signed
by Areva and EDF who took decisions, are now signed by Areva and EDF as founding
members of AFCEN, but this signature only testifies that a consensus (as defined above) was
obtained on this instruction decision within the group.
As AFCEN is open to participation of foreign members, the AFCEN internet site will be used
for communication with these members and the decision to organize annual workshops is
being discussed as this would provide the opportunity to share discussions on evolutions as
well as on future evolution needs with those members who could not attend regular meetings.
INDEX çç Sommaire
A training committee has also recently been added to the AFCEN organization in order to
harmonize training content and programmes within the various companies organizing
trainings, in the context of a specific agreement with AFCEN.
It is worth mentioning that the first new industrial members to join AFCEN were generally
those interested in the new public AFCEN codes, in particular the civil works industry. A
specific workshop was organized in March 2010 [15] in order to present the new AFCEN
orientations and working methodology, followed by a Round Table and a meeting of the
enlarged AFCEN general assembly.
The general structure of the RCC-M [10] is presented in Figure 4. In compiling the RCC-M,
an analogy was deliberately created between its structure and that of the ASME Code –
Section III – Division 1 and related sections, so that persons working with either of these
codes could find the correspondence between them with minimum possible confusion.
Differences come essentially from the fact that the RCC-M scope is reduced compared to the
ASME code. When correspondence is given with RCC-M sections (for materials, welding and
NDE) in Figure 4, this applies only for equipment of nuclear islands. For other components
(conventional boilers, pressure vessels and piping), these topics are covered in the applicable
non-nuclear codes (such as National standards like CODAP or CODETI in France, or
European EN standards).
The various subsections of RCC-M section I have the same general structure and refer to
applicable provisions on procurement, manufacturing, examination or technical appendices as
illustrated in Figure 5.
The structure of Subsections A is given in Figures 6. Content and structure of Sections II, III,
IV and V are given in figures 7, 8, 9 and 10.
INDEX çç Sommaire
1000 chapters
. Scope Subsection A
. Documentation
. Identification
2000 chapters
. Prevention of corrosion
. Applicable procurement
specification Section II Materials
3000 chapters
. Sizing Subsection Z
. Analysis
4000 chapters
. Manufacturing and Section III Examination
examination
Section IV Welding
5000 chapters
. Hydrostatic tests Section V Fabrication
6000 chapters
. Overpressure protection
Chapter Title
A 1000 Objectives and structure of the Code
A 2000 General provisions
A 3000 Documents
A 4000 Components subjected to the RCC-M and component classes
A 5000 Quality Assurance
Chapter Scope
M 000 General
M 1000 Carbon steels
M 2000 Alloy steels
M 3000 Austenitic steels
M 4000 Special alloys
M 5000 Miscellaneous
M 6000 Iron castings
Chapter Scope
MC 1000 Mechanical, physical and chemical tests
MC 2000 Ultrasonic examination
MC 3000 Radiographic examination
MC 4000 Liquid penetrant examination
MC 5000 Magnetic particle examination
MC 6000 Eddy current examination of tubular products
MC 7000 Other examination methods
MC 8000 Qualification and certification of NDE personnel
MC 9000 Terminology
INDEX çç Sommaire
Chapter Scope
S 1000 General
S 2000 Acceptance of filler materials
S 3000 Welding procedure qualification
S 4000 Qualification of welders and operators
S 5000 Qualification of filler materials
S 6000 Technical qualification of production workshops
S 7000 Production welds
S 8000 Weld-deposited hardfacings on carbon, low-alloy or alloy steels
Chapter Scope
F 1000 Introduction
F 2000 Marking procedure
F 3000 Cutting repair without welding
F 4000 Forming and dimensional tolerances
F 5000 Surface treatment
F 6000 Cleanliness
F 7000 Mechanical joints
F 8000 Heat treatment (parts and components)
General: As previously mentioned, Codes continuously evolve, taking into account technical
and industrial progresses, new needs, standards updating and regulatory evolutions. The 2007
RCC-M edition did integrate the 2000 edition plus 2002, 2005 and 2007 addenda, facilitating
in particular the conformity assessment process to the regulatory demands [16]. It has been
since updated by the 2008 and 2009 addenda.
The first consequence of the European PED was a clarification of the presentation of the
RCC-M structure and the suppression of the classification rules which did lead to upgrading
the equipment class to class 3 or class 2 for components subjected to high pressure,
temperature, or cyclic loading conditions, even where they were not safety classified. This
was in contradiction with the PED and an evolution was judged appropriate taking into
account the progress made in the non-nuclear field on these topics. As a consequence,
consideration of fatigue risk is included in the specific subsections in a technically equivalent
way.
The second consequence comes from the 2005 Order [13], which classifies in the same N1
category the Main Primary and Secondary Systems. The RCC-M is consequently updated to
allow constructing the Main Secondary System according to Class 1 or to Class 2 rules,
depending on the regulatory context.
On technical aspects, an assessment of the compatibility of the nuclear codes with the PED
and 2005 Order was conducted, which led to Code evolution or to the preparation of
justification showing that RCC-M provisions are compatible with regulatory demands, in
particular where the PED allows to use "appropriate measures to achieve an equivalent overall
level of safety". This concerns in particular allowable stresses.
Concerning conformance with regulation, the RCC-M integrates provisions in the main part
of the code each time such provisions can be applied whatever the regulatory context. For
provisions specific to a given country or set of countries, non mandatory appendices are
prepared:
INDEX çç Sommaire
- To give the correspondence between regulatory provisions and code provisions,
- To provide additional requests when a given regulation has to be applied.
Two appendices ZY and ZZ are currently dedicated to the French ESPN Order and to the
European PED, respectively. This approach allows flexibility in adapting a RCC-M practice
to various regulatory contexts. It could be extended, if the need arises, to other contexts, such
as for the Chinese, South African, or UK regulations.
New definitions and new documents requested by the regulation, such as the hazard analysis
and the installation and operation instructions have also been covered.
A new subsection P was also added in the code to cover containment penetrations. In its
present stage, this subsection only provides a reference to the other applicable sections of the
code for the design and construction of the various penetration elements.
As an example of those provisions which could be split in two sets: technical provisions
which could be generalized to all equipments being integrated in the basic RCC-M sections,
whereas specific regulatory aspects are integrated in non mandatory appendices, one could
mention pressure test requirements where provisions are split between the basic B, C and
D.5000 chapters and the appendices ZY and ZZ updated according to the recent European and
French regulations. Two main reasons justify this choice:
- First of all, some components may be ordered according to the previous regulations, which
may be in conflict with the new ones (individual test pressure of 1.33 Pd instead of 1.43 Pd
for the primary system, 1.5 Pd instead of 1.43 Pd for the secondary system, for example in
France).
- In addition, previous RCC-M B, C or D.5000 requirements could be in conflict with
practices applicable in foreign countries, in particular where individual tests are covered by
system tests (as is the case of piping systems). System tests conducted according to RCC-
M could in this case be in conflict with individual equipment tests performed in shop,
where such equipment is built according to another practice and installed in the system.
In cases where there is no justification for different test practices in different countries based
on differences of quality, AFCEN has estimated that such regulatory practices should be
shifted in non-mandatory appendices (ZZ for PED, ZY for ESPN Order), and that the basic
test provisions in the mandatory sections should limit themselves to be consistent with
internationally recognized practices (in this case, the 1.25 Pd test specified in the ASME III
code). The basic test provision in the last RCC-M edition is consequently consistent with
ASME Section III, and eventual supplementary provisions shall be specified according to the
regulation in force in the concerned country. From this point of view, the RCC-M is now in
the same position as the ASME code.
It may be mentioned that the rule integrated in appendix ZZ on pressure testing is consistent
with the last modification of EN 13445 Standard [17].
Among the important RCC-M evolutions, one shall also mention the updating of the A.5000
chapter on Quality Assurance, which takes into account the recent revisions of ISO standards
as well as the IAEA rules GS-R-3/2006. The integration of a reference to Safety culture shall
be noted. This chapter could be updated in a future addendum following AFCEN discussions
in the context of the RCC-MRx under preparation for application to Fast Breeder and
Experimental Reactors.
Materials and procurement: The structure of the RCC-M material section is given in
Figure 7. Additional materials are included in the last edition, covering EPRTM project needs.
INDEX çç Sommaire
This concerns equivalent materials to existing RCC-M grades, or new grades needed for the
project. New specifications cover in particular alternatives to 17.4PH martensitic grade for
application beyond 250°C, in view of thermal ageing prevention, carbon steel tubes made of
P235GH, P265GH and P355NH grades for secondary system application, forged carbon steel
parts and fittings for secondary and auxiliary applications, and forged tubes and elbows made
from controlled nitrogen content austenitic stainless steel grades for primary and auxiliary
applications. A 20MND5 low alloy steel grade was also introduced for large primary
equipment construction. This allows procurements capable of fulfilling RCC-M and ASME
provisions, in particular where an industrial anticipation is appropriate. This concerns in
particular steam generators and pressurizers. Evolutions also deal with the introduction of
new procurement specifications, in particular RPV nozzle shell with integrated flange and
Nickel-Chromium-Molybdenum bars.
Procurement criteria have been updated to be in conformance with regulatory demands, in
particular on mechanical properties, especially elongation at fracture and toughness (figure
10), and extent of examinations.
Design and analysis: As far as sizing and analysis criteria are concerned, the regulatory
evolutions had few influences on the RCC-M. This is due to the fact that design rules are
primarily dedicated to preventing damages according to experience, and that the regulation is
expressed in flexible terms allowing the manufacturer to exercise its responsibility in the
context of the hazards analysis. The previous regulation did express margins for several
INDEX çç Sommaire
damages according to loading condition categories. The 2005 order refers to PED which
includes acceptable limits applying to the primary membrane stress under design condition. In
each case, the regulation allows using different industrial practices, provided it can be
demonstrated that they provide an equivalent level of safety.
In this context, the ASME III, KTA or RCC-M allowable stresses in particular are considered
as remaining acceptable by code committees. Dedicated justification files are prepared in
parallel for evaluation by safety authorities, to demonstrate that code rules provide
thicknesses consistent or more severe than those which would have resulted from the
application of the regulation.
One shall note that the safety coefficient between material ultimate tensile strength and
allowable stress S was changed from 4 to 3.5 in Appendix Z.III in the 2009 addendum of the
RCC-M, and the tables of allowable stresses in Appendix Z.I were updated accordingly. This
change follows the trend introduced by the ASME code for the determination of allowable
stresses S. A factor of 3 was justified by AFCEN [18] taking into account the precautions
applied in case a risk of fatigue is identified, nevertheless the final choice of 3.5 was retained
in order to improve code harmonization at an international level.
The other regulatory requirements concerning in particular the reduction of hazards related to
fatigue (low cycle, thermal fatigue, stress concentrations, vibration, local thermal-hydraulic
effects), bimetallic weld behaviour, corrosion and ageing including irradiation effects,
limitation of welds in irradiated zones and of socket welds in piping, are considered as
adequately covered by the code, which was completed in particular in the 2000 edition to
cover local thermal effects in mixing zones, following return of experience. Evolutions of
fatigue rules also concern the introduction in the last edition of specific material fatigue
curves applicable in crack-like discontinuities, such as partial penetration welds. An
alternative approach referring to conventional S-N curves is also permitted provided a plastic
correction and a reduction factor are applied to local stress ranges calculated near the
discontinuity
The design complements also concern the evolution of stress classification rules, in particular
the consideration of seismic loads, which is discussed in [19]. The approach introduced in the
RCC-M is the following. Before seismic (or other specified reversible dynamic) loads are
introduced in equation (9) of B.3652 for class 1 and equation (10) of C.3654 for class 2
components, the primary part of the inertial part of the resulting moments shall be identified.
This primary part may be directly taken as equal to the computed moment when the damping
ratio is higher than 10%, and a reduction factor may be applied to the computed moment
subject to justification if the damping ratio is lower. There is consequently a balance between
the developments which may be undertaken to justify a higher value for the damping ratio and
the developments which can justify a less severe stress classification rule. The two points of
views are linked. This applies if a linear response spectrum analysis is used, as stress
classification has only a meaning in the context of elastic analyses. Another condition is given
in the code: such analysis shall use widely broadened spectra.
Fast fracture prevention rules cover brittle (non ductile) failure and ductile tearing. Based on
experience, criteria are given which allow avoiding an explicit evaluation at design stages
when thresholds on materials and/or loadings are met. When an analysis is conducted, criteria
applied are consistent with a margin of 2 on instability risks in normal and upset conditions,
1.6 in emergency conditions, 1.2 in faulted conditions. These criteria are verified considering
conventional defects with a depth of 1/2 or 1/4 the wall thickness depending on wall
thickness, with an upper limit of 20 mm on defect depth. No crack arrest is allowed at design
stage.
INDEX çç Sommaire
Pressure test requirements are directly linked to the applicable regulation and have been
discussed above.
One point also worth mentioning is the introduction of new chapters B, C and D.6000 on
overpressure protection in the 2007 edition. These chapters cover a scope equivalent to the
one in the ASME III code, and are consistent with the French and European regulations,
referring where appropriate to European standards.
Fabrication and examination: The structures of Sections III, IV and V of the RCC-M are
given in Figures 8, 9 and 10 respectively. RCC-M fabrication and examination provisions
have evolved in parallel with the European and international standardization. Some alternative
possibilities can be addressed in the specification, covering in particular ultrasonic
examination using the so-called tandem technique examination, qualification of filler
materials, or cleanliness practices.
In the particular case of welding qualifications, self-contained RCC-M prescriptions were
replaced by a reference to EN and ISO standards, supplemented by additional prescriptions to
reach an equivalent level of guarantee. This approach was developed in particular in [20].
Qualification provisions based on industrial experience have been maintained in the code on
aspects not currently covered by harmonized standards, in particular on cladding, tube to tube
plate welding, canopy or omega joints, socket welds in piping and friction welding.
In addition, a complete revision of S.8000 on hardfacing was necessary to cover Cobalt-free
processes. In the first editions of the RCC-M, this chapter was self-contained. It was later
taken as the basis for the NF M 64-100 standard to which the following RCC-M editions did
refer. The new edition is again self-contained and integrates new Iron-based and Nickel-based
alloys grades, thus avoiding the risk of Cobalt and consequently radioactive release during
operation.
Aspects linked to regulatory conformance have also been considered, in particular on tensile
strength of filler materials. Non-destructive examination requirements have been updated, in
particular on radiographic examination of class 3 socket welds, examination of surfaces to be
welded on buttering, Eddy current examination apparatus, and roughness requirements for
ultrasonic examinations. Radiographic examination conditions for high wall thicknesses have
also been updated in order to reduce the exposure time to an acceptable value while keeping
adequate quality guarantees.
FUTURE EVOLUTIONS
In parallel with code evolutions, synthesis justifications are prepared to evaluate code
conformance with regulatory demands. The objective is to agree with the Safety Authority on
the correspondence between code provisions and essential requirements in the regulation, and
consequently on the complementary aspects to be covered in addition to the code.
Some of these aspects fall under the scope of appendices ZY or ZZ, or will be dealt with in
future appendices dedicated to other regulatory contexts. Other provisions will remain under
the responsibility of the Equipment specification, the scope and status of which will be
précised in the code.
In addition to these regulatory aspects, AFCEN analyses continuously the interpretation and
modification requests issued by code users. Evolution of procurement test sampling should
for example evolve for a better international harmonization of practices.
Future evolutions will also result from technical developments, in particular on part
fabrication (tube plates with integrated cylindrical shells, vessel heads manufactured in one
INDEX çç Sommaire
part…), non-destructive examination tools and criteria, such as the introduction of multi-
elements and TOFD techniques in the ultrasonic examination field. Numerical radiography is
also under evaluation.
Design aspects need also evolutions. This will concern the integration the results of French-
German discussions obtained in the context of ETC discussions, complements on fast fracture
prevention, flange design and improved analysis methods, linked with the need to codify the
corresponding materials properties which are necessary for their implementation.
Among the modifications which are under preparation, one can mention in particular the need
for consideration of environmental effects in fatigue evaluations. The evolutions shall
consider the current practices applied to existing projects as well as the last R&D results in
this field.
For the licensing of new plant generations in an international context, it is also important to
facilitate the licensing process by conducting comparisons between practices in various
countries. Such a work is done in the context of the MDEP (Multinational Design Evaluation
Procedure initiative of OECD) between ASME code provisions and equivalent Korean,
Japanese, or European codes, to start with. A format has been agreed on, starting with the
ASME III code structure and mentioning the corresponding sections of the other codes, such
as RCC-M. Comments on differences are integrated with the objective to identify if these
differences are justified by technical reasons, or if they are strictly related to regulatory
approaches in force in the concerned countries.
When they are of a regulatory nature, the AFCEN approach is to shift the corresponding
provisions in the non-mandatory appendices dedicated to the applicable regulatory contexts
(ZZ for the case of the European PED, ZY for the case of the French ESPN order). When they
are justified by technical reasons, one has to understand how such concern is handled
according to various codes.
In some cases, provisions covered in one code (the RCC-M for example) is covered in the
Owner's specification in the context of another code (the ASME code, for example). In such
case, one shall not forget to add provisions in the specifications when shifting from a RCC-M
practice to an ASME-based practice. Examples of additional specifications may concern
Carbon content for inter-granular corrosion prevention, limitation of Sulfur or Phosphorus
content for toughness or ageing properties, Boron content for welding properties, Cobalt
limitations for irradiation protection of personnel, specified mechanical properties, such as
toughness properties for fast fracture prevention, specification of non-destructive
examinations, or part qualification conditions where a risk of heterogeneity of chemical or
mechanical properties of large parts is identified.
In other cases, the approaches may be different due to different philosophies, but judged
globally equivalent subject to specific conditions, which may also in certain cases correspond
to contractor specifications. Typical examples are qualifications of welding processes or
welding personnel. Such practices may be recognized as globally equivalent, subject to
requirements which remain compatible with code qualification conditions, such as on pre- or
post-heat treatment conditions, welding precautions, or welding production coupons.
This work is considered important by AFCEN code committees in preparing code updating.
The objective is to retain evolutions going in the direction of code convergence while
maintaining quality objectives. Comparisons of codes shall not be limited to the code
themselves, but should integrate what is under specification responsibility according to the
code application traditions.
INDEX çç Sommaire
CONCLUSION
The RCC Codes have been established as a tool for the design and construction of the French
nuclear program and the abroad projects based on this experience. They evolve continuously,
in order to integrate the evolutions of the industrial practices, the return of experience, as well
as the needs of new projects, the numerous evolutions of the standards to which codes refer,
and the results of developments.
In order to establish these rules, AFCEN was first established by a limited number of
founding members. It has since enlarged its scope and has open its structures to members of
the concerned manufacturers and independent bodies at an international level, and has revised
its functioning rules.
The adaptation of code provisions to the new French and European regulatory evolutions did
take into account two mains concerns: the quality and cost objectives which can only be met
if codes make the best use of Manufacturers experience, and the compatibility with code
application in various regulatory contexts, within and outside Europe.
Future evolutions will result from developments, some examples of which have been given in
the paper, from standards and industrial experience, as well as from the results of international
exchanges where convergence of standards may be proposed without impairing the global
consistency of codes.
REFERENCES
[1] Letter from the Ministry of Industry (Délégation Générale à l'Energie), dated February
1st, 1978, following the Inter-ministerial Committee on Nuclear Energy, which took
place on January 17, 1978.
[2] RFS V.2.b "Règles générales applicables à la réalisation des ouvrages de génie civil"
(code RCC-G), July 30, 1981, and RFS V.2.h, June 4, 1986.
[3] RFS V.2.c "Règles générales applicables à la réalisation des matériels mécaniques"
(code RCC-M), April 8, 1981, updated September 12, 1986.
[4] RFS V.2.d "Règles générales applicables à la réalisation des matériels électriques"
(code RCC-E), December 28, 1982, updated September 23, 1986.
[5] RFS V.2.e "Règles générales applicables à la réalisation des assemblages de
combustribles" (code RCC-C), December 28, 1982, updated December 14, 1990.
[6] RFS V.2.f "Règles générales applicables à la protection contre l'incendie" (code RCC-
I), December 28, 1982.
[7] L. Valibus "General presentation of the codes used in the construction of French nuclear
power plants". Symposium on Nuclear Mechanical Components, French codification –
European prospects, SFEN – AFCEN, November 23, 1989, Paris.
[8] A. Hoffmann "The AFCEN Consultative Scientific Committee". Symposium on
Nuclear Mechanical Components, French codification – European prospects, SFEN –
AFCEN, November 23, 1989, Paris.
[9] J. Journet, W.J. O'Donnell "Operating nuclear plant feedback to ASME and French
codes" PVP-Vol.339, 1996.
[10] RCC-M Code, 2007 edition, plus 2008 and 2009 addenda, AFCEN, Paris.
[11] H.J. Frank "Development of Common Rules in the Frame of the EPR Project",
ASME/JSME/SFEN ICONE-5, paper 2275, Nice, May 1997.
INDEX çç Sommaire
[12] Directive 97/23/EC of the European Parliament and of the Council of 29 May 1997 on
the approximation of the laws of the Member States concerning pressure equipment.
[13] Arrêté du 12 décembre 2005 relatif aux équipements sous pression nucléaires. J.O. N°
19 du 22 janvier 2006.
[14] Décret n° 99-1046 du 13 décembre 1999 relatif aux équipements sous pression; Journal
Officiel de la République Française, Numéro 290 du 15 Décembre 1999.
[15] International Seminar on Rules to design and operate Nuclear Power Plants, Paris
March 29-31, 2010.
[16] JM. Grandemange, P. Renaut, D. Paris, C. Faidy, "Conformity of Nuclear Construction
Codes with the requirements of the French Order dated December 12, 2005 related to
Nuclear Pressure Equipment". ESOPE Conference, Paris, October 2007.
[17] Amendment N°10 to the harmonized European standard EN 13445 dated November
2008.
[18] JM. Grandemange "Adaptation of RCC-M design and construction rules to the
evolution of projects needs, regulatory evolutions and international exchanges"
PVP2009, July 26-30, 2009, Prague, Czech Republic.
[19] F. Le Breton, C. Petesch "Stress Analysis Criteria for Piping. RCC-M 2002 Rules and
Validation", SMiRT 20, Division 5, Paper 1790.
[20] M. Lemoine, B. Lainez, P. Anastassiades, "Rapprochement des règles relatives aux
qualifications de soudage dans les normes européennes et le RCC-M", Journées ESOPE
2007, AFIAP, Paris.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page19
RÉSUMÉ / ABSTRACT
105SP
DÉVELOPPEMENT DES MOYENS DE PRODUCTION D’ÉNERGIE
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page20
71J Développements les plus récents dans les codes et normes ASME PASTOR T. P.
dédiés aux équipements sous pression / Latest developments in ASME HSB Global Standards
pressure equipment codes and standards KARCHER G.
Consulting Engineer
SHEEHAN M.
ASME
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page21
76/82J Utilisation des normes EN 13445 et 13480 pour les équipements GRANDEMANGE J.M.
sous pression nucléaires / Use of EN 13445 and 13480 standards for PETESCH C.
nuclear pressure equipment LE BRETON F.
AREVA NP
FAIDY C.
EDF SEPTEN
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page22
RÉSUMÉ / ABSTRACT
28J
NOUVELLES ÉDITIONS DES NORMES EUROPÉENNES EN 13445 et EN 13480
NEW EDITIONS OF EUROPEAN STANDARDS EN 13445 AND EN 13480
D. Koplewicz
(UNM – F 92038 Paris La Défense cedex - e-mail : [email protected])
RÉSUMÉ
L’adoption, en mai 2002, des premières versions des EN 13445 « Récipients sous pression non soumis
à l’action de la flamme » et EN 13480 « Tuyauteries industrielles métalliques » a constitué le premier
pas de l’harmonisation technique européenne pour le secteur des équipements sous pression.
Mardi
Après 7 ans d’application, la nouvelle édition de l’EN 13445 a été rendue disponible en juillet 2009.
Cette nouvelle édition est constituée de 7 parties (une partie 8 sur les récipients en aluminium est venue
compléter l’édition 2002), et s’est enrichie de règles relatives au domaine du fluage, à la conception
par méthode expérimentale, au calcul des parois planes ; des contraintes de conception renforcées sont
introduites pour certains aciers austénitiques (voie alternative), et des règles de fatigue sont ajoutées
pour les récipients en fonte à graphite sphéroïdal ; les méthodes de prévention de la rupture fragile sont
étendues ; les contraintes de fabrication et de contrôle sont allégées : réduction du nombre de coupons-
témoins, nouvelle règle de détermination de la pression d’épreuve.
Quant à la norme EN 13480, la nouvelle édition devrait être disponible début 2011. Cette nouvelle
édition sera constituée de 7 parties (une partie 6 sur les tuyauteries enterrées et une partie 8 sur les
tuyauteries en aluminium sont venues compléter l’édition 2002), et sera enrichie de règles relatives à
la prévention de la rupture fragile, de méthodes alternatives pour la vérification des raccordements de
tubulures et pour le calcul des assemblages à brides boulonnés, de l’ajout d’une analyse de contrainte
simplifiée, et d’une nouvelle règle de détermination de la pression d’épreuve.
ABSTRACT
The adoption of the first issues of EN 13445 “Unfired pressure vessels” and EN 13480 “Metallic
industrial piping” in May 2002 was the first step of the European technical harmonisation in the
pressure equipment sector.
After a period of application of 7 years, the new edition of EN 13445 was made available in July 2009.
This new edition includes 7 Parts (a Part 8 for aluminium vessels has been added to the 2002 edition),
and is enriched with rules in the creep field, method for experimental design, formulae for flat walls ;
increased design stresses are included for certain austenitic steels (Alternative route) and fatigue rules
are added for pressure vessels made from spheroidal graphite cast iron ; methods for preventing brittle
fracture are enlarged ; manufacturing and inspection constraints are decreased : reduction of
production test plates, new rule for determining the proof test value.
Concerning EN 13480, the new edition should be available at the beginning of the year 2011. This new
edition will include 7 Parts (a Part 6 for buried piping and a Part 8 for aluminium piping have been
added to the 2002 edition) and will be enriched with rules for the prevention of brittle fracture, with
alternative methods for checking branch connections and for calculation of bolted flange connections,
with the addition of a simplified pipe stress analysis and a new rule for determining the proof test value.
INDEX çç Sommaire
NOUVELLES EDITIONS DES NORMES EUROPEENNES
EN 13445 ET EN 13480
D. Koplewicz
(UNM – F 92038 Paris La Défense cedex - e-mail : [email protected])
RESUME
L'adoption, en mai 2002, des premières versions des EN 13445 “Récipients sous pression non
soumis à la flamme” et EN 13480 “Tuyauteries industrielles métalliques” a constitué le
premier pas de l’harmonisation technique européenne pour le secteur des équipements sous
pression.
Après 7 ans d’application, la nouvelle édition de l’EN 13445 a été rendue disponible en juillet
2009. Cette nouvelle édition est constituée de 7 parties (une partie 8 sur les récipients en
aluminium est venue compléter l’édition 2002), et s’est enrichie de règles relatives au
domaine du fluage, à la conception par méthode expérimentale, au calcul des parois planes ;
des contraintes de conception renforcées sont introduites pour certains aciers austénitiques
(voie alternative), et des règles de fatigue sont ajoutées pour les récipients en fonte à graphite
sphéroïdal ; les méthodes de prévention de la rupture fragile sont étendues ; les contraintes de
fabrication et de contrôle sont allégées : réduction du nombre de coupons-témoins, nouvelle
règle de détermination de la pression d’épreuve.
Quant à la norme EN 13480, la nouvelle édition devrait être disponible début 2011. Cette
nouvelle édition sera constituée de 7 parties (une partie 6 sur les tuyauteries enterrées et une
partie 8 sur les tuyauteries en aluminium sont venues compléter l'édition 2002), et sera
enrichie de règles relatives à la prévention de la rupture fragile, de méthodes alternatives pour
la vérification des raccordements de tubulures et pour le calcul des assemblages à brides
boulonnés, de l’ajout d'une analyse de contrainte simplifiée, et d'une nouvelle règle de
détermination de la pression d’épreuve.
ABSTRACT
The adoption of the first issues of EN 13445 “Unfired pressure vessels” and EN 13480
“Metallic industrial piping” in May 2002 was the first step of the European technical
harmonisation in the pressure equipment sector.
After a period of application of 7 years, the new edition of EN 13445 was made available in
July 2009. This new edition includes 7 Parts (a Part 8 for aluminium vessels has been added
to the 2002 edition), and is enriched with rules in the creep field, method for experimental
design, formulae for flat walls ; increased design stresses are included for certain austenitic
steels (Alternative route) and fatigue rules are added for pressure vessels made from
INDEX çç Sommaire
spheroidal graphite cast iron ; methods for preventing brittle fracture are enlarged ;
manufacturing and inspection constraints are decreased : reduction of production test plates,
new rule for determining the proof test value.
Concerning EN 13480, the new edition should be available at the beginning of the year 2011.
This new edition will include 7 Parts (a Part 6 for buried piping and a Part 8 for aluminium
piping have been added to the 2002 edition) and will be enriched with rules for the prevention
of brittle fracture, with alternative methods for checking branch connections and for
calculation of bolted flange connections, with the addition of a simplified pipe stress analysis
and a new rule for determining the proof test value.
ABRÉVIATIONS UTILISÉES
INDEX çç Sommaire
INTRODUCTION
Les normes EN 13445 et EN 13480 résultent des travaux menés au sein des comités
techniques CEN/TC 54 et CEN/TC 267 dont la présidence et le secrétariat sont assurés
respectivement par le Royaume-Uni et la France.
La première édition de ces normes, en mai 2002, marque l'aboutissement de 10 ans de travail
d'une centaine d'experts européens, en provenance de l'industrie (fabricants, utilisateurs,
producteurs de matériaux), des organismes de contrôle, des laboratoires de recherche et des
laboratoires universitaires, des pouvoirs publics et des experts en normalisation. Elle constitue
également la première étape d'un processus continu de développement et d'amélioration.
Les normes EN 13445 et EN 13480 incluent des dispositions innovantes, fournissent des
solutions aux problématiques actuelles, et ne se contentent pas de juxtaposer les approches
retenues préalablement dans les différents codes européens. Pour maintenir le haut niveau de
qualité de ces normes, dont l'application permet de bénéficier de la présomption de conformité
à la réglementation européenne (directive 97/23/CE), des dispositions de mise à jour
permanente ont été adoptées par le CEN. Pour les aspects techniques, cette mise à jour est
réalisée au sein des comités CEN/TC 54 et CEN/TC 267 ; les questions de correction,
compréhension et amélioration rédactionnelle sont traitées par deux bureaux d'aide, l'EN
13445/MHD [1 ] et le CEN/TC 267/WG 8 [2].
EN 13445
La première version (mai 2002) de l'édition 2002 de la norme EN 13445 contenait 6 parties,
complétées par un document informatif, le CR 13445-7 :
Partie 1 : Généralités
Partie 2 : Matériaux
Partie 3 : Conception
Partie 4 : Fabrication
Partie 5 : Inspection et contrôle
Partie 6 : Exigences pour la conception et la fabrication des récipients sous pression et
parties sous pression moulés en fonte à graphite sphéroïdal
CR 13445-7 : Guide sur l’utilisation des procédures d’évaluation de conformité.
Entre juin 2002 et février 2009, 35 versions successives (versions 2 à 36) et deux parties
complémentaires sont venues enrichir cet ensemble. Ces différentes versions ont consisté soit
en corrections, soit en évolutions techniques du texte, ou encore en ajouts de nouvelles
dispositions, comme la conception par l’analyse ou le fluage. Les deux parties
complémentaires concernent les récipients en aluminium (EN 13445-8) et la conformité de
l’EN 13445 à l’ISO 16528 (CEN/TR 13445-9).
La nouvelle édition 2009, rendue disponible par le CEN en juillet 2009 est constituée de 7
parties, qui incorporent les amendements acceptés par le CEN et les pages corrigées jusqu’à la
INDEX çç Sommaire
version 36, sans autre modification technique complémentaire.
Les deux documents informatifs (CR 13445-7 et CEN/TR 13445-9) n’ont pas fait l’objet de
nouvelle publication.
Chacune des parties de l’EN 13445 contient une Annexe Y qui fournit des informations sur
les évolutions techniques majeures entre les deux éditions de la Norme Européenne, à savoir :
L’adjonction de règles pour les récipients et parties de récipients sous pression
fonctionnant dans le domaine du fluage. Les règles de conception font l’objet de l’Article
19 de la Partie 3, auxquelles s’ajoutent les règles complémentaires des Parties 1 à 5 de la
norme.
L’adjonction de règles pour les récipients et parties de récipients sous pression réalisés en
aluminium et alliages d’aluminium. Cette extension est contenue dans la Partie 8.
L’adjonction de règles pour la conception expérimentale, dans une nouvelle Annexe T à la
Partie 3. Ces règles ont d’abord été introduites comme Article 20 de l’EN 13445-3:2002.
L’adjonction de règles pour les parois planes, dans de nouveaux Articles 20 et 21 de la
Partie 3. Ces règles ont d’abord été introduites comme Articles 21 et 22 de l’EN 13445-
3:2002.
L’adjonction de règles pour des contraintes de conception renforcées pour certains aciers
austénitiques dans certaines conditions (voie alternative). Ces règles font l’objet du
paragraphe 6.3 de la Partie 3.
La prise en compte dans les Parties 2, 4 et 5 des récipients et parties de récipients sous
pression conçus suivant l’Annexe B de la Partie 3 (Conception par Analyse – Voie
directe).
L’adjonction de deux annexes à la Partie 1 : utilisation de la norme et index.
INDEX çç Sommaire
A l’occasion de cette nouvelle édition, ont également été passées en revue les normes de
matériaux, soudage, CND, etc, qui sont référencées dans l’EN 13445, de façon à permettre
aux utilisateurs de la norme de s’appuyer sur les documents les plus récents, dès lors que les
évolutions qu’ils contiennent n’ont pas pour effet de modifier techniquement les dispositions
de l’EN 13445. Ainsi, la norme EN 288-3 sur le soudage a pu être remplacée par son
équivalent ISO, l’EN ISO 15614.
Une nouvelle version de la norme doit être rendue disponible en juillet de chaque année. La
version juillet 2010 tient compte des demandes d’explication reçues par le bureau d’aide EN
13445/MHD [1]; elle intègre les réponses adoptées par le réseau d’experts de ce bureau
d’aide, tel qu’illustré ci-dessous :
Exemples de demandes d’explication
Référence de (2009) 1-01 (2009) 2-01
la question
Paragraphe EN 13445-1, 1.2 EN 13445-2, B.2.2.4
de la norme
Question I would be grateful if you could clarify the reason for exclusion of In the 1st paragraph, the symbol RT is not
items specifically designed for nuclear use, the failure of which may defined.
cause a release of radioactivity.
We have manufactured several vessels for use in the Nuclear
Industry to PD5500 which is the UK's standard for 'Unfired Fusion
Welded Pressure Ves els' where there is no such exclusion. My
understanding is that the intent of 13445 was to provide a standard
method of design for pressure vessels across the EU and that it
incorporates many aspects of national standards from other member
states such as France (CODAP), Germany (AD Merkblatter) etc...
We have a French customer in the Nuclear sector who wishes us to
design and build a pressure vessel in accordance with the
requirements of 13445.
It is also our understanding that manufacturers in other member
states of the EU have built vessels for Nuclear applications in
accordance with the Euro norm 13445 and have also applied CE
marking in accordance with the requirements of the Pressure
Equipment Directive (PED 97/23/EC).
Réponse The reason for the exclusion of items specifically designed for For the second issue of EN 13445-2:
nuclear use, the failure of which may cause a release of radioactivity - add at the end of the 1st paragraph of
are the following : B.2.2.4: “For other bolts and nuts than given
- the standard was developed to support the European directive in Table B.2-8 a specified impact energy of
97/23/CE which excludes such vessels minimum 40 J is required at TKV = RT for TM
- for these vessels, it is necessary to consider failure modes >= - 10 °C (where RT means Room
additional to those taken into account in the standard. Temperature)”.
However as written, the sub-clause exclusion can be misleading. - in table B.2-8, replace in the last column
CEN/TC 54 has entrusted EN 13445/MHD, by resolution 2010/628, “+ 20 °C” by “RT b)” with the following note :
to clarify the wording of subclause 1.2. “b) According to EN 10045-1:1990, the room
The following has been agreed at EN 13445/MHD meeting on 8 July temperature, RT, is equal to 23 °C +/-5 °C”
2010: Replace the content of subclause 1.2 by the following:
This European Standard is not applicable to pressure vessels of the
following types:
vessels of riveted construction;
vessels of lamellar cast iron or any other materials not included in
parts 2, 6, or 8 of the standard;
multilayered, autofrettaged or pre-stressed vessels;
This European standard may be applied to the following vessels,
provided that account is taken of additional and/or alternative
requirements resulting from the hazard analysis:
transportable vessels;
items specifically designed for nuclear use;
pressure vessels with a risk of overheating
Other European standards apply to industrial piping (EN 13480) and
to water tube and shell boilers (EN 12952 and 12953).
Furthermore, delete the titles 1.1 and 1.2
Chaque version linguistique (français, anglais, allemand) de chaque partie donne la liste des
pages corrigées de la version en cours ; les modifications sur ces pages sont identifiées par un
INDEX çç Sommaire
trait en marge ; les pages de l’EN 13445-1 et 2 correspondant aux deux questions
précédentes sont reproduites ci-dessous :
Plusieurs sujets sont actuellement discutés dans le cadre du CEN/TC 54, qui donneront lieu à
la publication d’amendements et/ou compléments à l’EN 13445 :
EN 13445-2 : rupture fragile : extension et aménagements de la méthode 2
EN 13445-3 : nouvelles règles d’exemption à la fatigue
EN 13445-3 : révision des articles 17 et 18 sur la fatigue
EN 13445-3 : domaine du fluage : analyse séparée des contraintes ; données multiaxiales
sur les matériaux ; extrapolation des données de rupture à long terme
EN 13445-3 : révision de l’article 16 sur les charges autres que la pression
EN 13445-3 : révision de l’annexe J sur les plaques tubulaires
EN 13445-4 : compléments sur les traitements thermiques après soudage
EN 13445-4 : qualification des modes opératoires d'assemblage autres que le soudage
EN 13445-5 : conditions de réalisation de l’épreuve sur base statistique
EN 13445-5 : conditions de réalisation d’une épreuve pneumatique
EN 13445-5 : contrôle des récipients fabriqués en série
EN 13445-5 : domaine du fluage : détermination par essais de la résistance des aciers à
court terme
EN 13445-10 : récipients en nickel et alliages de nickel
EN 13480
Edition 2002
La première version (mai 2002) de l'édition 2002 de la norme EN 13480 contenait 5 parties,
complétées par un document informatif, le CR 13480-7 :
Partie 1 : Généralités,
Partie 2 : Matériaux,
Partie 3 : Conception et calcul,
Partie 4 : Fabrication et installation,
INDEX çç Sommaire
Partie 5 : Inspection et contrôle,
CEN/TR 13480-7 : Guide pour l’utilisation des procédures d’évaluation de la conformité.
Une nouvelle édition de l’EN 13480 est prévue pour 2011. Elle sera constituée des 7 parties
existantes et incorporera les amendements acceptés par le CEN et les pages corrigées, sans
aucune autre modification technique. Par rapport à la version 11 de 2009, l’édition 2011
devrait intégrer trois amendements en cours d’adoption sur le système de groupement des
matériaux, les règles relatives à la prévention de la rupture fragile et la détermination de la
pression d’épreuve. La mise à jour des références normatives, en particulier pour les normes
de matériaux, sera également prise en compte.
BIBLIOGRAPHIE
[1] http://www.unm.fr/fr/general/en13445/
[2] http://portailgroupe.afnor.fr/public_espacenormalisation/CENTC267WG8/index.htm
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page23
RÉSUMÉ / ABSTRACT
43J
CODAP 2010, DIVISIONS 1, 2 ET 3
CODAP 2010, DIVISIONS 1, 2 AND 3
RÉSUMÉ
Cette nouvelle édition des Divisions 1 et 2 du CODAP propose de nombreuses évolutions qui tiennent
compte d’une part du retour d’expérience des utilisateurs et d’autre part des évolutions techniques les
plus récentes ainsi que, dans le cadre européen, des dernières interprétations réglementaires.
La partie Généralités offre notamment la possibilité de réaliser des équipements de catégorie de risque
élevée sans contrôles non destructifs hormis les contrôles visuels. Dans cette même partie, des recom-
mandations relatives à l’application du code dans le cadre de la réglementation russe ont été introduites.
Pour la partie Matériaux, en complément aux mises à jour liées aux évolutions normatives, le code sera
fourni avec une banque de données matériaux, développée en collaboration avec le CETIM, permettant
d’accéder à l’ensemble des données mécaniques des produits.
La partie Conception et calculs a été aussi complétée et développée. Notamment, deux nouvelles sec-
tions ont été ajoutées, l’une permettant de prendre en compte pour les piquages, en supplément à la
pression, les efforts extérieurs, l’autre introduit la conception par analyse sur la base des modes de dé-
faillance (DBA).
Les parties Fabrication d’une part et Contôles – Essais – Inspection ont aussi fait l’objet de nombreuses
mises à jour, en particuliers l’Annexe FA1 concernant les assemblages soudés dans l’optique d’une ho-
mogénéisation de l’ensemble des codes du SNCT ou encore l’ensemble des Annexes relatives aux
contrôles afin de prendre en compte les dernières évolutions en la matière (Radiographie, Radiographie
numérique,TOFD,… ,Phases array, Emission acoustique…).
Enfin, faisant suite et complément indispensable aux Division 1 & 2 du CODAP, une nouvelle Divion
3 a été crée avec pour objet de proposer des recommandations concernant le suivi en service et la main-
tenance de ces équipements. Ces recommandations sont destinées d’une part aux Exploitants et d’autre
part aux Constructeurs intervenants sur les équipements et sont organisées selon le plan suivant :
Partie G - Généralités spécifiant notamment le domaine d’application des présentes recommandations,
Partie SI - Inspection - Aptitude au service décrivant les inspections recommandées ainsi que les critères
d’interprétation des résultats de ces inspections,
Partie RE - Réparations des appareils,
Partie MO - Modifications des appareils,
Partie F - Fabrication et Montage définissant les exigences techniques relatives aux interventions sur
les appareils,
Partie CE - Contrôles, Essais et Inspection après réparations ou modifications.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page24
RÉSUMÉ / ABSTRACT
71J
DÉVELOPPEMENTS LES PLUS RÉCENTS DANS LES CODES ET NORMES
ASME DÉDIÉS AUX ÉQUIPEMENTS SOUS PRESSION
2008 - 2010
RÉSUMÉ
Lors des précédentes journées ESOPE, l’ASME a régulièrement tenu informé les auditeurs des dernières
évolutions et améliorations dans ses codes dédiés aux équipements sous pression. La présentation de
cette année portera sur les développements les plus importants dans le B&PV Code dédié aux
Chaudières et aux Récipients sous Pression, dans le B31 dédié aux tuyauteries et dans le BPE dédié
aux équipements de l’industrie agroalimentaire et pharmaceutique.
Les sujets abordés incluront les nouvelles technologies, les nouvelles applications et les dispositions
visant à être en conformité avec la DESP Européenne. La réorganisation des comités en charge du
développement des règles du B&PV Code sera présentée ainsi que la transition en cours de la plateforme
de publication. Les auditeurs qui utilisent les codes ASME pour la construction de leurs produits
(Fabricants) et dans les spécifications de leurs commandes (Utilisateurs) bénéficieront ainsi des
informations les plus récentes sur les évolutions des codes et sur leurs nouvelles possibilités.
ABSTRACT
In previous ESOPE events, ASME has provided the attendees with an update on the latest additions
and improvements to its codes and standards for pressure equipment. This year’s presentation will
cover major developments in the ASME Boiler and Pressure Vessel Code, B31 Pressure Piping Code,
and Bioprocessing Equipment standard. Topics will include new technologies, new applications, and
enhancements aimed at compliance with the EC PED. We will also discuss the organizational changes
that took place within the Boiler and Pressure Vessel Committees, and the transition to a new
publication software platform. Attendees who use ASME codes and standards for construction of
products (Manufacturers) or for specifications (Users) will benefit from the knowledge of the latest
developments and opportunities.
INDEX çç Sommaire
LATEST DEVELOPMENTS IN ASME
PRESSURE EQUIPMENT CODES AND STANDARDS
2008 - 2010
Thomas Pastor
HSB Global Standards
Guido Karcher
Consulting Engineer
Mark Sheehan
ASME
ABSTRACT
In previous ESOPE events, ASME has provided the attendees with an update on the latest additions and
improvements to its codes and standards for pressure equipment. This year’s presentation will cover
major developments in the ASME Boiler and Pressure Vessel Code, B31 Pressure Piping Code, and
Bioprocessing Equipment standard. Topics will include new technologies, new applications, and
enhancements aimed at compliance with the EC PED. We will also discuss the organizational changes
that took place within the Boiler and Pressure Vessel Committees, and the transition to a new publication
software platform. Attendees who use ASME codes and standards for construction of products
(Manufacturers) or for specifications (Users) will benefit from the knowledge of the latest developments
and opportunities.
INDEX çç Sommaire
2008 SIGNIFICANT REVISIONS
Section I (2008 Addenda) Weld Joint Strength Reduction Factor - These revisions are the result of a
Joint Task Group made up of B31 and BPV members, and who were charged with recommending weld
strength reduction factors for certain materials operating at elevated temperatures. There have been
several documented failures, mostly in high pressure steam piping, where it was determined that the
failure initiated in the welded joint at a point in its operating life well short of the predicted design life.
The basis for these changes is the assumption that the elevated temperature allowable stresses, which are
based on creep test of base material, may be unconservative for the weld and heat affected zone. These
revisions introduce a weld joint strength reduction factor that is now introduced into the thickness
equations for pipe, drums, headers, and heads, and effectively increases the components thickness at
elevated temperatures. This weld joint strength reduction factor only applies to certain welds, primarily
longitudinal and spiral, as well as butt welded joints within a formed head. Late in the development of
this proposal, it was decided to remove carbon steel material from the list of material affected by this
change. Further investigative work is currently under way to determine whether or not a weld joint
strength reduction factor is required for carbon steel components. These same revisions are also
introduced into B31.1, but they were not adopted by Section VIII, Division 1.
Section I (2008 Addenda), Section VIII-1 (2009 Addenda) Adopted SNT-TC-1A-2006 and CP-189-
2006 - These standards have been updated by the originating organization, requiring their references to be
updated to maintain currency. The 2006 Editions have been reviewed and found acceptable for adoption
by Sections I & VIII. The only substantial technical change is an increase in the recommended
certification period for Level I and Level II examiners from 3 years to 5 years. Both documents have been
reformatted to remove the detailed topical information for training and place this information into new
document CP-105-2006, which is referenced within the certification documents, and does not require
adoption by Section I. Ordering either of the updated certification documents includes a copy of CP-105-
2006.
Section VIII-1 (2008 Addenda) Rules for Mass-Production of Pressure Vessels - UG-90(c)(2) - This
major revision updates the rules governing mass production of pressure vessels under UG-90(c)(2). It is
recognized that in a high-volume manufacturing environment it is impracticable for the Authorized
Inspector to fulfill each of his required duties. In these situations, the Manufacturer then takes on some of
the inspection responsibilities that would normally be carried out by the AI. The requirements given in
the current UG-90(c)(2) are "sparse" at best, and over the years a number of Interpretations have been
published clarifying the intent of these rules. The primary goal of this revision was to bring into the book
many of the clarifications that have been published in Interpretations of UG-90(c)(2) over the years, and
to also incorporate the provisions of Code Case 2046, which is routinely used in UG-90(c)(2)
construction. These expanded rules were published in a new Mandatory Appendix 35. Below is a
summary of the major clarifications that took place:
1. The term "multiple duplicate vessels" was replaced by "mass produced vessels". This simple change
eliminates much of the confusion concerning when the rules of UG-90(c)(2) are applicable, since
some people thought that the construction of two or more duplicate or identical vessels qualified for
UG-90(c)(2) construction. Instead the intent of the code is that the rules of UG-90(c)(2) only apply in
a construction environment where the volume or pace of fabrication is so large that it is impracticable
for the AI to perform all of his required duties.
2. Lethal service vessels, unfired steam boilers and direct fired pressure vessels can not be constructed
under the rules of UG-90(c)(2).
3. Pressure vessels constructed under the provisions of UG-90(c)(2) shall be identical, except for
differences in fitting sizes and locations, shell lengths, and the location and configuration of non-
pressure attachments.
4. There is no size limitation on mass-produced pressure vessels.
INDEX çç Sommaire
5. The most significant clarification concerns the amount of inspection time required by the AIA during
mass production of pressure vessels. A minimum of one full-time (40 hours per week) Inspector shall
be present during mass production operations. For production less than 40 hours per week, the AI
shall be present at all times. When multiple shift mass production takes place, the required AI
presence beyond the full-time requirement shall be a matter of agreement between the AIA of Record
and the Manufacturer.
6. Appendix 35 now contains a list of responsibilities for both the AI as well as the Manufacturer, as
provided in his Inspection and Quality Control Procedure.
7. The provisions of Code Case 2046, permitting the substitution of a pneumatic test in lieu of a
hydrostatic test have been incorporated into Appendix 35.
8. Reference to ASME PCC-2 " Repair of Pressure Equipment and Piping" was added to Table U-3.
Section VIII-1 (2008 Addenda) UG-140, Overpressure Protection by System Design - Provisions for
overpressure protection by system design as published in Code Case 2211-1 have been incorporated into
new paragraph UG-140. These new rules cover two scenarios for overpressure protection by system
design: 1. the pressure in a vessel is self-limiting [UG-140(a)]; e.g. the maximum discharge pressure of a
pump or compressor is less than or equal to the maximum allowable working pressure of the vessel at the
coincident temperature. Note that the provisions of UG-140(a) may be used for vessels in air, water or
steam service. 2. When the pressure in the vessel is not self-limiting [UG-140(b)], the vessel may be
protected from overpressure by system design or by a combination of overpressure by system design and
pressure relief devices. The decision to limit the pressure in a vessel by system design is the
responsibility of the user. The user also has responsible to carry out a detailed hazard analysis to identify
and examine all potential overpressure scenarios. When the rules of UG-140 are applied, it shall be stated
on the Manufacturer's data report.
Section VIII-1 (2008 Addenda) Appendix KK - User Design Specification - Unlike Section VIII,
Division 2, Section VIII, Division 1 does not mandate the use of a Users Design Specification to provide
the detailed input to the vessel Manufacturer who is tasked with constructing the pressure vessel.
However feedback from both manufacturers as well as users of VIII-1 equipment indicates that the
industry would benefit from at least having a sample design specification available for use. Members of
the Subgroup on General Requirements who represent Users worked to develop a comprehensive users
design requirement form, and which is now published in the new Nonmandatory Appendix KK. This
Nonmandatory Appendix provides a template form that may be used to properly specify a pressure vessel
to be constructed to Section VIII, Division 1.
Section VIII-1 Cold Stretched Vessels - Case 2596 - It is known that austenitic stainless steel work
hardens and thereby increases the yield strength when deformed plastically at room temperature. The
increased yield strength results in higher allowable design stress. This concept has been used successfully
for many years in the Swedish Pressure Vessel Code and has been recognized for several years in some of
the European countries such as in Germany. In the recently developed European standard prEN 13458,
similar rules have been incorporated. Additionally, in the new ISO cryogenic vessel standard being
developed under ISO TC / 58/ 220 Committee, the same rules are being incorporated. These rules are
similar to Swedish, ADM, prEN 13458 and AS1210 Codes.
Case 2596 allows a manufacturer to fabricate pressure vessels of austenitic stainless steel (304, 316) with
allowable stresses as much as 90% higher than the current allowables published in Section II, Part D.
This is accomplished by taking advantage of the strain hardening properties of stainless steel. The
completed pressure vessel is "cold stretched" by filling with water and pressurizing to 1.5 times the
design pressure. This pressure is maintained till the strain rate has dropped to less than or equal to 0.1%
per hour, indicating that the entire vessel is plastically deformed at the applied cold stretch pressure.
There are several limitations on the use of this case with regard to material thickness, minimum and
INDEX çç Sommaire
maximum temperature, and vessel design features. This method has been used in Europe for over 25
years, and is primarily intended for vessels used in cryogenic service, although other service applications
may also benefit from this technology.
Section VIII-2, Code Case 2605 “Fatigue Evaluation of 2 ¼ Cr-1Mo-V for Operating
Temperatures Greater than 371°C (700°F) and Less Than or Equal to 454°C (850°F)” - This
recently approved code case represents a significant advance towards the design of heavy walled reactors
operating in the creep range, and subject to cyclic service. One of the shortcomings of the 2007 Edition of
VIII-2, was that the fatigue exemption screening rules in 5.5.2 could not be applied for materials with a
specified minimum tensile strength greater than 80 ksi (552 MPa) and with operating temperatures within
the creep range. This was also a shortcoming of the previous versions of Div. 2, and although we had
Case 2306, it was viewed to be inappropriate based on the larger allowable stress values in the 2007
Edition of Div. 2, so a new methodology based on a more rigorous theory was required. This case
provides fatigue screening and fatigue analysis rules for 2 1/4Cr-1Mo-V material up to 850°F (454°C).
A load histogram is required with both design and operating conditions specified. The design of the
vessel may be made using design by rule (DBR) or design by analysis (DBA) methods, and some special
design limitations are invoked. A creep analysis is required using the method given in API 579/ASME
FFS-1, Part 10. This method is based on the MPC Project Omega approach. Fatigue curves are provided
as a function of the design life as established in the creep analysis. The design fatigue curves include the
effects of creep-fatigue interaction and provide the number of permissible design cycles based on the
design life. All other design aspects associated with 2.25Cr-1Mo-V such as allowable stress, toughness
requirements, pressure-temperature ratings of standard pressure parts, and fabrication requirements,
examination requirements, and pressure test requirements are already covered in Section VIII, Division 2,
2007 Edition. The special examination requirements for vessels in fatigue service in Part 7 are also
invoked. With the approval of this important code case, the full benefit of constructing to the new VIII-2
can now be realized for heavy walled reactors used in petrochemical facilities.
Section VIII-3, Section X - Case 2579 “Composite Reinforced Pressure Vessel for Gaseous H2
Service” - This case provides the construction rules for composite reinforced pressure vessels for gaseous
hydrogen service, and is modeled after Case 2390-2 which covers hoop wrapped vessel (CRPV). Similar
to Case 2390, Manufacturers of these composite reinforced pressure vessels must hold both “U3” (Section
VIII, Division 3) and “RP” (Section X) stamps. These rules are primarily intended for the construction of
vessels to transport hydrogen gas up to an internal design pressure not exceeding 15 ksi (103 MPa).
Section I, Table PW-39 - PWHT Table for P-No.1 materials was revised to expand the current
exemptions from PWHT, based on carbon equivalence, nominal thickness, and the application of preheat
when required. Under General Note (a) (1), in the 2008 Addenda, PWHT was mandatory for all
longitudinal butt welds regardless of thickness. The revision in the 2009 addenda now extends the
exemption based on thickness up to ¾ inch for all groove butt or combination groove and fillet welds,
independent of whether it is a circumferential or longitudinal joint. Similar exemption criteria were
incorporated into the rules of ASME B31.1 2004 Edition (2006 Addenda) and Section VIII, Division 2
2007 Edition. PW-40.2 was revised to complement the revisions of Table PW-39 under this action. PW-
39.3 was revised to ensure Code users are correctly applying the thickness considerations when
performing welded repairs of fabrication and material defects.
INDEX çç Sommaire
Section I - Rules for Re-Stamping of Boilers - New Mandatory Appendix III provides criteria for
reapplication of an ASME Code symbol stamp and covers the situation where the ASME Code Symbol
stamping has become indistinct or that nameplate is illegible or lost, but traceability to the original
certification can be established. When requested by the Owner, the reapplication of the ASME Code
Symbol Stamp shall be performed by the original Manufacturer of the Code item or a successor
organization. The reapplication shall be authorized and witnessed by an AI, and it shall be documented on
a "Certificate of Conformance for Reapplication of the ASME Code Symbol Stamp" (Form III-1A). This
reapplication of the ASME Code Symbol is viewed only as restoring evidence of original compliance
with ASME Section I requirements.
One example to demonstrate the need for these rules is the case of establishing the MAWP of tubesheets.
Mandatory rules for the design of tubesheets in Part UHX were not published until the 2003 Addenda.
Prior to 2003, Manufacturers would invoke U-2(g) to design and establish the pressure/temperature rating
for their tubesheets, and one option they could have used is a proof test per UG-101. But such a proof test
could not be used today since mandatory design rules now exist for tubesheets in Part UHX. The bottom
line is that Manufacturers should review Code revisions on an annual basis to determine if any of these
changes affect past proof tests. If the review indicates that a change does affect a past proof test, then the
Manufacturer may need to update the proof test result, or in some cases conduct a new proof test. As part
of this action, revisions were made to Paragraphs 10-13(b)(3) and 10-15(d) adding Proof Test Reports to
the list of documents that must be stored and made available to the Authorized Inspector. Also Reference
Notes 39 and 40 were added to Line 18 of Form U-1, and an entry for Proof Test was added to Line 9 on
INDEX çç Sommaire
Form U-1A. In summary, vessel Manufacturers should plan on reviewing all Proof Test Reports against
current code rules to determine that these proof tests are still valid for use today. They can document their
findings on a Supplement to the original Proof Test Report. It can be expected that the Authorized
Inspector will look for evidence of this review and confirmation from the Manufacturer (Supplement to
Proof Test Report) that any proof test used to establish the MAWP for a vessel or part meets all current
Code requirements.
Section IX, P-No. Assignment for Creep Strength Enhanced Ferrous Alloys - With the incorporation
of the new Creep-Strength Enhanced Ferritic (CSEF) alloys into the Code, using the existing P-Number
groupings to specify PWHT parameters can lead to variations in heat treatments that may significantly
degrade the physical properties of these alloys; causing serious safety concerns. A CSEF alloy is a family
of ferritic steels whose creep strength is enhanced by the creation of a precise condition of microstructure,
specifically martensite or bainite, which is stabilized during tempering by controlled precipitation of
temper-resistant carbides, carbo-nitrides, or other stable phases. In the 2007 Edition of the Code, only P-
No. 5B Gr. 2 (Grade 91) met this definition and was approved for Code construction. Looking forward, a
number of CSEF alloys are already in use in Code Cases and drawing near to incorporation. To prevent
confusion, a new series of P-Numbers have been developed to address CSEF alloys. All the materials
presently assigned to P-No. 5B Group 2, are now assigned to P-No. 15E, Group No. 1. Manufacturers
who currently have PQR's qualified with P-No. 5b, Group 2 will need to
update their documentation to reflect the new P-No. assignment.
Section VIII-1, Part UHX Tubesheet Calculations - The alternative rules for calculating tubesheet
shear stress published in Case 2499 were incorporated into Part UHX. These rules permit the use of a
more representative equivalent diameter to calculate shear stress, effectively reducing its magnitude.
Additional revisions to Part UHX in the 2009 Addenda include updating the iteration instructions to
return to Step 1 in all cases, and clarify the value of "m" for the Kelvin functions as dependent on "Xa"
instead of "x". Also several equations in Part UHX were revised to correct the mistake in the use of the
term which accounts for the effect of the Poisson's ratio of the shell twice. Heat exchanger Manufacturers
should note that there were extensive revisions to Part UHX in the 2009 and that all tubesheet designs to
earlier versions of the code would need to be recalculated with these updated rules.
Section VIII-1, Part UIG – Pressure Vessels Constructed of Impregnated Graphite - This new Part
UIG represents 10 years of work to codify rules used in the construction of pressure equipment made
from impregnated graphite material. Impregnated graphite material has been in industrial use for the past
60 years, primarily in the construction of chemical processing equipment where exceptional corrosion
resistance and high thermal conductivity is particularly advantageous. Another excellent feature of this
material is that it is very stable and long-lasting, and vessels constructed of this material typically remain
in service for 20 to 30 years. Major elements of chemical process equipment are manufactured from
impervious graphite, including tubes for heat exchangers, tubesheets, headers, heads and nozzles. The
rules in Part UIG represent best practice within the industry that presently produces impervious graphite
components, and by publishing them within Section VIII, Division 1, it is now possible to construct
vessels fabricated from this material with code stamping, thereby directly allowing them to be placed in
service in jurisdictions that mandate Section VIII, Division 1 as their Pressure Vessel Law. Several
manufacturers of impregnated graphite equipment from both North America and Europe participated in
developing these rules.
Paragraph UIG-2 establishes the equipment and service limitations when constructing pressure vessels
from impregnated graphite material. Manufacturers who intend to construct impregnated graphite
pressure vessels under their "U" stamp will need to have their "U" Code Symbol Stamp scope revised to
include construction with impregnated graphite (see Appendix DD, Note 2, # 9) Also note that companies
that supply material for impregnated graphite pressure vessels (raw material, impregnation agent) do not
INDEX çç Sommaire
need to be a Certificate Holder, but must supply their material with a Certified Material Qualification
Form (Form CMQ). Due to the uniqueness of impregnated graphite material insofar as establishing design
properties, several mandatory appendices are also published containing standard test methods for
establishing the strength of the material.
Section VIII-2, Annex 2A & 2B- International Register of Professional Engineers - In the current
Code, two options are available for certifying the User's Design Specification and Manufacturer’s Design
Report; (1) certification may be carried out by a Registered Professional Engineer who is registered in one
or more states of the United States of America or the provinces of Canada; (2) an Engineer experienced in
pressure vessel design and who has received the authority to perform engineering work at the location
where he performs the work. This second option was revised to also require that not only shall the
Engineer identify the location where he will perform work but also the name of the licensing or
registering authority under which he has received the authority to perform the work. A third option has
been added under this action, whereby engineers who are registered with the International Register of
Professional Engineers (IRPE) of the Engineers Mobility Forum would be allowed to perform
engineering work in support of certification of the User's Design Specification. Background information
on the Engineers Mobility Forum can be found at: www.washingtonaccord.org/EMF/.
Section VIII-2, Part 3, 3.4.6 Mandatory UT of SAW Welds in 2 ¼ Cr-1Mo-V - API members have
recently reported reheat cracking in 2 ¼ Cr-1Mo-V SAW welds in thick wall reactors. The small
transverse cracks typically occurred after the first 600°C (1100°F) heat treatment cycle. It is only seen
with the submerged arc welding (SAW) process. While the industry performs root cause analysis of this
cracking, Section VIII has added paragraphs 3.4.6 and 7.5.5.1 which require mandatory ultrasonic
examination in accordance with 7.5.5 of all SAW welds in 2 1/4Cr-1Mo-V vessels.
Section VIII-2, Annulment of Code Case 2575 - Case 2575 was originally published to allow VIII-2
Manufacturers to continue to use the 2004 Edition/2006 Addenda of VIII-2 for an 18 month transition
period following the publication of the new 2007 Edition of VIII-2. This 18 month transition period ended
on June 30, 2009 with the annulment of case 2575. Effective July 1, 2009 Manufacturers who take orders
to construct VIII-2 vessels with stamping must use the latest edition/addenda of VIII-2.
Section VIII-3, Structural Stress Method for Fatigue Analysis of Welded Joints - The Structural
Stress method for fatigue analysis of welded joints has been added to Division 3. This method was
included in the 2007 Edition of Division 2 and is also applicable to Division 3 vessels. With the change
some of the old welded steel fatigue curves have been deleted as this new method makes them obsolete.
INDEX çç Sommaire
Section IX, Redesignating S-Numbers as P-Numbers - This revision reassigned all the existing S-
Number materials in Table QW/QB-422 and Nonmandatory Appendix D as P-Number materials for
welding and brazing procedure qualification purposes and deletes all references to S-Numbers from
appropriate Part QW and Part QB paragraphs of Section IX. P-Numbers are assigned to ASME materials
to reduce the number of welding and brazing procedure qualifications required of Certificate Holders. A
number of years ago the issue of reducing the number of qualification tests required for welding/brazing
non-ASME materials used in the B31 Code for Pressure Piping and selected BPVC Code Cases was
raised. Rather than anoint those materials with P-Numbers, the decision was made to assign a new
designation “S-Numbers” to these materials. Since 1989, BPV Committee IX has been assigning S-
Numbers to mostly ASTM materials in the B31 Codes and BPVC Code Cases.
Since the criteria for both P-Numbers and S-Numbers are based on chemical composition, mechanical
properties, weldability and brazeability, BPV Committee IX questioned the need for separating ASME
and non-ASME materials into P- and S-Numbers. The assignment of a material as a P-Number (rather
than an S-Number) rests solely on whether or not the base material has design allowable stress values in
Section II Part D or B31. The reality is that whether or not a base material has design allowable stresses,
does not alter the welding characteristics of that material. With this in mind, it was decided to delete the
S-Numbers and classify all weldable materials identified for use in pressure retaining or component
support construction, regardless of material specification, with P-Numbers.
Section IX solicited and received endorsement from BPV Committees I, III, IV, VIII and B31 regarding
the replacement of S-Numbers with P-Numbers. As for Code Case N-71, BPV Committee III has decided
that this document would continue the concept of S-Numbers for material supports as a standalone Case.
Section XII, New Modal Appendix 4 for Ton Tanks - This new Modal Appendix added specific
requirements to Section XII for Ton Tanks that are intended to transport compressed gases. It carries over
all the requirements of the applicable US DOT regulations, 49 CFR 179.300. A companion Mandatory
Appendix XIV providing the requirements for forge-welding the heads to the shells of the Category
106A tanks has been added together with this addition, showing the acceptability of the Category 106A
tanks, particularly the acceptability of exempting the SA-285 Grade A material of construction from
impact testing.
Section VIII-1, UG-45, Nozzle Neck Thickness - Paragraph UG-45 is often considered one of the more
difficult code paragraphs to understand. UG-45 establishes the minimum required thickness for nozzle
necks and inspection openings, and what often challenges code users is the fact that the thickness of a
nozzle neck is not only a function of pressure and other loads acting on it, but also consideration of the
thickness of the shell or head into which the nozzle is installed, as well as the minimum thickness of
standard wall pipe. Paragraph UG-45 has been presented in a couple of different formats over the years,
but in the 2007 Edition of VIII-2 the nozzle neck sizing rules were presented in a simpler form making
use of mathematical min/max expressions instead of excessive text. This revision to UG-45 now uses the
same presentation format used in VIII-2.
INDEX çç Sommaire
Section VIII, Div. 1, Appendix 2 - Bolt Spacing Correction Factor - It is a well known fact that if too
few bolts are used on a flange it is possible to develop secondary bending stresses in between the bolts
that would lead to non-uniform compression on the gasket. Excessive bolt spacing can result in a flange
that is more prone to leakage. The well-known Taylor Forge Bulletin 502 “Modern Flange Design”
contained an equation to calculate maximum bolt spacing as well as an equation to adjust the design
moment when the bolt spacing exceeded a certain limit. Many companies invoke these rules as part of
their internal specifications, but there has always been resistance within ASME Section VIII to add these
rules to Appendix 2. Many committee members argued that “if it’s not broke, don’t fix it”, meaning that
the Appendix 2 flange design rules have been in the book now for over 45 years without this bolt spacing
correction factor, so they fail to see the need to add it at this time. However from an engineering point of
view, it is hard to ignore the presence of this additional stress on the flange and potential for leakage of
the joint. With the current provision in UG-99(g), leakage is a Code issue for flanged joints, and the
spacing limitation is appropriate. The compromise that was reached during the balloting of this item was
that the bolt spacing and correction factor rules would be introduced into Appendix 2, but they would
only become mandatory when the vessel is designated for lethal service, or when the user or his
designated agent specifies the use of these rules.
Section VIII-2, Part 4.3, Stresses in: Cone-to-Cylinder Knuckles - The 2007 Edition VIII-2 contained
an updated set of rules for designing conical transitions with knuckles. These rules were recently
reviewed and some errors were identified. This action corrects these errors and introduces several new
variables, including betaf1, betaf2, betak1 and betak2 which are the angle used in the conical transition
calculation when a flare or knuckle is present.
In addition to these global guidance statements the following specific categories were also to be
addressed:
Using the above as metrics, a major review of the Code (Boiler and Pressure Vessel Code) organizational
and operational practices were extensively reviewed and studied to optimize the way these codes are
developed and maintained.
INDEX çç Sommaire
For many years the Code has functioned with an organizational hierarchy as briefly outlined in the
following.
Note: BNCS = Board on Nuclear Codes and Standards, BPTCS = Board on Pressure Technology Codes
and Standards, TOMC = Technical Oversight Management Committee.
Based on the studies and reviews over a two year period (2007 – 2009) the following organization was
implemented in the first quarter of 2009:
Technical Liaison
BNCS BPTCS
Technical Oversight
Management Committee
These are greatly simplified organization charts as the intent here is to simple show the new balloting
work flow through the Boiler and Pressure Vessel Code. The following are some highlights of these
organizational changes:
This major realignment is in its early stages. However implementation and planning are proceeding. The
bottom line objectives of assuring safe pressure containing structures via ASME and ANSI consensus
requirements for the Codes and Standards will be maintained. In addition, technical interchanges and
liaisons between the nuclear and non-nuclear Codes and regulatory organizations (e.g., NRC, National
Board, Jurisdictions, etc.) will also be assured.
INDEX çç Sommaire
NEW PUBLICATION FORMAT
Since the 1970’s, ASME has published the Boiler and Pressure Vessel Code in a format that includes
annual addenda issued in the form of replacement pages for the loose leaf Code. Several recent
international customer surveys have identified the addenda format as a barrier to easy use of the Code.
ASME undertook a project to eliminate addenda in a way that would better serve the users of the Code,
while continuing to provide timely updates and address a problem with the publishing system. The effort
to maintain addenda service has forced ASME to retain an outdated publishing system for more than 15
years. As the years have passed, it has become more difficult to retain competent programmers in the old
system, and the learning curve for new programmers has grown steeper. Over the last two years, ASME
has undertaken a multi-million dollar project to convert from the old system to a new system that utilizes
XML protocols.
XML provides several benefits unavailable under the current system. Volunteers on ASME committees
will be able to generate pdf and MS Word files to edit in the course of processing revisions. Because
these electronic files can then be imported directly into the publishing system, there will be fewer
opportunities for errata to take place. From a management standpoint, it will be easier to outsource
certain editing task as XML is the industry standard and there are many qualified individuals who can
work with it. In addition, the XML format will potentially allow ASME to offer the Code in electronic
format sometime in the future.
Moving away from annual addenda poses some challenges, particularly in the area of timely adoption of
new technologies and innovations in the construction of pressure equipment. ASME’s process includes a
delivery vehicle known as Code Cases. A Code Case allows new materials or methods of construction to
be adopted outside of the traditional revision process. Code Cases are issued four times a year and serve
as a fast track for innovation. Considering timeliness of Code Case issuances and concerns voiced by
users of the Code over annual addenda, the ASME Boiler and Pressure Vessel Committee voted to
convert to a two year edition cycle with no addenda commencing in 2013, with the following transition
plan.
The 2010 Edition of the Boiler and Pressure Vessel Code will be issued for purchase on schedule on July
1, 2010. Subsequently, on July 1, 2011, a complete reprinting of each Code Section with one year’s
worth of revisions will be issued to purchasers of the 2010 Edition as part of their subscription. After
that, there will be no issuance of Code changes in 2012 and on July 1, 2013, a new edition will be issued
for purchase. Following the 2013 Edition, new editions will be issued for purchase every 2 years, e.g., in
2015, 2017 and so on. Steps have been taken to ensure that the cost to purchasers of the new two year
editions will not exceed that for which they would have purchased the previous three year subscription.
In this way, ASME will serve the needs of Code users while continuing to provide the highest quality of
technical material in a timely manner.
INDEX çç Sommaire
Historically, ASME has periodically identified needs for specific research projects to support the codes
and standards development process. This research was previously performed by outside organizations
with ASME support. ASME ST-LLC has helped enhance the coordination and long range planning and
management of codes and standards development activities while strengthening volunteer participation in
developing the technology for codes and standards.
ASME’s approach to standards development for emerging technologies recognizes the important role of
technically relevant standards in advancing the commercialization, enhancing consumer confidence, and
protecting public health and safety. ASME’s involvement in R&D projects helps produce results that
respond to the needs of voluntary consensus committees in developing technically relevant codes and
standards. ASME identifies and prioritizes R&D needs to help focus the use of limited resources in these
priority areas. Collaboration in R&D projects helps to minimize individual investment while maximizing
benefits.
Projects can be initiated by anyone, but require a clear scope definition, a legitimate business need,
establishment of any funding requirements, and identification of applicable code, standard or committee.
Project Initiation Requests can be submitted online at http://stllc.asme.org/Initiate_Project.cfm.
ASME ST-LLC publishes project deliverables as Standards Technology Publications (STPs), which are
available through the ASME Catalog and Digital Store. (http://catalog.asme.org/).
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page25
RÉSUMÉ / ABSTRACT
75 J
COMPARAISON DES RÈGLES CODIFIÉES D'ANALYSE À LA FATIGUE
UNE PROPOSITION POUR AMÉLIORER LES EN
RÉSUMÉ
ABSTRACT
In connection with European Pressure Equipment Directive and the corresponding national regulation
(decree and order in France), it is required to the manufacturer to perform a risk study, taking into
account all the operating conditions defined by the end user, some margins in front of fatigue risk of
pressure equipments. Fatigue is a potential degradation mechanism connected to cyclic loads (constant
or variable amplitude or fluctuations)that has been studied by different industries in numerous countries.
This paper presents a comparison of existing rules for non-nuclear pressure equipments (EN13445,
EN13480, CODAP-CODETI, ASME VIII Div. 2) and nuclear equipments (ASME III, RCCM, RCCMR).
All the different aspects will be considered: stress-strain analysis, plasticity effects, thermal and
mechanical loads, exemption rules, cumulating, damage, fatigue curves, fatigue strength reduction
factors, environmental effects.
The conclusion will proposed 3 level methods for revision of EN standards: exemption rules, simplified
analysis and detailed analysis.
INDEX çç Sommaire
Comparaison des règles codifiées
d'analyse à la fatigue
Une proposition pour améliorer les EN
RESUME
ABSTRACT
In connection with European Pressure Equipment Directive and the corresponding national regulation
(decree and order in France), it is required to the manufacturer to perform a risk study, taking into
account all the operating conditions defined by the end user, some margins in front of fatigue risk of
pressure equipments. Fatigue is a potential degradation mechanism connected to cyclic loads
(constant or variable amplitude or fluctuations)that has been studied by different industries in
numerous countries.
This paper presents a comparison of existing rules for non-nuclear pressure equipments (EN13445,
EN13480, CODAP-CODETI, ASME VIII Div. 2) and nuclear equipments (ASME III, RCCM,
RCCMR). All the different aspects will be considered: stress-strain analysis, plasticity effects, thermal
and mechanical loads, exemption rules, cumulating, damage, fatigue curves, fatigue strength reduction
factors, environmental effects.
The conclusion will proposed 3 level methods for revision of EN standards: exemption rules, simplified
analysis and detailed analysis.
INDEX çç Sommaire
INTRODUCTION
L'ensemble des Centrales Nucléaires française est soumise tant pour la partie nucléaire (circuits
primaires, secondaires, auxiliaires…) que pour la partie eau-vapeur (salle des machines, poste d'eau…)
aux exigences du décret Equipements Sous Pression dans une première phase. Une seconde phase est
requise pour les équipements sous pression nucléaire via un arrêté spécifique.
La démarche générale est donc appliquer :
- définition de la commande par l'Exploitant
- le Fabricant responsable de la conception, de l'analyse de risque, de la justification de la
conception, de la protection contre le dépassement des limites (pression, température…), de la
visite finale qui comprend une épreuve hydrostatique, de la notice d'instruction, de l'évaluation
de la conformité et de la mise sur le marché
- l'Exploitant met ensuite en service, exploite, surveille, re-qualifie, répare… en respectant une
réglementation nationale
Le projet en cours de construction à Flamanville comprend plus de 15 000 équipements sous pression
nécessitant un dimensionnement, une analyse de risque et une justification de la conception.
Des difficultés sont rencontrées à ce jour sur la prise en compte de la fatigue dans les études de
conception (analyse de risque et justification pour une durée de fonctionnement associée à des
conditions de fonctionnement)
LA FATIGUE
La fatigue est le fait de soumettre un équipement à des sollicitations variables, parfois d'amplitude
constante, mais le plus souvent d'amplitude variable. En conséquence, après un certain cyclage
apparaît une ou plusieurs fissures en peau de l'équipement. Cette fissure peu se propager au cours de la
vie de l'équipement et peu donc réduire significativement les marges initiales à la pression, en
particulier dans le cas d'une tolérance faible de la structure aux défauts.
Différentes méthodes sont disponibles dans différents Codes et Normes. Ils sont basés, en général, sur
une estimation de l'histoire de contrainte équivalente élastique (de Tresca en général, parfois de Von
Mises) transformée en histoire de déformation en divisant la contrainte par le module d'Young du
matériau. Ensuite une comparaison entre le nombre de cycles admissibles dérivée d'une courbe de
fatigue du matériau et le nombre de cycles prévus permet de déterminer un facteur d'usage FU. Lors de
cette dernière étape différents facteurs de réduction de résistance à la fatigue peuvent être considérés:
les soudures, la contrainte moyenne, l'épaisseur, l'état de surface, l'environnement…
INDEX çç Sommaire
Ke,méca = 1 + (1-n)(Sn/3Sm-1)/(m-1)/n
Ke,therm = max {1 ; 1,86(1-(1/(1.66+Sn/Sm)))}
Avec m et n étant des coefficients caractérisant la courbe de traction cyclique du matériau, 3Sm égal à
2Sy pour les principaux matériaux et Sn la contrainte linéarisée dans l'épaisseur du composant au droit
du point considéré.
Le RCC-MR propose une alternative sous forme du coefficient K qui traduit le passage de 0.3 à 0.5
du coefficient de Poisson en fonction de la bi-axialité des contraintes (bi-axialité de 1 en thermique
pur).
L'histoire des chargements n'étant en général pas connu au niveau de la conception, des règles de
combinaison des cycles et sous-cycles sont proposées dans le RCC-M et dans ses annexes (annexe
ZH). Un calcul de facteur d'usage est ensuite effectué sur la base d'un cumul linéaire de type loi de
Miner.
Analyse des zones singulières
Un ensemble de règles spécifiques a été développé pour toutes les situations de type "entaille" ou
soudure non totalement pénétrée, pour lesquelles la contrainte élastique est infinie.
Une démarche est proposée dans l'annexe ZD du RCC-M, en remplaçant la contrainte au point
concerné par la contrainte à une distance d du fond d'entaille. d (qui est de l'ordre de 50 μm) et la
courbe d'amorçage correspondante est fournie dans le code. Le calcul de cette contrainte peut être
effectué soit par un maillage local raffiné, soit par l'intermédiaire du calcul des facteurs d'intensité KI et
KII de l'entaille analysé (les formules de calcul de KI et KII sont disponibles dans le RSE-M ).
Une méthode alternative est proposée en cohérence avec le RCC-MR; elle consiste à évaluer la
déformation à 50*m et à réduire cette déformation d'un facteur 2.5 puis de faire une analyse classique
de fatigue.
NDFP < N(C1Sm) DTN < Sa(N DTN) / C2Eya DTM < Sa(N DTM) / C2 (Ey1a1 – Ey2a2)
DPN < P/C1 (Sa (N DP)/Sm) DTR < Sa (N DTR) / C2Eya DSML < Sa(NDS)
Il est difficile de faire une comparaison générale mais néanmoins quelques différences majeures
méritent approfondissement :
- des courbes de fatigue très différentes, en particulier pour les joints soudés, mais également
pour le métal de base; certaines courbes n'ont pas de limite d'endurance
- différentes méthodes d'évaluation des effets de plasticité : Ke, Kn, KeRCCM
- des données matériaux différentes : certains ne donnent que des courbes monotomnes, d'autres
proposent des courbes de traction cyclique
- des méthodes différentes : effets de rugosité, effets d'épaisseur, effets d'environnement
Finalement il est difficile pour un Fabricant de sélectionner et de défendre une méthode d'analyse avec
un conservatisme limité.
INDEX çç Sommaire
BESOINS AUX NIVEAU DE L'ETUDE DE RISQUE ET DE CONCEPTION EN FATIGUE
CONCLUSIONS
Un besoin d'harmonisation et d'élargissement des méthodes est nécessaire pour faciliter la tâche des
Fabricant au niveau de l'étude de risque, de la conception et de la notice d'instruction.
Les méthodes proposées doivent être associées à l'ensemble des données nécessaires, si possible en 3
niveaux cohérents :
- règles d'exemption
- méthodes élastiques basés sur des calculs élastique et des formulaires
- méthodes non-linéaires de mécanique de la rupture.
References
[1] Directive 97/23/EC of the European Parliament and of the Council of 29 May 1997, Official Journal L 181,
09/07/97, pp. 0001-0055.
[2] NF-EN 13445 - Récipients sous pression non soumis à la flamme,Edition Septembre 2006, AFNOR, Saint-Denis
La Plaine, France
[3] ASME Boiler & Pressure Vessel Code.
Section III : Rules for Construction of Nuclear Facilities
Section VIII Division 2 : Rules for Construction of Pressure Vessels", Edition 2007, New York
[4] RCC-M – Règles de Conception et de Construction des matériels Mécaniques des îlots nucléaires REP, Edition
Juin 2000 + addenda jusqu'à juin 2007, AFCEN, Paris La Défense, France
[5] RSE-M – Règles de Surveillance en Exploitation des matériels Mécaniques des îlots nucléaires REP, Edition Juin
1997 + addenda jusqu'en juin 2007, AFCEN, Paris La Défense, France
[6] C. Faidy, "Fatigue et rupture : comparaison des règles codifiées", ESOPE 2004, Paris, 2004
[7] P. Dong, Z. Cao, JK Hong, "Low-Cycle Fatigue Evaluation using the Weld Master S-N curve", ASME Pressure
Vessel and Piping Conference, July 23-27,2006, Vancouver, BC, Canada
[8] C. Faidy, " Status of French and uropean Fatigue programs", ASME Pressure Vessel and Piping Conference, July
23-27,2006, Vancouver, BC, Canada
[9] C. Faidy, " Status of European Thermal Fatigue procedure", ASME Pressure Vessel and Piping Conference, July
22-26,2007, San Antonio, TX, USA
[10]
INDEX çç Sommaire
fs=Fs^(0.1ln N – 0.465)
avec Fs = 1 – 0.056 (ln Rz)^0.64 . ln Rm + 0.289 (ln Rz)^0.53
Rz : rugosité de surface en m
Figure 3-a : Coefficient lié à la contrainte moyenne, fm Figure 3-b : Coefficient lié à l'état de surface, fS
pour les tôles brut de laminage (Rz = 200m)
INDEX çç Sommaire
Figure 4 : Comparaison des courbes de fatigue pour le matériau de base
Figure 5 : Comparaison des courbes de fatigue pour les joints soudés de classe 32, 56 et 100
INDEX çç Sommaire
1
Ke th = max 1
186
.
1
Sn
+ 1.66
Sm
2 méthodes : une avec courbes de fatigue spécifiques
1 2 ou l'autre avec les courbes de fatigue (S, N) du
tot matériau
prim
.=cst
3
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page26
RÉSUMÉ / ABSTRACT
76J/82J
UTILISATION DES NORMES EN 13445 ET 13480
POUR LES ÉQUIPEMENTS SOUS PRESSION NUCLÉAIRES
USE OF EN 13445 AND 13480 STANDARDS
FOR NUCLEAR PRESSURE EQUIPMENT
RÉSUMÉ
Mardi
Selon l’approche qui est en particulier à la base de la réglementation française de 2005 relative aux équi-
pements sous pression nucléaires (arrêté ESPN), ces équipements sont des équipements sous pression
pour lesquels des exigences spécifiques doivent être satisfaites. Par conséquent, du point de vue de l’in-
dustrie, les équipements sous pression nucléaires peuvent être conçus et construits :
Soit selon des codes Nucléaires,
Soit selon des règles adaptées aux équipements non nucléaires, en particulier les normes harmonisées
EN, auxquelles des dispositions spécifiques sont ajoutées sur les points où de telles normes ne sont pas
adaptées à certains besoins spécifiques aux équipements nucléaires.
La communication présente le nouveau contexte réglementaire français dans ce domaine, ainsi que la
stratégie de sélection des codes retenus en fonction du classement des équipements.
Elle traite également des prescriptions techniques complémentaires à ajouter à une référence aux normes
EN 13445 et 13480, et présentera en particulier les critères proposés pour l’évaluation des équipements
sous pression nucléaires dans les situations accidentelles postulées, lorsque la prévention du risque d’in-
stabilité plastique est le seul objectif visé. Elle traite également des dispositions alternatives qui pourraient
être nécessaires afin de retenir une approche cohérente en matière de prévention des risques de fatigue.
ABSTRACT
According to the approach which is in particular the basis of the 2005 French regulation (ESPN Order),
nuclear pressure equipments are pressure equipments for which specific requirements shall be met. Conse-
quently, from the industry point of view, nuclear pressure equipments may be designed and constructed:
Either according to Nuclear codes,
Or according to non nuclear provisions, in particular EN harmonized standards, to which specific pro-
visions are added when such standards are not dedicated to specific nuclear needs.
The paper will present the new French regulatory context in this field and will present the code strategy
retained for nuclear equipment as a function of their classification.
It will address the complementary technical provisions which have to be added to the reference to the EN
13445 and 13480 standards, and will develop in particular the criteria proposed for the evaluation of
nuclear pressure equipment subjected to faulted conditions when plastic instability prevention is the only
objective to be fulfilled. It will also address the alternative provisions which could be necessary to provide
a consistent approach when dealing with fatigue risk prevention.
INDEX çç Sommaire
76 / 82
ABSTRACT
According to the approach which is in particular the basis of the 2005 French regulation
(ESPN Order), nuclear pressure equipments are pressure equipments for which specific
requirements shall be met. Consequently, from the industry point of view, nuclear pressure
equipments may be designed and constructed:
Either according to Nuclear codes,
Or according to non nuclear provisions, in particular EN harmonized standards, to which
specific provisions are added when such standards are not dedicated to specific nuclear
needs.
The paper will present the new French regulatory context in this field and will present the
code strategy retained for nuclear equipment as a function of their classification.
It will address the complementary technical provisions which have to be added to the
reference to the EN 13445 and 13480 standards, and will develop in particular the criteria
proposed for the evaluation of nuclear pressure equipment subjected to faulted conditions
when plastic instability prevention is the only objective to be fulfilled. It will also address the
alternative provisions which could be necessary to provide a consistent approach when
dealing with fatigue risk prevention.
RESUME
Selon l'approche qui est en particulier à la base de la réglementation française de 2005
relative aux équipements sous pression nucléaires (arrêté ESPN), ces équipements sont des
équipements sous pression pour lesquels des exigences spécifiques doivent être satisfaites.
Par conséquent, du point de vue de l'industrie, les équipements sous pression nucléaires
peuvent être conçus et construits :
Soit selon des codes Nucléaires,
Soit selon des règles adaptées aux équipements non nucléaires, en particulier les normes
harmonisées EN, auxquelles des dispositions spécifiques sont ajoutées sur les points où de
INDEX çç Sommaire
telles normes ne sont pas adaptées à certains besoins spécifiques aux équipements
nucléaires.
La communication présente le nouveau contexte réglementaire français dans ce domaine,
ainsi que la stratégie de sélection des codes retenus en fonction du classement des
équipements.
Elle traite également des prescriptions techniques complémentaires à ajouter à une référence
aux normes EN 13445 et 13480, et présente en particulier les critères proposés pour
l'évaluation des équipements sous pression nucléaires dans les situations accidentelles
postulées, lorsque la prévention du risque d'instabilité plastique est le seul objectif visé. Elle
aborde également les dispositions alternatives qui pourraient être nécessaires afin de retenir
une approche cohérente en matière de prévention des risques de fatigue.
INTRODUCTION
The European Pressure Equipment Directive (PED) [1] has been issued on May 29, 1997,
with the objective to suppress trade barriers within the European Union. Article 1.3.8 of PED
states that "the following are excluded from the Scope of the Directive: ... Items specifically
designed for nuclear use, failure of which may cause an emission of radioactivity".
The objective of this exclusion, which was also included in the corresponding European
harmonized standards, was to permit specific requirements justified by safety reasons to be
specified under the responsibility of the various countries. These countries had consequently
"the right to ask for more".
The PED was transposed in the French regulation by the December 13, 1999 Decree [2]. This
decree also covers topics excluded from PED, such as in-service surveillance and nuclear
equipments, for which it makes reference to an Order, which was later issued on December
12, 2005 [3], covering Nuclear Pressure Equipments (ESPN).
The approach was retained after an evaluation of the industrial "State of the Art" practices and
their applicability to nuclear components. It was then chosen to make reference to the PED,
with additional provisions, each time it was not in conflict with nuclear safety provisions.
This had the advantage of identifying specific requirements as a function of safety objectives,
such requirements being expressed so as to be compatible with the common PED approach.
Three levels of nuclear pressure equipments were defined depending on safety significance
and potential radioactive releases, with nuclear specific requirements modulated depending on
this classification. This has led the industry to retain a different strategy for the selection of
construction codes to be used for nuclear equipments in each of these levels, which will be
presented in the first part of this paper.
When a reference to European harmonized standards is chosen, there is a necessity to provide
complements to such standards in order to cover specific nuclear needs. This is in particular
the case for accident conditions which are postulated according to the so-called "defence-in-
depth approach". A second part of this paper will be dedicated to justifying the provisions
retained in this field.
Last but not least, when a reference is made to nuclear codes for some equipment classes and
to non-nuclear codes supplemented by nuclear-specific provisions for other classes, one has to
ensure consistency between the chosen approaches. The example of the fatigue prevention
risk is addressed in a third section of this paper and is more developed in another paper during
this conference.
INDEX çç Sommaire
GENERAL FRENCH REGULATORY APPROACH ON ESPN
If one considers the relation between the probability of an equipment failure and its potential
consequences, there is a level below which the result of the risk evaluation shall be located so
as to be considered "acceptable", as illustrated on figure 1.
There are two main approaches to ensure that the risk is "acceptable":
1. The first one consists in putting all necessary precautions at the prevention level in order
to reduce damage probability to a very small value, as for large equipments the failure of
which is not considered in the safety scenarios. By definition, those equipments are
classified N1, necessitating the highest level of quality guarantees.
2. The second one consists in introducing safety systems so as to "mitigate" the
consequences of potential failures. The corresponding pressure equipments are classified
N2 or N3 depending on radioactive release in case of component failure.
Consequences
2 – Safety
1 - ESPN
rules
unacceptable
acceptable
Hazard probability
The regulation considers that the pressure risk is conveniently covered by the PED. From this
point of view, pressure equipments are classified in categories I to IV according to PED
provisions. Those which are below category I thresholds are classified "0": figure 2. Those
which are below 0.5 bar are not pressure equipments according to PED.
Cat. IV
------ Not ESPN ------
Cat. I N3 N2 N1
Cat. 0
.5 bar
------------------ Not ESP --------------------
INDEX çç Sommaire
On the other hand, the nuclear risk is considered as follows:
Pressure equipments the failure of which would lead to a radioactivity release larger than
370 MBq (this corresponds to a currently existing threshold in nuclear installations of
10 mCu) are classified N3. Those which are below this threshold are normal pressure
equipments subjected to PED without additional request, except is they are safety
classified for other reasons.
Pressure equipments the failure of which would lead to a radioactivity release larger than
370 GBq (this corresponds to a currently existing threshold in nuclear installations of
10 Cu) are classified N2.
Pressure equipments the failure of which is not considered in the Plant safety report are
classified N1.
The regulation considers consequently that:
Below 370 MBq, equipments are considered as non-nuclear as far as the pressure
equipment regulation is concerned.
Between 370 MBq and 370 GBq, for N3 equipments, the pressure risk governs and
consequently PED approach applies. According to appendix 3 of the order, the hazard
analysis shall be conducted in a way consistent with the safety report and a specific
appendix 4 of the ESPN order covering radioprotection aspects is applicable in addition to
provisions based on PED.
Above 370 GBq, for N2 equipments, the radioactivity release risk governs, leading the
order to ask for specific precautions in appendices 2, appendix 4 applying for
radioprotection aspects.
For N1, the technical quality shall be of the highest quality. This category includes
equipments of the Main Primary and Main Secondary systems and the quality level shall
allow the possibility of applying the break preclusion concept.
INDEX çç Sommaire
Q2 are equipments for which nuclear risks shall govern design and manufacturing
precautions. They group in particular N2 equipment according to ESPN order. For these
components which are more "standard", a "manufacturer oriented" approach is best suited,
leading to opening to several possible standards, but nuclear codes founded similarly on an
ASME III-based philosophy should be applied. Opening to non-nuclear industrial
standards may only be considered for small components, as recognized in such codes.
Q3 are equipments for which the pressure risk governs. This class includes in particular N3
classified nuclear pressure equipments (as well as, in Finland, the non-nuclear classified
equipment for which the PED is mandatory in Europe). For these components, industrial
standards fulfilling the pressure regulations are applicable. This includes national standards
providing equivalent guarantees, in particular for abroad applications.
Figure 3 summarizes this global approach.
INDEX çç Sommaire
Radiation protection provisions.
These various provisions shall be covered in the Equipment specification in a way consistent
with the code(s) under consideration. An example was given in [4] presenting the structure of
such specifications, as well as the conditions to be met to ensure conditions of consistency at
interfaces between equipments. One shall note that using EN standards while keeping the
consistency with other nuclear practices implies nuclear engineering capabilities in order to
help component manufacturers applying the specified provisions in a consistent way.
INDEX çç Sommaire
conditions) and applied (these conditions being considered normal, upset or exceptional
according to EN meaning, depending on equipment required function).
When no specific functional requirement is specified, the only applicable requirement is to
demonstrate that there is no increase in accident severity, i.e. there is no bursting of such
equipment. Consequently, level D criteria shall be on the safe side compared to plastic
instability threshold.
Threshold stresses for RCC-M class 1 components
Excessive deformation: Analytical models exist to define excessive deformation threshold,
identified in the RCC-M or ASME III code as equivalent to the limit load calculated using an
elastic perfectly plastic stress-strain relationship for the material.
In paragraph III.4.1 of reference [5], the “Excessive Deformation” threshold stress ED
determined according to such model is given by the following expression:
ML 4 = S y cos P Do
P 4
ED = = S y cos (1)
Ze 4 S 4 2t S
y y
This “Excessive Deformation” threshold stress ED does not include any margin on yield
strength Sy. It corresponds to the maximum load which may be supported by the pipe, with a
material without strain hardening.
Reference [5] shows that class 1 and class 2 equations were different, but similar
linearizations of expression (1): figure 4 issued from [5].
ML = Limit moment
M0 = Elastic moment
INDEX çç Sommaire
Plastic instability: Such simple model, allowing to get rid of the various stress-strain curves
of materials, does not exist for plastic instability, which corresponds to the creation of a
plastic hinge in the pipework. The approach retained in [5] was simple: as the allowable
stresses in design condition provide a safety margin of 1.5 against the excessive deformation
threshold (Sm 2/3 Sy) and the allowable stresses in faulted conditions are defined as being
equal to 2 times these allowable stresses, one may conclude that the stress criteria in the RCC-
M code for faulted conditions correspond to the limit load determined using a theoretical yield
strength equal to 2 x 2/3 Sy = 4/3 Sy, leading to the following R threshold given in paragraph
III.4.3 of reference [5]:
MR 4 4 * 3 P 4 4 * 3 P Do
R = = S y cos *
= S y cos (2)
Ze 3 4 4 S 3 4 4 2t S*
y y
Sy was replaced by Sy* defined in paragraph III.4.3 of [5], where Sy* = 1.5 Sm, in order to
take into account the fact that the allowable stress Sm is determined by applying safety
margins to the yield strength and the ultimate tensile strength.
The verification of the existing margins between stress criteria and the plastic instability
damage may be obtained through elastic-plastic analyses or using experimental analyses. The
second way was followed in [5], where code criteria were compared to available experiments.
Reference [5] explained that experiments should not be interpreted using as they are the actual
material properties, but considering the code design philosophy. According to this approach,
the experimental results shall not be compared to the limit load evaluated taking into account
the actual material yield strength, but the limit load which would be determined taking into
account the allowable stress given by the code, which includes conditions on material ultimate
tensile strength.
This consideration is the basis of the introduction of a so-called "reference material" concept
in [5], with a yield strength Sy* taken equal to 1.5 Sm.
This rises nevertheless a difficulty as a lot of experiments were interpreted without taking into
account the significance of code rules, or, worse, were conducted without measuring the Rm
value necessary for interpreting the experiments according to code philosophy. Several papers
were obliged to limit themselves to analysing results corresponding to materials for which Sm
is directed by Sy, i.e. austenitic stainless steels essentially. This is the reason why reference
[5] was limited to analysing tests on materials for which sufficient data were available.
This effect was illustrated in figures 13 and 14 of [5], which show that excessive deformation
or plastic instability loads instead of being close to the thresholds given by formulas (1) and
(2) respectively, when results are interpreted using actual material properties, are generally
located far beyond code allowable limits according to the reference material approach.
Figure 14 of [5] is given in figure 5. It was obtained by replacing for illustrative purpose the
actual Sy by 1.5 Sm, with Sm taken equal to actual mean Rm value divided by 3 to reflect the
condition between allowable stress and material UTS. The real margin would be even more
important if the specified properties were taken instead of the mean measured Rm value of the
tested materials, as mentioned in [5].
These arguments supported the objective of the RCC-M which is to provide a margin of 1.5
against excessive deformation and 2.5 against plastic instability for class 1 equipment in
design condition:
INDEX çç Sommaire
Margins against excessive deformation are proven by limit analysis and confirmed by
excessive deformation thresholds measured on actual experiments interpreted using the
reference material approach.
Margins against plastic instability are proven by experiments, interpreted using the
reference material approach, showing that the instability load is beyond the R threshold
derived from a limit analysis conducted with a theoretical yield strength taken equal to 2
Sm (4/3 x 1.5 Sm), the safety margin between this threshold and the actual plastic instability
damage being close to 1.25.
The differences between the theoretical expressions (1) and (2) and its linearization leading to
B3600 equation (9) limits in design and faulted conditions were considered non significant in
the context of the discussions of the RCC-M rules. The only significant difference was the
non conservatism of the equation for low pressure values, which was covered by a specific
limitation of the bending stresses alone in RCC-M Appendix ZF, which is equivalent to
asking for the integration of a minimum pressure value when verifying code equation.
Threshold stresses for RCC-M class 2 components
Reference [5] did limit itself to the interpretation of the reference equation for class 2
components, compared to the class 1 equation, showing that class the 2 equation is basically a
different - but roughly equivalent - linearization of the same thresholds: figure 6.
INDEX çç Sommaire
The figure 15 of [5] reproduced in figure 4 of this paper was in particular built taking into
account the basic class 2 equation, without taking into account the code condition 0.75 i 1,
which the paper proposed to suppress. If condition 0.75 i 1 is applied, the class 2 rule for the
straight pipe case is obviously more severe than the class 1 equation, as it is in that case the
same equation with a more severe stress limit (2.4 Sy* instead of 3.0 Sy* in faulted
conditions).
In addition to that, the interpretation made on figure 14 (figure 5 of this paper) of
experimental data according to the reference material concept should be updated for class 2 to
take account of the fact that Sy* is not limited to Rm/2 as in class 1, but by Rm/2,67 (if S
Rm/4), or Rm/2.33 (when the condition S Rm/3.5 is applied), showing a larger margin
between the plastic hinge and the code allowable load.
The reference equation appears thus significantly more severe in class 2 compared to class 1,
as recognized by Landers [7] for the straight pipe case.
The general conclusion of [5] was that class 2 reference equation is generally more severe
than class 1 reference equation, except for class 1 level C due to the French Regulatory
request to demonstrate a safety margin of 1.2 (instead of 1 in the ASME III code) against
excessive deformation during emergency conditions for the Main Primary System
components, thus preventing satisfactorily against the excessive deformation and plastic
instability damages risks.
EN 13480 Stress Analysis Approach applied to level D
Excessive deformation: According to EN.13480 [6], for a pipe submitted to primary loads in
exceptional conditions, it shall be verified that pipe stress is lower than a limit depending on
material characteristics. This limit may be interpreted in terms of damage prevention.
The stress limitation to be met for exceptional conditions is defined as:
INDEX çç Sommaire
P Do M
2C = + 0.75 i 1.8 f h (EN 13480 equation 12.3.3-1)
4t Ze
where:
fh is the allowable reference stress at design temperature defined by:
fh = Min(Rm/3; Sy/1.5),
Sy is the yield strength at temperature
Rm is the ultimate tensile strength at room temperature.
This expression is very similar to the RCC-M and ASME expressions for class 2 components.
It may be considered consequently as preventing against the excessive deformation damage
without margin.
Plastic instability: By analogy with criteria in RCC-M and ASME codes, a criterion on
primary loads in faulted conditions, for “Plastic Instability” prevention, may be defined as:
P Do M
2D = + 0.75 i 2.4 f h (Extension of EN 13480 equation 12.3.3-1),
4t Ze
This limit corresponds to 2 times the limit applicable for upset conditions and to 4/3 times the
excessive deformation threshold.
The pressure Pc usable in EN equation is defined by: Pc = Max (P; P0).
For straight pipes (i = 1), the corresponding maximum moment stress M is thus defined by:
M 4 PD
M = = 2.4 f h c o (3)
Ze 3 4t
If the condition 0.75 i 1 applies, this translates into:
M P D
M = = 2.4 f h c o (4)
Ze 4t
INDEX çç Sommaire
One may note that RCC-M and ASME III condition at temperature is based on Su, which may
reach 1.1 Rm as mentioned in RCC-M Appendix Z III and ASME Section II. Under the
condition that the material UTS is not reduced by more than 10% between room temperature
and design temperature, the condition on Su is not likely to affect the allowable stress. This
RCC-M and ASME III code rule leads generally to a constant allowable stress as a function of
temperature until the allowable stress is governed by the yield strength.
This is illustrated below, where the yield strength and allowable stress are given as a function
of temperature according to EN standard and RCC-M for the example of carbon steel
P235GH used for Main Steam Relief Train discharge pipe:
250
200
Sm
150
fh
Rpt
100
Sy
50
0
50 100 150 200 250 300 350
Figure 7: Comparison of yield strengths Rpt and Sy and allowable stresses fh and Sm
used for P235GH grade in EN 13480 and RCC-M respectively
The main lesson learned from this exercise is the very limited practical effect of the fact that
condition Rm/3 is limited to room temperature condition in EN standard equations, compared
to test discrepancies and safety margins discussed above. Condition 0.75 i 1 appears to
provide an additional margin for the straight pipe reference case.
It shall be remembered that this approach remains theoretical, as the applied moments will be
limited in practice depending on the deformations allowed by supports and walls and a part of
the global safety factor in faulted conditions is to be associated with the accident hypotheses
and the load combinations taken into account.
INDEX çç Sommaire
stresses. Others, like RCC-M or EN standards, identify thresholds above which fatigue
risks should be explicitly evaluated.
The second one deals with the approaches used when conducting a fatigue analysis.
Fatigue analyses generally consist in comparing calculated stress or strain ranges (D) with
allowable ranges (R) depending on numbers of cycles: D R. Practical methods
nevertheless vary in the ways D and R are determined. Some practices consist in
determining local stress/strain ranges which integrate part geometry and comparing these
ranges to material intrinsic properties, whereas other practices consist in determining load
ranges compared to fatigue curves specific to given geometrical configurations.
When considering fatigue prevention at a system level, which may be later subjected to a load
follow during operation by the plant user, one shall ensure the consistency of fatigue
prevention thresholds and practices used at design stage, in particular when referring to
different codes such as the RCC-M nuclear code and the EN standards.
Criteria for the identification of significant fatigue risks
There shall be first a difference between exemption rules which are applied in case a fatigue
evaluation is mandatory, and rules for the identification of fatigue risks.
The fatigue exemption rules are applicable when, due to safety concerns, a fatigue analysis is
generally needed, but stress cycles are not severe. They shall consequently be envelope and,
as far as they shall be simple, they may be severe as a result.
Criteria for identification of significant fatigue risks are different. They intend to detect cases
where there could be a significant fatigue risk and, due to safety classification, the applied
design rules need not generally consider explicitly this risk. Such rules may be integrated with
investment protection objectives in mind. They may consequently be more realistic.
In the RCC-M for example, a fatigue analysis is always mandatory for class 1 (Q1)
equipments. For class 2 equipments, rules for identifying significant fatigue risks are given in
C.3112 and provisions for possible rapid temperature fluctuation effects are given in C.3625
for piping systems. When fatigue risks are judged significant, a design by analysis is
mandatory (C.3200 instead of C.3300 for vessels). When applying C.3200, criteria for fatigue
exemption may nevertheless be applied according to C.3214. For RCC-M class 3 (Q3)
components, the D.3000 design chapter refers to class 2 C.3000 rules, thus referring to the
same approach as for class 2 components, as far as fatigue is concerned.
The difference between identifying fatigue risks and fatigue exemption rules also concern the
state at which they are applied:
In the case of fatigue risk identification, the choice of the material and the design rules is
not yet done and the result may have an influence on the allowable stresses (and
consequently on component thickness) and allowable geometries. The information on
transients alone are used, and the hypothesis on component thickness can only be very
simplified.
In case of fatigue exemption, the component geometry and materials are known, which
allows applying criteria on more realistic thermal differences between adjacent points in
component wall.
One shall then be very careful when comparing fatigue risk identification criteria. From this
point of view, provisions in clause 17 of EN 13445 are judged too severe when considering
startup and shutdown pressure cycle transients, and are not sufficient to cover the other
effects, such as thermal transients effects.
INDEX çç Sommaire
Criteria for the detailed fatigue analysis
Rules for fatigue evaluation of pressure cycles are given in clause 17 of EN 13445 and more
detailed fatigue evaluation rules are included in clause 18. This last clause includes two
approaches:
One referring to fatigue curves specific to typical weld geometries,
One referring to more conventional stress-strain curves similar to the approach used in the
RCC-M or ASME codes.
The code practices may be compared on several levels:
The way the elastic stresses are determined (in particular the eventual stress indices),
The effects to be covered,
The plastic strain correction factors to be applied when elastically calculated stresses
exceed the yield strength of the material,
The rules for cycles combination,
The applicable fatigue curves for equivalent material grades.
An example of comparative review is provided in a separate paper [8]. The significant
differences identified when comparing nuclear codes and EN provisions and the necessity of a
global consistency of fatigue prevention at system level leads to the necessity to issue
specifications which can limit the reference to fatigue evaluation rules in EN standards and
refer instead to specified provisions defined in order to be consistent with those applied for
the other safety classified equipments.
CONCLUSION
This paper has shown that nuclear pressure equipments are firstly pressure equipments to
which specific requirements have to be met taking into account their nuclear application.
The industrial strategy has been to rely on nuclear codes each time the corresponding
equipments were specifically designed for nuclear use or the nuclear risk was the governing
concern. For equipments less important from the safety and radioactivity release points of
view, the pressure risk is the most important and a reference to non-nuclear codes is
acceptable, provided additional provisions are specified in order to cover nuclear aspects.
In this regard, nuclear engineering companies shall play a key role in order to provide the
conditions for consistency of provisions applied to various equipment classes subjected to
various codes, in order to provide appropriate inputs to manufacturers trained to meet the
PED provisions, but with reduced engineering capabilities.
This needs in particular the specification of appropriate criteria for the analysis of accident
conditions. On this topic, the paper has presented the level D criteria proposed when applying
EN 13480 to nuclear Q3 piping. It was shown that the proposed criteria are consistent with
the plastic instability damage prevention objective, with margins which are close to RCC-M
class 1, notwithstanding the condition on 0.75 i 1.
Fatigue prevention was also briefly discussed in the paper. Due to the necessity of defining a
global consistency of fatigue prevention at system level and the significant differences
identified when comparing nuclear codes and EN provisions, specification can limit the
reference to fatigue evaluation rules in EN standards and refer instead to specified provisions
defined in order to be consistent with those applied for the other safety classified equipments.
INDEX çç Sommaire
REFERENCES
[1] Directive 97/23/EC of the European Parliament and of the Council of 29 May 1997 on
the approximation of the laws of the Member States concerning pressure equipment.
[2] Décret n° 99-1046 du 13 décembre 1999 relatif aux équipements sous pression; Journal
Officiel de la République Française, n° 290 du 15 Décembre 1999.
[3] Arrêté du 12 décembre 2005 relatif aux équipements sous pression nucléaires, J.O n° 19
du 22 janvier 2006 page 1097.
[4] JM. Grandemange, D. Kreckel "Construction codes(s) applicable for the European
Pressurized water Reactor project", ESOPE 2004 Conference, September 2004, Paris.
[5] C. Heng, JM. Grandemange "Framatome view on the comparison between Class 1 and
Class 2 RCC-M Piping Design Rules", WRC Bulletin 361, February 1991.
[6] EN 13480 Standard, Part-3. Metallic Industrial Piping. 2002 Edition.
[7] D.F. Landers "Nuclear Piping Design – A critique". Teledyne, July 12, 1978.
[8] C. Faidy "Comparison of Codified Fatigue Analysis Rules. A proposal to improve EN
Standards", ESOPE 2010 Conference, Paris.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page27
RÉSUMÉ / ABSTRACT
40 J
ÉVOLUTION DES CODES ET NORMES AU CEA, PRÉSENTATION DU
RCC-MRx ET INTERNATIONALISATION
RÉSUMÉ
Un document de travail du RCC-MRx, établi en 2009 et 2010, est aujourd’hui disponible pour validation
afin de préparer la prochaine édition de l’Afcen (Association Française pour les règles de Conception
et de Construction des Matériels des Chaudières Electro-nucléaires). Ce Code (en Français et en An-
glais) réalisera la fusion du RCC-MX 2008 développé dans le contexte du réacteur expérimental RJH
(Réacteur Jules Horowitz) et du RCC-MR 2007 qui établi les règles applicables à la conception des
composants fonctionnant à haute température.
Cet article présente la structure générale et le contexte de ces trois Codes (RM 2007, MX 2008 et MRx)
et passe en revue les règles de conception, les conditions d’approvisionnement des pièces et produits,
les examens non destructifs, les règles de soudage et les règles de fabrication.
Enfin, cet article présente la stratégie de développement du Code pour les années à venir, son intention
de s’ouvrir à des coopérations européennes et internationales.
ABSTRACT
A private draft document, established during the years 2009 and 2010 is yet available for test to prepare
the next edition by Afcen (Association Française pour les règles de Conception et de Construction des
Matériels des Chaudières Electro-nucléaires). This RCC-MRx Code (in French and English) will be
the result of the merger of the RCC-MX 2008 developed in the context of the research reactor Jules
Horowitz Reactor project, in the RCC-MR 2007 which set up rules applicable to the design of compo-
nents operating at high temperature.
This paper will outline the general structure and the scope of these three codes (MR 2007, MX 2008,
and MRx), and give an overview of design rules, procurement conditions of parts and products, non
destructive testing, welding rules and manufacturing rules.
Finally this paper will outline the development strategy of the code for the future; it is intended to open
it to European and international cooperations.
INDEX çç Sommaire
EVOLUTION DES CODES ET NORMES AU CEA, PRESENTATION DU
RCC-MRx ET INTERNATIONALISATION
EVOLUTION OF CODES AND STANDARDS IN CEA, PRESENTATION OF
RCC-MRx AND INTERNATIONALIZATION
RESUME
Un document de travail du RCC-MRx, établi en 2009 et 2010, est aujourd'hui disponible pour
validation afin de préparer la prochaine édition de l'Afcen (Association Française pour les
règles de Conception et de Construction des Matériels des Chaudières Electro-nucléaires). Ce
Code (en Français et en Anglais) réalisera la fusion du RCC-MX 2008 développé dans le
contexte du réacteur expérimental RJH (Réacteur Jules Horowitz) et du RCC-MR 2007 qui
établi les règles applicables à la conception des composants fonctionnant à haute température.
Cet article présente la structure générale et le contexte de ces trois Codes (RM 2007, MX
2008 et MRx) et passe en revue les règles de conception, les conditions d'approvisionnement
des pièces et produits, les examens non destructifs, les règles de soudage et les règles de
fabrication.
Enfin, cet article présente la stratégie de développement du Code pour les années à venir, son
intention de s'ouvrir à des coopérations européennes et internationales.
ABSTRACT
A private draft document, established during the years 2009 and 2010 is yet available for test
to prepare the next edition by Afcen (Association Française pour les règles de Conception et
de Construction des Matériels des Chaudières Electro-nucléaires). This RCC-MRx Code (in
French and English) will be the result of the merger of the RCC-MX 2008 developed in the
context of the research reactor Jules Horowitz Reactor project, in the RCC-MR 2007 which
set up rules applicable to the design of components operating at high temperature.
This paper will outline the general structure and the scope of these three codes (MR 2007,
MX 2008, and MRx), and give an overview of design rules, procurement conditions of parts
and products, non destructive testing, welding rules and manufacturing rules.
Finally this paper will outline the development strategy of the code for the future; it is
intended to open it to European and international cooperations.
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
INTRODUCTION
The Afcen RCC-MRx Code of design and construction rules for mechanical component in Nuclear
Installations is one of the works in the collection of design and construction rules for nuclear plants
initiated by RCC-M for Pressurized Water Reactors (PWR).
The RCC-MRx design and construction code constitutes a single document that covers in a consistent
manner the design and construction of reactor components and the associated auxiliaries, examination
and handling mechanisms and irradiation devices.
The design rules were adapted to cover the mechanical resistance of structures close to neutron sources
that can, depending on the situation, also operate in significant thermal creep conditions.
This paper presents a general overview of this draft document and focuses on particularities of this
Afcen RCC-MRx Code.
Today, Afcen's board is composed by Areva and EDF (historical members), and CEA (since 2008).
The first code published by Afcen was, in 1980, the RCC-M code for PWR.
In 1985, Afcen published RCC-MR [1], written for components operating under high temperature,
high thermal gradient loading. This code was drafted jointly with a EdF, Novatome and CEA
Committee, and it was used for the design and construction of the new Fuel Storage Tank of
SuperPhenix1, the design of European Fast Reactor, the studies for the lifetime extension of PHENIX
and procurement of several spare components (Heat Exchangers, Pumps, Tanks, pipes,…), several
Advanced Fission Reactor projects or the Indian Prototype Fast Breeder Reactor, etc... Last edition of
this code was published by Afcen in 2007.
In 2005, CEA published RCC-MX [2], written for Research Reactors and Irradiation Devices,
developed in the context of the JHR. This code was drafted jointly with a EdF, Novatome and CEA
Committee, in connection with RCC-MR. The last edition of this code was published in 2008.
In 2008, it has been decided to merge RCC-MX in RCC-MR in order to create RCC-MRx code. A
draft of the fifth edition of the RCC-MR code, named RCC-MRx [3], has been established in a draft
document, in French and English.
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
The scope of application of the RCC-MRx design and construction rules is limited to mechanical
component:
considered to be important in terms of
Section I (RDG)
nuclear safety and operability,
General provisions
playing a role in ensuring Entrance Keys – Applicable sts of rules
Equipment specification
leaktightness, partitioning, guiding, Management system
MRx code
The RCC-MRx design and construction code constitutes a single document that covers in a consistent
manner the design and construction of reactor components and the associated auxiliaries, examination
and handling mechanisms and irradiation devices. The three classes of design and construction
proposed (N1MR, N2MR and N3MR) correspond to decreasing levels of assurance of ability to withstand
different types of mechanical damages to which the component might be exposed as result of loading
corresponding to specific operating conditions.
A broader choice of materials than the steels used in Pressurized Water Reactors (PWR) and Sodium
Fast breeder Reactors (SFR) such as alloys of aluminium and zirconium can meet requirements for
neutron transparent materials of research or irradiation reactors (RR). For instance, for reflectors, the
substitution of 10 mm of steel tube for 10 mm of zirconium or aluminium alloy tubes makes it possible
to potentially obtain 50% more thermal flux. It is therefore necessary to make sure that suitable trade-
offs are made between mechanical strength, thermal control (gamma heating of 10 to 20W/g in core)
and activation effect control (e.g. cobalt content).
Although the initial editions of the RCC-M and RCC-MR rely greatly on the French NF series
standards at the time, the rapid development of European standards (EN and EN ISO) covering all
subjects such as procurement, destructive and non-destructive testing, welding and now even design,
has resulted in it being necessary to carry out a considerable amount of work to integrate these new
standards into the RCC-MR 2007 and RCC-MX 2008. This update of the standards is found naturally
in the RCC-MRx code.
PARTICULARITIES OF RCC-MRx
Materials
According to the practice in RCC-M for Pressurized Water Reactors (PWR), material selection for the
various components should be listed in Tome 1 chapter 2000 in table format showing item by item the
Reference Procurement Specification or standard used. For prototype reactors or components, these
tables drawn-up in the preliminary study documents cannot be included in RCC-MRx Tome 1 chapter
2000.
The "presentation of the RCC-MRx", included in the code, summarizes the selections for Sodium Fast
Reactors (SFR) based on the most recent projects, and for Research Reactor (RR) « Jules Horowitz »
(RJH).
Procurement requirements for products for components covered by the Section III are given in
Subsections
INDEX B, C, D, K and L 2000 with reference to Tome 2.çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
For each component or part of a component, the Equipment Specification shall define the document
that the Manufacturer shall refer to for his procurement activity:
either one of the Reference Procurement Specifications in Tome 2,
or one of the Reference Standards listed in RM 030-0 or in RB, C, D, K, L 2000,
or Particular Procurement Specification (RM 011-5)
Considering the general design analysis rules are applicable to structures made of material whose
properties are listed in one of the Properties Groups in Appendices A3 and A9 for the specified
operating conditions, then in the case of a new product or new grade purchased under, for example, a
Particular Procurement Specification, it must be accompanied by a new Properties Group as shown in
Appendices A3 and A9, validated by a Material Report as well as a report justifying the applicability
of the general design analysis rules for the specified usage conditions.
Design
Design rules are contained in chapters 3000 of subsections B, C, D, K, L. They cover sizing and
material behavior analysis and they aim to ensure that components are sufficiently safe under the
various mechanical damages to which they could be exposed under loads in specified operating
conditions:
These were drafted with due regard for the corresponding chapter of the RCC-MR code in order to
introduce additional rules into it to cover conditions where irradiation is significant.
In addition to the negligible creep test in the RCC-MR which, if meet, allows the effects of creep to be
disregarded, a negligible irradiation test, based on a ductility criteria, has been added in RB 3216.2.
This test makes it possible to disregard the effects of irradiation if the fluence received by the
component in question remains below a value specified in Appendix A3 (Properties Group) at the
service temperature in question. If the test is meet, the fluence is considered to be negligible and the
rules “without irradiation” apply unrestrictedly. The same layout as RB 3200 of the RCC-MR code is
used with allowance, in additional sections, for the effects of irradiation. The structure of RB 3200 is
as follows:
Negligible creep Significant creep
Figure 2 : Design rules for each domain RB 3251.1 RB 3252.1
(Type P damages) (Type P damages)
Negligible
RB 3261.1 RB 3262.1
irradiation
(type S damages) (Type S damages)
RCC-MRx presents criteria to ensure identical to RCC-MR Identical to RCC-MR
sufficient safety under various failure modes RB 3251.2
(Type P damages)
RB 3252.2
(Type P damages)
of structures and thus justify their design and Siignificant
irradiation
RB 3261.2 RB 3262.2
sizing when these structures are made with (Type S damages)
New rules
(Type S damages)
New rules
materials with less ductility, either from the
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
To cover the area of significant irradiation (under negligible thermal creep), in the case of elastic
analysis, the primary stress is limited as shown previously by Sm, but all of the stresses (membrane
stress and total stress) is also limited. The allowable stresses called Sem and Set (listed in A3.57) are
based on standard tension tests, with margins applied, which results in specific values for this simple
structure: (standard tensile strength and associated percentage total elongation at maximum force) and ,
percentage total elongation at fracture. This curve and its values cannot be considered intrinsic to the
material but they are associated with this structure and the material values (local tensile curve and local
fracture variables) can only be accessed through interpretation or fine measurement, which is difficult
to perform systematically.
Therefore, it was agreed to:
use the maximum point of the tensile curve Rm, Agt to limit only the general membrane stress () to
avoid locating the deformation by reduction in area that appears after this maximum and is specific
to the uniform membrane load used in the test. This is very conservative because this type of stress
is rarely uniform in complex structures. Traditional construction starts from point and adds
margins leading to the membrane's elastic allowable stress Sem.
For other stress (total stress P+Q+F), a realistic limit for deformation should be based on the local
fracture variables taking into consideration the effects of the structure upon fracture (three
dimensional testing and elastic follow-up onto secondary stresses). Given the difficulty in accessing
these limit values and the effects on the structure upon fracture, use the instability limit provided by
the tensile specimen with additional elastic follow-up factor and section coefficient kb. This method,
which could be judged insufficiently close to reality (too conservative) has the benefit of using
easily accessible variables through experimentation with reasonable conservatism (before margins),
in conjunction with stress analysis on non-irradiated materials (presence of section bending
coefficient kb that varies from 1.5 for a non-irradiated material to 1). Traditional elastic-follow-up
construction starting from point , a form factor that is closer to 1 and the addition of margins lead
to total elastic allowable stress .
In addition, it has become necessary to keep the additional rule introduced in RCC-MX which aims to
predict a structure's susceptibility (and thus its material) to the crack which could initiate fast fracture.
The rule in RB 3250 for class 1 and 2 use mechanical fracture parameters and methods; it consists of
ensuring that the value of parameter J calculated in normal operationg conditions and postulating the
existence of a conventional defect remains less than JIC except for areas of discontinuity.
The accumulation of significant irradiation and significant thermal creep was studied in a specific case:
annealed
INDEX316 and 316L grade stainless steel under temperature çç ≤ ≤ 625°C.
conditions such as 450°CSommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
Under these conditions, the rules provided are applicable within a total irradiation limit of 25 dpa. In
evaluating creep usage factors, the reference times used for non-irradiated conditions, tr, should be
replaced by trir = tr /10. This also applies to creep damage calculations performed as part of a fatigue -
creep analysis.
Finally, RCC-MRx provides in Appendix A3 “Properties Groups” maximum irradiation limit curves
for various materials. These curves are based on ductility or swelling criteria (specifically in the case
of steels where a saturation effect appears in the changes in tensile properties due to irradiation.
Tome 3 covers:
in RMC 1000, the mechanical, physical and chemical tests,
in the other chapters, the various non-destructive examination techniques:
- ultrasonic,
- radiographic,
- penetrant,
- magnetic particle,
- eddy current tubes,
- other methods of inspection (visual, video, determination of surface finish, detection of leaks),
a last chapter covering qualification and certification of non-destructive examination staff.
Many standards based on nuclear and non-nuclear practices relate to these different techniques of
testing and examination. The standards are continuously evolving to allow for changes in the practices
and the types of equipment used, and above all to achieve European Union level harmonization (such
standards are designated NF EN in their French language versions) if not at a broader international
level (standards designated NF EN ISO in their French language versions). Tome 3 of the RCC-MRx
applies to the most recent versions of these standards. But it is not sufficient just to quote them, as
interfacing is necessary to resolve the options offered in the standards, but to indicate the parts of the
standards that are integrally applicable and to supply further information and additional requirement
elsewhere. The following sections cover certain aspects of this interfacing.
Welding
General provisions for welding of class 1, 2 or 3 components are given in RB 4400, RC 4400 and RD
4400. RK and RL 4400 present detailed rules applying respectively to examination or handling
mechanisms, and irradiation devices. These chapters refer to Tome 4 which deals in detail the elements
relating to welding.
The aluminium and zirconium alloys welding requires allowance for special precautions that are
provided by RCC-MRx.
Fabrication
Tome 5 gives the minimum requirements that apply to manufacture apart from welding of
components, the manufacturer being required to add any additional arrangements it considers
necessary to assure quality of manufacture. This is because no manufacture and implementation rule
can be expressed in sufficient detail to ensure a proper fabrication. This essentially depends on the
resources and the know-how of each manufacturer. Furthermore, the latter must make allowance for
any additional rules and tolerance limits that may be specified either in the various subsections of
Tome 1 or, in the Equipment Specification.
For all items of Tome 5, the RCC-MRx specifies special requirements concerning aluminium alloys,
zirconium alloys and irradiation devices in addition to the rules laid down in the RCC-M, the RCC-
MX and the RCC-MR.
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
INTERNATIONALISATION
Today, the RCC-MR Code has gained international recognition since it was chosen by India for the
construction of PFBR which should enter service by 2010 and its four future fast reactors whose
divergences are planned before the end of 2020. ITER, at the beginning of the Nineties, chose the
RCC-MR Code as the basis to develop IISDC (ITER Interim Structure Design Criteria, now ISDC-IC:
ITER Structural Design Criteria – Internal Components), for the design of the TOKAMAK internal
structures which are strongly irradiated and, similarly, by 2004, ITER has chosen the RCC-MR 2007
for the design and fabrication of the TOKAMAK vacuum vessel.
In order to raise international recognition, Afcen propose to choose the RCC-MRx-as a sound basis,
under the CEN umbrella, for design and fabrication of GEN IV prototypes or demonstrators for the
next ENSII challenging deadlines. Today, CEA, mandated by Afcen, proposes a CEN workshop based
on the RCC-MRx (English version) in order to have an European visibility of this Code:
- The Code is required to provide a common approach for the design and construction of
mechanical nuclear components for different types of Nuclear Installations.
- The Code must be linked to the existing non nuclear standardization system in Europe and to
international standards systems.
CONCLUSION
RCC-MRx code provides tools for design and construction of Nuclear Installations including High
Temperature applications such as Sodium Fast Reactors, the ITER Vacuum Vessel and Research
Reactors. Since the previous RCC-MR edition, the RCC-MRx includes a lot of improvements such as:
- Specific materials of research reactors and irradiation devices such as aluminium alloys or
zirconium alloys are now added in the code.
- Specific material properties and rules developed for such irradiated components.
The RCC-MRx will continue to evolve regarding industrial developments, supplier experience, project
needs, operation experience, and evolution of regulation and standards. Given the large number of
projects underway in Europe (MYRRHA, ASTRID, ALLEGRO …), the new edition which will be
published in French and English by AFCEN in 2012.
REFERENCES
[1] RCC-MR 2007: "Design and Construction Rules for Mechanical Components of Nuclear
Installations", 2007 AFCEN Code, Association Française pour les Règles de Conception et de
Construction des chaudières Electro-Nucléaires (www.afcen.org).
[2] RCC-MX 2008: "Design and Construction Rules for Mechanical Components of research
Reactors and Irradiation devices", 2008 CEA Code.
[3] RCC-MRx draft document: " Design and Construction Rules for Mechanical Components of
Nuclear Installations ", to be published by AFCEN, Association Française pour les Règles de
INDEX çç (www.afcen.org).
Conception et de Construction des chaudières Electro-Nucléaires Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page28
54F Problèmes de qualité rencontrés sur des accessoires sous pression / DUPOIRON F.
Quality issues met on pressure accessories TOTAL
PETROCHEMICALS
Mardi
RICHEZ M.
TOTAL RAFFINAGE
ET MARKETING
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page29
61F Comment justifier l’absence de contrôles internes dans des usines WINTLE J. / MOORE P.
à hauts risques / Establishing a justification for noninvasive TWI
inspection of highHazard Process Plant
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page30
RÉSUMÉ / ABSTRACT
54F
PROBLÈMES DE QUALITÉ RENCONTRÉS SUR DES ACCESSOIRES
SOUS PRESSION
F. Dupoiron
(TOTAL Petrochemicals – [email protected])
M. Richez
(TOTAL Raffinage & Marketing – [email protected])
Mardi
RÉSUMÉ
Le contexte économique actuel et d’un passé récent a contribué à renforcer les optimisations techniques
(simplification des gammes) et a entraîné une évolution des approches économiques (sous-traitance
vers des pays tiers) en partie à l’origine de pertes de savoir-faire. Ces pratiques sont observées, entre
autres, pour la réalisation d’accessoires sous pression moulés. Une alerte récente a mis en évidence des
non conformités sérieuses de réalisation de certains de ces accessoires.
Cette présentation a pour objet d’exposer l’approche adoptée par un industriel pour gérer la situation
créée par la découverte de vannes en acier au carbone moulé de qualité médiocre. La méthode d’analyse
du risque généré par ces accessoires est décrite, ainsi que les principes d’établissement du plan d’action.
Des essais mécaniques ont été réalisés pour fournir des données quantifiées et pour déterminer les
limites d’emploi de ces vannes.
Afin d’éviter le renouvellement d’un tel incident, les standard internationaux utilisés pour spécifier ces
équipements ont été amendés pour bien préciser la composition chimique de l’acier et les contrôles à
réaliser lors de la fabrication.
Une méthode de contrôle par ultrasons est en cours de développement pour vérifier que le traitement
thermique, qui est primordial pour l’obtention du niveau de résilience, a été correctement réalisé.
ABSTRACT
The current and past economic environment has contributed to the increase in the technical
(simplification of manufacturing steps) and economic optimizations (outsourcing) then loss of know-
how). These practices are observed, among others, during the manufacturing of cast pressure
accessories. A recent alert has highlighted the low fabrication quality of some of these accessories.
The object of this presentation is to explain the approach adopted by an Industrial to manage the situation
created by the discovery of cast steel valves of very low quality. The method of analysis of the risk
generated by these accessories is described, as well as the principles of establishment of the action plan.
Mechanical tests have been carried out so as to obtain quantified data and to determine the limits of use
of these valves.
To avoid the renewal of such incident, international standards used to specify this type of equipments
have been amended such as to to narrow the chemical composition of the steel and to detail the inspection
and follow-up during the manufacturing.
A non destructive test by ultrasound is being developed so as to ensure that the heat treatment, which
is essential to obtain the right level of impact strength, was correctly realized.
INDEX çç Sommaire
1. INTRODUCTION
Lors de la mise en service d’une nouvelle unité pétrochimique, 6 vannes sur environ 2000 installées
ont présenté des micro-fuites situées au niveau du corps. Il s’agit de vannes en acier carbone moulées
de type ASTM A 216 WCB, de dimension nominale supérieure ou égale à 50. Ces vannes, mises sur le
marché récemment, ont été livrées avec une attestation de conformité CE délivrée par le fabricant.
L’examen de ces vannes a révélé la présence de nombreux défauts de fonderie de type porosité ou
retassure. Lors de la fabrication un nombre anormal de vannes avait fait l’objet de réparations par
soudage, effectuées sans respecter les règles de l’art. Certaines de ces réparations étaient à l’origine des
fuites observées.
Les essais destructifs, qui ont été réalisés sur des corps et chapeaux de ces vannes, n’ont pas montré
d’anomalies importantes en matière de caractéristiques mécaniques (limite d'élasticité, résistance à la
rupture), par contre les essais de flexion par choc ont montré des valeurs d’énergie très basses,
certaines inférieures à 10 Joules aux températures minimales spécifiées. Selon le code de construction
ASME B31.3, les vannes en acier moulé de type ASTM A 216 WCB sont aptes au service jusqu’à des
températures atteignant -29°C, ceux en ASTM A 352 LCB sont réputés aptes jusqu’à la température de
-46°C. Les valeurs d’énergie à -29°C et -46°C devant être supérieures à 18 Joules selon l’ASTM ou le
code de construction américain ASME B31.3. Par contre la Directive Equipement Sous Pression
97/23/CE demande que « à moins que d'autres valeurs ne soient requises au titre d'autres critères qui
doivent être pris en compte, un acier est considéré comme suffisamment ductile… si son énergie de
flexion par choc sur éprouvette ISO V est au moins égale à 27 J, à une température au plus égale à 20
°C, mais non supérieure à la plus basse température de fonctionnement prévue ».
Des niveaux de résilience bas associés à la présence de défauts de dimensions supérieures à celles du
défaut critique au sens de la mécanique de la rupture pourraient conduire en cas de sollicitation
extrême de la vanne à une évolution en fatigue de ces défauts voire à une rupture brutale.
Les discussions menées avec le fabricant et les résultats des examens réalisés sur diverses vannes
fournies par le fabricant concerné ont montré que de tels non conformités n’étaient pas limitées aux
vannes des seuls lots livrés sur l'unité citée au premier alinéa, mais que d'autres lots pouvaient être
touchés, et ce depuis plusieurs années. Plusieurs milliers de vannes implantées sur différents sites, de
types vanne à coin, robinet-vanne ou clapet de non retour peuvent être concernés.
Pour maîtriser ce risque le groupe TOTAL a été amené à entreprendre des actions en vue de recenser
les vannes susceptibles d'être concernées, de les caractériser, d’évaluer leur criticité en fonction des
conditions de service, et d'élaborer un plan d’action.
En parallèle, des actions ont été prises pour sécuriser les approvisionnements.
Les six vannes défectueuses ont été expertisées en vue de caractériser leur niveau de conformité par
rapport aux différents points du cahier des charges notamment les exigences de la directive européenne
équipements sous pression.
Par la suite, des examens ont été conduits sur un millier de vannes, qui avaient été remplacées, de
dimension allant de 2’’ à 24’’ dans le but de disposer de données permettant d’évaluer la criticité des
composants en fonction du nombre et du type de non conformité . Ces examens ont été réalisés par
sélection successive dans le but d’obtenir une caractérisation des vannes les plus affectées.
Ces examens ont porté sur la métallurgie, la recherche de réparations par soudage, de défauts externes
INDEX
subsistants après réparation et de défauts de compacité. Pourççréaliser ce vaste programmeSommaire
d’inspection,
les techniques classiques d’investigation ont été mises en œuvre (visuel, ressuage, magnétoscopie,
réplique, radiographie, analyse chimique et essai mécanique).
Parmi le millier de vannes qui a été inspecté visuellement, une sélection de 320, représentative des
dimensions et types installés, a été démontée, puis examinée en externe et en interne par contrôle
visuel. Sur 228 d’entre eux, un contrôle par ressuage ou magnétoscopie a été effectué. Enfin sur une
sélection des 30 vannes, jugées les plus critiques, des examens non destructifs par radiographie et prise
de répliques ont été réalisés. A la suite de ce programme, des investigations destructives seront
réalisées. La figure ci dessous présente la stratégie de sélection :
Le tableau 1 ci-après présente une synthèse des résultats. Sur les 1075 vannes contrôlées, environ 30%
des produits semblent avoir subi des réparations par soudage. Sur les 320 vannes extraites de cette
population, et examinées visuellement en interne plus de la moitié présentent des réparations. Sur les
228 vannes contrôlées par ressuage / magnétoscopie, la quasi-totalité présente des indications plus ou
moins importantes. Enfin sur 30 vannes extraites de cette population, seules 13 % sont conformes en
compacité radiographique (selon EN 12681 et cliché ASTM E 446 vol 1) et 13% présentent une
microstructure optimum.
Les résultats de ces investigations, réalisées sur un nombre significatif de pièces, permettent d’avoir
une description assez complète des défauts les plus importants et les plus néfastes qui peuvent être
rencontrés. Ils sont très au-delà de ce qui est accepté pour ce type d’application. Ils justifient la mise en
place de plans d’actions concernant les produits installés, et les futurs approvisionnements.
Ces vannes disposent d'un repère fonctionnel porté sur les plans de circulation des fluides ou sur
des schémas isométriques, mais ils ne font pas l’objet d’un marquage individuel. Il est aisé
de retrouver les principales caractéristiques techniques de la robinetterie (DN, PN,
matière…) mais il n’est pas possible de disposer des données individuelles comme le repère
du fabricant, le numéro de coulée, la date de mise en place…
Dès lors que le nombre de vannes concernées dépasse quelques dizaines, et du fait qu'il est difficile
d'associer la sortie du magasin des vannes incriminées à une opération spécifique de maintenance, la
localisation des vannes sur base documentaire est une opération très difficile voire impossible. La seule
façon fiable de localiser les vannes susceptibles d'être concernées est de procéder à un relevé sur site,
INDEX çç Sommaire
circuit par circuit. Ce relevé est extrêmement long du fait de l’étendue des tuyauteries, de leur
complexité et parfois de difficultés d'accès sans échafaudage. Afin de procéder de façon méthodique,
les actions de localisation et de décision d'intervention (sécurisation, voire remplacement) ont été
soumises à une démarche préalable de hiérarchisation.
Les tuyauteries des unités sont regroupées par "système", un "système" regroupe les tuyauteries qui
contiennent un même fluide et qui sont exploitées dans des conditions de service proches. Pour les
relevés sur site une hiérarchisation simple a été faite. Le scénario majorant redouté est une rupture
guillotine brutale, ou une ouverture par fatigue de la vanne. Les paramètres dangerosité du fluide et
température minimale ont été retenus comme pertinents et, combinés, pour classer les circuits (voir
figure 1 employé par le Raffinage) :
Priorité 1 : lignes de GPL qui présentent la dangerosité maximum et qui sont susceptibles
d'engendrer les plus basses températures en cas de dépressurisation rapide
Priorité 2 : Les lignes de H2S, HF et fuel gaz présentent également un niveau de risque élevé
mais ne sont pas susceptibles d’atteindre un niveau de température aussi bas que les lignes
classées en priorité 1
Priorité 3 : les lignes de liquides facilement inflammables
Priorité 4 : les lignes de liquides peu inflammables (gazole)
Priorité 5 les lignes de produits peu inflammables chauds, la vapeur et les gaz utilité air et azote
Le recensement des vannes sur les circuits prioritisés 1, 2 et 3 a été mené en premier, des délais de
réalisation précis ont été fixés. Les vannes localisées doivent être :
repérées sur isométrie ou plan de circulation des fluides,
repérées sur site,
listées dans un fichier en précisant leurs caractéristiques, marquage et conditions effectives
d’exploitation.
INDEX çç Sommaire
La représentation matricielle des risques permet de prioritiser les actions de sécurisation des vannes à
mettre en œuvre lorsque nécessaire
3. Actions de sécurisation
Les vannes sont réparties en 3 catégories :
Le plan d’action et sa réalisation font l’objet d’un enregistrement et d’un suivi périodique.
Conformément aux recommandations de la normes EN 13445-2 [1] , et pour permettre de préciser les
niveaux de risque de rupture fragile générés par le niveau de résilience et la présence de défauts, une
approche d’aptitude au service a été développée, notamment en traitant le cas d’un robinet vanne de 8’’
série 600 lbs - PN 100 selon la méthodologie proposée par la norme BS 7910 « Guide on Methods for
Assessing the Acceptability of Flaws in Metallic Structures »[2]. Une première étude par éléments finis
a permis de décrire très complètement les niveaux de contrainte, puis le comportement de défauts plans
positionnés dans les zones de contraintes maximales en fonction de différentes hypothèses. Les
contraintes maximales se trouvent au niveau du rayon de raccordement des brides et du corps, ainsi
qu’au niveau du guide du coin mobile. Les premières conclusions montrent que ces organes sont
largement dimensionnés, ce qui conduit à des niveaux de contraintes globalement faibles. En
coordination avec cette étude, une approche expérimentale a été menée pour d’une part valider la
modélisation, et d’autre part évaluer de façon concrète les marges de sécurité.
Cette partie expérimentale à permis des tester une vanne présentant des niveaux de résilience de
l’ordre de 5 à 10 Joules aux températures d’essais (-10°C à -20°C). Un essai a été conduit jusqu’à
environ trois fois la pression nominale sans que la vanne ne montre de signe de dégradation. Après
introduction d’un défaut plan artificiel par fatigue, dans une des zones la plus chargée (congé de
raccordement de la bride), l’essai à été conduit à -20°C jusqu’à une pression de quatre fois la pression
nominale sans propagation du défaut. Ces essais ont donc permis de confirmer les résultats de la
modélisation, validant ainsi l’ordre de grandeur des marges de sécurité.
INDEX çç Sommaire
Figure -3- Schéma de principe et Vanne 8 300 lbs sur banc d’essai
Un essai de tenue en fatigue et des mesures de ténacité doivent encore être réalisées.
L’expérience récente a montré que la simple exigence d'application et de respect des standards ASME,
API et de la Directive européenne ne suffit pas pour assurer la qualité attendue des vannes. Il est
apparu nécessaire de compléter les standards ASTM A 216 et A 352 par des exigences techniques
complémentaires. Ces exigences portent avant tout sur la composition chimique, la métallurgie de
l'acier, les contrôles non destructifs à réaliser sur le corps et le chapeau de vanne, la représentativité des
résultats des essais de résilience réalisés sur lingot échantillon et la documentation.
1. Métallurgie
Tous les contrôles doivent être réalisés par du personnel certifié niveau 1 et interprétés par des
personnels certifiés niveau 2 suivant la norme EN 473 ou le document SNT-TC-1A. Un soin particulier
doit être apporté aux conditions de réalisation et d’interprétation des contrôles (par exemple la
luminosité pour l'examen visuel).
Un contrôle visuel de toutes les pièces doit être effectué sur la totalité des surfaces. Cet examen doit
être effectué en conformité avec le document MSS SP-55.
Les contrôles par magnétoscopie et par radiographie ci-après sont à effectuer sur les pièces brutes.
2.1. Magnétoscopie
Un contrôle magnétoscopique de 100% des surfaces (y compris des surfaces internes accessibles) doit
être réalisé sur 10% des pièces de chaque lot.
Dans le cas où il n’est pas possible de réaliser un contrôle magnétoscopique, sur justification et avec
accord du représentant de l'acheteur, celui-ci peut être remplacé par un contrôle par ressuage.
2.2. Radiographie
Un contrôle radiographique doit être réalisé sur 100% des zones critiques requises par le document
ASME B16.34 (ou la norme EN 12516-1), plus un contrôle des congés de la zone de raccordement fut
et cou de corps, il doit concerner 10% des pièces de chaque lot (100% des pièces pour CLASS > 600).
Il doit être effectué selon la norme EN 12681, ISO 4993 ou le document MSS SP-54.
Les critères d'acceptation sont ceux du document ASTM E 446. Les fissures et criques de retrait sont
inacceptables ; les soufflures, piqures, inclusions et retassures sont acceptables avec un niveau de
sévérité moyen (3 sur 5).
2.3. Ressuage
Après usinage, un contrôle par ressuage de 100% des portées de joint doit être réalisé sur 10% des
pièces de chaque lot, le contrôle des portées de joint RTJ doit pour sa part être réalisé à 100 % pour
toutes les pièces.
Numéro de coulée : Numéro correspondant à une charge unique fondue dans un four
INDEX çç Sommaire
Lot : Pièces moulées de même type, de même dimension nominale, de même série (de
pression), portant le même numéro de coulée et ayant subi le traitement thermique
simultanément dans un four donné (cette définition est proche de celle adoptée par les
normes européennes).
Lingot échantillon Echantillon d’un lot destiné à l’exécution d’essais destructifs et non
destructifs. Les lingots échantillons doivent être traités thermiquement dans le même
four et conjointement avec le lot qu’ils représentent.
Lingot échantillon attenant aux pièces Echantillon attenant à une pièce. Les lingots
échantillons attenant à une pièce ne doivent pas être détachés avant que le fabricant ait
terminé le traitement thermique des pièces moulées du lot d'appartenance. A défaut de
lingot échantillon attenant à une pièce, une pièce moulée appartenant au lot peut être
utilisée comme lingot échantillon attenant à une pièce.
Des recommandations générales ont été émises sur la réalisation du traitement thermique en insistant
sur son importance primordiale pour l’obtention d’une bonne résilience. La spécification ne contient
aucune recommandation quant au cycle du traitement thermique qui est de la seule responsabilité du
fondeur.
En l’absence d’indication précise, la température de l’essai de résilience est la plus faible des deux
valeurs suivantes : température spécifiée par l’ASTM et -10°C.
L’énergie de flexion par choc sur éprouvette ISO V doit être vérifiée sur 3 éprouvettes conformément à
la norme ISO 4990 à la température minimale spécifiée. La valeur moyenne de l’énergie absorbée des
trois éprouvettes ne doit pas être inférieure à 27 J, seule une des trois valeurs peut être inférieure à 27
J, mais ne doit pas être inférieure à 19 J.
Les essais doivent être effectués sur un lingot échantillon attenant aux pièces ou une pièce moulée du
lot. Il convient de prévoir pour chaque lot :
un lingot échantillon attenant aux pièces pour la réalisation des essais ;
Si demandé, un lingot échantillon attenant aux pièces à livrer avec le lot de vannes, lingot qui
aura été détaché après la réalisation complète du traitement thermique.
L’épaisseur des lingots échantillons attenants aux pièces doit être au moins égale à celle de la bride de
la pièce à laquelle elle est reliée et les dimensions du lingot échantillon attenant aux pièces doivent
permettre de réaliser l’ensemble des essais mécaniques requis. Un plan coté des lingots échantillons
attenants aux pièces et de leur zone d’attache doit être transmis pour accord.
Les résultats des essais effectués sur les lingots doivent être le reflet des caractéristiques mécaniques
des pièces moulées du lot.
TOTAL se réserve le droit de faire réaliser des essais mécaniques par un laboratoire indépendant sur
des pièces de vannes livrés. Si les résultats de ces essais n'étaient pas satisfaisants, l’ensemble du lot
concerné serait alors rejeté et une extension des contrôles serait réalisée pour vérifier la conformité des
autres lots.
Nous avons vu l’importance primordiale de la qualité du traitement thermique pour l’obtention d’un
bon niveau de résilience. Si le contrôle de la qualité de ces traitements peut être effectué via des
inspections spécifiques du fabricant et par l’enregistrement des cycles thermiques, il s’avère
INDEX çç Sommaire
néanmoins nécessaire de disposer d’un dispositif de contrôle non destructif de l’état microstructural
pour s’assurer de la qualité de réalisation des traitements thermiques.
Dans un premier temps de nombreux contrôles de microstructure ont été réalisés par répliques
métallographiques. Cependant cette méthode a montré ses limites tant d’un point de vue facilité de
mise en œuvre que du point de vue productivité, et coût associé
Une évaluation comparée de méthodes potentiellement plus efficaces a été conduite avec pour objectif
la possibilité de pouvoir réaliser aisément ce type de contrôle chez les fournisseurs. La méthode
retenue doit permettre la différenciation des microstructures, mais aussi de s’affranchir au mieux de la
variabilité des états de surface et de l’anisotropie des structures. Le matériel mis en œuvre doit être
aisément transportable. Afin d’être rapidement opérationnelle, il était nécessaire qu’elle fasse appel à
des technologies connues et disponibles immédiatement. Elle ne devait pas nécessiter de
développements trop importants.
Trois méthodes non destructives ont ainsi été testées par le CETIM : courants de Foucault, bruit
ferromagnétique (Bruit Barkhausen) et ultrasons. Des essais ont été réalisés à la fois sur des
microstructures issues de produits moulés industriels caractérisés, présentant plusieurs niveaux de
résiliences, et sur des microstructures spécifiquement retraitées thermiquement pour obtenir différents
états microstructuraux (normalisation, traitement inter-critique, trempe et revenu). Les évaluations
comparatives montrent que la méthode par ultrasons répond le mieux aux attentes.
Cette méthode en cours de développement par le CETIM est basée sur la modification du spectre
fréquentiel des ondes ultrasonores longitudinales lors de la traversée du matériau [5], [6], l’atténuation
des ondes ultrasonores étant fonction du diamètre du grain et de la fréquence.
Pratiquement, la mise en œuvre de cette méthode consiste à faire propager les ondes ultrasonores à
travers la bride de la vanne, et analyser le spectre fréquentiel du signal.
Cette méthode est encore en cours de développement. Elle pourrait être utilisée pour contrôler
aisément la présence de la structure attendue sur les nouvelles fabrications sans fournir directement les
caractéristiques de résilience, ces dernières étant aussi liées à la composition chimique et à la propreté
inclusionnaire.
6. Conclusion
Le contexte économique actuel et d'un passé récent a contribué à renforcer les optimisations techniques
(simplification des gammes) et a entraîné une évolution des approches économiques (sous-traitance
vers des pays tiers) puis perte partielle de savoir-faire. Ces pratiques sont observées, entre autres, pour
la réalisation d’accessoires sous pression moulés. Des alertes récentes ont mis en évidence des non
conformités de réalisation de certaines vannes.
Pour s’affranchir du risque que pouvait présenter ces équipements, un recensement systématique de
ces vannes a été réalisé sur les circuits critiques. Chaque vanne a fait l’objet d’une évaluation
spécifique de ses conditions de service pour élaborer un plan qui contient, si nécessaire, le
remplacement et/ou des mesures de sécurisation.
Un millier de vanne a été examiné pour les caractériser et fournir les données nécessaires à
l’évaluation de leur aptitude au service. Les contrôles ont visé à déterminer la nature et la dimension
des défauts présents, ainsi que les caractéristiques mécaniques du matériau. Des essais hydrauliques à
basse température et forte pression ont été réalisés sans effet remarqué sur la vanne. Des essais de
tenueINDEX
en fatigue doivent être réalisés. çç Sommaire
Pour éviter le renouvellement d’un tel incident, une spécification a été établie. Elle renforce les
contrôles effectuées lors de la fabrication et restreint la composition chimique de l’acier.
La bonne réalisation du traitement thermique est primordiale pour l’obtention d’une structure
métallurgique bénéfique à la résilience. Une méthode de contrôle non destructif basée sur les ultrasons
est en cours de développement pour vérifier l’obtention de cette structure sans recourir à des méthodes
plus lourdes, que seraient des répliques, ou des contrôles destructifs comme les essais de flexion par
choc.
Plus généralement, une démarche spécifique de qualification sur produits et d’agrément des
fournisseurs a été entreprise avec un suivi dans le temps basé sur les retours d’expérience et des
contrôles produit.
Bibliographie
[1] NF EN 13445-2:2009 (F) Version 1 (2009-07) - Récipients sous pression non soumis à la flamme -
Partie 2 matériaux
[2] Guidance on methods for assessing the acceptability of flaws in fusion welded structures, BS 7910:1999
[3] Tensile and impact properties of high purity Iron –carbon and Iron carbon Manganese alloys of low carbon
content. N.P. Allen et Al. J. Iron Steel Inst. , Vol 174, June 1953 P 108-120.
[4] Effect of alloying elements on notch toughness of pearlitic steels. J.A Rineholts and W.J Harris Jr . Trans
ASM, Vol 43, 1951, pp 1175-1214.
[5] Grain size influence on ultrasonic velocities and attenuation. A.Badidi Bouda and al. NDT&E International 36
1-5 – 2003
[6] Application of the acoustic resonance method to evaluate the grain size of low carbon steels. Bongyoung
Ahn and al. NDT&E International 32 85-89 – 1999.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page31
RÉSUMÉ / ABSTRACT
53F
ÉMISSIONS FUGITIVES DES ROBINETS INDUSTRIELS :
LE POINT SUR LA NORMALISATION ET LES MÉTHODES D’ESSAIS
RÉSUMÉ
Dans la pétrochimie, les robinets industriels sont considérés comme les principales sources d’émissions
fugitives de COV. Depuis quelques années des normes, des spécifications utilisateurs ou des
réglementations proposent des procédures (essais de type ou essais de production) visant à caractériser
et qualifier les performances des robinets ou des systèmes d’étanchéité des tiges de manœuvre.
Dans cet article sont d’abord présentées et comparées les principales références que sont la norme ISO
15848-1&2, les MESC Shell SPE 77-300 et 77-312, la spécification VDI 2440 liée à la TA Luft, la
norme API 622. On présente ensuite un bilan d’essais réalisés au CETIM sur différents types de robinets
et de garnitures (graphite et PTFE pour l’essentiel), en Hélium, selon l’ISO 15848-1, en terme de
faisabilité d’essai et de classe de performance atteinte.
Enfin, les fluides d’essai préconisés dans les différentes normes et spécifications sont l’hélium et le
méthane. Mais à titre de perspective, on décrit une technique de mesure basée sur l’utilisation d’un
mélange Hydrogène-Azote comme alternative notamment à l’hélium.
ABSTRACT
Industrial valves are considered as being the main source of fugitive emissions of VOC (volatile organic
compound) in the petrochemical industry. For the last few years, standards, end users’ specifications
and regulations describe some procedures (type testing or production testing) for the characterisation
and the qualification of valves or stem sealing systems.
This article presents and compares the main references which are standards ISO 15848-1&2 and MESC
Shell SPE 77-300 and SPE 77-312 and API 622, and specifications VDI 2440 related to TA Luft.
Then, a summary of some tests performed at CETIM is presented. These tests were carried out on dif-
ferent types of valves and packings (mainly graphite and PTFE), using helium, and according to ISO
15848-1 in terms of testing feasability and achieved performance class.
Finally, the recommended test fluids in the different standards and specifications are helium and
methane. However, so as to anticipate future possible practice, a measurement technique based on the
use of an hydrogen-nitrogen mixture is described. This mixture could be an alternative to the use of
helium mainly.
INDEX çç Sommaire
EMISSIONS FUGITIVES DES ROBINETS INDUSTRIELS:
LE POINT SUR LA NORMALISATION ET LES METHODES D’ESSAIS
RESUME
Dans la pétrochimie, les robinets industriels sont considérés comme les principales sources
d’émissions fugitives de COV. Depuis quelques années des normes, des spécifications
utilisateurs ou des réglementations proposent des procédures (essais de type ou essais de
production) visant à caractériser et qualifier les performances des robinets ou des systèmes
d’étanchéité des tiges de manœuvre.
Dans cet article sont d’abord présentées et comparées les principales références que sont la
norme ISO 15848-1&2, les MESC Shell SPE 77-300 et 77-312, la spécification VDI 2440
liée à la TA Luft, la norme API 622. On présente ensuite un bilan d’essais réalisés au CETIM
sur différents types de robinets et de garnitures (graphite et PTFE pour l’essentiel), en
Hélium, selon l’ISO 15848-1, en terme de faisabilité d’essai et de classe de performance
atteinte.
Enfin, les fluides d’essai préconisés dans les différentes normes et spécifications sont l’hélium
et le méthane. Mais à titre de perspective, on décrit une technique de mesure basée sur
l’utilisation d’un mélange Hydrogène-Azote comme alternative notamment à l’hélium.
ABSTRACT
Industrial valves are considered as being the main source of fugitive emissions of VOC
(volatile organic compound) in the petrochemical industry. For the last few years, standards,
end users’ specifications and regulations describe some procedures (type testing or
production testing) for the characterisation and the qualification of valves or stem sealing
systems.
This article presents and compares the main references which are standards ISO 15848-1&2
and MESC Shell SPE 77-300 and SPE 77-312 and API 622, and specifications VDI 2440
related to TA Luft.
Then, a summary of some tests performed at CETIM is presented. These tests were carried
out on different types of valves and packings (mainly graphite and PTFE), using helium, and
according to ISO 15848-1 in terms of testing feasability and achieved performance class.
INDEX çç Sommaire
Finally, the recommended test fluids in the different standards and specifications are helium
and methane. However, so as to anticipate future possible practice, a measurement technique
based on the use of an hydrogen-nitrogen mixture is described. This mixture could be an
alternative to the use of helium mainly.
INTRODUCTION
Dans la pétrochimie, les robinets industriels sont considérés comme les principales sources
d’émissions fugitives de COV. Depuis quelques années des normes, des spécifications
utilisateurs ou des réglementations proposent des procédures (essais de type ou essais de
productions) visant à caractériser et qualifier les performances des robinets ou des systèmes
d’étanchéité des tiges de manœuvre.
La société Shell propose une spécification MESC Shell SPE 77-300 (dec 2008) comparable
sur le principe de la qualification des robinets mais qui définit des
- conditions de mesures de fuite au presse étoupe différentes (elle préconise le reniflage),
- des classes d’étanchéité différentes
- et nombre de conditions d’essais (cycles mécaniques et thermiques) en fonction du type
de robinet, de sa dimension …
Plus récemment la norme API 622 a été publiée aux Etats-Unis. Elle concerne la qualification
des garnitures de presse étoupe à partir d’un montage d’essais représentatif d’une vanne de
sectionnement. La méthode de mesure de la fuite est le reniflage (panaché de flushing…) et le
fluide d’essais est le méthane.
Nous présentons dans cet article, tout d’abord, une brève étude comparative de ces normes et
spécification.
Ensuite, nous présentons un bilan d’essais réalisés au CETIM sur différents types de robinets
et de garnitures (graphite et PTFE pour l’essentiel), en Hélium, selon l’ISO 15848-1, en terme
de faisabilité d’essai et de classe de performance atteinte.
Enfin, si les fluides d’essai préconisés dans les différentes normes et spécifications sont
l’hélium et le méthane, nous décrirons, à titre de perspective, une technique de mesure basée
sur l’utilisation d’un mélange hydrogène-azote comme alternative notamment à l’hélium.
INDEX çç Sommaire
LA NORME EN ISO 15848
Elle a été publiée en 2006 après plus de dix ans de travaux. La France est à l’origine de ce
projet par une action conjuguée de ELF et de la commission robinetterie industrielle du
CETIM engagée en 1992.
La norme ISO 15848 spécifie des modes opératoires d'essai pour l'évaluation des fuites
externes des dispositifs d'étanchéité de la tige (ou de l'arbre) des appareils de robinetterie de
sectionnement et de régulation, et des joints de corps des robinets destinés à être utilisés au
contact de polluants atmosphériques volatils et de fluides dangereux. Les jonctions des
raccords d'extrémité, les applications sous vide, les effets de la corrosion et des rayonnements
sont exclus de l'ISO 15848.
L'ISO 15848 comprend les parties suivantes, présentées sous le titre général Robinetterie
industrielle — Mesurage, essais et modes opératoires de qualification pour émissions
fugitives:
La norme définit
• Des classes d’étanchéité
• Des classes d’endurance
• Des classes de température
Dans sa version actuelle, la norme précise que la méthode de mesure de fuite au niveau du
presse étoupe doit être la méthode globale par le vide ou par balayage de gaz (« flushing »).
Cette méthode impose une instrumentation ou un « encapsulage » du presse étoupe (Fig 1).
Associée à un spectromètre de masse hélium (Fig 1), elle est précise, fiable, indépendante de
l’opérateur et de l’environnement.
Cette méthode permet la mesure de la fuite globale (totale) du presse étoupe
INDEX çç Sommaire
La méthode par reniflage est préconisée pour la mesure des émissions des joints de corps. On
considère, à priori, que ces joints doivent être étanches et le reniflage est utilisé comme
moyen de contrôle. La méthode par reniflage est avant tout une méthode de localisation (cf
norme EN 1779)
1 chambre à vide
2 hélium sous pression
3 détecteur d'hélium à vide
Classes d’étanchéité
La norme définit 3 classes d’étanchéité pour les systèmes d’étanchéité de tige (tableau 1).
INDEX çç Sommaire
Dans la norme la classe d’étanchéité est exprimée en mg/s par m de circonférence par
référence à d’autres normes d’essais de produits d’étanchéité notamment l’EN 13555 pour la
caractérisation des joints plats pour assemblages à brides.
C <= 102 Généralement atteint avec des garnitures à base de graphite flexible
a Exprimé en mgs1m1 et mesuré avec la méthode de fuite gobale par la méthode au vide ou par flushing.
b La classe A peut être mesurée uniquement avec de l'hélium au moyen de la méthode sous vide.
Tableau 1 : Classes d'étanchéité pour les dispositifs d'étanchéité de la tige (ou de l'arbre)
exprimée en mg/s/m
Dans le cas de mesures en hélium, l’unité traditionnelle de mesure par spectromètre de masse
est l’atm.cm3/s (tableau 2)
Tableau 2: Classes d'étanchéité pour les dispositifs d'étanchéité de la tige (ou de l'arbre)
Le niveau d’émission des joints de corps du robinet doit être inférieur à 50 ppm d’hélium.
La norme définit des classes d’endurance incluant des cycles mécaniques de fonctionnement
et des cycles thermiques.
Elle distingue les robinets de sectionnement et les robinets de régulation.
Robinets de sectionnement
Le nombre minimal requis de cycles mécaniques pour les robinets de sectionnement doit être
de 500 cycles (course complète) avec deux cycles thermiques. Cette étape de classification
doit être identifiée comme CO1. Une extension jusqu'à la classification CO2 peut être
effectuée en ajoutant 1 000 cycles mécaniques et un cycle thermique. Une extension
supplémentaire jusqu'à CO3 peut être effectuée en répétant l'exigence définie pour CO2 (voir
Figure 2).
INDEX çç Sommaire
Ttest température d'essai, °C N nombres de cycles mécaniques P pression du fluide d'essai
L1 mesurage de la fuite du dispositif d'étanchéité de la tige
L2 mesurage de la fuite du dispositif d'étanchéité du corps
Robinets de régulation
Le nombre minimal requis de cycles mécaniques pour les robinets de régulation doit être de
20 000 cycles et deux cycles thermiques, excepté pour RT. Cette étape de classification doit
être identifiée comme CC1. Une extension jusqu'à la classification CC2 peut être effectuée en
ajoutant 40 000 cycles mécaniques et un cycle thermique. Une extension supplémentaire
jusqu'à CC3, peut être effectuée en répétant l'exigence définie pour CC2 (voir Figure 3).
INDEX çç Sommaire
Les Classes de température
La norme propose des températures de références auxquelles peuvent être testés les robinets.
Si un robinet doit être qualifié par exemple dans la gamme -46°C à 200°C le robinet doit être
testé séparément à chacune de ces températures.
Extension de la qualification
Pour minimiser le nombre d’essais la norme définit une extension de la qualification aux
robinets ayant un diamètre de tige compris entre et 2 fois le diamètre testé.
Dans l’esprit du WG10 de l’ISO TC153 SC1 en charge de son élaboration, l’objectif de la
partie 2 de l’ISO 15848 (essai de production) est de vérifier qu’un robinet dont un exemplaire
a été qualifié selon la partie 1 a été bien monté. Il en résulte des conditions d’essais moins
sévères que la partie 1 avec notamment
• Une pression d’essai de 6 bar
• Un essai à température ambiante
• Quelques cycles de fonctionnement seulement (5 au total).
De même, s’il s’agit de garder la notion de classes d’étanchéité, celles-ci sont vérifiées par la
méthode du reniflage, plus simple de mise en œuvre.
A <= 50
B <= 100
C <= 1 000
Dans le cadre de la révision quinquennale des normes, il a été décidé en Septembre 2009
d’amender la norme 15848 pour prendre en compte un certain nombre de commentaires émis
par ses utilisateurs. Les travaux de révision ont débuté fin juin 2010. Le projet révisé devrait
être envoyé en enquête courant 2011.
INDEX çç Sommaire
Il a été proposé de mesurer les classes d’étanchéité B et C par la méthode ou par accumulation
(bagging). On maintient ainsi la mesure de la fuite totale du système d’étanchéité. Le principe
de la méthode sera inscrit dans une nouvelle annexe.
Par ailleurs il a été décidé de créer une classe d’étanchéité D pour caractériser une mesure de
fuite par reniflage pour le cas ou la mise en œuvre des autres méthodes s’avérerait délicate
(robinets de très grande taille, très hautes pressions…)
La classe A serait décalée d’une décade pour s’harmoniser avec le niveau de fuite proposé
dans la VDI 2440, il n’y aurait donc plus qu’une décade entre le seuil de la classe A et celui
de la classe B.
Par ailleurs il a été convenu de distinguer les mesures en hélium et en méthane en exprimant
les fuites hélium en atm.cm3/s (ou l’équivalent mbar.l/s) et les fuites méthane en ppm.
Il a par ailleurs été proposé de rapporter les seuils des classes d’étanchéité par mm du
diamètre de tige et non plus par mètre de circonférence de la tige.
Il est proposé que le nombre de cycles mécaniques pour la qualification CO1 des robinets de
sectionnement soit ramené de 500 à 200 (par séries de 50). L’obtention des classes CO2 et
CO3 reste inchangée.
On conserve tels quelles les règles de qualification pour les vannes de régulation.
ISO 15848-2
Depuis de nombreuses années SHELL propose des MESC dédiées à la caractérisation aux
émissions fugitives des robinets. Pour les fabricants de robinets et les laboratoires en charge
des essais il n’a pas été facile de suivre les nombreuses évolutions (souvent importantes) des
différentes versions.
INDEX çç Sommaire
• la MESC SPE 77/312 d’avril 2010 « Fugitive emission production testing
(amendments/supplements to ISO 15848-2) »
Une partie seulement de cette MESC concerne la qualification aux émissions fugitives
(§ C.2.9, C.4.2.4, C.4 .4.4, C.4.6.4).
La MESC impose la mesure de fuite du système d’étanchéité de tige par reniflage (en faisant
référence à l’annexe B de l’ISO 15848-1). C’est un des points de divergence avec la norme
ISO 15848-1 qui propose une mesure globale de la fuite au niveau du presse étoupe.
Par contre la MESC préconise une méthode par accumulation (« bagging ») pour les joints de
corps.
La norme EN 1779/1999 précise que la mesure par reniflage n’est pas une méthode de mesure
de fuite mais une méthode de localisation de la fuite. Elle spécifie aussi que la méthode de
reniflage n’est pas compatible avec la class A de la MESC.
Il faut considérer que les classes d’étanchéité mesurée par reniflage ne sont représentatives
que d’une fraction de la fuite totale du presse étoupe. La confusion est accentuée avec
l’utilisation d’une même unité de fuite (atm.cm3/s ou mbar.l/s).
La MESC 77-300 définit deux classes d’étanchéité A(HS ) et B mesurées par reniflage.
La classe A(HS) est d’une décade supérieure à la class A de l’ISO 15848-1 (elle est
comparable en valeur à la nouvelle classe A proposée par le WG10 en juin 2010).
La class B à la même valeur que celle donnée par l’ISO 15848-1 mais une classe B MESC
SHELL mesurée par reniflage n’est pas comparable à une classe B ISO 15848-1 mesurée par
méthode globale !
Il faut noter que lors de la réunion de l’ISO TC153 SC1 WG10 de juin, les experts de SHELL
ont admis la mesure globale par le vide ou par accumulation comme méthode de référence
pour les classes B et C.
Pour une qualification d’un robinet pouvant être utilisé en haute et basse température l’essai
est réalisé en une seule fois avec une succession de cycles haute et basse température.
INDEX çç Sommaire
Le nombre de cyclages thermiques dépend de la température d’utilisation du robinet.
Pour une température de fonctionnement comprise entre 80°C et 200°C, un seul cycle
thermique est à appliquer, pour une température supérieure à 200°C, deux cycles thermiques
sont nécessaires.
INDEX çç Sommaire
VDI 2440 (NOV 2000)
Les mesures doivent être réalisées par une méthode globale (méthode globale sous vide, par
flushing ou accumulation), c'est-à-dire que la totalité de la fuite est prise en compte par la
méthode et peut ainsi être quantifiée de façon fiable. Pour ce faire, des instrumentations de
l’étanchéité de tige sont nécessaires (figure 7)
Figure 5 : Exemple d'instrumentation de tige pour mesure des émissions selon VDI 2440
INDEX çç Sommaire
Si les méthodes de mesures, les seuils à ne pas dépasser et les températures sont bien
déterminés, en revanche il n’est pas précisé
- ni l’amplitude et ni le nombre de cycles ouverture-fermeture,
- si la température est appliquée en continue ou en cycles, comme par exemple dans la
procédure ISO 15848-1.
Ce manque de précision conduit à des produits « qualifiés TA-Luft » qui ont été testés dans
des conditions qui sont parfois difficilement comparables.
Il semble néanmoins qu’il soit fait souvent références aux cycles mécaniques et thermiques
proposés par l’ISO 15848.
L’API 622 n’est pas une procédure de qualification aux émissions fugitives de robinets, mais
de garnitures (tout au moins dans sa version actuelle)
La norme définit un essai de garniture dans un montage dédié conçu sur la base d’un robinet
de sectionnement 4’’300 lbs.
Le fluide utilisé est le méthane, ce qui implique une méthode de mesure de type flushing ou
bagging (accumulation) avec un détecteur de vapeur organique de type FID, équipé d’une
sonde de reniflage (ou sniffer). Dans le cas de la méthode par flushing (fig 6), le débit du gaz
de balayage (flush) est fixé par le débit interne de la sonde (environ 1 l/min).
FID
INDEX çç Sommaire
Il n’y a pas de classe d’étanchéité préconisée. Il est simplement demandé de resserrer le
presse étoupe dès que le niveau d’émission dépasse 500 ppm (il ne semble pas y avoir de
limite du nombre de resserrages).
L’API étant dédiée à la qualification des garnitures, elle propose des essais spécifiques
• De corrosion à froid et à chaud
- Ces essais sont largement inspirés des essais de corrosion développés par EDF
et par TOTAL (SGM 2082 TUY) et pratiqués au CETIM depuis de longues
années. Les montages sont comparables. Les procédures se différencient
Par la durée de la corrosion à froid de 4 semaines (au lieu de 5
semaines )
Par la température et la pression de l’essai de corrosion à chaud : 147°C
et 45 bar pour l’API, 300°C et 90 bar pour la procédure TOTAL.
• De caractérisation du matériau des tresses
- Perte de poids
- Densité
- Teneurs en lubrifiant
Nous présentons ci après une synthèse d’une trentaine d’essais réalisés au CETIM en
application de l’ISO 15848-1. Ces essais mettent en évidence notamment que la procédure
d’essai est applicable, que les classes d’étanchéité définis sont atteignables notamment la
classe B dans de nombreux cas et avec des garnitures à base de graphite…
Cette analyse a été effectuée sur 31 essais réalisés sur 22 robinets de sectionnement et 9
robinets de régulation
Parmi ces robinets, on retrouve :
INDEX çç Sommaire
- 7 robinets papillon
- 13 robinets soupapes
- 8 à robinets tournant
- 3 robinets vannes
Ainsi cette analyse étudie un échantillon assez représentatif des robinets, équipés de systèmes
d’étanchéité de tige, que l’on retrouve sur des sites industriels. En revanche, la répartition des
types de systèmes d’étanchéité de tige n’est, elle, probablement pas représentative.
Nous avons testés des robinets allant de PN10 à PN250 avec deux maxima à PN10 (8 essais)
et PN50 (7 essais).
Les systèmes d’étanchéité réalisés à partir de composés en graphite sont bien représentés, et
ont été testés de façon assez homogène de PN20 à PN100.
Les joints O-rings en élastomère se retrouvent plutôt vers les faibles pressions et pour des
mouvements de tige en de tour.
Les presse-étoupe en PTFE et en composés à base de PTFE se retrouvent en majorité vers les
pressions moyennes.
Quant à la taille des robinets qualifiés, celle-ci évolue du DN40 au DN250 avec un maximum
de 8 essais pour le DN 80 (6 essais pour le DN100).
Si l’on analyse la répartition de ces systèmes d’étanchéité avec la température, nous pouvons
faire les remarques suivantes :
o Le graphite et les composés à base de graphite se retrouvent à tous les niveaux de
température.
o Les joints toriques ont principalement été testés à température ambiante (RT), et le
PTFE principalement entre 100 et 200°C.
o Un seul essai est effectué en température négative, avec des garnitures composées à
base de graphite.
On peut tout d’abord remarquer que la majorité des robinets a atteint la classe B. Quelques
robinets ont atteint la classe A, mais pour une pression interne inférieure à 20 bar. Au-delà de
100 bar, les robinets sont classés C et sont équipés de garnitures base graphite .
INDEX çç Sommaire
4
C3
Classes d'étanchéité
B2 C 3
2
B
1
A1 A
0
0 10 20 30 40 50 60 70
0
0 50 100 150 200 250
Pression (bar)
Les robinets classés A ont certes été testés à basse pression, mais à des températures de 250 et
450°C. Les robinets sont équipés soit de garnitures à base PTFE, soit de garnitures à base de
graphite (graphite dopé), soit d’un soufflet avec garniture en graphite pur en aval
(configuration TA Luft).
Les robinets classés B sont assez bien répartis entre -46°C et 425°C. Les types d’étanchéités
utilisés vont des garnitures bases graphite (essai à -46°C), aux joints élastomères et garnitures
PTFE (température ambiante jusqu’à 200°C), avant de revenir vers les garnitures en graphite
(ou graphite dopé).
C3
Classe d'étanchéité
B2
A1
0
-100 0 100 200 300 400 500
Temperature d'essai (°C)
INDEX çç Sommaire
En fonction du nombre de cycles appliqués (classe d’endurance)
La figure 10 ci dessous montre tout d’abord un nombre important de robinets [10] classés B-
CO3.
Ceux-ci présentent majoritairement un mouvement de tige de tour (robinet à boule), et sont
équipés d’O-rings en élastomère (faible pression, température proche de l’ambiante) comme
étanchéité de tige.
CO36 5 10 2
CO25 1
CO14 4
CC33 2
CC22 4
CC11 1 2
0
0 1A 2 B 3 C 4
Contrairement aux vannes de régulation, la classe intermédiaire CO2 pour les robinets de
sectionnement n’apparait qu’une fois (4 fois pour CC2). En effet, une fois le cap que
représente la classe CO1 dépassé, les robinetiers visent en priorité la classe CO3. Le robinet
qui n’a pas pu atteindre les 2500 cycles fatidiques est alors classé CO2.
Tous les robinets classés en CO1 présentent un mouvement de tige en translation, neuf
robinets sur dix classés CO3 ont un mouvement de tige en de tour. Cette observation
montre qu’un mouvement linéaire de tige est probablement plus contraignant pour les
étanchéités qu’un mouvement de rotation.
Tous les robinets classés A sont des robinets de sectionnement, et ont atteint la classe CO3.
Ce résultat laisse supposer qu’il est plus difficile d’atteindre ce haut niveau de qualification
pour les robinets de régulation, tous classés B ou C. La procédure des vannes de régulation
serait-elle plus sélective ? Avant d’arriver à cette conclusion, il est important de signaler que
la majorité des robinets de régulation qui ont été testés était à mouvement linéaire, condition
jugée précédemment plus contraignante en termes d’endurance.
INDEX çç Sommaire
INVESTIGATION D’UN FLUIDE DE TRAÇAGE ALTERNATIF BASE SUR
L’HYDROGENE
Contexte
Après avoir réalisé un état des lieux (principe, matériel existant, domaines d’utilisation,
sécurité, limites éventuelles) sur la technologie de mesure utilisant un tel mélange azote-
hydrogène comme gaz traceur, le CETIM a testé la faisabilité de réaliser des mesures
d’émission fugitives sur un presse-étoupe de robinet en utilisant le protocole d’essai défini
dans la norme ISO 15848-1. Un des avantages annoncés d’un tel mélange azote-hydrogène est
son coût inférieur à celui de l’hélium.
Le protocole d’essai implique la réalisation d’essais ISO 15848-1, de type CO1, à deux
niveaux de température (-50°C et +200°C) sur une vanne 4’’, CLASS 300, équipée d’un
presse étoupe à base de graphite. Les essais sont réalisés avec de l’hélium et avec un mélange
Azote/Hydrogène,
Pour chaque configuration, on réalise chaque essai 3 fois afin d’étudier la répétabilité de la
procédure.
Pour chaque essai on mesure un taux de fuite en utilisant successivement différentes
méthodes:
• reniflage,
• globale sous vide,
• par accumulation
• par flushing.
Résultats d’essai
Les figures 11 et 12, montrent un exemple de résultat obtenu lors d’un essai impliquant une
température de -50°C, avec l’hélium et le mélange Azote/Hydrogène, pour les différentes
méthodes.
INDEX çç Sommaire
! " #
"
! #
!
"
! $ #
$
%
! % &
% ! !% " "% # #% $ $% %
)!#(* +(
*(
$""
)(
$&%
%
)!#(+ (
#)(
#*(
)!#(, #+(
#,(
#-(
)!#(- #.(
( )(( *(( +(( ,(( -((
Pour chacun des fluides, les mesures par le vide, par accumulation et par flushing donnent des
valeurs comparables. La mesure par reniflage donne un niveau de fuite inférieur d’une décade
ce qui confirme que cette méthode est une mesure d’une fuite locale non représentative de la
fuite totale du presse étoupe.
Les mesures en hélium et avec le mélange azote-hydrogène donnent des résultats comparables
en ce qui concerne le niveau d’émission. Il faut toutefois préciser que l’utilisation du mélange
azote-hydrogène ne permet pas, à priori, de mesurer des niveaux de performance de la classe
A de l’ISO 15848-1.
INDEX çç Sommaire
A l’heure de la rédaction de ce document, l’ensemble du programme d’essai n’a pas été
réalisé. Toutefois, sur la base des mesure, il n’a pas été rencontré d’obstacle à la mise en place
de cette méthode. De plus, les dispersions de mesures obtenues avec le mélange
azote/hydrogène et les valeurs moyennes sont du même ordre que celles obtenues avec
l’hélium. Ces premières constations seront à confirmer avec la réalisation complète du
programme d’essai.
CONCLUSION
En ce qui concerne la qualification des robinets aux émissions fugitives, il y a semble t il une
volonté de rapprocher, autant que faire se peut, les deux références actuelles que sont l’ISO
15848-1 et la MESC SHELL SPE 77/300.
Le point majeur de la discussion concerne la méthode de mesure des émissions mais il semble
admis aujourd’hui que la classe d’étanchéité des robinets doive être définie à partir d’une
mesure de la fuite totale du système d’étanchéité de la tige.
Il y a une demande forte des robinetiers (voire des utilisateurs) d’une unique norme de
référence.
La VDI 2440 et l’API622 concernent la caractérisation des garnitures de presse étoupe. Ces
deux références ont peu de points communs et leur différence essentielle concerne le fluide
d’essai (l’hélium pour la VDI 2440, le méthane pour l’API 622). L’API622 est plus complète
que la VDI 2440 car elle impose des cycles de fonctionnement (mécaniques et thermiques)
ainsi qu’un essai de corrosion.
Il semble que la VDI 2440 puisse se rapprocher de l’ISO 15848-1 amendée en termes de
classe d’étanchéité (et de méthode de mesure des émissions) et des cycles de fonctionnement.
Ainsi une qualification ISO 15848-1 d’un robinet pourrait conduire directement à une
qualification TA Luft de sa garniture d’étanchéité !
Enfin, il faut noter qu’il est possible de qualifier, sur bancs dédiés, des garnitures de presse
étoupe selon la norme ISO 15848-1.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page32
RÉSUMÉ / ABSTRACT
50F
ÉVOLUTION DU PANORAMA NORMATIF DE LA ROBINETTERIE
ET DES ACCESSOIRES DE SÉCURITÉ
J. Amaraggi
(UNM – F 92038 Paris La Défense cedex - e-mail : [email protected])
Mardi
RÉSUMÉ
Ce parc de normes est en constante évolution : Les publications les plus récentes et l’avancée des
travaux en cours seront présentés en exposé.
ABSTRACT
The list of European and international standards dealing with industrial valves and safety devices is
long. Among others, we’ll focus on standards for industrial valves addressing specifications on sizing,
pressure tests and means to determine valve shell mechanical resistance and those on safety devices.
27 standards are currently referenced on the Official Journal of the European Union (OJEU) presuming
conformity with the 97/23/CE directive on pressure equipment (PED), 10 of them coming from
international ISO standards.
This set of standards is continuously evolving: last publications and projects underway will be reviewed
during presentation.
INDEX çç Sommaire
Introduction
Qui normalise
Le comité technique en charge de produire et entretenir ces normes est le comité CEN/TC 69
« robinetterie industrielle ». Tous les pays européens y ont droit de siège. De présidence et
secrétariat français, ce comité gère un parc de normes organisé par famille de produits et/ou
application, avec pour chaque famille un groupe de travail (WG) voire un sous groupe de
travail (SG) dédié lorsqu'une révision ou une production de nouveau sujet est au programme
du comité.
INDEX çç Sommaire
CEN/TC 69/WG 10/SG 2 = Disques de rupture
CEN/TC 69/WG 10/SG 3 = Dispositifs de sûreté combinant soupapes de sûreté et disques de
rupture
CEN/TC 69/WG 10/SG 4 = Soupapes de sûreté pilotées
CEN/TC 69/WG 10/SG5 = Dispositifs de sécurité à décharge controlés contre les surpressions
CEN/TC 69/WG 11 = Robinets de régulation
CEN/TC 69/WG 12 = Robinets pour l'industrie de process
CEN/TC 69/WG 15 = Robinets à membrane
CEN/TC 69/WG 16 = Robinets thermoplastiques
CEN/TC 69/WG 17 = Robinets métalliques et thermoplastiques
On constate qu’à ce jour sur les 27 normes concernées, 10 sont issues de la filière ISO.
Les 27 normes harmonisées portent en tout ou partie sur les exigences de conception, choix
des matériaux, essais de performance et marquage des familles de produits concernées.
4 normes traitent des "accessoires de sécurité", à savoir les parties 1, 3, 4 et 5 de la série des
EN ISO 4126 concernant les dispositifs de soupapes de sûreté à action directe combinés
éventuellement avec des disques de rupture, les soupapes de sureté pilotées et les soupapes de
sureté asservies.
Nota : l’harmonisation de la partie 2 de la série EN ISO 4126 traitant des disques de rupture est en cours de
discussion et la norme n’est donc pas pour l’instant référencée au JOUE.
Chacune de ces normes précise en dernière page dans son annexe ZA la correspondance à
rechercher entre les exigences de la directive 97/23/CE et les paragraphes de la norme
apportant réponses techniques et donnant présomption de conformité à ces exigences. A titre
INDEX çç Sommaire
d’exemple les annexes ZA des normes EN 1171, EN 12266-1 et EN ISO 4126-1 sont
rapportées en annexe (2).
Evolutions en cours
En fonction des besoins exprimés par le comité CEN TC 69 ou les comités ISO/TC 138 SC 7
et 185, les normes peuvent être amenées à être révisées ou amendées pour répondre à l’état de
l’art en vigueur. Lorsque la révision ou l’amendement est décidé, on parle d’inscription au
programme du TC. Par ailleurs, tous les 5 ans à compter de la dernière date de publication de
chaque norme, une enquête appelée « enquête systématique », est menée pour évaluer
l’opportunité de la révision.
Sur les 27 normes harmonisées, au premier semestre 2010, 12 sujets sont inscrits au
programme. Aucun examen systématique sur ces normes harmonisées n’est planifié pour
l’exercice 2010.
La norme transversale sur les essais EN 12266-1 :2003 est en révision depuis 2008 pour
notamment adapter les conditions d’essais statistiques ainsi que les prescriptions d’essais en
général et actualiser en conséquence l’Annexe ZA. Le projet a passé favorablement l’étape
de l’enquête et devrait être soumis au vote final au second semestre 2010. L’enjeu est
important puisque, hormis la série des normes EN ISO 4126 (accessoires de sécurité), toutes
les autres normes harmonisées « produit » y font référence.
Concernant les robinets à soupapes et robinets à clapet libre, l’EN 13709 :2002 (acier) a été
révisée pour tenir compte de l’évolution de certaines références normatives et actualiser
l'annexe ZA. De même l’EN 13789 :2002 (fonte) a été révisée pour mettre en cohérence le
choix des matériaux et la résistance mécanique des enveloppes avec la série des normes
transversales EN 12516 sur les enveloppes de robinetterie. La publication des deux sujets a
été entérinée au CEN et les deux normes publiées en France en Août 2010.
Concernant les robinets vannes, la norme EN 12288 :2003 (cuivre) a été révisée pour mettre à
jour les références normatives et actualiser l'annexe ZA. Les tableaux sur les extrémités de
corps ont été modifiés pour adapter les exigences de choix de matériaux et les relations
Pression-Températures correspondantes. La norme EN 1984 :2000 (acier) a également été
révisée avec une actualisation de l’annexe ZA. Les deux révisions ont été entérinées au CEN
et publiées en France en Août 2010.
La norme EN 593 :2009 sur les robinets papillons, publiée en Août 2009 fait l’objet d’un
amendement pour donner une meilleure lisibilité sur un tableau de correspondance pression -
vitesse d’écoulement maxi du fluide en circulation applicable aux produits. Ce sujet, inscrit en
Mars 2010 au programme devrait être soumis à vote final au second semestre 2010.
INDEX çç Sommaire
Enfin concernant les accessoires de sécurité, les parties 1, 4 et 5 de la série des normes EN
ISO 4126, dont la première publication date de 2004, est en cours de révision. Pilotée par la
France, la révision porte entre autres sur un renforcement de la cohérence entre les différentes
parties de la série (définitions, essais, choix des matériaux, etc.), une meilleure définition des
types d’accessoires relevant de chacune des parties, une clarification du concept de
surpression des soupapes s’inspirant de l’ASME, ... Elle a passé à ce jour favorablement
l’étape de l’enquête (approbation entre autres de : Allemagne, Corée, Etats-Unis pour la partie
4,, France, Italie, Japon, Pays-Bas, Royaume-Uni, Russie). L’annexe ZA sera remaniée en
conséquence. La publication est prévue pour fin 2011, début 2012.
Pour 2011 les examens systématiques sur les normes suivantes sont à prévoir:
Pour la robinetterie industrielle :
o EN 12334 « Clapets de non-retour en fonte »
o EN 13397 « Robinets métalliques à membrane »
o EN 14341 « Clapets de non-retour en acier »
o EN 1983 « Robinets à tournant sphérique en acier »
o EN ISO 16135 à 16139 et EN ISO 21787 « Robinetterie industrielle en matériaux
thermoplastiques »
Pour les accessoires de sécurité :
o EN ISO 4126-3 'Dispositifs de sécurité pour protection contre les pressions
excessives - Partie 3 : dispositifs de sûreté combinant soupapes de sûreté et disques
de rupture'
INDEX çç Sommaire
(1)
Références des normes harmonisées
1 - robinetterie
Référence titre
EN 19 Marquage des appareils de robinetterie métalliques
EN 593 Robinets métalliques à papillon
EN 1171 Robinets-vannes en fonte
EN 1349 Robinets de régulation des processus industriels
EN 1983 Robinets à tournant sphérique en acier
EN 1984 Robinets-vannes en acier
Essais des appareils de robinetterie - Partie 1 : essais sous pression, procédures
EN 12266-1 d'essai et critères d'acceptation - Prescriptions obligatoires
EN 12288 Robinets-vannes en alliage de cuivre
EN 12334 Clapets de non-retour en fonte
Résistance mécanique des enveloppes - Partie 1 : méthode tabulaire relative aux
EN 12516-1 appareils de robinetterie en acier
Résistance mécanique des enveloppes - Partie 2 : méthode de calcul relative aux
EN 12516-2 enveloppes d'appareils de robinetterie en acier
Appareils de robinetterie - Résistance mécanique des enveloppes - Partie 3 : méthode
EN 12516-3 expérimentale
Résistance mécanique des enveloppes - Partie 4 : méthode de calcul relative aux
EN 12516-4 enveloppes d'appareils de robinetterie en matériaux autres que l'acier
EN 13397 Robinets métalliques à membrane
EN 13709 Robinets à soupape et robinets à clapet libre blocable en acier
EN 13789 Robinets à soupape en fonte
EN 14341 Clapets de non-retour en acier
EN ISO 16135 Robinets à tournant sphérique en matériaux thermoplastiques
EN ISO 16136 Robinets à papillon en matériaux thermoplastiques
EN ISO 16137 Clapets de non-retour en matériaux thermoplastiques
EN ISO 16138 Robinets à membrane en matériaux thermoplastiques
EN ISO 16139 Robinets-vannes en matériaux thermoplastiques
EN ISO 21787 Robinets à soupapes en matériaux thermoplastiques
2 - accessoires de sécurité
Référence titre
Dispositifs de sécurité pour protection contre les pressions excessives - Partie 1 :
EN ISO 4126-1 soupapes de sûreté
Dispositifs de sécurité pour protection contre les pressions excessives - Partie 3 :
EN ISO 4126-3 dispositifs de sûreté combinant soupapes de sûreté et disques de rupture
Dispositifs de sécurité pour protection contre les pressions excessives - Partie 4 :
EN ISO 4126-4 soupapes de sûreté pilotées
Dispositifs de sécurité pour protection contre les pressions excessives - Partie 5 :
EN ISO 4126-5 dispositifs de sécurité à décharge contrôlés contre les surpressions (DSDCS)
La liste des normes harmonisées avec leurs amendements et corrigenda éventuels est disponible à
l’adresse :
http://ec.europa.eu/enterprise/policies/european-standards/documents/harmonised-standards-legislation/list-
references/pressure-equipment/index_en.htm
INDEX çç Sommaire
(2)
Exemples d’annexes ZA
INDEX çç Sommaire
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page33
RÉSUMÉ / ABSTRACT
55F
RÉFÉRENTIEL D’AUDIT POUR LA SÉLECTION DES
RÉPARATEURS DE SOUPAPES DE SÛRETÉ
E. Allart
(Total Petrochemicals France - 6 allée Irène Joliot-Curie 69792 Saint-Priest - e-mail : [email protected])
R. Dunoyer
(Total Petrochemicals France - Usine de Gonfreville BP 86 76700 Harfleur - e-mail : [email protected])
RÉSUMÉ
Les soupapes de sûreté sont des accessoires de sécurité et doivent par conséquent être maintenues de
façon à leur permettre un fonctionnement fiable.
Les opérations de réparation et de révision des soupapes de sûreté doivent donc être confiées à des
entreprises ayant le savoir-faire et les compétences requises.
Afin de pouvoir sélectionner de telles entreprises, l’UFIP et l’UIC ont élaboré, à l’intention de leurs
adhérents, le document technique DT 88.
Il s’intéresse à la révision et à la réparation des soupapes de sûreté à ressort, de pression de début
d’ouverture supérieure à 0,5 bar.
Il comprend deux parties : La partie A qui définit les exigences « Qualité » comportant en annexe une
grille d’évaluation et la partie B qui définit les exigences « Techniques » qui comporte également une
grille d’évaluation en annexe.
ABSTRACT
Safety valves are part of safety accessories and have to be maintained in a manner that ensures their
reliability.
Bearing this in mind, the repair and maintenance of safety valves must be put into the hands of
competent Companies.
So as to enable the selection of competent Companies UFIP and UIC have developed technical
document DT88 for use by their members.
This document is destined to cover the repair and maintenance of spring operated safety valves with
an opening pressure greater than 0.5 bar.
The document is in two parts : The part A defines the “Quality” requirements and includes an
evaluation grid and the part B defines the “Technical” requirements and also includes an evaluation
grid
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
INTRODUCTION
Les soupapes de sûreté sont des accessoires de sécurité et doivent par conséquent être maintenues de
façon à leur permettre un fonctionnement fiable.
Les opérations de réparation et de révision des soupapes de sûreté doivent donc être confiées à des
entreprises ayant le savoir-faire et les compétences requises.
Afin de pouvoir sélectionner de telles entreprises, l'UFIP (Union Française des Industries Pétrolières)
et l'UIC (Union des Industries Chimiques) ont élaboré, à l'intention de leurs adhérents, le document
technique DT88.
Il comprend deux parties : La partie A qui définit les exigences « Qualité » comportant en annexe une
grille d’évaluation et la partie B qui définit les exigences « Techniques » qui comporte également une
grille d’évaluation en annexe.
Pour qu’un atelier de réparateur puisse être sélectionné, il faut qu’il ait obtenu une note supérieure ou
égale à la note minimale à la partie A et une note supérieure ou égale à la note minimale à la partie B.
Cependant, pour les réparateurs qui sont certifiés NF EN ISO 9001 (Systèmes de management de la
qualité – Exigences), il est possible de prendre en compte cette certification et donc, dans ce cas, de ne
pas utiliser la partie A de ce document.
D’autre part, chaque établissement qui désire utiliser ce document technique doit établir des règles
d’utilisation de ce document technique précisant en particulier la durée pour laquelle la sélection du
réparateur est valide.
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
En France, il existe plusieurs sociétés de réparation de soupapes, de petite taille, qui fournissent
des prestations très correctes mais qui ne sont pas certifiées ISO 9001. Il n’a pas paru opportun
d’imposer comme préalable cette certification ISO 9001 qui aurait été difficilement acceptable par
ces sociétés de petite taille. Cependant, il n’était pas envisageable d’occulter le management de la
qualité. Aussi, cette partie a pour objet de préciser les exigences minimales relatives au système de
management de la qualité pour les réparateurs de soupapes de sûreté.
Elle est complétée en annexe (annexe A1) par une grille d’évaluation « Qualité » du réparateur.
La partie qualité du référentiel a été adaptée aux réparateurs de soupapes de sûreté puisqu’elle
s’intéresse, par exemple, aux points suivant :
- contrôle et essais à la réception avant démontage ;
- contrôle et essais lors de la révision et en final ;
- documentation technique du fabricant de soupape ;
- fiches suiveuses de réparation.
INDEX çç Sommaire
Page 3
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
Cette partie a donc pour objet de préciser les exigences techniques minimales à respecter par les
réparateurs de soupapes de sûreté.
Elle est complétée en annexe (annexe B7) par une grille d’évaluation « Technique » du réparateur.
Mais, pour qu’un réparateur puisse exécuter une prestation de qualité, il importe que son client
(utilisateur de la soupape) lui fournisse un certain nombre d’informations essentielles et c’est la
raison pour laquelle cette partie débute par les informations à fournir par le client.
Le client doit fournir au réparateur un cahier des charges détaillé comportant en particulier :
- l’identification de la soupape, les conditions de service et les valeurs de PDO (Pression de
début d’ouverture), PR (pression de réglage) et CPI (contre-pression initiale) ;
- pour les produits figeants, la température à laquelle doit être réalisé l’essai de déclenchement
avant démontage et nettoyage ;
- dans la mesure du possible, le dernier compte-rendu d'intervention avec les préconisations
émises ;
- tout renseignement important relatif à l’historique de la soupape depuis sa dernière révision ;
- le nom et les coordonnées du correspondant technique à contacter en cas de besoin ;
- la méthode à utiliser pour vérifier l’étanchéité de la soupape ;
- l’indication de la présence ou non d’un représentant du client aux essais et au plombage ;
- le conditionnement particulier de la soupape s’il y a lieu ;
- les systèmes de peinture à utiliser ;
- l’indication de l’utilisation ou non de la vis de blocage de la tige ;
- la nuance du matériau à utiliser pour la plaquette servant à l’étiquetage ;
- les modalités d’envoi du procès-verbal d’intervention.
De même il convient que des précautions soient prises pour le transport et la manutention des
soupapes et des rappels sur ces précautions sont faits.
Ces deux points (informations à fournir par le client et transport et manutention des soupapes) ne
rentrent pas dans la grille d’évaluation puisqu’elles sont de la responsabilité du client.
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
Ces contrôles permettent d’avoir une information détaillée sur l’état de la soupape après sa
période de service.
Un document de suivi, dont le contenu minimal est défini en annexe (annexe B1), doit
accompagner chaque soupape tout au long des opérations de révision. Une copie de ce
document de suivi est adressée au client.
Dans le cas où des contrôles non destructifs ont été réalisés, les procès-verbaux de ces
contrôles doivent être transmis au client.
Il ne sert à rien d’avoir un constat sur l’état des tubulures si celui-ci est trop vague ou
ambigu pour pouvoir l’exploiter correctement par la suite. Aussi, il convient de rendre
compte de l’état des tubulures en utilisant la terminologie définie en annexe (annexe B2).
1.2.2 Démontage
Après avoir noté des informations relatives à la vis et à la bague de réglage, toutes les pièces
doivent être démontées et aussitôt identifiées par marquage ou étiquetage et les internes de la
soupape placés dans un bac répertorié. Un nettoyage des pièces doit être effectué.
Chaque pièce constitutive de la soupape doit être expertisée par examen visuel ou contrôle
non destructif adapté conformément aux dispositions définies en annexe (annexe B3) en
utilisant la terminologie de l’annexe B2. L’expertise doit être réalisée par une personne
compétente et habilitée et les contrôles par ressuage et magnétoscopie réalisés et interprétés
par des contrôleurs certifiés niveau 2 suivant la norme NF EN 473 (Essais non destructifs -
Qualification et certification du personnel END - Principes généraux). En cas de doute ou de
difficulté d’expertise, le réparateur fait appel au correspondant technique du client pour
l’assister.
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
1.2.3.1 Ressort
Outre la réalisation des contrôles définis en annexe (annexe B3), il convient de s’assurer,
de plus, que la pression de réglage de la soupape entre bien dans la plage de réglage du
ressort indiquée par le fabricant en fonction de l’orifice. Il convient également de
s’assurer que la durée de vie préconisée par le fabricant pour certains ressorts, ne sera pas
dépassée.
1.2.3.2 Corps
1.2.3.3 Internes
Les internes doivent être contrôlés conformément aux dispositions définies en annexe
(annexe B3). A la fin de chaque contrôle, doit être précisée, pour chaque élément
constitutif de la soupape, sur le document de suivi de révision, la suite à donner (à laisser
en l’état, à remplacer, à rectifier, à roder, etc.) en indiquant l’éventuelle dégradation
constatée.
Les travaux de re-conditionnement des pièces doivent être effectués conformément aux
exigences définies en annexe (annexe B4).
1.2.5 Remontage
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
Le distinguo est fait entre les services « Gaz ou vapeur » et « Liquide », sachant qu’une
soupape réglée sur un banc pneumatique et qui fonctionne en service « Liquide » s’ouvre
avec le liquide à une pression supérieure.
Les instruments de mesure doivent être gérés suivant la norme NF EN ISO 10012
(Systèmes de management de la mesure - Exigences pour les processus et les
équipements de mesure). Leur étalonnage doit être réalisé au moins tous les six mois et
être enregistré sur le registre des contrôles conservé par le réparateur
Il est rappelé l’importance de la propreté du banc et les différentes étapes du réglage sont
précisées. Pour la vérification de l’étanchéité des soupapes sur service « Gaz ou vapeur »,
le client est tenu de préciser la méthode à utiliser : KELLOG M 24 15 (Essais des
soupapes de sûreté) ou API Standard 527 (Seat Tightness of Pressure Relief Valves). La
méthode KELLOG M 24 15 est rappelée à titre informatif en annexe (annexe B5).
1.2.6.4 Plombage
Dès que la soupape a été reconnue « conforme » par la personne habilitée du réparateur,
elle est plombée immédiatement. Le matériau du plombage doit être choisi en accord
avec le client et adapté à la température de fonctionnement de la soupape et à son
environnement.
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
1.2.7 Protection
Lorsque le client demande l’emploi d’une vis de blocage, avant de déposer la soupape du
banc, la tige doit être immobilisée par la vis de blocage extérieure (cette dernière doit être
clairement apparente). Un étiquetage spécifique doit être apposé sur le dispositif pour alerter
le client sur la nécessité de dépose du dispositif avant mise en service. La vis de substitution à
mettre en place dès repose de la soupape sur l'installation est attachée au chapeau avec le joint
correspondant
Un séchage des internes doit être effectué et les orifices entrée et sortie doivent être obstrués.
1.2.8 Peinture
Après s'être assuré que les protections sont bien en place sur les orifices d'entrée-sortie, sur les
plaques signalétiques et joints divers, le revêtement spécifié par le client est appliqué sur tout
le corps et les faces de bride hors portée de joint.
1.2.9 Etiquetage
Chaque soupape doit être munie d’une plaquette métallique solidement attachée ou fixée sur
une partie inamovible de la soupape et comportant les indications suivantes :
- Nom du réparateur ;
- Date du réglage ;
- Valeur de la PR ;
- Repère de la soupape et/ou numéro d’immatriculation.
Les informations sur les lieux de livraisons et les documents de colisage doivent être clairs et
explicites.
Tout choc doit être évité pendant les manutentions et le transport. Les soupapes doivent
voyager en position debout et calée. Les orifices « Entrée-Sortie » sont maintenus obstrués par
les bouchons en plastique jusqu'à la mise en place sur le site.
Pour chaque soupape révisée, un procès-verbal d’intervention, dont le contenu minimal est
défini en annexe (annexe B6), est établi. Il est envoyé aux clients suivants les modalités (fax,
courrier électronique, courrier postal, … et délai) fixées par ce dernier.
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
La prise en compte du retour d’expérience sur l’utilisation du référentiel permettra de faire évoluer
celui-ci.
La parution de ce référentiel étant relativement récente (septembre 2009), le retour d’expérience sur
son utilisation est encore limité, mais malgré cela il a paru important d’en faire part dès à présent.
La partie Qualité reprenant la plupart des exigences de la norme NF EN ISO 9001, certaines
impositions peuvent paraître assez lourdes pour des petites sociétés et il conviendra de réfléchir à
la nécessité de maintenir celles qui ne paraissent pas strictement indispensables.
Cependant, la mise en application de ces exigences oblige à formaliser des pratiques, bien
souvent, peu usitées : revue de direction, indicateurs de fonctionnement et de performance,
définition des critères d’acceptation pour les soupapes, procédure de traitement des non-
conformités, …
Ainsi, au travers de ces différents points, les processus associés au suivi des soupapes se trouvent
dans une démarche d’amélioration permanente.
A ce jour, aucune difficulté majeure dans l’application de ce document ne nous a été signalée.
Il convient de souligner que pour l’élaboration du référentiel, les bonnes pratiques écrites ou non
écrites, en vigueur ou plus anciennes, ont été prises en considération. En outre, le projet a été
soumis à deux fabricants de soupapes de sûreté et leurs remarques pertinentes prises en compte
lors de la rédaction du document final.
Le point principal qui est apparu à l’utilisation est le suivant : Certains réparateurs, par manque
d’habitude, ont quelque difficulté à utiliser la terminologie sur l’état interne des tubulures d’entrée
et de sortie, terminologie qui figure à l’annexe B2.
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page34
RÉSUMÉ / ABSTRACT
51F
CONTRÔLE DE FONCTIONNEMENT
DE SOUPAPES DE SÛRETÉ SUR SITE
J. Delage
PRESTO CONTROLE
BP 104 Parc d’activités de l’aérodrome 76430 Saint Romain de Colbosc
Mardi
E-mail : [email protected]
RÉSUMÉ
Cette activité permet de s’assurer du bon fonctionnement des soupapes de sûreté sans avoir besoin de
les déposer et sans perturber la marche des installations qui restent en fonctionnement normal, tout ceci
en ouvrant la soupape de quelques dixièmes de millimètre pendant quelques dixièmes de seconde. La
précision de la mesure de la pression de début d’ouverture est inférieure à 0.3 % alors que la norme
française E 29-411 (remplacée par NF EN ISO 4126-1 en juillet 2004) tolère des écarts de ± 3 %. Cette
précision à été certifiée après essais sur installation, par le service métrologie de l’Apave.
ABSTRACT
This technique allows to control on line the good operating and the correct calibration of the safety
valve without any dismountling and without any production’s disturbance. The process consists in, the
line remaining in full operation and without disturbance, opening the safety valve by a few 1/10
millimeters during a few 1/10 seconds. The precision of the set pressure measurement is less than 0,3%
despite the French Norm E 29-411 (followed by Norm NF EN ISO 4126-1 dated July 2004) is
specifying gaps of ± 3 %. After tests at site, the precision of the equipment has been certified by the
Measurement Department of APAVE.
INDEX çç Sommaire
Presto Controle
INDEX çç Sommaire
INTRODUCTION
Cette activité permet de s’assurer du bon fonctionnement des soupapes de sûreté sans avoir
besoin de les déposer et sans perturber la marche des installations qui restent en
fonctionnement normal, en provocant l’ouverture maitrisée de la soupape de quelques
dixièmes de millimètre pendant quelques dixièmes de seconde.
Le contrôle sur site avec le procédé Acressor, utilisé par Presto Controle, permet de s’assurer
du bon fonctionnement de la soupape de sûreté sur une installation en service, avec la reprise
éventuelle de la valeur de réglage et permettant de respecter :
• les réglementations en vigueur (A. M. du 04/12/98- A. M. du 15/03/00)
• les préconisations des constructeurs (Notices des constructeurs suivant DESP 97/23
CE)
Il évite la dépose systématique des soupapes, imposée lors des visites périodiques des
équipements sous pression quelles protègent. Seulement, en moyenne, 11 % des soupapes
feront l’objet d’une dépose pour remise en état en atelier, suite aux résultats de nos essais.
Le service métrologie de l’APAVE à certifié que la tolérance sur la mesure de la Pdo par le
procédé Acressor est inférieure à 0,3 %, alors que la norme française NF EN ISO 4126-1
tolère des écarts de ± 3 %.
Le procédé Acressor peut être utilisé sur les soupapes vapeur à ressort, à contrepoids, les
soupapes gaz ou liquide munies d'un soufflet ou ayant une faible contre pression ainsi que les
soupapes pilotées isolables jusqu'à des pressions de début d’ouverture de 400 bar.
Il est utilisé sur 250 fluides différents dont l’hydrogène, gaz naturel, air, ammoniac, azote,
benzène, butadiène, butane, brut, eau, essences, éthane, éthylène, fuel, gaz de craquage, GPL,
huiles, H2S, méthane, pentane, propane, propylène, solvants, styrène, soude, vapeur, xylène…
Par ailleurs, des équipements classés Atex ou Ex sont disponibles pour contrôler les soupapes
fonctionnant sur gaz et vapeur en zone explosive ou Atex.
INDEX çç Sommaire
SOUPAPES DE SURETE : NORMES, REGLEMENTATIONS
"Article 4 : à l’occasion des visites périodiques des appareils, la surveillance des soupapes de
sûreté comprend :
1 / Une vérification, en accord avec les états descriptifs des appareils, montrant que les
soupapes installées sont bien d’origine ou à défaut des soupapes assurant une protection des
équipements adaptée au processus industriel développé
2 / La réalisation en accord avec le processus industriel et les fluides mis en œuvre, d’un
contrôle de l’état des éléments fonctionnels des soupapes ou d’un essai de manœuvrabilité
adapté, montrant que les soupapes sont aptes à assurer leur fonction jusqu’à la prochaine
visite "
Arrêté Ministériel du 15 mars 2000 relatif à l’exploitation des équipements sous pression.
Lors des inspections périodiques :
" Article 11 :
Un examen des accessoires de sécurité et des investigations complémentaire en tant que de
besoin."
" Article 12 :
L’inspection périodique porte également sur le fonctionnement des accessoires et dispositifs
de sécurité … "
Normes françaises.
INDEX çç Sommaire
• Pression de début d'ouverture (Pdo) (1):
Pression effective à laquelle le clapet de la soupape de sûreté commence à s'ouvrir
dans les conditions d'exploitation.
La tolérance de réglage est de ± 3 % pour une pression de réglage > 5 BAR.
La tolérance de réglage est de 0,15 BAR pour une pression de réglage 5 BAR.
• Surpression (3):
Augmentation de pression par rapport à la pression d’ouverture généralement
exprimée en pourcentage de la pression de début d’ouverture.
• Accumulation (6):
Augmentation de pression par rapport à la pression de calcul et exprimée en
pourcentage de la pression de calcul.
Pression d’ouverture
(7)
Accumulation (6) Surpression
Pdo max (3)
Pression maximale admissible Ps
Pdo
mini
Chute de pression
Pression de refermeture à la refermeture
(5)
Pression d’étanchéité
Pression d’utilisation de l’équipement (P) (2)
INDEX çç Sommaire
LE PROCEDE DE CONTROLE DE SOUPAPES SUR SITE
Pour une soupape réglée, le clapet est maintenu sur le siège par le ressort d’une force (F)
équivalente à la section moyenne au niveau de l’étanchéité du siège et du clapet (S) multiplié
par la pression de début d’ouverture (Pdo).
Ps
FORCE SUR LE CLAPET
AVEC PRESSION DE SERVICE
F = Pdo x S - P x S
Lors de la mise en service de l’équipement, le clapet de la soupape est maintenu sur le siège
par le ressort, d’une force (F) équivalente à la section moyenne au niveau de l’étanchéité du
siège et du clapet (S) multiplié par la pression de début d’ouverture (Pdo) diminué de l’effort
de la pression dans l’équipement (P) multiplié par la section (S).
Un dispositif mécanique ou hydraulique est installé sur la soupape, permettant d’exercer une
traction sur la tige ou pointal et décoller le clapet, ce dispositif est instrumenté avec :
• un capteur de force (mesure de la force F).
• un capteur de déplacement (détection du décollement du clapet du siège).
La pression dans l’équipement est mesurée en continu à l’aide d’un capteur.
L’ensemble de ces capteurs utilisés sont reliés aux étalons Nationaux. Leurs certificats
d’étalonnage bénéficient de la garantie du COFRAC. L’ensemble des chaines de mesures de
Presto Controle sont étalonnées annuellement.
Schéma de principe
POMPE
HYDRAULIQUE
TRACEUR
CIRCUIT HYDRAULIQUE
DE SECURITE
CAPTEUR DE PRESSION
CIRCUIT HYDRAULIQUE
AIR SERVICE
POUR L'OUVERTURE
INDEX çç Sommaire
Conditions d’intervention
Soit une soupape ayant une Pression de début d'ouverture (Pdo) de 15,4 Bar, une section (S)
de 100 cm, installée sur un appareil dont la pression de service (P) mesurée est de 8,8 Bar.
Nous avons dessiné deux graphes, ci-dessous, montrant l'influence d'un réglage :
INDEX çç Sommaire
1) Analyse du graphe avant réglage (graphe rouge)
La force de traction sur la tige au moment du début d'ouverture est de : 540 daN
D’où :
2) Après réglage de la vis de compression d' tour, nous obtenons un nouveau graphe
que nous analysons. (graphe vert)
La force de traction sur la tige au moment du début d'ouverture est maintenant de : 650 daN
D’où :
INDEX çç Sommaire
Inspection visuelle
Lors des essais de contrôle de fonctionnement, l’opérateur de Presto Controle effectue une
inspection visuelle qui consiste à vérifier tous les éléments extérieurs visibles de la soupape
ainsi que son environnement.
Cela commence par le piquage process de la soupape et prend en compte les problématiques
de l'échappement qui sont souvent source de dysfonctionnements majeurs.
1
10
4
11
6 12
7 13
8 14
INDEX çç Sommaire
Après essai, l’opérateur de presto Controle plombe la soupape avec mise en place d’une
étiquète gravée indiquant la date de l’essai, la valeur de la pression de début d’ouverture et le
numéro d’identification de la soupape dans notre base de données.
Lors de son retour en agence, l’opérateur de Presto Controle rédige un procès verbal de
contrôle de fonctionnement reprenant les résultats des essais et inspections visuelles. Ce
document peut être complété d’une fiche d’anomalie ou d’un rapport d’expertise.
INDEX çç Sommaire
AVANTAGES DU PROCEDE
CONCLUSION
Le contrôle de fonctionnement de soupapes avec le procédé Acressor proposé par Presto
Controle a permis, depuis les 20 dernières années, de valider l’état de plusieurs dizaines de
milliers de soupapes sur la pluparts des sites industriels Français et à l’étranger, et reste une
alternative fiable dans l’application de la réglementation actuelle, demandant de plus en plus
de contrôles des accessoires de sécurité, tout en permettant à l’industriel de maintenir sa
production pour un coût 3 fois moindre en moyenne que les autres solutions.
INDEX çç Sommaire
VALIDATIONS DU PROCEDE DE CONTROLE DE FONCTIONNEMENT DE
SOUPAPES SUR SITE.
L’activité de contrôle de soupapes sur site était proposée par Presto Fuites jusqu’en 2002
(date de la création de Presto Controle
Apave en 1994
INDEX çç Sommaire
Apave en 1998
INDEX çç Sommaire
AIB VINÇOTTE en 1999
INDEX çç Sommaire
APAVE en 2001
INDEX çç Sommaire
DRIRE en 2003
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page35
RÉSUMÉ / ABSTRACT
61F
COMMENT JUSTIFIER L’ABSENCE DE CONTRÔLES INTERNES
DANS DES USINES À HAUT RISQUES
J. B. Wintle, P. Moore
TWI, Granta Park, Great Abington, Cambridge, CB21 6AL
Email: [email protected]
ABSTRACT
Non-invasive examination has become the standard practice for inspection of vessels used for storing
pure liquefied gases at cryogenic temperatures. Certain sectors of industry are now considering whether
it may be possible to justify such an approach for other categories of process vessel containing pure
products under pressure. This paper describes the results of a study investigating when a change to the
nature of inspection from internal examination to a non invasive regime, and /or extended intervals
between internal examinations, may be justified.
The views of leading companies active in applying risk based inspection were obtained on the factors
that a risk based inspection (RBI) assessment should take into account. It is proposed that any change
of the requirement for internal examination should be justified on three legs: technical evidence,
adequate safeguards and proactive management. The paper will describe these legs and the practical
limitations and implications of applying this approach.
INDEX çç Sommaire
INTRODUCTION
For process vessels containing non-corrosive, clean and dry products under stable and benign
conditions, there may not be any degradation mechanisms affecting the inside of the vessel for
the foreseeable future. When this has been shown, then it may be possible to justify a change to a
more non-invasive inspection regime with longer intervals between invasive internal
examinations. Even so, it is important that this judgement is made carefully, according to a
defined process, and that adequate safeguards are in place so appropriate action may be taken if
the expected conditions of operation or process fluids change.
The UK Health and Safety Executive commissioned TWI to investigate under what
circumstances if any a less invasive inspection regime could be justified for process plant
containing high hazard fluids, given that this had been the practice for cryogenic gas plant [1].
This work has led to publication of a six-step process for assessing and justifying internal
examination requirements for high hazard process plant [2]. It has been developed partly through
consultation with leading petrochemical companies and inspection bodies operating in the UK,
and builds on earlier UK work on risk based inspection as a part of plant integrity management
[3, 4].
From the visits instances were found of pressure vessels where longer intervals between internal
examinations had been specified than might have conventionally been the case. Such decisions
not to routinely open and enter a vessel for inspection were subject to periodic review. This
meant that the interval between invasive examinations could be extended for a considerable
period, although in practice intervals beyond 12 years were uncommon. In some cases, the period
between outages was due to the restrictions of the process, such as residue build-up or reduced
efficiency or catalyst change. The opportunity to undertake some internal examination during
these outages was normal practice consistent with risk based inspection.
An overview of the six step process that can lead to sound decisions on the requirements for
internal examination is summarised in Figure 2. A RBI assessment team is formed and a rigorous
technical assessment is made of the possibility of active internal degradation of the vessel from
knowledge of the chemistry between the materials of construction and the process fluids. Where
INDEX çç Sommaire
no active degradation is foreseeable, the past operating and inspection history and other evidence
supporting (or countering) this conclusion is considered.
It is important that a robust management policy is in place to review the justification periodically
as well as in rapid response to any change to expected conditions. A company with an
organisational culture that has good communications between the operations and integrity
engineering teams is more able respond appropriately to any changes. A comprehensive
information management system can improve access to the operational and inspection history of
the equipment needed to make risk based inspection assessments. The RBI assessment must
outline clear limits on the conditions of operation over which it is valid.
There are inherent risks from carrying out invasive internal examination, which might actually
increase a previously low risk of vessel failure or endanger inspectors if suitable precautions are
not taken. In practice these risks can be effectively managed and are not on their own sufficient
justification to avoid an invasive examination if there are reasons to do one. An examination of
internal surfaces to which there is direct access would normally give better quality data than use
of a non-invasive NDT technique.
For some high hazard situations, where the scale and consequences of failure would be very high,
moving to a completely non-invasive inspection scheme is unlikely ever to be tenable consistent
with reducing risks ALARP (as low as reasonably practicable). In these cases there are grounds
for continuing to carry out some direct internal examination for reasons of defence in depth, even
though no active degradation mechanisms are identified.
INDEX çç Sommaire
Research [5] has shown that different RBI teams, be they from different sections of a single
company, or from different companies, can give very different results from their RBI assessments
due to their different levels of experience and rigour throughout the process. Provided all
approaches have conservative assumptions, some inconsistency does not necessarily matter.
Following authoritative guidance and training can help the RBI team achieve a consistency of
approach in line with good practice elsewhere.
When RBI teams lack sufficient independence from their company’s management and operations
the results of an RBI assessment may become skewed towards particular findings. In a similar
way, an RBI assessment may fail to assess all the potential damage mechanisms or risk factors.
This can occur if the RBI team has a ‘closed mind’ and assesses only what it expects to find, or
has insufficient imagination or thoroughness to consider less likely, but potentially dangerous,
scenarios. An independent consultant to chair or be part of the team is a possible solution to this.
ABSENCE OF DEGRADATION
Equipment for which a change to a less invasive inspection regime might be considered would
primarily be that described as in ‘clean non-corrosive service’. This means equipment that is not
expected to suffer from internal degradation mechanisms active over the time in normal service
or from transient events within the design basis. Where a RBI assessment identifies degradation
mechanisms (such as wall thinning, stress corrosion or fatigue cracking) that could affect the
inside of a vessel, a level of continued internal examination would be generally be expected.
Equipment in clean non-corrosive service’ is where the product is continuously sealed so that
water or other impurities cannot reside in the system, where stresses cycles are below code
fatigue thresholds, and where welding is of documented high quality. The kinds of chemicals
under consideration are mainly the range of organic chemical products, which in their pure forms
tend to be non-corrosive. It is important to assess the risk of contamination or ingress of water
into such systems, for example from leaks, openings or other parts of the system, which can leads
to degradation.
Many very pure chemicals that are considered corrosive are stored within vessels made of
suitable materials which do not corrode. Where there are no other degradation mechanisms, these
vessels can be treated as being in clean and non-corrosive service. However, the threat of
corrosion from even minor contamination with water or other impurities may be significantly
higher for these products, such as in the case of highly concentrated acids. The integrity and
operations of storage systems containing these products need very careful RBI assessment.
WHAT IF…?
If there are no identified corrosion mechanisms and nothing, other than pure product within the
operational pressure and temperature ranges, is ever contained in a vessel, then there can be high
confidence that no degradation will occur and internal examination will not find anything.
However, for a robust justification to be made, we must ask what happens if something changes
or goes wrong?
INDEX çç Sommaire
The most likely contamination of a clean product is from water. This can be from the production
process, leaks in steam heating or water cooling systems, moisture ingress from air as the product
is stored or transferred to/from storage, from steam cleaning during outage, or possibly as a result
of degradation of the product. Unlike cryogenic gas plant and other low temperature equipment,
where water is not a threat atmospheric storage tanks would normally require continued regular
internal examination. The risk from atmospheric water entering the system is that it will contain
oxygen and/or salt, which can result in rusting, corrosion, cracking or pitting.
Evidence of the absence of active degradation will be supported by the RBI assessment, in
conjunction with previous inspection records over the duration of the equipment’s lifetime. This
evidence can be difficult to obtain for third party inspectors when the contracts may be changed
every turnaround. Further evidence may come from other equipment operating under an
equivalent service, but the weight placed on this evidence must not be too high.
It is also possible that new understanding of degradation mechanisms can come to light for
products normally assumed to be non-corrosive. An example exists in the US experience of
ethanol stress corrosion cracking of steel [6]. Keeping up to date with world knowledge along
with regular review of the RBI assumptions is important. There are important roles which
regulators, industry bodies, professional institutions and international organisations can take in
improving the dissemination of experience and research.
SAFEGUARDS
Given that a less invasive inspection regime may be justified, there are safeguards and other
actions which may be taken to compensate for the benefits of more frequent internal examination.
Non-invasive inspection is not always cost saving. The methods can be slower and more
expensive to apply than direct internal examination, and require more highly trained operatives.
Welds, and in particular weld roots, are usually the most critical locations to inspect, but when
the weld caps have been ground flush, the roots can be hard to locate from the outside. External
ultrasonic examination can detect wall thickness loss or hydrogen induced cracking, but should
not be relied upon to find internal stress corrosion cracking or pitting.
INDEX çç Sommaire
intended simply to measure product quality may not, however, be suitable for detecting rogue
constituents or have the accuracy required for integrity assurance.
It is also important to ensure that bolts, seals and other ancillary components are inspected,
tightened, and overhauled at the required intervals, even though these may not coincide with
shutdowns for inspection. Consideration should be given to the opportunity provided to make an
internal examination whenever a vessel is to be taken out of service and opened for other reasons.
Arguments or measures that mitigate the hazard arising from a failure in the worst case identified
are a further safeguard. This may be a leak-before-break argument or possibly secondary
containment. Where the possibility of leakage is contemplated, duty holders should always assess
the means and time to detect a leak and the consequences.
MANAGEMENT OF CHANGE
Particular care is needed when feed, product or process conditions change, or when changes are
made to other parts of the plant, which could alter the fluid composition and level of impurities.
Changes to operations might be planned or they could be a result of unexpected upset or fault
conditions, which may be transient or permanent. The impact of such changes on the
susceptibility to degradation and the requirement for internal examination needs to be assessed.
Modern companies manage such changes through a Management of Change (MOC) policy.
A Management of Change committee should include relevant experts who can discuss the effects
of any changes or incidents on the RBI strategy. In leading companies, unexpected incidents
requiring the full Management of Change committee to meet may occur only infrequently.
However, the policy ensures that the RBI decisions remain valid under all eventualities.
Leading companies have systems in place to warn appropriate people of any relevant change to
the established parameters, such as level gauges, pH alarms, processing pressure and temperature
gauges. The warning alerts someone who has the understanding of the implications of the change
and the authority to convene the MOC team (or a suitable sub-set of the RBI team).
Good communications between the operations and integrity engineering teams is a key
requirement at this time. A responsive system that can act quickly is important. This can be more
INDEX çç Sommaire
difficult where the full RBI team may be spread across a number of companies or sites, or where
poor communications may create delays in response.
INFORMATION MANAGMENT
The large amount of equipment at a refinery, chemical or petro-chemical plant means that
managing the data about the equipment is best done using specialist software. The information
management system ideally needs a way of recording operational history, including for example
the number of start-ups/shutdowns, pressurisations, temperature transients and changes to
contained product. Trends or anomalies in the conditions of the equipment or product can
indicate ahead of time where issues might occur.
Handling, validating and maintaining data is a costly and time consuming business, but it does
allow on-going fact-based decision making. Having detailed past inspection reports readily
accessible is part of this process. Companies that have made this investment are more likely to
find it easier to justify inspection decisions.
The intention of the peer review is to examine the stages of the process and the factors considered
in reaching the inspection decisions. It is not intended that the peer review effectively does the
whole job a second time. A review of how the members of the team have been selected, how their
competencies and knowledge have been demonstrated and maintained, and what information has
been gathered for the RBI assessment is an essential part of the audit process.
CONCLUSION
1. When an RBI assessment foresees no internal degradation, then a move towards a more non
invasive inspection regime subject to periodic review may be justified for some process
plant containing high hazard products. This technical basis for the decision needs to be
robust and sufficient non-invasive inspection achievable and/or other safeguards employed.
2. In such cases it is important to have a proactive management regime that will review the
justification in response to any changes or excursions from expected operating conditions or
to process fluids, and to take opportunities for internal examination when they arise.
3. There may be some very high hazard processes where a non-invasive inspection approach is
unlikely ever to be tenable. High potential failure consequences may demand the need for
defence in depth consistent with reducing risks as low as reasonably practical.
INDEX çç Sommaire
References
[1] BCGA, 2004: ‘Approved Code of Practice on revalidation of cryogenic static storage tanks’,
BCGA CP25 Revision 2, 2004.
[2] HSE, 2009: ‘Establishing the requirements for internal examination of high hazard process
plant’, HSE Research Report 729/2009, HSE Books 2009. Download at:
http://www.hse.gov.uk/research/rrpdf/rr729.pdf
[3] HSE, 2006: ‘Best practice for risk based inspection as a part of plant integrity management’,
HSE Research Report 363/2001, HSE Books 2001. Download at:
http://www.hse.gov.uk/research/crr_pdf/2001/crr01363.pdf
[4] EEMUA, 2006: ‘Risk Based Inspection - A Guide to Effective Use of the RBI Process’,
EEMUA Technical Publication 206, 2006.
[5] HSL, 2002: ‘Risk Based Inspection -A Case Study Evaluation of Onshore Process Plant’
Health and Safety Laboratory Report HSL/2002/20. Download at:
http://www.hse.gov.uk/research/hsl_pdf/2002/hsl02-20.pdf.
[6] API, 2007: ‘Stress Corrosion Cracking of Carbon Steel in Fuel-Grade Ethanol: Review,
Experience Survey, Field Monitoring and Laboratory Testing’, API Technical Report API
939-D, 2007, Second Edition.
[7] Babcock, 2002: ‘Recommended practice for non-invasive inspections’ Mitsui Babcock
Energy Ltd. Group Sponsored Project 235.
[8] DNV, 2007: ‘Non-intrusive inspection’, Det Norske Veritas Recommended Practise DNV
RP G103, October 2007.
INDEX çç Sommaire
Technical Assessment
Less frequent
invasive
inspection?
Adequate Proactive
Safeguards Management
INDEX çç Sommaire
Form RBI team
• Include people with the all the required knowledge, experience and
competencies to carry out Risk Based Inspection (RBI)
• Choose a team leader with appropriate independence and authority
+
Show no active internal degradation
• Analysis of all possible damage mechanisms
• Combination of process fluids and containment material is stable and
inert
• Good quality (documented) fabrication
• Operating/inspection history of vessel
+
Justify inspection requirements
• Assess other evidence for internal degradation and possible threats
• Set date to review inspection decision
• Set operating/process limits for policy validity within a Risk Based
Management (RBM) framework
R d h i k l +bl i l (ALARP)
+
Implement Management of Change policy
• Good knowledge of limits to RBI assessment validity
• Reactive to equipment modification or changes in product
• Live process ie fast response to upset or abnormal conditions
+
Create an organisational and management awareness culture
• Maintain awareness from world experience
• Sustain the competencies and knowledge of the RBI team
• Peer review and audit the RBI process
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page36
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page37
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page38
RÉSUMÉ / ABSTRACT
27H
LE ROBINET DES BOUTEILLES DE GAZ,
LE DISPOSITIF DE CENDRILLON
D. W. Birch
(BOC Ltd, A Member of the Linde Group, The Priestley Centre,
10 Priestley Road, Guildford Surrey, GU2 7XY, UK )
Mardi
RÉSUMÉ
Depuis plus de cent ans les bouteilles de gaz comprimé ont été le fer de lance de l’industrie du gaz offrant
une manière pratique de transporter les gaz industriels de la compagnie de gaz vers l’utilisateur et son
lieu de travail.
L’histoire de la bouteille de gaz a été liée à plusieurs reprises, à l’évolution des aciers lourds vers les
aciers légers de haute résistance d’aujourd’hui et aux composites avec ou sans le recouvrement de
métaux qui représente le futur.
Un composant presque oublié dans cette histoire est celle du robinet. Avec la progression de la
technologie de bouteille, celle des robinets a suivi. Les robinets d’aujourd’hui empêchent l’humidité,
qui peut entraîner une rapide corrosion, d’entrer dans les bouteilles en acier de haute résistance. Elles
peuvent, ainsi être utilisées facilement à haute pression par les clients, elles offrent une plus grande
résistance au coup de feu avec l’oxygène à des pressions de 300 Bar et elles donnent une plus grande
résistance aux conditions difficiles de transport et d’utilisation.
Les robinets sont très différents des conceptions simples utilisées si on regarde les premières générations.
Cette présentation passera en revue le progrès accompli dans la technologie de robinet à partir du 19ème
siècle jusqu’à aujourd’hui avec une description des robinets utilisés sur le marché de nos jours et
soulignera le rôle essentiel joué par les normes ISO dans la conception, les tests et la compatibilité des
matériaux.
ABSTRACT
For well over one hundred years compressed gas cylinders have been the work-horse of the industrial
gases industry providing a convenient way to transport industrial gases from the gas company to the
end user and his workplace.
The history of the gas cylinder has been related many times, through its progression from heavy carbon
steels to today’s lightweight high strength alloyed steels and on to composites with and without metallic
liners that represent the future.
An almost forgotten component in this story of achievement is the cylinder valve. Just as cylinder
technology has progressed, so has that of cylinder valves. Today’s valves prevent moisture from entering
high strength steel cylinders and causing rapid corrosion, they can be operated easily at high pressures
by customers, they are very resistant to ignition in oxygen at pressures of 300 bar and seal tight even
under difficult conditions of transportation and use.
They are very different from the simple designs used when gas cylinders were in their infancy.
This presentation will combine a review of the progress made in cylinder valve technology from the 19th
century to present day with a description of cylinder valves used in the market now and will emphasise
the vital role played by ISO standards for design, testing and materials compatibility.
INDEX çç Sommaire
GAS CYLINDER VALVES, THE CINDERELLA DEVICE
D W Birch
(BOC Ltd, A Member of the Linde Group, The Priestley Centre, 10 Priestley Road,
Guildford Surrey, GU2 7XY, UK )
ABSTRACT
For well over one hundred years compressed gas cylinders have been the work-horse
of the industrial gases industry providing a convenient way to transport industrial
gases from the gas company to the end user and his workplace.
The history of the gas cylinder has been related many times, through its progression
from heavy carbon steels to today’s lightweight high strength alloyed steels and on to
composites with and without metallic liners that represent the future.
An almost forgotten component in this story of achievement is the cylinder valve. Just
as cylinder technology has progressed, so has that of cylinder valves. Today’s valves
prevent moisture from entering high strength steel cylinders and causing rapid
corrosion, they can be operated easily at high pressures by customers, they are very
resistant to ignition in oxygen at pressures of 300 bar and seal tight even under
difficult conditions of transportation and use.
They are very different from the simple designs used when gas cylinders were in their
infancy.
This presentation will combine a review of the progress made in cylinder valve
technology from the 19th century to present day with a description of cylinder valves
used in the market now and will emphasise the vital role played by ISO standards for
design, testing and materials compatibility.
RESUME
Depuis plus de cent ans les bouteilles de gaz comprimé ont été le fer de lance de
l'industrie du gaz offrant une manière pratique de transporter les gaz industriels de la
compagnie de gaz vers l'utilisateur et son lieu de travail.
L'histoire de la bouteille de gaz a été liée à plusieurs reprises, à l’évolution des
aciers lourds vers les aciers légers de haute résistance d'aujourd'hui et aux
composites avec ou sans le recouvrement de métaux qui représente le futur.
Un composant presque oublié dans cette histoire est celle du robinet. Avec la
progression de la technologie de bouteille, celle des robinets a suivi. Les robinets
d'aujourd'hui empêchent l’humidité, qui peut entraîner une rapide corrosion, d'entrer
dans les bouteilles en acier de haute résistance. Elles peuvent, ainsi être utilisées
facilement à haute pression par les clients, elles offrent une plus grande résistance
INDEX çç Sommaire
au coup de feu avec l’oxygène à des pressions de 300 Bar et elles donnent une plus
grande résistance aux conditions difficiles de transport et d'utilisation.
Les robinets sont très différents des conceptions simples utilisées si on regarde les
premières générations.
Cette présentation passera en revue le progrès accompli dans la technologie de
robinet à partir du 19ème siècle jusqu’à aujourd’hui avec une description des
robinets utilisés sur le marché de nos jours et soulignera le rôle essentiel joué par les
normes ISO dans la conception, les tests et la compatibilité des matériaux.
INTRODUCTION
Gas cylinder valves have been in existence as long as gas cylinders themselves, which
is more than 120 years. Just as the technology of gas cylinders has advanced in the
intervening years, so has that of the cylinder valve.
Acting as the last process valve in a gas company’s cylinder filling system and the
first process valve in a customer’s system for use of the gas, the humble cylinder
valve has a tough life and has also to meet the leak tightness and strength
requirements needed when being used at high and low ambient temperatures as well
as being transported, often over great distances, via bad roads, by air or by ship.
The objective of this paper is to chart the course of development of cylinder valves
from their original introduction to the present day and to explain the importance of the
various ISO standards which shape their design and testing.
INDEX çç Sommaire
Here is a valve that is typical of those
earliest designs. The body of the valve is
made of high tensile brass and the spindle
of a martensitic stainless steel.
Such valves have very few components
and the gland seal is able to be tightened
by customers via a large gland nut.
It is worth noting that the connection
thread between the cylinder valve and the
gas cylinder is a tapered thread of a
nominal size around 1 inch (25.4mm) and
would have been to a national standard.
Very similar threads are used today and
are now standardised by ISO globally.
This particular valve appears to have
suffered high levels of applied torque to
the spindle.
VALVE DESIGN
Component count is at the heart of cylinder valve design, reliability and cost of
manufacture:
– 1921 5
– 1980 12
– Today 17
INDEX çç Sommaire
Modern valves are made on highly automated production machinery with cylinder
valve bodies being made on a single “trunnion” machine with multiple spindle heads
that can typically fully machine 10 pieces per minute.
The requirement for durability of cylinder valves has changed as gas cylinder periodic
test frequencies have been extended. Early gas cylinders were subjected to inspection
and test every 2 years. Modern cylinders are normally on a 10 year test period. This
has influenced the quality of machining, screw cutting, and general limits and fits as
well as lubrication along with the gas compatibility and durability of elastomeric
materials such as valve seats and O ring seals.
In recent times gas cylinder valves have increasingly incorporated a residual pressure
device (RPV) that prevents gas in the customer’s process from entering the cylinder
and also retains a positive differential of 3-5 bar between cylinder pressure and valve
outlet pressure. Both these functions improve the quality of the gas supplied from the
cylinder in terms of contaminants (usually air and moisture). Additionally they
prevent damage being caused to the cylinder internally by such contamination.
Modern steel gas cylinders use high strength alloys that corrode readily in the
presence of moisture. Internal corrosion of such cylinders can take place rapidly and
lead to premature cylinder failure if undetected. Valves equipped with an RPV device
can prevent this problem and have enabled cylinder design to advance further with
lighter and lighter cylinders able to be filled to pressures of 300 bar and beyond. Such
RPV devices need to be functionally checked before every cylinder fill to reassure the
gas company that the cylinder is fit to fill. Such valves must be type tested to comply
with ISO 15996.
Such large diameter gas passages can compromise the strength of the valve in
customer service and in transport where valves can be subjected to large impact loads
sufficient to break them off of the cylinder.
During cylinder filling, gas is often admitted to the cylinder at very high velocities
and this imposes significant mechanical and temperature loadings (both high and low)
on items in the gas stream such as O ring seals. The sealing performance of valves is
critical after filling when, depending on the gas product and filling protocol they may
be hot (up to 65 degC) or cold (down to -20 degC). They must remain sealed as they
return to ambient temperature.
INDEX çç Sommaire
A recent development in cylinder valve design is the so-called Valve with Integrated
Pressure Reducer commonly abbreviated to VIPR. These are complex valves which
incorporate a pressure regulator that historically has been a separate (usually customer
owned) device.
VIPRs offer customers a very convenient
cylinder package and have found wide
acceptance, particularly in the medical
gases field.
They are more challenging to gas
companies because of their complex
filling protocols and, in the current
generation, their long term reliability.
Cylinder gases are used extensively in the silicon device industry where gases are
used for doping and etching silicon wafers. The extreme cleanliness and often
corrosive properties of these gases has resulted in a type of cylinder valve dedicated to
this arduous job. Such valves are made from electropolished materials such as 316L
stainless steel or pure nickel and have minimal wetted area within the valve seat zone.
Often used in gas cabinets for safety reasons, such cylinders may be equipped with
pneumatic actuators incorporated into the cylinder valves connected into site
emergency shut-down systems.
Connection standards such as BS 341, DIN 477 and NF E 29-650 gave details of the
connections used in UK, Germany and France respectively. In USA standard V1 from
the Compressed Gas Association was written in 1949 to address problems of outlet
thread compatibility identified during World War 2.
A later initiative has been the introduction of standards relating to the performance of
cylinder valves. With the development of European Norms has come the requirement
for prototype cylinder valves to meet a harmonised set of test conditions that will
ensure safe and reliable service throughout the wide range of conditions encountered
in service and during transportation.
INDEX çç Sommaire
EN 849 Transportable gas cylinders- Cylinder valves-Specification and type testing
was published in 1996 and is a landmark in the history of cylinder valve technology.
It incorporated type tests that had previously been used in different European
countries such as Germany, Belgium, France and UK into one document and was the
result of much discussion, intense argument and compromise. Subsequently it has
been replaced with a new standard, ISO 10297 Gas cylinders-Refillable gas cylinder
valves-Specification and type testing, which engendered similar levels of discussion
with additional USA, Japanese, South African and Australian input.
Special standards have been prepared to type-test RPV cylinder valves (ISO 15996)
and also the VIPR type valves both industrial (ISO 22435) and medical (ISO 10524-
3).
These standards require valves to be tested for endurance (2000 opening and closing
cycles) against pressures of 1.2 times working pressure.
Valves used in oxygen service are additionally subjected to an adiabatic pneumatic
impact test to prove that the internal design of the valve, with its use of suitable soft
sealing materials and heat-sink provision, can withstand harsh sudden exposure to
oxygen at elevated temperature and pressure. The rise time for the twenty pressure
pulses from 0 to 1.2 times working pressure is 20 milliseconds, which will cause the
valve to become very hot and exposes the materials within to sudden shock.
The choice of metallic and non-metallic materials in cylinder valves is guided by ISO
standards 11114-1 and 2.
The majority of cylinder valves for air gases such as oxygen, nitrogen, argon and
carbon dioxide are made from brass but those for many “special gases”, especially
corrosive gases, are made from stainless steel, aluminium silicon bronze and even
nickel based alloys such as Hastelloy TM. ISO 11114-1 gives details of compatibility
between a very large number of gases and metallic materials that guides the valve
designer and gas company technician who must specify valves.
Compatibility of gases with non-metallic materials such as used for O rings and soft
seat seals is of particular importance for safety and longevity. Manufacturer’s
literature is often vague regarding compatibility of their products with gases but the
gas companies and valve manufacturers have built up significant experience over
many years and this is embodied in ISO 11114-2.
INDEX çç Sommaire
CONCLUSION
Cylinder valves have progressed in the last 120 years from being very simple, rugged
devices operating at pressures of about 100 bar to sophisticated mechanisms operating
at 300 bar and incorporating residual pressure and non-return functionality. The
competitive nature of the gases industry has resulted in such devices being less
expensive in real terms today than their simple counterparts were at the end of the 19th
century. Design of modern valves ensures longevity to match the periodic
examination frequency of gas cylinders which has changed over the years from 2 to
10 years for most gases. Prototype testing and the recent requirement in Europe to Pi
mark cylinder valves under the Transportable Pressure Equipment Directive (TPED)
has ensured great emphasis is placed on the tests and detailed test protocols of ISO
10297 and similar standards.
The reliable ability of residual pressure and non-return devices in cylinder valves has
allowed gas cylinders to be used safely at 300 bar filling pressure as the device
protects the vulnerable high strength steel alloys from the corrosive effects of
moisture.
The latest advances including integrated pressure regulation take cylinder valves to a
new level of customer focus where the customer is able to use a fully checked
cylinder package operating at the peak of its efficiency and does not have to worry
about making connections at very high pressures between the cylinder package and
his user equipment.
For nearly a century cylinder valves changed very little and were apt to be ignored. In
the last 20 years enormous changes have been made and there is no reason to think the
pace of change will reduce.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page39
RÉSUMÉ / ABSTRACT
56H
LES ÉQUIPEMENTS SOUS PRESSION TRANSPORTABLES :
UNE RÈGLEMENTATION EN COURS D’ÉVOLUTION
O. Lefort
(MEEDDM/DGPR/BSEI – 92055 La Défense
e-mail : [email protected])
RÉSUMÉ
Le transport des équipements sous pression (par exemple, des récipients, des fûts et des bouteilles) est une composante
importante du secteur du transport de marchandises dangereuses. La directive 1999/36/CE du conseil du 29 avril
1999 garantit un niveau élevé de sécurité pour le transport des équipements sous pression transportables, tout en
permettant leur libre circulation et leur utilisation sur le marché européen grâce à des normes communes en matière
de conception, de construction et de contrôles ultérieurs. En ce qui concerne les exigences techniques, la directive
renvoie aux dispositions des accords internationaux sur le transport de marchandises dangereuses (ADR(1) et RID(2)).
Cette directive est en cours de révision dans le but de simplifier et de clarifier les exigences réglementaires
applicables aux équipements sous pression transportables. Cette révision répond également aux attentes des
autorités françaises qui considèrent qu’il est indispensable de renforcer les outils qui permettront d’assurer la
surveillance du marché et la qualité de l’intervention des organismes notifiés. En effet, elle intègre les évolutions
récentes de la législation communautaire relative à la commercialisation des produits sur le marché unique
européen, c’est-à-dire le « nouveau cadre législatif », qui comprend le règlement (CE) n° 765/2008 et la décision
n° 768/2008/CE. La proposition précise notamment les obligations de tous les opérateurs économiques intervenant
dans la chaîne de fourniture d’équipements sous pression transportables.
(1)
ADR: accord européen relatif au transport international des marchandises dangereuses par route.
(2)
RID: règlement concernant le transport international ferroviaire des marchandises dangereuses.
ABSTRACT
The transport of pressure equipment, e. g. tanks, receptacles, drums and cylinders, constitutes an important area of
the transport of dangerous goods. Council Directive 1999/36/EC of 29 April 1999 ensures a high level of transport
safety for transportable pressure equipment whilst allowing their free movement and use on the European transport
market through common norms for their design, construction and subsequent checks. For the technical requirements,
the existing Directive 1999/36/EC refers to the requirements contained in international agreements on transport of
dangerous goods.
This Directive is in revision in order to simplify and clarify administrative requirements relating to transportable
pressure equipment. This revision also meets french authorities expectations who consider that it’s necessary to
enhance tools to ensure market surveillance and the quality of the performance of notified bodies. Indeed, it includes
recent developments in Community legislation relating to the marketing of products on the single European market,
i.e. the «New Legislative Framework» Regulation No 765/2008/EC and Decision No 768/2008/EC. The proposal
details, in particular, the obligations of all economic operators involved in the supply chain of transportable pressure
equipment.
(1)
ADR: European Agreement concerning the International Carriage of Dangerous Goods by Road.
(2)
RID: Regulations concerning the International Carriage of Dangerous Goods by Rail.
AFIAP - ESOPE 2010 Programme
INDEX çç Sommaire
LES EQUIPEMENTS SOUS PRESSION TRANSPORTABLES, UNE
REGLEMENTATION EN COURS D'EVOLUTION
O. LEFORT
(MEEDDM/DGPR/BSEI – 92055 La Défense – e-mail : Olivier.Lefort@developpement-
durable.gouv.fr)
RESUME
Le transport des équipements sous pression (par exemple, des récipients, des fûts et des
bouteilles) est une composante importante du secteur du transport de marchandises
dangereuses. La directive 1999/36/CE du conseil du 29 avril 1999 garantit un niveau élevé de
sécurité pour le transport des équipements sous pression transportables, tout en permettant
leur libre circulation et leur utilisation sur le marché européen grâce à des normes communes
en matière de conception, de construction et de contrôles ultérieurs. En ce qui concerne les
exigences techniques, la directive renvoie aux dispositions des accords internationaux sur le
transport de marchandises dangereuses ( ADR ( 1 ) et RID ( 2 ) ).
Cette directive est en cours de révision dans le but de simplifier et de clarifier les exigences
réglementaires applicables aux équipements sous pression transportables. Cette révision
répond également aux attentes des autorités françaises qui considèrent qu'il est indispensable
de renforcer les outils qui permettront d'assurer la surveillance du marché et la qualité de
l'intervention des organismes notifiés. En effet, elle intègre les évolutions récentes de la
législation communautaire relative à la commercialisation des produits sur le marché unique
européen, c’est-à-dire le « nouveau cadre législatif », qui comprend le règlement (CE) n°
765/2008 et la décision n° 768/2008/CE. La proposition précise notamment les obligations de
tous les opérateurs économiques intervenant dans la chaîne de fourniture d’équipements sous
pression transportables.
INDEX çç Sommaire
ABSTRACT
The transport of pressure equipment, e. g. tanks, receptacles, drums and cylinders, constitutes
an important area of the transport of dangerous goods. Council Directive 1999/36/EC of 29
April 1999 ensures a high level of transport safety for transportable pressure equipment
whilst allowing their free movement and use on the European transport market through
common norms for their design, construction and subsequent checks. For the technical
requirements, the existing Directive 1999/36/EC refers to the requirements contained in
international agreements on transport of dangerous goods.
INTRODUCTION
Le 29 avril 1999, le Conseil a adopté une directive 1999/36/CE sur les équipements sous
pression transportables. Cette directive garantit un niveau élevé de sécurité pour le transport
des équipements sous pression transportables, tout en permettant leur libre circulation et leur
utilisation sur le marché européen des transports grâce à des normes communes en matière de
conception, de construction et de contrôles périodiques.
INDEX çç Sommaire
Évolutions réglementaires
En ce qui concerne les exigences techniques, la directive 1999/36/CE en vigueur s'appuie sur
les directives 94/55/CE ("directive ADR" relative au rapprochement des législations des États
membres concernant le transport des marchandises dangereuses par route) et 96/49/CE
("directive RID" relative au rapprochement des législations des États membres concernant le
transport de marchandises dangereuses par chemin de fer). Ces directives ont été abrogées par
la directive 2008/68/CE relative au transport intérieur des marchandises dangereuses avec
effet au 1er juillet 2009.
ECOSOC
Règlement type des Nations Unies
Recommendations pour le transport
de marchandises dangereuses
INDEX çç Sommaire
International Union européenne
Recommandations
de l ’ONU
Directives AD R/RID
Transport
Simplification et clarification
A la suite de ces récents changements législatifs, qui sont eux-mêmes la conséquence des
évolutions techniques des dix dernières années, les règles en vigueur dans l’Union européenne
sont devenues complexes et très difficiles à comprendre. En outre, en ce qui concerne
plusieurs points techniques, le chevauchement des règles européennes et internationales a créé
des incohérences qui rendent leur application plus difficile encore.
Il était donc nécessaire de simplifier et d’harmoniser les règles lorsque cela était possible.
L’objectif principal est d'éliminer les contradictions entre la directive en vigueur sur les
équipements sous pression transportables et les règles internationales en matière de transport
de marchandises dangereuses, notamment celles qui ont été incorporées dans la législation
communautaire au moyen de la directive 2008/68/CE. Les règles techniques et les procédures
administratives sont désormais contenues de manière plus logique dans une seule source – les
accords internationaux. La directive proposée porte essentiellement sur les aspects qui ne
peuvent être traités de manière satisfaisante que par la législation européenne.
INDEX çç Sommaire
Par exemple, les dispositions existantes concernant les modules relatifs aux procédures
d’évaluation de la conformité sont simplifiées et le projet de directive renvoie désormais pour
ce point aux accords internationaux applicables.
Ce «nouveau cadre législatif» définit un cadre général dont l’application doit, dans la mesure
du possible, être étendue à tous les secteurs industriels. Les principes qu’il énonce étant
applicables au marché des équipements sous pression transportables, ils ont été inscrits dans
le projet de révision de la directive. Cette inscription, si elle n’introduit pas de changements
fondamentaux, contribue à l’accomplissement de l’objectif majeur du «nouveau cadre
législatif» qui est d’harmoniser les règles du marché dans le plus grand nombre possible de
secteurs industriels tout en simplifiant l'application de ces règles par l'industrie.
En intégrant ces règles dans la proposition concernant le secteur particulier des équipements
sous pression transportables, la Commission ne définit pas de règles sectorielles spécifiques
lorsque les règles générales remplissent déjà cette fonction, contribuant ainsi à simplifier la
réglementation. On retrouve notamment dans la proposition une définition précise des
différents opérateurs économiques qui doivent être responsables de la conformité des
équipements sous pression transportables aux règles de sécurité et d’accès au marché eu égard
à leur rôle respectif dans la chaîne d’approvisionnement.
Les obligations opérationnelles des organismes notifiés sont définies, y compris leur
reconnaissance mutuelle ainsi que les obligations en matière d’information, notamment
lorsqu’un certificat a été refusé ou retiré. Il est prévu un échange d’expérience et une
coordination entre les autorités responsables de la politique de notification, celles chargées de
la surveillance du marché et les organismes notifiés.
Enfin, des procédures sont définies en matière de sécurité et de prévention des risques aux
niveaux national et communautaire.
INDEX çç Sommaire
CONCLUSION
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page40
RÉSUMÉ / ABSTRACT
57H
BOUTEILLE COMPOSITE HAUTE PRESSION
RÉSUMÉ
Cette présentation fournit des informations sur le passé, le présent et le futur des «bouteilles
composites». Elle donne une vue générale sur les types de bouteilles, les normes de conception, les
possibilités de fabrication et leurs performances. Elle présente le nombre important de bouteilles
utilisées aujourd’hui dans un large panel de marchés et d‘applications. Elle aborde également les
opportunités pour le futur de ces bouteilles ainsi les possibilités de développement.
ABSTRACT
This presentation will provide information on the past, present and future of «Composite Cylinders».
It will provide an overview of the different cylinder types, design standards, manufacturing capabilities
and their performance. It will report on the very significant quantities of composite cylinders being
used in the market place today and show the wide variety of markets and applications that they are
used in. Highlighting the opportunities for composite cylinders and what developments are likely for
the future.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page41
RÉSUMÉ / ABSTRACT
58H
TECHNOLOGIE DES BOUTEILLES DE GAZ
H. Barthélémy
Air Liquide, 75 quai d’Orsay, Paris, 75321, France
[email protected]
RÉSUMÉ
L’objet de ce papier est de présenter un exposé historique et technique des bouteilles à gaz ainsi que de
fournir des détails concernant les problématiques et contraintes spécifiques. Depuis le début du 20è
siècle, les gaz industriels sont stockés dans les bouteilles d’acier sans soudure. A la fin des années
1960, les bouteilles en aluminium ont également été utilisées pour le stockage de gaz, mais à un coût
plus élevé et pour de plus petites capacités en eau par rapport aux bouteilles en acier. Les bouteilles
métalliques peuvent être frettées afin d’augmenter leur pression de service ou de diminuer leur poids.
Dans les années 1980, des bouteilles entièrement bobinées sont apparues pour des développements
spécifiques aux applications spatiales ou militaires. Grâce à leur poids réduit, elles ont commencé à
être utilisées pour des applications portables : véhicules (stockage à bord de gaz naturel), applications
de loisir (paint-ball), etc… Ces bouteilles composites entièrement bobinées, dites types III et IV, sont
aujourd’hui développées pour le stockage d’hydrogène énergie. Les problématiques spécifiques à la
pression de service très élevée (de 700 à 850 bar) sont également abordées.
Chaque technologie est décrite en termes de ses matériaux, ses technologies de fabrication et ses
épreuves. Les problématiques spécifiques à la haute pression sont décrites.
ABSTRACT
The topic of this paper is to give an historical and technical overview of gas cylinders and to detail the
specific issues and constraints. Since the beginning of the last century, industrial gases are stored in
seamless steel cylinders. At the end of the 60s, tubes also made of seamless steels were used. Aluminum
cylinders were also used for gas storage since the end of the 60s, but their cost was higher compared
to steel cylinders and smaller water capacity. To further increase the service pressure of gas tanks or
to slightly decrease the weight, metallic cylinders can be hoop-wrapped. Then, with specific
developments for space or military applications, fully-wrapped tanks started to be developed in the
80s. Because of their low weight, they started to be used in for portable applications: for vehicles (on-
board storages of natural gas), for leisure applications (paint-ball) etc… These fully-wrapped composite
tanks, named types III and IV are now developed for hydrogen energy storage; the requested pressure
is very high (from 700 to 850 bar) leads to specific issues which are discussed. Each technology is
described in term of materials, manufacturing technologies and approval tests. The specific issues due
to very high pressure are depicted.
INDEX çç Sommaire
COMPRESSED GAS STORAGE
Introduction
The following 4 types of high pressure vessels are classified (figure 1):
Type II: pressure vessel made of a thick metallic liner hoop wrapped with a fiber-resin
composite.
Type III: pressure vessel made of a metallic liner fully-wrapped with a fiber-resin
composite.
Type IV: pressure vessel made of polymeric liner fully-wrapped with a fiber-resin
composite. The port is metallic and integrated in the structure (boss).
The pressure vessels are generally cylinders, but composite vessels can also be polymorph or
toroid (figure 2).
Gases can be stored in the four types of pressure vessels. The choice of the storage is based on
the final application which requires a compromise between technical performances and cost-
competitiveness. H2 as industrial gas is stored in type I tanks, the pressure of which is from
150 to 300 bar (usually 200 bar). These are the most spread high pressure vessels today and
are the cheapest. When only higher pressures are required – mainly for stationary applications
– the type II tanks are preferred. Type III and type IV vessels are intended for portable
applications, for which weight savings is essential. However these vessels are much more
expensive. Technical performances will be presented in the last paragraph of this paper.
INDEX çç Sommaire
Type I cylinder Type II vessel Type III or IV vessel Toroid composite vessel
Figure 2 : Pressure vessels pictures
Some history
The oldest pressure vessel type is the type I, when made of steel. It was introduced in the 19th
century between 1870 and 1880. It was lead to the development of a new industry at that time:
the industrial gas business. More particularly, it seems to be first linked to the high
consumption of carbon dioxide for beverage, with the necessity to store it safely, in the liquid
state and in high quantities. The probable date of such storage for carbon dioxide seems to be
1874. Note that low pressure vessels (< 20 bar) were already used in the mid 19th century:
they were made of copper, generally low capacity and used for the very first breathable
apparatus; but they cannot be qualified as high pressure vessels. Regarding the hydrogen, it
was mainly used during the last 19s century for military observation balloons in Asia and in
Africa (around 1880). It was stored in 120 bar pressure vessels in wrought iron vessels in
1880; of course, these cylinders were very heavy (500 kg of steel were needed to store 25
Nm3 of hydrogen), but they seem to be the first to be able to transport safely hydrogen at high
pressure. Vessels made with seamless steel were introduced in 1885, manufactured by
drawing and forming of plates (Lane & Taunton British patent) with a special development
for making the neck. Meanwhile, the manufacturing of high pressure cylinder from seamless
tubes was developed (Mannesmann German patent). High pressure cylinders manufactured
from billets were also perfected in the late 1880s.
At the beginning of the 20th century, the three main processes, still used today for industrial
gas cylinders production, were existing. Only the wrought iron first process was given up. Of
course, some improvements were then performed during the 20th century. These
improvements concerned the end closing, the materials, the thermal treatments etc… For
example, the aluminum type I cylinders were introduced in after the First World War. The
motivation was to have a non magnetic storage for sub-marines. Their use was only really
spread after 1960.
INDEX çç Sommaire
Until the 60s, the service pressure of these storages was 150 bar. From 1960, the service
pressure was increased up to 200 bar. Today both 200 and 300 bar cylinders are co-existing
for industrial gases.
It is interesting to note that the core processes developed in the 19th century for type I vessels
are still used today.
High pressure composite vessels were introduced more than 70 years after and were
especially developed for space and military applications, for which technical performances –
especially weight – was a very important criteria. The experimentation of composite vessels
started in the 50s but the first high pressure vessels used for space and military applications
started actually in the 60s in the US (rocket motors and other pressure vessels for space
shuttles, sonar equipment etc…). The market of these new high pressure vessels was very low
in the 60s and the production was not regular at that time: manufactured batches could count
from about ten to a few hundreds vessels (no series production). They were made of a
metallic or polymeric liner wrapped with glass fiber composite. The civil market started to be
penetrated in the first 70s and was linked to the will to increase the market of these high tech
products. When compared to the type I conventional vessels for industrial gases, the cost of
these composite vessels and the lack of regulation for composite vessels slowed down this
penetration. For example, in the late 70s, 100 000 cycles were requested by the ASME code
for pressure vessels – what composites tanks could not fulfill. As a consequence, for each
model and each application, the composite vessels had to obtain a special authorization and
their lifetime was limited. However, the first important civil market in the 70s was the
breathable apparatus for firemen. From 80s on, these composite vessels started to be used for
skin diving, fuel storage (mainly natural gas) and leisure applications (like paint-ball). The
storage pressures were conventional: from 100 to 300 bar.
Many improvements were also performed since the first developments: for example, weight
has decreased, cycling performance has increased by using thin liners with adequate
mechanical properties, and other fibers than glass (kevlar, carbon). Moreover the regulation
was set up for both industrial gases and fuel gas storages.
Except for very specific applications (military) and type II for H2 trailers since 90s, the
composite vessels were not used for gas storage as industrial gas, because of their high cost.
Hydrogen started to be stored in composite vessel when the potential use of hydrogen as an
energy carrier began.
For all pressure vessels, the design shall take into account the service and test pressures, the
external stresses which are specific to the use (like impacts, aggressive media, vibrations,
temperature of service, weight of connectors etc…), the real lifetime (cycling) and the safety
coefficients defined for both static and dynamic conditions. The failure modes like plastic
deformation, buckling, creeping, fatigue etc… for metals, delaminations, fiber ruptures,
cracks, ageing etc… for composites are also taken into account for the design. All these
parameters define the mechanical design and the choice of the materials. The materials shall
INDEX çç Sommaire
also be compatible with the gas when in contact. It is important to note that metallic vessels
and composite vessels are very different:
The metal is isotropic, the composite is anisotropic: the mechanical properties are
concentrated in the fiber direction for the composite.
The failure modes are different.
The ageing is different.
For example, figure 3 gives the main strains which are generally considered for metallic
pressure vessels. In general, the domes are over designed. That is why the type II vessels, with
their hoop reinforcement on the only the cylindrical part of thick liners, can easily withstand
higher pressures.
For the composite wrapping, an analytical and simplified calculation is generally performed
(figure 4) to have a first estimation of the lay-up design. Then a complete study with a finite
elements software is necessary for a correct and optimized design (which should be coherent
with the filament winding machine code).
Figure 3 : Main strains considered for the metallic pressure vessels design (type I and metallic
liner)
INDEX çç Sommaire
Type I vessels can be manufactured from 3 different processes (figure 5 for the principles):
From plates: the process consists in deep-drawing metallic plated to form the shape (this
step can be performed many times to have the desired diameter and thickness); the neck
is formed by hot-spinning and the port is machined in the excess of metal coming from
the spinning step. A one port cylinder is thus obtained. The heat treatments are then
applied to have the desired mechanical properties.
From billets: the billet is firstly heated to allow the drawing to be performed. The
process is then similar to the previous one.
From tubes: Tubes are purchased and in general the original thickness is kept for the
hoop. The domes are formed by hot spinning and a 1 or a 2 ports cylinder can be
obtained. The process is then similar to the first one.
For each technology, quality controls of the materials used and of each step of manufacturing
is performed and traced. The liners of type II and type III vessels can be manufactured in the
same ways.
From the polymer or the monomers by the rotomolding process: the polymer (or the
monomers) is introduced in a mold the shape of which is the final liner shape. The liner
is made by heating and then cooling the mold while rotating (the fusion temperature or
the polymerisation temperature have to be reached). It can be one or two port liner. The
metallic bosses are introduced during the rotomolding step or stick on the liner before
wrapping.
From tubes: polymeric tubes (made by extrusion blow molding) and domes (equipped
with the metallic boss) are purchased at the desired diameter. Both shapes are welded to
form the liner.
INDEX çç Sommaire
For all composite vessels, the metallic or the polymeric liner is then hoop-wrapped or fully
wrapped with the composite with a filament winding machine. For cylinder vessels, 3
wrappings are possible: hoop, polar and helical (figure 6). Types II are only hoop-wrapped.
Type III and IV vessels are generally a combination of hoop and polar wrapping – but a
combination of the 3 wrapping can be considered. Many vessels can be wrapped in the same
winding machine if it is equipped for many winding heads. Once the liner is wrapped, the
resin must be cured. The curing is generally performed in ovens with the resin appropriate
heat treatment.
From a material compatibility point of view, the main issues for the hydrogen high pressure
vessels are:
The risk of hydrogen embrittlement of the steel: this phenomena leads to a premature
crack of the steel due to H atom dissolution and trap (stress corrosion cracking). The
main risk is the burst of the tank. A lot of efforts on H2 gas pressure embrittlement
understanding and prevention rules were conducted in the 70s and 80 s after many
accidents occurred with steel pressure vessels. The prevention rules based on 200 bar
cylinders were then defined.
The parameters which influence the embrittlement behavior of the pressure vessel are:
H2 purity, storage pressure, temperature, stresses and strain and time of exposure
(environment effects)
Microstructure, chemical composition, inclusion, mechanical properties and welding
quality (steel properties).
Stress levels, stress concentration, surface defects (design quality)
The first high pressure steel vessels were made of soft steel (carbon content < 0,25 %) and for
a long time, only this steel was authorized for pressure vessels. Today and since the 50s,
INDEX çç Sommaire
thanks to the steel metallurgy development, various compositions are suitable. Table 1
presents the conventional steels used for type I and type II vessels in the gas industry. By
extrapolation, the same requirements are defined for type III vessel steel liners. It is important
also to note that if the metallic boss of type IV is made of steel, the same requirements are
also required. For more information on the risk of hydrogen embrittlement of steels, see [1].
The permeation rated through the polymeric liner. The permeation is specific of type IV
vessels and is an inherent phenomenon for all gases in contact with polymers. It is the
results of the H2 gas dissolution and diffusion in the polymer matrix. It had been
identified at the beginning of composite vessels development and low permeable
polymers are one topic of research.
Because H2 is a small molecule, the diffusion and thus the permeation are enhanced. For
safety reason, the permeation shall be lower to a certain rate. This leads to the
development of special polymers which are suitable for hydrogen liners. In 2006,
polyethylene and polyamide (specific semi-crystalline grades) are the most used liners
for hydrogen energy type IV tanks.
No specific issue brought by hydrogen gas is met with aluminium alloys (except if presence
of mercury). The aluminium alloy grades used for the high pressure vessels are 6061 and
7000. They are not specific to H2 and are used for the types I, II and III vessels and also for
the metallic boss of type IV tanks.
Either glass, aramide or carbon fibers can be used for composite tanks wrapping. These fibers
are characterized by their tensile modulus, tensile strength and elongation. Table 2 gives the
usual ranges of these mechanical properties for each category of fiber. Hydrogen is stored in
very high pressure vessels when used for hydrogen energy applications (service pressure ≥
350 bar). As a consequence, from a mechanical point of view, carbon fiber is preferred. In the
same way, various resins can be used (polyester, epoxy, phenolic etc…). But, for pressure
vessels, due to its good mechanical properties and good stability, the most used is the epoxy
resin. Pre-impregnated fibers are commercially available, but expensive. Thus, mainly for
cost reasons, the fiber is impregnated just before the winding step.
INDEX çç Sommaire
To conclude, when compared to the storage of other gases, hydrogen requires special
attention for the choice of the steel for types I, II and III tanks & for the polymer choice for
type IV tanks. A material test is generally requested to prove that the embrittlement is low:
tensile test, disc test, fracture mechanism tests. A full permeation measurement is required on
one vessel to prove that the permeation is below a specified rate (1 cm3/l/h in 2005). A
measurement on a sample alone could also be sufficient to assess the permeation rate.
Detailed information concerning the sources of accidents involving gas cylinders are given in
Annex B.
Today industrial hydrogen is delivered as a function of the customer consumption. The supply
modes are presented in Table 3. In this paragraph, we will then only consider pressure
storages in vessels.
Compared to industrial gas, hydrogen energy has brought new constraints for pressure vessels
and this, mainly for the transportation field. Before entering into details for pressure vessels
dedicated to hydrogen energy, it is important to remind the main hydrogen energy
applications. They are listed below and are depending on the fuel cell development (PEMFC
mainly):
Fuel for transportation: buses, car, scooters, other leisure vehicles. These vehicles can
be powered by a fuel cell or by an internal combustion engine fuelled with H2. Use in
boats is also considered. For this application, he main constraint at the moment is the
weight and volume savings. As a consequence, when pressure storage is considered,
only types III and IV could the cost of the storage system is also important.
Stationary applications: back-up power supply or power generator for residential. For
this application, the cost of hydrogen supplied is the main parameter.
Portable applications: portable back-up power supply, portable power generators,
electronics (computers, mobile phones, etc…).
Weight and volume savings are primordial.
The performances which are generally used to compare the pressure vessels are Cm and Cv
defined as:
INDEX çç Sommaire
Cm: weight performance: mass of H2 stored divided by the mass of the vessel (%wt)
Cv: volume performance: mass of H2 stored divided by the external volume of the
vessel (g/l)
cost
The safety requirements are of course the same for all of them. Figure 7 shows the technical
performances as a function of service pressure for the today technology of type III and type
IV vessels. The performances are given with 10% of incertitude. Note that the Cm for type I
and type II vessels are respectively 1 and 1,5 % at 200 bar (the weigh is ~60-70kg to store 10
Nm3 of hydrogen).
The general features for type I vessels and that they are heavy (Cm ~1% max) and pressure
limited. Indeed, to have high service pressures of type I tanks, the thickness of the wall shall
be increased. But, heat treatment during the manufacturing cannot be as efficient as on thinner
walls: the very high pressure walls will have strong material properties heterogeneity through
the thickness. It generally reaches the point where steels are very sensible to hydrogen
embrittlement and thus defect growth unpredictable. This has resulted in restriction for 350
bar all metallic vessel use in Europe. As a consequence, for higher-pressure applications (700
bar refuelling pressure) type I technology, shall definitely be avoided to reach an appropriate
safety level. For on board storages in vehicles, both the weight and the pressure limit will
restrict their use. On the other hand, they are the cheapest high pressure vessels. Thus, when
no space or weight savings are required, they are the more competitive and technological
sufficient supply mode. It is the case for stationary applications like back-up power supply or
power generator.
Type II vessels are still heavy (no significant improvement compared to type I), but they can
easily withstand very high pressures. Due to their weight, they are not a solution for on board
hydrogen storage. But they have other uses in the hydrogen energy. For example, they are
used as high pressure buffers in hydrogen fast filling stations. 450 and 800 bar type II buffers
are used today for the demonstration and the deployment of H2 fuel stations. Because few
INDEX çç Sommaire
fibers are needed, they remain cost competitive for such applications. An example of fast
filling station is given in figure 9.
Figure 8 : Stationary fuel cell power supply equipped with conventional 200 bar type I vessels
(Axane technology)
Type III and type IV tanks are one of the most suitable solutions today for on-board storages.
These technologies are widely used for other gases (air, natural gas), but the main difference
for on-board hydrogen is the need of very high pressures: 350 to 700 bar for H2 instead of 200
bar for natural gas or 300 bar for breathable apparatus. This pressure increase meets
technological issues.
However, even at very high pressures, these storages do not fulfill the automotive industry
requirements in term of technical performances and cost. The Department of Energy in the
US (DoE) gives the requirements of Table 4 for the whole on-board storage system (that is
including valves, pressure regulators, pipes etc… from the vessel itself up to the fuel cell or
the thermal engine). It is clear that these requirements won’t be achieved with today
technology. For a car system and a 700 bar pressure storage, the weight performance at the
INDEX çç Sommaire
moment seems to be 5-6% at the vessel level (not system). From a pure technical point of
view, maybe a change in the fiber type could allow to reach these values (for example by
using a very high performance carbon fiber ; a prototype 700 bar vessel have reached 11%
weight performance by using very high cost materials in 2001). But the cost will highly
increase. One other solution could be to re-define the safety coefficient. At the moment for
carbon fiber composite vessels, it is equal to 2.25. The proposal would be to decrease it,
especially for 700 bar on board storage. This would allow both to save weight and volume
and to decrease the cost. Some initial studies have been undertaken on that topic both in North
America and in Europe.
With the usual design, 350 bar type III and type IV tanks seem to be commercially available
in 2005. They are used in FC or IC H2 powered buses for which 350 bar seems sufficient. For
cars, due to the low energetic density of hydrogen and specific car constraints – mainly room
and autonomy equivalent to gasoline cars – higher pressures are required. A consensus seems
to agree for the 700 bar (10 000 psi). Very high pressure composite tanks started to be
developed a few years ago and they are not fully available from a commercial point of view.
Testing and development are still on going ; it is expected that the increasing interest for
hydrogen energy, and the first large scale demo and commercial application will open the
market, allowing the manufacturers to increase their efforts for more reliable and cost
effective solutions in the coming years.
At the moment, In Europe, information concerning the high pressure composite vessels
developments can be found in the European integrated project: STORHY. In this project,
work in on going on 700 bar tank technical performances improvement, on increasing the
wrapping step in the production cycle and on very new vessel design to decrease cost. In
North America, In DoE programs, developments are also on going of vessel and
manufacturing cost improvements.
For small portable applications (computers etc…), no pressure storage is considered at the
moment.
Conclusion
The main features (state-of-the-art) of the 4 pressure vessels are summarized in Table 5.
Each pressure vessel type can be used in the hydrogen energy supply chain. This depends
mainly on the application type of the fuel cell. The main developments today concern type III
and type IV at very high pressure. Even with the today design, some improvement has to be
performed for the 700 bar pressure in order to fulfill the hydrogen energy standards. It is
important to underline that these standards are still under discussion.
INDEX çç Sommaire
Table 5 : Main features for H2 pressure vessel types in 2006
Type I Type II Type III Type IV
Technology mature : Technology mature Technology mature Technology mature
++ :+ for P≤350 bar; 700 for P≤350 bar; 700
Pressure limited to Pressure not limited bar under bar under
300 bar ( density : ( density : +) development. development.
) Cost performance : Cost performance : Cost performance :
Cost performance : ++ +
Weight performance : Weight Weight Weight
performance : 0 performance : + performance : +
INDEX çç Sommaire
ANNEX A - SAFE AND EFFICIENT GAS CYLINDERS PRODUCTION
CYLINDERS TYPES
For more than a century high pressure gas cylinders, like many other industrial products, have
been obtained from steel and later on from aluminium alloys.
About twenty years ago high pressure gas cylinders from new steel, aluminium alloys and
incorporating new fibres started to become available from traditional cylinder manufacturers
and from smaller manufacturers, who also started their activity in those days.
Prototypes for aerospace had been manufactured and used earlier. Nevertheless only in the
last few years concepts like the classification of the cylinders in "types" have become popular:
"Type 1" to indicate steel or aluminium alloy cylinders, "Type 2" to indicate a cylinder with a
metal liner (capable of withstanding a pressure higher than the working pressure of the
cylinder) with a hoop fibre wrapping, "Type 3" to identify cylinders with a metal liner and a
full fibre wrapping and "Type 4" for the so called fully composite cylinders (figure A1).
Such a classification was introduced for the first time in the ISO standard for CNG cylinders
for automotive use and these days a new "Type 5" category appears in some draft standards
for cylinders for compressed gas vehicles.
Advanced production methods for Type 1, Type 2 and Type 3 gas cylinders are described
here.
Production methods
Billet, tube and plate are the three production methods available today for the production of
Type 1 cylinders and for the production of the liners for Type 2 and Type 3 cylinders.
INDEX çç Sommaire
The different production methods are referred only to the first part of the whole production
process. Once the "shell" has been obtained from billet, or tube or plate it is processed in the
same way (figure A2).
shells are obtained from tube by spinning and shaping the base;
manufacturing from plate is mainly a cold process even though the annealing and
lubrication of the material are necessary at each step of the production (deep drawing)
(figure A3).
INDEX çç Sommaire
Advantages and disadvantages of production methods
Each production method has its advantages and disadvantages (see Table A1).
It seems that today and for the next few years, production from plate offers more advantages
when compared with cylinder production from billet and from tube. Especially if we take
advantage of the new improvements, already available, like for example the start of the
production process from steel coil instead of steel plate.
Production from tube offers the relevant advantage of a smaller capital investment.
Nevertheless there are only three or four industrial groups able to supply the seamless steel
tubes of the dimensions and of the quality needed for the production of gas cylinders.
The high tolerance on the wall thickness affects the weight of cylinders and liners
manufactured from tube, especially when the design wall thickness is thin.
The diameter of the cylinders manufactured from billet usually does not exceed 280 mm,
which is not enough for some applications. Also the capital investment is high, probably too
high for the present market value of the cylinders that can be manufactured using that method.
Cylinder production from plate requires a relatively high investment, but it is extremely
flexible and efficient, also in case of small production lots.
That is not the case for production from billet which needs large production lots of a few
thousand pieces to guarantee an acceptable level of efficiency.
INDEX çç Sommaire
NEW CYLINDERS
Today new cylinders are requested for industrial, medical and special gases, breathing, fire
fighting applications, natural gas vehicles and we all know the expectations about the use of
compressed hydrogen.
The new materials, high strength steel and high strength fibres, must be compatible with the
contained gas, environment and handling.
More and more the strategic know-how is in the production of shells with the highest
efficiency according to the design. Which practically means obtaining the lightest possible
cylinder or metallic liner according to the reference design standard?
The task of the manufacturers is to make those new cylinders available at an "equivalent"
safety level so that all users can afford to take advantage of their higher efficiency, thanks to
their lighter weight, without compromising safety.
Often those new standards are dedicated standards for different applications. And they now
make available cylinders whose design is based also on Finite Element Analysis and
Prototype Performance Tests.
The scope of those tests is to simulate the use conditions of the cylinders during their life.
Cycling tests simulate the filling cycles. LBB (leakage before burst) proves that the cylinder
leaks and does not burst in case of damage. Dynamic and environment tests are performed as
well.
European Directives about high pressure cylinders and their relative standards have defined
the rules that will govern a good part of the business in the next few years.
That is not the case for hydrogen. Standards for compressed hydrogen (land vehicle fuel
tanks) are still under discussion. Those who are optimistic, expect to have those standards for
the cylinders approved in two/three years time but the agreement inside the different
international technical committees, which are competent about the standards for compressed
hydrogen cylinders, has still to be reached.
The technical standards will certainly leave open to the final users the choice of the type of
cylinder they want to adopt for their applications. Nevertheless the acceptance criteria for the
performance tests set by the future standards will have a relevant influence on that choice.
Metal cylinders and liners offer the absolute guarantee against permeation, which is not the
case for type 4 cylinders, but the compatibility with hydrogen at the design mechanical
characteristics has to be proved. International standards are today being compiled for the
assessment of such a compatibility.
INDEX çç Sommaire
There are no doubts that a relevant part of the future production of high pressure gas cylinders
will be for compressed hydrogen for industrial purposes and for vehicles, although the
development of the "Hydrogen Economy" might be slower than expected.
Production of cylinders and liners from steel plate is very rare among cylinder manufacturers,
but the process itself is not new.
The novelty is having brought in-house the annealing, the flattening and the cutting of the
steel coils allowing a better control of the initial heat treatment of the raw material.
That initial annealing of the raw material is usually not available from the steel mills and is
carried out by service companies which are not always reliable.
It is a process which takes around seventy hours in a protective nitrogen atmosphere (figure
A4).
The correct choice of the steel alloy and the tuning of the initial annealing of the raw material
are essential in order to obtain light steel cylinders and light steel liners.
The production process from steel plate (now from steel coil) is long because it requires up to
five cold deep drawings, and for each of them a lubricating cycle and an annealing process
(figure A5).
INDEX çç Sommaire
Figure A5 : Production process from steel plate (coil)
Flow forming is used to reduce the wall thickness to the design shell thickness which, in case
of liners, can be very thin.
The neck spinning must result in a flawless dome in some cases with large necks starting from
a thin shell thickness. In the case of compressed hydrogen, for example, the request from the
motor companies is now for 2 inch necks threads and it might be that soon an even larger
thread is needed.
The quenching and tempering must achieve a uniform condition at the new high mechanical
characteristics of the steel set by the design.
The whole production process must be achievable for the cylinders and the liners of the
different dimensions and pressures that the market now requires (up 400 mm diameter, 200
litres water capacity and 800 bar working pressure).
Composite cylinders
Composite Type 2 and Type 3 cylinders are obtained by winding with E-glass fibre or with
carbon fibre the steel liner (figure A6).
Carbon fibres are available from very few suppliers in the world and the alliance or the
agreement with the fibre supplier is essential for the cylinders manufacturer to ensure that the
raw material remains available in the future. Type approvals of the cylinders are in fact valid
INDEX çç Sommaire
only for the specific fibre used to manufacture the prototypes. A change of the supplier would
mean a change of fibre type and new type approvals of the cylinders.
Fibres are impregnated with resin. Resin is cured by a thermal or UV process depending on
the type of fibre that is used.
The pressure applied with auto-frettage strains the liner past its yield point, sufficient to cause
a permanent plastic deformation. The result is the liner having compressive stresses and the
fibres having tensile stresses at zero internal pressure.
Cylinders efficiency
A cylinder’s efficiency is expressed in this paper in terms of weight per litre of water
capacity.
For a given standard and working pressure it depends on the diameter of the cylinders and on
the water capacity of the cylinders.
CONCLUSIONS
The gas cylinders available on the market and used by the gas industry have standard
diameters which come from the diameter of the seamless tubes used as raw material or from
the limitation of the forging process from billet.
Production from steel plate (coil) allows us to obtain truly seamless “efficient” cylinders with
a large range of diameters that can meet the specific need of users.
The graphs hereunder show the efficiency of the different types of cylinders with different
diameters.
The design of the cylinders is according to the most recent European norms available for each
type. The range of water capacity is the one available for production today.
INDEX çç Sommaire
Efficiency vs. Capacity
200 bar WP, 356 mm diameter
1,10
1,00
0,90
0,80
0,70
0,60
0,50
0,40
40 50 60 70 80 90 100 110 120 130 140 150 160 170
EN1964-1 Type 1 Steel Cylinders (from plate-coil)
EN1964-2 Type 1 Steel Cylinders (from plate-coil)
EN12257 Type 2 Steel/E-Glass Fibre Composite Cylinders
EN12257 Type 2 Steel/Carbon Fibre Composite Cylinders
EN12245 Type 3 Steel/Carbon Fibre Composite Cylinders
0,90
0,80
0,70
0,60
0,50
0,40
20 30 40 50 60 70
EN1964-1 Type 1 Steel Cylinders (from plate-coil)
EN1964-2 Type 1 Steel Cylinders (from plate-coil)
EN12257 Type 2 Steel/E-Glass Fibre Composite Cylinders
EN12257 Type 2 Steel/Carbon Fibre Composite Cylinders
EN12245 Type 3 Steel/Carbon Fibre Composite Cylinders
INDEX çç Sommaire
Efficiency vs. Capacity
200 bar WP, 203 mm diameter
1,20
1,10
1,00
0,90
0,80
0,70
0,60
0,50
0,40
0 10 20 30 40 50
EN1964-1 Type 1 Steel Cylinders (from plate-coil)
EN1964-2 Type 1 Steel Cylinders (from plate-coil)
EN12257 Type 2 Steel/E-Glass Fibre Composite Cylinders
EN12257 Type 2 Steel/Carbon Fibre Composite Cylinders
EN12245 Type 3 Steel/Carbon Fibre Composite Cylinders
1,10
1,00
0,90
0,80
0,70
0,60
0,50
0,40
4 5 6 7 8 9 10 11
EN1964-1 Type 1 Steel Cylinders (from plate-coil)
EN1964-2 Type 1 Steel Cylinders (from plate-coil)
EN12257 Type 2 Steel/E-Glass Fibre Composite Cylinders
EN12257 Type 2 Steel/Carbon Fibre Composite Cylinders
EN12245 Type 3 Steel/Carbon Fibre Composite Cylinders
INDEX çç Sommaire
ANNEX B – LEARNING FROM GAS CYLINDERS ACCIDENTS –
A GENERAL OVERVIEW
INTRODUCTION
Gas cylinders are in use for more than one hundred years. For Industrial Gas Companies, the
stock of these cylinders represents a major capital investment.
Seamless steel cylinders (the most commonly used cylinders for high pressure industrial
gases),
Seamless aluminium cylinders (increasing usage since more than 40 years),
Welded steel cylinders (mainly used for low pressure gases such as LPG and acetylene),
Composite cylinders (introduced more recently when the efficiency justifies the used of
more expansive cylinders ; examples of use of these cylinders for medical and hydrogen
energy applications will be given during this Symposium),
Large cylinders called “tubes” (they are mainly used for tube trailers),
And pressure drums (large welded cylinders).
In this paper, we will mainly cover seamless gas cylinders.
For steel cylinders, three basic processes for fabricating are used:
From billets (figure B1). The different forging steps are described in figure B2,
From tubes. This process consists of closing one of the extremities of the tube by
hammering (forging) or spinning,
From plates. A specific paper will be given during this symposium.
INDEX çç Sommaire
Figure B1 : Stock of steel bars for cylinders made from billets
For the three processes, the most widely method used to form the neck and shoulder of the
cylinder is spinning process (figure B3).
Aluminium cylinders are made from billet (figure B4). The most common process is the cold
extrusion (figure B5), but hot extrusion is also used (figure B6).
INDEX çç Sommaire
Figure B3 : Steel cylinders spinning process
INDEX çç Sommaire
Figure B6 : Aluminium cylinders made from hot extrusion
Nowadays, most of the steel cylinders are made of chrome-molybdenum quenched and
tempered steels. Aluminium cylinders are made of a number of alloys:
AA 6061
AA 7060
AA 7032
AA 2001
AA 5283
AA (6351)
The different types of composite cylinders are hoop wrapped and fully wrapped made of glass
fiber, amaride fiber and carbon fiber.
In this process, light and high strength fibers are wrapped onto a liner (see figure B7).
Finally, some types of applications of gas cylinders are illustrated in figures B8 and B9.
INDEX çç Sommaire
Figure B8 : Medical application
Gas cylinder incidents can result from manufacturing defects or from misuse in service.
Manufacturing defects
Manufacturing defects can be attributable to the original materials, the forging operation, the
heat treatment, or during final manufacturing operation.
INDEX çç Sommaire
Defects from the original materials
For steel, such defects can be the result of a poor “cleanliness” of the material (presence of
non-metallic inclusions such as sulfur or oxides) or a bad distribution of these inclusions
which have concentrated during the manufacturing process of the billet, plate or tube (figure
B10, an example of defect in the cylinder base due to segregations resulting from continuous
casting).
For aluminium alloys, it is now well known that this alloy AA 6353 which is susceptible to
Sustained Load Cracking (SLC) can lead to the formation of neck or shoulder cracking
(figure B11). Lead contamination in AA has also been a source of cylinders failures, the lead
content must be limited to 30 ppm.
Figure B10 : Defect in the cylinder base due to segregations resulting from continuous casting
FOLDS
Figure B11 : Neck and shoulder cracks due to Sustained Load Cracking
For billet piercing, the main defect could be draw marks (if the tool used for piercing is not
regularly changed). These draw marks can initiate fatigue cracks due to pressure cycling
(filling/emptying of the cylinders). Other defects can be thin wall thickness due to excessive
eccentricity or eventually excessively thin cylinder base (figure B12).
For cylinders manufactured from tube, the main defect is an improper closure of the base,
which could lead to leak in service (figure B13). It is also somewhat difficult to obtain a
progressive increase of the cylinder wall at the cylinder to base transition zone (makes the
cylinder more susceptible to failure by pressure cycling).
INDEX çç Sommaire
Finally, for the cylinders made from plate, transverse cracks may be formed if extremely
severe deformation are applied (figure B14).
As far as the shoulder shaping is concerned, the main defect to avoid is neck folds due to
improper preheating of necking operation. Pre-existing defects in the material may also be a
source of neck defects (folds, cracks).
For aluminium alloys cylinders, the shoulder shaping is a critical operation ; in addition to the
risks of neck fold, the operation which is performed at high temperature can generate a coarse
grain structure (figure B15) making this part of the cylinder prone to developing various
defects (e.g. risk of damaging the neck threads during valving operation).
INDEX çç Sommaire
Figure B14 : Cylinders made from plate – Transverse cracks due to extremely severe
deformation
Figure B15 : Aluminium cylinders with coarse grain structure in the neck/shoulder region
For steel, improper heat treatment may lead to brittleness of the material at low or even room
temperature.
For aluminium alloys, improper heat treatment may make some materials sensitive to inter-
crystalline stress corrosion cracking (figure B16).
INDEX çç Sommaire
Figure B16 : Aluminium cylinders – Materials sensitive to inter-crystalline corrosion
Some final operations such as local grinding, stamp marking, shot blasting may generate
surface defects which could be the source of failure in service (figure B17). Such defects must
be carefully evaluated especially because these operations are often done after non destructive
examination and batch tests.
In service defects
The in service defects can be due to overfilling (liquefied gases and C2H2), improper gas and
material compatibility, external or internal corrosion, external impact, arc burns, cylinders
involved in a fire or internal extermination of the cylinders.
INDEX çç Sommaire
Overfilling
This risk is particularly relevant for high pressure liquefied gases such as CO2 or N2O. To
avoid the occurrence of failures due to overfilling, the Industry recommended the use of
bursting disk for all CO2 and N2O cylinders [3].
If a gas not compatible with the cylinder material is used, this may result in the cylinder
failure. This is the case with hydrogen imbrittlement. Several hydrogen cylinders and tubes
failed in the 70’s because of the use of very high strength steel. To solve this issue, EIGA
developed TN 26/81 which has been recently republished [4].
The effect of other potentially damaging gases is treated in ISO 11114-1 [5].
Corrosion
It could either be External Corrosion when cylinders are used in certain environments
(seaside, ships…), EIGA republished recently a document on this issue [6] or
Internal Corrosion
In this later case, this is the result of water ingress for cylinders used with oxidizing gases
(e.g. O2) or mixtures thereof or used with acid gases (e.g. CO2) or mixtures thereof (figure
B18). To avoid the occurrence of such failure, a document has been published by EIGA [7].
In some cases, small traces of water can lead to stress corrosion cracking. This is the case for
CO/CO2 gas mixtures (figure B19). Several cylinder failures of this type were reported and
EIGA recently republished a document on this issue [8]. To avoid such failure, (because it is
in practice very difficult to avoid the presence of traces of water), it is best to reduce the
service pressure of the steel cylinders (which reduces the wall stress), or to use aluminium
cylinders which are not sensitive to this SCC phenomenon.
Aluminium is not always the right solution in case of water presence. Tap water by itself
(without gases) can initiate after few days inter-crystalline corrosion on the surface of AA
6061. (These small surface cracks reduce the cycle life of composite cylinders made of AA
6061 liners, see [9]).
INDEX çç Sommaire
External impacts
External impacts can make defects on the external surface. Cylinders shall not be exposed to
external impacts. This is particularly true for soft materials (e.g. AA) or brittle materials
(carbon fiber).
Fires
The question of full cylinder fire engulfment has been treated previously [10]. In Europe,
contrary to North America, the cylinders are not protected by Pressure Relief Devices, PRDs.
It is considered that the fire itself is the biggest hazard and that the cylinders may fail anyway.
Localized fires could be more detrimental in a way that it could lead to a drop of the local
mechanical properties, thinning of the wall or local swelling (figure B20). All these defects
may lead to subsequent cylinder failure. The temperature needed to weaken the cylinders wall
is much lower for aluminium than steel cylinders and consequently not easily detectable from
the external appearance. Steel cylinders are however subjected to “arc burns” (figure B21).
Cylinders presenting such defects shall be rejected.
INDEX çç Sommaire
Figure B21 : Initial cracking under arc burn
Finally, cylinder failure or burst can be the result of introduction of contamination in the
cylinders (such as oil and grease).
In Table B1, we have compared the situation as it was 20 years ago, compared to nowadays.
In the left side column, is given the list of cylinder incidents reported by Michel Marchal
during the EIGA symposium held in Paris in 1985 (only the incidents due to a cylinder defect
are reported and not the ones due to incidents resulting from improper cylinder handling).
In the right column, are listed the incidents reported to EIGA/WG2 during the years 2001-
2002. This 2-year period was chosen in order to have in both columns approximately the same
number of incidents, to determine which type is the most frequent.
Note: this table cannot be used to indicate any increase or decrease in the cylinder incident
frequency rate.
This is the case for the defects in the billets and in the material in general. This is due to the
fact that continuous casting is not used anymore, and that ultrasonic testing (UT) check of
both the original material (billet) and the finished cylinder is performed.
INDEX çç Sommaire
Table B1 : Comparison from 1985 to 2002
1985 2001 - 2002
TYPES Number Observations Number Observations
Manufac- Materials 10 6 OLD – 4 2 Neck and body cracks
turing H2 of AA 6351
General 3 11 6 leaks at the first fill
Heat 2 failures during
treatment - retest
Extrusion 3 bursts during or
after filling
Failure 6 Including 2 Softening of AA
due to bad grinding and 6061, during painting
treat- local and local
ment in heating
service
Corrosion 6 5 internal – 4 Internal corrosion
1 external
Overfil- 2 1 Drum
ling
TOTAL 27 20
Other 2 Explosive mixtures
types (CO/NO and H2/O2)
3 C2H2 cylinders –
Explosion in Russia
C2H2 cylinders batch
of sunken masses
This is also due to the fact that “cleaner” materials are used with low sulphur and
phosphorous content.
Failures due to the embrittlement have also disappeared. This is due to the application of the
EIGA recommendation for H2 cylinders.
This is the case of failure due to overfilling thanks to the EIGA recommendation to use
bursting disks for CO2 and N2O cylinders.
This is also the case for the failure of CO and CO/CO2 cylinders by stress corrosion cracking
(see [8]).
Progress is also expected in the near future in the avoidance of internal corrosion problems
thanks to the recommendation to use “residual pressure valves” (RPV).
INDEX çç Sommaire
But some progress is still to be made; there is still a too high number of cylinders leaking at
the first filling. EIGA must ensure that the manufacturers do not release cylinders not suitable
for use.
REFERENCES
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page42
RÉSUMÉ / ABSTRACT
64H
LA NORMALISATION AU SERVICE DE LA CONSTRUCTION
DES BOUTEILLES À GAZ
J. Loumeto
AFNOR
RÉSUMÉ
Mardi
Les normes sont devenues en un peu plus d’un demi siècle des outils de référence dans les relations
commerciales notamment dans le secteur industriel. Ces documents techniques créés par les experts
d’un secteur donné n’ont pas vocation à être d’utilisation obligatoire mais visent, par leur mode de
réalisation basé sur le consensus, à donner un cadre de travail communément acceptable pour le bénéfice
mutuel de toutes les parties prenantes d’une activité spécifiée.
La présentation qui sera faite vise à faire un rapide état des lieux du monde de la normalisation en
général et de celui plus spécifique des normes de construction dans le domaine des bouteilles à gaz qui
entrent dans le cadre des EQUIPEMENTS TRANSPORTABLES.
1. Généralités et enjeux
¢
Principe et définitions
¢
De l’intérêt de normaliser
¢
Un joint stratégique pour l’économie
2. Les référentiels normatifs
¢ Symboles et descriptions
¢
Une palette de choix documentaires
¢ Le monde normatif et les bouteilles à gaz
3. Acteurs de la normalisation
¢
Répartition des rôles
¢
Organisation normatif autour des « bouteilles à gaz »
¢
La commission française : AFNOR/E29E
¢ Normalisation internationale: ISO/TC 58 « Bouteilles à gaz »
¢
Normalisation Européenne: CEN/TC 2358 « Bouteilles à gaz transportable »
4. Les normes de constructions des bouteilles à gaz
¢
Bouteilles composites
¢ Bouteilles sans soudures
¢
Bouteilles soudées
¢
Autres types de normes « construction »
¢ Focus sur deux normes pour bouteilles en acier soudé
5. Conclusion
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page43
RÉSUMÉ / ABSTRACT
64H
ABSTRACT
The standards have become, after a little more than an half of century, a reference tools in the trade re-
lationships particularly in the industrial sector. These technical documents created by experts are not
intended to be compulsory but by the way they are created based on consensus, they are required to
give a commonly acceptable framework for the mutual benefit of all stakeholders of a specified sector.
In France, AFNOR is the main standards body which represents the interests of the French industry to
ISO and CEN. We provide and / or follow the completion of a large number of normative documents
thank to the highly collaborative work with ISO and CEN structures.
The presentation to be issued aims to make a quick overview on the world of standardization in general
and more specifically in the design standards in the field of gas cylinders which are part of TRANS-
PORTABLE EQUIPMENT.
The plan below summarize a presentation, subdivided in four parts, to be made on this subject;
1. Generals and issues
¢
Principle and definitions
¢ The importance of making standards
¢
A strategic joint for the economy
2. Normative documents
¢
Symboles et descriptions
¢ A large range of normative documents
¢
Standards and gas cylinders world
3. Stakeholders in standardization field
¢ Organization of standardization in the field of « Gas cylinders »
¢
French mirror committee: AFNOR/E29E
¢ International Standardization Organization: ISO/TC 58 « Bouteilles à gaz »
¢
Comité Européen de Normalisation: CEN/TC 23 « Bouteilles à gaz transportable »
4. Gas cylinders standards for design
¢
Composites cylinders
¢
Seamless cylinders
¢
Welded cylinders
¢
Other « design » standards
¢
Spotlight on two standards for welded steel cylinders
5. Conclusion
INDEX çç Sommaire
La normalisation au service de la
construction des bouteilles à gaz
1. Généralités et enjeux
5. Conclusion
2
1 - Généralités et enjeux
Principe et définitions
De l’intérêt de normaliser
3
Principes et définitions
4
De l’intérêt de normaliser
Rationaliser la
production
Transférer les
Clarifier les
technologies
transactions
nouvelles
Innovation
Échanges
Management
Développer Aider aux choix
des marchés Normalisation stratégiques
Transparence
Choisir Protéger les
des produits consommateurs
Marketing Î Gage de
Sortir des contrats les Intelligence
savoir-faire technique
éléments techniques économique
Symboles et descriptions
7
Symboles et descriptions
NF = Norme française
Très peu de normes encore utilisées
EN = Norme Européenne
Norme ratifiée mais non publiée
8
Une palette de choix documentaires
9
3 - Acteurs et structures de la
normalisation
10
Le monde normatif et les bouteilles à gaz
SAC
BSI
UNI 2%
AENOR 11%
Autres AFNOR DIN
9% 2%
15% 19% 18%
JISC
SIS 7%
3%
BSI ISO/TC 58
18%
« Bouteilles à gaz »
NEN DIN
7% 27%
11
Organisation normative autour des
« bouteilles à gaz »
CEN/TC
23
ISO/TC
58
E29E
12
Répartition des rôles
Chef de projet
Experts
Secrétariat des commissions
Contenu technique
Suivi des projets
Commentaires et positions
Reportings/communication
Validation technique
Support méthodologique
Validation des traductions
Organisation/support réunion
Animation des réunions
Obligation de consensus
Faciliter le consensus
Animation du réseau d’experts
Networking
Veille réglementaire/normative
Rechercher des alliances
Assurer l’équilibre financier
Financement
14
Normalisation internationale: ISO/TC 58
15
Normalisation internationale:
ISO/TC 58/SC 2
16
Normalisation internationale:
ISO/TC 58/SC 3
Groupe 34
«xxx »
17
Normalisation internationale:
ISO/TC 58/SC 4
18
Normalisation Européenne:
CEN/TC 23
19
4 - Les normes de constructions
des bouteilles à gaz
Bouteilles composites
Bouteilles soudées
20
Bouteilles composites
21
Bouteilles sans soudure
ISO 11120: Tubes en acier sans soudure rechargeables d'une contenance en eau de 150 l à 3000 l - Conception,
construction et essais
EN ISO 7866: "Bouteilles à gaz sans soudure en alliage d'aluminium destinées à être rechargées - Conception,
construction et essais
ISO 3500: "Bouteilles à CO2 en acier, sans soudure, pour installations fixes de lutte contre l'incendie à bord des
navires
EN ISO 9809-1: "Bouteilles à gaz rechargeables en acier sans soudure - Conception, construction et essais -
Partie 1 : bouteilles en acier trempé et revenu ayant une résistance à la traction inférieure à 1 100 MPa
EN ISO 9809-2: "Bouteilles à gaz rechargeables en acier sans soudure - Conception, construction et essais -
Partie 2 : bouteilles en acier trempé et revenu ayant une résistance à la traction supérieure ou égale à 1100 MPa
EN ISO 9809-3: "Bouteilles à gaz rechargeables en acier sans soudure - Conception, construction et essais -
Partie 3 : bouteilles en acier normalisé
EN ISO 9809-4: "Bouteilles à gaz rechargeables en acier sans soudure - Conception, construction et essais -
Partie 4: Bouteilles en acier inoxydable avec une valeur Rm inférieure à 1 100 MPa
EN 13293:2002: Spécifications pour la conception et la fabrication de bouteilles à gaz rechargeables et
transportables sans soudure en acier au carbone manganèse normalisé, de capacité en eau jusqu'à 0,5 litre pour
gaz comprimés, liquéfiés et dissous et jusqu'à 1 litre pour le dioxyde de carbone
EN 1975:1999: Spécifications pour la conception et la fabrication de bouteilles à gaz rechargeables et
transportables en aluminium et alliage d'aluminium sans soudure de capacité comprise entre 0,5 l et 150 l inclus
EN 12257:2002: Bouteilles sans soudure, frettées composites
EN 1964-3:2000: Spécifications pour la conception et la fabrication de bouteilles à gaz rechargeables et
transportables en acier sans soudure, d'une capacité en eau comprise entre 0,5 l et 150 l inclus - Partie 3 :
bouteilles en acier inoxydable sans soudure ayant une valeur Rm inférieure à 1100 MPa
22
Bouteilles soudées
EN 13322-1: Bouteilles à gaz transportables - Bouteilles à gaz rechargeables soudées en acier -
Conception et construction - Partie 1 : acier au carbone
EN 13322-2: Bouteilles à gaz rechargeables soudées en acier - Conception et construction - Partie 2 : acier
inoxydable
EN 1442: Équipements pour GPL et leurs accessoires - Bouteilles en acier soudé transportables et
rechargeables pour gaz de pétrole liquéfiés (GPL) - Conception et fabrication
EN 14638-1: Récipients soudés rechargeables d'une capacité inférieure ou égale à 150 litres - Partie 1 :
bouteilles en acier inoxydable austénitique soudées, conçues par des méthodes expérimentales
EN 14140: Équipements pour GPL et leurs accessoires - Bouteilles en acier soudé transportables et
rechargeables pour gaz de pétrole liquéfié (GPL) - Autres solutions en matière de conception et de
construction
ISO 11513: Bouteilles à gaz (en acier soudés) rechargeables destinées à des matériaux stockés et fournis dans
des gaz sous pression inférieure à la pression atmosphérique (sauf acétylène)
ISO 18172-1: Bouteilles soudées en acier inoxydable rechargeables - Partie 1: Pression d'épreuve de 60 bars et
inférieure
ISO 18172-2: Bouteilles soudées en acier inoxydable rechargeables - Partie 2: Pression d'épreuve supérieure à
60 bars
ISO 20703: Bouteilles rechargeables soudées en alliage d'aluminium - Conception, construction et essais
ISO 26444: Bouteilles à gaz rechargeables en alliage d'aluminium soudées avec une pression d'essai supérieur à
60 bars - Conception, construction et essais
ISO 4706-1: Bouteilles en acier soudées rechargeables - Partie 1: Pression d'essai de 60 bar et moins
ISO 21172-1: Fûts soudés de capacité inférieure ou égale à 3000 litres destinés au transport des gazes - Partie 1:
Capacité jusqu'à 1000 litres
EN 12862: Spécification pour la conception et la construction des bouteilles à gaz rechargeables transportables
soudées en alliage d'aluminium
EN 14208: Spécification pour les fûts soudés de capacité inférieure ou égale à 1000 litres destinés au transport
23
des gaz - Conception et fabrication
Focus sur deux normes de construction de
bouteilles en acier soudé
EN 14140: Équipements pour GPL et leurs EN 1442: Équipements pour GPL et leurs
accessoires - Bouteilles en acier soudé accessoires - Bouteilles en acier soudé
transportables et rechargeables pour gaz de transportables et rechargeables pour gaz de
pétrole liquéfié (GPL) - Autres solutions en pétrole liquéfiés (GPL) - Conception et
matière de conception et de construction fabrication
Méthodes relativement
neuves Æ peu de RETEX
Méthodes/ le calcul
Sur-épaisseur (-)
Méthodes/ l’expérimentation Essai d’homologation (+)
Flexibilité de fabrication (+)
Traçabilité des pptés mécaniques (-) « Patte » française:
animateur du GT Mr Gierling
24
Autres normes de construction
25
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page44
RÉSUMÉ / ABSTRACT
78H
INCIDENCE DE L’APPLICATION DE LA NORME EN 15085
POUR LES FABRICANTS DE PRODUITS CHAUDRONNÉS
G. Bourgeois
Mardi
RÉSUMÉ
La Directive sur l’Interopérabilité des Systèmes Ferroviaires demande, d’une manière indirecte, la mise
en application de la série de normes EN 15085. Ce package contient beaucoup d’informations, tant au
niveau de la conception, de la fabrication, de l’inspection, du contrôle mais également de la certification
du constructeur. Les fabricants de produits chaudronnés ne sont pas exclus de l’application de ces
normes. Cette courte présentation va essayer de résumer les dispositions normatives et réglementaires
applicables.
ABSTRACT
The Directive on the Interoperability of the Railroad Systems requests, in a indirect way, for the appli-
cation of the series of standards EN 15085. It package contains a lot of information, at the level of the
design, manufacturing, inspection, control but also certification of the manufacturer. The manufacturers
of vessels and tanks as well are not excluded from the application of these standards. This short pre-
sentation tries to summarize the mandatory and technical requests.
INDEX çç Sommaire
Incidence de l’application de la
norme EN 15085 pour les fabricants
de produits chaudronnés
Domaine d’application
EN 15085-1 Généralités
EN 15085-4 Production
Directives RISC
Spécifications
Technique ERA
d’interopérabilité
Normes
européennes CEN, CENELEC,ETSI
Révision du 23/01/2009
Niveau de Certification CL 1 à CL 4
Classe de Performance de CP A à CP D
Soudure
Classe de Contrôle de CT 1 à CT 4
Soudure
Organisation qui
CL 1 Certification exigée
CL 2 Certification exigée
CL 1
Les normes produits spécifiques doivent être prises en compte en priorité par rapport
aux exigences de la norme
VT QUI ? Documents
CT 1 100% Pers. Cert. EN 473 Requis
CP B B B(restrictions)
CP D D D
Evaluation
Personnel EN 473 niveau 2
Supervision
QS DMOS QMOS
CL 1 Exigé Exigé Exigé (sauf CP D)
CL 2 Exigé B ou C C
Un suppléant présent
Validité 3 ans
Vérification annuelle
Audit de suivi
Niveau de certification
Champ d’application
Plage de Certification
9Ea Guide des Bonnes Pratiques pour le contrôle par émission BAOUDOUR J.-L.
acoustique des équipements sous pression / Guideline for acoustic EXXONMOBIL
emission testing of pressure equipment CHERFAOUI M.
CETIM
Mardi
1Ea Contrôle d’un ESP par émission acoustique dans le cadre d’une VIAUD A.
requalification aménagée : Ne pas perdre de vue l’objectif ! / ANVIXED
Acoustic emission examination as a part of the requalification
process of a pressure equipment by acoustic emission
47Ea Surveillance en service et à distance des ESP par émission LENAIN J.-C. /
acoustique / On line and remote monitoring of pressure PROUST A.
equipment : Keep the purpose in mind ! EPA
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page47
17Ea Contrôle par émission acoustique de réservoirs pour GPL enterrés TSCHELIESNIG P. /
recouverts de béton / Acoustic Emission Testing (AT) of SCHAURITSCH G.
underground concrete convered LPG tanks TÜV AUSTRIA
SERVICES GmbH
87Ea Equipement sous pression sous talus: Contrôle par émission CATTY J.
acoustique selon la méthodologie Cetim / Buried pressure CHERFAOUI M.
vessels : Acoustic emission testing by Cetim methodology CETIM
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page48
RÉSUMÉ / ABSTRACT
9Ea
GUIDE DES BONNES PRATIQUES POUR LE CONTRÔLE PAR ÉMISSION
ACOUSTIQUE DES ÉQUIPEMENTS SOUS PRESSION
RÉSUMÉ
Mardi
Ce guide élaboré au sein de l’AFIAP permet de faciliter la pratique de l’émission acoustique, notamment
lors de requalifications d’équipements spécifiques, en permettant le remplacement de certaines épreuves
hydrauliques par des essais en pression hydrostatique ou pneumatique surveillés par émission
acoustique. Le retour d’expérience et l’introduction de nouvelles annexes ont permis la réalisation
d’une seconde édition 2009 du guide composée d’un corps de texte et d’annexes. Chacune d’entre elles
est applicable à une famille d’équipements (sphères, petits vracs enterrés, ESP cylindrique, réacteurs
et autoclaves).
Ce document fait l’objet de la décision BSEI n° 09-102 du 29 juin 2009, émise par le Ministère de
l’écologie, de l’énergie, … (MEEDDAT), relative au remplacement par l’épreuve hydraulique, lors de
la requalification périodique de certains équipements sous pression, par un essai sous pression gaz
contrôlé par émission acoustique.
Cette présentation permet de montrer d’une part, les nouveautés de cette nouvelle édition et d’autre
part, de faire le bilan des retours d’expériences sur les 3 dernières années.
ABSTRACT
This guide elaborated within the AFIAP, allows to facilitate the practice of the acoustic emission, in
particular during requalifications of specific equipments, by allowing the replacement of certain hy-
draulic tests by pressure tests in hydrostatic or pneumatic pressure by acoustic emission monitoring.
The experience feedback and the introduction of new appendices allowed the realization of the second
edition 2009 of the guide consisted of one text and of appendices. Each of them is applicable to a family
of equipments (spheres, small buried LPG tanks, cylindrical vessels, reactors and autoclaves).
This document is the object of the decision BSEI N 09-102 of June 29th, 2009, emitted by the Ministry
of Ecology, Energy, (MEEDDAT), relative to the replacement by the hydraulic test, during the periodic
requalification of certain pressure equipments, by the test under pressure gas monitored by acoustic
emission.
This presentation allows to show on one hand, the news of this edition 2009 and on the other hand, to
assess experience feedback over the last 3 years.
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page49
RÉSUMÉ / ABSTRACT
1Ea
CONTRÔLE D’UN ESP PAR ÉMISSION ACOUSTIQUE
DANS LE CADRE D’UNE REQUALIFICATION AMÉNAGÉE :
NE PAS PERDRE DE VUE L’OBJECTIF !
A. Viaud - ANVIXED
RÉSUMÉ
La décision BSEI n° 07-107 du 13 avril 2007 prévoit que la requalification aménagée d’un ESP, avec
mise en charge pneumatique surveillée par émission acoustique, soit effectuée en application des
dispositions du « Guide AFIAP des Bonnes Pratiques pour le contrôle par émission acoustique des
équipements sous pression ». Si ce texte balise clairement la démarche, la pratique depuis 2004 a montré
que, tant pour la réalisation du contrôle que pour la rédaction du rapport, certains aspects restent mal
compris ou mal appliqués par rapport à l’objectif fondamental : une contribution au diagnostic de
l’équipement, compréhensible et exploitable par l’Expert de l’Organisme Habilité en charge de la
requalification. La présentation reprendra les étapes du processus en mettant en évidence les règles à
appliquer et les erreurs à éviter.
INDEX çç Sommaire
<3
#(&#%(' #%%' (( &%'$(&9:# (% '%!
#% && (&'$( & $(#'& & (& #%&& 1 (& &# & & '' 4(
%' (%4+#%$('& ##' &%&%$(' &&&
%#'&& (&34&'#&('%##%$(4 %* (('%(%%'
(' (&&'(%&4&& (&'$(1&&* %# (% '#%#'%'%
&&%$(' &%'%&3 &%#'1&'##%($( (''
#%% ' (' & # %' & '' ##' )1 $( & "(*% 4(
#%(*,%($(# &&('&1%%4(& (' '%'*#&&#%(
&&#('$(&(%*#%&& (&'$(9:3& %&1 '%!&'
#(& ( &' ' $(4 #(' 4''% #%'$( 4( && &'%('1 &
*' #%' #%' ( #% &&(& %$(' 3 #' & $('% * '&
'(+$(& '%&%&1#% (% '%!14&&(8'
%## %' '%!3
(& (& #% # & & % ( ' (% 4 %- ' &
#%'$(&% '%&3
INDEX çç Sommaire
E;
./ % (0&'/ &('/ ,0! +-!. . +(!'/ !&+(-/'/. //'0. % +-(0-9 0
('/-)% / 0 -++(-/ !'%; !/('. /!/- <3&+% % +-..!(' %<..!9 - . 2%0-
('&'/% +(0- % +-..!(' -,0%!!/!('9 ' +0/ /- (+/ +- % +-.//!-
.0%: %% !&+%!,0 0 +-&!- %<
-'!.& !%!/ B
C;
!- . -. 2 0..!
+-!.- %. !.+(.!/!('. ,0 2 +-'- %<3+%(!/'/ +(0- +-+-- %<++-!% +-(02- /
&//- ' +% %<(0/!%% &!. ' - +'0&/!,0;
. &(%!/. ('-''/
%<..0-' 0%!/ / % %' -2'/!(' /-(02'/ '/0-%%&'/ %0- +% '. % !-
. -.; %+0/0..!!'!,0-%.('!/!('..0-2!%%'(0.0+-2!.!('0('/-)%+-
0'(-'3/-';
'-.0&9% !-. -.+-&/<2(!-0'%'(&&0'03/0-.0+-($/
-,0%!!/!(': 3+%(!/'/ / +-.//!- !' .1-9 &!. 0..!
-'!.& !%!/ /
&!'!./-/!(';
+-(0- ('/-)% ,0! !/ 0..! +-/! 0 (..!- &' <&'&'/
!'/-(0!-0+-.%<&!'!./-/!('B
C9(!/('/'!-/(0.%.%&'/.'..!-.%
-%!./!(' 3 0./!2 0 ('/-)% / % -/!(' 0 -++(-/; /!/-9 %% +/ 0 .
+-/!0%!-/-!/-%.3!'.0 !-. -.0('/-)%9,0!./.('(0&'/=
&('/>;'('.,0'9(''(!/+./-(02-'.%+-(0-%.(-&0%/!('.<0'(
B9(0,0!$(0-)%C;
<3&+%/4+!,0./%<!'./-0&'//!(':%+-(0-(!/A
('/-!-&'/,0!.2(!/+-(!.@!'!,0-+-!.&'/,0%./%&/-!%,0!.-&!.
'*02-+(0-%/./;<./%-)%%<'/!/.0-2!%%'B.<!%4'0'C2-!!-,0%.
&(4'.0/!%!...('/('(-&.033!'.0((00;
'-/!'%(0++-"/-0%!-&'/.<!..'/.(0&'/.--':+--+
!/0'%!/'!(0&'/.,0!'.<++%!,0'/+.(0,0!+-.-!2'/.!.+(.!/!('.,0!'
.('/+.(&+/!%.9(0,0! '('-''/,0-/!'..+/ 0 ('/-)%9(0,0!.('/.
--'.-%&'/!-.; %0/'(-&%&'/!./!'0-'!/G/(-!.(0&'/.:
%.(0&'/.--'0.'.%!%!(-+ !9%(0&'/&('/B'+-!'!+0'
.0%:% !-. -.8C('/(0%%+-(0-0.'.%0%!//%.(0&'/.
/-2!%,0!2('//-0/!%!..+(0-%('/-)%;
&.0-%+-..!('&('/-0..!.('0.!('.:%0!3!0'&.0-<!'-/!/0
&3!&% F L;
<0/!%!./!(' <0' "' &.0- % +-..!(' F L '<./ (' +.
INDEX çç Sommaire
=4
)+)#*2 ) $# #!$ *$+*) !) +)) &$))!) 5#(**+4 &($+( $* +))
#'+('+!5#)*(+"#**$#+*!))*$#$("$#*$##!!"#**"*($!$'+"#*2
#) ! ( ! "*() ) "$.#) ")+( * 5)) &( ! &()**(4 # *$+*
(++(2!)(#()6*!$##)7$**("#*$##4
($#$!$+"$$&(*$(*),!*$##5&&(*&)*$+ $+()!("#*4
&$#* )* &$+(*#* ))#*! &$+( !5#)&*+( !5
(#)" !* '+2 #5*#* # #(!
&))&!)*!5"))$#$+)*'+2$*##"$#)*(#")+()+,()+(!*((#
!)*&)+&($))+)$#*(%!4
5$ *)*(!)(+#$#*(%!#*$+)&$#*)$#$("+-&()(&*$#)!&($+(
$#*(%!2")+))+#$#*(%!&$(*(!"#*('+("&!!5&(+,.(+!'+4
* (2 ! &()# !5#)&*+( !5
)* #)&#)!2 !5)) )$+) &())$#
&#+"*'+$#+**#*+#,(*!))())*#:*#$#+#$#*(%!5*#*;
'+))+)**+!5&(+,!))'+#))$#*$#)##)4
&(*'+"$#*('+)(*)&+,#*)&($+(4$#!)((&$)&($))
&($!")&('+!)#*(,##*)"#'+#*)$+,#*+!*+(6745
$+!5#**
)+(,!!# #*(,#* )* &$+( ,!+( !5"&$(*# !5(* :(") # +)
,#*+!! + $#*(%!; $+ )5! )5* 5+# (* "#+(2 -"#( ! &($))+) *(*"#*
&($&$)&(!&()**(4!)5*5)*"(!5"&*)+(!)()+!**)+$#*(%!*"$#*((
$""#*!#)**#+$"&*#)!5#!.))$##)4
5'+& 6"))$# $+)*'+7 $* *( # ")+( ( +# +* #*(
#$&#*)&$)(#&(*+!(*$+)!)$+"#*)*(,!+$##(,)$#4)
#)*(+"#*)$,#*&$(*(!),#**)65*!$##7,!*!"*5+*!)*$#4
(&&$(*$*$(*($)#"#*)$(*)3*(!(!*-*+$#*(%!(!)2$##(
!) *&) + &($))+) *(*"#* * 5-&!$**$# ) $##) '+ ,$#* $#+(
!5*!))"#* ! 6)#*+( $+)*'+7 !5
*)* * ## &()#*( !) ()+!**)
5+#"#(!(*$($##2#&("**(+!*+(#$#)&!)*$"&(#(
*5-&!$*(!(&&$(*4!+*$"&(#(!5-&())$#6)#*+($+)*'+7#)+#)#)
!(3 *$+) !) !"#*) '+ $#$+(#* '+!( * 6$(7 ! &$&+!*$# )
INDEX çç Sommaire
E;
INDEX çç Sommaire
;2
$( %'!& )(%& #% '%& 8%% 7' #% * #9 #% * # ! &"!' #&
!$(&2
&& !'&"!&'"%&((!"'#&'%(!&& !'4#%#(51"!
&3%'%%%(!&& !'#(&&)%#"(%!#%!%((!%&$(1"!)(!&&
"!(&!' 3 #"%'!'&
)% "('0 #% $( 3)!0 #"#('"! &
)! !'&('&!3)'#&'*'%'&(& !'2
"(% '% !% &(% #%" '$( ( %##"%'0 (%7'3 '% #"&% (! &'%('(%
#%&%##"%'0$(!!&&'"(&&"!'!(&/
3&& #%&&"! ) &&"! "(&'$( "!&''( 4# '%&&5 ( &#"&'
%$('"! "!' &#"& 3!&#'(% 3%!& ' #"(% & #%"!"!% ! '"('
"!!&&!(&2 3 &&"!"(&'$(%&'(! '"*!'0$( !(!
*#%'&#"!'(2(&$(#"(% &('%& '"&
0! &&"!"(&'$(4%
& #03&'" #$(52(&&0"!)!'$(''*#%'&%&'&%)3*#"'!''
&(%'"(' 30 $( "' #"()"% " #%!% #"(% "!% ) #%'!! &"! ( !'
%$('"!2
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page50
RÉSUMÉ / ABSTRACT
47Ea
SURVEILLANCE EN SERVICE ET À DISTANCE DES ESP PAR ÉMISSION
ACOUSTIQUE
RÉSUMÉ
Mardi
ABSTRACT
Because of the development of the instrumentation and data analysis software, the long experience of
applications for in-service testing of pressure equipments (PE) in various industries, now Acoustic
Emission (AE) technique is mature to be used as an on-line system to monitor the mechanical integrity
of PE while in operations. The solution provides data locally for use by operations personnel as well
as remotely via internet for detailed analysis by AE experts. AE on-line asset monitoring provides plant
personnel the ability to detect anomalies at an early stage, the growth of a defect and warning. This is
made possible by correlating the AE activity with other process variables. Several examples are des-
cribed in this paper (reactors, PSA, ammonia converters, exchangers, columns ...) showing the benefits
of this approach to enhance process safety, prevent catastrophic failures, reduce maintenance and
repair costs, reduce number of inspections, prevent operation interruptions, improve risk management,...
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page51
RÉSUMÉ / ABSTRACT
17Ea
CONTRÔLE PAR ÉMISSION ACOUSTIQUE DE RÉSERVOIRS GPL
ENTERRÉS RECOUVERTS DE BÉTON
P. Tscheliesnig, G. Schauritsch
(TÜV AUSTRIA SERVICES GMBH - Deutschstrasse 10 - 1230 Vienna, Austria –
e-mail: [email protected])
RÉSUMÉ
Le contrôle par émission acoustique des réservoirs de GPL enterrés, réalisé conformément à la norme
prEN 12817, sera utilisé par de plus en plus de pays européens comme une méthode pertinente et
économique de contrôle périodique. L’évaluation et la classification sont principalement basées sur de
nouvelles normes européennes et des données quantitatives qui sont précisées dans la procédure de
contrôle et résultent de l’expérience de l’organisation en charge du contrôle. Les paramètres de base sont
principalement l’intensité et l’activité en conjonction avec une combinaison plus complexe de différents
paramètres d’émission acoustique (EA).
Pendant le contrôle des réservoirs recouverts de béton, les craquements du béton génèrent une énorme
quantité de signaux EA, rendant impossible une évaluation pertinente de l’état de l’équipement métallique
sous pression.
Pendant le programme de recherche engagé en France avec le support du CFBP, nous avons commencé
à tester des réservoirs de GPL présentant des défauts, recouverts ou non de béton, et nous avons pu
développer un système de reconnaissance du domaine de fréquence avec lequel nous avons pu distinguer
les signaux EA émis par les défauts de l’enveloppe métallique de ceux émis par le revêtement du béton.
En appliquant cette méthode de reconnaissance du domaine de fréquence, nous sommes maintenant à
même de réaliser par EA une évaluation sûre et économique de l’état du réservoir de GPL lors de son
contrôle périodique. L’application de cette méthodologie permet de préserver la compétitivité de
l’industrie.
ABSTRACT
AT of underground LPG tanks according prEN 12817 will be used in more and more European countries
as a meaningful and economic method for the periodic inspection. The evaluation, assessment and
classification are based mainly on new European standard with quantitative numbers given in the test
instruction based on the experience of the testing organisation. The basis is mainly the intensity and the
activity in conjunction with a more complex combination of different AE parameters.
During the testing of concrete covered tanks an enormous mass of AE signals occur due to cracking in the
concrete, which make a meaningful evaluation of the condition of metallic pressure equipment impossible.
During a research project performed in France with the support of the CFBP we started to test plain and
concrete covered defective LPG tanks and were able to develop a frequency domain pattern recognition
system with which we have been able to distinguish AE signals coming from the defects in the metallic wall
and the concrete coverage.
With the application of this frequency domain pattern recognition method we are now able to perform a
safe and economic assessment of the condition of the LPG tank during the periodic inspection with AE. By
the application of this methodology the competiveness of the industry will be ensured.
AFIAP - ESOPE 2010 Programme
INDEX çç Sommaire
ACOUSTIC EMISSION TESTING (AT) OF UNDERGROUND
CONCRETE COVERED LPG TANKS
ABSTRACT
AT of underground LPG tanks according prEN 12817 will be used in more and more
European countries as a meaningful and economic method for the periodic inspection. The
evaluation, assessment and classification are based mainly on new European standard with
quantitative numbers given in the test instruction based on the experience of the testing
organisation. The basis is mainly the intensity and the activity in conjunction with a more
complex combination of different AE parameters.
During the testing of concrete covered tanks an enormous mass of AE signals occur due to
cracking in the concrete, which make a meaningful evaluation of the condition of metallic
pressure equipment impossible.
During a research project performed in France with the support of the CFBP we started to test
plain and concrete covered defective LPG tanks and were able to develop a frequency domain
pattern recognition system with which we have been able to distinguish AE signals coming
from the defects in the metallic wall and the concrete coverage.
With the application of this frequency domain pattern recognition method we are now able to
perform a safe and economic assessment of the condition of the LPG tank during the periodic
inspection with AE. By the application of this methodology the competiveness of the industry
will be ensured.
RESUME
Le contrôle par émission acoustique des réservoirs de GPL enterrés, réalisé conformément à
la norme prEN 12817, sera utilisé par de plus en plus de pays européens comme une méthode
pertinente et économique de contrôle périodique. L’évaluation et la classification sont
principalement basées sur de nouvelles normes européennes et des données quantitatives qui
sont précisées dans la procédure de contrôle et résultent de l’expérience de l’organisation en
charge du contrôle. Les paramètres de base sont principalement l’intensité et l’activité en
conjonction avec une combinaison plus complexe de différents paramètres d’émission
acoustique (EA).
Pendant le contrôle des réservoirs recouverts de béton, les craquements du béton génèrent
une énorme quantité de signaux EA, rendant impossible une évaluation pertinente de l’état de
l’équipement métallique sous pression.
Pendant le programme de recherche engagé en France avec le support du CFBP, nous avons
commencé à tester des réservoirs de GPL présentant des défauts, recouverts ou non de béton,
INDEX çç Sommaire
et nous avons pu développer un système de reconnaissance du domaine de fréquence avec
lequel nous avons pu distinguer les signaux EA émis par les défauts de l’enveloppe métallique
de ceux émis par le revêtement du béton.
En appliquant cette méthode de reconnaissance du domaine de fréquence, nous sommes
maintenant à même de réaliser par EA une évaluation sûre et économique de l’état du
réservoir de GPL lors de son contrôle périodique. L’application de cette méthodologie
permet de préserver la compétitivité de l’industrie.
INTRODUCTION
The most European countries apply for the requalification of underground LPG tanks the EN
12817. One testing possible testing method (Group 1) will be Acoustic Emission (AE), which
will be applied from several countries. The TÜV AUTRIA SERVICES has been applying the
AE testing, for LPG tanks, as a replacement of the hydraulic pressure test and the inside
inspection, since 1992 and have been tested since this time more than 40.000 tanks with this
methodology all over Europe (1,2). From the beginning the methodology has been changed
several times and the rapid development of the data processing opened us a new approach for
the testing of this vessels. At the beginning the tests have to be evaluated separately and “by
hand”. Due the enormous mass of date we were able to develop a so-called “Cluster
Evaluation Factor (CEF)”, in which all relevant AE-parameters were combined in this way,
that we have been able to define the status (integrity) of a LPG tank with a number. We were
able to elaborate with other countries the new Annex C “Acoustic Emission Testing) which
specifies the minimum requirements for the test equipment, thy type and sequence of loading,
data acquisition and data evaluation, for the performance of an AT on LPG tanks up to a
volume of 13 m3. Nevertheless the methodology (AT) can be applied also for LPG tanks over
a volume of 13m3, which is mentioned and declared in the EN 12819.
Based upon the regulations of EN 13554 and beside other requirements the prEN 12817
requires the following evaluation criteria. In any case the ∆t – value takes into account that a
tank never collapse in total and that always one or several defects are the reasons for failing
(3,4). Based upon a proper application of the sensors, which take into account the wave
attenuation for the maximum propagation distance between source and sensor, the following
evaluation criteria shall be based upon the following parts:
INDEX çç Sommaire
5) For LPG vessels containing juggle joints a more complex evaluation criterion is required.
This should be based upon evaluation criteria, combined to produce an evaluation factor
C, which takes into account at minimum
number of the AE bursts within a ∆t – cluster
peak amplitude and/or energy of the AE bursts within a ∆t – cluster
burst activity within a ∆t – cluster over the test period and/or shorter time and/or
pressure periods sequentially distributed over the test time.
In paragraph C.6.2 of the standard it’s mentioned, that the stop criteria during the real time
control can be reduced to number 1) and 5), if the calculation of the evaluation factor C is
done automatically in real time.
The idea behind this factor C, which in the testing procedure of the TÜV AUSTRIA is
mentioned as “Cluster Evaluation Factor (CEF)”, has been to combine the different AE-
parameters, see number 5) of the evaluation criteria plus some other parameters in such
manner, that on one hand side the combination minimise the influence of the distance
between the AE source and the point of detection (sensor) and on the other hand side that
critical situations (e.g. instable crack growing) will be identified on time. Of course such a
factor needs a theoretical works and an extensive data base. Finally the cluster with the
highest CEF value, which is the most dangerous for the integrity of the structure, will be used
for the tank grading.
AE-data-record
AE-parameter
amplitude
hit-rate
energy
further
.............
Evaluation - matrix
computer based calculation (on-line)
1 n= z
Pa
Pa
Pa
Pa
.... CEF = • ∑ f Pn
ra
ra 2
ra 3
ra n
1
Level 1 Evaluated z n =1
Level 2
Level ..
Database
Level n (Prototyp)
....
f1
f2
f3
fn
CEF - Calculation
Alarm - Level 1
p/t
The TÜV AUSTRIA SERVICES developed a sub-programme, which offers the testing
personnel based upon the calculation on the AE-equipment (Vallen Systeme), an automatic
acceptance or rejection of the LPG tank.
INDEX çç Sommaire
APPLICATION OF AT ON LPG TANKS
With the application of prEN 12817 and the TÜV AUSTRIA technology we and our licenses
tested more than 40.000 small LPG tanks (above- und underground) in Europe successful.
The first test-trials of AT of concrete covered LPG tanks showed, that such tanks are very
active, because of occurrence of many cracks within the concrete. This “very noisy” condition
prevents any meaningful test, because it is impossible to distinguish the AE coming from the
metallic tank and concrete. The reason for the activities is caused from the elastic deformation
of the steel shell during the pressurisation, which leads besides the cracking from the concrete
coverage to a high number of friction activities between the surface of the steel and the
concrete.
In year 2006 the TÜV AUSTRIA SERVICES started together with the “Comite Francais du
Butane et du Propane (CFBP)” a R&D project to develop an AT methodology for the testing
of concrete covered LPG tanks. The solution should be the separation of the AE signals
caused by real defects and disturb noise with a frequency domain pattern recognition system
“Visual class TM”.
After a comparison of the wave propagation in normal and concrete covered tanks, specimen
with well defined conditions were produced. These test specimens include defect free and
defective, blank and concrete covered tanks. With these tanks we were able to gather
databases for different situation and create classifiers for STELL CRACKs and CONCRETE.
CREATION OF A CLASSIFIER
For the creation of a classifier the specific data for the different situation have to be acquired.
For this work step the different tanks have had to undergo a pressurisation with gas and the
data were acquired. The time signals were checked according their frequency content by a
Fast-Fourier-Transform (FFT) and normalised. The normalised frequency spectra were
transferred into specific features and these features creates an “n-dimensional” space, where
the different situations can be divided. For this work step the data were distributed into
training and validation data. With the training data a supervised learning mode will start,
where at the end a classifier will be produced, which represents the specific situation in a
statistical manner. This classifier can only be applied on situations, which were trained
properly in advance. The remaining validation data were used to confirm the applicability and
also quality of the classifier (6,7). The validation showed a right assignment of the data over
90%, which is rather good for statistical system with this harsh environment.
For the validation of the technology tests without and with defect were arranged and the
automatically evaluation (CEF) were applied on raw and filtered test data. The filtering was
done on-line with the classifier where all CONCRETE data were eliminated.
The first example a tank (2 to) with artificial defect inserted at the 5 o-clock position in the
longitudinal weld was tested with 80 % water and pressurised with nitrogen. The burst test
was supervised with AT (2 sensors, clustering, classification by Visual class, automatically
CEF-calculation). For the unfiltered raw data any discrimination between activities was
impossible and according the large number of AE signal any evaluation was impossible.
INDEX çç Sommaire
Fig.: 2 time correlated events and CEF-value of these events via pressurisation of raw data
When only the filtered data were evaluated and fed into the CEF calculation system the
evaluation of the existing imperfection (Crack) was possible. The pre-warning period before
the burst at 1,76 MPa was sufficient. Under test conditions the test would be stopped at 1,55
MPa.
Fig.: 3 time correlated events and CEF-value of these events via pressurisation of
filtered data (removal of all data: class CONCRETE)
An example for good tanks showed, that the AT can be applied also on LPG tanks in the field.
By the application of the methodology we were able perform the test properly.
INDEX çç Sommaire
with classifier (filtered Class 2) with classifier (filtered Class 2)
Fig.: 4 comparison of the results (detected clusters and resulting CEF) for a good concrete
covered Tank without and with filtering
CONCLUSION
The AT of LPG tanks in Europe become more and more the common methodology. The
methodology is well standardised and by application of well established procedures and
trained personnel the advantage for the LPG industry is meaningful from the economical side
as well as from the safety point of view.
At the beginning of the described project the AT of concrete covered LPG tanks were
impossible. The paper describes how and with which tools it was possible to overcome the
problems. Since the finalisation of the project in 2007 approximately 1250 LPG tanks were
tested till the end 2009.
The TÜV AUSTRIA SERVICES and the Comite Francais du Butane et du Propane (CFBP)
own together the methodology. It is clear, that this methodology can be applied also on other
type of vessel e.g. spherical gas storage tanks, which are covered for safety and for fire-
protection reason with concrete.
REFERENCES
(1) Tscheliesnig P., Schauritsch G., Krenn G., Kragulj W. (1991) „Die wiederkehrende
Untersuchung von Flüssiggasdruckbehälter durch Schallemissionsprüfung“, DACH –
Jahrestagung pp. 44; Luzern, Switzerland
(2) Schauritsch G., Tscheliesnig P. (2000) “Die Schallemissionsprüfung als integrales
Prüfverfahren für die Beurteilung von Flüssiggaslagerbehältern – Erfahrungsbericht”,
DACH-Jahrestagung pp. 243-252; Innsbruck, Austria
(3) EN 13554:2004 „Non-destructive testing – Acoustic Emission – General principles“
(4) prEN 12817:2008 Annex C „LPG equipment and accessories – Inspection and
requalification of LPG tanks up to and including 13 m3 overground/underground“
INDEX çç Sommaire
(5) P. Tscheliesnig, G. Schauritsch (2007) „Applikation eines automatischen AT-
Auswertesystems bei der Prüfung schwer zugänglicher Strukturen“ DGZfP -
Jahrestagung; Fürth, Germany
(6) G. Schauritsch (2007) „Schallemissionsprüfung an betonummantelten, erdvergrabenen
Flüssiggaslagerbehälter – Wo herkömmliche Auswerteverfahren versagen“ DGZfP
Schallemissionsseminar, Puchberg, Austria
(7) P. Tscheliesnig (2008) “Periodic inspection of LPG vessels within Europe” EWGAE
conference; Krakow, Poland
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page52
RÉSUMÉ / ABSTRACT
45Ea
SUIVI EN SERVICE DES ÉQUIPEMENTS SOUS PRESSION A
COUVERCLE AMOVIBLE UTILISES DANS L’INDUSTRIE
AERONAUTIQUE ET SPATIALE
J.-C. Lenain (EPA), Y. Bouchaud (Airbus), L. Eas (RPS), D. Cloiseau et J.-L. Blossier (Socotec)
Mardi
RÉSUMÉ
L’Emission Acoustique (EA) est utilisée depuis de nombreuses années pour la requalification des ESP
en service. Pour les autoclaves utilisés dans l’industrie aéronautique et spatiale, une procédure particu-
lière a été développée à partir des expériences acquises dès 1981 et d’une analyse de risques. Depuis
1999, cette procédure est acceptée, pour une période expérimentale de 10 ans, par le BSEI en rempla-
cement des ré épreuves hydrauliques et des contrôles associés aux visites intérieures périodiques. A
partir du retour d’expérience (REX) de 140 examens réalisés sur 60 autoclaves différents, cette procé-
dure a été approuvée en 2009, a fait l’objet d’une nouvelle annexe (9) dans le guide de bonne pratique
de l’AFIAP et inclus dans cahier technique professionnel (CTP) du GIFAS pour le « suivi en service
des équipements sous pression a couvercle amovible utilisés dans l’industrie aéronautique et spatiale
». Cette communication décrit la méthodologie appliquée et le REX avec de nombreux exemples.
ABSTRACT
Acoustic Emission (AE) is widely used since several years for the in-service requalification of Pressure
Equipments. For the autoclaves used in the aircraft and aerospace industry, a special procedure has
been developed from the experience since 1981 and a risk based analysis. Since 1999, this procedure
is accepted by the French Authorities as an alternative of the hydraulic test and of the periodic internal
inspection for a 10 years experimental period. Based on the Return Of Experience (ROE) of 140 tests
performed on 60 different autoclaves, this procedure was approved in 2009, included, as a new annex,
in the guide of good practice for AE testing of pressure equipment from AFIAP and in a new profes-
sional technical guide for “in-service inspection and requalification of autoclaves in the aircraft and
aerospace industry”. This paper describes the applied methodology and the ROE with many examples.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page53
87Ea
ÉQUIPEMENT SOUS PRESSION SOUS TALUS : CONTROLE PAR
ÉMISSION ACOUSTIQUE SELON LA MÉTHODOLOGIE CETIM
BURIED PRESSURE VESSELS: ACOUSTIC EMISSION TESTING
BY CETIM METHODOLOGY
RÉSUMÉ
Le contrôle par émission acoustique des appareils à pression est appliqué depuis des décennies en
France, en Europe et dans le monde. Il permet en effet un contrôle global et rapide de grandes structures,
minimisant fortement les temps d’intervention et d’immobilisation des installations. Le Cetim déve-
loppe cette méthode depuis plus de 3 décennies.
Dans un souci de diminuer l’impact de la présence de grands réservoirs de stockage de gaz sur l’envi-
ronnement (facteur de risque en cas d’incident), les projets de ce type sont soumis à des contraintes
nouvelles depuis plusieurs années. Ces stockages de gaz, cylindriques ou sphériques, sont alors soit
recouverts par un talus, soit placés dans un sarcophage rempli de sable.
La question du contrôle périodique se pose pleinement, puisque la surface de l’équipement sous pression
n’est plus accessible. En tant que mesure compensatoire à ce problème d’accessibilité, l’émission acous-
tique est alors envisageable.
Cet article présente la méthodologie développé au Cetim en illustrant les étapes de mise en œuvre de
l’émission acoustique sur un équipement sous pression de stockage de gaz.
Dès la conception, les équipements sous pression ont pris en compte les contraintes liées à la mise en
œuvre de l’émission acoustique. En effet, la solution retenue consistait à utiliser des guides d’onde,
permettant la surveillance de la partie rendue inaccessible par le talus. Des essais en laboratoire jusqu’au
test final après ensevelissement des sphères, cette technologie a été validée. Le suivi par émission
acoustique des épreuves hydrauliques initiales a notamment permis d’obtenir une première signature
acoustique de l’équipement, qui servira de référence pour les essais de requalification futurs.
C’est un des plus grands projets industriels de ce type si l’on considère que près de 400 capteurs d’émis-
sion acoustique ont été nécessaires pour le contrôle de ces équipements de stockage.
ABSTRACT
Acoustic emission testing of pressure vessels has been applied for decades in France, Europe and world-
wide. It allows overall and fast monitoring of large structures, greatly minimizing the time of mainte-
nance and shutdown of plants. Cetim develops this method for more 3 decades.
In order to reduce the impact of the presence of large gas storage tanks on the environment and to mi-
nimize the risk level, projects of this type are subject to new constraints for several years. These gas
storage vessels, cylindrical or spherical, are then covered by either an embankment slope or placed in
a sarcophagus filled with sand. The question of periodic testing arises fully, since the surface of the
pressure equipment is no longer available for inspection. As a compensatory measure to the problem
of accessibility, acoustic emission is then possible and appears of great interest.
The experience reported in this article present Cetim methodology and illustrates all stages of acoustic
emission implementation on pressure equipment. From design, pressure equipments have taken into
account the constraints related to the integration of acoustic emission. Indeed, the solution was to use
waveguides, allowing the monitoring of inaccessible areas due to the presence of the embankment.
From laboratory experimentation until final tests after burying, this technology has been validated. The
monitoring of hydraulic pressure test by acoustic emission provided a first acoustic signature of the
vessel, which will serve as reference for the future.
This successful project is one of the largest industrial projects in the world if we consider that nearly
400 acoustic emission transducers were needed to test these pressure storage vessels.
AFIAP - ESOPE 2010 Programme
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page54
25/26E Evaluation des méthodes de contrôle non destructif pour la LE NEVE C. / RICHEZ M.
détection de l’attaque par hydrogène à chaud / Evaluation TOTAL RAFFINAGE ET
of non destructive testing for the detection of hot temperature MARKETING
hydrogen attack BEIGE J. / BLANCHARD S. /
DEMONTE S. /
LELIEVRE C.
INSTITUT DE SOUDURE
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page55
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page56
RÉSUMÉ / ABSTRACT
25/26E
ÉVALUATION DES MÉTHODES DE CONTRÔLE NON DESTRUCTIF
POUR LA DÉTECTION DE L’ATTAQUE PAR HYDROGÈNE À CHAUD
C. Le Nevé, M. Richez
(TOTAL Raffinage & Marketing, C.E.R.T. - 76700 Harfleur)
J. Beige, S. Blanchard, S. Demonte, C. Lelievre
(INSTITUT DE SOUDURE INDUSTRIE - 90, rue des Vanesses - 93420 Villepinte )
RÉSUMÉ
De nombreuses unités de raffinage construites jusque dans les années 70 ont utilisé des aciers à 0.5%
de Molybdène pour la fabrication des équipements fonctionnant en température et sous pression
d’hydrogène. Sous certaines conditions, ils peuvent être endommagés par l’attaque par l’hydrogène à
Mercredi
chaud. Ces équipements doivent faire l’objet d’un suivi en service spécifique.
La présente étude a évalué la capacité de différentes méthodes de contrôle non destructif, basées sur
les ultrasons (TOFD, AUBT, Phased Array), à détecter cet endommagement complexe, à partir
d’échantillons provenant d’équipement attaqué en service. Préalablement, les dégradations ont été
identifiées et caractérisées par des examens métallurgiques. Ils constituent désormais des cales de
référence.
Ces évaluations ont permis de développer des modes opératoires d’examen non destructifs combinant
plusieurs de ces techniques ultrasonores pour contrôler un équipement (métal de base, soudure et ZAT)
vis-à-vis des dégradations par l’hydrogène à chaud.
Les cales de référence pourront être utilisées ultérieurement pour vérifier l’aptitude du matériel de
contrôle à détecter les dégradations et à vérifier la compétence du personnel en charge des contrôles et
de l’interprétation des résultats.
ABSTRACT
Many refining units, built up to the beginning of the 70s, include equipment made of 0.5% molybdenum
steel operating at high pressure and temperature under hydrogen. Under these severe operating
conditions, this metallurgy can be sensitive to high-temperature hydrogen attack . The inspection of
the equipment shall take into account HTHA risks.
This study assessed the capability of NDE techniques based on ultrasonics to detect HTHA on material
samples coming from damaged vessels. These samples heve been investigated to characterize HTHA
damages. They now constitute reference blocks.
These evaluations have allowed to develop NDE procedures using some of these ultrasonic techniques
to check equipment (base metal, weld and HAZ) towards the HTHA damages.
Reference blocks can be later used to verify the ability of NDE to detect the damages and to verify the
competencies of the personnel in charge of the controls and their interpretation.
INDEX çç Sommaire
Evaluation des méthodes de contrôle non destructif
pour la détection de l’attaque par hydrogène à chaud
C. Le Nevé, M. Richez
(TOTAL Raffinage & Marketing, C.E.R.T. - 76700 Harfleur)
J. Beige, S. Blanchard, S. Demonte, C. Lelievre
(INSTITUT DE SOUDURE INDUSTRIE - 90, rue des Vanesses - 93420 Villepinte )
Introduction/résumé
De nombreuses unités de raffinage construites jusque dans les années 70 ont utilisé des aciers
à 0.5% de Molybdène pour la fabrication des équipements fonctionnant en température et
sous pression d’hydrogène. Sous certaines conditions, ils peuvent être endommagés par
l’attaque par l’hydrogène à chaud. Ces équipements doivent faire l’objet d’un suivi en service
spécifique.
Ces évaluations ont permis de développer des modes opératoires d’examen non destructifs
combinant plusieurs de ces techniques ultrasonores pour contrôler un équipement (métal de
base, soudure et ZAT) vis-à-vis des dégradations par l’hydrogène à chaud.
Les cales de référence pourront être utilisées ultérieurement pour vérifier l’aptitude du
matériel de contrôle à détecter les dégradations et à vérifier la compétence du personnel en
charge des contrôles et de l’interprétation des résultats.
C. Le Nevé, M. Richez
(TOTAL Raffinage & Marketing, C.E.R.T. - 76700 Harfleur)
J. Beige, S. Blanchard, S. Demonte, C. Lelievre
(INSTITUT DE SOUDURE INDUSTRIE - 90, rue des Vanesses - 93420 Villepinte )
ABSTRACT
Many refining units, built up to the beginning of the 70s, include equipment made of 0.5%
molybdenum steel operating at high pressure and temperature under hydrogen. Under these
severe operating conditions, this metallurgy can be sensitive to high-temperature hydrogen
attack . The inspection of the equipment shall take into account HTHA risks.
This study assessed the capability of NDE techniques based on ultrasonics to detect HTHA on
material samples coming from damaged vessels. These samples heve been investigated to
characterize HTHA damages. They now constitute reference blocks.
These evaluations have allowed to develop NDE procedures using some of these ultrasonic
techniques to check equipment (base metal, weld and HAZ) towards the HTHA damages.
Reference blocks can be later used to verify the ability of NDE to detect the damages and to
verify the competencies of the personnel in charge of the controls and their interpretation.
INDEX çç Sommaire
1. Rappel de la situation
L’API 941 [1] donne des recommandations générales sur le choix des matériaux en service
hydrogène. Elle propose des courbes expérimentales qui indiquent les domaines d’emploi de
l’acier carbone et des aciers faiblement alliés au chrome et au molybdène.
Le cas particulier des équipements en acier à 0,5 % de molybdène y est longuement traité, car
la courbe d’emploi à été plusieurs fois abaissée.
Les cas reportés de dégradation d’équipements en acier 0,5 % Mo sont indiqués sous la forme
de triangles noirs (voir figure 1). Les deux retours d’expérience proposés dans la présente
étude sont reportés en rouge.
Figure 1
INDEX çç Sommaire
2. Rappel sur l’attaque par l’hydrogène à chaud (HTHA)
Pour des températures supérieures à 230°C, la diffusion de l’hydrogène dans un métal peut
provoquer de l’attaque à chaud (high temperature hydrogen attack). Cette attaque se présente
sous deux formes : une décarburation de surface et/ou une décarburation interne [3] .
La décarburation interne est liée à la diffusion de l’hydrogène atomique qui déstabilise les
carbures pour former du méthane. Cela provoque une baisse des caractéristiques mécaniques
provoquée par la perte de carbures et une dégradation ‘’mécanique’’ de la structure avec
l’apparition de micro bulles de méthane piégées aux joints de grain. Ces micro bulles peuvent
provoquer des décohésions importantes dans la structure pouvant engendrer des micros puis
macros fissures et/ou des blisters.
Dans les deux cas étudiés, des dégradations ‘’mécaniques’’ de la structure ont
systématiquement été observées. La décarburation de la structure au niveau des zones
attaquées n’a pas toujours été mise en évidence avec les examens métallurgiques traditionnels
utilisés. Dans tous les cas, cette décarburation n’est pas généralisée dans ces zones mais plutôt
localisée.
Les techniques de contrôle évaluées pour la détection de l’HTHA sont toutes basées sur les
ultrasons :
- contrôle ultrasons par TOFD (Time Of Flight Diffraction)
- contrôle ultrasons par AUBT (Advanced Ultrasonic Back-scattered Technique)
- contrôle ultrasons par Phased-Array (sonde multiéléments)
Un descriptif de ces techniques a été proposé lors d’ESOPE 2007 [4] et elles sont explicitées
dans plusieurs références bibliographiques [5], [6], [7]
3.1. Le TOFD
Le TOFD (Time Of Flight Diffraction) repose sur le phénomène de diffraction généré par les
extrémités de défauts. Cette méthode peut aussi bien être mise en œuvre sur la soudure que
sur le métal de base.
En l’absence de défaut dans la zone insonifiée, il y a détection de deux ondes, l’onde latérale
se propageant sous la surface et l’écho de réflexion sur le fond. Lorsqu’il y a présence d’un
défaut, les extrémités de celui-ci diffractent les ondes ultrasonores émises et un signal
supplémentaire est alors détecté.
L’exploitation de l’image représentant une coupe latérale de la soudure, permet d’identifier et
de caractériser des indications de défauts.
INDEX çç Sommaire
Figure 2 : Principe du TOFD
INDEX çç Sommaire
3.3. Le PHASED ARRAY
Les 3 techniques de CNDs sélectionnées ont été utilisées pour contrôler l’épaisseur complète
des échantillons, en accédant par la face externe non exposée au produit, c'est-à-dire par
l’extérieur des équipements.
Leur mise en œuvre sur des équipements claddés par de l’ acier inoxydable, suit donc le
même mode opératoire, pour la détection de dégradation par HTHA dans l’acier faiblement
allié.
Les échantillons de la présente étude ne possédant pas de clad, des contrôles par
Magnétoscopie (fluorescente) ont également été réalisés en face interne.
INDEX çç Sommaire
5. Choix des échantillons :
Sur le reformeur, c’est le fluage qui est le mode d’endommagement prédominant avec
d’importantes dégradations en ZAT (fissures type IV), mais des dégradations par l’hydrogène
sont également visibles en métal de base .
Sur le réacteur d’hydrofinissage, c’est l’HTHA qui est le mode de dégradation reconnu. Il est
bien visible sur le métal de base mais pas en ZAT comme souvent mentionné dans la
littérature. De plus, des dégradations importantes par HTHA en métal fondu sont également
visibles.
Pour la présente étude, le choix retenu est une sélection de trois échantillons présentant 3
niveaux d’endommagement différents détectés par les CNDs en métal de base :
- Une zone sans endommagement
- Une zone avec un endommagement faible (6 à 7 mm d’HTHA en métal de base, côté
paroi interne)
- Une zone avec un endommagement important (30 à 60% d’ HTHA en métal de base, côté
paroi interne)
Sur l’ensemble de ces échantillons, les trois techniques retenues ont détecté l’HTHA, ou plus
précisément des dégradations ‘’mécaniques’’ de la structure du fait de la présence de micros
et macros fissures. Cependant, plusieurs nuances apparaissent entre ces techniques.
Sur les échantillons non et faiblement attaqués, le TOFD s’est montré d’avantage ‘’perturbé’’
par le bruit de structure (correspondant à une propreté inclusionnaire médiocre) que les deux
autres techniques, ceci s’expliquant par une sensibilité de détection plus importante de cette
méthode par rapport aux deux autres (en raison de l’exploitation du phénomène de diffraction
et l’emploi de traducteurs fortement divergents).
Sur des niveaux de faible endommagement, l’AUBT est la seule technique à discerner
l’HTHA sans confusion possible avec le bruit de structure. Le phased array permet un
meilleur discernement de ce faible endommagement que le TOFD.
Dès lors que l’endommagement par HTHA est détecté et identifié par ces techniques, elles
mesurent toutes les trois des épaisseurs de métal dégradé assez proches.
L’annexe 1 présente des exemples d’indications relevées par TOFD, AUBT et Phased Array
en métal de base sur les zones faiblement et fortement endommagées par HTHA.
INDEX çç Sommaire
7. Expertise des échantillons de métal de base contrôlés par CNDs
Des expertises métallographiques ont permis de confirmer les épaisseurs d’HTHA décelées en
CNDs.
Des microfissures intergranulaires ont été observées sur une épaisseur de 6mm sur
l’échantillon faiblement attaqué et sur une épaisseur de 28mm sur l’échantillon fortement
attaqué.
Sur les mêmes zones, les CNDs mettent en évidence des indications d’HTHA sur une
épaisseur de 4 à 7mm sur l’échantillon faiblement attaqué et sur une épaisseur de 27 à 31mm
sur l’échantillon fortement attaqué.
Ces décohésions intergranulaires ne sont donc pas décelables par les méthodes CNDs
actuelles.
Des essais mécaniques comparatifs ont permis d’appréhender leur nocivité. Des éprouvettes
de traction ont été prélevées dans 3 zones d’une même coupe, à différentes épaisseurs: au
niveau de la peau externe sans défaut, dans l’épaisseur avec des décohésions intergranulaires
mais sans microfissures intergranulaires et au niveau de la peau interne avec des microfissures
intergranulaires.
Les résultats montrent que les caractéristiques mécaniques (Rm, Rp0,2%) à l’ambiante et à une
température représentative de la température de service sont proches entre les zones
présentant des microfissures intergranulaires et celles présentant des petites décohésions
intergranulaires à l’échelle du grain. Les résultats obtenus pour ces deux zones sont
légèrement inferieurs à ceux obtenus sur l’éprouvette prélevée dans une zone saine proche de
la surface extérieure de la virole (-10% pour Rp0,2% et -15%à -20% pour Rm) mais restent
superieurs au minimum requis par les normes.
Par contre, les caractéristiques de ductilité (A%, Z%) sont affectées par ces dégradations. Les
résultats obtenus sur la zone avec des microfissures intergranulaires sont nettement inférieurs
à ceux obtenus sur la zone avec décohésions intergranulaires, et très inferieurs à ceux de la
zone sans défaut.
Compte tenu des sensibilités de détection des 3 techniques ultrasonores pour la détection de
l’HTHA, deux d’entre elles sont retenues pour le contrôle du métal de base :
- contrôle ultrasons par Phased-Array
- contrôle ultrasons par AUBT
INDEX çç Sommaire
envisageable. Le pourcentage de surface exploré doit être adapté à la criticité de l’équipement.
Cette technique permet d’enregistrer les contrôles effectués et de faire un suivi de l’évolution
dans le temps.
L’AUBT permet la détection d’un niveau de dégradation par HTHA Il permet surtout de
discerner l’HTHA du bruit de structure. Son usage doit être maintenu sur les zones les plus LENEVE Charles 7/5/10 14:40
Mis en forme: Français
critiques d’un équipement (par exemple au niveau des nœuds de soudures), et pour confirmer
LENEVE Charles 7/5/10 14:41
les zones potentielles d’HTHA décelées en phased array. Supprimé: , quand celui-ci atteint 10% de
l’épaisseur du métal de base.
La recherche d’HTHA dans les soudures (ZAT et métal fondu) utilise classiquement les
techniques UT disponibles, afin de mettre en évidence des indications de dégradation.
Seules les techniques TOFD et Phased Array ont été comparées, car l’AUBT n’est pas
envisageable sur ces zones présentant d’importantes variations structurales.
Les tests ont été menés sur le réacteur d’hydrofinissage qui ne présente que des défauts dans
le métal fondu des soudures (rien en ZAT).
La technique TOFD permet de déceler des indications provoquées par l’HTHA dans le métal
fondu. La technique Phased Array est nettement moins pertinente actuellement, du fait de
traducteurs non adaptés (manque d’amortissement des traducteur).
Les indications d’HTHA décelées en métal fondu par les CNDs sont des microfissures
interdentriques. Les décohésions interdendritiques beaucoup plus fines observées en
micrographie sur des épaisseurs plus importantes ne sont pas décelables avec les techniques
ultrasonores actuelles.
L’annexe 4 présente des exemples d’indications relevées par TOFD, AUBT et Phased Array
en métal fondu sur les zones faiblement et fortement endommagées par HTHA.
L’annexe 5 présente des exemples d’illustrations micrographiques de zones faiblement et
fortement endommagées par HTHA.
Dans tous les cas, cette recherche d’HTHA nécessite des contrôleurs avertis et connaissant ce
type de dégradation. D’autres examens comme la magnétoscopie fluorescente et les répliques
métallographiques peuvent être mis en œuvre pour confirmer le diagnostic CND, quand la
surface dégradée est accessible.
L’usage de cales de référence est indispensable pour fiabiliser le diagnostic des contrôles en
métal de base. En effet, s’agissant de dégradations difficilement décelables en CND
INDEX çç Sommaire
traditionnel (microfissures intergranulaires), les indications relevées en UT doivent être
clairement identifiées afin de ne pas les confondre avec d’autres indications non nocives
(bruit de ‘’fond’’ lié à la propreté inclusionnaire par exemple).
Ces cales de référence servent également pour former et qualifier les contrôleurs, et pour
continuer à développer ces techniques pour la détection d’HTHA.
11. Conclusion
Les développements récents des techniques de CND par ultrasons (TOFD, AUBT, Phased
Array) et les techniques d’imagerie, permettent de déceler des indications produites par de
l’endommagement par l’HTHA, aussi bien en métal de base qu’en métal fondu.
Les possibilités actuelles de ces techniques permettent de mettre en évidence les microfissures
présentes dans la structure, mais pas encore les petites décohésions à l’échelle du grain.
Cependant, les essais mécaniques réalisés montrent que ces petites décohésions ont un impact
limité sur les caractéristiques de résistance des matériaux.
A partir de plusieurs retours d’expérience, des cales de référence ont été réalisées et des
modes opératoires de contrôle performants pour détecter l’éventuelle apparition d’HTHA ont
été élaborés. Ces cales seront employées pour qualifier de nouvelles méthodes de contrôle ou
pour qualifier les opérateurs.
Ces progrès récents seront mis à profit pour améliorer l’inspection des équipements concernés
par ce mode de dégradation et contribueront à l’amélioration de la sécurité de nos unités.
INDEX çç Sommaire
BIBLIOGRAPHIE
[1] API Recommended Practice 941, 7ème édition aout 2008, Steels for hydrogen service at
elevated temperatures and pressures in petroleum refineries and petrochemical plants
[2] RICHEZ M., C9 - Gestion des appareils en acier faiblement alliés contenant 0.5%Mo et
exploités en service hydrogène, Congrès AFIAP-ESOPE, 2007
[3] PRESCOTT G.R., History and basis of prediction of hydrogen attack of C-0.5Mo steel,
a state of the art review presented at the second international conference on interaction
of steels with hydrogen in petroleum industry pressure vessel and pipeline service, the
materials properties council, inc, Vienna Austria 19.10.1994.
INDEX çç Sommaire
ANNEXE 1
TOFD
Phased Array
AUBT
AUBT
INDEX
LENEVE C. / RICHEZ M. - TOTAL RAFFINAGE ET MARKETING
LELIEVRE C. / BEIGE J. / DEMONTE S. /BLANCHARD S. - INSTITUT DE SOUDURE çç Sommaire
Phased Array
TOFD
AUBT
INDEX
LENEVE C. / RICHEZ M. - TOTAL RAFFINAGE ET MARKETING
LELIEVRE C. / BEIGE J. / DEMONTE S. /BLANCHARD S. - INSTITUT DE SOUDURE çç Sommaire
TOFD
Phased Array
AUBT
INDEX
LENEVE C. / RICHEZ M. - TOTAL RAFFINAGE ET MARKETING
LELIEVRE C. / BEIGE J. / DEMONTE S. /BLANCHARD S. - INSTITUT DE SOUDURE çç Sommaire
ANNEXE 2
53 MM
28 MM
FIGURE 1 : ILLUSTRATION D’UN COUPON PRESENTANT UNE SUSPICION FORTE D’ENDOMMAGEMENT PAR HTHA
EPAISSEUR ENDOMMAGEE PAR MICROFISSURES INTERGRANULAIRES : 28 MM
EPAISSEUR ENDOMMAGEE PAR DECOHESIONS INTERGRANULAIRES A L’ECHELLE DU GRAIN : 53 MM
EPAISSEUR
AFFECTEE PAR LES
MICROFISSURES
INTERGRANULAIRES
28 MM
INDEX
LENEVE C. / RICHEZ M. - TOTAL RAFFINAGE ET MARKETING
LELIEVRE C. / BEIGE J. / DEMONTE S. /BLANCHARD S. - INSTITUT DE SOUDURE çç Sommaire
500 : 1
50µm
FIGURE 3 : DETAIL DES DOMMAGES EN ZONE FORTEMENT ENDOMMAGEE PAR DES MICROFISSURES
FIGURE 4 : DETAIL DES DOMMAGES EN ZONE AFFECTEE PAR DES MICROFDECOHESIONS DE LA LONGUEUR D’UN GRAIN
INDEX
LENEVE C. / RICHEZ M. - TOTAL RAFFINAGE ET MARKETING
LELIEVRE C. / BEIGE J. / DEMONTE S. /BLANCHARD S. - INSTITUT DE SOUDURE çç Sommaire
FIGURE 6 : DETAIL DE LA MICROSTRUCTURE AVEC DES INDICES D’UNE LEGERE DECARBURATION DE LA STRUCTURE
INDEX
LENEVE C. / RICHEZ M. - TOTAL RAFFINAGE ET MARKETING
LELIEVRE C. / BEIGE J. / DEMONTE S. /BLANCHARD S. - INSTITUT DE SOUDURE çç Sommaire
43 MM
6 MM
FIGURE 7 : ILLUSTRATION D’UN COUPON PRESENTANT UNE SUSPICION MOYENNE D’ENDOMMAGEMENT PAR HTHA
EPAISSEUR ENDOMMAGEE PAR MICROFISSURES INTERGRANULAIRES :6 MM
EPAISSEUR ENDOMMAGEE PAR DECOHESIONS INTERGRANULAIRES A L’ECHELLE DU GRAIN : 43 MM
FIGURE 8 : ILLUSTRATION D’UN COUPON PRESENTANT UNE SUSPICION MOYENNE D’ENDOMMAGEMENT PAR HTHA
EPAISSEUR ENDOMMAGEE PAR MICROFISSURES INTERGRANULAIRES :6 MM
EPAISSEUR ENDOMMAGEE PAR DECOHESIONS INTERGRANULAIRES A L’ECHELLE DU GRAIN : 43 MM
INDEX
LENEVE C. / RICHEZ M. - TOTAL RAFFINAGE ET MARKETING
LELIEVRE C. / BEIGE J. / DEMONTE S. /BLANCHARD S. - INSTITUT DE SOUDURE çç Sommaire
ANNEXE 3
FIGURE 1 : DETAIL DE LA MICROSTRUCTURE AVEC DES DECOHESIONS INTERGRANULAIRES PARALLELLES AUX SURFACES
INDEX
LENEVE C. / RICHEZ M. - TOTAL RAFFINAGE ET MARKETING
LELIEVRE C. / BEIGE J. / DEMONTE S. /BLANCHARD S. - INSTITUT DE SOUDURE çç Sommaire
ANNEXE 4
TOFD
Phased Array
INDEX
LENEVE C. / RICHEZ M. - TOTAL RAFFINAGE ET MARKETING
LELIEVRE C. / BEIGE J. / DEMONTE S. /BLANCHARD S. - INSTITUT DE SOUDURE çç Sommaire
TOFD
Phased Array
INDEX
LENEVE C. / RICHEZ M. - TOTAL RAFFINAGE ET MARKETING
LELIEVRE C. / BEIGE J. / DEMONTE S. /BLANCHARD S. - INSTITUT DE SOUDURE çç Sommaire
ANNEXE 5
34 MM
23 MM
FIGURE 2 : ILLUSTRATION D’UN COUPON PRESENTANT UN ENDOMMAGEMENT PAR HTHA EN METAL FONDU
INDEX
LENEVE C. / RICHEZ M. - TOTAL RAFFINAGE ET MARKETING
LELIEVRE C. / BEIGE J. / DEMONTE S. /BLANCHARD S. - INSTITUT DE SOUDURE çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page57
RÉSUMÉ / ABSTRACT
4E
DIAGNOSTIC GLOBAL DES ÉQUIPEMENTS SOUS PRESSION
EXEMPLE D’APPLICATION CONCERNANT UNE SPHÈRE DE STOCKAGE
RÉSUMÉ
L’âge moyen des appareils à pression en service dans le monde augmente d’année en année. Une préoc-
cupation majeure des industriels exploitants est d’assurer la sécurité de leurs installations et des per-
sonnes, tout en prolongeant la durée de vie de leur parc d’appareils à pression. Dans ce contexte, il est
nécessaire de maîtriser l’état de santé de ce type de structure, et également être capable d’évaluer la
durée de vie résiduelle d’un appareil.
Au travers d’un cas réel, l’étude présentée dans cet article illustre toutes les étapes de mise en œuvre
d’un processus ayant permis dans un premier temps d’effectuer un diagnostic précis de l’état de santé
d’une sphère de stockage et ainsi évaluer sa durée de vie, puis, dans un second temps, assurer les opé-
rations de remise en conformité pour assurer une remise en service de cette structure.
Ce processus, entièrement assuré par le CETIM, a impliqué les opérations suivantes :
• Une épreuve hydraulique avec suivi par émission acoustique,
• Des contrôles CND complémentaires pertinents (TOFD, ultrasons, Magnétoscopie …),
• Des contrôles métallographiques par la méthode de répliques et / ou prélèvements,
• Une étude de durée de vie, estimée par calculs, basée sur les CND et les études métallurgiques,
• La rédaction d’une procédure de recommandations de réparation par soudage des défauts critiques,
• La supervision des réparations et le contrôle après réparations
ABSTRACT
The average age of pressure vessels in service worldwide is increasing from year to year. A major con-
cern of industrial operators is to ensure the security of their facilities and people, while extending the
life of their fleet of pressure vessels. In this context, it is necessary to control the health of this type of
structure, and also be able to assess the residual life of a device.
Through a real case, the study presented in this article shows all the stages of implementation of a
process that allowed initially to make an accurate diagnosis of the condition of a sphere of storage
and to evaluate its lifetime, then in a second time, to ensure the operations back into compliance to en-
sure a reactivation of this structure.
INDEX çç Sommaire
Diagnostic Global des équipements sous pression - Exemple
d’application concernant une sphère de stockage
Global Diagnosis of pressure equipment - Example of application for a spherical storage tank
RESUME
L’âge moyen des appareils à pression en service dans le monde augmente d’année en année. Une
préoccupation majeure des industriels exploitants est d’assurer la sécurité de leurs installations et des
personnes, tout en prolongeant la durée de vie de leur parc d’appareils à pression. Dans ce contexte, il
est nécessaire de maîtriser l’état de santé de ce type de structure, et également être capable d’évaluer la
durée de vie résiduelle d’un appareil.
Au travers d’un cas réel, l’étude présentée dans cet article illustre toutes les étapes de mise en œuvre
d’un processus ayant permis dans un premier temps d’effectuer un diagnostic précis de l’état de santé
d’une sphère de stockage et ainsi évaluer sa durée de vie, puis, dans un second temps, assurer les
opérations de remise en conformité pour assurer une remise en service de cette structure.
ABSTRACT
The average age of pressure vessels in service worldwide is increasing from year to year. A major
concern of industrial operators is to ensure the security of their facilities and people, while extending
the life of their fleet of pressure vessels. In this context, it is necessary to control the health of this type
of structure, and also be able to assess the residual life of a device.
Through a real case, the study presented in this article shows all the stages of implementation of a
process that allowed initially to make an accurate diagnosis of the condition of a sphere of storage
and to evaluate its lifetime, then in a second time, to ensure the operations back into compliance to
ensure a reactivation of this structure.
INDEX çç Sommaire
This process is fully insured by the CETIM, involved the following:
• A hydraulic test with acoustic emission monitoring,
• Testing additional relevant NDT (TOFD, Ultrasonic, Magnetic ...)
• Checks by the method of metallographic replicas and / or levies,
• A study of life, estimated by calculations based on NDT and metallurgical studies,
• Writing a procedure recommendations repair by welding of critical defects,
• The supervision during repairs and the testing after repairs.
INTRODUCTION
L’âge moyen des appareils à pression en service dans le monde augmentant d’année en année, il
devient essentiel de pouvoir effectuer des diagnostics fiables et complets de ces installations, afin de
prolonger leur durée d’exploitation tout en assurant un niveau de sécurité satisfaisant. Maîtriser l’état
de santé de ce type de structure nécessite des contrôles réguliers, faisant appel à de nombreuses
technologies, complémentaires, aboutissant à une description précise des éventuels endommagements.
Lorsqu’une structure présente une quelconque dégradation, il est alors nécessaire d’évaluer sa durée de
vie résiduelle. Ainsi, effectuer un diagnostic précis sur la santé d’une structure, et donner les
informations pertinentes sur les éventuelles réparations nécessite des champs de compétences couvrant
des domaines de larges et variés.
Au travers d’un cas réel réalisé par le CETIM, l’étude présentée dans cet article illustre toutes les
étapes de mise en œuvre d’un processus ayant permis dans un premier temps d’effectuer un diagnostic
précis de l’état de santé d’une sphère de stockage et ainsi évaluer sa durée de vie, puis, dans un second
temps, assurer les opérations de remise en conformité pour assurer une remise en service de cette
structure.
La sphère de stockage de GPL ayant fait l’objet de l’expertise se trouve dans un grand parc de
stockage de près de 15000 m3.
L’objectif recherché dans le cadre de ce projet était de réaliser toutes les opérations nécessaires pour
évaluer l’intégrité de cette sphère, et à terme pouvoir la remettre en service. Les différentes étapes sont
décrites dans les paragraphes suivants.
L’épreuve hydraulique a été réalisée et suivie par émission acoustique. La sphère a donc été portée à
une pression maximale de 22.5 bars, en suivant un cycle de pressurisation défini dans la procédure de
contrôle.
La configuration adoptée par le CETIM pour la réalisation de ce contrôle, est de type planaire
(couverture globale de l’équipement, avec une localisation par triangulation en tout point des sources
d’émission acoustique). Ainsi, 77 capteurs ont été placés sur la structure. Toutes les opérations de
vérification préalables et postérieures au test ont permis de s’assurer de la sensibilité du contrôle,
comme par exemple la vérification de la localisation à partir de sources artificielles de type Hsu-
Nielsen.
Figure n°1 : Vérification de la localisation sur un pied de la sphère et sur une portion de soudure
circulaire
Un certain nombre de régions émissives ont été mises en évidence lors du test. Celles-ci ont été
clairement identifiées, caractérisées, et repérées sur la structure. Suivant leur degré d’émissivité (selon
le Guide des Bonnes Pratiques), ces régions ont nécessité des investigations complémentaires (CND
adapté à chaque cas) ou ont fait simplement l’objet d’un archivage.
INDEX çç Sommaire
Photo n°1 et figure n°3 : Région émissive ‘K’ – Analyse détaillée du comportement EA, photo de la
région suite au repérage réalisé postérieurement à l’épreuve hydraulique
Toutes les régions émissives détectées lors du contrôle par émission acoustique (de catégorie 2 ou 3)
ont fait l’objet de contrôles CND complémentaires. Les méthodes utilisées ont été les suivantes :
o Ultrasons - TOFD : pour soudure circulaire et verticale bout à bout
o Ultrasons : pour piquage, trou d’homme, zones pleine tôle
o Magnétoscopie : toute zone extérieure
o Ressuage : toute zone intérieure
o Radiographie : uniquement sur certaines régions nécessitant des investigations poussées.
Trois zones fissurées situées au niveau de cordons de soudure avaient été détectées par magnétoscopie
(zones n’ayant pas été révélées par le contrôle par émission acoustique). Afin d’augmenter les
informations relatives à la dimension de ces fissures le CETIM a mis en œuvre plusieurs techniques
(Ultrasons classiques, Ultrasons Multi-éléments, TOFD). Compte-tenu de la difficulté à distinguer ces
indications, le CETIM a préconisé l’arasage du cordon de soudure (avec un état de surface
parfaitement plan) afin de réaliser des contrôles UT plus fins, et tenter de dimensionner plus
précisément ces indications.
La combinaison de ces techniques a permis de dimensionner précisément deux zones fissurées. Notons
que la dernière fissure n’est détectable que par magnétoscopie.
Ce cas précis démontre qu’il est impératif de combiner les technologies pour aboutir à un diagnostic
complet et précis. Rester objectif et tenter de comprendre les résultats obtenus permet de progresser
techniquement. Les investigations complémentaires permettront de donner des explications pertinentes
sur ces défauts, et leur capacité à être détectés par telle ou telle méthode de contrôle non destructif.
Cette partie a apporté beaucoup d’éléments techniques et objectifs. Elle a permis de répondre à
certaines questions soulevées au cours des étapes précédentes.
Des répliques métallurgiques ont été faites sur les zones fissurées détectées par magnétoscopie et
ultrasons. Elles ont permis de confirmer la présence de défauts débouchants et de vérifier l’état
métallurgique de la zone soudée.
Pour caractériser complètement ces défauts et comprendre leur origine, il a été décidé de réaliser un
prélèvement (aussi appelé « barquette »). Les analyses et examens faits dans les laboratoires du
CETIM sur ces barquettes ont été :
- Ouverture des fissures et étude des plans de fissure,
- Analyse du mode de décohésion par microscopie électronique à balayage,
-INDEX
Examens macrographiques et micrographiques, çç Sommaire
- Réalisation d’éprouvettes de résilience sur la barquette afin d’estimer une valeur de ténacité
nécessaire pour le calcul.
Après réalisation de ces différents examens, les conclusions sont les suivantes :
- Les morphologies et les dimensions mesurées ont montré que nous sommes en présence de 2 types
de défauts :
o fissure en surface et débouchante. Celle-ci est peu profonde, probablement là depuis la
fabrication et ayant très peu évolué. Les plans de fissuration étant matés, et n’étant pas
évolutive, cela explique pourquoi elle n’a pas été détectée par émission acoustique.
o Défaut interne, qui est un manque de fusion probablement lié à une mauvaise
configuration du chanfrein de soudage. De même, son éventuelle évolution étant très
lente, cela explique pourquoi elle n’a pas été détectée par émission acoustique.
o Ces examens métallurgiques, révélant des fissures peu profondes, expliquent et
confirment les examens US, TOFD ou multi-éléments, qui voyaient relativement bien le
démarrage de la fissure, ainsi qu’une réponse un peu plus en profondeur avec de la
difficulté à bien « voir » entre les 2. Cela explique pourquoi il a été particulièrement
difficile de détecter et dimensionner ces défauts distincts.
o Les résultats des essais mécaniques sur éprouvettes tirées des barquettes montrent un
bon niveau de résilience. Cependant, de manière conservative, les valeurs de résilience
issues des QMOS seront prises pour la phase calcul.
.
1.5. Phase 4 : Calcul de la criticité des défauts et durée de vie
A partir de ces éléments, il est possible de déterminer la taille critique vis-à-vis du risque de rupture
fragile des défauts à partir d’une approche analytique faisant appel notamment à des fonctions
INDEX çç Sommaire
d’influence associées d’une part au type de défaut (débouchant, interne) et d’autre part aux termes de
contraintes.
Cette caractérisation consiste à :
- évaluer un coefficient d’intensité de contrainte en mode I, KI (mode d’ouverture du défaut par
traction sur ses lèvres) correspondant à la taille du défaut constaté,
- comparer ce coefficient KI au facteur d’intensité critique pour ce mode, KIC ou ténacité.
Cette méthodologie est utilisée dans diverses industries, notamment pétrolière et nucléaire.
Le logiciel utilisé pour l’étude proposée est le logiciel CETIM SECURE développé sur la base de
l’API RP 579 “fitness for service”.
En conclusions, les défauts de la zone K ne présentent pas de danger particulier, compte tenu d’un
certain nombre de paramètres, le nombre minimum de cycle résiduel en cas :
- d’une vidange totale de l’appareil avec mise à la pression atmosphérique est de 41 000 cycles,
- d’une variation uniquement de la pression hydrostatique de l’appareil correspondant à une
vidange et un remplissage partiel est de 449 000 cycles.
En ce qui concerne les défauts de type ‘fissure débouchante + défaut interne’ et compte tenu d’un
certain nombre de paramètres, le nombre minimum de cycle résiduel en cas :
- d’une vidange totale de l’appareil avec mise à la pression atmosphérique est de 32 000 cycles,
- d’une variation uniquement de la pression hydrostatique de l’appareil correspondant à une
vidange et un remplissage partiel est de 530 000 cycles.
INDEX çç Sommaire
1.6. Phase 5 : Recommandations de réparation par soudage – Supervision des
réparations
Les résultats obtenus suite à l’étude de nocivité ont montré que la durée de vie résiduelle de la sphère
permettait un maintien en service et qu’aucun défaut ne présentait de caractère nocif ;
Néanmoins, compte-tenu des prélèvements réalisés pour l’expertise métallurgique, et par précaution, il
a été décidé d’effectuer des réparations sur les zones suivantes :
- zones contenant des indications de surface : réparation par meulage ou affouillement et
rechargement ;
- zone du prélèvement de barquette : rechargement par soudage.
Dans la continuité des phases précédentes, le CETIM a assuré le suivi des opérations de soudage.
Ces travaux de soudure ont été réalisés en suivant les préconisations du CETIM, conformément à la
qualification de mode opératoire de soudage en respectant les exigences du code de construction
ASME sect. IX.
Le mode opératoire, ainsi que les soudeurs avaient préalablement été qualifiés.
Toutes les dispositions pouvant permettre une parfaite maîtrise de cette opération ont été prises :
- Examen visuel préalable,
- Préparation par un meulage de la surface afin d’éliminer tous les oxydes,
- Mesure des profondeurs d’affouillement,
- Contrôle par magnétoscopie avant et durant l’opération de soudage,
- Réalisation du soudage en respectant les phases de chauffage, le nombre de passes, les paramètres
de soudage, …
- Parachèvement par meulage.
Une dernière étape consistait à contrôler les zones de la sphère ayant subi des réparations. Ainsi les
différentes méthodes ont été mises en œuvre :
- Contrôle TOFD réalisé par l’intérieur et l’extérieur,
- Ultrasons réalisés avec capteurs 45°, 60° et 70°,
- Contrôle MT réalisé par l’intérieur (uniquement zone réparée ou rechargée).
INDEX çç Sommaire
Lors de ces contrôles, des petites indications linéaires ont été vues par MT, faiblement visibles par UT
et non détectables par TOFD. Très superficielles, un léger meulage a suffit à les éliminer.
En conclusions, aucune indication non acceptable n’a été détectée par les CND réalisés, sur les zones
concernées. Les réparations ont donc été validées, et la sphère redevient alors apte à subir une épreuve
hydraulique suivie par un organisme habilité (OH), dans le but d’une requalification décennale en
conformité à la réglementation locale en vigueur. Par la suite, un plan de contrôle de la sphère, c'est-à-
dire des contrôles CND périodiques, sera mis en place par l’exploitant de manière à maintenir la
maîtrise de l’équipement
CONCLUSIONS
Maîtriser l’état de santé de structures à risque nécessite des contrôles réguliers, faisant appel à de
nombreuses technologies, complémentaires, aboutissant à une description précise des éventuels
endommagements. Lorsqu’une structure présente une quelconque dégradation, il est alors nécessaire
d’évaluer sa durée de vie résiduelle. Ainsi, effectuer un diagnostic précis sur la santé d’une structure,
et donner les informations pertinentes sur les éventuelles réparations nécessite des champs de
compétences couvrant des domaines de larges et variés.
Au travers du cas réel réalisé par le CETIM, et présenté dans cet article, on peut prendre conscience de
la nécessité de compétences multiples nécessaires à la mise en œuvre d’un tel processus :
- Contrôles non destructifs,
- Analyse métallurgique,
- Calcul,
- Soudage, …
Afin d’aboutir à une opération réussie, toutes ces multiples actions devront cependant être connectées,
synthétisées, coordonnées. Le CETIM, de par son très large champ de compétences, est naturellement
structuré pour assurer ce type d’opérations.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page58
RÉSUMÉ / ABSTRACT
44E
CONTRÔLE DES TUYAUTERIES OFFSITES
A. Houdaïbi
Raffinerie TOTAL Donges
RÉSUMÉ
Les tuyauteries offsites de la raffinerie de Donges font actuellement l’objet d’une campagne globale
d’inspection. L’importance du périmètre de cette campagne et les nombreuses spécificités de ce type
de tuyauteries ont conduit la raffinerie à opter pour une gestion de type « projet ». Dans un premier
temps il a été nécessaire de prioriser les lignes en particulier en intégrant les conséquences
environnementales d’une éventuelle défaillance. Une équipe pluridisciplinaire a ainsi été mise en place,
Mercredi
des choix techniques adaptés aux modes de dommages rencontrés ont été arrêtés, des indicateurs de
suivi ont été définis et un comité de pilotage permet de s’assurer du bon déroulement de la campagne.
Cette présentation a pour but d’exposer l’organisation mise en place pour mener à bien cette campagne.
Il présente également les choix techniques adoptés par la raffinerie de Donges en terme de méthodes
d’inspection et de contrôle non destructif ainsi que les principaux enseignements qui en ont été tirés à
ce jour.
ABSTRACT
The refinery TOTAL of Donges is carrying out a global inspection program of its offsite piping. The
importance of the scope and the numerous specificities of this type of piping led the refinery to manage
this program as a specific project. In a first step, piping has been prioritized particularly by introducing
the environmental consequences of an eventual failure. A multidisciplinary team was set up, technical
choices adapted to the observed damage modes were chosen, monitoring indicators were defines and
a steering committee is in charge of the follow-up of the project good progress.
The present report aims at explaining the organization that has been put in place to carry out this
global inspection program. It also presents the technical choices adopted by the refinery of Donges in
term of inspection method and non destructive testing as well as the main learning up to this day.
INDEX çç Sommaire
1. INTRODUCTION
La raffinerie TOTAL Donges a entrepris depuis 2008 une importante campagne d’ inspection
de ses tuyauteries offsites (hors unité de production). Compte tenu de l’ ampleur du périmètre
et des moyens humains nécessaires à cette campagne, une organisation de type projet a été
mise en place. Celle-ci permet de mener parallèlement les inspections et les travaux en
découlant. Cette présentation a pour but de décrire l’ organisation mise en place dans le cadre
de cette campagne, les moyens techniques utilisés ainsi que les principaux enseignements
retenus en terme de corrosion.
2. PRINCIPES DE BASE
Pour la majeure partie des lignes, aucun calcul de criticité, type API 581, n’ était disponible au
moment du lancement de cette campagne. Ce sont les premiers résultats de la campagne qui
ont permis de dégager les grands principes pour adapter les inspections aux risques de
corrosion.
- Les corrosions sont essentiellement externes et détectées au droit des « défauts » des
tôles de calorifuge, au droit des supports propices à la rétention d’ eau (ex: colliers
boulonnés), des dégradations de revêtement.
Ces premiers résultats ont donc conduit à adapter la nature et l’ étendue des inspections (voir
annexe 1) aux modes de dégradation, aux types de supports, à la présence ou non de
calorifuge et à la criticité environnementale des lignes (voir annexe 2). Cette dernière a été
définie en prenant en compte la distance aux clôtures douanières de la raffinerie ainsi qu’ au
cours d’ eau la traversant. Ceci passe donc par une caractérisation préalable des différentes
zones exposées.
Enfin l’ importante superficie couverte par les tuyauteries a conduit à adopter une approche
géographique par rack (optimisation des moyens d’ accès et des ressources humaines).
INDEX çç Sommaire
3. ORGANISATION DE LA CAMPAGNE INSPECTION :
La raffinerie a fait le choix de mettre en place une structure de type projet en nommant un
chef de projet et en détachant des ressources internes (maintenance, sécurité, exploitation et
inspection). Ci-dessous sont décrites les différentes cellules du projet :
• Cellule Inspection qui mène les inspections suivant une stratégie clairement définie
(l’ homogénéisation des pratiques de terrain et des comptes-rendus est importante dès
lors que l’ équipe compte plusieurs inspecteurs)
INDEX çç Sommaire
3.2 Organigramme du Projet (2009)
En 2009, ceux sont en moyenne 150 à 200 personnes qui étaient engagées dans le projet
RACKS
INDEX çç Sommaire
4. SOLUTIONS TECHNIQUES POUR INSPECTION AU DROIT DES SUPPORTS
Dans un premier temps il a été nécessaire de créer une bibliothèque des différents types de
support avec une classification propre à la raffinerie TOTAL Donges, des supports propices à
la corrosion externe (voir annexe 4), ceci afin d’ adapter l’ étendue des inspections aux type de
supports.
Ensuite, compte tenu de la difficulté d’ apprécier les défauts au droit des supports, la raffinerie
a adapté la méthode d‘inspection aux types de supports. Ainsi, 3 moyens complémentaires
aux visuels ont été utilisés :
Le retrait du support
Cette technique permet le contrôle de plusieurs supports simultanément. Elle met en œuvre
des ondes émises depuis des couronnes de diamètres adaptés à ceux des tuyauteries et dans
lesquelles sont disposés une multitude de capteur US.
INDEX çç Sommaire
On notera que cette méthode n' est initialement pas prévue pour le
contrôle de supports. Dans notre application, elle se limite aux lignes
150 mm
reposant directement sur poutre. Les réglages pour l'interprétation sont
spécifiques à la recherche de ce type de corrosion. La méthode USOG
donne un résultat de perte d' épaisseur par section. Dans le cas des
contrôles que nous avons réalisés, il n' y a pas de perte d' épaisseur
interne et externe sauf au niveau des supports. La perte de section est
rapportée à une plage de corrosion de largeur 150 mm
Plus de 20000 m de tuyauteries ont été contrôlés en USOG au niveau de plus de 1000
supports. Les classements «moyen» (perte entre 20 et 40%) ou «sévère» (perte > 40%) ont été
vérifiés par un visuel à 100%. Le classement en « mineur » des défauts qui s’ avèrent en réalité
moyens ou sévères représente 10 à 20% des cas. Après analyse, ces classements erronés sont
liés à des problèmes de réglage d’ appareillage, d’ interprétation ou de limite technique sur les
lignes de DN élevé par exemple…
Il faut éviter de se reposer uniquement sur les USOG pour les lignes critiques.
INDEX çç Sommaire
5. ILLUSTRATIONS DES PRINCIPAUX MODES DE DOMMAGE RENCONTRES
Sont présentées ci-dessous les principales dégradations rencontrées au cours de ces 2 années
complètes d’ inspection. Comme expliqué dans le chapitre « 2- principe de base », il s’ agit
essentiellement de corrosions externes :
Corrosion à l’ interface des tuyauteries et des supports (Conception, réalisation des
supports, passage de route sans fourreau ou centreur)
INDEX çç Sommaire
Corrosion entrée/sortie de terre, lignes enfouies, Mauvais drainage sous les pipe-
ways, accumulation d’ eau, de dépôts, de végétation
6. CONCLUSION
Le contrôle des lignes offsites sur une raffinerie est une opération lourde, du fait de l'
étendue
du réseau, de sa complexité, de la variété des points sensibles à contrôler (supports, passages
de route, entrées en terre) et du caractère parfois parcellaire de la documentation; sa
réalisation nécessite la mise en place d' une structure de type projet avec des équipes
maintenance et inspection intégrées, en effectif suffisant pour :
INDEX çç Sommaire
ANNEXE 1 : Stratégie Inspection
ANNEXE 2 : Criticité environnementale
PS*DN I (Intensité)
<15 1
>15 2
>50 3
100-150 4
>150 5
G Conséquence
environnementale
0à4 A
5à9 B
10 à 14 C
15 à 19 D
20 à 25 E
INDEX çç Sommaire
ANNEXE 3 : Plan de coupe
géo
localisation
ANNEXE 4 : Types de support
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page59
RÉSUMÉ / ABSTRACT
3D
LES ULTRASONS MULTI-ÉLÉMENTS
PHASED ARRAY
RÉSUMÉ
Le contrôle non destructif par ultrasons multi-éléments est de plus en plus présent dans le domaine du
CND depuis quelques années. Le CETIM s’est rapidement impliqué dans cette technique, en
investissant dans deux types de systèmes :
Ces investissements ont permis aux CETIM d’évaluer cette technique et de proposer des axes
d’amélioration ou de remplacement des contrôles volumiques traditionnels. En effet, la possibilité de
traçabilité, le gain de temps et l’amélioration de la probabilité de détection de discontinuités offrent
une réelle plus-value sur les ultrasons conventionnels. De plus, cette technique combinée à la technique
du TOFD, permet dans certains cas, de se substituer à la radiographie.
Ces constatations sont le résultat de campagnes d’essais effectuées par le CETIM sur différents cas
industriels :
¢
Caractérisation de pertes d’épaisseur liée à l’érosion/corrosion
¢
Contrôle de soudures sur les équipements sous pression
¢ Recherche de fissure dans un arbre mécanique
INDEX çç Sommaire
Les ultrasons multiéléments
(phased array)
Une plus-value sur les techniques volumétriques
conventionnelles
F. BERTHELOT (CETIM)
B. DUPONT (CETIM)
R. MATHIEU (CETIM)
Introduction
La technique des ultrasons multiéléments est de plus en plus présente dans le domaine du
CND depuis quelques années. Cette technique est actuellement en cours de normalisation.
Ces actions menées depuis quelques années, ont permis au CETIM de positionner les
ultrasons multiéléments comme une nouvelle technique de contrôle volumétrique. En effet,
la possibilité de traçabilité, le gain de temps et l’amélioration de la probabilité de détection de
discontinuités offrent une réelle plus-value sur les ultrasons conventionnels. De plus, cette
technique combinée à la technique du TOFD permet, dans certains cas, de se substituer à la
radiographie.
Ces constatations sont le résultat de campagnes d’essais effectuées par le CETIM sur
différents cas industriels :
The UT phased array technique is more and more present in the field of NDT for some years.
This technique is now standardized.
CETIM, as a technical centre for mechanics, was early involved in UT phased array:
For some years, CETIM is developing UT phased array as a volumetric technique. Indeed,
the possibility of traceability, saving of time and increase of probability of detection, offer an
INDEX çç Sommaire
real improvement compare with conventional UT. In certain cases, the phased array
technique combined with TOFD may replace radiographic testing.
These observations are the result of tests achieved by CETIM on different industrial cases:
INDEX çç Sommaire
1. PRINCIPE
La technologie ultrason multiéléments repose sur les mêmes principes physiques que le
contrôle par ultrasons conventionnels.
La particularité est liée à l’utilisation d’un réseau de traducteurs multiéléments élémentaires
(électriquement et acoustiquement indépendants) pilotables indépendamment et séparément
les uns des autres.
Les traducteurs multiéléments sont pilotés par des appareils multiéléments ayant les 3
caractéristiques suivantes :
ˇ multivoies
ˇ déphaseurs
ˇ sommateurs
Elle permet donc, avec un seul traducteur multiéléments, de disposer virtuellement d’un grand
nombre de traducteurs mono-éléments.
Les appareils multiéléments permettent de définir les paramètres ultrasonores, de piloter
l’électronique et de visualiser et enregistrer les acquisitions.
Les traducteurs multiéléments sont définis selon l’agencement des éléments constituant le
réseau (par exemple linéaire, matriciel …). Les arrangements les plus courants sont les
traducteurs à réseau linéaire (fig 1), c’est-à-dire que les éléments sont agencés selon une ligne
droite et unique.
INDEX çç Sommaire
Comme les traducteurs classiques mono-éléments, les traducteurs multiéléments peuvent être
utilisés avec un sabot droit ou incliné. Il est à noter que le sabot peut être mis en forme pour
s’adapter à la géométrie de la pièce à contrôler et ainsi assurer un bon couplage du traducteur.
1.3 Lois de retards
ˇ combiner ces lois de retards pour dévier le faisceau suivant une direction donnée tout en
focalisant le faisceau en un point donné. L’association de ces deux lois de retards est
schématisée sur la figure 4.
INDEX çç Sommaire
1.4 Types de balayages électroniques utilisés
Multiséquentiel : Ce type de balayage comporte plus d’une séquence. Chaque séquence peut
être constituée d’un ou plusieurs tirs.
Le contrôle par ultrasons en multiéléments peut être mis en œuvre suivant deux types de
balayage différents :
INDEX çç Sommaire
2. CAS N°1 : Relevés de profil
- Fournir aux industriels une solution plus avantageuse que la méthode usuelle
couramment utilisée de manière industrielle, qui elle, réside en des relevés de points de
mesure d’épaisseur réalisés par quadrillage de quelques dizaines ou centaines de
millimètres d’espacement.
- Rendre possible la réalisation d’un relevé de profil sur des matériaux très pénalisant
pour les ultrasons (ex fonte). L’utilisation des UT multiéléments assure d’une manière
générale une meilleure détection du fond de la pièce.
La pièce examinée est une tôle (fig 7) de 300x300 sur laquelle des usures ou des saillies ont
été réalisées par meulage sur quelques millimètres de profondeur.
2.1.1 Matériels
Le balayage utilisé est de type linéaire par séquence de 8 éléments avec un pas de 1 élément.
Une focalisation est appliquée à 9 mm de profondeur, c'est-à-dire 1 mm avant le fond de la
pièce. Un filtrage à 5 MHz est appliqué au signal en réception.
L’acquisition des cartographies est faite par pas de 45mm, l’ouverture totale de la sonde étant
de 45,6 mm, il y a un recouvrement de 0,6 mm entre 2 cartographies (fig 8).
INDEX çç Sommaire
Y
recouvrement 0,6
45,6
0 X
2.1.3 Résultats
En sélectionnant les coordonnées à l’aide d’un curseur, on peut obtenir un relevé de profil
très précis dont un exemple est présenté sur la figure 9 (une dilatation de l’échelle est réalisée
sur le dernier millimètre) :
INDEX çç Sommaire
2.2 Exemple sur pièce fonte
Le contrôle des épaisseurs a été réalisé sur un équipement sous pression en fonte. L’objectif
était de vérifier que les épaisseurs présentes étaient supérieures à une valeur limite. Cette
mesure n’a pas pu être réalisée par ultrasons conventionnels. Seule la technique UT
multiéléments a permis la mesure, ainsi que l’acquisition de cartographies.
Le matériel utilisé est un système OLYMPUS, Omniscan MX16-128PR. On utilise une sonde
OLYMPUS 2L64, 2,25 MHz de fréquence, 64 éléments avec un pitch de 0,75 mm.
L’ouverture totale est 48 mm. Elle est montées sur un sabot 0° en rexolite de hauteur 20mm.
On effectue un balayage linéaire à 0° avec une séquence de 16 éléments par pas de 1 élément
ainsi qu’une focalisation à 42 mm de profondeur. La loi de retard appliquée sur une séquence
en émission et en réception est présentée sur la figure 11.
Le contrôle d’une soudure circulaire bout à bout de diamètre 171 mm et d’épaisseur 28,6 mm
peut être réalisé par les trois méthodes CND volumétriques différentes :
La mise en œuvre du contrôle de soudure par TOFD est depuis peu normalisé et des critères
d’acceptation sont proposés. Cette technique est très séduisante pour le contrôle de soudure
bout à bout. Cependant, certaines zones du volume à contrôler sont peuvent apparaître comme
des zones de faible ou de non détection. Il s’agit en général des zones très proches des
surfaces. En fonction du niveau d’examen demandé, il convient de compléter ce contrôle par
un autre CND.
INDEX çç Sommaire
3.1 Pièce contrôlée
Dans les deux cas, les sabots sont adaptés au rayon de courbure.
La zone de couverture est décrite sur la figure14. Le TOFD couvre une grande partie du
volume à contrôler. Les ultrasons multiéléments utilisés avec un balayage angulaire 40° - 65°
permettent la couverture des zones proches des surfaces, ainsi que l’ensemble du volume à
contrôler. Le contrôle est réalisé, pour les multiéléments, de part et d’autre du cordon de
soudure.
La pièce d’essai contient une soudure et des défauts de référence (entailles) sur les bords et au
centre du cordon de soudure circulaire. La définition de ces défauts est présentée sur la figure
15 et 16.
INDEX çç Sommaire
H (mm) W (mm) L (mm) location Position dans la
circonférence (mm)
Milieu de cordon 80
Entaille 1 1,6 1 25
Racine
Bord de cordon 285
Entaille 2 1,6 1 25
Surf sup
Bord de cordon 365
Entaille 3 1,6 1 25
Surf sup
Milieu de cordon 430
Entaille 4 1/3 T 1 25
Racine
H : hauteur de l’entaille
W : largeur de l’entaille
L : longueur de l’entaille
T : épaisseur de la pièce (28,6 mm)
3
3
2 4 1 4
INDEX çç Sommaire
3.5 Résultats des contrôles
La paire de traducteurs TOFD permet l’acquisition d’une seule cartographie B-SCAN, alors
que chaque sonde ultrasons multiéléments donne lieu à l’acquisition de 25 cartographies D-
SCAN. On peut aisément reconstruire un C-Scan pour chaque sonde. Les deux vues
présentées sont donc des C-SCANS.
Les cartographies (figure 17) obtenues montrent que toutes les entailles sont détectées et
caractérisées correctement. La cartographie TOFD (niveau de gris) ne met en évidence qu’une
seule entaille (4).
Les UT multiéléments sont ici indispensables pour la détection des défauts proches des
surfaces (entailles 1,2 et 3).
Le contrôle permet une très bonne reproductibilité des résultats, un enregistrement numérique
des cartographies et un gain de temps puisque l’acquisition des 3 cartographies se fait en un
seul passage, c'est-à-dire en un seul un tour de tube.
INDEX çç Sommaire
4. CONCLUSION
Les ultrasons multiéléments apportent une réelle plus-value sur les technique volumétriques
usuelles par
ˇ La rapidité du contrôle
ˇ Un gain dans la probabilité de détection des défauts
ˇ Un enregistrement et une traçabilité des résultats
Cette technique permet dans certains cas de remplacer la radiographie, et dans d’autre de
remplacer les ultrasons. Elle permet aussi pour certaines applications, de rendre le contrôle
possible.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page60
RÉSUMÉ / ABSTRACT
46D
CONTRÔLE DES CANALISATIONS PAR DIFFERENTES METHODES
ULTRASONORES COMPLEMENTAIRES
RÉSUMÉ
La technique des Ondes Ultrasonores Guidées (GUL) permet une scrutation rapide des canalisations
telles que les pipelines de liquides ou de gaz ainsi que les tuyauteries des sites industriels. Elle permet
une détection et une localisation des zones de corrosion. De récents développements ont permis d’amé-
liorer les capacités de détection, l’interprétation des données et l’évaluation semi-quantitative de l’en-
Mercredi
dommagement par corrosion. Elle demande une formation et une grande expérience des opérateurs.
Un des points critiques concerne la zone de contact entre les canalisations et les supports. Ces zones
ne sont pas accessibles pour l’inspection visuelle et les mesures d’épaisseur par ultrasons, elles sont
très souvent exclues du plan de contrôle, le levage de la conduite étant la seule solution, alors qu’elles
sont des zones à risques. Une technique particulière (TPC), basée sur la génération d’ondes ultrasonores,
a été développée afin de détecter et d’évaluer la perte de matière dans cette zone. Elle est applicable
seule pour le contrôle des contacts aux supports et/ou en complément des ondes guidées lorsque cette
technique nécessite un complément de mesure pour améliorer l’interprétation. Cette communication
présente des résultats obtenus sur différentes canalisations avec ces deux techniques dans le cadre de
plan de contrôle afin de prévenir des risques liés au vieillissement.
ABSTRACT
Guided Ultrasonic waves (GUL) is a rapid screening technique of pipes such as liquid and gas pipelines
and general piping systems in industrial plant facilities. It allows detection and a localization of the
corroded areas. Recent developments have greatly enhanced the detection capability, the interpretation
of data and a semi quantitative evaluation of the damage by corrosion. A high level of training and ex-
perience is needed for the operators. One of the critical point is the contact between the support and
the pipe. These zones are not accessible for visual inspection and UT thickness measurements, they are
often excluded of the control plan, and the solution is to lift up the pipe, while they are potential area
of critical damage. A special ultrasonic technique was developed to detect and estimate wall thickness
loss in this area. This technique is applicable alone or in complementary of the guided waves when
more data are needed to improve the diagnosis. This paper shows results obtained by both techniques
in different situations, during the application of inspection plan to prevent the ageing damage.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page61
RÉSUMÉ / ABSTRACT
72D
RETOUR D’EXPÉRIENCE DE LA MÉTHODE END PAR ONDES GUIDÉES
ULTRASONORES AU CONTRÔLE DES TUYAUTERIES
C. Laenen – J. Leherpeur
CETE Apave Nord-Ouest - France
RÉSUMÉ
La corrosion des tuyauteries et des canalisations de transport est un problème majeur pour les secteurs
de la pétrochimie, de l’énergie et d’autres industries. Des exigences réglementaires récentes imposent
le suivi périodique des tuyauteries pour atteindre les objectifs de sécurité. Fréquemment, les tubes sont
calorifugés ne permettant de détecter la corrosion externe sans enlèvement du calorifuge, occasionnant
ainsi des coûts élevés. Des situations particulièrement difficiles sont rencontrées pour l’inspection des
parties non accessibles, comme les traversées de route en partie enterrées et pour les tuyauteries placées
en fourreau ou caniveau où le contrôle ne peut être effectué sans excavation. Le contrôle au moyen des
ondes ultrasonores guidées est une solution attractive pour contribuer à l’inspection des tubes à partir
d’un seul point accessible de la tuyauterie. Les ondes ultrasonores en se propageant sur une longue
distance permettent par scrutation volumétrique la détection de la corrosion dans toute la section. Cette
technique est dorénavant applicable sur chantier, cet article présente le retour d’expérience d’un
prestataire de service sur la mise en œuvre de cette méthode de scrutation des tuyauteries utilisant la
technologie piézoélectrique du système GWUT Wavemaker ⎪ GUL G3.
ABSTRACT
Corrosion in process pipes and pipelines is a major problem in the petrochemical, utilities plants and
other industries. New regulations request the inspection surveillance of piping in order to gain safety
.Frequently pipes are insulated which means that even external corrosion cannot be seen without
removing the insulation, which can be prohibitively expensive. Particularly severe problems are
encountered at road crossings or sleeved pipes entrance where the pipe cannot be inspected without
major excavation. The Guided Waves Ultrasonic Technique provides an attractive solution to this
problem because they can be excited at one location on the pipe and will propagate many metres along
the pipe allowing screening corrosion detection. The technique is now applied on site, this paper
discusses operational return of practical experience from a service company using the piezoelectric
pipe screening instrument GWUT Wavemaker ⎪ GUL G3.
INDEX çç Sommaire
RETOUR D’EXPERIENCE DE LA METHODE END PAR ONDES
GUIDEES ULTRASONORES AU CONTROLE DES TUYAUTERIES
PRACTICAL CONSIDERATION FOR GUIDED WAVES ULTRASONIC TESTING NDT FOR PIPING
INSPECTION
C. Laenen – J.Leherpeur
CETE Apave Nord-Ouest - France
SOMMAIRE
La corrosion des tuyauteries et des canalisations de transport est un problème majeur pour les
secteurs de la pétrochimie, de l’énergie et d’autres industries. Des exigences réglementaires
récentes imposent le suivi périodique des tuyauteries pour atteindre les objectifs de sécurité.
Fréquemment, les tubes sont calorifugés ne permettant de détecter la corrosion externe sans
enlèvement du calorifuge, occasionnant ainsi des coûts élevés. Des situations particulièrement
difficiles sont rencontrées pour l’inspection des parties non accessibles, comme les traversées de
route en partie enterrées et pour les tuyauteries placées en fourreau ou caniveau où le contrôle ne
peut être effectué sans excavation. Le contrôle au moyen des ondes ultrasonores guidées est une
solution attractive pour contribuer à l’inspection des tubes à partir d’un seul point accessible de la
tuyauterie. Les ondes ultrasonores en se propageant sur une longue distance permettent par
scrutation volumétrique la détection de la corrosion dans toute la section. Cette technique est
dorénavant applicable sur chantier, cet article présente le retour d’expérience d’un prestataire de
service sur la mise en œuvre de cette méthode de scrutation des tuyauteries utilisant la
technologie piézoélectrique du système GWUT Wavemaker GUL G3.
ABSTRACT
Corrosion in process pipes and pipelines is a major problem in the petrochemical, utilities plants
and other industries. New regulations request the inspection surveillance of piping in order to
gain safety .Frequently pipes are insulated which means that even external corrosion cannot be
seen without removing the insulation, which can be prohibitively expensive. Particularly severe
problems are encountered at road crossings or sleeved pipes entrance where the pipe cannot be
inspected without major excavation. The Guided Waves Ultrasonic Technique provides an
attractive solution to this problem because they can be excited at one location on the pipe and
will propagate many metres along the pipe allowing screening corrosion detection. The
technique is now applied on site, this paper discusses operational return of practical experience
from a service company using the piezoelectric pipe screening instrument GWUT Wavemaker
GUL G3.
INDEX çç Sommaire
1 - INTRODUCTION
La méthode des Ondes Guidées (Guided Waves Ultrasonic Testing GWUT) est un examen non
destructif utilisé pour la détection de la corrosion des tuyauteries d’usines [1]. Elle utilise des
ondes ultrasonores propagées le long des tubes dans une direction axiale pour mesurer les
variations de la section tubulaire. Les ondes guidées ne permettent pas la mesure directe de
l’épaisseur ou la perte de métal par la corrosion mais assurent des possibilités de détection et de
localisation par scrutation globale de grande longueur de tuyauterie à partir d’un seul point
d’accès. Cette méthode est maintenant largement utilisée pour les inspections en service [2]. En
particulier le challenge de l’inspection des tuyauteries calorifugées sans dépose majeure du
calorifuge et des parties difficiles d’accès ou les autres techniques sont invisibles étend sa
demande. La technique est actuellement particulièrement appréciée pour l’inspection des
tuyauteries en traversée de murs, de la partie enterrée des traversées de routes et pour la
surveillance des pipelines en entrée de terre ou à partir de fouilles espacées. Comme toute
nouvelle méthode, il est normal d’évaluer les performances de détection et la fiabilité obtenue en
conditions réelles de terrain. Ce document indique le retour d’expérience observé sur des
applications sur site industriel d’un prestataire de service. L’expérience acquise permet ainsi de
mieux connaitre les limites du système et les configurations de tuyauteries considérées les plus
adaptées à la mise en œuvre de la technique.
La technique doit être considérée comme un outil de détection par scrutation des zones de
corrosion pouvant exister sur une tuyauterie ou canalisation. Une grande longueur des tubes et
accessoires est alors inspectée rapidement dans son volume de métal et sur toute l’épaisseur de
paroi pour identifier et positionner les problèmes potentiels. Les positions des zones de corrosion
généralisée et localisée sont déterminées avec précision à partir du point unique de mesure où est
installé l’élément capteur.
Le système fonctionne en émettant des ondes de basses fréquences (en général de 15 – 70 kHz) se
propageant sur la longueur de la tuyauterie. Les variations de la section du tube conduisent à une
réflexion des ondes ultrasonores, ces réflexions sont analysées pour déterminer la position des
anomalies et estimer les dimensions des événements ayant causé ces variations. La technique ne
permet pas directement de mesurer l’épaisseur résiduelle mais de relever des amincissements par
estimation de la variation de section que la corrosion soit interne ou externe ou combinée.
Lorsque les zones où des problèmes sont identifiés, une inspection complémentaire par les
techniques conventionnelles est ensuite réalisée pour quantifier avec précision les dimensions des
anomalies. La technique est idéale pour inspecter des grandes longueurs de tuyauteries (> 10
mètres dans chaque direction).
A chaque changement de la section de la tuyauterie, il y a une réflexion des ondes guidées.
L’amplitude de la réflexion dépend de la variation de section à chaque point d’acquisition du
signal le long du tube. Les soudures, les brides sont des réflecteurs importants retournant vers
l’anneau capteur et sont visibles sur toute la circonférence. Une soudure bout à bout présente en
général une surépaisseur du cordon en passe de finition et de pénétration constituant une variation
positive de la section, une partie de l’énergie ultrasonore est alors réfléchie mais continue à être
transmise en subissant une atténuation exponentielle de par la distance. L’atténuation est
maitrisée par une courbe de décroissance distance-amplitude (DAC).Les réflexions sont
INDEX çç Sommaire
classifiées en mode axialement symétrique. Une corrosion localisée avec perte de l’épaisseur de
métal est une situation où des ondes non axialement symétriques seront formées. Elles se
propageront en retour vers le capteur pour être analysées conjointement avec le mode symétrique.
Cette conversion de mode non symétrique est un élément clé pour la stratégie d’identification des
anomalies.
La propagation des ondes en mode de torsion le long du tube conduit donc aux points de
variations de la section plutôt que seulement par une perte de l’épaisseur à des réflexions du
signal. Les études ont démontrées que le rapport entre la perte de section transversale due à un
défaut et l’amplitude de la réflexion est consistant. La forme et la position circonférentielle ainsi
que la position dans l’épaisseur n’affecte pas de façon notable l’amplitude de la réflexion. La
longueur axiale du défaut à un impact sur le signal en fonction de longueur d’onde et de son
étendue mais peut être analysé en variant la fréquence.
3 – SYSTEME DE CONTROLE
Le système choisi par le CETE Apave Nord Ouest est l’instrument commercialisé par Guided
Ultrasonics Ltd, modèle Wavemaker G3 [3] avec la fonction avancée de focalisation EFC
L’instrument électronique fonctionnant sur batterie est connecté avec des câbles de liaison sur un
anneau comportant les palpeurs piézoélectriques multiéléments, enroulé autour de la tuyauterie
(Figure 1). L’acquisition de la mesure est effectuée à partir d’un ordinateur portable relié par un
câble de type USB. A chaque diamètre nominal des tubes correspond un anneau capteur
spécifique. Il existe des anneaux rigides pour les tubes de petits diamètres et pour les diamètres
allant de 6’’ à 36’’voir plus, des anneaux souples raccordables instrumentés avec la population
de capteurs. Un montage par ressort ou par une chambre mise en pression assure le bon contact et
le couplage de chacun des capteurs sur la surface métallique.
L’arrangement des modules contenant les capteurs émetteur-récepteur est organisé en fonction du
diamètre et de la longueur souhaitée pour les investigations. La configuration permet d’ajuster le
régime des fréquences. Par exemple pour un diamètre de tube de 8’’ (219 mm), 24 modules de 2
rangées de capteurs piézoélectriques seront installés autour du tube.
INDEX çç Sommaire
Les ondes guidées sont envoyées de part et d’autre de l’anneau se propageant ainsi dans les 2
directions à partir du point d’installation (Figure 2). Une zone de longueur environ 500 mm est
requise pour la mise en place de l’anneau capteur. Aucune préparation particulière de la surface
n’est nécessaire sauf l’enlèvement des revêtements épais et le brossage des oxydes ou les écailles
de peinture. Un tube peint peut être examiné directement. L’enlèvement du calorifuge est limité
au seul point d’installation de l’anneau. La zone située directement sous l’anneau n’est pas
incluse dans l’essai. Les anneaux rigides sont utilisables jusqu'à une température de 120°C et les
anneaux souples jusqu'à 70°C permettant le contrôle des tuyauteries en service dans ces
conditions.
4 – RETOUR D’EXPERIENCE
Le système utilisé est un outil END particulièrement intéressant pour sa rapidité d’installation et
les possibilités d’inspecter plusieurs mètres voir dizaines de mètres de la tuyauterie depuis une
seule position. Il faut cependant bien prendre en compte les limites de la technologie car plusieurs
conditions existantes peuvent limiter la distance d’inspection et introduire des perturbations
affectant le signal. Ces paramètres sont par exemples : la configuration géométrique des lignes, les
accessoires comme les piquages, supports, les revêtements de surface, l’oxydation superficielle, les
dépôts, les vibrations et la viscosité des fluides présents dans les tuyauteries.
La technique est fréquemment mise en œuvre pour inspecter des parties difficiles d’accès des
réseaux de tuyauterie comme les parties pénétrantes dans les murs en béton, en passage dans les
fourreaux, les traversées enterrées sous route et les tuyauteries aériennes implantées en rack.
Les conditions peuvent être classifiées idéales, bonnes ou mauvaises pour réaliser l’examen.
Dans certains cas, des réglages particuliers sont nécessaires pour les situations difficiles et pour
étendre la longueur du contrôle en faisant appel à l’expérience de l’opérateur. Un niveau de
confiance des résultats est évalué et indiqué dans le rapport d’examen.
Nous présentons ici quelques paramètres influant sur la qualité de l’inspection illustrés par le
retour d’expérience de cas rencontrés en situation de chantier.
INDEX çç Sommaire
4.1 – LONGUEUR INSPECTEE
Il est difficile de prédire la longueur inspectée pour un essai sans connaître la tuyauterie, car la
géométrie de la ligne, l’atténuation du signal sont des paramètres importants. La longueur
inspectée interprétable est déterminée lorsque l’acquisition vient d’être réalisée. La détection des
anomalies est principalement gouvernée par le ratio du niveau du signal et du bruit de fond
(S/N).Le bruit de fond est causé par 2 mécanismes possibles. Le premier provient de
l’environnement externe comme les vibrations et les perturbations électriques de l’électronique
appelé incohérent, l’amplitude du signal ne décroissant pas avec la distance. Le second
correspond aux réflexions de discontinuités le long du tube comme l’oxydation en peau, les
piqures de corrosion ou les variations de l’épaisseur des tolérances de fabrication. Ce bruit de
fond cohérent décroit avec la distance depuis l’anneau capteurs selon le niveau d’atténuation. En
pratique on considère que le ratio S/N doit être suffisamment grand pour faciliter l’identification
et l’interprétation de petites variations de section ou des anomalies de petites dimensions. La
règle commune est d’avoir un écho de défaut au moins de 6dB au dessus du niveau du bruit de
fond pour obtenir une bonne sensibilité de détection. La position de fin de test déterminée par
l’opérateur est choisie lorsque le niveau de bruit de fond est de moitié la hauteur entre la base et
la courbe d’appel de limite des anomalies réglée à 7 % de variation de section (ECL ou perte de
section estimée).
Les figures 3 et 4 montrent des extrêmes de longueurs exploitables en fonction du S/N. Lorsque
que l’atténuation est limitée et le bruit de fond faible, la propagation est étendue comme visible à
gauche avec une couverture de plusieurs dizaines de mètres de longueur. Il apparait comme
présenté à droite sur un tronçon d’une ligne fortement corrodée en surface ou endommagée par
une corrosion généralisée interne (comme les réseaux d’eau d’incendie) la propagation est réduite
à quelques mètres. La forte atténuation et le bruit de fond notable sont des paramètres
préjudiciables pour la méthode.
Corrosion généralisée
atténuation
seuil détection
Bruit de fond
Figure 3 : Propagation sur plus de 80 mètres de Figure 4 : Propagation réduite due à l’atténuation
longueur en contrôle (Affichage A-Scan) par une corrosion généralisée uniforme
INDEX çç Sommaire
longueur correspondant à la partie exploitable de l’essai. Une sensibilité de l’ordre de 5 % au
minimum doit toujours être recherchée au point le plus éloigné de l’anneau capteur. Une courbe
d’appel correspondant à une variation de section de 7 % est typiquement choisie en référence à
l’amplitude des soudures dont les dimensions estimées correspondent à une variation de section
connue théorique ou réelle. Le seuil de détection est placé au-dessus du bruit de fond soit autour
de 3 % pour les parties droites de la ligne. L’opérateur va alors rechercher les signaux des
réponses symétriques et non-symétriques dont l’amplitude est au dessus du bruit de fond et entre
la mi- hauteur, tangent ou au dessus de courbe appel pour classifier les anomalies.
Etendue de la corrosion
Des plages étendues de corrosion peuvent cependant conduire à sous estimer la classification de
la profondeur des anomalies. Le suivi et la caractérisation des anomalies détectées par des
moyens conventionnels permettent alors une expertise dimensionnelle précise, ce qui est
d’ailleurs une action requise par la procédure.
4.2 – FREQUENCE
INDEX çç Sommaire
d’analyser postérieurement à l’acquisition une analyse des données dans une gamme étendue de
fréquence et non pas sur une fréquence unique. Le système utilise en remplacement d’une
fréquence conventionnelle un nombre normalisant la longueur d’onde au diamètre du tube.
L’affichage à la fréquence élevée va pouvoir donner une sensibilité élevée mais avec une
longueur de propagation réduite. Vers la basse fréquence la sensibilité sera réduite pour offrir
une propagation plus étendue. L’opérateur devra rechercher un compromis pour les essais. Un
autre intérêt est la dépendance de certains événements de varier en amplitude le signal selon la
fréquence et ainsi d’assister à la classification des réflecteurs comme pouvoir séparer une
corrosion d’un support simple. La bande passante peut également être ajustée afin d’améliorer la
résolution.
L’exemple présenté sur la figure 6 correspond à une application sur une tuyauterie verticale en
unité de raffinage. Une zone de corrosion a été relevée au droit d’un petit piquage d’injection
d’eau de lavage. L’utilisation de l’outil informatique permettant l’affichage du signal sur une
gamme de fréquence étendue à permis de séparer dans une zone contenant plusieurs événements
comme l’orifice du piquage, la bride de la zone de corrosion.
Corrosion
Brid
e Soudure
Orifice piquage
Les réflecteurs les plus éloignés de l’anneau seront de plus faible amplitude que ceux proches en
raison de l’atténuation du signal même si ils ont les mêmes caractéristiques (soudures par exemple)
Les soudures sont donc intéressantes pour régler sur une acquisition la courbe de correction
distance amplitude DAC (Figure 7). Les soudures représentent environ 22% de variation de
section. Si les dimensions du cordon en largeur et hauteur sont visibles on peut affiner cette
variation par défaut et ainsi placer la courbe d’appel des anomalies avec précision [4].
INDEX çç Sommaire
Figure 7 : Tracé de la courbe d’atténuation sur une longueur de 60 mètres
Les soudures symétriques sont représentées sur l’affichage en déroulé
Une anomalie de corrosion est relevée au début de l’acquisition
La DAC va prendre en compte l’atténuation par les revêtements, les dimensions des éléments de la
tuyauterie, la taille des soudures selon le procédé de soudage.
Il apparait donc très important de pouvoir afficher plusieurs réflecteurs comme les soudures pour
régler correctement la décroissance et la pente de la courbe d’atténuation. Cette ligne de référence
va permettre d’évaluer les anomalies à tout point de la distance depuis l’anneau. Il est donc
nécessaire d’obtenir sur l’affichage au minimum 2 réflecteurs de type soudure. Si la DAC n’est pas
correctement ajustée le seuil de détection fixé sera incorrect. Sans réflecteur de référence il est
impossible d’analyser le signal en sensibilité. Il reste possible de faire appel à d’autres moyens de
réglages avec des réflecteurs artificiels.
Dans la zone immédiatement proche de l’anneau et ceci de part et d’autre les capteurs recevant et
transmettant les signaux sont affectés dans une zone perturbée (zone de champ proche). Directement
sous l’anneau la zone est invisible et devient une zone morte. Pour couvrir la zone perturbée il faut
alors faire une acquisition supplémentaire décalée. La longueur de la zone perturbée (Figure 8)
dépend de la fréquence d’excitation et de la bande passante lors de l’acquisition (de quelques
dizaines de cm à moins d’un mètre). Dans certains cas d’examens comme pour les traversées de
murs où les anomalies sont souvent rencontrées a l’interface, il est très important alors de pouvoir
avoir suffisamment de recul par rapport à la zone en inspection pour échapper à la zone perturbée ou
l’amplitude des signaux est affectée.
Plusieurs situations ont été rencontrées sur le terrain où la distance (D) de mise en place de l’anneau
ne permettait pas de s’éloigner suffisamment par manque de place de la zone souhaitée en examen
comme des entrées en mur (Figure 9: tuyauterie en fosse, présence de supports soudés, gros
INDEX çç Sommaire
piquages…). Les anomalies sont détectables mais la classification en amplitude est affectée.
Anneau
Champ proche Zone morte
Anomalie
D Tube
Figure 8 : Zone morte et champ proche Figure 9 : Entrée en mur pour une inspection GWUT
4.5 – REVETEMENTS
Les revêtements selon le type ont une influence directe sur l’atténuation du signal affectant la
longueur de l’examen. La raison est le transfert et la perte de l’énergie dans le matériau solide
entourant la tuyauterie. Les revêtements de type bitumeux sont particulièrement atténuants (environ
3dB/m) pouvant limiter la propagation des ondes guidées peut être limitée sur seulement quelques
mètres. Le ratio de l’épaisseur du revêtement avec l’épaisseur du tube est également un facteur
prépondérant. Les revêtements de type polyéthylène sont plus modérés.
Des plages étendues de corrosion peuvent cependant conduire à sous estimer la classification de la
profondeur des anomalies. L’expertise postérieure des anomalies détectées par ondes guidées doit
faire l’objet d’examens conventionnels pour caractériser les dimensions des anomalies, ce qui est
d’ailleurs une action requise par la procédure.
La sévérité de la corrosion peut être déterminée selon une procédure de classification prenant en
INDEX çç Sommaire
compte l’amplitude des signaux symétriques et non symétriques en référence à la courbe DAC. La
perte de section est exprimée en % avec la possibilité d’estimer la répartition circonférentielle des
anomalies (Figures 10-11). Il s’agit aujourd’hui d’une approche semi-quantitative car l’information
obtenue ne permet cependant pas de calculer directement par une entrée de données la durée de vie.
Dans le cas des parties de tuyauteries cachées comme les traversées de route enterrées ou en
fourreaux, le coût de confirmation des anomalies est élevé. La méthode GWUT peut permettre de
faire un tri des lignes sans - avec signaux d’anomalies comme une pratique Go – No Go. Un suivi
des anomalies par la mise en place d’un capteur fixe est possible.
4.7 – SUPPORTS
Les supports constituent des réflecteurs d’amplitude modérée ou élevée selon les types. La
distinction entre la réflexion du support et une corrosion est facilitée par une analyse du
comportement du signal selon plusieurs fréquences. Les supports soudés forment cependant des
réflexions d’amplitudes élevées rendant difficile la détection d’anomalies localisées. De plus les
supports dont les cordons de soudures sont longs forment des échos de traine perturbant la partie
tubulaire située derrière le support. Pour cette raison la longueur de contrôle est affectée par le
rapprochement des supports soudés. Le choix d’une fréquence élevée permet de mieux passer sur
les supports et de pouvoir distinguer les corrosions. L’affichage en mode surfacique permet de
corréler la position des anomalies et points de contact des supports. L’expérience sur site pour le
contrôle de grande longueur de tuyauterie mais également l’examen des petites zones de contact de
quelques centimètres avec les supports a montré les possibilités et limites de la méthode. La
dépendance des réflexions est fonction de plusieurs paramètres comme la pression de contact des
supports avec la tuyauterie, de la longueur et répartition des soudures sur les plats, de la corrosion et
produits de corrosion. Des procédures couvrent ces applications spécifiques (Figures 12-13).
Figure 12 : Ligne calorifugée Ø 4’’ Figure 13 : Corrosion sévère détectée aux supports
4.6 – ARTEFACTS
Des faux échos sont la cause d’anomalies détectées ne correspondant pas à des anomalies de
corrosion. Certains effets reconnaissables de type réverbération et miroirs sont rencontrés mais
certains peuvent être difficile à classifier et doivent être alors considérés comme anomalie une fois
détectée restant à analyser par moyens conventionnels. L’expérience de l’opérateur est bénéfique
INDEX çç Sommaire
pour identifier les fausses indications. Il est requis que l’opérateur spécialiste soit présent sur le site
pour assurer une pré-analyse à la suite de chacune des acquisitions.
5 - CONCLUSIONS
La technique non destructive avancée GWUT utilisant les ondes guidées est maintenant un outil
industriel permettant la détection par scrutation de la corrosion sur les tuyauteries d’usines. La
rapidité de mise en œuvre permet d’inspecter de grandes longueurs de tuyauteries dans une seule
journée.
Le contrôle au moyen des ondes ultrasonores guidées devient une solution attractive pour
contribuer à l’inspection des tubes à partir d’un seul point accessible de la tuyauterie comme les
traversées de route en partie enterrées sont rencontrées et pour les tuyauteries placées en hauteur sur
rack, en fourreau et en caniveau, le contrôle peut alors être effectué. Même si la méthode reste une
approche semi-quantitative elle peut assurer la détection des anomalies sur des zones non
contrôlables par d’autres techniques.
L’expérience acquise permet ainsi de mieux connaitre les limites du système et les configurations de
tuyauteries considérées les plus adaptées à la mise en œuvre de la technique.
REFERENCES
[1] Alleyne, D.N., Pavlakovic, B., Lowe, M.J.S. and Cawley, P. (2001) 'Rapid long-range
inspection of chemical plant pipework using guided waves', Insight, Vol 43, pp93-96,101.
[2]. C.Laenen, T.Vogt, Contribution de la technique des ondes guidées ultrasonores (GWUT) à la
surveillance des tuyauteries d’usine, Conférence COFREND Toulouse 2008
[4] Reliability of Guided Wave Testing Thomas VOGT, Mark EVANS, Guided Ultrasonics Ltd,
London, UK
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page62
30/33B Exigences pour les tôles d’acier allié CrMo(V) destinées à des DETEMPLE I. KRAUT R.
applications dans des réacteurs pétrochimiques / Demands for DILLINGER HUTTE GTS
alloy CrMo(V) steel plates in petrochemical reactor applications
42B Perspectives offertes par un nouvel alliage de nickel à forte HOULLE P. / MECK S.
teneur en molybdène et chrome / Shell courses made of HAYNES
Mercredi
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page63
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page64
RÉSUMÉ / ABSTRACT
30/33B
EXIGENCES POUR LES TÔLES D’ACIER ALLIÉ CrMo(V) DESTINÉES A
DES APPLICATIONS DANS DES RÉACTEURS PÉTROCHIMIQUES
I. Detemple - R. Kraut
Dillinger Hütte GTS – postfach 1580 – 66748 Dillingen/Saar
[email protected]
RÉSUMÉ
L’efficacité des processus et l’optimisation des coûts sont généralement des facteurs importants pour les
opérateurs s’installations. En conséquence de quoi des températures de fonctionnement et des pressions
de plus en plus élevées conduisent à des réacteurs toujours plus grands, avec des parois très épaisses. En
Mercredi
outre, les exigences élevées de sécurité (conduisant à des spécifications de matériaux exigeantes) des
opérateurs, lesquels veulent garantir une productivité élevée au cours du cycle de vie d’une installation,
contribuent au fait que les matériaux de base pour les composants critiques des grandes installations ne
peuvent être produits que par des sidérurgistes de renommée mondiale. En particulier au cours des deux
dernières décennies, les producteurs de tôles en acier CrMo(V) ont mis au point des produits améliorés
destinés à répondre aux exigences des réacteurs modernes.
Cet exposé montrera comment les producteurs d’acier ont suivi les exigences des opérateurs et leur besoins
en aciers de plus en plus sophistiqués. Il sera également montré comment les différentes impositions, qui
interfèrent partiellement entre elles, peuvent être respectées et comment elles peuvent conduire aux limites
de la faisabilité technologique, particulièrement quand elles sont spécifiées simultanément. A ce propos,
des critères tels que le comportement en traction après un fort traitement de détente après soudage, l’état
de livraison, la ténacité, la dureté la composition chimique l’épaisseur et avant tout leur interdépendance
seront des points clés à prendre en compte lors du processus de production à l’aciérie ainsi qu’au laminoir.
ABSTRACT
Process efficiency and cost optimisation are generally important issues for plant operators. As a result
increasing process temperatures and pressures are leading to ever bigger reactors with high wall thickness.
Additionally, the high safety demands (resulting in demanding material specifications) of the operators,
who want to ensure high productivity over the life cycle of a plant, contribute to the effect that the base
material for these core components of major plants can only be produced by a few world-wide leading
steel manufacturers. Especially over the last two decades the plate producers for CrMo(V) steels made
enhanced products fulfilling the requirements of modern reactors.
This paper will outline how steel producers followed the demands of the operators and their needs for
more and more sophisticated steels. It will also be shown how the different requirements, partly interfering
each other, can be met and where they will lead to the edges of technological feasibility, especially when
specified all together. In his regard, criterions like tensile behaviour after strong PWHT, delivery condition,
toughness, hardness, chemical composition, thickness and foremost their interdependencies will be
highlighted in the production process of the steel plant as well as the rolling mill.
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page65
RÉSUMÉ / ABSTRACT
42B
PERSPECTIVES OFFERTES PAR UN NOUVEL ALLIAGE
DE NICKEL À FORTE TENEUR EN MOLYBDÈNE ET CHROME
Dr P. Houlle, Dr S. Meck
RÉSUMÉ
Ce nouvel alliage a été développé pour pallier les inconvénients des alliages Ni-Mo (type B). Il est en
effet beaucoup plus tolérant à la présence d’éléments oxydants. Par ailleurs, cet alliage est de manière
significative plus résistant que tous les alliages Ni-Cr-Mo (type C) dans des milieux très utilisés comme
les acides chlorhydrique et sulfurique. Il possède aussi une remarquable résistance aux corrosions
localisées piqures, corrosion caverneuse et fissuration sous contrainte. Sa mise en œuvre est semblable
à celle des alliages Ni-Cr-Mo.
INDEX çç Sommaire
INTRODUCTION
Les alliages Ni-Cr-Mo sont très connus pour apporter des solutions dans de nombreux milieux
corrosifs comme les milieux sulfuriques, les milieux acides chlorhydriques, à moyenne
température et les acides organiques. De plus, ils sont largement résistants à la formation de
piqures et de cavernes en présence d’halogénures principalement chlorures, et ils offrent une
bonne résistance à la fissuration sous contrainte dans ces mêmes milieux [1-4]. Ces alliages base
nickel contiennent en général un maximum de molybdène de16% avec une teneur de chrome de
16% à 23% selon les alliages. Ils sont parfois identifiés comme «type-C», on peut citer les
alliages C-2000®, C-22®, C-276, 59 et 686. Toutefois, dans certains cas leurs résistances
n’apportent pas la solution recherchée par l’utilisateur, ce qui oblige à considérer d’autres
solutions : les alliages base nickel à forte teneur en molybdène «type-B», comme par exemple
l’alliage B-3®, sont alors parfois utilisés. Bien que leur résistance à l’acide sulfurique, par
exemple, soit très remarquable [5,6], ils présentent certaines limitations dues à la quasi absence
de chrome dans leurs compositions: les alliages Ni-Mo ne résistent pas en présence d’éléments
oxydants (oxygène dissous, ions ferriques, cuivriques…) [7-11]. C’est pourquoi un nouvel
alliage a été développé afin de pallier cet inconvénient. Ce nouveau matériau, commercialement
appelé HASTELLOY HYBRID-BC1, peut être identifié comme appartenant au système des
alliages Ni-Mo-Cr. La présence substantielle de chrome (15%) permet d’assurer la formation
d’une couche passivante en présence d’éléments oxydants, mais la haute teneur en molybdène
(23%) lui confère une excellente résistance en milieu acide sulfurique, mais aussi en milieu
acide chlorhydrique De plus, sa stabilité thermique et sa soudabilité. sont excellentes, En
conséquence, sa mise en œuvre, reste conforme à celles des alliages Ni-Cr-Mo.
RESISTANCE A LA CORROSION
Matériaux testés
INDEX çç Sommaire
Alliage Famille Ni Mo Cr Fe W Mn Si C Autres
Conditions expérimentales
L’alliage HYBRID-BC1 a été testé sous forme d’échantillons de tôles de dimensions 25,4 mm
x 25,4 mm x 3,2 mm. Des tests d’immersion ont été réalisés dans différentes milieux
caractéristiques. La vitesse de corrosion mesurée pour chaque condition (température,
concentration) est la moyenne obtenue sur deux spécimens immergés. Les courbes d’iso-
corrosion de 0,1 mm/an et 0,5 mm/an sont réalisées avec une incertitude de ± 7 °C sur les
valeurs de température. Chaque test d’immersion a une durée de 4 fois 24 heures avec une
interruption toutes les 24 heures pour procéder à une pesée. Les ions ferriques ont été ajoutés,
lorsque nécessaires sous forme de sulfates ferriques. Pour caractériser la résistance de l’alliage
HYBRID-BC1 aux corrosions localisées induites par les chlorures (piqures et corrosion
caverneuse), les tests standards ASTM G-48 méthode C et D [12] ont été utilisés. Ces méthodes
utilisent une solution à 6% de chlorure ferrique acidifiée par une solution à 1% d’acide
chlorhydrique. A partir de 20°C, la température est augmentée par incrément de 5 °C jusqu’à
obtenir, pour chaque alliage, la température la plus basse qui initie en 72 heures, des crevasses
ou des piqures d’une profondeur minimum de 0,025 µm selon conformément à la spécification.
Ces températures sont identifiées comme étant la température critique d’apparition de crevasses
(CCT) et la température critique d’apparition de piqures (CPT). Les tests sont réalisés dans des
systèmes à reflux pour les températures inférieures à l’ébullition sous pression atmosphérique,
et dans des autoclaves en alliage C-276 pour les températures supérieures au point d’ébullition
Les échantillons utilisés pour les tests de piqures sont identiques aux tests d’immersion. Les
échantillons pour les tests de corrosion caverneuse sont des pièces de tôles 50 mm x 25,4 mm x
3,2 mm, les sites de crevasses sont créés par présence d’une rondelle rainurée en PTFE fixée sur
l’échantillon.
Les tests de fissuration sous contrainte sont réalisés dans une solution de chlorure de magnésium
(45%) selon les procédures définies par la spécification ASTM G-36 sur des échantillons
conformes à la spécification ASTM G-30 [13]. Les échantillons sont examinés après 24, 48, 72,
168 h puis toutes les 168 heures jusqu’à un maximum de 1008 heures. Toute fissure impose la
fin du test.
INDEX çç Sommaire
RESULTATS ET DISCUSSION
Les courbes d’iso-corrosion ont été déterminées pour l’alliage HYBRID-BC1 à partir de tests de
laboratoire réalisés sur un acide sulfurique pur et dilué (Figure 1) de 10 à 90% .Les valeurs de
corrosion de 0,1 mm/an et 0,5 mm/an représentent des valeurs généralement acceptées par les
utilisateurs comme respectivement «bonnes» et «acceptables».
240
210
90
60 <0,1 mm/an
30
0
0 10 20 30 40 50 60 70 80 90
Conc. acide
. pds%
Des vitesses de corrosion supérieures à 0,5 mm/an sont considérées comme inacceptables sauf
conditions exceptionnelles. Les figures 2 et 3 montrent que l’alliage présente une vitesse de
corrosion de 0,1 mm/an vers 110 °C relativement indépendant de la concentration et que cette
température va au-delà de 120 °C, si on accepte une vitesse de 0,5 mm/an. Sa résistance est
supérieure à l’ensemble des alliages Ni-Cr-Mo testés dans les mêmes conditions, y compris
l’alliage C-2000 connu comme étant le plus performant. L’alliage HYBRID-BC1 ne présente
pas d’augmentation des vitesses de corrosion vers 60-80% en acide sulfurique comme les aciers
inoxydables et les alliages de nickel de type-C. Son comportement est plus proche de l’alliage
B-3 de la famille Ni-Mo, mais présente des valeurs de corrosion plus élevées. Cependant pour
une concentration de 90% les taux de corrosions tendent vers une même valeur.
INDEX çç Sommaire
280 Courbes 0,1mm/an 280 Courbes 0,5mm/an
Temp., °C
Temp., °C 160 160
B-3 HYBRID-BC1
40 40
0 20 40 60 80 100 0 20 40 60 80 100
Conc. Acide % pds Conc. Acide % pds
L’acide sulfurique concentré est le produit chimique le plus utilisé. L’alliage C-276 est
largement employé dans ce milieu, mais la température maximum d’utilisation, en particulier
sur des échangeurs à plaques, est de 90 °C.
Des alliages à base nickel ou base fer qui contiennent une certaine quantité de silicium peuvent
convenir en milieu sulfurique concentré jusqu’à 130°C ou 110°C respectivement, avec des
vitesses de corrosion inférieurs à 0,1 mm/an [14], mais ils ne résistent pas lorsque l’alliage est
dilué. Il était donc important de connaître les performances de ce nouvel alliage en milieu
concentré (supérieur à 90% en acide sulfurique). Les tests de laboratoire effectués à une
concentration de 96% à différentes températures sont représentés dans le tableau 2.
L’alliage HYBRID-BC1 offre une résistance supérieure aux alliages de type Ni-Cr-Mo.
La limite de 0,1 mm/an semble reportée vers 110 °C soit 20 °C supérieure comparée à l’alliage
C-276. Il est moins résistant que l’alliage B-3, et équivalent aux aciers inoxydables avec
silicium mais présente un avantage considérable en terme de résistance dans l’acide sulfurique
dilué et aux corrosions localisées en milieu chlorure (voir chapitre suivant)
Vitesse de corrosion
Conc, H2SO4 (%) Temp, °C
mm/an
96 107 0,11
96 121 0,21
96 135 0,45
96 149 1,0
INDEX çç Sommaire
Influence de la présence d’ions ferriques en milieu sulfurique
Lors des tests précédents, il a été mis en évidence, le bon comportement de l’alliage HYBRID-
BC1 en milieu sulfurique «propre», c'est-à-dire sans impuretés oxydantes. Ces conditions ne se
rencontrent pratiquement jamais dans l’industrie chimique. C’est ce qui explique que
l’utilisation d’alliages de type Ni-Mo reste très faible. Pour simuler des conditions oxydantes,
des solutions d’acide sulfurique de concentration de 50 à 96% à 121°C ont été additionnées de
sulfate ferrique. La concentration en ions ferriques variant de 0 à 1000 ppm. Les résultats
obtenus dans les différentes conditions sont repris dans le tableau 3.
L’addition d’ions ferriques jusqu’à 50 ppm dans les différentes solutions d’acide sulfurique ne
provoquent pas d’évolution majeure de la vitesse de corrosion à 121°C. Au delà, on constate
que les vitesses de corrosion sont plus affectées dans les solutions diluées, limitant à 100 ppm
les ions ferriques pour une concentration de 50% d’acide sulfurique, à 250ppm pour 70% et
80% et acceptable jusqu’à 1g/l pour les solutions supérieures à 90%. Ce nouvel alliage a un
comportement très différent des alliages Ni-Mo qui sont corrodés à toute concentration d’acide
sulfurique en présence d’ions ferriques même à faible concentration (<50 ppm). Compte tenu du
fait que ces tests ont été faits à la température limite d’utilisation de l’alliage en milieu
sulfurique, on peut conclure que cet alliage ne sera pas affecté par la présence d’ions ferriques
résultats de corrosions sur acier en amont de son utilisation.
3+
Conc. Fe Vitesses de Corrosion (mm/y)
(ppm)
50% H2SO4 70% H2SO4 80% H2SO4 90% H2SO4 96 H2SO4
0 0.15 0.28 0.15 0.14 0.21
10 0.19 0.34 0.18 0.17 -
25 0.22 0.35 0.15 0.17 -
50 0.27 0.40 0.18 0.19 0.23
100 0.40 0.60 0.19 0.23 0.26
250 0.96 0.42 0.37 0.27 0.28
500 1.50 0.87 0.59 0.40 0.40
1000 2.88 1.67 1.05 0.44 0.35
L’acide chlorhydrique est un milieu connu comme étant très corrosif pour les matériaux
métalliques en particulier pour les aciers inoxydables.[Y] Les alliages de nickel sont parmi les
solutions acceptables mais montrent aussi des limites d’utilisation. L’acide chlorhydrique étant
non oxydant, les alliages Ni-Mo offrent les meilleures solutions, à la condition toutefois qu’il
n’y ait pas présence d’impuretés oxydantes (oxygène dissous ou autres), ce qui est loin d’être
toujours le cas. En présence d’éléments oxydants, qui permettent leur passivation, les alliages
Ni-Cr-Mo peuvent être utilisés, mais principalement pour les solutions diluées (>8% HCl) ou
très concentrées. Il ya donc un besoin important pour d’autres matériaux. C’est pourquoi,
l’alliage HYBRID-BC1 a été testé dans de nombreuses conditions différentes de manière à
déterminer aussi son comportement dans l’acide chlorhydrique.
INDEX çç Sommaire
Résultats comparatifs des courbes d’iso-corrosion dans l’acide
chlorhydrique
Les figures 4 & 5 offrent une comparaison des différentes courbes d’iso-corrosion (0,1 et
0,5 mm/an) dans les solutions d’acide chlorhydrique des alliages Ni-Cr-Mo, Ni-Mo et du nouvel
alliage. L’influence du molybdène est très importante dans les milieux chlorhydriques, c’est
pourquoi, la meilleure performance est offerte par l’alliage B-3. Sa vitesse de corrosion est en
fait plus dépendante de la température que de la concentration de l’acide. En revanche, les
vitesses de corrosion des alliages Ni-Cr-Mo sont fonction de la concentration et limitent
l’utilisation du meilleur de ces alliages (alliage C-2000) à une concentration maximum
d’environ 12%. L’alliage HYBRID-BC1 a un comportement voisin des alliages Ni-Cr-Mo
jusqu’à 10% HCl et n’apporte pas d’amélioration sensible. Par contre, quand la corrosion
devient plus important par augmentation de la concentration ou par augmentation de la
température, l’alliage HYBRID-BC1 se différencie sensiblement des alliages Ni-Cr-Mo en
offrant une résistance égale ou légèrement inférieure à l’alliage B-3 comme le montrent la figure
4 pour les concentrations supérieures à 12% et la figure 5 comparant les courbes iso-corrosion
0,5mm/an .
80 C-22 59
686
Temp.,°C
C-22
60
C-2000 59
60
HYBRID-BC1
40
686
40
B-3 20
20 0
0 4 8 12 16 20 0 4 8 12 16 20
Conc. Acide % pds Conc. Acide % pds
Parmi les éléments oxydants très importants au regard de leur influence sur la tenue à la
corrosion des matériaux, l’oxygène dissous dans les solutions joue un très grand rôle, car il est
souvent responsable de la passivation du matériau. Enfin d’en comprendre l’influence sur la
passivation de l’alliage HYBRID-BC1, des tests préliminaires ont été réalisés dans une solution
à 2,5% HCl à 121°C en autoclave pressurisé pendant 96 heures. La solution a été préalablement
purgée soit à l’azote (ce qui conduit à une solution non oxydante) soit à l’oxygène (ce qui
conduit à une solution saturée en oxygène. Des tests similaires sur l’alliage B-3 montre une
grande résistance en l’absence totale d’oxygène et une très grande vitesse de corrosion dans la
solution saturée. Inversement, les alliages C-2000 et C-22 montrent une très grande résistance
dans la solution saturée par l’oxygène et une attaque significative en absence d’oxygène.
INDEX çç Sommaire
L’alliage C-276 s’avère résistant dans aucune des deux solutions. En fait le seul alliage capable
de résister dans les deux conditions est l’alliage HYBRID-BC1 comme le montre le tableau 4.
Sa teneur de 15% Cr conduit à la passivation en milieu oxydant, mais la teneur de 22% Mo
assure la résistance en milieu non oxydant.
Tableau 4 : Influence de l’oxygène sur les vitesses de corrosion d’alliages de Nickel dans une
solution 2,5% HCl à 121°C sous pression
Les solutions d’halogénures sont connues pour initier des corrosions localisées sur les
aciers inoxydables (piqures et corrosions caverneuses). La tenue des alliages de nickel est bien
supérieure et c’est l’une des raisons courantes du choix de ces matériaux. Toutefois, il existe de
grandes différences de tenue en fonction particulièrement de la teneur en molybdène et
tungstène qui sont les deux éléments essentiels de la résistance à la formation de corrosions
localisées. Pour caractériser, la tenue des matériaux dans ces solutions on a recours à des essais
normalisés qui permettent une bonne comparaison des alliages en déterminant des températures
critiques d’apparition de piqures (CPT) ou de cavernes (CCT).
La présence est aussi à l’origine d’un autre type de corrosion localisée ; la fissuration
sous contrainte Les alliages de nickel sont reconnus pour être très résistant à ce type de
corrosion et l’alliage HYBRID-BC1 ne fait pas exception comme le montrent les tests qui ont
été réalisés.
Le tableau 5 montre les valeurs de températures obtenues dans une solution à 6% FeCl3+ 1%
HCl pour l’apparition de piqures (CPT) sur l’alliage HYBRID-BC1 et quelques autres alliages
Ni-Cr-Mo de type C. L’alliage HYBRID-BC1 ne montre aucune corrosion par piqures à 85 °C,
limite recommandée par l’ASTM en raison de l’instabilité du chlorure ferrique. Toutefois des
tests à plus haute température sont permis par le standard ASTM [15]. Les alliages Ni-Cr-Mo
sont réputés très résistants même dans ces conditions extrêmes au-delà de 100 °C. L’alliage
HYBRID-BC1 s’est montré encore plus résistant, ne développant des piqures qu’à 130 °C.
Toutefois ce résultat est modulé par une attaque générale supérieure à 0,5 mm/an à 120 °C. Il est
aussi connu que les alliages Ni-Cr-Mo offrent généralement une bonne résistance à la corrosion
en caverne. Même s’il existe une hiérarchie entre les différents matériaux, l’alliage C-2000 est
le plus performant avec une CCT à 80 °C. Il a été trouvé que l’alliage HYBRID-BC1 se montre
INDEX çç Sommaire
plus résistant que tous les alliages testés. La température de CCT excède les 100 °C avec une
corrosion généralisée de l’ordre de 0,15 mm/an.
La résistance à la fissuration sous contrainte en milieu MgCl2 45% (Test ASTM G-30 & G-36) à
ébullition ne montrent aucune initiation de fissuration après 1008 heures, sur l’alliage HYBRID-
BC1, comme sur tous les alliages de nickel, Ni-Mo ou Ni-Cr-Mo. (Tableau 6).
Tableau 6 : Résultats des tests de fissuration sous contrainte milieu MgCl2 à ébullition
MISE EN ŒUVRE
L’alliage HYBRID-BC1 se met en œuvre selon les mêmes procédures que les alliages à base
nickel Ni-Cr-Mo et Ni-Mo déjà commercialisés. Le lecteur pourra se référer aux brochures
fournies par les fabricants [19]. En particulier les soudages TIG et MIG sont parfaitement
adaptés à ce matériau. Des essais de corrosion ont été réalisés comparativement sur un
INDEX çç Sommaire
échantillon de métal de base et un échantillon soudé dans différents milieux chimiques. La
soudure a eté réalisé par une passe de fond autogène et une seconde passe en TIG avec fil de
3,2mm. L’élément soudé a ensuite été découpé en éléments plus petits pour réaliser les
différents essais de corrosion. Une série a été utilisée pour les essais de corrosion localisées dans
la solution standard ASTM G-48 méthodes C&D [12] pour déterminer les CPT et CCT comme
décrits précédemment. Une autre série a été utilisée pour la réalisation de tests d’immersion
dans différents milieux Le tableau 7 ci-dessus donne les valeurs de vitesses de corrosion
obtenues. La simple lecture de ce tableau permet d’apprécier que l’alliage conserve ses
propriétés de résistance à la corrosion après soudage sans traitement thermiques particuliers
avant ou après l’opération.
CONCLUSION
Un nouvel alliage a été conçu pour remplir un manque entre les alliages Ni-Mo et Ni-Cr-Mo. Le
nouvel alliage développé s’avère beaucoup plus tolérant à la présence d’éléments oxydants que
les alliages Ni-Mo tout en approchant leurs performances en milieux réducteurs, et par
conséquent il est aussi plus performant dans certains milieux que les alliages Ni-Cr-Mo, en
particulier en milieu acide sulfurique .Les courbes d’iso corrosion montrent que cet alliage est
plus performant que les alliages Ni-Cr-Mo. C’est aussi un alliage très résistant aux corrosions
localisées en présence de chlorures du fait de sa forte teneur en molybdène. Sa stabilité
thermique est pour le moins suffisante à assurer la continuité de la résistance à la corrosion
après soudage. L’alliage est actuellement commercialisé sous toutes les formes nécessaires à la
fabrication d’appareils de génie chimique. Parmi les applications potentielles de cet alliage
HYBRID-BC1, on peut noter des applications dans des milieux contenant de l’acide sulfurique :
récupération de calories sur acide sulfurique concentré, par eau de mer, dilueur d’acide
sulfuriques.
INDEX çç Sommaire
REFERENCES
[1] P. Crook, N.S. Meck, J. Crum. R.B. Reback. ASM Handbook, vol 13B, Corrosion: Materials
(2005) 231-238.
[2] W.Z. Friend Corrosion of Nickel and Nickel-Base Alloys, John Wiley & sons, (1980) p 292-
367.
[3] D.L. Klarstrom, P. Crook, A review of the Corrosion Resistant HASTELLOY® Alloys, 11th
International Symposium on Advanced Super Alloys-Production and Applications. The
Chinese Society for Metals, Shangaï, China, May (2007).
[4] P. Crook, “Advanced Materials and Processes”, Vol. 165 No. 6, 7, 8, and 9, June-July 2007,
p. 37, 31, 45, and 45
[5] Haynes online literature-Corrosion Resistant Alloys, HASTELLOY B-3®, Brochure H-
2120b, Haynes International, Inc
[6] P. Crook, Guide to Corrosion-Resistant Nickel Alloys, Brochure H-2114b, Haynes
International, Inc (2001).
[7] R.B. Reback, CORROSION, Paper N° 99219, NACE International, Houston (1999).
[8] R.B. Reback, J.R. Dilman, P. Crook, C.V.V. Shawber, Mater. Corros. 52 (2001) 289-297.
[9] N.S. Meck, P. Crook, D.L. Klarstrom, R.B. Reback, CORROSION, paper N° 04430, NACE
International, New Orleans (2004).
[10] N.S. Meck, P. Crook, 3rd Stainless Steel World 2004, KCI Publishing BV, Paper N° 0448,
p 316-323, Houston (2004).
[11] N.S. Meck, P. Crook, CORROSION, paper N° 05334, NACE International, Houston
(2005).
[12] Standard Tests Methods for Pitting and Crevice Corrosion Resistance of Stainless Steels
and Related Alloys by Use of Ferric Chloride Solutions, ASTM-Standard G-48-03, ASTM
International (2003).
[13] Standard Practice for Evaluating Stress-Corrosion Cracking Resistance of Metals and
Alloys in a Boiling Magnesium Chloride Solution, ASTM G 36-94 (2006), ASTM
International 2006
[14] Haynes literature-corrosion resistant alloys, Hastelloy D-205, Brochure H-2103, Haynes
International, Inc (1993).
[15] P. Crook, N.S. Meck, N. Koon, CORROSION, paper N° 08190 NACE International, New
Orleans (2008).
[16] A.C. Lloyd, PhD, Thesis, University of Western Ontario, London, Canada, Dec (2004).
[17] A.C. Lloyd, D.W. Shoesmith, N.S. Mc Intyre, J.J. Noel, J. Electrochem. Soc. 150 (2003)
B120-B130.
[18] A.C. Lloyd, D.W. Shoesmith, N.S. Mc Intyre, J.J. Noel, J. Electrochem. Acta 49 (2004)
3015.
[19]Haynes literature Fabrication of HASTELLOY Corrosion Resistant Alloys , Brochure H-
2010,Haynes International, Inc (2003)
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page66
RÉSUMÉ / ABSTRACT
23B
EFFET DU VIEILLISSEMENT THERMIQUE ET DE LA FRAGILISATION PAR
L’HYDROGÈNE SUR LES PROPRIÉTÉS DE MÉCANIQUE DE LA RUPTURE ET DE
RÉSILIENCE DES NUANCES TYPE 2,25CR1MO – APPLICATION À LA DÉTERMINATION
DE LA TEMPÉRATURE MINIMALE DE PRESSURISATION (MPT)
3
Industeel Belgium, ArcelorMittal Group, Belgium, [email protected]
RÉSUMÉ
Les tôles en nuances de type 2,25Cr1Mo standard (EN 10028-2 12CrMo9-10 et ASTM A387 gr 22) et avec addition de vanadium
2,25Cr1MoV (EN 10028-2 13CrMoV9-10 et ASTM A542 gr D) sont très largement utilisées lors de la construction de gros
réacteurs de raffinage de produits pétroliers.
Ces réacteurs sont réalisés à partir de composants de fortes épaisseurs (épaisseurs pouvant atteindre 300-350mm) sous forme de
tôles fortes, de viroles forgés, de fonds forgés, de piquages et de raccords. Ces appareils à pression sont sujets à des cycles
thermiques (stop and go – arrêts et redémarrages) et à des conditions de service très sévères (hautes températures, hautes pressions
et hautes pressions partielles d’hydrogène). Un des problèmes principaux pour les utilisateurs finaux est la définition de la
température minimale de pressurisation (MPT - Minimum Pressurizing Temperature) correspondant à la température la plus
basse à laquelle l’appareil peut être remis sous pression après un arrêt tout en évitant tout risque de rupture fragile (brutale) de
l’enveloppe. Cette définition se fait sur la base des propriétés de résilience et de mécanique de la rupture.
Cet article présente la méthodologie de détermination de la MPT et l’impact du vieillissement thermique ainsi que du service
sous hydrogène sur les propriétés mécaniques des matériaux et par conséquent sur la MPT. Cette détermination est explicitée
dans le cas d’un appareil à pression virtuel mais représentatif des réacteurs existant dans les raffineries.
ABSTRACT
Standard and Vanadium enhanced 2.25Cr1Mo plate steels (ASTM A387 gr. 22 / EN 10028-2 12CrMo9-10 and A542 type D / EN
10028-2 13CrMoV9-10) are commonly used for the manufacturing of heavy reactors for applications in petroleum refining
plants.
These reactors are fabricated from heavy wall plates, forged shells (thickness up to 300-350mm), forged nozzles and fittings.
They are subjected to thermal cycles (stop and go) and to severe service conditions (high temperatures, high pressure, high
hydrogen partial pressure). A primary concern for end-users is defining the Minimum Pressurizing Temperature (MPT). This
temperature is the lowest temperature at which the vessel can be repressurized after shutdown. This minimum temperature insures
no risk of brittle fracture, and is defined by fracture mechanics and/or CVN approaches and calculation.
This paper presents the methodology of MPT determination and the impact of ageing and exposure to hydrogen on material
mechanical properties and on the value of the MPT. The MPT determination method is explained by using a virtual pressure
vessel that is representative of reactor vessels found in petroleum refineries.
INDEX çç Sommaire
EFFECT OF TEMPER AND HYDROGEN EMBRITTLEMENT ON
FRACTURE MECHANICS AND CVN PROPERTIES OF 2,25CR1MO
STEEL GRADES – APPLICATION TO MINIMUM PRESSURIZING
TEMPERATURE (MPT) ISSUES
1
Industeel France, Site du Creusot, ArcelorMittal Group, France, [email protected] /
[email protected] / [email protected]
2
Industeel France, Site de Châteauneuf, ArcelorMittal Group, France,
[email protected] / [email protected]
3
Industeel Belgium, ArcelorMittal Group, Belgium, [email protected]
* Corresponding author
ABSTRACT
Standard and Vanadium enhanced 2.25Cr1Mo plate steels (ASTM A387 gr. 22 / EN 10028-2
12CrMo9-10 and A542 type D / EN 10028-2 13CrMoV9-10) are commonly used for the
manufacturing of heavy reactors for applications in petroleum refining plants.
These reactors are fabricated from heavy wall plates, forged shells (thickness up to 300-
350mm), forged nozzles and fittings. They are subjected to thermal cycles (stop and go) and
to severe service conditions (high temperatures, high pressures, and in particular high
hydrogen partial pressures). A primary concern for end-users is defining the Minimum
Pressurizing Temperature (MPT). This temperature is the lowest temperature at which the
vessel can be repressurized after shutdown. This minimum temperature insures no risk of
brittle failure of the containment body, and is defined by fracture mechanics and/or CVN
approaches and calculations.
This paper presents the methodology of MPT determination and the impact of ageing and
exposure to hydrogen on materials’ mechanical properties and consequently on the value of
MPT. MPT determination methodology is explained by using a virtual pressure vessel
representative of reactor vessels found in petroleum refineries.
INDEX çç Sommaire
RESUME
Les tôles en nuances de type 2,25Cr1Mo standard (EN 10028-2 12CrMo9-10 et ASTM A387
gr 22) et avec addition de vanadium 2,25Cr1MoV (EN 10028-2 13CrMoV9-10 et ASTM A542
gr D) sont très largement utilisées lors de la construction de gros réacteurs de raffinage de
produits pétroliers.
Ces réacteurs sont réalisés à partir de composants de fortes épaisseurs (épaisseurs pouvant
atteindre 300-350mm) sous forme de tôles fortes, de viroles forgées, de fonds forgés, de
piquages et de raccords. Ces appareils à pression sont sujets à des cycles thermiques (stop
and go – arrêts et redémarrages) et à des conditions de service très sévères (hautes
températures, hautes pressions et en particulier hautes pressions partielles d’hydrogène). Un
des problèmes principaux pour les utilisateurs finaux est la définition de la température
minimale de pressurisation (MPT - Minimum Pressurizing Temperature) correspondant à la
température la plus basse à laquelle l’appareil peut être remis sous pression après un arrêt,
en évitant tout risque de rupture fragile (brutale) de l’enveloppe. Cette définition se fait sur la
base des propriétés de résilience et de mécanique de la rupture.
INTRODUCTION
One of the main risks in the petrochemical / refining industry is the risk of brittle (i.e.
unstable) failure of reactors. Thick wall low-alloy pressure vessels are designed to operate at
high temperatures, and at high pressures and hydrogen partial pressures. This combination can
lead to numerous potential problems.
During unit shutdowns, for example for maintenance operations, pressure and temperature are
reduced to levels much lower than exposed to in operation. After the end of the maintenance
procedures, the operation staff restarts the unit, raising the vessel pressure and temperature to
the operational conditions, thus starting a new production cycle. During the early stages of the
start-up process, strict attention must be paid to metal temperature and internal pressure to
insure sufficient toughness of the reactor material. This is done by controlling the heating and
pressurization rate of the vessel, avoiding conditions that could cause brittle failure. This
dramatic and unacceptable situation could occur if the low-alloy steel temperature is not high
enough to reach a level of fracture toughness sufficient to mitigate crack instability and
catastrophic failure.
INDEX çç Sommaire
2
steps of temperature-pressure couples that must be abided by when the production cycle is
restarting.
In the case of a newly fabricated reactor, the definition of MPT is quite easy because all
needed material properties can be required at the delivery of the pressure vessel and then the
reactor is designed to avoid the risk of brittle failure. Steelmakers, fabricators, engineerings,
and end-users are aware of the potential problems and safety margins are used. Evaluation of
alloy embrittlement is also taken into consideration. Steelmakers provide guarantees to their
customers that the steels they produce have a satisfactory resistance to temper embrittlement.
This guarantee stems from confidence gained in testing of the steels that comprise of furnace
heating of coupons and subsequent Charpy V-notch impact testing. In the furnace, the
coupons are exposed to an accelerated ageing simulation program, also called Step Cooling.
The situation becomes more complex when the low-alloy steels are exposed to hydrogen at
elevated temperatures. The complexity arises because of the lack of data characterizing the
effects of hydrogen on fracture toughness. The mechanical tests required to assess fracture
toughness of low-alloy steels exposed to high hydrogen pressures at elevated temperatures are
not very common, and moreover are not easy to perform. These tests are important because of
the strong embrittling effect of hydrogen on reactor steels. Another major problem concerns
the definition of MPT in the case of vintage reactors that have been operating for years and
for which historical operating data may not be fully available.
This paper focuses on the degradation mechanisms that promote embrittlement of low-alloy
steels. The main actors are temper embrittlement and the embrittlement due to hydrogen. For
pressure vessels operating in the creep range, creep embrittlement should also be taken into
consideration. This paper also presents the findings of recent tests performed in hydrogen
environments to assess mechanical properties.
INDEX çç Sommaire
Mechanical properties as tabulated in manufacturers material test reports generated by
steelmakers are affected by subsequent fabrication processes. All further thermo-mechanical
operations performed by vessel fabricators (e.g. shell rolling, head forming, welding of the
shell courses, heads, and nozzles) will affect the given initial mechanical properties, and as a
consequence modify global behaviour of the structure. In general, customers ask steelmakers
to assure given properties after minimum PWHT (corresponding to delivery state of the
pressure vessel to final customer) and after maximum PWHT (corresponding to pressure
vessel state after some weld repairs).
Furthermore, tempering of a low alloy steel can also modify mechanical properties. Special
attention has to be paid to this influential parameter. Initial tempering done by steelmakers as
well as further fabrication heat treatments (Dehydrogenation Heat Treatment (DHT),
Intermediate Stress Relieving Treatments (ISR) and final Post Weld Heat Treatment
(PWHT)) must be performed with great care. Final properties of pressure vessels’ constitutive
steels are not only a function of their chemical composition, but is also a function of all heat
treatments performed during fabrication. A paper by CHAUVY et al. focuses on this point
[1].
Figure 1 : Effect of tempering on CVN and Figure 2 : Effect of tempering on CVN for
Drop Weight transition temperature for 2,25Cr1MoV grade (SAW and SMAW weld
2,25Cr1MoV grade (Base metal). metal).
Recommendations and guidelines concerning heat treatments have been published by API
(American Petroleum Institute) [3] to control and master the whole pressure vessel fabrication
process and hence to assure sufficient properties to reactors. These properties can be
considered as a basis for design, at the beginning of service life of pressure vessels or
reactors. However these properties will evolve with time and exposure to operating
environment. This change in material behaviour is reported in next chapter.
INDEX çç Sommaire
2 – EFFECTS OF TEMPERATURE AND HYDROGEN ON MATERIAL
PROPERTIES
Low alloy pressure vessels used in high-pressure refining processes are generally thick wall
(typically 150-350mm), and built from forged rings or rolled and welded plates each forming
shell courses. Inner diameters of these reactors typically range from 4 to 5 meters. Total
weight of these very large components ranges from 500 to 800 metric tons. Reactor wall is a
composite structure made of two different metallic materials. The primary pressure boundary
is fabricated from low-alloy steel (e.g. Chromium-Molybdenum or Chromium-Molybdenum-
Vanadium steel), while the other component is a thin corrosion resistant layer of austenitic
stainless steel. While small components can be fabricated from clad plates, the liner in large
and thicker components is typically a weld overlay.
The primary function of low-alloy steel is to provide strength and toughness necessary for a
pressure boundary at high pressures and temperatures. In the following text, this part of the
reactor wall will be referred to as “Base Material”. The vessel’s mechanical behaviour is
defined by base material. The function of the stainless steel weld overlay is to protect the low-
alloy steel from corrosion and its minor influence on mechanical behaviour of the structure is
therefore not taken into consideration.
The extent of temper embrittlement is measured by the shift of CVN transition to higher
temperatures and was first studied in Nickel-Chromium-Molybdenum-Vanadium steels used
for forged rotors devoted to low-pressure turbines in electrical power plants. It has been
proven that temper embrittlement is linked to the co-segregation of impurities at prior
austenitic grain boundaries. Impurities, such as Tin (Sn), Antimony (Sb), Arsenic (As),
Phosphorus (P), co-segregate with alloying elements (primarily Silicon (Si) and Manganese
(Mn), but also Chromium (Cr) and Nickel (Ni)). The result leads to grain boundary
embrittlement, and subsequently to a risk of intergranular fracture. Temper embrittlement
occurs at temperatures below 600°C, but mainly in the range 350-550°C (typical C-Curves of
iso-embrittlement for a given temperature and time couple as shown in figure 5).
Some metallurgical parameters can influence the extent of CVN transition in low-alloy steels.
Increasing grain size can be very detrimental. Additionally, microstructures that promote high
hardness values are more sensitive than softer microstructures. This is why weld heat affected
zones (HAZ), and in particular, coarse grain heat affected zone (CG-HAZ) are locations
where problems are the most likely to occur. As previously stated, high Manganese and
Silicon contents are known to promote temper embrittlement. The role of Molybdenum is
more complex. Low concentrations of Molybdenum (about 0.5%wt, as in the case of P11) is
much better than no addition, and better than higher concentrations (about 1%wt, for 2,25%Cr
INDEX çç Sommaire
steels, such as P22). Some references on this subject are given in a paper by BOCQUET [4]. It
is well known from end-users that standard 2,25Cr1Mo is much more sensitive to temper
embrittlement than 1,25Cr0,5MoSi and Vanadium modified 2,25Cr1Mo (2,25Cr1Mo0,25V).
Different parameters based on chemical composition have been created to evaluate low-alloys
(and in particular CrMo steels) sensitivity to temper embrittlement. These chemistry
parameters have been developed using a statistical analysis approach, based on evaluation of
CVN parameters and Ductile to Brittle Transition temperature (DBTT), also called Fracture
Appearance Transition Temperature (FATT). This method involved testing heats with minor
differences in chemical composition after isothermal treatments, both laboratory samples and
actual reactor steels were used.
The two most widely used parameters are Watanabe’s J-Factor (standard and simplified) for
base materials, and the Bruscato X-bar Factor for weld metals. A third parameter, called
“Equivalent Phosphorus” content can be used for both base and weld metal but is not widely
used. The definition of these parameters is given hereafter by equations 1 to 4.
Figure 3 shows FATT scattering as a function of J-Factor. It can be noticed that FATT
increases dramatically with increasing concentration of impurities. It suggests that a low level
of impurities (low J factor) must be specified and obtained in order to get acceptable CVN
toughness. Figure 4 shows the evolution of FATT at specific exposure times, at service
temperatures for impurities concentrations (J-Factor). It can be seen that CVN properties of
the 2,25Cr1Mo steels with low concentrations of impurities retain acceptable FATT below or
just above room temperature. In the case of high J-Factor steels that have been exposed to
elevated temperatures for very long time of service conditions, the CVN ductile-brittle
INDEX çç Sommaire
transition temperature can approach 200°C. More results can be seen in papers by
PRESCOTT [5, 6].
Last revision of API RP 934 generalizes the use of Bruscato Factor for both base materials
and weld metals, instead of Bruscato Factor for welds and J-Factor for bases. This recent
evolution of the standards is due to the general misunderstanding of J-Factor values for
Silicon containing steel grades such as A387 gr 11 (1,25Cr0,5MoSi) or A387 gr 5
(5Cr0,5MoSi) that can not be compared to the values of non Si-containing steels.
200
Base Material [API] 20000Hrs
Base Material [JSW] 30000Hrs
Base Material [Chevron] 35000Hrs
Base Material [Chevron] 75000Hrs 99% confidence limit
150 Base Material [ ] 150000 to 200000Hrs
Weld Metal [API] 20000Hrs
Weld Metal [JSW] 30000Hrs
Weld Metal [Chevron] 35000Hrs
Weld Metal [Chevron] 75000Hrs 95% confidence limit
100 Weld Metal [ ] 150000 to 200000 Hrs
FATT (°C)
50
-50
0 50 100 150 200 250 300 350 400 450 500
J-Factor
250
Base Material - J<150
Base Material - J<250
Base Material - J<400
Base Material - J>400
200 Weld Metal - J<150
Weld Metal - J<250
Weld Metal - J<400
Weld Metal - J>400
150 J=500
100
FATT (°C)
J=350
50
J=250
-50
J=150
-100
1000 10000 100000 1000000
Exposing Time (Hrs)
Figure 4 : Effect of impurities (J-Factor) on CVN properties after very long exposing time at
service temperatures (2,25Cr1Mo steel grades) – historical data
INDEX çç Sommaire
heat treatment program is to provide a relatively quick and cost-effective method to simulate
the embrittlement behaviour that occurs after long-term isothermal exposure. It is based on
the “iso-embrittlement” curves shown on the time-temperature diagram depicted in figure 5
While many different variations exist, a typical step cooling test program takes approximately
two weeks of total time. In refining applications, the most traditional step-cooling program
makes use of the following cycle: heat-up to 593°C, soaking for 1 hour, cooling at 5.6°C/hour
with soaking steps at 538°C (15 hours), 524°C (24 hours), 496°C (60 hours) and 468°C (125
hours). Coupons exposed to this step cooling are then machined into Charpy V-notch samples
and impact tested at various temperatures as required by API 934-A [3] (Materials and
Fabrication of 2 1/4Cr-1Mo, 2 1/4Cr-1Mo-1/4V, 3Cr-1Mo, and 3Cr-1Mo-1/4V Steel Heavy
Wall Pressure Vessels for High-temperature, High-pressure Hydrogen Service) to determine
an appropriate transition curve.
Historically it has been shown that step-cooling test programs do not simulate the full extent
of embrittlement that has been observed in samples removed from retired reactors that have
been in service for many years. Nevertheless, step-cooling testing does provide a relative
measure of an alloy tendency to temper embrittlement. In order to add a measure of
conservatism to the design, the differences in temperature between the PWHT steel and the
step cooled coupons at 54J fracture energy (“temperature shift”) is multiplied by a factor of
2.5 or 3. Multiplying the test coupon shift temperature by 2.5 provides a useful correlation to
reactors that have operated at 400-450C for periods up to 30 years.
Figure 5 : Iso-embrittlement curves used to define Step Cooling Heat Treatments (taken from
reference [2]).
have special chemistry: most common rules for base materials are J<150 in the case of
1,25Cr0,5MoSi steel grade or J<100 for 2,25Cr1Mo(V) grades. Additional
INDEX çç Sommaire
requirements for P+Sn<0,012%wt, can be added. For weld consumables, the most
common criterion is Xbar <15 or 12.
have coupons tested before and after Step Cooling heat treatment to evaluate the
embrittlement of the given material.
Nevertheless, usefulness of Step Cooling must be balanced by noticing that 2,25Cr1Mo steels
purity is now very good (J-factor always below 100 due, to strong improvements in
steelworks) and then 54J impact energy is achieved in general close -100°C for base
materials. For this very low J-factor steels, shift due to temper embrittlement is not significant
anymore and then Step Cooling becomes not mandatory.
Discussion between steelmakers and API were ongoing for years to remove this
embrittlement test which is very detrimental to delivery schedules. Since May 2008 revision
of API RP 934, one complementary paragraph has been added concerning relevancy of Step
Cooling. This paragraph takes into account quality evolutions in steel shops. Simulated ageing
tests can be now waived in the case of high quality steelmakers provided that Watanabe’s J
factor remains below 100 and 54J Impact energy temperature is below -60°C. If one on those
two conditions is not achieved, Step Cooling tests have to be performed.
Figure 6 depicts steady state concentration gradient of hydrogen through a reactor wall during
operation and after cooling. Note that the concentration in the low-alloy steel is highest at the
interface with the austenitic stainless steel overlay. One common problem associated with the
diffusion of hydrogen is Hydrogen Induced Disbonding (HID) phenomenon. This is discussed
in details in a paper from COUDREUSE et al. [7].. Also discussed in this latter paper is a
method to compute hydrogen concentrations and gradient through the wall thickness.
INDEX çç Sommaire
Figure 6 : Profiles of hydrogen concentration in Material and Overlay before and after cooling
down of a reactor in refining plant.
3.0
2.5
Local Hydrogen Content (ppm)
2.0
1.5
1.0
0.5
0.0
0 25 50 75 100 125 150 175 200
Figure 7 : Profiles of H2 content in Base Material before and after cooling down.
The embrittling effect of hydrogen in Cr-Mo steel is very close to that attributed to temper
embrittlement. Considering either CVN transition curves (with and without hydrogen) or
fracture mechanics toughness transition curves, the effect of hydrogen results in an upward
shift of transition temperature.
INDEX çç Sommaire
Recent CVN transition curves have been published by SAKAI et al [9] for 2,25Cr1Mo steel
grade and PILLOT et al [10] for 2,25Cr1Mo, 2,25Cr1MoV and C-Mn steels. Figures 8 and 9
show the shift in TK54J (temperature for which impact energy equals 54J) values in the case of
2,25Cr1Mo and 2,25Cr1MoV steel respectively. Values are shown for base metal, HAZ, and
weld metal, in “as-delivered” and after step-cooling conditions) as a function of hydrogen
concentration in the bulk.
Figure 10 shows the shift in CVN transition temperature due to hydrogen concentration after
various heat treatments. This corresponds to the slope of curves given in figures 8 and 9.
These slopes indicate the loss of toughness as a function of ppm of hydrogen in steels.
Figure 11 shows the decrease in fracture toughness with increasing hydrogen content. It can
be seen that the higher the hydrogen content, the higher the transition temperature. Another
way to evaluate the effects of hydrogen is to consider the toughness at a given temperature
and then, the higher the hydrogen content, the lower the toughness.
0 0
0 1 2 3 4 5 6 0 1 2 3 4 5 6
Base Material
-20 -20
Base Material SC
Martensitic HAZ
-40 -40
Martensitic HAZ SC
-60 -60
TK 54J (°C )
TK 54J (°C )
-80 -80
-100 -100
Base Material
Base Material SC
-120 -120
Weld Metal
Weld Metal SC
-140 -140
Martensitic HAZ
Martensitic HAZ SC
-160 -160
H2 content (ppm) H2 content (ppm)
Figure 10: CVN transition curve shift as a function of H2 content (ppm weight) in CrMo(V)
and C-Mn steels
INDEX çç Sommaire
250
Base Materials
MC
1%
99%
200 3 ppm H2
4 ppm H2
Aged - no hydrogen
Aged - 1,5 to 4ppm
KIC ( MPa.Vm)
150
100
KISH
50
0
-150 -100 -50 0 50 100 150
T-To (To=FATT-80°C) (°C)
Figure 11: Review of K1C data for 2,25Cr1Mo steel (as delivered, aged, hydrogen charged and
aged + hydrogen charged conditions).
The main database concerns KIc values of base materials in as delivered conditions. Some
data relative to aged material and hydrogen charged material are also plotted in this figure. In
this plot, KIC is represented as a function of T-T0, where T0 is the temperature for which static
toughness is equal to 100 MPa.√m. In this case, T0 is indexed on FATT temperature obtained
with CVN specimens.
On one hand, in the case of new pressure vessels, mandatory documentation supplied to end
users is clearly sufficient to get initial materials’ properties and initial design is done to insure
a safe MPT at a given temperature for given service conditions (operating temperature,
operating pressure, environmental temperature). Final service conditions are defined at the
very beginning of a project and then MPT is given as a basis. Engineers have to design the
pressure vessels both to support service conditions and to insure the given MPT.
For example, in the case of refining plants located in cold climates where temperatures can be
very low, toughness of material has to be much better than in the case of warmer regions to
insure safe behaviour and allow re-pressurization at or close to ambient temperature.
Embrittlement is taken into consideration to maintain stable or quasistable MPT during the
whole life of the reactor.
INDEX çç Sommaire
In this case, then, a method for calculating MPT can be summarized by the following
formulae (5 and 6):
Tk 54 J + x × ∆ Tk54 J SC + C H 2 × ∆ Tk54 J H2 < TK CVN ( MPT ) (eq. 5)
if we consider methodology based on CVN properties, or
Where:
TK54J is the temperature for which we get CVN impact energy equal to 54J just after
building of the reactor
∆TK54J_SC is the shift of temperature due to temper embrittlement in Step Cooling
x is the correlation factor between Step Cooling embrittlement and actual
embrittlement (in general taken equal to 2,5 or 3).
∆TK54J_H2 is the shift of temperature due to hydrogen embrittlement (per ppm of
hydrogen)
CH2 is the hydrogen concentration in the reactor wall
TKCVN(MPT) is the indexation temperature computed during design of the pressure
vessel to link CVN properties required at the given MPT temperature (it takes into
consideration safety margins).
Or:
TKIC is the temperature for which we get fracture toughness equal to 100MPa.√m just
after building of the reactor
∆TKIC_SC is the shift of temperature due to temper embrittlement in Step Cooling
x is the correlation factor between Step Cooling embrittlement and actual
embrittlement (in general taken equal to 2,5 or 3).
∆TKIC_H2 is the shift of temperature due to hydrogen embrittlement (per ppm of
hydrogen)
CH2 is the hydrogen concentration in the reactor wall
TKKIC (MPT) is the indexation temperature computed during design of the pressure
vessel to link KIC properties required at the given MPT temperature (it takes into
consideration safety margins).
In the case of vintage reactor, the problem is clearly different. The pressure vessel is already
operating, sometimes for years or decades, and end-users have to insure safe use of this
INDEX çç Sommaire
reactor. Material properties are not always perfectly known, and therefore a reverse way of
thinking, compared to previous case, is required.
Properties of materials cannot be changed and then the final issue is to evaluate the safest
temperature of re-pressurization, taking into account already accumulated degradation
(temper embrittlement and hydrogen embrittlement) as well as economic point of view.
Figure 12 explains how temper embrittlement and hydrogen embrittlement affects toughness
properties of a given material. The black curve corresponds to toughness mastercurve of as
delivered material at the very first start of the reactor. This initial state of material properties
can be known or not, depending on the quality of initial documentation delivered to end-user.
This point will be discussed later. The blue curve represents the toughness mastercurve of the
same material, taking into account temper embrittlement. Some considerations on this curve
will also be given hereafter. Pink (to purple) to red curves concern the effects of hydrogen on
toughness. The cumulative effect of these two phenomena applied on initial properties leads
to a toughness transition curve that allow the calculation of MPT for a given reactor
geometry, a given flaw size and shape and then for given service conditions.
Figure 12: Cumulative effect of ageing and hydrogen embrittlement of quasi static fracture
toughness.
Concerning initial properties, it is only necessary to get them if end user wants to estimate the
level of degradation due to temper embrittlement. If initial documentation of pressure vessel
does not contain this information, an estimation of the properties can be obtained by removing
a test coupons from the reactor and then apply to it a de-embrittlement heat treatment (600°C
for one hour followed by quick cooling – in general water cooling – to avoid re-embrittlement
while crossing 350-550°C range).
This method avoids too much conservatism, but it is clearly not a perfect method and only an
estimation of initial properties can be obtained. Scattering is very large and then special care
must be taken to analyze the results. It is also based on the hypothesis that tests coupons have
INDEX çç Sommaire
been put inside the pressure vessel during construction of the plant. In some case it has been
done, in some others, it hasn’t and then this solution is not available.
The second method to assess properties of material, for cases with no coupons to be tested, is
clearly conservative but is based on large return of experience. In a study performed by API
in the early 80’s [11], it has been shown that for reactors built before 1975 not fully in
accordance with API RP 934A [3] requirements concerning chemistry, embrittled materials
can exhibit very poor CVN properties. In particular some reactors with electro-slag welds
(ESW) have TK54J values up to about 170°C (300°F). For pressure vessels containing such
welds, this maximum value should be taken as a basis. In all other case, maximum transition
temperature has been evaluated to about 120°C (250°F).
For more recent pressure vessels with more restrictive chemistries and compliant with API RP
934-A requirements, the highest transition temperature has been evaluated to be about 40°C
(100°F).
For very new reactors, CVN or fracture toughness data available in the initial documentation
(as delivered properties and embrittled by Step Cooling heat treatment) are sufficient to
calculate material toughness.
Concerning hydrogen embrittlement, two philosophies are spread. The first one consists in
thinking that hydrogen has a real effect on the material only if its content is above a threshold
estimated to 3ppm. For lower contents, it is considered that fracture mechanics properties and
especially fast “brittle” properties are similar to the case with no hydrogen. Figures 13 and 14,
taken from PRESCOTT’s literature review [6] and KOBE STEEL’s work on 2,25Cr1Mo
[12], show that hydrogen levels below 2ppm and up to 5ppm actually do affect fracture
toughness. It has already been proved by PILLOT et al. [10] and SAKAI et al. [9] that low to
medium contents of hydrogen also present a detrimental effect on CVN toughness (see Figure
8).
400
KIC valid
KJc Valid
350 KIC
KIC at Da=0
KJC Valid H2
300 KIC H2
lower bound without hydrogen
lower bound with hydrogen
KIC, KJC (MPa.Vm)
250
200
1,17ppm
1,38ppm
150
1,19ppm
100
1,08ppm
50
1,20ppm
0
-150 -100 -50 0 50 100 150
Temperature (°C)
Figure 13: Effect of low hydrogen content on quasistatic fracture toughness (2,25Cr1Mo).
INDEX çç Sommaire
Figure 14: Effect of high hydrogen content on quasistatic fracture toughness (2,25Cr1Mo).
Figure 15 attempts to correlate CVN transition shift due to temper embrittlement and to
hydrogen embrittlement with data coming from fracture mechanics tests. Shift measurements
have been obtained using Figure 10 values and WALLIN correlation between KJC and TK28J
as described in ASTM E1921 [13] test procedure or in ASTM Publication on Mastercurve
Method [14]. For one inch thick specimen, this relationship is given by equations 7 and 8:
Equation 8 gives the fracture mechanics mastercurve with a given probability of failure Pf :
1/ 4
1
TK JC (T ) = 20 + (11 + 77. exp(0,019(T − T0 )). ln (eq. 8).
1 − Pf
One other typical problem is the evaluation of TK28J. In general, only TK54J or FATT is
known by end-users. Some relations exist to extrapolate this value, but it includes further
scattering in data. Equations 9 and 10 are devoted to extrapolation of TK28J knowing TK54J
and tensile properties (or other CVN values at one given temperature TKV) while equations
11 to 13 allow extrapolations of TK28J knowing FATT (only valid for 2,25Cr1Mo). These
formulae come from the European Research Project QUALYTOUGH [15]. A security of
10°C can be added to be sure to be conservative in every situation. Figure 16 taken from [15]
shows errors in TK28J estimation regarding initial CVN value considered (TK41J, TK68J or
FATT).
INDEX çç Sommaire
C CVN.(US - 28)
TK 28J = TKV - .ln (eq. 9)
4 28.(US - CVN)
The constant C is given by C = 34 + YS/35,1 - US/14,3 (eq. 10)
with:
US the Upper Shelf CVN toughness (can be estimated typically to 250J for
2,25Cr1Mo steel)
TKV the temperature for which CVN impact energy is known
CVN the impact energy at temperature TKV
YS the yield strength at room temperature
C CVN - 2
T = FATT + .ln (eq. 11)
2 US - CVN
C 54.(US - 28)
Or TK 28J = (FATT - 7 ) - .ln (eq. 12)
4 28.(US - 54)
with C given by equation 10.
These conventional relationships used in a pressure vessel are applied here in the case of
hydrogen charged material. Figure 15 shows a good correlation between test data and the
models, but very few data are available. A more accurate study should be done to further
validate these relationships in the case of hydrogen charged materials.
300
Base Materials
MC
1%
Stable propagation of crack
99%
3 ppm H2
4 ppm H2
200
Aged - no hydrogen
KIC ( MPa.Vm)
100
KISH
0
-150 -100 -50 0 50 100 150 200 250
T-To (To=FATT-80°C) (°C)
Figure 15: effect of temper and hydrogen cumulative embrittlement on fracture mechanics
data (2,25Cr1Mo).
INDEX çç Sommaire
Figure 16: scattering in evaluation of TK28J.
In the next section, an application of these rules to a virtual pressure vessel will be developed
as an example.
First of all, the geometry of the pressure vessel can be assimilated to a cylinder shell with the
following dimensions:
Outer diameter is 4 meters (Ro), inner diameter is 3,4 meters (Ri)
Total wall thickness is 0,3 meter
Length (for information only) is 20 meters.
INDEX çç Sommaire
Flaw geometry is assumed to be at the surface of the inner side of the reactor with semi-
elliptical shape in longitudinal orientation (half penny shaped defect). Dimensions of this
defect are:
Shortest axis (a) is very small regarding total wall thickness.
Longest axis (c) is greater than shortest axis.
Shortest axis is radial and longest axis is longitudinal.
After cooling of the reactor, estimated maximal content of hydrogen in base and weld metal is
about 3ppm, in accordance with Figure 7.
Material properties, measured on the coupon removed from the reactor follow (specimen
thickness is 25mm):
Quasistatic toughness of base material: KIC at 20°C is 150 MPa.√m (aged material, no
hydrogen).
Quasistatic toughness of base material: KISH at 20°C is 50 MPa.√m and KISH at 100°C
is 75 MPa.√m (aged material, about 3ppm hydrogen).
Quasistatic toughness of weld metal: KIC at 20°C is 50 MPa.√m (aged material, no
hydrogen).
Quasistatic toughness of weld metal: KISH at 20°C is 45 MPa.√m and KISH at 100°C is
70 MPa.√m (aged material, about 3ppm hydrogen).
All calculation are done using API 579:2007 Fitness-For-Service standard (FFS) [16].
Calculation of KI value following level I of FFS can be made using Annex C and in particular
paragraph C.5.10 (Cylinder – Surface Crack, Longitudinal Direction – Semi-Elliptical Shape,
Internal Pressure). KI is given by formulae 14 to 18:
2 3 4
πa
2G − 2G a a a a
2
P.Ro
KI = 1
+ 3G2 − 4G3 + 5G4 (eq. 14)
− Ri Q
2 2 0
Ro Ri Ri Ri Ri
With:
1, 65
a
Q = 1,0 + 1,464
c (eq. 15)
INDEX çç Sommaire
Let’s consider the outer side of the flaw, (the location of the defect the most inside the wall). φ
is the angle of deepest point of the crack regarding the surface of the wall (equals to π/2).
Coefficients Ai,j are taken from FFS in tables C.12 or C.10 for elongated defects.
Two cases are taken into consideration and lead to equations 19 and 20:
Ratio a/c is close to 0 (e.g. the defect is a scratch or similar), and then G0=1,12 and
2
2,24.P.Ro
G1=0,682 and K I ≈ πa ≈ 8.07 P πa (eq. 19).
Ro − Ri
2 2
Ratio a/c is close to 1 (semi-circular defect), and then G0=1,044 and G1=0,741 and
πa
2
2,088.P.Ro
KI ≈ ≈ 4.79P πa (eq. 20).
Ro − Ri
2 2
2.464
The worst case is clearly an elongated defect and then only this case will be considered in
further calculation to be conservative.
Figure 17 and 18 give the evolution of toughness of the wall material as a function of
temperature. Upper shelf toughness is limited to 220 MPa.√m in the case of base material and
to 125 MPa.√m in the case of weld metal to take into account the risk of lowering of this
plateau for very strong temper and/or hydrogen embrittlement.
250 250
150 150
100 100
Base metal - Pf=5%
Base metal with H2 - Pf=5%
KIC
KISH
50 50
0 0
0 50 100 150 200 250 0 50 100 150 200 250
Metal Temperature (°C) Metal Temperature (°C)
Figure 17: Evolution of material toughness Figure 18: Evolution of material toughness
(with and without hydrogen) regarding metal (with and without hydrogen) regarding metal
temperature (Base material) (25mm temperature (Weld metal) (25mm thickness
thickness specimens). specimens).
INDEX çç Sommaire
Thickness correction is made using Weibull’s law described by Wallin et Al. in [14]. This
leads to equation 21:
Weld metal is the location where toughness is the lowest and then assessment of MPT must
be done in this region to be safe.
Figure 19 shows the evolution of toughness in the case of the full thickness wall (300mm). A
shift to higher temperature (about 40°C) is then applied.
140
80
60
40
20
0
0 50 100 150 200 250
Metal Temperature (°C)
Figure 19: Evolution of material toughness (with and without hydrogen) regarding metal
temperature (Weld metal) (Full thickness wall – 300mm).
INDEX çç Sommaire
160
maximal a=2mm
maximal a=3mm
140 maximal a=4mm
maximal a=5mm
Maximal admissible pressure (Bara)
maximal a=10mm
120
100
80
60
40
20
0
0 50 100 150 200 250 300 350
Metal temperature (°C)
Figure 20: Evolution of maximal admissible pressure regarding metal temperature (Weld
metal) (Full thickness wall – 300mm).
Figure 20 represents the MPT diagram, which corresponds to the maximal admissible
pressure in the reactor at a given temperature for a given flaw size. It appears that in this
example, if defects with radius greater than about 3,5mm are detected, a reduction of maximal
admissible pressure will be necessary. This implies reduction of in-service conditions.
Some optimisation can be done to improve efficiency. More sophisticated calculations, taking
into account plasticity can also be done to improve accuracy of this diagram and then reduce
maintenance cycle duration and costs.
As explained before and shown in figure 6, during operation of reactors, a high content of
hydrogen can enter the wall due to dissociation of hydrocarbon molecules. Diffusion of
hydrogen inside the wall is favoured by high temperature and high pressure and then
equilibrium between hydrogen wall content and internal hydrogen partial pressure inside
reactor is reached following Sievert’s law. Diffusion mechanisms are described more deeply
in COUDREUSE et al. paper [7] and KOERS paper [17]. During shutdown, in-service
equilibrium is not reached anymore and then hydrogen has to diffuse to match equilibrium. In
a perfect wall, with no defects, hydrogen will diffuse to interface with overlay or to inner
surface of reactors and then will be rejected outside the wall. This problem is well known in
INDEX çç Sommaire
refining industry because it can lead to hydrogen induced disbonding (HID) phenomenon. In
the case of an actual reactor, with embedded defects, hydrogen pressure inside the defects,
due to hydrogen recombination, has to be also in equilibrium with local hydrogen content in
steel matrix. This internal pressure is relatively low under operation conditions but it can
increase very quickly during shutdown and then lead to extension of defects if local toughness
of constitutive steel (at given temperature and hydrogen content) is not sufficient to support
the stress level due to pressure increase.
Figure 21 hereafter summarizes the growth mechanism during shutdown of a pressure vessel.
As it is a dynamic phenomenon, the process described in the figure must be iterated for every
combination of pressure/temperature parameters achieved during cooling of the reactor.
Moreover, this mechanism must be repeated at each shutdown. Defect’s size evolution must
be assessed for MPT calculations described in the first part of this paper.
Shutdown procedure
Decrease of H2
solubility
Diffusion of H2 into
defect increase of Temper
Local H content embrittlement
pressure into defect
Growth of defect
KI<KIH+TE
No growth of defect
KOERS made a sensitivity analysis of several parameters in his paper [17]. A hydrogen
assisted crack growth model has been implemented in FEA software to take into account
diffusion of hydrogen into defects and then their stable extension during the cooling of
reactors. Various materials have been compared as well as various reactor geometries.
INDEX çç Sommaire
It appears that the worst configuration is very thick wall reactor for which a huge amount of
hydrogen is stored in the bulk and is able to migrate to defects. It also appears that
constitutive material of walls has a strong incidence on crack growth rate. Vanadium added
materials are much better than conventional grades, mainly due to trapping ability. These two
points are classical and are very close to conclusions that can be drawn on HID propensity.
The location of the has also a strong incidence on its criticism. It appears clearly that inner
side of the wall, roughly between the overlay interface and 30-50% of its thickness, is the
worst location for embedded defects. This is mainly due to the hydrogen content (see figure 7
– after cooling curve) that is very high in steel and then pressure inside defects can increase
very quickly to reach equilibrium.
Simulation of optimised shutdown procedure, with degassing step (no pressure inside the
reactor and maintain at service temperature) shows that it has only a limited effect on crack
growth incubation and propagation in the case of thick wall reactors. Walls are in general so
thick that degassing step should have very long duration to permit hydrogen to diffuse out of
the critical zone and reduce sufficiently to limit stable crack extension. The major problem is
then the economic cost of such procedure.
CONCLUSIONS
The synergistic effect of in-service ageing and embrittlement of steel material under high
temperature and hydrogen atmosphere conditions have been presented in this paper,
emphasizing the importance of taking into account effect of hydrogen even in low quantities
in the metal. A method has been described to assess the present-day service and maintenance
conditions to assure safe shutdown and startup cycles. It has also been demonstrated that a
knowledgeable steel supplier is able to deliver a material that will be much more resistant to
embrittling effects thanks to rigid control of the delivered metallurgy.
Some assumptions on effects of hydrogen on fracture or CVN toughness have also to be taken
into considerations and well validated, in particular the effect of high content of hydrogen on
fast fracture mechanics. Hydrogen embrittlement is a very complex phenomenon and very
few data are available. Then, particular attention must be paid to this matter.
MPT definition is then a complex matter and a lot of parameters have to be taken into
consideration. It must be understood that MPT definition changes regarding reactor history
and every cycle (shutdown / restart) must be analysed to assess MPT evolution. Presence of
defects in the wall should be controlled with accuracy as well as defect size. In the case of
strong conditions (high thickness, non-trapping material, high service H2 pressure and
temperature, quick cooling , embrittlement material, …) small defects can evaluate to bigger
size by stable (slow) propagation and then reach critical size regarding brittle failure.
INDEX çç Sommaire
BIBLIOGRAPHY
[1] C. CHAUVY, L. COUDREUSE, Ph. BOURGES, Consequences of PWHT requirements
on service properties for petrochemical grades, IIW 2008, Graz,
[3] American Petroleum Institute, API RP 934-A - Materials and Fabrication of 2 1/4Cr-1Mo,
2 1/4Cr-1Mo-1/4V, 3Cr-1Mo, and 3Cr-1Mo-1/4V Steel Heavy Wall Pressure Vessels for
High-temperature, High-pressure Hydrogen Service, May 2008
[6] G. R. PRESCOTT, Operating Issues of Aging Reactors, Joint Industry Project Aging
Hydroprocessing Reactors, Tech. Report n°6, December 1998
[8] T. L. ANDERSON, REACT - Software for Evaluating Aging Reactors, version 1.1,
User’s Manual, Joint Industry Project Aging Hydroprocessing Reactors, Tech. Report n°12,
May 2000
[9] T. SAKAI and al., Effect of Hydrogen on MPT and De-Hydrogenation During Shut Down
in Hydroprocessing Reactors, PVP – vol 344, High Pressure Technology, ASME 1997
INDEX çç Sommaire
[12] KOBE STEEL, Report on Fast Fracture Testing of Mid X-bar weld metal, JIP Meeting,
April 15th, 2004
[13] ASTM standards, ASTM E1921-05 Standard Test Method for Determination of
Reference Temperature, To, for Ferritic Steels in the Transition Range, 2005
[15] Development of a Fracture Toughness Quality System Tool Applicable for Steel
Producers and Steel Users of Heavy Plates, Profiles and Weldments – QUALYTOUGH”,
RFCS Contract No. 7210-PR-375, Final Report.
[16] American Petroleum Standard, API 579-1/ASME FFS1 2007 Fitness For Service, 2007
[17] R. W. J. KOERS, Hydrogen Assisted Crack Growth in the Wall of Pressure Vessels
Made of Conventional and V-modified Steels, Second Creusot Loire Industrie Workshop in
Burgundy – “Present Situation on Steels for Hydrogen Pressure
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page67
RÉSUMÉ / ABSTRACT
31/32B
PROPRIÉTÉS SPÉCIFIQUES DES TÔLES FORTES EN ACIERS ;
INDISPENSABLES A UN MEILLEUR RENDEMENT POUR LES FABRICANTS
D’ÉQUIPEMENTS SOUS PRESSION
RÉSUMÉ
Lors de la dernière décennie, les impositions pour les équipements sous pression dans le secteur de
l’énergie tout comme celui de la pétrochimie ont constamment augmenté. Le besoin d’améliorer
l’efficacité du process au moyen de températures et de pressions plus importantes a conduit à des
appareils plus lourds à parois plus épaisses.
Comme conséquence de ces conditions de service plus sévères, la sélection de matériaux chez les
designers et ingénieries est devenue un problème de métallurgiste. Le choix du matériau fait partie
intégrante du design de l’appareil mais aussi de l’analyse des risques préalable à effectuer par le
fabricant et/ou l’exploitant. Les résultats de ces analyses sont ainsi des spécifications pour la fourniture
du matériau.
L’aciériste doit réaliser la meilleure combinaison parmi toutes ces impositions afin d’assurer le meilleur
design d’acier possible en accord avec les spécifications. La marge de manœuvre entre les possibilités
technico-économiques des aciéristes et les exigences des spécifications est diminuée constamment, du
fait des impositions croissantes.
Le processus partant de l’appel d’offre jusqu’aux expériences dans les marchés sera analysé d’un point
de vue d’un fournisseur de plaques. De l’analyse détaillée des impositions lors de la phase de
consultation au processus d’assurance qualité, il sera expliqué comment un fabricant d’aciers de qualité,
avec ses suggestions, s’y prend pour améliorer l’efficacité de ce processus.
ABSTRACT
Within the last decade the material requirements for pressure equipment in the energy as well as the
petrochemical industry are steadily rising. The demand to increase the efficiency of the operating pro-
cess by means of higher operating temperatures and pressures leads to bigger and heavier equipment
with higher wall thicknesses.
As a consequence of more severe operating parameters, material selection has become moiré and more
an issue for metallurgists in design and engineering companies. Material selection is a part of pressure
vessel design and basically a function of operating conditions and anticipated risk analysis which has
to be carried out by the fabricator and / or the plant operator. This result in the material manufacturer
has to tailor the production process to match with the steel design the requirements of the specification.
Due to increasing requirements the gap between the techno-economic feasibility and the requested ma-
terial properties narrows constantly.
The process from bidding stage to final experiences in the markets is highlighted from the material ma-
nufacturers point of view. From the thorough checking of requirements in the inquiry stage to the quality
assurance system accompanying the production process it is described how the quality manufacturers
handle the challenges including suggestions for improved process efficiency.
AFIAP - ESOPE 2010 Programme
INDEX çç Sommaire
1. Introduction
Primary energy consumption in the world is still largely driven by the use of fossil
combustibles. This fact is not likely to change in the near future. In contrary expectations are
that the amount of fossil combustibles will rise in the future. Bases of the energy supply are
and will be oil and gas, which are largely used also as base for the production of e.g. synthetic
materials. The amount of oil and gas already consumed in the past and the actual demand
leads to the necessity of exploring new sources. Of course easy to exploit sources are used
first followed by those causing more technical and financial challenges. This means that it is
expected that the share of oil and gas that need special treatment before use will rise more and
more leading to more complex methods and more expensive installations. As the market
prices for oil and gas rise, sources formerly left out of production because of high cost
become attractive. The rather easy to exploit like e.g. onshore oil and gas fields in the Middle
East offer increasing challenges in the chemical composition as they contain a lot of harming
elements, which have to be removed requiring great effort in the refining process. Others
being more “suitable” in chemical composition are extremely hard to exploit, e.g. pre-salt,
arctic. Highly corrosive media (sour oil and gas feeds) as well as e.g. tar sands are nowadays
processed on a regular basis providing some challenges for the engineers concerning refining,
environmental hazards, installation safety, cost efficiency and so on. Investments are based on
various assumptions and are calculated very tightly to ensure the profitability. Only with
materials meeting the requirements in terms of quality and longevity the calculated efficiency
can be achieved.
It is obvious that with those conditions the requirements for materials are tough to comply
with and the quality oriented material producers will strive to supply their customers with the
materials they will need to achieve their goals. Therefore a rising necessity for specialised
respective customised material supply is seen in the future which calls for tailor made
solutions for each application to fit the special demands of that equipment. The quality steel
producers follow with their business conduct this trend. They will not offer unless the
requirements are thoroughly checked, ensuring the overall quality of the offered materials and
will also have a close look to their quality assurance system and certification practise. By
definition the testing of the test samples has to be representative for the material as a whole.
This paper will outline the challenges that come with demanding specifications and the
strategy to supply the customer in a co-operative manner for the long term mutual benefit of
the partners. All relevant steps along the order chain of a plate in the mill and their specific
requirements are commented from the bidding stage to the certification of the plate. Although
valid for all kinds of plates it is showcased for a HIC (hydrogen induced cracking) resistant
material as it enables to demonstrate how to deal with additional requirements.
INDEX çç Sommaire
specific values like tensile strength, yield strength, chemical composition, minimum heat
treatment to adhere to. The manufacturer will have to produce a material that complies with
all those design requirements at the same time.
But that is not all. Apart from those design requirements we find additional exigencies. Due to
H2S containing media in the relevant temperature-pressure range, there are requirements for
HIC testing in accordance with the standard test methods. Apart from that usually further
restrictions exist e.g. for hardness or chemical composition, that are more restrictive than
those already given by the design requirements. Reasons for that can be various. The operator
e.g. had bad experience with higher values in the past, which however are/were covered by
the standard and are not in any case linked with chemical composition but were provoked by
e.g. general quality problems. Often some sort of standardisation is a cause for such
restrictions or it is intended to keep carbon equivalent low for improved weldability. Such
additional restrictions may come from operating companies, engineering companies and from
vessel manufacturers. They are all reasonable requirements, but unfortunately they tighten the
allowed production window for the base material, consequently impairing the feasibility in the
context of all requirements.
All these requirements have to be checked thoroughly in the bidding stage in order to judge
the technical, or better said the industrial feasibility, of the requested product. In case the
production without restriction can be confirmed in the process the commercial offer based on
the pricing elements of the technical examination will be generated.
Especially for the steadily rising number of demanding specifications to handle e.g. sour
media, high pressures and so on the check for technological feasibility reveals more and more
often some restrictions that needed to be clarified as the specifications leave space for
interpretations. In other cases however the restrictions are that strict that the technological
feasibility is impaired, meaning the feasibility on a reasonable (for customer and
manufacturer) techno-economic basis is not given anymore. As a quality manufacturer will
always strive to clear such points as early as possible, he will mention such things in the
technical amendment to the commercial offer in order to get comments and deviations to the
specifications clarified before order placement. This is because any negotiation later on will
cause massive project delays or compromises that endanger safe operation of the equipment.
The typical occurrence of deviations to specifications in the pressure vessel business arises
usually in the stress field of chemical composition, toughness requirements, plate thickness
INDEX çç Sommaire
and requirements for post weld heat treatment (PWHT). In case of materials intended to be
used under sour service other restrictions (mainly additional restrictions for chemical
composition, and mandatory PWHT treatment) are added to the material profile. Figure 1
gives you some overview about the connection of the parameters.
It can be seen that as soon as there is a deviation from the specification the manufacturer is in
a difficult position. He wants to supply a high quality product. But he has also to keep in mind
that he has to ensure the material enables to complete the project to satisfaction, which is only
possible in case the whole material adheres to the specification and not only the material of
the test coupon. What are his options?
Agree to specifications although he knows the test coupon might be ok but a testing at other
locations would potentially reveal unpleasant results? – Absolutely no option for quality
producers!
The way is, although not convenient to both the manufacturer and all involved other parties,
to point out potential risks for the whole manufacturing chain of the pressure vessel and try to
work together to find a solution that will allow a save and economic realisation of the project.
Often only small alterations in the specifications lead to save constructions that adhere to all
the requirements of the production chain. But it is very important, that anybody affected feels
as comfortable as possible with the modifications that need to be done.
To overcome deviations some parameters normally would have to be altered in order to
improve the technological feasibility. For pressure vessels designed under ASME code
thickness and basic chemical restrictions can’t be altered as these are design requirements.
Usually toughness requirements can’t be changed as well, because as per minimum design
metal temperature the brittle failure of the vessel has to be excluded. The minimum PWHT
conditions are defined by the construction code as well. So the only parameters left where
there is a realistic option to adopt the material for feasibility are: PWHT exceeding the code
requirement, additional chemical restrictions and potential additional requirements set by the
specification that exceed those of the standards.
In a close co-operation between plate manufacturer and the involved parties it is usually not
difficult to come to a compromise to the mutual benefit of the parties involved, which
guarantees a save operation of the pressure vessel and leaves no doubt about a save
production at all stages that adheres to the initial time frame in case the discussions are carried
out at an early project stage. Issues like e.g. doubt about formability or weldability can be
resolved without much ado even if formerly fixed parameters of the specifications have to be
modified. The essential thing is that anybody in the production chain understands where the
other party is coming from and when this is done experience shows that often a compromise
is not far away, because steel manufacturer, vessel manufacturers or the operating/engineering
companies come up with some good alternative from their experience. In any case the parties
can learn from these action in order to enable further progress in the whole business conduct.
INDEX çç Sommaire
necessary. The production route is applied to obtain a homogenous HIC-resistance level over
the whole plate and consists of special technical treatments and improved quality assurance
measures. Figure 2 shows the main steps in the steel refining process.
The hot metal delivered from the blast furnaces undergoes a pre-desulphurisation process
before it is filled into the converter. In the converter process the contents of carbon,
phosphorous and nitrogen are brought down. Afterwards, in the ladle, the analysis adjustment
is performed. Next, together with adjustment of temperature of the heat for later casting a
rough adjustment of the chemical analysis is carried out.
Subsequently the fine adjustment of the chemical analysis under vacuum is done. Also under
vacuum conditions the removal of nitrogen, hydrogen and sulphur to extremely low levels is
carried out. For HIC-resistant plates some additional restrictions concerning chemical
composition have to be respected in order to achieve high HIC resistance. Unfortunately these
have direct influence on the combination of toughness level and tolerability of PWHT.
Measures usually considered to be advantageous like micro-alloying were proven to be
inappropriate in this case. The central compactness has to be created with high rolling forces
in order to achieve good toughness levels in combination with high HIC resistance and
simultaneously good tolerability of PWHT.
• Casting
During casting measures to avoid an oxygen and sulphur absorption in the liquid steel must be
applied. One of the main targets in casting – besides obtaining the initial geometry for rolling
– is to reach high cleanness in the steel. [3] Both continuous casting and ingot casting can be
used for the production of HIC-resistant plates, but require special know-how for the entire
process in order to control e.g. segregations and their detrimental effects.
In continuous casting vertical caster types show best results in regard to homogeneous low
level distribution of non-metallic inclusions and segregations, which is essential for HIC-
resistant steels. The reason is shown in figure 3. In vertical casting the cast strand will only be
bent after complete solidification. This allows non-metallic inclusions – mostly oxides – in
the liquid steel to float-up to the surface. In the curved caster type the inclusions get caught at
the solidus line in roughly quarter thickness of the slab. The centre line segregation and
shrinkage cavities can be minimised due to exact alignment of the rolls of the caster by the
use of force. This process is called soft reduction. [3] Plates from continuous casting can
nowadays be produced to a thickness of about 130mm respecting the code requirement of a
reduction ratio of minimum 3.0 to 1 (ASME SA20 5.3), which should be indicated in the test
certificates. Using ingot casting for HIC resistant plate production requires long
INDEX çç Sommaire
Figure 3: Differences between different continuous casting types
experience and special know-how. Again the avoidance of segregations is a key issue for this
production route applied for very thick plates. It is essential to have optimised geometry of the
ingot formats in order to produce a homogeneous microstructure in the final plate. A basic
necessity for this production route is a very strong rolling high stand to ensure also a
sufficient deforming in the plate core during rolling.
• Plate rolling:
After casting the slabs will be prepared for the following rolling process This is necessary in
order to cut the slabs to length for an optimised rolling process and to take only the cast strand
sections most suitable due to homogeneity reasons. The level of homogeneity is ensured by
keeping all of the monitored casting parameters in specially adjusted target windows.
After being reheated in the pusher furnaces to 1100 to 1250°C the slabs are rolled to plates.
The application of the “high shape factor rolling” ensures a complete deformation even in the
core of the final plate by using very high forming forces and creates a most homogenous
microstructure [4]. After hydrogen effusion of the plate the final step in the plate mill is the
normalising. However the full HIC resistance of the plate can only be obtained after a stress
relieving treatment. As normally the plates undergo a stress relieving after welding at
customer’s work shop this is usually not done in the mill.
• Quality assurance and testing
Parallel to the production for high sophisticated plates manufacturers have to run a highly
efficient quality assurance program in order to detect any deviation from the pre-scheduled
production plan which could create a risk of not meeting the product requirements. From the
whole steel refining process until the final delivery of the product monitoring of quality
parameters is applied to document the production quality. In case the target window for
certain parameters is not met, additional measures to ensure no major influence has been
generated are taken. E.g. during the continuous casting process more than thousand individual
values observe the conformity to the scheduled values. In case of detection of critical process
parameters automatically a standard increased test procedure is launched. The position in the
cast strand is marked and additional tests are run. Those areas not meeting the product
requirements are discarded from production. Not only the incident position undergoes such
testing, but also some positions before and after the marker are tested and get approval for
INDEX çç Sommaire
further use only in case the testing shows acceptable test results. Figure 4 shows this
procedure.
Dillinger Hütte operates a testing house and acceptance department which is not only ISO
9001 certified but also accredited to ISO 17025. This accreditation for test house, chemical
laboratories, acceptance department including non-destructive testing is so far unparalleled in
this extend in the heavy plate market, which shows the dedication to quality. Some of the
main differences between ISO 9001 and ISO17025 include e.g. mandatory estimation of
possible environmental influences on test results, documented calibration of test equipment in
regular intervals and documentation thereof, usage of certified reference materials, and so on.
In total the accreditation of the whole testing independent of the production brings the
certainty to the customer that each value shown in the test certificate has been tested under
most stringent conditions to state of the art procedures and measuring equipment. Together
with the strict production plan and the tight quality assurance system the overall quality of
the product in accordance to customer’s requirements is ensured.
INDEX çç Sommaire
for a small amount of time to reach the full HIC resistance of a normalised pressure vessel
steel. That is why HIC-resistant steels according to EN10028-3 (e.g. P275NH) work also as
typical PWHT temperatures for those materials are 560 to 580 °C. An other reason for high
PWHT temperatures are potential residual stresses, which might exceed in addition with the
applied stresses the yield stress of the material. As the level of residual stresses after a PWHT
of 600 °C is considered to be less than 60 MPa (it seems to be more realistic in the region of
30 to 45 MPa) this is no issue as well [5]. Additionally the gain in minimising residual stress
by rising the applied temperature is not worth the loss of mech.-tech. properties that are
bought with it.
Over the last years with rising demand for highly sophisticated tailor made products it can be
seen that apparently the number of companies offering to produce such materials is
increasing. For the customer this is a good thing as there seem to be more material sources.
But it is not all that easy to compete in this market as there is a necessity to operate special
installations and apply complex processes to satisfy customer’s requirements. Vessel
manufacturers recently reported an increasing number of quality problems for their projects.
These were e.g. resulting from returned orders of the material manufacturers, negotiations
trying to relax specification requirements after placing the order or the disclosure of mech.-
tech. values not matching the requirements during production although claimed to do so.
The effect of the latter can be shown with the example of HIC-resistant plates. There you can
distinguish between intentionally designed plates for this application and so called pseudo-
HIC plates. The results of the special production route on the product properties is shown in
figure 5. In this picture the HIC values obtained in the HIC test according to NACE TM 0284
between intentionally produced HIC resistant plates (dark coloured marking) and plates
specified with some (e. g. low Sulphur, low Phosphorus, ...) elements of the special
production route (ligth coloured marking) are compared. The so called pseudo-HIC plates
show a probability of only about 37% for obtaining values of less than 2% CLR in the HIC-
test.
INDEX çç Sommaire
Plates produced following the entire special production route (HIC resistant) are likely to
produce CLR values below 2% in approximately 88% of the cases. This means that the
special route provides not only clearly better results in regard to the HIC properties but also
the lead times are safer as the risk in failing the HIC test is very small thus avoiding
reproduction. In addition intentionally produced HIC-resistant plates offer these quality level
throughout the whole plate, whereas the pseudo-HIC plates are only tested at the coupon
taken and nobody exactly knows how the situation will be at any other location of the plate as
homogeneity has not been a concern during production.
Overall the customer has the choice between high quality products really fulfilling the
requirements of a specification or some materials that are claimed to do so but in the end of
the day show weaknesses at very critical points. It is recommended to not only have a look at
the offered price and the claimed properties. Ask the questions you need to have answered as
soon as possible in order to clarify the whole supply situation. As soon as leading plate
manufacturers show concerns about technological feasibility, there should be taken special
care about the intended order. There are many reported examples in the markets of massive
delays, returned orders, not acceptable vessels and even worse things due to bad supply
decisions in the early stages of projects. This clearly shows that the offered price level is not
the only criterion for buying decisions.
Consultation of the steel manufacturers is the key to specialised plate supply, if possible as
early as possible in the process. The quality steel manufacturers will be glad to assist in the
selection process for materials and come up with advise and solutions to concrete challenges.
In the context of the actual technological feasibility statements like “non-challenging
production prospects” and agreement to whatever the specification stipulates is not serious
business conduct and will most probably result in trouble. In case projects are carried out like
that hope for the best would the best bet.
Especially when a purchase due e.g. to emergency reasons has to be done at stockists, it is
strongly recommended to ask basic questions about the material offered as this material is not
necessarily designed for that specific purpose or due to the stocking concept some
compromises had to be taken. The serious stock partner will discuss concerns, suggests and
seeks the consultation with the manufacturer and will, if it is really is a very “hot” issue not
do the business just for business sake.
5. Conclusion
It has been shown that supplying heavy plates for current projects in the oil and gas industry
is a very challenging business. With the more demanding specifications that gain larger parts
in the markets also new market participants are working in this sector. Experience shows that
supplying for specialised projects needs to be handled with special care. The whole supply
chain has to adhere to thorough checking of requirements, strict design of the material to be
produced, following a defined production plan based on the manufacturers experience and
know-how - also willingness - to further develop processes. All this can only work with the
trusted co-operation between all parties involved. Often project delays result from unforeseen
incidents or initial concerns that were thought to be negligible. The laws of physics can not be
overruled and have to be respected. Therefore it is necessary to find manageable compromises
as soon as possible in the process in order to enable the timely production of vessels and put
into operation of equipment. It has been demonstrated how quality steel manufacturer analyse
requirements and comment in their technical offer to point out where potential risks to the
project may be hidden from clear sight. After this has been done the design is carried out and
production according to very strict technological parameters is done which is accompanied by
a well structured and state of the art quality assurance management. This quality assurance
INDEX çç Sommaire
management itself works with state of the art processes installations and measuring devices,
that have been accredited in accordance to EN17025 ensuring the documentation of the
produced materials shows the real quality level achieved by the production.
However it can be seen that not all suppliers work in accordance to that ethics and any
purchaser has to be careful about assessing commercial offers. It is a fine line to walk on
whether to choose the alternative being a bit more expensive but clearly shows all potential
risks and technical support or to solely go for the cheaper solution with a unknown risk. This
is in the first place not so much a problem to the purchaser, who’s work usually is measured
by the commercial facts, but at the end of the day his colleagues in the work shops have to
process the material and have to take the responsibility for potential failures. Often workshop
people made the experience that spending some more money with quality producers in the
first place saves money at the end of the day, because unnecessary taking of risk can be
avoided. Often those people are very thankful to receive direct technical product support,
advise which is offered in most cases only by quality manufacturers. Finally big and
challenging projects can only be accomplished under constant co-operation and trust between
the project partners for the benefit of every party involved.
References:
[1] I. Detemple: “ Alloying concepts”; Hydrocarbon Engineering; December 2003; Farnham
(England)
[2] F. Gottwalles, I. Detemple, P. Wolf: “Tailor made heavy plates for pressure vessels”; The
international Oil&Gas Engineer; February 2008; London
[3] K. Harste, J. Klingbeil, V. Schwinn, N. Bannenberg, B. Bergmann: “Die neue
Stranggießanlage der Dillinger Hütte als Vorstufe zur Erzeugung von Grobblechen mit
höchsten Qualitätsanforderungen”; Stahl und Eisen, February 2000;Verlag Stahleisen GmbH;
Düsseldorf; 2000
[4] A. Streißelberger, V. Schwinn: “ Grobblechherstellung aus verfahrenstechnischer Sicht”;
Stahl-Informations-Zentrum - Merkblatt 570 “Grobblech – Herstellung und Anwendung”,
page 7ff.; Düsseldorf, 2002
[5] M. Haneke, W. Middeldorf: “Einfluss des Spannungsarmglühens auf die mech.
Eigenschaften von Baustählen” DVS Bericht 49; S. 8-18, Düsseldorf 1978
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page68
RÉSUMÉ / ABSTRACT
19B
EXPÉRIENCE ACQUISE EN COMPORTEMENT DES ACIERS RÉSISTANT À LA
CORROSION DANS DES MILIEUX AQUEUX CONTENANT DU SULFURE
D’HYDROGÈNE – CONSÉQUENCES
DES IMPOSITIONS DE PWHT (POST WELD HEAT TREATMENT)
lormittal.com
Mercredi
RÉSUMÉ
Les équipements utilisés lors des opérations d’extraction et de raffinage du pétrole et du gaz sont souvent
exposés à des milieux aqueux contenant du sulfure d’hydrogène (H2Saqueux), ce composant étant
naturellement présent dans le pétrole brut et le gaz naturel. Une grande partie de ces équipements est
élaborée avec des aciers au Carbone-Manganèse susceptibles de présenter des fissures dues à
l’H2Saqueux.
Le sulfure d’hydrogène est responsable de différents types d’endommagement dans les aciers, désignés
par les termes anglais tel que « Hydrogen Induced Cracking » (HIC), « Sulfide Stress Cracking » (SSC),
and « Stress Oriented Hydrogen Induced Cracking » (SOHIC). Plusieurs paramètres influencent le
comportement des aciers au Carbone-Manganèse dans des milieux aqueux contenant du sulfure
d’hydrogène.
Cet article expose l’expérience acquise en comportement des tôles en aciers résistants à la corrosion
dans des milieux aqueux contenant du sulfure d’hydrogène, utilisés lors de la construction des appareils
à pression. L’influence de la composition chimique, des conditions de fabrication des aciers, des
traitements thermiques et de l’épaisseur des tôles sur la résistance à l’endommagement en milieu
H2Saqueux sont analysés. Des solutions pour des aciers optimisés résistants à la corrosion dans des
milieux aqueux contenant du sulfure d’hydrogène sont proposées.
ABSTRACT
Oil and gas treatment and petroleum refining processing equipments are often working in sour envi-
ronments containing wet hydrogen sulfide (H2S), which is a natural component of crude oil and natural
gas. A large part of this equipment is fabricated in Carbon-Manganese steels, which can be sensitive
to wet H2S cracking.
Hydrogen sulfide is responsible for numerous embrittlement phenomena in steels, such as Hydrogen
Induced Cracking (HIC), Sulfide Stress Cracking (SSC), and Stress Oriented Hydrogen Induced Crack-
ing (SOHIC). Some parameters influence Carbon-Manganese steel components behavior in sour service
conditions, where wet H2S cracking mechanisms can operate.
This paper exposes the experience achieved on sour service resistant steel plates for pressure vessel
construction. The effects of chemical composition, steel fabrication rout, heat treatment and steel plate
thickness on cracking resistance in sour service environments containing wet hydrogen sulfide are an-
alyzed. Optimized solutions for wet H2S service resistant steels are suggested.
INDEX çç Sommaire
1. INTRODUCTION
With the world oil and gas consumption growth trend, more and more gas and oil fields with H2S
content are exploited. Wet hydrogen sulfide cracking of steel components has been a major issue in the
petroleum refining and oil and gas industry processing for about fifty years.
A large share of oil and gas equipment is fabricated in Carbon-Manganese steels (according to ASTM
A516 Standard Specification for Pressure Vessel Plates), which could be sensitive to cracking phenomena in wet
H2S environment, as follows:
• Hydrogen Induced Cracking (HIC) or stepwise cracking [1] which manifests along the rolling
direction in pipelines and pressure vessel plates in the form of surface blisters and/or internal cracks in the
absence of an applied stress (Figure 1);
• Sulfide Stress Cracking (SSC) [2] occurring under the combined action of internal or applied
tensile stress and aqueous environments containing H2S at hard martensitic or bainitic areas of the base metal or
associated with weldments, including both the weld deposit and the heat affected zones; the cracks are
perpendicular to the tensile stress direction (Figure 1);
• Combination of HIC and SSC - Stress Oriented Hydrogen Induced Cracking (SOHIC) which can
be categorised as SSC caused by a combination of external stress and the local straining around hydrogen
induced cracks [3]. SOHIC usually occurs in the base metal, adjacent to the HAZ of a weld (zone of high
residual stresses from welding) or in a hard weld heat affected zone (Figure 1).
A lot of research has been done to develop materials with improved resistance to wet H2S cracking
phenomena. In spite of works performed and progress obtained development of wet H2S resistant steels still need
to be investigated.
Following service experience and research results, the National Association of Corrosion Engineer
(NACE) recommended in the NACE Standards MR0175 [4] and MR0103 [5] to select material according to
hardness requirements, and proposed a list of acceptable materials to be used in sour environments.
However, being included in the NACE MR0175 and MR0103 standards is not a guarantee for a good in-
service behavior. There is a typical example of the steels for pressure vessel where the hardness guarantee (less
than 22HRC or 248HV10) is not sufficient, because some additional parameters have to be taken into account to
insure the H2S resistance of material.
Industeel has done extensive research to solve wet H2S related problems in pressure vessel construction.
So, first this paper briefly outlines some theoretical considerations behind wet H2S cracking mechanism and then
highlights recent experiences and results obtained on sour service resistant steel plates for pressure vessels. The
influence of the metallurgical parameters, such as chemical composition, steel fabrication, microstructure and
INDEX çç Sommaire
heat treatments on cracking resistance in sour service media, are analyzed. Optimized solutions for wet H2S
service resistant steels are suggested.
In contact with aqueous environments containing H2S, steel corrodes producing hydrogen at its surface.
Sour corrosion in steel can be represented by the following anodic reaction:
Cracking
The H2S inhibits the hydrogen atoms from combining to form hydrogen gas and promotes their
absorption into the steel. At low temperature (T<200°C) hydrogen atoms can be trapped on defects of the
structure. The presence of hydrogen in the steel may cause embrittlement and possibly cracking.
Cracking in carbon steels occurs on a given defect in the structure when locally amount of hydrogen
(CH) is greater than a critical concentration for cracking (CK):
- Parameters acting on the amount of hydrogen CH that can be introduced into the material;
- Parameters acting on the critical concentration for cracking CK.
INDEX çç Sommaire
The parameters influencing H2S cracking of carbon steels are listed in Table 1. It is also indicated if the
parameter acts on CH or CK.
Parameters Influence on
Environmental
Mechanical
Metallurgical
The external parameters (environment, time, etc) act mainly on CH value, while metallurgical parameters
have an effect on both CH and CK. CH and CK are not independent values. For example, the chemical composition
influences not only CK values, due to effects on the microstructure, but also CH values because of
microstructure’s role in trapping mechanisms. CK values depend on the size and shape of inclusions, but CH
values depend also on the number and the distribution of these inclusions.
Following the precedent point, H2S related problems must be solved either by increasing of the CK value
for the material, or by decreasing of the amount of hydrogen CH trapped on potentially dangerous defects in the
structure.
The steelmaker can only act upon metallurgical parameters. In order to improve the resistance to H2S
cracking, steels must be developed with not only high CK values, but also with a low concentration of trapped
hydrogen for a given external hydrogen activity.
Metallurgical parameters such as chemical composition and cleanliness, steel fabrication, microstructure
and heat treatments influence C-Mn steels behavior in sour service conditions, where wet H2S cracking
mechanisms can operate.
INDEX çç Sommaire
3-1. Influence of cleanliness
• Influence on HIC behavior: It is now well known that hydrogen induced cracks can initiate on
defects in the structure where high hydrogen pressure can be developed. Among these defects, non-metallic
inclusions (endogenous or exogenous) are the most dangerous (figure 3). The reduction of the size and number
of inclusions reduces the number of crack initiation sites. On the other hand, globular inclusions are less
dangerous than elongated and aligned ones [6].
FIGURE 3 – Some examples of HIC cracks in C-Mn steels initiated on the non-metallic elongated inclusions
Elongated sulfide inclusions have been found to be the most dangerous ones (figure 4). The reduction of
the sulfur content leads to a decrease in the number and the size of sulfide inclusions. In addition, their shape
change and they become more globular [7]. This dependence is shown in Figure 5, where the sulfide shape
parameters are plotted against the sulfur content.
FIGURE 4 – Example of elongated sulfide FIGURE 5 - Influence of sulfur content and calcium
inclusions in a C-Mn steel structure treatment on the sulfide inclusions shape [8]
Calcium treatment is often proposed as an efficient way to obtain globular inclusions. The influence of
Ca treatment on the sulfide shape is also given in Figure 5. It appears that Ca treatment is effective for S content
superior to 0.002%. In the case of very low S content (≤0.001%) Ca treatment seems to be less effective because
of sulfide inclusions are rare and already globular.
The correlation between the HIC behavior and the sulfur content is exposed on Figure 6. Following this
figure the decrease in S content is effective in improving of HIC behavior.
INDEX çç Sommaire
FIGURE 6 – Correlation between the HIC behavior and the sulfur content [8]
Whilst for S content superior to 0.002% the Ca treatment is effective in HIC sensitivity reducing, for the
very low S content (≤0.001%) the situation seems to be different. The worst HIC results obtained in low S steels
are associated with high oxide content. Thus Ca treatment can be harmful for very low sulfur steels, because of
calcium high tendency to form oxides [9-10].
Oxide inclusions are a little bit less dangerous than sulfide ones. However little cracks can be found
initiating at oxides and it can be expected that when these cracks are numerous, the mechanical properties of
steels are affected even if the HIC criteria have been respected.
Following the above steels with improved HIC resistance are very clean steels with:
- Very low sulfur content (<0.002%);
- Very low oxygen content (<0.0015%).
• Influence on SSC behavior: The influence of sulfur content on the SSC behavior is illustrated
on Figure 7 where the parameter σ720 (applied stress) / YS (yield strength) has been reported versus the sulfur
content, for steels with high yield strength (YS > 400MPa). Results are shown for SSC specimens taken in three
directions – longitudinal, transverse and through thickness.
The correlation between sulfur content and SSC behavior is very marked for the through thickness
direction. In fact concerning this orientation, if there are hydrogen induced cracks, they will propagate easily
under the effect of stress [7]. The correlations between SSC behavior and sulfur content for the longitudinal and
INDEX çç Sommaire
transverse directions are not so good, the scatter is large. However the better results are obtained with the low
sulfur steels.
During SSC tests it has also been observed, that steels with high oxide content have a poor performance
compared to steels with a low oxide content (oxides act as crack initiation sites). In fact the SSC behavior
depends on both stress and defect orientations (Figure 8).
If the applied load is parallel to the major axis of the defect, the hydrogen induced cracks will not be the
cause of the final rupture. Otherwise, if the applied load is perpendicular to the defect major axis, hydrogen
induced cracks nucleated at inclusion tips will directly lead to the final rupture. Thus an applied load in the
through thickness direction will be very harmful. There is no effect of load direction on SSC behavior in case of
very low sulfur content (because of shape isotropy of inclusions - Figure 8).
• Influence on SOHIC behavior: SOHIC - a combination of HIC and SSC - usually occurs in the
base metal, adjacent to the HAZ of a weld, where there are high residual stresses from welding or in a hard weld
heat affected zone. So, all the above considerations work for providing of improved SOHIC behaviour.
After crack initiation, propagation clearly depends on the microstructure around the crack tip. So, it is
necessary to improve the microstructure of the steel. In particular, the segregation areas with the hard bainitic or
martensitic structure have to be avoided. Generally the sensitivity of steels to hydrogen embrittlement decreases
with the decreasing of microstructure hardness following the rank below:
Untempered martensite > Untempered bainite > Normalized structures > Tempered martensite
and bainite
Carbon-Manganese steels generally used for sour service have the normalized microstructures. The hard
microstructures sensitive to cracking can be found in C-Mn steel plates in heat affected zones of weldments
(relevant for SSC and SOHIC resistances) or due to the segregation of the elements relating to carbon equivalent
value (Ceq = C + Mn/6 + (Cr+Mo+V)/5 + (Cu+Ni)/15), such as carbon and manganese, etc., phosphorus
(relevant for HIC and SOHIC resistances).
INDEX çç Sommaire
FIGURE 9 – Influence of C content on segregation ratio of Mn [8]
Phosphorus also shows a strong segregation capability during solidification. The high C, Mn, P and
other elements relating to carbon equivalent value segregation areas thus formed during solidification could be
the potential sites for cracking as illustrated on Figures 10-11.
Tests have been performed in order to correlate the intensity of segregation to the HIC behavior. In
segregated areas a correlation between the C, Mn, P content and the occurrence of cracking has been found.
Qualitative electron microanalysis performed across the segregated areas of different plates (Figure 12)
shows that the main difference between cracked and uncracked areas is the segregation of phosphorus (Figure
13). The results on Figure 12 show the increasing of P segregation in cracked areas with increasing in carbon
content. Figure 14 illustrate the increasing of P segregation with increasing in C and Mn contents.
INDEX çç Sommaire
P content (10-3%)
40
P content (10-3%)
30
30
Cracked Uncracked
20
20
10 10
0 0
0 50 100 150 200 250 300 350 400
profile (microns) 0 50 100 150 200 250 300 350 400
profile (microns)
800
Uncracked
Segragated rates, [(Pseg-
Cracked
Pheat)/Pheat]*100%
600
400
200
0
80 114 142 165
Carbon content, 10-3%
The results presented on Figure 15 show the influence of P content on HIC results. These results are
issued from 111 low sulfur C-Mn steel plates without significant variation of other elements (C, Mn…) than P. It
is clear, that steels resistance to HIC is associated with P segregation. Then a decrease of the average amount of
P below 0.008% increases the HIC resistance.
4
Average CLR value, %
0
<6 [6;7[ [7;8[ [8;9[ >9
P 10-3 weight, %
INDEX çç Sommaire
The reducing of the amount of alloying elements such as C and Mn or more generally the carbon
equivalent (Ceq = C + Mn/6 + (Cr+Mo+V)/5 + (Cu+Ni)/15) content also decrease the segregations rates and the
risk of crack propagation.
But, in the case of normalized steels, to guarantee the required mechanical characteristics (especially for
thick plates), these elements cannot be reduced under a certain level.
The addition of microalloying elements, such as niobium and vanadium can allow a reduction of Ceq
content, but on the other hand the Nb and V precipitates (like Nb(C, N)) could be the potential sites for cracking
(figure 16). Moreover, as it will be explained below, microalloying elements additions have other consequences
on the hardness of heat affected zones.
The presence of segregation can be reduced by an optimized casting process. It is well known, for
example, that for plates obtained through continuous casting, the segregation is located at mid-thickness.
Cracking is effectively often found in this mid-thickness segregation. With ingot casting, segregation is mainly
found at the top of the ingot. An optimized cropping of ingot top can therefore reduce the presence of
segregation. It is also important to adapt the size and the shape of the ingot to the final size of the product.
A tempering of the plates, such as the tempering performed by the Post Weld Heat Treatment (PWHT),
has also a positive effect on HIC resistance of segregated areas through diffusion of elements and reduction of
potential sites of cracking. Typically, ASTM A516 Standard Specification for pressure vessel plates requires
minimum tempering temperature of 595°C (1100°F).
INDEX çç Sommaire
• Influence of hardness variation (HAZ) on SSC and SOHIC behaviors: Heat affected zones are
considered to be very sensitive to SSC and subsequently to SOHIC phenomena. Particularly hard bainite or
martensite coarse-grained areas are the potential sites for cracking initiation. This is illustrated on Figure 17,
where the threshold for stress sulfide cracking (σ720h) is given as function of the maximum hardness in the
welded area. The results obtained on two steels are given, an A516 grade 70 and a BS 1501-225 grade 490
(which is Nb modified A516 grade 70 steel). The trend shows that hardness is one of the main parameters for the
resistance to SSC of welded areas.
450
BS1501.225
400
A516 gr.70
SSC threshold stress, MPa
350
300
250
200
150
100
50
0
100 150 200 250 300 350 400
Maximum hardness, Hv 10
It is clear that for hardness levels higher than 250HV10, the SSC (and subsequently SOHIC) sensitivity
increases. For this reason to avoid sulfide stress cracking (and stress oriented hydrogen induced cracking) a
maximum hardness of 22HRC (248HV10) is requested in heat affected zones (according to NACE Standard
MR0175 requirements for materials to be used in sour environments). There are several solutions to achieve this
goal.
- The decreasing of carbon and manganese contents (and reduction of Ceq) allows to obtain a maximum
hardness of 22HRC (248HV10) in HAZ in the as-welded conditions. The Figure 18 shows some limits in
chemical composition of carbon-manganese steels to insure this result in the as-welded heat affected zones.
0,16
10kJ/mm
0,14
20kJ/mm
Carbon content, %
0,12
0,1
0,08
0,06
0,04
0,02
0
0,4 0,6 0,8 1 1,2 1,4 1,6 1,8
Manganese content, %
INDEX çç Sommaire
These carbon and manganese content limits are obviously insufficient to guarantee the required
minimum mechanical characteristics for slightly thick plates. To obtain a maximum hardness of 248HV in the
HAZ another solution exists.
- Post Weld Heat Treatment is effective in reducing of the hardness in HAZ. A PWHT tempers the
obtained microstructures, homogenizes the properties in the different parts of the material and relieves the stress.
The choice of steel has a strong effect on the PWHT behavior. This point appears clearly on Figures 19-
20, where the influence of the PWHT conditions on HAZ maximum hardness is shown for an A516 grade 70 and
a BS 1501-225 grade 490 (Nb modified A516 grade 70 steel).
380 380
0,99kJ/mm 1,98kJ/mm 1,35kJ/mm 1,85kJ/mm
2,78kJ/mm Max allowed HV 2,7kJ/mm Max allowed HV
HAZ max hardness (HV5)
340 340
260 260
248HV 248HV
220 220
180 180
As welded 600°C/3h 625°C/3h 650°C/3h As welded 600°C/3h 625°C/3h 650°C/3h
PWHT conditions PWHT conditions
FIGURE 19 – Influence of the PWHT conditions FIGURE 20 – Influence of the PWHT conditions
on the HAZ maximum hardness on the HAZ maximum hardness
of A516 grade 70 steel of BS 1501-225 grade 490
The heat affected zones of the microalloyed grades (with vanadium or niobium such as BS 1501-225
grade 490) are less sensitive to the tempering effects and it appears difficult to aim the 248HV10 level required
by NACE Standard MR0175 even for a high temperature (650°C). On the other hand, for carbon-manganese
grades (such as A516 grade 70) the PWHT at 600°C is sufficient to achieve the 248HV10 hardness level.
• Influence of heat treatment on SOHIC behavior: Following the previous point, the resistance
to SOHIC of welded areas is strongly correlated to the hardness level. The recent experiences show that the
favorable heat treatment and microstructure obtained can provide the increasing of SOHIC resistance. This is
illustrated on Figure 21, where SOHIC test results have obtained on an A516 grade 70 in following heat
treatment conditions:
- normalized;
- normalized and PWHT;
- quenched and tempered;
- quenched, tempered and PWHT.
INDEX çç Sommaire
45 A516 gr 70 N
40 A516 gr 70 N + PWHT
35 A516 gr 70 Q + T
TCL value, %
30 A516 gr 70 Q + T + PWHT
25
20
15
10
5
0
260 300 400 450
Stress level, MPa
It is clear from the Figure 21 that quenched and tempered or quenched, tempered and PWHT
microstructures (fine bainitic microstructure) allow to obtain best SOHIC results in comparison with normalized
or normalized and PWHT ones (ferrite and pearlite microstructure). Even for the high stress levels applied any or
very low SOHIC cracking is observed for quenched and tempered or quenched, tempered and PWHT A516
grade 70 steel.
The analysis of the parameters influencing steel behavior in sour service has allowed Industeel to bring
the answer to H2S related problems in pressure vessel construction and to suggest the optimized solutions for wet
H2S service resistant steels. Following all previous considerations, the real improvement of H2S resistance of
carbon-manganese steels can be provided by:
Effective techniques have been developed by Industeel in order to satisfy these requirements and to
suggest the optimized solutions for wet H2S service resistant steels.
The steelmaking using Heated Ladle Refining Process under vacuum, after initial melting of scrap in an
electric arc furnace, enables Industeel to obtain very clean high quality carbon-manganese steels with:
INDEX çç Sommaire
- Very well controlled chemical composition and optimized alloy elements addition for
guarantee low value of Ceq (but sufficient to achieve the required minimum mechanical characteristics) for the
thicker plates range.
The steels are obtained by well controlled continuous casting, or by casting in ingot moulds under argon
protection. Optimized ingot shapes have been developed in order to control the solidification conditions and to
limit the occurrence of segregation that affects the H2S resistance. After rolling the top of obtained from ingots is
cropped in order to eliminate the macrosegregation.
In the final state, after forming and welding operations, a well defined PWHT of the entire fabrication is
strongly recommended by Industeel in order to avoid any hard structures. For carbon-manganese grades
produced by Industeel according to ASTM A516 Standard Specification, PWHT between 600 and 610°C is
enough to achieve the 248HV10 hardness level.
These effective techniques have allowed Industeel to solve H2S related problems in pressure vessel
construction and to suggest the optimized solutions for wet H2S service’s resistant steels.
Industeel produces and supplies the very clean high quality C-Mn CarElso® [12] steel grades up to
280mm of thickness for wet hydrogen sulfide service from both continuous and ingot casting ways.
The steels are available in normalized (for HIC/SSC resistance) or in quenched and tempered conditions
(for higher strength and/or enhanced resistance against the most severe H2S conditions of the Stress Oriented
Hydrogen Induced Cracking). The following characteristics of the steel are guaranteed:
- Chemical composition according e.g. to ASTM A516 Standard Specification for Pressure Vessel
Plates;
- Perfectly balanced low carbon equivalent value from 0.38 up to 0.45 (for very thick plates) in order to
minimize the PWHT needed (Table 2):
- High level of HIC quality (average CLR value inferior to 5% - Figures 22-23);
- Additional guarantees of HIC quality in case of specific customer request (average CLR and CLR max
per specimens values inferior to 5%) - CarElso® HIC Premium+ standardized steel offer in the thickness range from 6 to
105mm.
INDEX çç Sommaire
70,0
60,0
Results on 749 plates
50,0
% of values
40,0
30,0
20,0
10,0
0,0
0 ]0;1] CLR max per specimen
]1;2] ]2;3] ]3;4] ]4;5] Average CLR values
]5;7] ]7;10]
>10
CLR range
70,0
60,0
Results on 358 plates
50,0
% of values
40,0
30,0
20,0
10,0
0,0
0 ]0;1]
]1;2] CLR max per specimen
]2;3]
]3;4] Average CLR values
]4;5]
]5;7]
]7;10]
CLR range >10
FIGURE 23 – Statistic HIC test results for C-Mn steel plates with thickness
range from 80 to 280mm obtained by casting in ingot moulds
- Sulfide Stress Cracking resistance guarantee without further tests (thanks to maximum hardness level
248HV10 in heat affected zones after Post Welded Heat Treatment between 600 and 610°C according to NACE
INDEX çç Sommaire
Standards MR0175 [4]). However the additional SSC tests could be performed upon customers request with the
agreed applied stress given by customer specification;
- Special steel grade CarElso®70 SOHIC [12] available only in quenched and tempered conditions with
the Stress Oriented Hydrogen Induced Cracking resistance guarantee (maximum TCL level of 15% with the test
load at 80% of actual yield strength value);
- For offshore applications where pressure vessels weight is particular stringent and a more significant
reduction of thickness and weight are needed, a new high strength steel grade SuperElso®500 HIC [12] (yield
strength of 500MPa) with optimized welded properties and resistance to H2S (average CLR value inferior to 5%
and SSC resistance guarantee) has been developed in accordance with the usual sour service requirements for
lower grade steels. This grade has been patented.
5. CONCLUSIONS
For better resistance of pressure vessels to embrittlement phenomena such as Hydrogen Induced
Cracking (HIC), Sulfide Stress Cracking (SSC), and Stress Oriented Hydrogen Induced Cracking (SOHIC) in the
sour environments containing wet hydrogen sulfide, a lot of metallurgical parameters must be taken into account
by the steelmaker.
The extensive research done by Industeel in order to solve hydrogen sulfide related problems in pressure
vessel construction has shown that the real improvement of H2S resistance of carbon-manganese steels can be
provided by:
- Improvement of steel cleanliness (reduces of sulfur, oxygen and phosphorus contents,
very low inclusion content);
- Improvement of steel chemical composition homogeneity (decreases of segregations
presence) and the steel microstructure (definition of the right post weld heat treatment in order to avoid any hard
structures in welded areas and keep the hardness level below 250HV).
The analysis of the parameters influencing steels behavior in sour service has allowed Industeel to bring
the answer to hydrogen sulfide related problems in pressure vessel construction and to suggest the optimized
solutions for wet H2S service resistant steels. Taking into account all the above considerations, Industeel
produces and supplies the very clean high quality carbon-manganese CarElso® steel grades up to 280mm of
thickness for wet hydrogen sulfide service according to ASTM A516 Standard Specification.
6. REFERENCES
1. NACE Standard TM0284 (latest revision), “Evaluation of Pipeline and Pressure Vessel Steels for
Resistance to Hydrogen-Induced Cracking”
2. NACE Standard TM0177 (latest revision), “Laboratory Testing of Metals for Resistance to Sulfide
Stress Cracking and Stress Corrosion Cracking in H2S Environments”
3. NACE Standard TM0103 (latest revision), “Laboratory Test Procedures for Evaluation of SOHIC
Resistance of Plate Steels Used in Wet H2S Service”
4. NACE Standard MR0175 (latest revision), “Petroleum and Natural Gas Industries – Materials for Use
in H2S-containing Environments in Oil and Gas Production”
5. NACE Standard MR0103 (latest revision), “Materials Resistant to Sulfide Stress Cracking in
Corrosive Petroleum Refining Environments”
6. C. Zmudzinski, L. Coudreuse, J. Charles, “Le phénomène de HIC : point des connaissances et
résultats actuels; garanties possibles”, Rapport Interne, CRMC, 1985
7. V. Lemoine, G. M. Pressouyre, “Fragilisation par H2S des aciers C-Mn fiblement allies”, Rapport
Interne, CRMC, 1984
8. L. Coudreuse, J. Charles, “Hydrogen Embrittlement Mechanisms; Influence of Metallurgical
Parameters”, Internal Report, CRMC, 1998
INDEX çç Sommaire
9. S-R. Chen, F-I. Wei, H-Y. Liou, C-L. Chou, “Inclusion Shape Control and Its Application on the
Property Improvement of Steel Plate”, China Steel Technical Report, N°4, 1990, pp. 13-23
10. M. F. J. Galis, G. C. Guntz, “Study of Metallurgical Parameters Influencing the Behavior of Line
Pipes in H2S Medium”, Corrosion 88, Paper N°195, March 21-25, 1988
11. T. Taira, K. Tsukaka, Y. Kobayashi, M. Tanimura, H. Inagaki, N. Seki, “HIC and SSC Resistance of
Line Pipes for Sour Gas Service”, Nippon Kokan Technical Report, N°31, 1981
12. Industeel ArcelorMittal Technical Data, “CarElso®”, www.industeel.info
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page69
RÉSUMÉ / ABSTRACT
91B
MATÉRIAUX EN TITANE POUR LES APPLICATIONS CHIMIQUES
K.-M. Koch
(ThyssenKrupp VDM GmbH – Plettenberger Str. 2 – 58791 Werdohl – Allemagne -
e-mail : [email protected]
RÉSUMÉ
La production industrielle de titane, dont on peut situer le point de départ dans les années 50, s’est développée tant
et si bien qu’il est devenu impensable de concevoir l’industrie chimique actuelle sans elle. En raison de sa
remarquable résistance à la fois à la corrosion et à l’érosion, cela même à de hautes températures, et à sa bonne
capacité à être travaillé, le titane représente la solution idéale pour de nombreuses applications dans l’industrie
chimique.
Le titane fait preuve d’une excellente résistance à la corrosion, tout particulièrement en présence de milieux
oxydants. De ce fait, les ingénieurs et les concepteurs sont en train de réaliser que le titane a pour effet de réduire
de manière considérable les coûts liés au cycle de vie pour une large étendue d’appareils et de procédés chimiques.
Les grades de titane pur, d’utilisation commerciale, sont particulièrement adaptés pour la fabrication d’appareils,
d’autant que les résistances maximales couvertes par celles-ci vont de 290 à 740 MPa. Les grades de titane
faiblement alliés ont, quant à eux, une meilleure résistance à la corrosion uniforme et caverneuse. Ils sont
recommandés pour des applications avec des milieux légèrement réducteurs ou des milieux fluctuant entre oxydant
et réducteur. En ce qui concerne la fabrication d’appareils, le soudage TIG (Tungsten Inert Gas) est le processus
le plus utilisé pour le titane, et les procédures ainsi que l’équipement sont généralement similaires à ceux utilisés
pour le soudage des aciers inoxydables. La bonne ductilité des grades de titane pur et faiblement alliés vient de sa
faible sensibilité à l’écrouissage.
Cette présentation donne un aperçu sur la résistance à la corrosion du titane, sur les avantages que cela donne pour
les applications chimiques, sur les matériaux en titane utilisés et quelques informations sur la fabrication.
ABSTRACT
The industrial production of titanium, beginning in the 1950’s, has been further developed so successfully that the
modern chemical process industry would be inconceivable without titanium. Owing to its outstanding corrosion
and erosion resistance and its high strength even at elevated temperatures and in addition to its good workability
titanium is the ideal solution for numerous applications throughout the chemical process industry.
Titanium displays excellent corrosion resistance particularly against oxidizing media. Due to this fact engineers
and designers are discovering that titanium significantly reduces lifecycle costs in a wide range of equipment and
chemical processes. For the fabrication of equipment the commercially pure titanium grades are particularly
suitable and cover a range of tensile strength from 290 – 740 MPa. The low alloyed titanium grades have a better
resistance to general and crevice corrosion than the commercially pure grades and they are recommended for
applications involving environments that are moderately reducing or that fluctuate between oxidizing and reducing.
In the fabrication of equipment the tungsten inert gas welding is the most applied process for titanium and the
procedures and equipment are generally similar to those used for welding stainless steel. The good formability of
commercially pure and low alloyed titanium is based on the low susceptibility to strain hardening.
This presentation gives an overview about the corrosion resistance of titanium, advantages resulting of this in
chemical applications, the used titanium materials and some information about fabrication.
AFIAP - ESOPE 2010 Programme
INDEX çç Sommaire
MATERIAUX EN TITANE POUR LES APPLICATIONS CHIMIQUES
M. Koch
(ThyssenKrupp VDM GmbH – Plettenberger Str. 2 – 58791 Werdohl – Allemagne -
e-mail : [email protected]
RESUME
La production industrielle de titane, dont on peut situer le point de départ dans les années 50,
s’est développée tant et si bien qu’il est devenu impensable de concevoir l’industrie chimique
actuelle sans elle. En raison de sa remarquable résistance à la fois à la corrosion et à l’érosion,
cela même à de hautes températures, et à sa bonne capacité à être travaillé, le titane représente
la solution idéale pour de nombreuses applications dans l’industrie chimique.
Le titane fait preuve d’une excellente résistance à la corrosion, tout particulièrement en
présence de milieux oxydants. De ce fait, les ingénieurs et les concepteurs sont en train de
réaliser que le titane a pour effet de réduire de manière considérable les coûts liés au cycle de
vie pour une large étendue d’appareils et de procédés chimiques. Les grades de titane pur,
d’utilisation commerciale, sont particulièrement adaptés pour la fabrication d’appareils,
d’autant que les résistances maximales couvertes par celles-ci vont de 290 à 740 MPa. Les
grades de titane faiblement alliés ont, quant à eux, une meilleure résistance à la corrosion
uniforme et caverneuse. Ils sont recommandés pour des applications avec des milieux
légèrement réducteurs ou des milieux fluctuant entre oxydant et réducteur. En ce qui concerne
la fabrication d’appareils, le soudage TIG (Tungsten Inert Gas) est le processus le plus utilisé
pour le titane, et les procédures ainsi que l’équipement sont généralement similaires à ceux
utilisés pour le soudage des aciers inoxydables. La bonne ductilité des grades de titane pur et
faiblement alliés vient de sa faible sensibilité à l’écrouissage.
Cette présentation donne un aperçu sur la résistance à la corrosion du titane, sur les avantages
que cela donne pour les applications chimiques, sur les matériaux en titane utilisés et quelques
informations sur la fabrication.
ABSTRACT
The industrial production of titanium, beginning in the 1950’s, has been further developed so
successfully that the modern chemical process industry would be inconceivable without
titanium. Owing to its outstanding corrosion and erosion resistance and its high strength even
at elevated temperatures and in addition to its good workability titanium is the ideal solution
for numerous applications throughout the chemical process industry.
Titanium displays excellent corrosion resistance particularly against oxidizing media. Due to
this fact engineers and designers are discovering that titanium significantly reduces lifecycle
costs in a wide range of equipment and chemical processes. For the fabrication of equipment
the commercially pure titanium grades are particularly suitable and cover a range of tensile
strength from 290 – 740 MPa. The low alloyed titanium grades have a better resistance to
general and crevice corrosion than the commercially pure grades and they are recommended
for applications involving environments that are moderately reducing or that fluctuate
INDEX çç Sommaire
between oxidizing and reducing. In the fabrication of equipment the tungsten inert gas
welding is the most applied process for titanium and the procedures and equipment are
generally similar to those used for welding stainless steel. The good formability of
commercially pure and low alloyed titanium is based on the low susceptibility to strain
hardening.
This presentation gives an overview about the corrosion resistance of titanium, advantages
resulting of this in chemical applications, the used titanium materials and some information
about fabrication.
INTRODUCTION
Whereas in the early 1960s the non-aircraft uses of titanium represented only about 5 to 6% of
the market while the relative proportions today are nearly 50:50. The total amount of titanium
produced worldwide in 2009 was approximately 80,000 tons and the market shares were 49%
in aerospace applications and 44% in chemical applications. The balance is applications in
other areas like medicine, architecture, automotive, sporting goods and consumer goods. The
outstanding properties of titanium which use to advantages in all areas of application are:
High strength (500 – 1500 N/mm²)
Low density ( 4.5 g/cm³)
Outstanding biocompatibility
Low modulus of elasticity
Excellent corrosion resistant
Good processing characteristics
High availability
Good decorative properties
Good image
The excellent corrosion resistance and the good workability like weldability and formability
are the main advantages of titanium for the use in the chemical process industry.
CORROSION RESISTANCE
Even at room temperature, titanium forms an adhering, nonporous and self-healing oxide film
on its surface, the so-called passive layer. Therefore the outstanding protection against
corrosion of titanium can be expected in all media supporting the formation of this layer and
not decomposing it. In general this is the case in oxidizing and neutral media. Scratches and
defects in the layer are rapidly healed if oxygen is available in the surrounding environment.
Free oxygen is not necessary. The presence of small amounts of water is sufficient to heal a
scratch.
Thus titanium has a superb corrosion resistance to sea water, brine, polluted inland and
brackish waters. The material is very resistant to oxidizing acids in a broad range of
concentrations and temperatures. Examples are nitric acid, chromic acid and acetic acid. Also
titanium is not attacked by wet chlorine gas or by solutions of chlorine compounds such as
sodium, chlorate, chlorite and hypochlorite.
Titanium has a relatively poor corrosion resistance in reducing environments. Reducing acids
will break down the oxide film. Titanium can be used in sulphuric acid, hydrochloric acid and
phosphoric acid at low concentrations and temperatures.
Titanium is strongly attacked by hydrofluoric acid under all conditions and should not be used
in environments where hydrofluoric acid is present. Strong corrosion, even to the extent of
INDEX çç Sommaire
inflammation, is also observed in dry chlorine gas with water contents below 0.3%, and in
dehydrated red fuming nitric acid.
The generally excellent corrosion behaviour of titanium in oxidizing media could also occur
in reducing media that contain small quantities of oxidizing agents, so-called inhibitors. Thus
titanium is resistant to attack in mixtures of sulphuric/nitric and hydrochloric/nitric acids and
even in hydrochloric acid containing free chlorine.
A similar effect can be achieved by alloying titanium with palladium, nickel-molybdenum or
ruthenium, when the anode potential of the material is increased and the resistance to
corrosion of these titanium grades under reducing conditions improved.
A summary of all these fundamentally facts about the corrosion resistance of titanium is
shown in figure 1.
Special mention must be given to the corrosion resistance of titanium to chloride ions in
inorganic media such as salt soles, sea water and moist, gaseous chlorine as well as in many
electrolytic solutions. Pitting corrosion is limited to a few chlorides in high concentrations,
temperatures above 100°C and low ph-values. The susceptibility of titanium to crevice
corrosion is slight and can be widely avoided when gasket materials of the fluoride
hydrocarbon type like Teflon are not used in titanium equipments. Teflon gaskets can react
with the process media to build hydrofluoric acid and attack the titanium.
The commercially pure titanium grades and the low alloyed titanium alloys are generally not
susceptibility to stress corrosion cracking. However, attention must be paid to intercrystalline
stress corrosion cracking in dehydrated methanol. Additions of water to methanol inhibit
stress corrosion cracking and the necessary amount of water for inhibition increases with the
halogen content in the liquid.
INDEX çç Sommaire
Titanium also is very resistant against erosion and flow-induced corrosion. The erosion
resistance of titanium grade 2 is comparable to that of stainless steel and can tolerate flow
velocities up to 30 m/s or more.
In galvanic compounds titanium, whose potential in its passive state is only exceeded by the
noble metals, will not be attacked. In general, no serious galvanic corrosion problems will
result from coupling titanium to 18-8 type passive stainless steels, nickel, Monel, Inconel or
Cupro-nickel alloys in passivating environments. Corrosion of metals in the “middle range”
of the series like copper, brass, tin, lead and certain steels may or may not be a problem
depending on specific environmental factors. Aluminium, magnesium, iron and cadmium are
expected to show significantly accelerated attack where coupled to titanium.
Titanium’s natural corrosion resistance can give it a virtually unlimited lifespan in equipment
subject to aggressive acids, salt solutions alkalis and waters. It can reduce maintenance and
downtime, extend equipment life, and improve plant performance, so it frequently offers the
highest value from a life cycle point of view.
Titanium’s strength, erosion adaptability and resistance to higher temperatures and pressures
allow thinner equipment walls and higher flow rates for better operational heat transfer and
potentially increased production. The result is an improved plant performance.
The fabrication of titanium does not need special equipment. Conventional metal processing
tools and procedures can be used to form, machine and join. The requirements in welding and
forming are comparable to stainless steel.
The availability of titanium is good. It is widely found in all common mill product forms and
a variety of grades to meet the specific requirements of the chemical apparatus engineering.
Besides great corrosion resistance, titanium also displays lower electrical and thermal
conductivity than steel. Its low thermal expansion results in reduced thermal stresses in
titanium structures. In applications requiring good spring properties, titanium’s low modulus
of elasticity is a distinct advantage.
Special titanium grades displaying deep-drawing properties have also been developed for use
in plate heat exchangers: they are cold-rolled and fabricated in such a way as to guarantee
good stamping properties with an optimum tensile-to-yield strength ratio. Plate heat
INDEX çç Sommaire
exchangers fabricated from this material withstand high operating pressures and
hydrodynamic stress. Owing to the extremely high corrosion resistance such products are
suitable for applications where they are exposed to corrosive media that would rapidly destroy
steel and special brass grades by pitting.
Titanium qualities can be divided into three main groups: commercially pure titanium with
different contents of oxygen, low alloyed titanium also with different contents of oxygen plus
small contents of palladium, nickel-molybdenum and ruthenium, as well as titanium alloys.
The pure and low alloyed titanium grades are especially used due to their attractive corrosion
resistance and therefore mainly in the chemical process industry. The titanium alloys are
usually chosen because of their high strength-to-weight-ratio mostly in the aerospace industry
and only for special applications in the chemical industry. The different titanium qualities
have been standardized in many countries. The American ASTM and the German DIN
standards are most commonly used by the European apparatus engineering.
Table 1 gives an overview of the titanium grade produced by ThyssenKrupp VDM including
the material identification according to ASTM and DIN standards and the material number
according to the German aerospace standard (for comparison only).
The following is a brief characterization of the various qualities (material identification acc. to
ASTM / DIN).
Grade 1 / 3.7025
The lowest strength pure titanium with low oxygen content. This quality is used for highly
formed parts where cold workability is critical, including expanded metals, as well as for
deep-drawing and cladding.
INDEX çç Sommaire
Grade 2 / 3.7035
A moderate strength pure titanium with higher oxygen content than grade 1. This quality is by
far the most commonly used commercially pure grade as it has the best combination of
strength, weldability and formability. The strength is similar to austenitic stainless steels.
Grade 3 / 3.7055
A higher strength pure titanium with higher oxygen content than grade 2. This quality has a
reduced cold formability compared with grade 1 and 2 and is used for thinner wall pressure
vessels and tube sheets in heat exchangers.
Grade 12 / 3.7105
This titanium quality is alloyed with 0.8% Ni and 0.3% Mo with strength and formability near
to grade 3. These elements are added in order to improve the crevice corrosion resistance at
higher temperatures and also the general corrosion resistance. The corrosion resistance in
reducing environments is approximately in the middle of grade 2 and grade 7. At elevated
temperatures this grade performs better than grade 3.
The mechanical properties of the above described titanium grades are determined by the
chemistry, not by mechanical working or heat treatment. Table 2 shows the comparison of
strength for the titanium grades and the effect of the chemical composition on the mechanical
properties. Basis of the values in the table is the DIN standard.
INDEX çç Sommaire
WELDING
Since titanium has a high affinity to the atmospheric gases oxygen, nitrogen and hydrogen, it
cannot be welded by processes in which molten metal may come into contact with these
elements. An oxyacetylene welding is therefore impossible and the welding has to be carried
through in an inert gas atmosphere or in vacuum.
The inert-gas-shielded arc processes are chiefly employed for fusion welding. Thereof the
argon-arc process (TIG/MIG) is best suited for the welding of titanium. In the industrial
practice the tungsten inert gas welding (TIG) is the most applied process in the fabrication of
chemical equipment. A heat treatment after welding is normally not necessary.
During the welding process it has to be ensured that not only the weld, the heat-affected zone
and the back of the weld, but also the bordering area of the base material is protected from the
atmosphere by suitable shielding device until the temperature has dropped below 300°C.
In correctly performed welded joints the mechanical properties of the weld are close to those
of the base material. A slight decrease in ductility may occur due to the different
microstructure. Under passivating conditions titanium welds exhibit the same corrosion
resistance as the base material. Only in reducing media corrosion action exceeds the base
material so that the weld may be exposed to more severe corrosive attack.
Careful preparation of the joints is a prerequisite to achieving sound welds. To avoid porosity
in the weld it is good practice to remove - immediately prior to welding - all surface
impurities from the area of the weld by grinding with emery, brushing and pickling. Since
even the presence of finger prints will result in hardening of the weld area, the area should be
cleaned just before welding with a suitable solvent that leaves no residues.
Welds on sheets of up to 2.5 mm gauge may be composed of a single layer. Beyond this
thickness two or multi-layer welding should be employed, in which case the area of the weld
should be thoroughly cleaned after the deposition of each layer to remove any temper
colouring. This also applies to tack welds whose number should be kept to a minimum when
designing the welding fixture. To avoid local oxygen concentrations, oxygen products should
be removed whenever welding is resumed after an interruption and the tip of the filler rod
should be cut off if it shows signs of oxidation.
As filler rod an uncoated wire of the same grade or possibly of one grade lower than the base
material should be used. If titanium of different grades should be welded the choice of the
filler rod will be decided by the basic criterion of requirement of the weld like strength,
ductility, operating temperature or corrosion resistance.
Particularly a tungsten electrode with a content of 1 or 2% thorium should be used and the
welding has to be carried out with direct current and negative electrode. As inert gas argon
> 4.6 should be used, which has a purity of at least 99.99%. Also the moisture content of the
argon, not generally indicated in the analysis, is of importance. Only argon with particular
low moisture level can be considered. The dew point should be under -50°C.
The conditions must be such as to ensure that the finish weld shows a metallic lustre. Slight
temper colours ranging from light straw to pale blue may just be acceptable. The occurrence
of dark blue colours or worse of a grey oxide layer is an indicative of insufficient inert gas
INDEX çç Sommaire
shielding and consequently of weld embrittlement due the absorption of oxygen and/or
nitrogen.
The hardness of a good weld may exceed that of the fully recrystallized base metal by a
maximum of 50 hardness units. If also after light grinding of the surface a hardness test of the
weld gives a value beyond this limit, the weld must be regarded as embrittled, and such a
weld has to be completely removed.
Due to the formation of brittle phases, titanium cannot be welded to other materials such as
steel by the inert-gas-shielding arc processes.
FORMABILITY
At room temperature titanium grade 1 and 2 can be worked very well and grade 3 and grade
12 moderately well. When the degree of deformation is fairly high it is good practice to work
all grades in heated conditions. Depending on the amount of deformation and the titanium
grade used working temperatures range from 150 to about 500°C.
The applicability of a material for cold forming is given by its strain hardening behaviour.
The strain hardening diagram of titanium grade 1 is shown in the left of figure 2. The
moderate increase of the yield strengths and also of the tensile strength illustrates the slow
strain hardening of titanium grade 1. Titanium grade 2 shows the same behaviour but of
course at higher absolute values in strength.
The grade of strain hardening also depends on the rolling direction of titanium sheets. This is
shown in the right graph of figure 2 where the strain hardening exponent is given in the
dependence of the sheet rolling direction. It demonstrates that the highest strain hardening is
in the rolling direction and the lowest in 90° transverse to the rolling direction. This visualizes
well that the best cold workability behaviour of sheets is perpendicular to the rolling
direction.
The cold forming of sheets will be also affected by the anisotropy and the texture.
INDEX çç Sommaire
Due to the relative high yield point ratio - increasing with the strength of titanium - and also
due to the low modulus of elasticity, pronounced elastic behaviour must be reckoned during
the forming process. This causes a strong springback, so it is often preferred to carry out
forming at increased temperatures. For example the springback angle in sheet forming of
titanium grade 1 is approximately 5 – 8°.
For bending of sheets large radii and low forming speeds should be chosen to avoid cracks on
the surface or at worst a fracture. Advised minimum bend radii are listed in the standard DIN
9003-3. If, for constructional reasons, small radii and sharply angled folds cannot be avoided
hot-forming or cold pre-forming followed by hot sizing is recommended.
An essential condition for good forming results is a stable and uniform lubrication. By coating
sheets with plastic foil and at the same time lubricating with deep drawing oil, considerably
better drawing behaviour can be achieved without failure of the material. At high
temperatures colloidal graphite and the usual hot press greases with graphite or molybdenum
disulphide additives have been highly successful.
The forming can be carried out with the same equipment and tools used for stainless steel and
nickel alloys. However, the machinery should be carefully cleaned to avoid contamination of
the titanium by steel or foreign rust particles.
In the chemical process industry titanium is mostly used in the form of sheet, plate, tube and
welding wire and also bars as pre-material for forged flanges. However, heavy forgings are
also gaining importance, e.g. tube sheets for tubular heat exchangers and big shafts for
agitator vessels.
Summary
The modern chemical process industry operates at highly corrosive process environments in
many applications. Due to its outstanding corrosion resistance titanium is the ideal material in
INDEX çç Sommaire
such cases. The use of titanium for chemical process equipment results in very low life cycle
costs, reduced operating costs, and low upfront costs of chemical plants combined with an
improved plant performance. In order to benefit most from the excellent material properties
the choice of the right titanium grade is of eminent importance. The fabrication of titanium
components is not more expensive than those made from stainless steel or nickel-alloys. A lot
of examples of well working titanium equipment are demonstrated in several applications of
the chemical industry.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page70
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page71
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page72
RÉSUMÉ / ABSTRACT
66A
GÉNÉRATEUR DE VAPEUR EN SITUATION DE COLMATAGE –
STATUT DES INTERNES SUPPORTANT LE FAISCEAU TUBULAIRE
ET VÉRIFICATION DU RESPECT DES CRITÈRES DE TENUE MÉCANIQUE
DE CES INTERNES, AVEC LA RÉGLEMENTATION APPLICABLE
STEAM GENERATOR WITH TUBE SUPPORT PLATES CLOGGING. STATUS OF
THE TUBE BUNDLE SUPPORT AND VERIFICATION OF MECHANICAL
CRITERIA COMPLIANCE WITH APPLICABLE REGULATIONS
J.-F. Goetgheluck - P. Genette
Electricité de France – SEPTEN - 12, avenue Dutriévoz, - 69628 Villeurbanne CEDEX
[email protected] - [email protected]
D. Desombre - P. Malouines
AREVA NP - Tour Areva - 92084 Paris La Défense CEDEX
[email protected] - [email protected]
RÉSUMÉ
Mercredi
Le phénomène de colmatage des plaques entretoises des Générateurs de Vapeur se traduit par une
obstruction progressive des interstices présents entre les tubes GV et leurs plaques entretoises support.
La présence de ces dépôts lorsqu’ils sont importants, conduit à une évolution des caractéristiques de
fonctionnement des GV, qui est susceptible d’entraîner une dégradation de la situation vis à vis du
respect des critères de sûreté. L’examen détaillé montre que le colmatage peut conduire à des oscillations
thermohydrauliques créant un risque d’instabilité du niveau d’eau ou plus globalement un risque
d’instabilité du comportement d’ensemble des GV, à une réduction de leur inventaire en eau, et à des
sollicitations mécaniques particulières, comme des instabilités vibratoires des tubes, ou un
accroissement de l’effort appliqué sur les plaques entretoises lors de certains transitoires.
Dans cette publication, nous nous intéresserons pour des GV impactés par le phénomène de colmatage,
à la justification du comportement mécanique des tirants, plaques entretoises et faisceau tubulaire (en
cas d’interaction de ce dernier avec les plaques entretoises), qui sont confrontés aux situations et
chargements applicables.
Nous verrons comment dans les dossiers de conception était vérifiée la tenue mécanique des internes
support (plaques entretoises, tirants) et quels étaient les critères réglementaires appliqués.
Le phénomène du colmatage n’étant pas dimensionnant lors de la conception, il est nécessaire de
l’intégrer dans le Rapport De Sûreté s’il a un caractère permanent et/ou partiel. Cela impose un travail
sur la classification des internes, qui par conception, n’appartiennent ni directement au circuit primaire
principal, ni au circuit secondaire principal. Nous étudierons ainsi le statut réglementaire et les exigences
de sûreté associées applicables aux internes supportant le faisceau tubulaire. Nous verrons quels sont
les impacts de l’arrêté ESPN du 12 décembre 2005.
Par la suite nous exposerons sur quels critères mécaniques se transposent les exigences réglementaires
et de sûreté.
Enfin, cette publication comparera les résultats d’une étude des internes pour un cas concret de
générateur de vapeur dans une situation de colmatage.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page73
RÉSUMÉ / ABSTRACT
10A
D. A. Osage
The Equity Engineering Group, Inc.
Shaker Heights, OH USA
M. Prager
The Materials Properties Council, Inc.
New York, NY USA
RÉSUMÉ
Le développement de nouvelles règles sur la ténacité représenta une part importante de l’effort déployé
pour la réécriture de la Division 2 de la Section VIII du B&PV Code de l’ASME. Les nouvelles règles
du Code ont été établie en se basant sur les règles d’évaluation de nocivité des fissures données dans
la Partie 9 de l’579-1/ASME FFS-1 Fitness-For-Service (Aptitude au Service). Comparativement aux
règles concernant la ténacité actuellement dans la Section VIII Division 1 et à celles de l’ancienne
Division 2, les changements le plus notables sont pour les aciers au carbone ou faiblement alliés à
l’exclusion des aciers de boulonnerie. Les nouvelles règles de ténacité dans la Section VIII Division 2
sont basées sur une énergie de rupture en flexion par choc sur éprouvettes ISO-V de 27 Joules en
conformité avec la pratique Européenne, ceci en tenant compte des effets bénéfiques d’un éventuel
traitement thermique après soudage évalué conformément aux procédures de l’ API 579-1/ASME FFS-1.
Ce papier présente les fondements techniques des nouvelles règles de ténacité ainsi que le
développement des exigences qui en découlent et le développement des règles d’exemption d’essais
de rupture en flexion par choc.
ABSTRACT
The development of new toughness requirements was a major part of the effort to re-write the ASME
B&PV Code, Section VIII, Division 2. The new toughness rules in this code were established using the
fracture mechanics assessment procedures in API 579-1/ASME FFS-1 Fitness-For-Service, Part 9.
The major changes in the toughness rules when compared to older editions of Section VIII, Division 2
and the current edition of Section VIII, Division 1 are for carbon and low alloy steel materials excluding
bolting. The new toughness rules in Section VIII, Division 2 are based on a Charpy V-Notch impact
requirement of 27 Joules (20 ft-lbs) consistent with European practice and the beneficial effects of post
weld heat treatment are included consistent with the procedures in API 579-1/ASME FFS-1. This paper
provides a technical background to the new toughness rules including the development of material
toughness requirements and the development of impact test exemption rules.
AFIAP - ESOPE 2010 Programme
INDEX çç Sommaire
INTRODUCTION
In 2007, a new version of the ASME B&PV Code, Section VIII, Division 2 (VIII-2) that
incorporates many new technologies including an increase in allowable stress to be more globally
competitive was issued by ASME. A technical background and commentary on this new code is
provided by Osage [1]. One of the new technologies incorporated pertained to changes to the
toughness rules necessitated by the increased allowable stresses and the need for corrections as
discovered by a comprehensive review of the past practices and assumptions.
The development of the material toughness rules for VIII-2 is described by Prager and Osage [2]
and summarized in this paper. The approach described herein for calculating fracture mechanics
driving forces and setting the corresponding requirements for toughness to develop toughness
requirements for VIII-2 is provided. The approach utilizes the fracture mechanics methodology
in API 579-1/ASME FFS-1 2007, Fitness-for-Service (FFS). Thus a continuum is now provided
between design requirements and the long accepted procedures and models for FFS. The
foundations for the toughness requirements in the ASME B&PV Code, Section VIII, Division 1
(VIII-1) and VIII-2 were reported by Corten (see WRC 528 [2], Annex A), Barsom and Rolfe
[3], Selz [4], and Jacobs [5]. The VIII-2 and FFS applications of these foundations represent
updates that include modern fracture mechanics concepts such as Failure Assessment Diagram
(FAD), adjustments to calculation procedures for lower shelf energies, estimation of residual
stresses in welded equipment and interpolation between lower and upper shelf Charpy energies.
Also described herein are the concepts used to establish the exemption curves and the reductions
in allowable stresses required to permit lowering of Minimum Design Metal Temperature
(MDMT).
In this paper, the editions of the ASME B&PV Section VIII Codes are identified as follows:
• VIII-2 – Section VIII, Division 2, 2007 Edition and later
• Old VIII-2 – Section VIII, Division 2, 2004 Edition, 2006 Addenda and earlier
• VIII-1 – Section VIII, Division 1, 2007 Edition
BACKGROUND
Impact testing requirements for materials that are classified into groups are provided in VIII-1,
paragraph UCS-66. For each group, identified with the labels A, B, C, and D, there is a thickness
dependent curve in UCS-66 showing the MDMT for which the material is exempt from impact
testing. Below the curves materials must be tested unless lower than normal allowable stresses or
other specified conditions are met. The VIII-1 curves for various material groups are shown in
Figure 1 in which temperature of exemption can be seen to decrease with decreasing thickness.
Impact testing is required for the specific combinations of design temperature, material
classification and thickness below the respective curves. Thin materials have lower allowable
design exemption temperatures and the curves become quite steep as the thickness decreases as
shown in Figure 1. The curves as shown are truncated at the thickness of a full size Charpy
impact specimen, 0.392 in. The procedures herein can be applied for extension of the curves to
lower thicknesses.
The VIII-1 toughness rules described above were also used in Old VIII-2. However, these rules
required significant updating for VIII-2 because the decision was made by the PVRC and ASME
committees in charge of the re-write for this new code to adopt a minimum of 27 Joules (20 ft-
lbs) for toughness for ferritic materials that is reasonable for modern steel making practices and
expectations and close to European standards.
APPROACH
In developing the toughness rules for VIII-2 the entire technical and historical basis for the UCS-
66 exemption curves was examined, understood, checked, corrected and upgraded to modern
INDEX çç Sommaire
fracture mechanics standards. The relevant original equations documented by Corten (see WRC
528 [2], Annex A) were obtained and applied to verify and understand the UCS-66 exemption
curves. The equations were then modified to develop exemption curves applicable to the higher
design allowable stresses and the special needs of VIII-2. The result of that effort was a
systematic approach that can be applied to other design codes and criteria and even modified for
particular geometries and assumed flaws if desired. Specifically, the method was updated to use
the most modern fracture mechanics approach for welds in API 579-1/ASME FFS-1 including
stress intensity factor and reference solutions for the crack driving force, estimation schemes for
residual stress, and appropriate Failure Assessment Diagrams.
Another element worth noting here is the importance of a systematic and reasonable scheme for
correlating fracture toughness with Charpy energy. The approach fashioned to smoothly
correlate fracture toughness with Charpy impact test values from the lower shelf to the upper
shelf was adjusted to compensate for yield strength and nonlinear behavior. Using this approach,
it would be possible to create new exemption curves (e.g. E, F, G, etc.) based on the performance
of new materials that may have been improved by control of their composition, melting practice
etc., or to justify moving a material from one curve to an existing curve if its classification is now
deemed to be inappropriate.
INDEX çç Sommaire
Primary and Residual Stress
To compute the crack driving force, the following membrane stresses are assumed for the applied
primary stress, σ mP , and residual stress, σ mSR . Note that although allowable design stress is based
on the specified minimum yield strength, the specified minimum ultimate tensile strength and
time-dependent properties, the primary membrane stress is conservatively approximated using
Equation (3), i.e. it is a function of the specified minimum yield strength only.
2
σ mP = σ ys (3)
3
2
σ mSR = σ ys ( not subject to PWHT ) (4)
3
σ mSR = 0.20σ ys ( subject to PWHT ) (5)
Setting the residual stress to 20% of the yield strength for components subject to post weld heat
treatment is consistent with API 579-1/ASME FFS-1, Annex E.
Required Material Fracture Toughness
The toughness ratio for the FAD-based fracture mechanics approach discussed above is given by
Equation (6). This equation is taken from API 579-1/ASME FFS-1, Part 9, paragraph 9.4.3.2.l).
K I P + ΦK I SR
Kr = (6)
K mat
The FAD described using Equation (7) was used, see API 579-1/ASME FFS-1, Part 9, Figure
9.20. This is a conservative FAD diagram when LPr (max) = 1.0 .
Combining Equation (6) with Equation (7) and solving for the required material toughness, K mat ,
the following expression is obtained.
K1P + ΦK1SR
K mat (t ) = (8)
(1.0 − ( L ) )
0.2
p 2.5
r
In this expression the required material fracture toughness is a function of the thickness of the
material.
The fracture toughness parameters, K1P and K1SR in Equation (8), are defined by Equations (9),
Cylinder
(10) and (11). In these equations, the parameter K RF was derived using API 579-1/ASME
FFS-1, Annex C using the KCSCLE2 Solution with a 1 ksi membrane stress and the reference
flaw defined in Equations (1) and (2). The membrane stresses in equations (9) and (10) are set in
accordance with Equations (3), (4) and (5). The resulting equations for the fracture toughness
parameters are a function of the cylinder wall thickness ( t ) and radius-to-thickness ratio ( R / t ) .
K1P = σ mP ⋅ K RF
Cylinder
(9)
K1SR = σ mSR ⋅ K RF
Cylinder
(10)
INDEX çç Sommaire
⎛ ⎞
⎜ ⎟
⎜ 76.127581 + 29.28642 ⋅ t ⋅ ln [t ] − 190.73697 ⋅ t + ⎟
⎜ 122.7235 12.153937 ⎟
Cylinder
K RF = ⎜120.62934 ⋅ ln [t ] + − + ⎟ (11)
⎜ t t ⎟
⎜ ⎟
⎜13.33629 ⋅ exp [ −t ] − 0.17604576 − 0.42094737 ⋅ ln [ R t ] ⎟
⎜ Rt (R t)
2
⎟
⎝ ⎠
Equation (11) utilized in the development of the toughness rules herein was developed using the
data in API 579, 2000 Edition, Appendix C and is valid for a thickness range of 0.25 in ≤ t ≤ 4 in
.
The plasticity interaction, Φ , defined by Equation (12) was derived by curve fitting the plots
shown in Figure 9.19 of API 579-1/ASME FFS-1.
⎛ 0.99402985 − 0.34259558 ⋅ LP + 0.07849594 ⋅ LSR + 1.3153525 ⋅ ( LP )2 − ⎞
⎜ r r r
⎟
⎜ ⎟
Φ = ⎜ 0.035075224 ⋅ ( Lr ) + 0.2222982 ( Lr )( Lr ) − 0.97610564 ⋅ ( Lr ) +
SR 2 P SR P 3
⎟ (12)
⎜ ⎟
⎜ 0.0041367592 ⋅ ( LSR
r ) − 0.0062624497 ⋅ ( Lr )( Lr ) − 0.16970127 ⋅ ( Lr ) ( Lr ) ⎟
3 P SR 2 P 2 SR
⎝ ⎠
The load ratio parameters, LPr and LSR
r in Equation (8) and Equation (12), are defined by
Cylinder
Equations (13), (14) and (15). In these equations, the parameter RRF was derived using API
579-1/ASME FFS-1, Annex D using the RCSCLE2 solution with a 1 ksi membrane stress and
the reference flaw defined in Equations (1) and (2). The membrane stresses in Equations (13)
and (14) are set in accordance with Equations (3), (4) and (5). The resulting equations for the
load ratio parameters are a function of the cylinder wall thickness and radius to thickness ratio.
σ mP ⋅ RRF
Cylinder
LPr = (13)
σ ys
σ mSR ⋅ RRF
Cylinder
r =
LSR (14)
σ ys
Cylinder
The parameter RRF is evaluated using the following equation.
⎛ ⎞
⎜ 0.99829577 + 0.0071541778 ⋅ t + 1.3018206 − 0.0019047184 ⋅ t 2 − ⎟
⎜ (R t) ⎟
⎜ ⎟
⎜ 4.3132859 0.042484369 ⋅ t ⎟
+ 0.00011487158 ⋅ ( t ) +
3
Cylinder
RRF = − (15)
⎜ ( R t )2 ( )
R t ⎟
⎜ ⎟
⎜ 5.4284626 0.12675001 ⋅ t 0.0033013072 ⋅ t 2 ⎟
⎜⎜ + − ⎟⎟
⎝ (R t)
3
(R t)
2
(R t) ⎠
Equation (15) utilized in the development of the toughness rules herein was developed based on
API 579, 2000 Edition, Appendix D and is valid for a thickness range of 0.25 in ≤ t ≤ 4 in .
INDEX çç Sommaire
Derivation of Material Charpy V-Notch Impact Test Requirements
Required Material Fracture Toughness
The required material toughness, K mat (t ) , as a function of thickness is based on the reference
flaw and applied stress is given by Equation (8). When K mat (t ) is evaluated, the fracture
toughness parameter given by Equation (11) and reference stress parameter given by Equation
(15) were evaluated at R t = 100 . In addition, the primary and residual stress membrane stresses
in the fracture toughness equations, Equations (9) and (10), and in the load ratio Equations (13)
and (14), are given by Equations (3), (4) and (5).
MPC CVN and Fracture Toughness Model
The Materials Properties Council (MPC) has developed a Charpy impact energy correlation and a
methodology to relate this energy to both the dynamic and static fracture toughness, see API 579-
1/ASME FFS-1, Annex F, paragraph F.4.5.3. This correlation uses a two-step procedure for
correlating static fracture toughness with Charpy test results in the transition region using
temperature shift of the dynamic fracture toughness correlation.
K1C (T − ΔTs ) = K1d (T ) (16)
where:
K1d = 15 CVN ( ksi in , ft − lb ) (17)
C = 66 o F (24)
The dynamic fracture toughness for a material for a given yield strength, σ ys , temperature, T ,
and ASME exemption Curve may be estimated by Equation (19). The materials assigned to each
of the exemption curves, A, B, C, and D are shown in Table 1.
The static fracture toughness is determined by applying the temperature shift indicated by
INDEX çç Sommaire
Equation (16), or
⎧⎪ ⎛ 27 ⎞ ⎡ (T − (T0 − 75 ) ⎤ ⎫⎪
K1C = σ ys ⎨ 3 + ⎜ 3 −
⎜ ⎟
σ ys ⎟⎠
⋅ tanh ⎢
⎣ C
⎥⎬
⎦ ⎭⎪
( ksi in , ksi, o F ) (25)
⎩⎪ ⎝
INDEX çç Sommaire
Upper Shelf Region CVN Requirement
For extreme values of temperature, T >> T0 , i.e. upper shelf behavior, an estimate of the upper
shelf fracture toughness may be inferred directly from Equation (19), or:
(
K us = K1d = 2 3 ⋅ σ ys − 27 ) (31)
The Rollfe-Novak-Barsom correlation given by Equation (32) and API 579-1/ASME FFS-1,
Annex F, paragraph F.4.5.2 provides a correlation between the upper shelf CVN and the upper
shelf fracture toughness and the yield strength.
K us2 σ ys
CVN us = + (32)
5σ ys 20
Substituting Equation (31) into Equation (32), and expression for the upper shelf Charpy energy
is obtained.
(2 )
2
3 ⋅ σ ys − 27 σ ys
CVNus = + (33)
5σ ys 20
Note that the required CVN for a material is a function of the yield strength and nominal
thickness as shown in Figures 2 and 3 for components not subject to PWHT and components
subject to PWHT, respectively.
INDEX çç Sommaire
by Equation (8). When K mat (t ) is evaluated, the fracture toughness parameter given by Equation
(11) and reference stress parameter given by Equation (15) are evaluated at R / t = 100 . In
addition, the primary and residual stress membrane stresses in fracture toughness equations,
Equations (9) and (10), and the load ratio equations, (13) and (14), are given by Equations (3),
(4) and (5).
For all materials covered by the four exemption curves labeled A, B, C, and D, the yield stress in
equation (36) is conservatively set as σ ys = 80 ksi . In addition, the cut-off limit for the lower
bound of the curve is taken as -58°F or the temperature at which the thickness is equal to 0.4 in,
( t = 0.4 in ) , whichever is less.
It should be noted that the new ASME exemption curves for Section VIII, Division 2 are
independent of the strength of the steel because the yield strength in Equation (36) is set at
σ ys = 80 ksi . This assumption is made to simplify use of the exemption curves. If the yield
dependence of the exemption were honored, then multiple set of exemption curves labeled A, B,
C, and D would need to be generated for different values of the yield strength.
In earlier editions of Section VIII, Division 2 and the current edition of Section VIII, Division 1,
the strength independence of the exemption curves lies in assumptions regarding the material
property model used when the exemption curves were developed, see WRC 528 [2], Annex A,
Equations (1), (2), and (3). The main assumption that was used in the development of the
exemption curves was that the fracture toughness at the indexing temperature should be
proportional to the yield strength of the material and vary with temperature in a strength
independent manner. This assumption can be assured through application of yield strength
dependent Charpy test requirements.
⎛2 ⎞
σ mP = Rts ⎜ ⋅ σ ys ⎟ (37)
⎝3 ⎠
With these assumptions, Equation (36) may be written as follows:
⎡ ⎤
⎢ ⎥
K ( Rts ) − 3 ⋅ σ ys ⎥
−1 ⎢ mat
TR ( Rts ) = tanh ⎢ ⎥ ⋅ C + T0 (38)
⎛ 27 ⎞
⎢ ⎜⎜ 3 − ⎟⎟ ⎥
⎢ ⎝ σ ⎠ ⎥
⎣ ys ⎦
Note that in Equation (38), the temperature is now a function of the stress reduction ratio, Rts ,
INDEX çç Sommaire
rather than the wall thickness. The final equation for the temperature reduction, ΔT , in Figures
6 and 7 is given by:
ΔT ( Rts ) = TR (1) − TR ( Rts ) (39)
This equation is evaluated for both the non PWHT and PWHT condition, and for the two yield
strengths shown in the figures.
In ASME Section VIII-2, if the computed value of the Rts ratio is less than or equal to the 0.24,
see Figure 6, then the MDMT may be set to -155°F and impact testing is not required unless a
lower MDMT is required. This requirement essentially stipulates that if the operating stresses
are equal to or less than 10% of the ultimate tensile strength, then operation for ferritic materials
is permitted on the lower shelf. This rule is approximately consistent with older versions of
ASME Section VIII-2 where the limit for the Rts ratio is 0.3 and for ASME Section VIII-1 where
the limit for the Rts ratio is 0.35. The justification for lower shelf operation is that the stress is
low enough such that brittle fracture is not likely. However, lower shelf operation may not be
justified for welded components as explained in WRC 528 [2].
INDEX çç Sommaire
πa
K VIII −2 = M σ VIII − 2 (40)
Q
πa
K Old VIII − 2 = M σ Old VIII − 2 (41)
Q
where
1.65
⎛a⎞
Q = 1 + 1.464 ⎜ ⎟ (42)
⎝c⎠
For equal fracture resistance or an equal reference flaw size,
K VIII −2 = K Old VIII −2 (43)
σ VIII −2 = σ VIII
P
− 2 + σ VIII − 2
SR
(45)
where
⎡ σ uts σ ys ⎤
σ VIII
P
− 2 = min ⎢ , ⎥ (46)
⎣ 2.4 1.5 ⎦
σ VIII
SR
− 2 = 0.15 ⋅ σ ys ( subject to PWHT ) (47)
σ VIII
SR
− 2 = σ ys ( not subject to PWHT ) (48)
⎡ σ uts σ ys ⎤
σ Old
P
VIII − 2 = min ⎢ , ⎥ (50)
⎣ 3.0 1.5 ⎦
σ Old
SR
VIII − 2 = 0.15 ⋅ σ ys ( subject to PWHT ) (51)
σ Old
SR
VIII − 2 = σ ys ( not subject to PWHT ) (52)
From API 579-1/ASME FFS-1, the following equation may be used to determine the relationship
between fracture toughness and Charpy V-Notch impact energy.
K1c = 9.35 ( CVN )
0.63
( ksi in , ft − lbs ) (53)
INDEX çç Sommaire
1
The Lateral Expansion (LE) in a Charpy V-Notch test may be related to the corresponding
Charpy V-Notch impact energy (CVN) using the Equation (56). This equation was obtained
from actual test data on SA 353, SA 517 and SA 645.
CVN = 1.84 ⋅ LE − 1.56 ( ft − lbs, mils ) (56)
In terms of VIII-2 and Old VIII-2:
CVNVIII − 2 + 1.56
LEVIII − 2 = ( ft − lbs, mils ) (57)
1.84
CVNOld VIII − 2 + 1.56
LEOld VIII − 2 = ( ft − lbs, mils ) (58)
1.84
Combining Equations (44), (54), (55), (57) and (58), an expression for LE for VIII-2 can be
derived as:
[ B + 1.56]
LEVIII − 2 = ( ksi, ft − lbs, mils ) (59)
1.84
where
1
⎛ σ ⎞ 0.63
(
B = ⎜ VIII −2 ⎟ ⋅ 1.84 ⋅ LEOld VIII −2
⎜ σ Old VIII −2 ⎟ ) (60)
⎝ ⎠
Calculations were performed for SA 517, grade F and SA 353 for vessels with a wall thickness
up to 3 in. The results are presented in Table 2.
Approach 2 for Determining CVN LE Requirements
The second approach is based on establishing the toughness required to resist the reference flaw
that is a semi-elliptical surface crack with a 0.25t depth and 1.5t length. TWI software program
CW4, based on BS7910:1999 was used in the analysis. The use of the failure assessment
diagram (FAD) has become the standard FFS method to evaluate crack-like flaws found in in-
service components. The FAD provides a methodology to evaluate the interaction between a pure
fracture mechanics approach and a pure limit load failure. The limit load condition in the FAD
approach addresses the condition flow-stress controlled failure of a component. Both the API
579-1/ASME FFS-1 and BS 7910 Standards have adopted the FAD methodology for the
evaluation of crack like flaws. The failure criterion for both standards is given by Equation (61).
(
K r = (1 − 0.14 L2r ) 0.3 + 0.7 exp ⎡⎣ −0.65L6r ⎤⎦ ) (61)
For all practical purposes, the API 579-1/ASME FFS-1 Level 2 procedure is identical to that of
BS 7910 Level 2. The two standards differ in the solutions provided for the applied stress
intensity factor that is used in the Kr calculation and the reference stress used in the Lr
calculation. A comparison by Rana [12] of the 579-1/ASME FFS-1 Level 2 procedure and of BS
7910 indicates that BS 7910 would result in a conservative fracture strength prediction.
INDEX çç Sommaire
The following assumptions were used in the analysis.
a) The API 579-1/ASME FFS-1 Level 2 FAD is used.
b) The ratio of the vessel radius to thickness is set equal to 50.
c) A surface elliptical crack on the outside surface is assumed. The crack depth is set equal to
0.25t and the crack length is set equal to 1.5t where t is the thickness of the vessel.
d) The crack is located in the center of a longitudinal weld.
e) The material data is:
1) SA 517, Grade F: UTS=115 ksi, YS=100 ksi, the allowable stress S=115/2.4=47.9 ksi,
the stress in the hydrostatic condition is 1.43(47.9 ksi)=68.5 ksi, and
2) SA 353: UTS=100 ksi, YS=85 ksi, the allowable stress S=100/2.4=41.7 ksi, the stress in
the hydrostatic condition is 1.43(41.7 ksi)=59.6 ksi.
f) The residual stresses from welding are computed by the TWI CW4 software based on the
PWHT and non-PWHT conditions for a given thickness and material considering the
hydrotest conditions.
Using the above assumptions and input data, the required toughness was computed using the
TWI CW4 software. The resulting thickness was adjusted for plane stress conditions using
Equation (62).
0.5
⎡ 4
⎤
⎛ 1.0 ⎞ ⎛ K1c ⎞
2
CONCLUSIONS
The development of the material toughness rules for carbon and low alloy steels, i.e. required
CVN, impact test exemption curves, and the additional reduction in the impact test temperature
based on loading condition incorporated into the new ASME Section VIII Division 2, 2007
Edition is covered in this paper. Changes to the toughness rules in earlier version of this code
were necessitated by increased allowable stresses under that ASME Code and the need for
corrections as discovered by a comprehensive review of the past practices and assumptions. The
new toughness rules in this code were established using the fracture mechanics assessment
procedures in API 579-1/ASME FFS-1 Fitness-For-Service, Part 9 and a new correlation that
determines the required Charpy V-Notch impact energy (CVN) from the fracture toughness. A
methodology to relate this energy to both the dynamic and static fracture toughness using the
model is also described.
INDEX çç Sommaire
REFERENCES
1. Osage, D.A., ASME Section VIII – Division 2 Criteria and Commentary, ASME PTB-1
2007, ASME, New York, N.Y.
2. Prager, M. and Osage, D.A., Development of Material Fracture Toughness Rules for the
ASME B&PV Code, Section VIII, Division 2, WRC Bulletin 528, The Welding Research
Council, New York, N.Y., August, 2010.
3. Barsom, J.M. and Rolf, S.T., Fracture and Fatigue Control in Structures, Applications of
Fracture Mechanics, Third Edition, ASTM, West Conshohocken, PA.
4. Selz, A., “New Toughness Rules in Section VIII, Division 1 of the ASME Boiler and
Pressure Code,” 88-PVP-8, ASME, New York, N.Y., 1988.
5. Jacobs, W.S., “ASME Code Material Toughness Requirements for Low Temperature
Operation, Section VIII, Division 1 and Division 2, 1998, 1999 Addenda,” PVP Vol. 407
Pressure Vessel and Piping Codes and Standards – 2000, ASME 2000, pg 23-38.
6. Scott, P.M., Anderson, T.L., Osage, D.A., and Wilkowski, G.M., A Review Of Existing
Fitness-For-Service Criteria For Crack-Like Flaws, WRC Bulletin 430, The Welding
Research Council, New York, N.Y., April, 1998.
7. Anderson, T.L., Fracture Mechanics – Fundamentals and Applications, 3nd Edition, CRC
Press, Boca Raton, Florida, 2005.
8. Anderson, T.L. and Osage, D.A., “API 579: A Compressive Fitness-for-Service Guide”,
International Journal of Pressure Vessels and Piping, 77 (2000) 953-963.
9. WRC, PVRC Recommendations on Toughness Requirements for Ferritic Steels, PVRC Ad
Hoc Group on Toughness Requirements, WRC Bulletin 175, The Welding Research
Council, New York, N.Y., August, 1972.
10. Rana, M.D., Private Communication
11. Gross, J.H., “Effect of Strength and Thickness on Notch Ductility,” Welding Research
Supplement, The Welding Research Council, New York, N.Y., October, 1969.
12. Rana, M.D. and Rawls, G.B., “Prediction of Fracture Stresses of High Pressure Gas
Cylinders Containing Crack-Like Flaws,” ASME, Journal of Pressure Vessel Technology,
November 2007, Volume 129, Issue 4,639 (5 pages).
INDEX çç Sommaire
NOMENCLATURE
a reference flaw depth.
2c reference flaw length.
C K1d parameter.
CVN Charpy V-Notch impact energy
CVN code−min minimum CVN requirement of the code.
CVNls CVN of lower shelf.
CVNus CVN requirement for the upper shelf.
CVN Old VIII − 2 representative CVN for Old VIII-2
CVNVIII − 2 representative CVN for VIII-2
CVN (t ) CVN requirement as a function of thickness.
CVN nls (t ) CVN requirement for the near lower shelf as a function of thickness.
CVNtrans (t ) CVN requirement for the transition region as a function of thickness.
ΔTR ( Rts ) temperature reduction as a function of Rts .
ΔTS temperature shift from dynamic to static toughness.
MDMT Minimum Design Metal Temperature.
Kc fracture toughness
K ls fracture toughness estimate for the lower shelf.
K mat value of the material fracture toughness.
K mat (t ) value of the material fracture toughness as a function of thickness.
Kr toughness ratio.
K us fracture toughness estimate for the upper shelf.
K Old VIII − 2 representative fracture toughness for Old VIII-2
KVIII − 2 representative fracture toughness for VIII-2
K1c plane strain static fracture toughness.
K1d dynamic fracture toughness.
P
K I stress intensity factor based on primary stresses.
SR
K I stress intensity factor based on secondary and residual stresses.
Cylinder
K RF stress intensity factor
K nls (t ) fracture toughness requirement for the near lower shelf region as a function of
thickness.
Lr load ratio.
LPr load ratio based on primary stress.
SR
L r load ratio based on secondary and residual stresses.
LE linear expansion from a Charpy impact test
LEOld VIII − 2 representative LE for Old VIII-2
LEVIII − 2 representative LE for VIII-2
M stress intensity factor coefficient
MDMT acronym for Minimum Design Metal Temperature
INDEX çç Sommaire
Φ plasticity correction factor
R radius of the cylinder
Rts stress ratio defined as the stress for the operating condition under consideration
divided by the stress at the design minimum temperature. The stress ratio may
also be defined in terms of required and actual thicknesses, and for components
with pressure temperature ratings, the stress ratio is computed as the applied
pressure for the condition under consideration divided by the pressure rating at the
MDMT .
Cylinder
RRF reference stress factor.
σ ys engineering yield stress evaluated at the temperature of interest.
σ uts engineering ultimate tensile stress evaluated at the temperature of interest.
σ m ,Old VIII − 2 representative membrane stress for Old VIII-2
σ m ,VIII − 2 representative membrane stress for VIII-2
σm primary membrane stress
P
INDEX çç Sommaire
TABLES
INDEX çç Sommaire
Table 2 – CVN LE Requirements Using Approach 1, Equal Fracture Resistance
CVN
Crack Crack CVN
Thick CVNVIII-2 LEold VIII-
Material PWHT Depth Length KIc,VIII-2 Kc,VIII-2 LEVIII-2
(in) (Ft-lbs) 2
(in) (in) (mils)
(mils)
SA 517,
No 0.31 0.08 0.48 51 73 15 9 15
Grade F
SA 517,
No 0.58 0.15 0.9 65 85 22 14 15
Grade F
SA 517,
Yes 1.0 0.25 1.5 63 70 21 12 15
Grade F
SA 517,
Yes 1.25 0.31 1.86 71 78 25 15 15
Grade F
SA 517,
Yes 2.2 0.55 3.3 94 104 39 22 20
Grade F
SA 517,
Yes 3.0 0.75 4.5 110 122 50 28 25
Grade F
SA 353 No 0.31 0.08 0.48 43 59 11 7 15
SA 353 No 1.0 0.15 0.9 76 103 28 16 15
SA 353 No 1.25 0.25 1.5 84 114 33 19 15
SA 353 No 2.0 0.31 1.86 107 146 48 27 19
SA 353 Yes 2.2 0.55 3.3 85 97 33 18 20
SA 353 Yes 3.0 0.75 4.5 95 106 40 22 25
INDEX çç Sommaire
FIGURES
Figure 1 – VIII-1, Paragraph UCS 66 Exemption Curves Are Shown. Indicated Notes Define the Materials
Covered
INDEX çç Sommaire
70
Cv, ft-lb (average of three specimens)
60
≥ 80 ksi
40
65 ksi
30
50 ksi
20
≤ 38 ksi
10
0.00 0.25 0.50 0.75 1.00 1.25 1.50
Figure 2 (VIII-2 Figure 3.3) – Charpy V-Notch Impact Test Requirements for Full-Size Specimens for Carbon
and Low Alloy Steels As a Function of the Specified Minimum Yield Strength – Parts Not Subject to PWHT
INDEX çç Sommaire
70
Cv, ft-lb (average of three specimens)
60
≥ 80 ksi
50
65 ksi
40
50 ksi
30
20
≤ 38 ksi
10
0 1 2 3 4 5 6 7
Figure 3 (VIII-2 Figure 3.4) – Charpy V-Notch Impact Test Requirements for Full-Size Specimens for Carbon
and Low Alloy Steels For Selected Specified Minimum Yield Strength – Parts Subject to PWHT
INDEX çç Sommaire
120
60
B
40
C
20
D
0
-20
-40
Impact Testing Required
-60
0.00 0.25 0.50 0.75 1.00 1.25 1.50
Figure 4 (VIII-2 Figure 3.7) – Impact Test Exemption Curves – Parts Not Subject to PWHT
120
Minimum Design Metal Temperature, °F
100
A
80
60
B
40
20
C
0
D
-20
-40
-60
Impact Testing Required
-80
0 1 2 3 4
Figure 5 (VIII-2 Figure 3.8) – Impact Test Exemption Curves - Parts Subject to PWHT and Non-welded Part
INDEX çç Sommaire
1.0
0.9
0.8
0.7
Stress Ratio - Rts
0.6
0.5
≤ 50 ksi
0.4
> 50 ksi, ≤ 65 ksi
0.3
0.2
0.1 See paragraph 3.11.2.5.a.5.i when Rts is less than or equal to 0.24
0.0
0 20 40 60 80 100
Figure 6 (VIII-2 Figure 3.12) – Reduction in the MDMT without Impact Testing – Parts Not Subject to PWHT
1.0
0.9
0.8
0.7
≤ 50 ksi
Stress Ratio - Rts
0.6
0.3
0.2
0.1 See paragraph 3.11.2.5.a.5.i when Rts is less than or equal to 0.24
0.0
0 20 40 60 80 100
Figure 7 (VIII-2 Figure 3.13) – Reduction in the MDMT without Impact Testing - Parts Subject to PWHT and
Non-welded Parts
INDEX çç Sommaire
Figure 8 – Charpy V-Notch Lateral Expansion Requirements (CVN LE) in Old VIII-2 and the 2007 Edition of
VIII-2.
34
32
30
CV Lateral Expansion, mils
28
26
24
22
20
18
16
14
12
10
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Figure 9 (VIII-2, Figure 3.6)– Charpy V-Notch Lateral Expansion Requirements (CVN LE) in VIII-2, 2008
Addenda.
INDEX çç Sommaire
Figure 10 – Required CVN LE for Materials with a UTS> 95 ksi Using Approaches 1 and 2
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page74
RÉSUMÉ / ABSTRACT
2A
CONCEPTION SELON L’EN 13445 D’UN RÉACTEUR-ÉCHANGEUR À TUBES
BAÏONNETTES DE GRANDES DIMENSIONS ET À HAUTE TEMPÉRATURE
RÉSUMÉ
Mercredi
Dans le cadre du projet de recherche européen CACHET (CArbon dioxide Capture via Hydrogen
Energy Technology), TECHNIP et l’IFP ont développé le concept d’un nouveau type de réacteur-
échangeur répondant aux conditions très sévères du procédé HyGenSys (Hydrogen Generation
System) : grand volume de fumées pressurisées à hautes températures.
L’exposé présentera les différentes étapes de la conception du réacteur : la sélection des matériaux, les
concepts innovants et le dimensionnement selon le code européen EN 13445.
ABSTRACT
In the scope of the European research project CACHET (CArbon dioxide Capture via Hydrogen Energy
Technology), TECHNIP and IFP have developed the concept of a new reactor-heat exchanger, in respect
to the process constraints of HyGenSys (Hydrogen Generation System), which are mainly large
pressurised volume and high temperatures.
The paper will describe the different steps of the reactor design: material selection, innovating concepts
and calculations according to European code EN 13445.
INDEX çç Sommaire
CONCEPTION SELON L’EN 13445 D’UN REACTEUR-ECHANGEUR
A TUBES BAÏONNETTES DE GRANDES DIMENSIONS ET A
HAUTE TEMPERATURE
RESUME
Dans le cadre du projet de recherche européen CACHET (CArbon dioxide Capture via
Hydrogen Energy Technology), TECHNIP et l’IFP ont développé le concept d’un nouveau
type de réacteur-échangeur répondant aux conditions très sévères du procédé HyGenSys
(Hydrogen Generation System) : grand volume de fumées pressurisées à hautes températures.
L’exposé présentera les différentes étapes de la conception du réacteur : la sélection des
matériaux, les concepts innovants et le dimensionnement selon le code européen EN 13445.
ABSTRACT
In the scope of the European research project CACHET (CArbon dioxide Capture via
Hydrogen Energy Technology), TECHNIP and IFP have developed the concept of a new
reactor-heat exchanger, in respect to the process constraints of HyGenSys (Hydrogen
Generation System), which are mainly large pressurised volume and high temperatures.
The paper will describe the different steps of the reactor design: material selection,
innovating concepts and calculations according to European code EN 13445.
INTRODUCTION
Nowadays, there are several options for meeting the growing hydrogen needs. The most
common hydrogen production technology is steam reforming which can handle a variety of
feedstocks including natural gas, LPG, naphtha. Steam methane reformers (SMR) are based
on very mature technologies and are offered by large number of licensors such as Foster
Wheeler Corp, Haldor Topsoe AS, Technip, Lurgi, and Uhde. However SMRs are large and
expensive to construct mainly because of the large fire box that is required to avoid tube
damage by very hot flames and upper convection section that recover heat by steam
generation. Another conventional SMR constraint is the huge amount of generated steam that
could be difficult to valorise in some site.
INDEX çç Sommaire
This paper presents the concept and the design according to EN 13445 of a new reactor-heat
exchanger, developed by IFP and Technip, "heart" of an innovative steam reforming process
called HyGenSys, developed by IFP. HyGenSys is a "CO2 ready" co-generation process
(Hydrogen production through Steam Reforming reactions and power generation with
facilitated CO2 capture for sequestration). It is developed within the European CACHET
project (Carbon Dioxide Capture and Hydrogen Production from Gaseous Fuels) [1, 2].
2 HYGENSYS CONCEPT
The HyGenSys process is a Hydrogen Generation System that comprises a steam methane
reformer coupled with a gas turbine. The energy necessary for natural gas reforming is
supplied by convective exchange involving the hot flue gas leaving the gas turbine. The main
characteristics of the process are a very high degree of heat integration and consequently a
reduction of CO2 emission.
The objective of steam methane reforming process is to maximise the quantity of hydrogen
produced. The synthesis gas can either be used by refining and petrochemical applications or
by power generation plants. The feedstock used here is natural gas, the two mains reactions
are steam reforming reaction (1) and water gas shift reaction (2):
CH 4 + H 2 O CO + 3H 2 (1)
CO + H 2 O CO2 + H 2 (2)
The steam methane reforming reaction (1) is strongly endothermic; the hydrogen formation is
promoted by low pressure and high temperature, these reactions occurs only in the steam
reforming reactor. In order to minimize the quantity of unconverted methane, called methane
slip, the outlet temperature of steam reforming is chosen as high as possible, that is more or
less 900°C due to material constraint. The water gas shift reaction (2) is exothermic and
thermodynamically favoured by low temperature, whereas pressure has no influence.
In HyGenSys process depicted in Figure 1, the steam reforming reactions are achieved within
a reactor-exchanger. The heat source is provided to this reactor by the hot pressurized flue gas
leaving a turbine at an intermediate level of pressure (about 3-5 bara). This flue gas has been
reheated at about 1300°C in a secondary combustion chamber prior entering the reactor-
INDEX çç Sommaire
exchanger shell. At the outlet of the reactor-exchanger shell, the flue gas is again reheated
before entering the expander that generates power.
Natural
SMR
gas Syngas to purification
MP-steam
Flue gas
Fuel Fuel
Exhaust
Fuel HRSG gases
Air
~ Electricity
3 HYGENSYS REACTOR-EXCHANGER
The steam reforming reaction is globally very endothermic. Furthermore, for thermodynamic
reasons, a very high temperature is required to achieve the intended conversion of natural gas
and steam into syngas. In conventional steam reforming units, the conversion is achieved in
tubes filled with catalyst. Those tubes are heated in a furnace or less often by heat exchange
with a hot stream such as an ATR effluent. However, these existing reactor concepts are not
applicable to HyGenSys which relies on a devoted SMR reactor exchanger which shall ensure
production of syngas and heat exchange with hot pressurized flue gas.
The HyGenSys reactor concept is based on a vertical arrangement of double pipes (also called
bayonet tubes). In the bayonet tube concept, natural gas and steam are fed to the annular
section which is filled with steam reforming catalyst. Bayonet tubes have already been used
by Topsøe for steam methane reforming (Haldor Topsøe Convection Reformer) in the 1990's
[3], but those reactors are limited to small capacity units. That reactor uses a tube-sheet at
high temperature with a large pressure drop. That tube-sheet becomes very thick when the
reactor diameter is large.
The main innovation of the HyGenSys reactor is the specific arrangement of the bayonet
tubes in the pressurized vessel, with external connections and no tube-sheet. By the use of this
patented solution, there is no size limitation in the reactor diameter.
The conversion of the feed into syngas is achieved in this annular section while the stream is
being heated. The substantial heat required by the reaction is mainly brought by heat
INDEX çç Sommaire
exchange through the wall of the outer tube, thanks to a hot flue gas stream which flows in
counter-current to the process side. The resulting hot generated syngas provides additional
heat to the catalytic section while being routed back and cooled through the internal pipe.
Figure 2 shows the bayonet tube concept and the different flow streams:
Feed Syngas
Charge Syngas
Cooled syngas
H2 +CO+CO2+CH4+H2O
Manifold
System de collecte
Spool pieces
Spool
pieces
Reactor top
Plaque tubulaire
bombée
Cooled
flue gas
Fumées refroidies
Catalyst
Catalyst
Catalyseur
Hot syngas
H2+CO+CO2 +CH4 +H2O
900°C
Hot flue gas Hot flue gas
Hot flue
Fumées gas
chaudes
The bayonet tubes are contained inside a pressurized and insulated shell ( Figure 3). The
bundle of tubes is fixed to the reactor top head and hangs vertically. The catalyst loading, the
feed distribution and the syngas collection are ensured by an original arrangement of the
upper part of the bayonet tube. The inner tube passing through the external pipe avoids the
use of tube-sheet and allows the connections to be carried out outside the reactor shell. Each
bayonet tube is inserted in a flue gas sheath (or chimney) as shown in Figure 2. The narrow
section between the chimney walls and the bayonet outer tube imprisons the hot flue gas
around the bayonet tubes and induces high velocity. The high turbulence developed favours
convective heat transfer to the catalyst.
The reactor-exchanger concept described above has been adapted to the HyGenSys process
characteristics: its maximum capacity is limited by the flue gas flow rate that can be provided
by the biggest Siemens two-shaft gas turbine (SGT-700) that will be used as hot gas
generator. Therefore, the reactor shall work with both process side and heating side in
pressure (process side operating pressure is chosen so that hydrogen can feed the turbine
without any recompression). In addition, the pressure balance between process side and flue
gas side is significant (about 30 bars). The reactor internals and tubes have to be suitable for
the mechanical stress induced by such a pressure difference. This constraint is worse with a
standard SMR (flue gas at atmospheric pressure) but is not shared with most of available
reactor/exchanger which are used to ensure heat exchange between streams of similar
pressure and composition. As a consequence of these two points and for economical reasons,
a compact arrangement of the reactor is awaited.
INDEX çç Sommaire
3.2.2 Material constraints
The extreme operating conditions inside the reactor lead to very high thermal and mechanical
constraints (flue gas temperature can achieve values of about 1300°C). The materials
selection is then essential. Metals and alloys are generally susceptible to carbon attacks when
exposed at high temperature to an environment containing mixtures of CO, H2 and
hydrocarbons. Carbon attack refers to the metal damaging due to coke deposition,
carburization and metal dusting. These three phenomena act at different temperature ranges
between 400 and 1000°C [4, 5], so each one will be an issue for a different part of the
metallic bayonet tube. Specific studies have been carried out to find the most adapted alloys
for the design.
Catalyst loading and unloading is also essential. Since no specific catalyst development was
planned in the CACHET project, the reactor concept has to be compatible with commercial
catalyst. This also applies to the process flow conditions across the catalyst pellets. Loading
and un-loading operations of the bayonet tubes must be feasible and reproducible (uniform
feed distribution through the tube and also between each tube).
An iterative work has been realized to take into account all the factors described above,
essentially to determine the optimum operation conditions, the materials selection, as well as
the tubes and reactor geometry. Final results are presented and discussed in the next sections.
Because of the chemical composition of the syngas which favours the metal dusting
phenomenon, the high nickel alloy 602CA™ 1 has been selected for the internal tube of the
bayonet. As this tube is not under pressurised design conditions (about same pressure inside
and outside the internal tube), only the strength due to its own weight has been checked.
With respect to the external tube, the very high temperature is obviously the main constraint
with regard to creep strength for the material selection, associated with a corrosive
environment. Therefore, after comparison of several materials and thicknesses calculations
according to EN 13445, the external tube is made of several sections of Manaurite™ 2 with
different thicknesses with regard to the temperature variation. The external diameter is set
1
602CA: trademark of Thyssenkrupp VDM material supplier
2
Manaurite: trademark of Manoir Industries material supplier
INDEX çç Sommaire
constant to preserve a correct flue gas flow and then the catalytic space is more important for
smaller thicknesses i.e. in the upper part of the bayonet (Figure 5).
Finally, the flue gas sheaths experience also a very high temperature (up to 1300°C) and very
oxidizing conditions. A simple configuration using tubesheet (metallic and ceramic tubes)
was dismissed due to thermal constraints, junctions' feasibility and ceramic fragility.
Technip and IFP have worked on a new concept, using a hybrid solution: at the bottom of the
tube, a concrete structure ensures the sheath support and a high temperature resistance. The
main upper section of the sheaths, at lower temperature, can be made of metal (hybrid
concrete-metal concept). This solution has the advantage of supporting the high flue gas
temperature at the chimney inlet thanks to the refractory concrete, but avoids the supporting
and thermal inertia problems that a complete concrete filling may have caused. The metallic
tubes for the upper part can be made in 602CA™ with regard to environmental conditions.
Tube
Refractory Internal
concrete bayonet tube External
bayonet tube
Figure 4: Flue gas sheaths – hybrid solution
Several refractory suppliers have been consulted for the construction of the concrete block for
flue gas sheaths. This solution is confirmed by all of them as technically feasible. The detailed
engineering will be done by the chosen refractory manufacturer.
The main problem of the concrete block solution is the associated thermal inertia, which may
generate very long cool down periods and disturb periodical stops of the reactor and delay
maintenance operations. Thus it must be verified how the reactor is cooled once it has been
INDEX çç Sommaire
stopped. Two cases have been studied: free convection (stop of the reactor, with no gas
circulation) and forced convection inside the reactor (mixture circulation with gas at 1300°C
and air at 20°C). Calculated estimations lead to about 35 days necessary to cool concrete from
1300°C to 20°C in the case of free convection, and about 60 hours in the case of forced
convection.
A Computational Fluid Dynamics (CFD) model has been built to help the SMR tubes design
and to guarantee that the heat transfer from hot flue gas towards the SMR reaction section is
high enough without leading to very high metal temperatures causing damage of the tube.
Because the SMR process is made of a bundle of identical SMR reactor tubes, only one tube
has been simulated. The main tube geometry that has been used in the simulation is presented
in Figure 5. The simulated part of the SMR reactor focuses on the tube section that includes
the catalyst section plus the heating sections. The simulator takes 5 domains into account:
1. Dom1: The internal bayonet channel ("syngas" in the centre).
2. Dom2: The wall between the internal bayonet channel and the SMR catalyst ("Metal").
3. Dom3: The SMR catalytic bed for SMR reactions.
4. Dom4: The external wall of the overall tube ("Metal").
5. Dom5: The flue gas that heats up the tube.
The purpose of the inner tube of the bayonet is to heat the catalyst: no chemical reactions take
place in that tube. The wall between the inner canal and the catalyst conducts heat from the
produced syngas to the SMR catalytic bed. To be realistic, the simulator includes the chemical
reactions, the gas temperature and the hydrodynamics in the catalytic bed.
The metallic wall surrounding the catalytic bed is in contact with the very hot flue gas. This
flue gas heats the metallic wall that conducts heating energy to the SMR catalytic bed.
INDEX çç Sommaire
Because SMR performances are closely connected to the energy provided for the chemical
reactions, the modelling of this heating has been performed with great care especially because
the flue gas flow is highly turbulent.
The main degrees of freedom are the tube length and the space opened for the flue gas flow
under the main following constraints: a 900°C syngas temperature at the catalytic bed exit,
injected flue gas temperature not exceeding 1300°C and the wall temperature compatible with
the material limits.
It results from simulator optimisation that the section of the tube in contact with the flue gas
should be 14 m in length to ensure the required heat transfer from the flue gas to the syngas.
The space for the flue gas flow has been limited in radius to induce a high flue gas velocity
necessary to achieve a high heat transfer. This space is enlarged at the entrance to reduce the
flue gas velocity at that point and hence reduce the heat transfer. In the absence of this
enlargement, the metallic wall would have to withstand a too high temperature. This results in
lowering the wall temperature at bottom part of the bayonet.
541 °C 371 °C
450 °C 661 °C
According to the simulator, the flue gas inlet temperature should be equal to 1293°C to
provide enough heating to the SMR and the flue gas temperature at the upper exit is closed to
660°C. This temperature is in the good range to be directed outwards the reactor.
The temperature of the syngas at the bottom of the reactor is 900 °C, both in the SMR section
and in the syngas section which is the exact required temperature for the SMR equilibrium.
The metallic wall is at 960 °C which is in good agreement with the specified max temperature
that can stand this metallic wall.
INDEX çç Sommaire
As a result, according to the simulator, with 320 tubes of 14 meters in length for the reactive
section plus 5 meters for heat exchange purpose, it is possible to reach the expected
conversion: the SMR temperature can reach the expected value of 900 °C to guarantee a high
conversion. The syngas composition leads to more than 72% of H2 in the dry syngas.
The European standard EN 13445 is obviously the preferred code in the context of this
European project. EN 13445 specifies the requirements for design, construction, inspection
and testing of unfired pressure vessels. Design requirements and rules are given in the part 3.
Requirements for creep phenomenon are given in the paragraph 19: the creep allowable stress
is allow = Rm(T) / 1.5, where Rm(T) is the average value of the rupture stress for 100 000
hours at temperature T.
In this section, the whole tube mechanical design including the upper heat exchange part is
considered. The following scheme reminds the tube configuration:
Syngas Colder flue gas
Internal tube
Feed
CH4+H2O
Catalytic
space
Flue gas
chimney
External tube
Hot flue gas
Figure 7: Tube configuration
INDEX çç Sommaire
3.5.2.1 Internal tube
As previously explained, due to the chemical composition of the Syngas which favours the
metal dusting phenomenon, the high nickel alloy 602CA™ has been selected for the internal
tube.
The considered dimensions of this tube in the preliminary model are the following:
- internal diameter of 24 mm,
- thickness of 5 mm.
As this tube is not under pressurised design conditions (same pressure inside and outside the
internal tube), only the strength to its own weight has to be checked.
Tube thickness is then decreased to 3 mm which is sufficient with regard to the strength to its
own tube weight (mass of 40 kg for a length around 20 m).
Several calculations loops have been computed, taking into consideration internal tube design
and results from SMR simulations.
The initial conditions which have been considered in terms of pressure and temperature for
the design of the external tube are the following:
- internal pressure of 35 bar abs,
- tube wall temperature of 1000°C for the bottom part (10% of the tube) and 930°C for
the main part of the tube.
This very high temperature of 1000°C is obviously the main constraint with regard to creep
strength for the material selection, associated with a corrosive environment. The first
calculations are based on the required tube thickness with regard to the creep resistance (100
000h). For these preliminary calculations, no corrosion allowance was considered. They have
been made under these conditions, with several materials adapted to high temperatures in
more or less corrosive environments: 602CA™, HP40 modified and Manaurite™ alloys (XM,
XTM, XOR and 40XO designations).
INDEX çç Sommaire
- The smaller thicknesses are found with Manaurite™ XM and XOR. Calculated
thicknesses with the alloy 602CA™ are not acceptable (several hundreds of mm for
the bottom part at 1000°C).
Following these results, calculations have been revised, considering a target value of 930°C as
the maximum wall temperature, with the possible help of thermal protection. Hypotheses and
design conditions have been updated. The following table synthesises the evolutions:
With these hypotheses, the calculated thickness for the external tube is then 20 mm. The
associated mass for a length around 20 m is then 1.6 tons: a 20 mm thickness is sufficient
with regard to the strength to the own tube weight.
Some optimizations and improvements on the previous design have been made considering
the following assumptions:
- Final internal tube design, with internal diameter of 24 mm and thickness of 3 mm,
- The external tube is made of several sections of Manaurite™ with different
thicknesses with regard to the temperature variation: the external diameter is set
constant to preserve to correct flue gas flow and then the catalytic space is more
important for smaller thicknesses,
- The bottom part of the external tube is made of Manaurite™ XOR (first section of 4 m
length), while upper sections are made of Manaurite™ XM,
- Design pressure has been revised to the differential design pressure, between process
side and flue gas side.
The design pressure considered for these last calculations has been defined from operating
pressures as following:
- Operating pressure for flue gas side:
o Inlet = 4.27 bara
o Outlet = 3.97 bara
- Operating pressure for process side:
o Inlet = 33.5 bara
o Outlet = 31.5 bara
- Design pressure for process side = 37 bara (110% of the maximum operating pressure)
- Design pressure for flue gas side - min / max = 2.5 / 5.1 bara
INDEX çç Sommaire
The internal design pressure for the calculation of the external tube thicknesses is then the
differential P of 34.5 bar.
With these new assumptions, the maximum wall temperature can reach 960°C. The final
design of the external tube is then:
For the spherical bottom part in Manaurite™ XM, the same dimensions as the bottom part in
Manaurite™ XOR are considered.
The unitary mass for one external tube is 1008 kg. Including the internal tube mass and the
catalyst mass, the total mass for one complete bayonet tube is then 1250 kg.
The reactor diameter is determined with regard to the tubes layout. The maximum number of
tubes by reactor is estimated to 327. Two main criteria are considered to implant the tubes:
- An operator should access to each tube, for checking flanges for example,
- Each tube should be independently dismantled.
Finally the suggested tubes layout considers “corridors” between tubes rows:
INDEX çç Sommaire
Figure 8: Tubes layout
NOTE – In the tube layout (Figure 8), one tube is drawn with its corresponding flange and
appears to be in bold.
The considered material is P355GH, with a corrosion allowance of 3 mm. The reactor has
been defined with the following process conditions:
- design pressure of 5.1 bara,
- design temperature of 400°C (defined by the temperature of the inlet gas in the
bayonet tubes).
From the previous layout, the reactor external diameter of 10.4 m has been deduced with
regard to the code rule, which imposes to implement nozzles on heads in a central area limited
by 0.8 x diameter. In addition, an optimization with regard to the previous rule has been
considered, allowing one nozzle beyond this limit of 0.8 x diameter (see Figure 8).
Considering the process conditions, but also the weight of bayonet tubes, the preliminary
reactor dimensions are then:
- Bottom tangent line to top tangent line length: 17 m,
- Heads thicknesses: 40 mm,
INDEX çç Sommaire
- Shell thickness: 34 mm,3
- Support: skirt of 3.5 m long and thickness of 35 mm,
- Mass: about 225 tons (skirt weight of 32 tons included)
The ability of the top head to withstand the tubes bundle weight has been checked by Finite
Element Analysis (FEA). Calculations have been done with ANSYS software.
For reminder, in the way to optimize the reactor diameter, some deviations to the
requirements given in EN 13445 for the design by formulas have been considered provided
the design is validated by FEA.
The reactor diameter is indeed defined by the tubes layout and mainly two deviations would
allow minimising this dimension:
- Waiver for the 0.8 x diameter criteria for the nozzles implementation,
- Waiver for the ligament checking between two adjacent openings.
The nozzles are implemented according to Figure 8. Nozzles external diameter is 250 mm,
with thickness of 30 mm. The minimum outside projection (at the top of the head) is 300 mm.
3
The shell thickness is set to 34 mm in the way to avoid PWHT according to EN 13445 requirements.
INDEX çç Sommaire
The kind of chosen elements is ANSYS SHELL63. Next to 129000 elements are used for the
mesh:
Material model
The considered material is P355GH according to standard EN 10028-2. Its properties are
given in the table hereafter:
Rm @ Tamb 510 MPa (minimal value)
Rp0.2 @ Tdesign = 400°C 196 MPa (minimal value, normalised)
Creep properties for 100000 h Rupture 179 MPa
@ Tdesign = 400°C 1% strain 118 MPa
Considering the Annex O of EN 13445, the Young’s modulus is 182600 MPa at 400°C and
Poisson’s ratio is 0.3. Density is set to 8000 kg/m3.
According to the Annex B of EN 13445 (Design by Analysis – Direct route), the allowable
stresses are defined as following:
3 / 2 R p 0.2 @ T R p 0.2 @ Tdesign
- Without creep consideration, all = min , = 130 MPa 4
design
1.25 1.5
Rrupt @ Tdesign
- Under creep consideration, all = = 143 MPa
1.25
4
First term is for plasticity check (factor 3 /2 used with regard to Von Mises stresses), second term is for
stability check.
INDEX çç Sommaire
Boundary conditions and loads
The base of model is fixed in the way to allow only radial displacement. At the end of each
nozzle, a weight of 1250 kg is applied to simulate the bayonet tube weight. It is distributed
along the outline of the nozzle:
Results
Maximum
displacement
in mm
INDEX çç Sommaire
Minimum
stress
Maximum
stress
Figure 13: Von Mises stresses (MPa) on top head – Bottom side (internal)
The maximum stress of 120.3 MPa is on the internal side of the head, in the knuckle junction:
it is acceptable with regard to the allowable stresses, with creep consideration (see Material
model paragraph). The strength for 1% plastic creep strain is 118 MPa (for 100000 h at
400°C). The areas above this value are very limited (grey areas on Figure 13).
INDEX çç Sommaire
Figure 14: Deformed shape
Figure 15: Von Mises stresses (MPa) on top head – Bottom side (internal)
INDEX çç Sommaire
The maximal stress for the case without internal pressure is 46 MPa for the top head: it is
acceptable with regard to the allowable stresses.
In conclusion, the design of the top head is acceptable with regard to the previous
calculations: it can withstand the tubes bundle weight. Further simulations should be done to
confirm these results, in particular seismic simulations.
CONCLUSION
The concern about efficiency and sustainability was not limited to the process development
itself, but also impact the detailed development of the main equipment of the scheme that is
the reactor-exchanger. This equipment that concentrated many technological challenges was
also a good opportunity for efficiency improvement and CO2 emission reduction.
ACKNOWLEDGMENTS
The authors thank and acknowledge the European Commission (project no 019972) and every
one of the 28 partners [6] in the consortium and especially to BP, NTUA and SIEMENS for
their contribution in HyGenSys development.
REFERENCES
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page75
RÉSUMÉ / ABSTRACT
29A
ABSTRACT
Vacuum vessels belong to a special class of pressure vessels. This is particular the case for those in the
particle accelerators. After an introduction to the High Energy Physics world and its flagship, CERN,
this paper will present particle accelerators, their diverse vacuum vessels and their design and
construction in relation with the European standards and finally some examples of realizations.
- CERN and High energy physics
The European Organization for Nuclear Research (CERN) is an intergovernmental organization with
20 Member States. It has its seat in Geneva but straddles the Swiss-French border. Its objective is to
provide for collaboration among European States in the field of high energy particle physics research
and to this end it designs, constructs and runs the necessary particle accelerators and the associated
experimental areas.
- Particle accelerators
The aim of an accelerator is to accelerate particles up to the speed of the light in order to break these
particles and understand their composition. Particles (protons, electrons, heavy ions) travel for hundreds
of meters to kilometers through chambers with ultra high vacuum (UHV). The particles are accelerated
by radio frequency cavities and guided by electromagnets, often superconducting and therefore enclosed
in cryogenic vessels directly linked with them. Both types of vacuum vessels (cryogenic and UHV)
will be presented.
- Vacuum vessels for cryogenic equipment and UHV chambers
While UHV beam pipes mainly consist of twice 27 km of the so-called cold bore tubes of diameter 53
mm, insulation vacuum for the LHC superconducting magnets and cryogenic distribution lines
represents a total volume of the order of 15000 m3. Large volumes, cleanliness, temperature range of
functioning are that many obstacles to overcome. The application of European technical references,
European Directive (97/23/EC) and standards (i.e. EN13445, EN13458), to these particular types of
equipment is addressed.
Methodologies in the domain of vacuum chambers used in the Particle Accelerators community, in
particular at CERN, are explained. This includes details on design, buckling calculation, materials,
assembly techniques, inspection and testing. Examples of vacuum chambers are given.
INDEX çç Sommaire
A.F.I.A.P.
GENERAL
Vacuum vessels belong to a special class of pressure vessels. This is in particular the case for
those in particle accelerators. After an introduction to the High Energy Physics world and its
flagship, CERN, this paper will present particle accelerators, their diverse vacuum vessels
and their design and construction in relation with the European standards and finally some
examples of realizations.
In the LHC, the last element of this chain, each particle beam will be accelerated up to the
record energy of 7 TeV, before they are made to collide with each other. In addition, each of
the LHC’s injectors has its own experimental hall, where the beams are used for experiments
at lower energies, aiming at analyzing collisions with stationary targets. Detectors observe
and record the results of these collisions. The largest detectors (ATLAS, CMS, ALICE and
LHCb) are installed in four huge underground caverns located around the ring of the LHC.
The aim of an accelerator is to accelerate particles up to the speed of the light in order to
break these particles and understand what they are made of. Particles (protons, electrons, their
anti-particles but also heavy ions) travel for hundreds of meters to kilometers through
chambers under ultra-high vacuum (UHV) [1]. The particles are accelerated by radio
frequency cavities and guided by electromagnets, often superconducting and therefore
enclosed in cryogenic vessels directly linked with them.
INDEX çç Sommaire
A.F.I.A.P.
In the LHC (Large Hadron Collider), the world largest accelerator, while UHV beam pipes
mainly consist of twice 27 km of the so-called cold bore tubes of diameter 53 mm, insulation
vacuum for the superconducting magnets and cryogenic distribution lines represents a total
volume of the order of 15,000 m3.
The specificity of a vacuum vessel in the world of pressure vessels is its negative internal
pressure which induces a different mechanical behavior: instead of being mostly under tensile
stresses, a vacuum vessel exhibit compressive stresses and therefore is systematically
subjected to buckling. This common specificity leads to common design principles and
methodologies. But differences are coming from the physical vacuum: a cryogenic vessel
may operate at 10-5 mbar whereas UHV chambers should usually operate below 10-7 and
often at a much lower pressure. This leads to different selection of material and
manufacturing techniques. The parameters used for the design of vacuum vessels are
presented, the design methodology, the materials selection, the requested quality and the
manufacturing techniques are then given.
In all cases, the conceptual design, which usually does not need accurate studies, is followed
by the detailed design, the latter strongly associated with the manufacturing, itself controlled
by a quality assurance plan. Difficult to be exhaustive and it will not be since what the subject
is extensive. The vacuum vessels considered here are static, the ones which move being
subject to specific rules, are not presented.
Determining the environment is clearly the first step. Although subjects of each sub-chapter
are not necessarily detailed in an exhaustive manner, the list of physical phenomena
influencing the design of vacuum vessels is probably complete.
2. Location
As briefly explained in the introduction, vacuum vessels can sit in large experimental halls if
one considers the situation of colliding particles on fixed targets ; their integration is in this
case not very much of a problem.
INDEX çç Sommaire
A.F.I.A.P.
The beam envelope is the main parameter for the definition of the vacuum chamber geometry
of an accelerator. The conductance plays a role in case of lumped pumps: the pressure
distribution between two pumping ports is parabolic and the maximum value of this parabola
defines the operating pressure. The distributed pumping is a solution, at least partial.
For the vacuum casing of a cryogenic system, the inner piping and its insulation layers define
the inner envelope; extra space also is left free to allow for movements resulting from
pressure and temperature variations.
4. Pressure, forces
The differential pressure on a vacuum chamber is obviously one bar during operation and it is
necessary to check what are all the intermediate steps like, for example, an over-pressure test
for qualification. Very often tests of experimental chambers cannot be fully representative of
the final conditions. As example, the presence of a closed end for the pressure tests, which
will not be installed during operation, will influence the value of resulting loads on the
supports of the chamber and their transmission through the chamber wall.
5. Temperature
Although cryogenic vessels do not require pressure testing at a temperature different from
room temperature, UHV is often obtained by applying a temperature difference during
commissioning, say before operation: e.g. a bake-out or an activation of Non-Evaporable
Getters (NEGs) at temperatures of several hundreds of degrees Celsius for an operation at
room temperature (RT) or an operation at cryogenic temperature after a cleaning at room
temperature.
6. Specific cases
INDEX çç Sommaire
A.F.I.A.P.
Important local heat loads, like synchrotron radiations generated by the accelerated particles,
should be evacuated thanks to good thermally conductive materials. Electrical impedance is
reduced with a good electrical conductivity but, on the opposite, non-conducting material is
required for electric insulation.
Activation of material by the particles hamper the functioning of the equipment, delaying the
access to an accelerator after being switched off and producing radioactive waste delicate to
handle. Elements with long half-lifetime must be avoided in the materials in order to reduce
the activation level and minimize the decay time.
MATERIALS SELECTION
7. General
The materials for a vacuum vessel are similar to the ones for pressure vessels; however,
specific parameters have to be taken into account for vacuum applications. This leads often to
discussions about the required properties.
The material parameters in terms of mechanics are quite numerous but, since the design is
mostly conservative, the three main used are: the modulus of rigidity (Young’s modulus: E),
the elastic (0.2) and rupture (r) limits. Other important parameters for specific elements can
be fatigue limit, creep for bellows or fracture toughness especially for low temperature
applications.
Physical properties are also taken into account depending upon the environment:
- magnetic properties at room temperature as well as at low temperatures;
- electrical or thermal conductivity. Copper and aluminum are the usual candidates when
high conductivity is required and ceramics, like aluminum oxide, when not.
Besides the physical characteristics, specific technological properties should not be forgotten.
Leak tightness of a vacuum chamber is a must but it may be difficult to obtain in case of poor
weldability of the material which can lead to the use of sophisticated and expensive
techniques.
If most of the above parameters are applicable to vacuum vessels, requirements for UHV
chambers can be even more constraining.
The vapor pressure, which depends upon the temperature and the pressure, is to be taken into
account. Materials like zinc, cadmium, magnesium or standard resins which have significant
INDEX çç Sommaire
A.F.I.A.P.
vapor pressure at low temperature are not acceptable when a bake-out is foreseen to obtain an
UHV.
Leak tightness for UHV can only be achieved with specific, and severe, requirements in
terms of metallurgical processes. They are applied to obtain adequate purity, absence of
inclusions, and fineness of the microstructure. Whenever leak tightness has to be guaranteed
across a thin wall, an adequate microstructure in terms of grain size and inclusion content is
specified.
Major advancements in metallurgy and surface physics allow creating more optimized
material which exhibits specific properties. The plain material is replaced by a combination
of materials, each for a specific function based on particular physical parameter. Two
examples among many combinations are the adjunction of a layer of a conductive material
(copper) on a structural material (stainless steel) and, as a barrier to gas diffusion, a thin layer
of aluminum on a structural material (carbon fiber composite).
Besides the material quality, in particular the cleanliness in terms of inclusions and
impurities, the availability (often more than one year of lead time for delivery) and cost are
predominant factors. It is strongly recommended to favor materials of general use in
industrial applications.
Even when raw material of general use is chosen, complex and expensive manufacturing
techniques can totally hamper the final cost. Machining can be very efficient, fast and
therefore cheap although a large amount of costly raw material is lost in chips. In addition,
the precision required for a welded part could need expensive tooling not included in the
initial estimate.
Steel is the most common material for the construction of vacuum vessels, usually austenitic
stainless steel (316LN, 316L, 304L) but also simple carbon steel in some standard cases.
Aluminum alloys (5000 and 6000 series) and copper (OFHC, Glidcop) are used as well.
INDEX çç Sommaire
A.F.I.A.P.
8. Stainless steels
AISI 304L is a general-purpose grade stainless steel. For vacuum applications, the grade
should be purchased through a careful specification, aimed to achieve a substantially
austenitic microstructure and a controlled maximum level of inclusions. However, owing to
the limited amount of alloying elements, its magnetic susceptibility can be subject to increase
by martensitic transformation upon cooling to room or cryogenic temperatures, and following
work hardening. The specific price of this grade is the lowest of this category for a general-
purpose version and can go up by 70% for a vacuum specified wrought product.
AISI 316L is a Mo-containing grade to enhance corrosion resistance and austenitic stability
versus martensitic transformation. Adjustments in Cr and Ni achieve an almost fully
austenitic microstructure. Based on its formability, ductility, increased austenitic stability
compared to 304L and extremely low inclusion content, this grade is favored for the
manufacturing of bellows convolutions. The price of thin sheet products is higher up by a
factor 20 to 30 compared to general-purpose 304L. Wrought products for vacuum
applications have prices in between 304L and 316LN for equivalent metallurgy and
metalworking processes.
AISI 316Ti is an example of a Ti ‘stabilized’ grade. Stabilized grades, which contain higher C
than low C grades AISI 316Ti or similar stabilized grades (AISI 321, AISI 347) are offered
by suppliers as an alternative to low-carbon grades. Nevertheless stabilized grades should
generally be avoided for demanding vacuum applications, since the addition of the stabilizer
elements reduces the cleanliness of the steel and toughness at low temperature.
9. Aluminium alloys
Wrought aluminum alloys are very attractive materials for ultrahigh vacuum systems for
accelerators. The residual radioactivity after machine operation is strongly reduced compared
to steels. It shows high transparency to radiation. It can be shaped into complicated profiles
by extrusion, is completely non-magnetic and, for given components, can be competitive in
terms of cost. High electrical and thermal conductivity are also assets.
INDEX çç Sommaire
A.F.I.A.P.
Extrudable (EN AW-6061, -6082, -5083, 7020, etc.) and weldable alloys are important for
vacuum applications for accelerators. Several alloys within the different families are
considered as readily weldable.
Non heat-treatable alloys, such as EN AW-5083, can be supplied in a work hardened state,
where they can reach significant strength at bake-out temperature. The same alloys in fully
annealed or as fabricated temper show moderate to low tensile properties which might be of
limited interest in several applications.
A relatively high corrosion resistance, machinability and formability into different shapes,
ease of weld for several grades make copper and copper alloys very attractive for vacuum
components, in particular for those applications where high electrical and/or thermal
conductivity are relevant properties.
For very specific types of vacuum chambers (transparent to particles) quoted above like beam
pipes for the experiments installed in colliders e.g. LHC, and windows, the selection of
materials becomes a multi-parameter problem which can be eased by the use of non-
dimensional parameters [2], [3]. It has been shown that the selection can be based on the non-
dimensional parameter (X0E1/3) that gives a figure of merit for the materials (Table 1) and, for
this specific case of beam pipes, Beryllium and Carbon Fiber Composite outrun by a large
factor aluminum, titanium and steel.
Table 1
Figures of merit of materials in terms of transparency
INDEX çç Sommaire
A.F.I.A.P.
As conform to what has been described above, one can distinguish two types of vacuum
vessels: the cryostats, in particular the ones of the magnets aimed at thermally insulate these
from the outside, and vacuum chambers, under UHV, aimed at the circulation of particles.
The first type of vessel, although not submitted to a pressure higher than 0.5 bar, is set to fall
under the European Directive 97/23/EC. Indeed, from the structural integrity of this vessel
will depend load cases of the pressurized vessel, in this case a cryo-magnet. The global risk
analysis of the equipment shall take into account the possible failure cases of the vacuum
casing and their consequences on the pressurized equipment shall be assessed. On the other
hand, the consequences of failure of the pressurized vessel on the design of the outer casing
shall be assessed as well.
A way to be compliant with the Safety requirements defined in the Directive is to apply
European Standards harmonized with these. In the framework of pressure vessels, the
standards, prepared by normalization committees, do contain rules for ensuring that the
equipment are properly designed and documented; the rules address among other subjects,
the design, material choice, manufacturing, inspection, documentation and certification
processes by notified bodies.
Part of European Standards dealing with the design of cryogenic vessels, e.g. EN13458 [4],
focus on the design of vessels submitted to an external pressure. These specialized standards
list the load cases to be taken into account either for the outer casing or the internal vessel,
which is under pressure. Some of these, for example the loss of vacuum insulation, have to be
complemented by special load cases typical to superconducting magnets, like the analysis of
consequences in case of quench of the magnet.
At the time LHC cryo-magnets were designed, the European Directive for pressure vessels
was not of application yet and the European Standards were either not existing or much less
developed. Load cases were defined through a risk analysis and, for what concerns the outer
vessel rules defined in CODAP 95 were applied for the magnets designed by CERN, namely
the dipole and quadrupole magnets. Other magnets, under the responsibility of US and
Japanese Laboratories, were designed according to the ASME code. Formulas, still present in
today’s codes, were applied.
The second type of vessel usually cannot be considered as directly linked to pressure
equipment. Own regulations should therefore be applied and the common practice is to treat a
vacuum chamber as a pressure vessel with some arrangements.
It is important that at least guidelines are provided. This is in particular the case at CERN
with the Safety Regulation SR-M for mechanical equipment, for which the design of special
equipment has to be assessed by CERN’s Safety Commission.
Among vacuum chambers, those which geometry is rather simple (cylindrical with usual
dome ends) can be treated as above, and the “design by formulas” proposed in standards like
CODAP or EN13445 can be strictly applied.
INDEX çç Sommaire
A.F.I.A.P.
Those vacuum chambers which geometry is made complex for the purpose of particle physics
have to be treated separately. Even though the design of such chambers is addressed in the
next chapter, it is worth mentioning here that developments in Finite Element Analysis codes
played an important role in the way such chambers were approached. A procedure based on
the non-linear analysis of the complex chambers was applied; this procedure is in many
points similar to what is proposed in the Annex B of the European Standard EN13445.
Not to forget a special mention to the European standard EN 13458-2 [4] which treats
specifically the “static vacuum insulated vessels” of the “cryogenic vessels” in a way very
similar to the codes mentioned above.
DESIGN
12. Basics
As basic rule in mechanics and structural resistance of equipment, stresses generated by the
loads should remain in the elastic range of the material used. The equivalent stress (according
to Von Mises or Tresca criterion for example) shall therefore not exceed the elastic limit ( 0.2)
anywhere in the structure, taking into account safety factors which aim at cover uncertainties
on the knowledge of the loads and the material properties. The directive 97/23/EC states in
particular that “the permissible membrane stress for predominantly static loads…must not
exceed…2/3 of the yield limit (stainless steel and aluminum alloys)”.
Vessels under vacuum can be considered as submitted to an external pressure and the
resulting membrane stresses are compressive. Under these conditions and considering
elementary radial displacements exist, the solving of the system of equations of equilibrium
of an elementary part of the shell submitted to an external pressure leads to find a singular
point in the behavioral curve pressure-displacement of the element. This point corresponds to
a theoretical critical pressure over which the theory of small displacements cannot be applied
anymore. A steep change in the first derivative of the behavioral curve pressure-displacement
of the element occurs. Membrane strain energy is converted to bending strain energy, leading
to instability and potential buckling of the vessel.
INDEX çç Sommaire
A.F.I.A.P.
Buckling of equipment can be encountered, independent to what the stress state is. Even
equipment for which stresses are well under the elastic limit can fail due to buckling.
Buckling is a non-linear phenomenon and is strongly influenced by defects or geometry
imperfections of equipment inherent to the manufacturing processes.
Although in practice it is quite easy to elaborate a design to stay in the elastic range of a
material, buckling is more difficult to tackle.
14. Procedure
In the pre-design phases and when controlling the final design, somewhat simple analytical
methods could be more efficient than the structural analysis programs. They allow for a quick
analysis of the effects of various geometrical parameters.
The use of structural analysis packages, based on the Finite Element Method (FEM) is
prescribed at a second stage and is a must for the detailed design of vacuum vessels of
complex shape or which have manufacturing tolerances over the range accepted by a design
code. Two main families of elements ought to be used for these computations: the solid
elements for the “thick” machined parts and the shell elements more appropriate for thin shell
structure < 100). Readers used to carry out stress analyses for pressure vessels know that the
through-thickness stresses of shell elements are assumed linear and integrated in membrane
(constant term) and bending (linear term). Some of these elements can be used for buckling
and non-linear analyses.
In these cases, when the predesign is complete, the procedure applied is the following:
1. The first step of the analysis is to carry out a linear elastic analysis which gives all the
information on a structure: displacements, strains and stresses.
2. The second step consists in determining the eigenbuckling values and geometric
shape of the main modes of instability. The stress state given in step one is taken into
account for step two.
3. The geometry of the vessel under study is deformed according to the most relevant
buckling modes given by step two, in general those characterized by a low buckling
eigenvalue. For a conservative analysis, the geometry is deformed as much as to
correspond to the highest manufacturing tolerances one can achieve.
INDEX çç Sommaire
A.F.I.A.P.
4. Based on the deformed geometry, a non-linear analysis is carried out, with a gradually
increasing external pressure on the vessel. The buckling phenomenon is reached when
the computation fail to converge. Plotting the deformation v/s load will allow for
determining the shake-down factor on the linear buckling value (given at step 2); it
can be as high as 10 for very thin shells (see figures 2 and 3).
15.
Designed to allow particle beams to escape the vacuum chambers without interaction,
windows are thin and often manufactured with a material adequate for the function. Special
care is required for these critical items.
The best shape to resist to the external pressure is a spherical dome but the manufacturing is
difficult and the external ring must be quite rigid. A flat window is the other common option
but, when put under pressure, the deflection is not negligible.
Non-circular windows exhibit high loads concentrated in the corners; somewhat difficult to
overcome in case of presence of a weld and a source of nightmare when clamped inside
flanges.
Bellows are another critical item providing the capacity to minimize stresses due to
displacements during commissioning and operation and easing the assembly of misaligned
parts. Designed to bring flexibility, they are thin and therefore fragile. They are either
mechanically formed or hydroformed from thin tubes or assembled by welding series of
individual annular rings.
The design of bellows is treated in the general codes for pressure vessels but a specific code
is issued by the manufacturers, the Expansion Joint Manufacturers Association, Inc.
(EJMA), [7] recognized as the authority on metallic bellows type expansion. A norm EN
14917 [8] was recently published by CEN.
Bellows are designed to withstand a given number of cycles of axial expansion/compression
over the expected life of the system. They accept an angular movement but large offsets
dramatically decrease the fatigue limit. A double bellow “à la cardan” (two bellows with a
tube in-between) is the recommended solution for the latter case when space exists. If the
neighboring tubes are not correctly supported, the bellows may buckle immediately or after
cycling.
The imagination of the designer is without limits. Any shape can be designed and analyzed
with the available tools and following the procedure described. But one should be careful
with the analysis of the results, in particular with the interpretation of the local stresses (true
or generated by the FEM mesh). Curved surfaces are preferred to flat panels wherever
possible.
INDEX çç Sommaire
A.F.I.A.P.
However, it turns out that special shapes are also more difficult to manufacture. Major
progress in the extrusion technology allows now for fancy shapes with multi-channels.
Welded tubes can also be ring-stiffened to compensate for thin walls or for large dimensions
leading to large deflections.
Materials for manufacturing are available in various states: raw products like blanks or
sheets, or semi-finished products like extruded elements (aluminum, copper), molded, forged
or sintered.
The choice is usually led by cost; however the final quality in terms of vacuum is function of
the manufacturing techniques.
Defects due to internal impurities of the material can provoke leaks; appropriate elaboration
techniques, properly carried out, give solutions. Forging, extruding and laminating will
squeeze the impurities but also enclose them. Laminated metal sheets will usually remain
leak-tight unless submitted to large deformations due to forming (e.g. deep drawing, 180°
folding…) or in the heat affected zones next to the welds.
Machining has the tendency to open impurities with a risk of leak, even starting with a good
quality of material.
The choice of joining technique will depend on the need and frequency of dismounting the
assembly in the future.
Knowing this, permanent joining techniques, like welding, or mechanical joining techniques,
by flanges for example, have to be compared. The cost in terms of money and space of
flanges is to be put in balance with the one of welding, which will require subsequent cutting
and rewelding in case the area needs to be accessed. Pieces of equipment, tooling,
accessibility and assembly operations have their own cost.
19. Flanges
Flanges are industrial standard parts available on the market. However, metal-sealed flanges
have to be developed for specific needs. The main parameters for their design are the strength
of the material, to withstand high forces for bolting or clamping, and the quality and hardness
of the surface where the seal is located. The seal itself should keep its overall rigidity (elastic
INDEX çç Sommaire
A.F.I.A.P.
behavior) while its surfaces in contact with the flanges should plastify to enforce the leak-
tightness.
20. Welding
Welds are sources of impurities and defects. Therefore, they should be carefully managed.
Welding techniques will not be treated in details but hints in order to avoid potential
problems are given.
The first point, which sounds obvious but is sometimes forgotten due to the importance of the
other arguments, is the weldability of the material. To fulfill their role of mechanical
resistance, the welds should be designed and executed according to the rules of the
construction codes. Depending on the requirements, a reduction factor (0.85, 0.7) is applied
in the calculation of the stress level and it is therefore recommended avoiding the presence of
welds in a highly stressed zone.
Specific rules must be applied to UHV chambers to avoid contamination and virtual leaks
generated by gas pockets. Welds have to be performed on the side in contact with vacuum
and trapped volumes between two welds are forbidden. The use of filler metal is not
recommended. Grooves help to minimize heat propagation along the walls and can be useful
in case of subsequent cutting and rewelding. Crossing welds lead to a remelting of the
material, which creates defects. Figures 4 and 5 show typical weld configurations for vacuum
components.
A classical source of leaks in stainless steel vessels is the macro inclusions from entrapped
slag or refractories added to protect the steel bath during the melting phases. These inclusions
are segregated during the solidification and cannot be fully removed because of an
insufficient discard of the top of the steel ingot. Slag trapped in the ingot gives elongated
macro inclusions during the hot working (forging, rolling) of the steel.
The preferred and most used welding technology is the TIG (Tungsten Inert Gas). A very
good inert gas (argon) protection is a prerequisite. Electron beam welding is another solution
but it is not as flexible due to the need for a vacuum envelope. In case of thin shells, plasma
welding is recommended.
INDEX çç Sommaire
A.F.I.A.P.
Finally, one should remember that a joint of quality will also depend on the mechanical
preparation of the parts to be welded and where the welding process will be carried out,
whether in a workshop or in-situ. Whatever the techniques and the quality, not more than
99 % of a series of welds will be leak-tight. Visual inspection must be systematic and X-rays
are highly recommended if not compulsory.
21. Brazing
Brazing is an expensive solution which requires a furnace filled with a protective gas and jigs
to keep the parts in position during the thermal cycle. It is often the only solution to join
dissimilar metals or ceramics together. The design of the interfaces to be brazed should be
done according to strict rules taking into account the expansion of the materials allowing the
brazing material to flow through the gap. The surfaces are etched before and a high quality
cleaning after brazing is very much important. Indeed, the appearance of corrosion on thin-
walled stainless steel components is very often due to a deficient cleaning of the chloride-
based flux.
22. Gluing
Gluing is quoted here for completeness. Configuration of joints is similar to the ones used for
brazing but the curing temperature is lower and sometimes even ambient. Degassing glued
joints is a problem, even for high performances glues (epoxies …) which are commonly used
for repair of leaks. There is still room for development of the gluing of vacuum parts.
CLEANLINESS
Cleanliness of the delivered parts is essential to minimize degassing. When particles circulate
in the beam pipes, the surfaces are bombarded by particles generating degassing and therefore
increasing the “dynamic” pressure. Heavy molecules, carbon like ones, are more harmful
than, for instance, hydrogen molecules which might be degassed from the constituent metallic
parts. To reduce residual gas pressure, the surfaces under vacuum of all the constituent parts
must be free of organic additives, oils, greases, packaging residues. Cleaning can be
performed by well-known processes which should be compatible with parts to be assembled
afterwards (materials, temperature, abrasion…). A better solution lies in avoiding the
contamination of the parts during the fabrication (as for example, avoiding machining with
silicon oils which are difficult to remove). The last step of the cleaning process is to fully
empty and dry the parts; the design should take this into account by suppressing the zones
where puddles will remain and avoiding closed volumes. A procedure meant to measure and
checks the amount of contaminant still present on the surface or a quantity related to it and
compare it with the acceptance level ends the process. This limit defines the cleanliness.
INDEX çç Sommaire
A.F.I.A.P.
LEAK DETECTION
An important step of the commissioning of a vacuum system is the leak detection. Simple
actions will ease that work: provide access to the vacuum chamber, in particular to the joints
(helium spraying, clam shell), avoid virtual leaks by suppressing closed volumes like the
presence of bolts.
The time of flight technique will be used for large chambers; the goal is to provide an easy
way for the helium to flow through radially in order to minimize the perturbation when
measuring time of flight over long distances.
QUALITY ASSURANCE
A good design is the first step for a good quality but it might be entirely spoiled by a bad
quality of manufacturing. What is required should be clearly identified in the technical
specification and in the drawings, using international standards. All the steps of the
procurement, manufacturing and assembly should be detailed beforehand and controlled
according to agreed procedures. Design codes for pressure vessels contain specific chapters
on that subject, which are sometimes referred to for the construction of vessels under
vacuum.
In addition and even if it is not full proof, the certification of a manufacturer according to
ISO9001 will ease the application of a strict quality assurance plan and shall constitute one of
the requirements for choosing a manufacturer.
Including regular inspections at the premises of the contractor either by technicians and/or by
an accredited third party could be considered as a quite heavy workload but it is the only way
to obtain the requested quality. Look in particular at the material used versus the specified
one, at the tolerances of the parts versus the manufacturing drawings, at the procedures and at
the cleanliness.
CONCLUSION - EXAMPLES
Particularities of building vacuum vessels for Particles Accelerators have been outlined. Their
design, materials, manufacturing and assembly techniques and the link to Standards and
Codes, mainly used for pressure vessels, were clarified. Giving examples of realizations is
probably the most appropriate manner of illustrating and coming to a conclusion.
Among so many different vacuum chambers built at CERN over half a century, three
representative examples are given.
Each of the eight continuous cryostats of the LHC is 2.8 km long. It consists of vacuum
vessels designed to enclose the main magnets operating at 1.9K in superfluid helium (figure
6). These vessels of a diameter of 1.1 m have been manufactured in carbon steel with
stainless steel flanges.
INDEX çç Sommaire
A.F.I.A.P.
Figure 6: The LHC continuous cryostat with one interconnections zone under welding
Particle beams often circulate in more fancy UHV chambers. The example given is the rib-
stiffened chamber (70mm x 175mm) manufactured in thin Inconel 625 and developed for the
medical accelerator MedAustron (figure 7).
The accelerating cavities are UHV vessels working at cryogenic temperatures aimed at
transferring energy to accelerate the particles. High purity copper is the preferred material.
The LEP (predecessor of LHC) RF cavities were thin niobium vessels hydroformed and EB-
welded with a diameter of 0.76m (figure 8). Niobium-coated copper cavities were also
largely developed for this application.
These three examples give only a small idea of the variety of vacuum chambers in the
Particle Accelerator world. A limited number of norms exist to design and manufacture them.
But common rules have been established to ensure their design is reliable and safe.
ACKNOWLEDGEMENTS
The authors thank all their colleagues with whom they learned how to design vacuum
chambers and all those who provided documents for this paper.
REFERENCES
[1] CAS - CERN Accelerator School and ALBA Synchrotron Light Facility: Course on Vacuum
in Accelerators, Platja d'Aro, Spain, 16 - 24 May 2006
[2] Ashby, M.F. – ‘Materials Selection and Process in Mechanical Design’ –
Butterworth Heinemann, Oxford, (1999 ISBN 0-7506)
[3] C. Hauviller – Design of Vacuum Chambers for Experimental Regions of Colliding
Beam Machines – IEEE 1993 Particle Accelerator Conference (May 1993)
[4] European Committee for Standardization (CEN) – Cryogenic vessels – Static
vacuum insulated vessels – EN 13458-2
[5] Syndicat National de la Chaudronnerie, de la Tôlerie et de la Tuyauterie Industrielle
(SNCT) – CODAP – French code for construction of unfired pressure vessels
[6] European Committee for Standardization (CEN) – Unfired pressure vessels – EN
13445
INDEX çç Sommaire
A.F.I.A.P.
[7] Expansion Joint Manufacturers Association Inc. (EJMA) – 2003 Eight Edition
EJMA Standard and 2005 Addenda
[8] European Committee for Standardization (CEN) – Metal bellows expansion joints
for pressure application – EN 14917
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page76
RÉSUMÉ / ABSTRACT
5A
MÉTHODES DE VÉRIFICATION DE LA RÉSISTANCE PAR ANALYSE DES
MODES DE DÉFAILLANCE
RÉSUMÉ
La conception des appareils à pression a été historiquement fondée sur l’utilisation de formules
analytiques (Design By Formula – DBF). En complément à l’approche DBF la plupart des Codes de
conception proposent depuis longtemps un mode de dimensionnement par analyse. Le dimensionnement
par analyse a été basé à l’origine sur la méthode dite « d’analyse des contraintes » de l’ASME qui date
des années 1960. Cette méthode est souvent d’une application délicate du fait des difficultés inhérentes
Mercredi
aux choix de catégories de contraintes qu’elle impose d’effectuer, de plus elle prend en compte
essentiellement les effets de la pression et dans une moindre mesure ceux de la température. Pour palier
aux inconvénients propres à l’analyse des contraintes une méthode de vérification de la résistance dite
« par analyse des modes de défaillance » a été développée dans le cadre de l’écriture de la norme
Européenne EN13445. Cette méthode a été introduite en 2002 dans l’EN13445-3 dont elle constitue
l’annexe B, elle a également été introduite depuis la version 2010 du CODAP® dont elle constitue le
chapitre 10.3. L’article s’efforce de présenter les grandes lignes de la méthode dite « par analyse des
modes de défaillance », pour cela il redéfinit les modes de défaillances qui sont considérés et replace
d’abord la méthode par rapport aux autres procédures de dimensionnement possibles, il introduit ensuite
les principes de pondération des charges qui sont utilisés, il se termine par une description, à des fins
d’illustration, du traitement du mode de ruine par déformation excessive tel qu’il est codifié dans le
CODAP® 2010 ou l’EN13445.
ABSTRACT
Design of pressure vessels has been historically based on analytical solutions, called Design-By-
Formula. In addition to this approach, the most of construction Codes propose the alternative approach,
called Design-By-Analysis (DBA). This approach was originally proposed by ASME. The application
of DBA is not an easy task for non expert analyst due to the choice of primary and secondary stress
categories. In addition, this approach takes into account only the effect of pressure and temperature.
To overcome theses difficulties, a method of resistance verification has been developed and introduced
in European standard EN13445 (Annex B, part 3), and CODAP® version 2010 (chapter 10.3), called
analysis based on damage modes.
This paper presents the foundation of this approach with definition of damaged modes considered and
compared with other possible design procedures. Then it introduces the basic rules of applying the
partial safety factor. Finally to illustrate the DBA approach, an example based on gross plastic
deformation proposed by CODAP® 2010 and EN13445 is presented.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page77
RÉSUMÉ / ABSTRACT
89A
APPROCHES BASÉES SUR LA DÉFORMATION EN CONCEPTION DES
APPAREILS SOUS PRESSION – UNE EXTENSION DE LA MÉTHODE DBA
BASÉE SUR LA RUPTURE DUCTILE
RÉSUMÉ
Pour la construction d´équipement sous pression il existe dans le monde différentes normes de
construction. Outre les normes nationales comme les directives pour les équipements sous pression,
ou la norme américaine ASTM, il existe aussi depuis environ dix ans la norme européenne EN 13445,
qui peut être consulté pour la construction de réservoirs sous pression. On y fait la différence entre la
construction par formules (DBF = Design by Formulae) et la construction par méthode d´analyse (DBA
= Design by Analysis). Ces deux procédés ont cependant le même inconvénient : les réserves en
contrainte des aciers à haute résistance pour réservoirs sous pression ne peut être mis à profit que
insuffisamment, étant donné que des exigences concernant le rapport Y/T sont formulées, exigences
qui ne peuvent guère être satisfait en vertu du critère de Considère pour les aciers à haute résistance.
D´ailleurs, ces exigences sont néanmoins basées sur simple considérations non scientifique étaient trop
sur la coté sécure et ne peuvent pas décrire la résistance réelle ou sa tolérance de dommage. Pour cette
raison un concept d´élongation limite, avec lequel on peut décrire la résistance d´aciers pour réservoirs
sous pression, est actuellement en développement. Ce procédé doit à l´avenir être intégrer à la norme
EN 13445 pour pouvoir optimiser la construction de réservoirs sous pression avec des aciers ferritiques
à hautes résistance. Dans l´article suivante les fondements théoriques de cette nouvelle approche sont
décrits. De plus, les réserves non utilisées de matériau sont montrées numériquement avec un
dimensionnement d´un réservoir sous pression selon ce nouveau procédé.
ABSTRACT
Worldwide several design codes provide different rules for the design of pressurised components. Apart
national design codes, e.g. such as the pressure equipment directives or the American ASTM, the Eu-
ropean Standard EN 13445 can be used for the dimensioning of pressure vessels. This code has been
recently developed and introduces two different approaches called Design by Formulae (DBF) and
Design by Analysis (DBA). However, the utilisation of the potentials of high strength steels is generally
being restricted by both approaches. Load bearing capacities of these materials cannot be fully exploi-
ted by far as they can hardly meet the prescribed limits for the Y/T – ratios. However, these requirements
are strongly artificial and originate from times where scientific knowledge and modern steels were not
available. They describe neither the effective load bearing capacity nor the damage tolerance. Due to
this reason a new strain-based damage design concept is being developed to describe the load bearing
capacities of ferritic steels. The proceeding shall be transferred into the EN 13445 to improve the design
of pressure vessels with high strength steels. The article describes the theoretical background of the
new concept. Furthermore, the so far unexploited material resources are quantified for a model pressure
vessel made of steel P500Q.
AFIAP - ESOPE 2010 Programme
INDEX çç Sommaire
INTRODUCTION
Design by Formulae
The European standard EN 13445 [1] provides two general approaches for the design of pressure
vessels. According to the Design by Formulae, for a given material the required wall thickness of the
vessel is calculated from the design stress f which can be calculated by the following equation:
According to this the tensile strength of the material becomes the relevant value for the design
whenever the Y/T - ratio exceeds 0.625. Hence, for a Y/T – ratio of 1 the allowable stress drops down
to 41.7 % of the tensile strength. By this it is quite obvious that the application of high strength steels
is strongly hindered and more realistic safety factors have to be developed.
Design by Analysis
A research project sponsored by FOSTA [2] ties up to that problem by determination of the real
service capacity of pressure vessel steels with the objective to optimise the design code and improve
the application of high strength steels. Resulting from the scientific developments of the last two
decades, micromechanical damage models (e.g. the Gurson-Tvergaard-Needleman model [3-7]) could
be used as appropriate approaches to compute the critical loading situations for pressure vessels made
of ductile materials. However, the application of such micromechanical damage models for safety
assessments is actually not widely accepted as these models show pronounced mesh size dependence
in case of the common locally formulated model definitions [8-10]. Furthermore, the selection of
model parameters strongly influences the simulation results, but a reliable and reproducible
quantification of material parameters is connected to extensive experimental and numerical
investigations. It can thus be concluded that a simple failure criterion, which can be derived easily and
reproducibly would be a more beneficial tool for the design of pressure vessels. The damage curve
[11-15] can probably fulfill these demands in case of ductile failure behaviour. It is an empirical
failure criterion which expresses the equivalent plastic strain to ductile crack initiation as a function of
the stress triaxiality h. In addition, the damage curve can more easily be implemented into the DBA
procedure than micromechanical damage models.
The project aims to give more realistic limit strains for the DBA procedure. By the damage curve the
burst pressure of a vessel finally can be estimated– consequently, the limit strains given by the damage
curve concept should be subjected to adequate safety factors so that a safe design scenario can be
provided. The derivation of such safety factors is based on numerical investigations on model pressure
vessels made of high strength steels. Initially, these investigations aim to predict the amount of the so
far not utilised material capacities by comparing the predicted burst pressure with that value of
INDEX çç Sommaire
pressure yielding from the DBA procedure. These results can be used for the definition of new,
improved safety factors and limit strains. Considering metallurgical aspects it has become evident that
the micro structural influence on the equivalent plastic strain to ductile crack initiation and hence on
the upper shelf toughness properties is rather pronounced. Here, mainly the material´s internal
cleanliness is important as both non-metallic inclusions and precipitations are nucleation sites for voids
and cavities during ductile failure. Thus, the micro structural influence on the upper shelf toughness
has to be taken into account for the definition of new limit strains. Here, the Gurson-Tvergaard-
Needleman (GTN) damage model is favourable. Due to its micromechanical character, it describes the
evolution of damage in the material´s micro structure and considers the effect of damage on the strain
hardening behaviour by a modification of the von Mises yield potential. As the void evolution law
relies on the volume fraction of void nucleating micro structural constituents, the GTN model is able to
reflect the micro structural influence on upper shelf toughness properties. In consequence, the GTN
model can also be used to compute the course of the damage curve for steels of a given strength level
containing larger amounts of non-metallic inclusions and precipitates.
The investigations have started with the determination of damage curves for the selected high strength
pressure vessel steels. As a second task, it was shown that the GTN model can be applied to predict the
course of the damage curve using realistic information on the amount of inclusions and precipitations.
After this validation of the general approach, the GTN simulations were repeated with parameter sets
representing steels with poor internal cleanliness resulting in a low upper shelf toughness level. Like
this, the “worst case” damage curves could be generated which form the basis for the definition of
more realistic limit strains for the DBA procedure. Finally, the damage curves which are also valid for
steels with rather poor upper shelf toughness properties are used to compute the burst pressure of a
model vessel so that the actual safety margin can be expressed.
Damage curve
INDEX çç Sommaire
coalescence. This is at least approximately represented by stress triaxialities h > 0.4, and in that range
the damage curve can be described by equation 1.
pl pl
e c1 e c2 h e i (Eq .1)
with:
c1, c2: fit parameters,
h: stress triaxiality,
pl
ei : equivalent plastic strain necessary to provoke ductile failure at hydrostatic stress state.
In case of ductile damage, using crack initiation criteria is not sufficient for toughness modelling as
due to strain hardening, even after ductile crack initiation the sample´s loading can be increased.
Hence, the behaviour of elements that have exceeded the critical loading defined by the damage curve
has to be modelled as well. This can be done by a modification of the stress-strain relationship. As
long as no crack initiation criterion is fulfilled, the material´s stress-strain response is according to the
given flow curve, but once the crack initiation criterion is fulfilled, the modified stress-strain
relationship *(e) should be used which reads:
* e 1 D e , (Eq. 2)
with:
D: damage variable,
(e): material´s flow curve.
The damage variable D is calculated according to a damage evolution law. As D = 1 leads to *(e) = 0,
a simple possibility to define the damage evolution would be to specify a characteristic fracture strain.
However, in FE - calculations the value of the equivalent plastic strain at failure depends on the
element size and should hence not be understood as a material characteristic value for the specification
of the damage evolution law. Instead, the damage evolution law should be specified in terms of
fracture energy dissipation Gf [15], which is the energy required to open a unit area of crack.
The GTN damage model belongs to the group of micromechanical ductile damage models. Gurson [3-
4] analysed a spherical void in an infinite perfectly-plastic matrix material in order to define a yield
potential for a porous plastic material. This yield condition was later modified by Tvergaard and
Needleman [5-7]. It can be used to describe isotropic strain hardening and it is expressed as follows:
2
* 3 *20
v 2 q f cosh q
2 2 H
1 q3 f (Eq. 3)
GTN 1
y y
with:
v: von Mises equivalent stress,
y: matrix material yield stress,
H: hydrostatic stress,
q1, q2, q3: model parameters to define the shape of the yield surface,
f* effective void volume fraction.
INDEX çç Sommaire
The effective void volume fraction f* is a function of the void volume fraction f, which is the ratio of
the total volume of all cavities to the volume of the body. The effective void volume fraction f* was
established in order to consider void interaction effects starting at a critical void volume fraction fC.
The function to calculate the effective void volume fraction f* reads:
f; f fC
f *f (Eq. 4)
f c f f C ; f fC ,
with:
fu* fc
,
f f fc
fc : critical void volume fraction,
ff: void volume fraction at macroscopic failure,
fu*=f*(ff): effective void volume fraction at macroscopic failure.
In this function, the accelerated damage evolution after reaching the critical void volume fraction fC is
described by the scalar value .
The void volume fraction f is calculated according to a void evolution law. It considers growth of
existing voids with the initial volume fraction f0 as well as nucleation and growth of a second void
population. The nucleation of new voids of the volume fraction fN is assumed to be a strain controlled
damage process [16]. The damage evolution law reads [17]:
1 e pl e
2
f fgrowth fnucleation 1 f e kk
fN
pl
exp N
e pl (Eq. 5)
SN 2 2 SN
with:
e kkpl : plastic volume dilatation,
fN: fraction of void nucleating particles,
SN: standard deviation of strain controlled void nucleation,
pl
e : equivalent plastic strain,
eN: characteristic strain of void nucleation,
e : rate of equivalent plastic strain.
pl
Hence, damage modeling in numerical simulations including the GTN model requires the
quantification of the parameters q1, q2, q3, f0, fN, eN, SN, fC and .
RESULTS
Material
Three different steel grades are investigated: a normalised steel P355N with ferritic – pearlitic
microstructure, a bainitic steel P500Q and a martensitic steel P690Q. Plates of these materials were
delivered in gauges of 20mm, 50 mm and 80mm. In addition, the investigations also include
examinations of weldments both in the as welded and in the post weld heat treated condition. Due to
the variety of materials the following chapters exemplify the investigations for the base material of
steel P500Q in 50 mm thickness.
INDEX çç Sommaire
The chemical composition of the steel P500Q is presented in table 1. It can be concluded that the C
content was selected to adjust the austenite stability and the critical cooling rate for the quench and
temper process. The contents of S and P are low as these elements are known to be detrimental for
toughness properties.
For the derivation of damage curve, samples shall be loaded until crack initiation. These experiments
are accompanied by Finite Element (FE) simulations. For the time step corresponding to the loading at
crack initiation, the equivalent plastic strain and the stress triaxiality are evaluated in the spot where
the micro crack has nucleated. Thus, it is evident to perform a series of such investigations in order to
generate enough points along the damage curve to be able to mathematically describe its function. For
this purpose, fracture mechanics tests with pre-cracked CT25 samples in varying thickness were
conducted applying the direct current potential drop technique in order to monitor the ductile crack
initiation.
Figure 3 shows the FE – model for a CT sample of 10 mm thickness. It consists of hexahedral 3-D -
elements with 20 nodes. To capture the inverse square root singularity of stress and strain in front of
the crack tip most effectively, the midside nodes of the elements around the crack tip were duplicated
and relocated to their ¼ point position during the calculation. Altogether, eight CT25 specimens were
tested in gauges of 7.5mm, 10mm, 12.5mm, 15mm, 17.5mm, 20mm, 22.5 and 25mm. As the
experimental results did not cover the whole range of stress triaxialities h > 0.4 the damage curve was
matched by a least-squares fit from experimental results. Figure 4 shows the damage curve for the
P500Q pressure vessel steel.
INDEX çç Sommaire
1.5
0.5
0.0
1.0 1.5 0.02.0 0.5
2.5
stress triaxiality h, -
Figure 3: FE – model of 10 mm thick Figure 4: Damage curve of P500Q base material in 50 mm
CT sample. thickness.
GTN calculations
The scientific approach of the investigations is based on the assumption that the GTN model is able to
express the micro structural influence on the damage curve and consequently on the upper shelf
toughness properties. In order to validate this approach, the volume fractions of non-metallic
inclusions and carbides were determined in metallographic investigations as fractographical analysis
proved these micro structural constituents being responsible for the nucleation of two void populations.
For this purpose micro sections were examined by light optical microscopy and evaluated statistically.
In figure 5 a polished micro section is shown at a magnification of 200:1. As the initial void volume
fraction f0 equals the void volume fraction of inclusions in the material, the average volume fraction of
inclusion (0,185 Vol.%) was identified by computer-aided image analysis and afterwards used for the
numerical simulations. Furthermore the cavitation of secondary voids is associated with the content of
small carbides in the matrix. These precipitations are uncovered by Sodium Picrate or Murakami
etching (figure 6). Afterwards, the volume fraction of carbide precipitations (0,21 Vol.%) was used as
volume fraction of newly nucleated voids fN in the GTN simulations.
Figure 5: evaluation of inclusion volume fraction. Figure 6: evaluation of carbide volume fraction.
The model parameters q1 = 1.5, q2 = 1.0 and q3 = 2.25 were selected as suggested by Tveergard and
Needleman while the critical void volume fraction fC = 3.3 Vol.-% was identified with cell model
calculations according to [18]. The parameters N = 0.7, SN=0.2 and =2 were selected after an
iterative parameter calibration using the results of tensile tests with notched cylindrical samples.
INDEX çç Sommaire
The validation of the experimentally derived damage curves is carried out by GTN simulations with
unit cells [19]. The model in figure 7 consists of a quadratic axis-symmetric cell with 8 nodes and
reduced integration. For modeling of ductile damage, a GTN subroutine is implemented in the
simulation. The stress triaxiality h is kept constant at h = 0.5, 1, 1.5 and 2 due to a constant ratio of
loading F1/ F2 during the whole calculation. In a first step values of fN and f0 are passed together with
all other parameters to the solver. As shown in figure 8, simulation results are nearly congruent to the
experimental damage curve. Consequently, the GTN simulations can be used to predict the course of
the damage curve for the given material.
1.5
6 7 8 GTN simulations
1.0
1 2 3 0.0
1.0 1.5 0.0
2.0 2.5 0.5
stress triaxiality h, -
Figure 7: Cell model for GTN Figure 8: Validation of experimental damage curve.
calculation.
As a further step, Charpy tests with V-notch samples were simulated using the GTN model with the
presented set of model parameters using Abaqus/Explicit. For reasons of simplicity, the adiabatic
loading conditions were not considered. Instead, it was assumed that the effects of strain rate and
temperature on the flow curve approximately balance each other. For the simulations, the hammer
mass of 19.8 kg and the striker´s initial velocity of 5.5m/s were considered.
Figure 9 shows the calculated load-deflection plots together with the corresponding experimental
results. Obviously, the results sufficiently agree with each other. An impact energy of 223 J was
measured, while the GTN simulations resulted in a value of 225 J.
30 1.5
equivalent plastic strain epl,-
20 1.0
load, kN
simulation results
experimental damage curve
15
10 0.5
5
0 0.0
0 10 20 30 0.0 0.5 1.0 1.5 2.0 2.5
deflection, mm stress triaxiality h, -
Figure 9: Load deflection plots of Charpy tests at Figure 10: Experimentally derived damage curve
room temperature from experiment and GTN and calculated general lower bound damage curve
simulation. for steel P500Q.
INDEX çç Sommaire
Again, it can be concluded that the GTN model is capable for the quantitative description of micro
structural influence on upper shelf toughness properties. Consequently, the model can also be used in
an inverse way. Then, the investigations aim to calculate as a theoretical value the amount of void
nucleating micro structural constituents to achieve a certain low level of upper shelf impact toughness.
It was assumed that KVmax = 70 J is a good approximation for the lower borderline of Charpy values of
P500Q steel produced in Europe. Thus, the GTN simulations were repeated with varying parameter
sets until finally, an upper shelf impact toughness of 68 J resulted. In these simulations, the following
parameter set was identified:
This parameter set was understood to be representative for the “worst case” micro structure
configuration. Using this parameter set in the cell model simulations, the lower borderline of damage
curve was derived numerically afterwards, as presented in figure 10.
To demonstrate the safe capacities of the investigated material (but so far permitted to utilise), a model
pressure vessel was designed according to EN 13345 – part 3. The vessel consists of a cylindrical shell
with an outer diameter of 3.000 mm and spherical bottoms at both ends. The wall thickness for
analysis was chosen to 50 mm with respect to the prior investigated material P500Q. To keep the
subsequent design models simple only a single nozzle of an inner diameter of 500 mm was placed in
the cylindrical shell by a butt weld. The plate thickness of this nozzle was restricted from the
consideration that the design pressure of the nozzle is equal to the shell. The so derived thickness of deb
= 72.5 mm is based upon a design check by formula. Further the gross plastic design check for the
overall vessel was performed using the concepts of EN 13445 – Part 3 (DBF, DBA - Annex C and
DBA – Annex B).
Using the DBF concept taking into account a nominal yield strength of 500 MPa and a lower bound
limit of the ultimate tensile strength of 590 MPa the nominal design stress was calculated to f = 245.8
MPa, which results in a design pressure of 8.3 MPa for both shell and nozzle.
For the design check according to Annex C using the DBA concept based upon stress categorisation a
numerical model with shell elements and elastic material properties was generated. The value for the
design pressure was determined to 8.1 MPa which is even lower than the value derived by DBF.
In a further step a numerical model was created using solid elements and linear elastic – perfect plastic
material behaviour to perform the design check DBA according to Annex B. For P500Q the partial
safety factor on the resistance side R equals 1.32 so that the nominal yield strength is calculated to
Rm,d = 327 MPa when using the Mises yield condition. By increasing the internal pressure the
numerical calculation is driven until the equivalent plastic strain reaches a limit of 5% at the critical
point of the vessel for normal operating load cases. At this limit the model of the selected pressure
vessel bears a pressure of 11.8 MPa. Applying the partial safety factor on the action side (P = 1.2)
results in the design pressure of 9.9 MPa.
To compare the pressures derived by the design checks according to the standard with the realistic
behaviour of a pressure vessel in terms of the maximum pressure and the point of ductile crack
initiation the numerical model used for DBA – Annex B was modified as follows:
INDEX çç Sommaire
During the calculation geometric nonlinearities are taken into account.
Realistic material behaviour in terms of the flow curve derived from tensile tests was
implemented.
The “worst case” damage curve was used as ductile failure criterion to assess the point of crack
initiation.
The modified RIKS method was applied to allow a calculation beyond the maximum value of the
pressure-displacement curve.
As shown in figure 12 the elastic range of the vessel is limited by a pressure of 22.6 MPa which is
more than twice as much as estimated by DBA. The maximum pressure reaches a value of 25.8 MPa.
The development of the equivalent plastic strain at this point is shown in figure 12. The load path
derived from the values of stress triaxiality and equivalent plastic strain for the relevant location is
given in figure 13. By comparison with the “worst case” damage curve the point of crack initiation was
estimated at a stress triaxiality of 0.81 and an equivalent plastic strain of 41.0%.
30 1,0 1,0
25,8 MPa
pressure
25
0,8 0,8
equivalent plastic strain, epl, -
equivalent plastic strain epl, -
point of
crack initiation
20
pressure p, MPa
0,6 0,6
15 point of crack initiation
Figure 12: Pressure and equivalent plastic strain vs. Figure 13: Load path at relevant location vs.
displacement “worst case” damage curve
INDEX çç Sommaire
As it can be seen from figure 12 for this pressure vessel and the investigated steel P500Q the point of
crack initiation and the maximum pressure coincidence. From these investigations the exaggerated
actual safety margins can be seen. Proposals for establishing realistic safety margins are under way.
The European pressure vessel standard EN 13445 belongs to modern design codes for unfired
pressure vessels. However, the application of high strength steels to pressure vessels is still
hindered by this design rule although these materials offer excellent strength and toughness
properties. However, the application of such modern materials is useful because they can
significantly contribute to lightweight and resource-efficient design yielding to a better economy.
Especially as European steel producers have concentrated on the development of quenched and
tempered structural steel grades, the disadvantageous treatment of these materials should be
avoided by the European design standard.
The DBA design approach already allows to use a limit strain concept – which gives the
opportunity to employ empirical damage mechanics models relying to a critical plastic strain to
ductile crack initiation. For the future it is planned to suggest individual limit strains and safety
factors for the different pressure vessel steel grades. Like this, the relation between strength and
ductility can be taken into account, and the limit strains and safety factors shall be defined with
respect to the expected minimum upper shelf toughness properties of the different steel grades.
It was possible to predict the course of the damage curve for h > 0.4 using cell model simulations.
In these simulations, both the initial void volume fraction f0 and the volume fraction of newly
nucleated voids fN were derived in metallographic simulations.
Using the same set of GTN parameters, it was possible to predict the upper shelf impact toughness
of steel P500Q by simulation of Charpy test. Like this, the micro structural influence on toughness
properties could be expressed.
In an inverse approach, the GTN parameters for steel P500Q were determined which result in an
expected minimum upper shelf toughness of 70 J. These GTN parameters were afterwards used to
predict the general lower bound damage curve of steel P500Q.
Using this “worst case” damage curve a pressure vessel designed according to the EN 13445 was
investigated in view of maximum pressure and the point of ductile failure. For the investigated
combination of pressure vessel and steel the point of failure in terms of a ductile crack initiation
was located near the point of maximum pressure.
From the comparison of these pressures with the design pressures derived by DBF or DBA it
could be seen that the actual safety margins are unrealistic, i.e. far too large and even not a half of
the elastic range of the pressure vessel is utilised.
ACKNOWLEDGEMENT
The authors gratefully acknowledge the financial support of the research project which was granted
from FOSTA, Forschungsvereinigung Stahlanwendung, Sohnstraße 65, D - 40237 Düsseldorf.
REFERENCES
[1] EN 13445: Unfired pressure vessels – Part 3: Design. German version EN 13445-3:2009, only
on CD-ROM.
[2] FOSTA, Forschungsvereinigung Stahlanwendung, Sohnstraße 65, D - 40237 Düsseldorf
INDEX çç Sommaire
[3] Gurson AL. Plastic Flow and Fracture Behavior of Ductile Materials Incorporating Void
Nucleation, Growth and Interaction. PhD Thesis, Brown University; 1975.
[4] Gurson AL. Continuum theory of ductile rupture by void nucleation and growth. 1. Yield criteria
and flow rules for porous ductile media. J Eng Mater Technol-Trans ASME. 1977;99(1):2-15.
[5] Tvergaard V. Influence of voids on shear band instabilities under plane-strain conditions. Int J
Fract. 1981;17(4):389-407.
[6] Tvergaard V. On localization in ductile materials containing spherical voids. Int J Fract.
1982;18(4):237-252.
[7] Tvergaard V, Needleman A. Analysis of the cup-cone fracture in a round tensile bar. Acta
Metallurgica. 1984;32(1):157-169.
[8] Needleman A. Material rate dependence and mesh sensitivity in localization problems. Comput
Meth Appl Mech Eng. 1988;67(1):69-85.
[9] Needleman A, Tvergaard V. Mesh effects in the analysis of dynamic ductile crack-growth. Eng
Fract Mech. 1994;47(1):75-91.
[10] Skallerud B, Zhang ZL. Effects of finite element mesh on numerical prediction of ductile tearing.
In: Panontin TL, Sheppard SD, editors. 29th National Symposium on Fatigue and Fracture
Mechanics. Stanford, Ca: American Society Testing and Materials; 1997: 201-214.
[11] Arndt J, Majedi H, Dahl W. Influence of strain history on ductile failure of steel. EUROMECH -
MECAMAT 96 / 1st European Mechanics of Materials Conference on Local Approach to
Fracture 86-96. Fontainebleau, France: Editions Physique; 1996.: 23-32.
[12] Schluter N, Grimpe F, Bleck W, Dahl W. Modelling of the damage in ductile steels. 5th
International Workshop on Computational Modelling of the Mechanical Behavior of Materials.
Aachen, Germany: Elsevier Science Bv; 1995:. 27-33.
[13] Schiffmann R, Bleck W, Dahl W. The influence of strain history on ductile failure of steel.
Comput Mater Sci. 1998;13(1-3):142-147.
[14] Norris DM, Reaugh JE, Moran B, Quinones DF. A plastic-strain, mean-stress criterion for
ductile fracture. J Eng Mater Technol-Trans ASME. 1978;100(3):279-286.
[15] Oyane M, Sato T, Okimoto K, Shima S. Criteria for ductile fracture and their applications.
Journal of Mechanical Working Technology. 1980;4(1):65-81.
[16] Lemaitre J. A continuous damage mechanics model for ductile fracture. J Eng Mater Technol-
Trans ASME. 1985;107(1):83-89.
[17] Chu CC, Needleman A. Void nucleation effects in biaxially stretched sheets. J Eng Mater
Technol-Trans ASME. 1980;102(3):249-256.
[18] Steglich, D., Bestimmung von mikrostrukturellen Parametern in Schädigungsmodellen für
duktile Metalle; PhD Thesis, TU Berlin; 1999
[19] Weber, U., Mohanta, A., Schmauder, S., Numerical determination of parameterised failure
curves for ductile structural materials, Int. J. Mat. Res.,Vol. 98, 2007:1071-1080
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page78
14C Effet de l’énergie de chauffage et du revenu sur la ténacité des GAMBHAVA D. / LEWIS A. /
joints soudés en CrMoV / Effect of Heat Input & Tempering PATEL R. / GHOSH M.
parameter on NotchToughness Property of CrMoV weldments LARSEN & TOUBRO Ltd
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page79
98C Etude de soudabilité de joints en matériaux différents utilisés dans MORAN D. / PINTOS V.
la fabrication de réchauffeurs de production d’énergie / Weldability LOPEZ A. / SIEIRO F.
study of dissimilar metal joints used in power generation reheaters /GALVEZ J.
manufactory AIMEN TECHNOLOGY
CENTRE
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page80
RÉSUMÉ / ABSTRACT
14C
EFFET DE L’ÉNERGIE DE CHAUFFAGE ET DU REVENU SUR LA
TÉNACITÉ DES JOINTS SOUDÉS EN CR-MO-V
RÉSUMÉ
2.25Cr -1Mo-0.25V aciers 2.25Cr ont gagné en popularité au cours des aciers Cr-Mo classiques dans les deux dernières décennies en raison de ses
excellentes propriétés à haute température en termes de fluage et attaque de l’hydrogène. Cependant, l’ajout de vanadium rend le matériel plus
Mercredi
critique en termes de soudabilité et de traitement thermique. Cela conduit à mettre davantage l’accent sur le choix de paramètres de soudage
appropriés ainsi que soulager le stress des traitements thermiques. Ce document vise à expliquer l’effet de tempérer les comportements en matière
de propriété résilience des soudures avec différents apports de chaleur durant le post traitement thermique de soudure (PWHT).
Pour-1Mo-0.25V acier 2.25Cr, ASME VIII Div sec 1 et 2 indique la température minimale PWHT 677 ° C. Ces soudures sont normalement stress
soulagé à la température cible de 705 ° C et trempage / loding temps ho est de 8 - 10 h, indépendamment de l’épaisseur. Ceci a été sur base de
l’expérience des fournisseurs de consommables et les fabricants. Même si le code de la construction permet de baisser la température et faible
temps de PWHT, le temps de trempage a augmenté et la température est nécessaire pour tempérer suffisamment la soudure de haute dureté.
Dans ce travail,-1Mo-0.25V plaques 2.25Cr ont été soudés avec deux entrées différentes de la chaleur et soumis à des conditions différentes PWHT.
Réduire le déficit de soudage à l’arc submergé (SAW NG) a été utilisé avec 2 passes par couche technique, pour obtenir au maximum / raffinée
grain zone tempérée dans la soudure. Ces soudures ont été soumis à des tests mécaniques comme l’impact et les essais de dureté ainsi que l’analyse
de microstructures.
Résultats de résilience ainsi que des données la dureté a été tracée en fonction de Larson Miller paramètre pour évaluer l’effet de la trempe pendant
PWHT. Fractographie par SEM et microstru comparaison cture optique a également été effectuée sur l’échantillon impact cassé. Résultat montre
que la résilience de la soudure augmente avec l’augmentation des Larson Miller paramètres (augmentation de la durée PWHT fois et le temps).
Dureté montre une légère diminution avec une augmentation de Larson Miller paramètre.
ABSTRACT
2.25Cr-1Mo-0.25V steels have gained popularity over conventional Cr-Mo steels in the last two decades due to its superior high temperature properties
in terms of creep & hydrogen attack. However the addition of vanadium makes the material more critical in terms of weldability & heat treatment.
This leads to increased emphasis on choosing appropriate welding parameters as well as stress relieving heat treatments. This paper aims to explain
the effect of tempering behavior on notch toughness property of welds with varying heat inputs during post weld heat treatment (PWHT).
For 2.25Cr-1Mo-0.25V steel, ASME sec VIII Div 1 & 2 states the minimum PWHT temperature as 677°C. These welds are normally stress relieved
at targeted temperature of 705°C and Soaking/ ho loding time is 8- 10 h irrespective of the thickness. This has been derived on basis of experience
of consumable suppliers and fabricators. Even though construction code permits lower temperature and lower time for PWHT, the increased
soaking time and temperature is required to sufficiently temper the weld of high hardness.
In this work, 2.25Cr-1Mo-0.25V plates were welded with two different heat inputs and subjected to different PWHT conditions. Narrow Gap
Submerged Arc Welding (NG SAW) was used with 2 pass per layer technique, to get maximum tempered/grain refined zone in the weld. These welds
were subjected to mechanical tests like impact and hardness tests as well as microstructural analysis.
Impact toughness results as well as Hardness data was plotted as function of Larson Miller parameter to evaluate the effect of tempering
during PWHT. Fractography by SEM and Optical microstructure comparison was also carried out on the broken impact specimen.
Result shows that notch toughness of the weld increases with increase in Larson Miller Parameter (i.e. increase in both PWHT duration
and time). Hardness shows a marginal decrease with increase in Larson Miller Parameter.
KEY WORDS: 2.25Cr-1Mo-0.25V steels, Cr-Mo, Larson Miller Parameter, Narrow Groove, post weld heat treatment, SAW process,
Fractrography
INDEX çç Sommaire
Effect of Heat Input & Tempering parameter on Notch-Toughness
Property of Cr-Mo-V weldments
a b c d
Darshan Gambhava , Allwyn Lewis , Ritesh Patel , Manas Ghosh
a ,b ,c d
Larsen & Toubro Ltd, Heavy Engineering Division, Hazira
[email protected], [email protected], [email protected], [email protected]
Résumé
2.25Cr -1Mo-0.25V aciers 2.25Cr ont gagné en popularité au cours des aciers Cr-Mo
classiques dans les deux dernières décennies en raison de ses excellentes propriétés à haute
température en termes de fluage et attaque de l'hydrogène. Cependant, l'ajout de vanadium
rend le matériel plus critique en termes de soudabilité et de traitement thermique. Cela
conduit à mettre davantage l'accent sur le choix de paramètres de soudage appropriés ainsi
que soulager le stress des traitements thermiques. Ce document vise à expliquer l'effet de
tempérer les comportements en matière de propriété résilience des soudures avec différents
apports de chaleur durant le post traitement thermique de soudure (PWHT).
Pour-1Mo-0.25V acier 2.25Cr, ASME VIII Div sec 1 et 2 indique la température
minimale PWHT 677 ° C. Ces soudures sont normalement stress soulagé à la température
cible de 705 ° C et trempage / loding temps ho est de 8 - 10 h, indépendamment de
l'épaisseur. Ceci a été sur base de l'expérience des fournisseurs de consommables et les
fabricants. Même si le code de la construction permet de baisser la température et faible
temps de PWHT, le temps de trempage a augmenté et la température est nécessaire pour
tempérer suffisamment la soudure de haute dureté.
Dans ce travail,-1Mo-0.25V plaques 2.25Cr ont été soudés avec deux entrées
différentes de la chaleur et soumis à des conditions différentes PWHT. Réduire le déficit de
soudage à l'arc submergé (SAW NG) a été utilisé avec 2 passes par couche technique, pour
obtenir au maximum / raffinée grain zone tempérée dans la soudure. Ces soudures ont été
soumis à des tests mécaniques comme l'impact et les essais de dureté ainsi que l'analyse de
microstructures.
Résultats de résilience ainsi que des données la dureté a été tracée en fonction de
Larson Miller paramètre pour évaluer l'effet de la trempe pendant PWHT. Fractographie par
SEM et microstru comparaison cture optique a également été effectuée sur l'échantillon
impact cassé. Résultat montre que la résilience de la soudure augmente avec l'augmentation
des Larson Miller paramètres (augmentation de la durée PWHT fois et le temps). Dureté
montre une légère diminution avec une augmentation de Larson Miller paramètre.
Abstract
2.25Cr-1Mo-0.25V steels have gained popularity over conventional Cr-Mo steels in
the last two decades due to its superior high temperature properties in terms of creep &
hydrogen attack. However the addition of vanadium makes the material more critical in terms
of weldability & heat treatment. This leads to increased emphasis on choosing appropriate
welding parameters as well as stress relieving heat treatments. This paper aims to explain the
effect of tempering behavior on notch toughness property of welds with varying heat inputs
during post weld heat treatment (PWHT).
INDEX çç Sommaire
For 2.25Cr-1Mo-0.25V steel, ASME sec VIII Div 1 & 2 states the minimum PWHT
temperature as 677°C. These welds are normally stress relieved at targeted temperature of
705°C and Soaking/ ho loding time is 8- 10 h irrespective of the thickness. This has been
derived on basis of experience of consumable suppliers and fabricators. Even though
construction code permits lower temperature and lower time for PWHT, the increased
soaking time and temperature is required to sufficiently temper the weld of high hardness.
In this work, 2.25Cr-1Mo-0.25V plates were welded with two different heat inputs
and subjected to different PWHT conditions. Narrow Gap Submerged Arc Welding (NG
SAW) was used with 2 pass per layer technique, to get maximum tempered/grain refined
zone in the weld. These welds were subjected to mechanical tests like impact and hardness
tests as well as microstructural analysis.
Impact toughness results as well as Hardness data was plotted as function of Larson
Miller parameter to evaluate the effect of tempering during PWHT. Fractography by SEM
and Optical microstructure comparison was also carried out on the broken impact specimen.
Result shows that notch toughness of the weld increases with increase in Larson Miller
Parameter (i.e. increase in both PWHT duration and time). Hardness shows a marginal
decrease with increase in Larson Miller Parameter.
KEY WORDS: 2.25Cr-1Mo-0.25V steels, Cr-Mo, Larson Miller Parameter, Narrow Groove, post
weld heat treatment, SAW process, Fractrography
Introduction
2.25Cr-1Mo- 0.25V steel or modified 2.25Cr-1Mo steel has crucial role in processes
in refineries and petrochemical plant such as desulphurization of hydrocarbons and cracking
of heavier hydrocarbon fractions in to lighter molecules. These processes are carried out at
high temperature and pressure in presence of catalyst. This calls for increasing steel strength
which reduces in construction cost. However all the benefits of the steel are exploited only if
the properties of the weldments match with the properties of the conventional steel
only.welding of this steel required precaution and careful processing and Welds are always
post weld heat treated before service.[1]
For 2.25Cr-1Mo-0.25V steel, ASME sec VIII Division 1 & 2 states the minimum
PWHT temperature as 677°C. These welds are normally stress relieved at targetted
temperature of 705°C and Soaking/ holding time is 8- 10 Hrs. irrespective of thickness.This
has been derived on basis of consumable supplier and fabricator’s experience. Even though
construction code calls for a lower temperature and lower time PWHT, the increased soaking
time and temperature is required to sufficiently temper the high hardness weld.
INDEX çç Sommaire
Experimental
The studies were performed on experimental submerged arc welds deposited on 144mm thick
2.25Cr-1Mo-0.25V steel plate. The composition of the weld is listed Table 1. J factor for this
plate is material is 33. J factor is defined as J= (P + Sn) (Mn + Si) x 104. It relates temper
embrittlement susceptibility of the base material. It should be lass than 100. In this case it is
very low that is low susceptibility of temper embrittlement. Two pass per layer welding
technique is used for experimental. Schematic diagram of bead placement for two pass per
layer technique is shown in figure 1. Joint configuration for 144 mm thick welds is generally
double “J” type having J-radius of 12mm and included angle of 1 deg. For experiments joint
configuration has been modified to the J-radius of 10mm and included and is 1 deg. Joint
configuration is shown in the figure 2. This modification has been done to accommodate the
two pass per layer technique with chances of less defects.
Elements C S Mn P Si Cu Mo Ni Cr V Ti As Sb Sn
Wt % 0.14 0.001 0.53 0.004 0.049 0.12 1.00 0.12 2.23 0.3 0.001 0.003 0.002 0.0017
Elements C S Mn P Si Cu Mo Ni Cr V Ti As Sb Sn
Wt % 0.072 0.003 1.37 0.005 0.081 0.57 1.01 0.33 2.24 0.25 0.001 0.002 0.002 0.0023
INDEX çç Sommaire
Table 3: Welding condition and heat input
HITP I HITP II
Base material 2.25Cr –1Mo –0.25V 2.25Cr –1Mo –0.25V
Consumable(s) EG Filler wire + flux EG Filler wire + flux
Process SAW SAW
Position 1G 1G
Polarity AC AC
Heat input (KJ/mm) 1.81 2.04
To find out the effect of notch toughness on tempering parameter, plate welded by both heat
input under go different tempering parameter that is different PWHT. We have used Larson
miller parameter as tempering parameter. Larson miller parameter can be calculated by
equation [5][6]
LMP = T X (20 + log t )
Where T = PWHT temperature in oK
t = PWHT time in Hr.
Different Larson miller parameter and corresponding PWHT condition is shown in the table 4
Testing
Chemical testing of base and weld metal was done to find out the chemical composition of
the base metal and weld to find out the J factor of base metal and X- factor for the weld
metal.
Charpy V – Notch testing was carried out for the specimen taken from Weld and HAZ of
HITP I and HITP II in as welded condition and in varying Larson miler parameter condition.
Samples were taken out from the 1.5mm from the surface and at the half of the thickness
(T/2) of the weld of each heat input. Total 9 samples were taken from each location. Testing
was carried out at -30oC. Acceptance value for the testing is 48 J (Avg.) and 54 J (Min) in a
set of 3 specimens. [3] To study the effect on notch toughness of LMP Impact Vs LMP graph
has been prepared for both heat input.
Hardness testing was carried out as per ASTM E 92 by Vickers hardness tester with 10 kg
load. At each condition of the heat input and LMP, Six values of hardness was taken on three
zone of the weldment that is Weld, HAZ and base Metal. To study the effect of LMP,
Hardness Vs LMP graph has been prepared.
INDEX çç Sommaire
Macro of the weld has been taken to conform, the two pass per layer technique through
thickness. The microstructures were characterized by using optical and scanning electron
microscope. Grain size measurement was also done to find out the difference with respect to
both heat input
Result
Result of chemical testing of base metal and weld metal are shown in the table 1 and table 2.
The J factor for base metal is 33 as per watanabe formula, which is quite low for base metal.
X- Factor for weld metal as per Bruscato formula is 7.12 ppm which is also quite low for
good weld metal properties
695-705°C
for 6 Hr 123 180 166 0 128 196 164 0
(LMP 20217)
695-705°C
for 8 Hr 142 283 201 302 131 243 162 309
(LMP 20338)
695-705°C
for 10Hr 165 252 193 299 167 220 202 267
(LMP 20433)
705-715°C
for 8Hr 189 273 239 296 209 282 225 331
(LMP 20548)
705-715°C
for 10Hr 186 290 215 267 181 271 227 308
(20643)
The average value of notch toughness of all nine samples is shown in table 5.It shows
impact values of both heat input at each LMP. Testing was carried out of nine samples to
minimize the error in results
The average value of hardness is shown in the table 6. These values are in HV 10 of both heat
input at each LMP.
INDEX çç Sommaire
Table 6 : Hardness of HITP I and HITP II
HARDNESS HV10
HITP I HITP II
HEAT
TREATMENT
WELD HAZ BM WELD HAZ BM
695-705°C
for 6 Hr 227 248 222 229 243 221
(LMP 20217)
695-705°C
for 8 Hr 211 216 213 213 223 216
(LMP 20338)
695-705°C
for 10Hr 204 220 205 205 218 204
(LMP 20433)
705-715°C
for 8Hr 211 221 210 205 228 196
(LMP 20548)
705-715°C
for 10Hr 207 212 210 213 216 210
(20643)
Figure 3: Weld joint configuration of modified ‘J’
Macro of the welding is shown in the figure 3.it clearly shows that welding of two pass per
layer with very thin HAZ. The height of bead is around 2.5mm so that previous pass can be
sufficiently tempered. And due to that good impact property can be achieved.
INDEX çç Sommaire
400X 1000X
400X 1000X
INDEX çç Sommaire
Discussion
Figure 6 reveals similar trend in the impact toughness values in the weld at the two locations,
T/2 always exceeding 1.5mm. Also there is very marginal change in the impact values for the
two heat inputs. The values are almost similar with HITP I showing slightly higher values as
compared to HITP II with a few exceptions.
Figure 6 shows the comparison of the impact toughness values for different Larson-Miller
parameter. The figure clearly reveals that the impact toughness values increases with
increasing LMP. An increase of impact toughness values in the order of 25-30 J is observed
for an increase in LMP of approx. 100 unit. Increasing the LMP from 20548 to 20643 shows
a very slight or negligible decrease in the impact toughness values for both the heat inputs as
well as for both the locations.
Conclusion
Changing the heat input from 1.81 to 2.04 kJ/mm does not significantly affect the impact
toughness property. However, it influences the grain size & produces a fine grain structure in
the lower heat input sample. ASTM Grain size No. for 1.81 kJ/mm heat input is 7-9 whereas
for 2.04 kJ/mm is 5-6.
Impact toughness values observed at T/2 are substantially higher than that observed at 1.5mm
from top for both the heat inputs due to insufficient tempering of the top layer as compared to
mid thickness.
Impact toughness values increases with increasing LMP. An increase of impact toughness
values in the order of 25-30 J is observed for an increase in LMP of approx. 100 units.
Increasing the LMP from 20548 to 20643 shows a very slight or negligible decrease in the
impact toughness values for both the heat inputs as well as for both the locations.
Impact toughness requirements have been met for LMP of 20217 (695-705°C for 6 Hrs),
however, HAZ hardness is beyond the acceptable limit. Therefore, LMP of 20338 (695-
705°C for 8 Hrs) is essentially required to meet overall impact-hardness requirements.
References:
1. Advanced vanadium modified steels for high pressure hydrogen reactors, Joanna
Hucinska, Gda sk University of Technology, Faculty of Mechanical Engineering,
Poland
2. American petroleum institute (API) TECHNICAL REPORT 934-B First Edition,
Fabrication Consideration for Vanadium Modified Cr- Mo Steel Heavy Wall
Pressure Vessels.
3. Boiler and pressure vessel code ASME Section VIII Division 1 (UG 84)
4. Dr Siegfired Dittrich, Thyssen Draht AG, Hamm, West Germany 21/4Cr-1Mo Filler
Wire Metal With High Toughness Properties For The Construction Of Hydrocrackers
5. Welding Research Council (WRC) - 275 (Publication - 1971)
INDEX çç Sommaire
6. Materials and Fabrication of 2 1/4Cr-1Mo, 2 1/4Cr-1Mo-1/4V, 3Cr-1Mo, and 3Cr-
1Mo-1/4V Steel Heavy Wall Pressure Vessels for High-temperature, High-pressure
Hydrogen Service, Downstream Segment,API recommended Practice 934-A second
edition, May 2008
7. New Generation of Cr-Mo-V steel Grades for the use in Petrochemical Reactors
8. by Dr. I. Detemple, Dr. F. Hanus, Dr. G. Luxenburger, R.J. Cawelius,Dillinger
Hüttenwerke.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page81
RÉSUMÉ / ABSTRACT
20C
RÉSUMÉ
Pour assurer la sécurité, la fiabilité et la rentabilité des produits soudés, il faut que les soudures soient
d’une qualité acceptable. Le soudage est un processus particulier qui nécessite une maîtrise particulière
pour atteindre la qualité requise. Si on ne maîtrise pas le soudage d’une manière correcte, on risque
des ruptures en service catastrophiques et de nombreuses victimes. De plus, les entreprises qui n’ont
pas mis en place la maîtrise du soudage éprouveront les conséquences négatives dues à la nécessité de
faire des réparations : par exemple augmentation des coûts, retards et problèmes contractuels.
La maîtrise du soudage, comme processus particulier, est le sujet de l’ISO 3834 – exigences de qualité
en soudage par fusion. A son tour, cette norme a des exigences spécifiques pour la compétence de tout
personnel qui a les responsabilités en soudage ou dans les domaines associés. Il est évident que tout le
personnel qui peut influencer la qualité des soudures dans la fabrication doit avoir la compétence
appropriée.
ABSTRACT
The safety, reliability and cost effectiveness of welded products requires the welds to be of adequate
quality. Welding is a special process which needs special controls in order to achieve the required
quality. Failing to control welding correctly can result, in the worst case, in catastrophic failure and
loss of life. Furthermore, companies that do not adequately control welding, experience debilitating
factors arising from having to make repairs: cost overruns, delays, contractual problems, etc.
Controlling welding as a special process is dealt with in ISO 3834, quality requirements for fusion
welding. In turn, this standard has specific requirements about the competence of all people with
welding, or welding-related, responsibilities. It is clear that everyone who can have an impact on the
final quality of the welds in the product must have appropriate competence.
The paper tracks the growth in recognition with regard to the importance of competence in welding
and indicates the current recommended best-practice approach for companies to deal with it.
INDEX çç Sommaire
THE SPECIAL IMPORTANCE OF PERSONNEL COMPETENCE IN
WELDING FABRICATION
ABSTRACT
The safety, reliability and cost effectiveness of welded products requires the welds to be of
adequate quality. Welding is a special process which needs special controls in order to achieve
the required quality. Failing to control welding correctly can result, in the worst case, in
catastrophic failure and loss of life. Furthermore, companies that do not adequately control
welding, experience debilitating factors arising from having to make repairs: cost overruns,
delays, contractual problems, etc.
Controlling welding as a special process is dealt with in ISO 3834, quality requirements for
fusion welding. In turn, this standard has specific requirements about the competence of all
people with welding, or welding-related, responsibilities. It is clear that everyone who can have
an impact on the final quality of the welds in the product must have appropriate competence.
The paper tracks the growth in recognition with regard to the importance of competence in
welding and indicates the current recommended best-practice approach for companies to deal
with it.
RÉSUMÉ
Pour assurer la sécurité, la fiabilité et la rentabilité des produits soudés, il faut que les soudures
soient d’une qualité acceptable. Le soudage est un processus particulier qui nécessite une
maîtrise particulière pour atteindre la qualité requise. Si on ne maîtrise pas le soudage d’une
manière correcte, on risque des ruptures en service catastrophiques et de nombreuses victimes.
De plus, les entreprises qui n’ont pas mis en place la maîtrise du soudage éprouveront les
conséquences négatives dues à la nécessité de faire des réparations : par exemple augmentation
des coûts, retards et problèmes contractuels.
INDEX çç Sommaire
La maîtrise du soudage, comme processus particulier, est le sujet de l’ISO 3834 – exigences de
qualité en soudage par fusion. A son tour, cette norme a des exigences spécifiques pour la
compétence de tout personnel qui a les responsabilités en soudage ou dans les domaines
associés. Il est évident que tout le personnel qui peut influencer la qualité des soudures dans la
fabrication doit avoir la compétence appropriée.
BACKGROUND
The safety, reliability and cost effectiveness of welded products requires the welds to be of
adequate quality. Welding is a special process which needs unique controls in order to achieve
the required quality. Such processes are recognised in the quality management systems standard,
ISO 9001 [1], which requires them to be treated in a particular way. Ignoring the need for
special, competent control of welding can result, in the worst cases, in catastrophic failure and
loss of life. Whilst such occurrences are rare these days, they are not unknown. More commonly
companies that do not adequately ensure competent control of welding may find themselves
suffering from cost overruns, delays, contractual problems, etc, arising from the need to make
repairs to bad welds. The cost penalties can be significant: it has been established that repairing a
faulty weld incurs more than five to six times the cost of doing it correctly the first time.
The importance of competence in welding was recognised a long time ago. As an example, on 28
November 1918, the Technical Committee of Lloyd’s Register of Shipping reviewed a report [2]
on the application of arc welding to shipbuilding. This report was generally positive about the
introduction of the process and proposals regarding additional regulations were made. These
included the following:
• The Committee must be satisfied that the operators engaged are specially trained, and
are experienced and efficient in the use of the welding system proposed to be employed.
These proposals were adopted and appeared in the 1920 issue of Lloyd’s Rules.
Therefore, when we talk about competence in welding today, we are not talking about anything
new! Our predecessors’ opinion on the matter is evident from the above quotations; although it is
only by viewing these statements through the prism of history that we can see how prescient they
were.
In the years that followed, specifications and standards on welder approval (or ‘qualification’ as
it is now called) emerged from a number of countries. In the UK, a new professional body, the
Institute of Welding (now part of The Welding institute – TWI) was formed in 1923.
INDEX çç Sommaire
Later on, in the 1960s and beyond, specifications dealing with the competence of non-destructive
testing (NDT) and welding inspection personnel emerged.
Nowadays, controlling welding as a special process is dealt with in ISO 3834 [3], quality
requirements for fusion welding (which was first published as ISO 3834, and as EN 729, in
1995). This standard sets specific requirements for the competence of people with welding
responsibilities and this includes: welders, NDT personnel, welding inspectors and welding
coordinators. In other words, it is clear that anyone and everyone who can have an impact on the
final quality of the welds in the product must have the required competence. It is like a chain - a
‘welding quality chain’ - if only one link is missing or faulty, the chain will break and the kinds
of problems highlighted above may occur. This is recognised in ISO 14731 [4], welding
coordination – tasks and responsibilities (first published in1994 as EN 719) which is referred to
in ISO 3834.
The correct implementation of these two standards, ISO 3834 and ISO 14731, is the key to the
cost-effective manufacture of safe and reliable welded products.
DEFINITIONS
The subject of competence is emotive because it is a personal attribute. In order to analyse and
deal with the subject objectively, it is important in the first instance to be clear about the
definition of words and phrases that are used in the context of competence. In the English
language, this is not a straightforward matter and therefore considerable difficulties arise when
attempting to transfer competence concepts into different languages. However, in order to make
a start, the following definitions are offered, although they may not be the true ‘Oxford
Dictionary’ definitions:
Competence - the state of being competent; that is, being adequately capable of performing
certain defined tasks.
Training - the process of imparting knowledge and/or a set of skills which is narrowly defined
and usually employment related. The emphasis is generally on the attainment of practical skills
together with relevant underpinning knowledge.
It should be noted that a person who undergoes a course of education or training does not
automatically acquire the knowledge and/or skill that the course was designed to impart.
Therefore it is common to have some kind of test at the end of a course.
Qualification - confirmation that a person has been successful in a test (usually theoretical) to
confirm the attainment of knowledge. A university degree is an example of an academic or
INDEX çç Sommaire
professional qualification. Qualifications may help in determining the competence of a person
but they do not normally, in themselves, give adequate assurance of competence.
Experience - working in a job in such a way as to make use of any education, training and
qualifications gained. The type of experience can be determined from the person’s job
specification and of particular importance for people with professional qualifications is the level
and type of decision-making authority that the person has been given.
Certification - a testimony that a person has proved an ability to carry out certain defined tasks
and has the underpinning knowledge necessary to perform those tasks. Certification usually
requires the prior achievement of specific qualifications, training and experience; and practical
testing that is in addition to the qualification testing mentioned above. In cases where the scope
of certification matches the requirements of the job to be carried out, certification is the best way
of demonstrating competence.
Authority to work - the authority to carry out a job of work can only be granted by the employer
according to the employment law of most countries. This implies that the employer is also
responsible for the competence of the person to do the job that he/she has been engaged to do.
Employers are encouraged to use personnel with relevant certification, where available and
appropriate, in order to help with this.
As mentioned above the special importance of competence in welding has been recognised for
many years. Since the 1980s a great deal of effort has gone in to the development of an
internationally harmonised education, training, qualification and certification system [5]. The
system is now maturing, and is implemented and recognised widely throughout the world. Over
150,000 diplomas have been awarded. Some of these harmonised qualifications are also
mentioned in ISO 3834 and ISO 14731.
The system is a valuable tool for individuals seeking wide recognition of their knowledge and
skills and for employers seeking to employ competent people. Qualification and certification that
are widely recognised inspire confidence in holders, in employers and in clients who purchase
welded products. However, the system continues to be misunderstood in some quarters and
therefore it may be helpful to remind ourselves what these qualifications mean.
The harmonised system is largely focused on education and examination leading to the issue of a
‘Diploma’. In the definition above this is clearly a qualification. The key thing is that these
INDEX çç Sommaire
qualifications do not require any experience at all and therefore they do not, in themselves,
provide any assurance of competence. For example, a person can achieve the
European/International Welding Engineer (E/IWE) Diploma at the age of about 22, on the basis
of having completed a three year university degree in an engineering or materials subject, and the
14 week E/IWE course and examination. Such a person possesses a valid E/IWE Diploma but
without any understanding of how to apply the knowledge in a real working environment.
Unfortunately some employers are not aware of this and assume that a person with an
International/European Diploma is a competent person.
An International/European personnel certification scheme is in place and available [5]. This uses
the International/European Diplomas as proof of knowledge but sets additional requirements in
terms of experience and CPD. It is unfortunate that the scheme is much less popular than the
qualification scheme with only around 1500 certificates having been awarded world-wide. More
must be done to inform employers about the value of this certification scheme.
ISO 14731 requires the manufacturing organisation to appoint at least one Responsible Welding
Coordinator (RWC). It is useful that the standard makes reference to the International/European
qualification system as a way of dealing with the technical knowledge requirements for RWCs;
but what about RWCs who do not possess an International/European qualification? EWF/IIW
have developed guidance on this for certification bodies conducting compliance assessments of
manufacturers to ISO 3834, see Appendix 1. Routes (a) and (b) in Appendix 1 are
straightforward because of the presence of the international harmonised
certification/qualification. Route (c) is more complicated because it deals with RWCs without
such certification/qualification. Some experience of implementing route (c) in the UK is given in
the section below.
Welding Coordination as described in ISO 14731 covers many different welding and welding
related tasks covering the entire ‘welding quality chain’ mentioned in above. This includes:
Review of requirements
Technical review
Sub-contracting
Welding personnel
Equipment
Production planning
Qualification of welding procedures
Welding procedure specifications
Work instructions
Welding consumables
Materials
Inspection and testing before, during and after welding
Post-weld heat treatment
Non-conformance and corrective actions
Calibration and validation of measuring, inspection and testing equipment
Identification and traceability
Quality records.
INDEX çç Sommaire
In order to comply with the standard, the manufacturer must be able to show that employees with
any of the tasks listed in ISO 14731 possess appropriate competence. EWF/IIW have developed
a wide range of so-called ‘guidelines’ that describe courses of training designed to underpin
different welding related jobs, see Appendix 2. Therefore for many of the activities covered by
welding coordination in ISO 14731, a harmonised and recognised International/European course
is available. However, these courses are not widely used at present and, again, more must be
done to convince industry of their value.
In the field of non-destructive testing (NDT) there is a European and an International standard
[6,7] that deal with the independent certification of NDT personnel. These standards have not
resulted in the harmonised approach to NDT like the one that has been adopted in the
international welding community. However, they do provide a measure of assurance of
competence of NDT practitioners in possession of certificates that are in accordance with one of
the standards, especially if the certificates are backed by a recognised national accreditation
body. ISO 3834 requires NDT personnel to be qualified in accordance with ISO 9712 or an
equivalent standard.
There is no European or International standard for welding inspectors at present. However, there
is an International/European qualification scheme, see Appendix 2, and IIW has recently
approved a document that deals with the certification of welding inspectors. National
certification schemes for welding inspectors have existed for many years, some of which are
available internationally [8,9]. ISO 3834 requires welding inspectors to be competent but does
not require any particular qualification or certification. However, the take up of the welding
inspector certification schemes has generally been extraordinarily high (many tens of thousands)
reflecting the fact that certified welding inspectors is a common feature in the specification of
manufacturing or construction contracts world-wide.
It can be argued that welding is one of the most regulated occupations in the world. As we have
seen from the historical evidence above, the importance of welding skill was recognised over 90
years ago; and today, for all ‘serious’ welding, the welders are subject to a strict regime of
testing and re-testing in accordance with national, European and/or International standards.
These standards are, in many respects, similar; and it is a pity that attempts to achieve a single
harmonised international standard for the testing of welders have not been successful.
Nevertheless, for the manufacturer, a reasonable assurance of competence of welders is achieved
by adopting an appropriate standard, and welder testing has become embedded in the culture of
today’s manufacturing by welding.
Similarly, manufacturers usually take measures to ensure the competence of personnel involved
in NDT and welding inspection often by using one of the schemes mentioned above.
For all the other welding activities and related tasks and responsibilities identified in ISO 3834
and ISO 14731 there is no such culture of competence assessment. In recent years, TWI has been
INDEX çç Sommaire
involved in the assessment of a number of companies against the requirements of ISO 3834 and
this has led to a number of interesting issues with regard to competence assurance.
One manufacturer challenged the link between ISO 3834 and ISO 14731. The view of EWF/IIW
(and TWI) is that, in order to comply with ISO 3834 Part 2 (‘comprehensive’ quality
requirements) or Part 3 (‘standard’ quality requirements), the manufacturer must also comply
with ISO 14731. It is interesting that neither Part 2 nor Part 3 refers to ISO 14731 directly: the
issue is covered only by reference to Table 2 of Part 5 (normative references). This table does
indicate the need to comply with ISO 14731 but, in the general section of Part 5, in section 2.1
(according to a corrigendum) it is stated that ‘other documents’ or ‘different supporting
standards’ can be used in certain circumstances. Unfortunately this provides a basis for a
manufacture to challenge the need for compliance with ISO 14731.
Once the manufacturer has accepted the need for compliance with ISO 14731, the first issue to
deal with is the competence of the appointed RWC. As mentioned above IIW/EWF have issued
guidance on this (see Appendix 1) and, in dealing with route (c) people, TWI has also found it
beneficial to follow the requirements of document EA-6/02 [10] which state:
Special care should be taken by the EN ISO 3834 Assessment Team in evaluating the
competence of the manufacturer’s welding coordinators in accordance with EN ISO
14731. The Certification Body should have procedures, which demonstrate that this
important aspect of ISO 3834 is properly evaluated.
Such procedures should include a peer review and challenge process* in which the
company’s Responsible Welding Coordinators are interviewed and their work examined.
The EN ISO 3834 Assessment Team should be able to demonstrate complete evaluation
of welding co-ordination (functions and individuals) in the company. Records of this
process should be maintained.
* This means that technical discussions must take place between each Responsible
Welding Coordinator and the relevant Assessor about the detailed technical scope of the
Responsible Welding Coordinator's responsibilities. This process will require the
Assessor to examine evidence of completed work done by each Responsible Welding
Coordinator and to investigate his/her knowledge and understanding of it.
The peer review process should involve the examination of specific contract(s) to assess
compliance with the customer’s specification in, for example, the following areas:
INDEX çç Sommaire
In TWI’s experience there have been several instances where the appointed RWC has fallen
short of the requirements in terms of technical knowledge requirements of ISO 14731. In the
UK, RWCs are often people who started their careers in practical welding and, over a number of
years, have been successively developed their careers away from practical welding, for example
into welding inspection, quality assurance, etc, until they become the most senior welding person
in the company. These people are often perfectly competent to follow the procedures and
systems that have proved successful in the company over a period of time, but often they do not
have the underlying knowledge as to why things are done in the way that they are. This can be a
serious problem if there are changes, for example to the materials, welding processes or products
used by the company, or if something runs outside of the applicable specification. The person
could tell you, for example, what the preheat temperatures are for the all the joints to be welded
and how those temperatures are achieved, but may be completely ignorant as to why preheat is
necessary. In such cases, TWI recommends that the person attend the relevant
European/International Diploma modules in order to acquire the important underlying technical
knowledge.
If it is clear that there is no one in the company with the capability to meet the ISO 14731
requirements regarding the RWC, the company can engage an external consultant to fulfil this
role. In such cases, TWI advises that additional control measures are applied:
i) there must be a contract between the company and the external RWC which includes a
clear description of the RWC’s role and responsibilities
ii) the RWC must visit the company regularly (at least once per month) and there must be
record of the work carried out by the RWC in each visit
(iii) the RWC must have the right to visit the company anytime unannounced.
Anyone with welding responsibilities, no matter how narrow is categorised by ISO 14731 as a
welding coordinator and must be able to demonstrate competence in the allocated tasks. TWI
recommends that the manufacturer draws up a matrix of all the welding tasks identified in ISO
14731, together with the names of the people who are responsible for each one. As an example, a
storekeeper may be responsible for the correct storage of a wide range of equipment, tools, etc,
and this may include welding consumables. In such a case, the storekeeper would be expected to
know about the importance of correct storage of welding consumables and to ensure that correct
procedures are adhered to. If the storekeeper does not possess the competence to fulfil these
tasks, there is a weak link in the ‘welding quality chain’ and, even if the company has the best
welders and the best RWC in the world, problems may arise. The storekeeper is a welding
coordinator according to ISO 14731 and should be trained and recognised as such by the
company.
INDEX çç Sommaire
The final example is competence in NDT. As indicated above, ISO 3834 is clear about the need
for some kind of independent assessment of competence in accordance with ISO 9712 or an
equivalent standard. But is it really necessary? In one company visited by TWI, a person was
found performing magnetic particle inspection on a steel framework component for a fork lift
truck. The person appeared to be using appropriate techniques but on questioning him it was
found that he had been shown what to do by a colleague and did not know anything about the
principles of the NDT method he was using. Furthermore, when asked what action was required
if indications appeared the answer was that he did not know because he had never found
anything! The lesson to be learned from this is that, like welding, NDT is a special process and
care is needed to ensure that operators are trained and competent.
The competence requirements in ISO 3834 and ISO 14731 are a major step forward in
recognising the special importance of welding and in ensuring that welding production is under
competent control. What is important now is for standards writing bodies to call up the standards
in applications standards and for client organisations to specify the standards in their contracts.
Where possible it would also be beneficial to write in the need for some kind of recognised third
party assessment of compliance. Certification bodies are free to offer ISO 3834 assessment and
certification services, but these are only likely to bring value of they are under national
accreditation in accordance with EA-6/02. Even so there is evidence that accredited certification
to ISO 3834 is being carried out that is not in compliance with EA-6/02. The accreditation bodies
should address this: EA-6/02 is their document and they should be more rigorous in controlling
its implementation.
EWF/IIW has developed and implemented a system for ISO 3834 certification which fully
embraces EA-6/02 together with additional controls to ensure that certificates issued to
manufactures follow both the letter and the spirit of ISO 3834. This is exemplified by the special
requirements for RWC assessments as indicated in Appendix 1. Furthermore, assessing the
competence of RWCs can be facilitated by adopting the EWF/IIW personnel certification
scheme. All sides of industry should be made aware of the added value offered by these
certification schemes.
EWF/IIW have issued a wide range of training and qualification guidelines in addition to the
main ones that deal with the knowledge requirements for RWCs. Adoption of these guidelines,
where relevant to the tasks required, can also facilitate the assessment of competence in more
specialist welding and related areas of work. Further promotion of this is required.
With regard to the standards themselves, the link between ISO 3834 and ISO 14731 should be
made more solid. There should be no scope for challenging the applicability of ISO 14731 in the
case of compliance with ISO 3834 Part 2 or Part 3.
In the case of welding manufacture, health and safety, and environmental management are
especially important but neither of them is mentioned in ISO 3834 or in ISO 14731. It is
recommended that this is addressed by the bodies responsible for these standards. The omission
INDEX çç Sommaire
has been recognised by EWF and documents have been published about how to take care of
heath, safety and the environment from the welding point of view [12]. Manufacturers have the
option to undergo assessments against these requirements as part of their EWF ISO 3834
assessment.
CONCLUSIONS
1. Competence in welding is key to the safety, reliability and cost effectiveness of welded
products. This applies to all personnel to whom welding, or welding related, tasks have been
assigned.
3. Compliance with ISO 3834 and ISO 14731 is a way for manufacturers to ensure that they deal
with welding competence in a proper way.
4. EWF/IIW have developed qualification and certification systems which facilitate the adoption
of best practice in terms of competence demonstration.
5. More effort should be made to raise industry awareness of the importance of welding
competence and of the routes to achieve it.
6. ISO 3834 and ISO 14731 should be more widely specified in applications standards and in
contract documents.
7. In the light of recent experience some improvements to ISO 3834 and ISO 14731 with regard
to competence should be considered, particularly:
8. Accreditation bodies should do more to ensure that certification bodies fully satisfy the
accreditation bodies’ requirements in relation to ISO 3834 assessment and certification.
BIBLIOGRAPHY
INDEX çç Sommaire
[2] Lloyd’s Register of Shipping, London. Remarks submitted by the chief ship surveyor on
the application of electric arc welding to shipbuilding. Document appended to the
minutes of a meeting of the Technical Committee, 28 November 1918.
[3] ISO 3834:2005, in five parts, Quality requirements for fusion welding of metallic
materials.
[4] ISO 14731:2006, Welding coordination - Tasks and responsibilities.
[5] Quintino, L; Ferraz, R; and Fernandes, I; ‘International education, qualification and
certification systems in welding’. 5/2008. European Welding Federation web site:
www.ewf.be.
[6] BS EN 473:2008 Non-destructive testing. Qualification and certification of NDT
personnel. General principles.
[7] ISO 9712:2005, Non-destructive testing -- Qualification and certification of personnel.
[8] CSWIP Welding Inspector certification scheme. TWI Certification Ltd, web site:
www.cswip.com.
[9] AWS Certified Welding Inspector scheme. American Welding Society, web site
www.aws.org.
[10] EA Guidelines on the Use of EN 45 011 and ISO/IEC 17021 for Certification to EN ISO
3834. EA-6/02. July 2007 rev 1. EA web site: www.european-accreditation.org.
[11] UK-SPEC 2008, UK Standard for Professional Engineering Competence, Engineering
Technician, Incorporated Engineer and Chartered Engineer Standard. Engineering
Council UK web site: www.engc.org.uk.
[12] Quintino, L; Ferraz, R; and Fernandes, I; ‘International certification systems in welding’.
7/2008. European Welding Federation web site: www.ewf.be.
INDEX çç Sommaire
APPENDIX 1 - EXTRACT FROM IAB DOCUMENT IAB-340-08 (AND EWF
DOCUMENT 639-07) ABOUT ASSESSMENT OF RWCS (WWW.IIW-IIS.ORG)
Concerning the welding coordination personnel according to ISO 3834...... the following criteria
shall be considered:
(a) Certification according to IIW Personnel Certification Scheme (CIWE, CIWT, CIWS and
CIWP) shall be accepted provided the schedule of the person’s certificate is consistent with the
type of products being manufactured.
(b) If an IIW qualification is available (IWE, IWT, IWS and IWP) the welding coordinator(s)
could be accepted provided there is an adequate experience and competence in the products
being manufactured to be verified by means of a professional interview and curriculum vitae.
(c) If no certifications or qualifications are available, the ANBCC shall verify compliance by
means of a professional interview covering knowledge, experience and competence of the
welding coordinators with particular emphasis on welding technology, materials and their
behavior during welding, design fundamentals of welded construction, as well as fabrication and
inspection aspects (including knowledge of standards). If such a professional interview is
satisfactory, the welding coordinator(s) can be recognized by the ANBCC for this role and for
the products and processes consistent with the current production.
Due to the principle of mutual recognition between IIW and EWF whenever IIW qualification
and/or certification of welding coordination personnel is mentioned, the correspondent EWF
qualification and/or certification is considered equivalent.
The manufacturer shall define tasks and responsibilities of the welding coordination team, based
on ISO 14731 “Welding Coordination personnel – Tasks and responsibilities”.
INDEX çç Sommaire
APPENDIX 2 - LIST OF ALL IIW/EWF EDUCATION/TRAINING GUIDELINES
(WWW.EWF.BE)
16. EWF-494-01, Special Course in Laser Welding (Engineer, Technologist and Specialist
levels)
17. EWF-544-01, Special Course for Welding Reinforcing Bars at the Specialist level
19. EWF-627-07, Special Course on Personnel with responsibility for Macroscopic and
Microscopic Metallographic Examination
INDEX çç Sommaire
20. EWF-628-08, Special Course on Personnel with Responsibility for Heat Treatment of
Welded Joints
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page82
RÉSUMÉ / ABSTRACT
90C
RÉCENTS PROBLÈMES DE FISSURATION AU RÉCHAUFFAGE DANS
LES SOUDURES DE RÉACTEURS FABRIQUÉS EN ACIER CrMoV
RÉSUMÉ
Début 2008, lors de la fabrication de réacteurs de forte épaisseur en acier 2¼ Cr-1Mo-¼V, des micro-fissures au réchauffage furent détectées
dans les joints soudés réalisés à l’arc submergé (SAW). Ces micro-fissures furent décelées après traitement thermique par des méthodes
ultra-sons à haute sensibilité développées par les chaudronniers pour leur usage propre. La fissuration intervient toujours en métal fondu
des joints à l’arc submergé après traitement thermique intermédiaire de détente (ISR) ou post-soudage (PWHT). Les micro-fissures sont de
nature transverse à la direction de soudage et de tels défauts furent trouvés simultanément chez plusieurs chaudronniers, principalement en
Europe, sur plus de 25 réacteurs de forte épaisseur en 2¼ Cr-1Mo-¼V, à la fois sur des soudures longitudinales, circulaires ou de piquage
Mercredi
réalisées depuis début 2008. De telles observations de ce type de fissuration avaient déjà été reportées dans le passé mais il s’agissait toujours
de cas isolés, assez aléatoires dans leurs apparitions et limitées dans leur étendue.
Ce papier présente comment ces fissures furent tout d’abord découvertes et souligne les efforts réalisés par de nombreux laboratoires et
autres intervenants afin d’analyser et de déterminer les facteurs susceptibles de conduire à cette situation. Les différentes pistes suivies par
les fournisseurs de métal d’apport et les chaudronniers pour venir à bout de ce problème de fissuration sont également expliquées. Ce papier
décrit comment l’essai Gleeble® de traction à chaud, développé par le centre de recherche d’ArcelorMittal Industeel en France, a permis de
constater un creux de ductilité à chaud entre 620°C et 650°C, via l’abaissement des valeurs de striction (ROA), sur des échantillons prélevés
en métal fondu.
Ce papier fournit également des recommandations pour la réalisation d’essais de contrôle visant à assurer que le phénomène de fissuration
au réchauffage ne se produira pas lors des traitements thermiques post-soudage. De nouvelles procédures de contrôle non-destructif par
ultra-sons (UT) ont également été développées sont désormais recommandées à l’utilisation chez les chaudronniers afin de disposer d’une
méthode fiable qui assure la détection de ces potentielles micro-fissures dans les soudures SAW et le cas échéant de pouvoir procéder aux
réparations nécessaires.
ABSTRACT
In early 2008, reheat micro cracks were found in heavy wall 2¼ Cr-1Mo-¼V reactor submerged arc (SAW) welds. These micro cracks were
found by reactor fabricators using their own high sensitivity ultrasonic examination procedures after heat treatment. The cracking only
occurred in SAW deposited weld metal after intermediate stress relieving (ISR) or postweld heat treatment (PWHT). The micro cracks were
transverse in nature and were found by different fabricators, mainly in Europe, in over 25 heavy wall refinery hydroprocessing 2¼ Cr-1Mo-
¼V reactors in longitudinal, girth and nozzle welds welded since early 2008. There had been reports of intermittent findings of similar
cracking prior to 2008, but these past cases had been very limited and random in nature.
The paper outlines how these micro cracks were first discovered and highlights the efforts of numerous laboratories and investigators in
trying to isolate and determine the factors which may have attributed to the cause of the cracks. It explains the different avenues that reactor
and consumable manufacturers proceeded with in trying to overcome the micro cracking problem. The paper discusses how hot tensile
Gleeble® tests carried out by the ArcelorMittal Industeel laboratories in France indicated a significant dip in the reduction of area (ROA)
of weld test specimens between 620º-650ºC.
The paper provides recommendations for using welding consumable screening tests that were developed to ensure that future welding would
preclude reheat cracks occurring during subsequent heat treatment cycles. New ultrasonic testing (UT) procedures were also developed
and were recommended for use by the fabricators during fabrication to try to ensure that these micro cracks in reactor SAW weld could be
readily and reliably detected for subsequent repair during the fabrication process.
INDEX çç Sommaire
RÉCENTS PROBLÈMES DE FISSURATION AU
RÉCHAUFFAGE DANS LES SOUDRES DE RÉACTEURS
FABRIQUÉS EN ACIER CrMoV
L Antalffy,
Fluor Enterprises, Inc., One Fluor Daniel Drive, Sugar Land, Texas, 77478, USA.
C.Shargay
Fluor Enterprises, Inc., 3 Polaris Way, Aliso Viejo, California, 92698, USA.
D Smythe
Fluor Enterprises, Inc., One Fluor Daniel Drive, Greenville, S.C., 29607-2762, USA.
K Moore
Fluor Enterprises, Inc., 3 Polaris Way, Aliso Viejo, California, 92698, USA.
RESUME
Ce papier présente comment ces fissures furent tout d’abord découvertes et souligne les
efforts réalisés par de nombreux laboratoires et autres intervenants afin d’analyser et de
déterminer les facteurs susceptibles de conduire à cette situation. Les différentes pistes
suivies par les fournisseurs de métal d’apport et les chaudronniers pour venir à bout de ce
problème de fissuration sont également expliquées. Ce papier décrit comment l’essai
Gleeble® de traction à chaud, développé par le centre de recherche d’ArcelorMittal
Industeel en France, a permis de constater un creux de ductilité à chaud entre 620°C et
650°C, via l’abaissement des valeurs de striction (ROA), sur des échantillons prélevés en
métal fondu.
INDEX çç Sommaire
traitements thermiques post-soudage. De nouvelles procédures de contrôle non-destructif
par ultra-sons (UT) ont également été développées sont désormais recommandées à
l’utilisation chez les chaudronniers afin de disposer d’une méthode fiable qui assure la
détection de ces potentielles micro-fissures dans les soudures SAW et le cas échéant de
pouvoir procéder aux réparations nécessaires.
ABSTRACT
In early 2008, reheat micro cracks were found in heavy wall 2 Cr-1Mo-V reactor
submerged arc (SAW) welds. These micro cracks were found by reactor fabricators using
their own high sensitivity ultrasonic examination procedures after heat treatment. The
cracking only occurred in SAW deposited weld metal after intermediate stress relieving
(ISR) or postweld heat treatment (PWHT). The micro cracks were transverse in nature and
were found by different fabricators, mainly in Europe, in over 25 heavy wall refinery
hydroprocessing 2 Cr-1Mo-V reactors in longitudinal, girth and nozzle welds welded
since early 2008. There had been reports of intermittent findings of similar cracking prior to
2008, but these past cases had been very limited and random in nature.
The paper outlines how these micro cracks were first discovered and highlights the efforts of
numerous laboratories and investigators in trying to isolate and determine the factors which
may have attributed to the cause of the cracks. It explains the different avenues that reactor
and consumable manufacturers proceeded with in trying to overcome the micro cracking
problem. The paper discusses how hot tensile Gleeble® tests carried out by the ArcelorMittal
Industeel laboratories in France indicated a significant dip in the reduction of area (ROA) of
weld test specimens between 620º-650ºC.
The paper provides recommendations for using welding consumable screening tests that
were developed to ensure that future welding would preclude reheat cracks occurring during
subsequent heat treatment cycles. New ultrasonic testing (UT) procedures were also
developed and were recommended for use by the fabricators during fabrication to try to
ensure that these micro cracks in reactor SAW weld could be readily and reliably detected
for subsequent repair during the fabrication process.
INTRODUCTION
The evolution of modern hydroprocessing technology can be traced back to the mid 1920’s
in Germany, where quenched and tempered chrome molybdenum steels were used for
reactors in high pressure hydrogenation plants. These vessels were fabricated from 2.25% to
3.8% chrome molybdenum alloys with operating pressures in the 28 to 70 MPa range(1).
Today, hydroprocessing reactor technology has evolved into the fifth generation(2) with each
new generation introducing significant improvements. Currently, operating conditions for
modern hydroprocessing reactors can range from 400°C and 10 MPa hydrogen partial
pressure to above 482°C and hydrogen partial pressures as high as 35 MPa.
These high hydrogen partial pressures and temperatures were the focal point for the
introduction of new grades of vanadium-modified 2Cr-1Mo and 3Cr-1Mo steels with
higher strength levels, increased hydrogen attack resistance, and with guaranteed 54 Joule
step-cooled transition temperatures (TT54) toughness as low as -29°C in the deposited weld
metal. The evolution of hydroprocessing reactors has not only been enhanced by these
recent material advances, but also significant advances in design, fabrication, inspection,
INDEX çç Sommaire
testing and ultrasonic examination techniques have contributed to provide reliable and
longer operational service lives.
Today, there are over six hundred and fifty vanadium-modified reactors that are either in
service, or have been delivered for service, or are presently under fabrication around the
world. This has been a great achievement since the first commercial vanadium modified
reactor only went into service in 1989. Because of the above characteristics, vanadium-
modified alloys have become the materials of choice for hydroprocessing reactors as
operating data has become more available, as fabrication technologies and techniques have
improved and as fabrication experience has grown. The 2007 ASME Section VIII Division
2 Pressure Vessel Code has also permitted even higher allowable stresses than permitted by
the previous editions and this event will inevitably lead to even greater acceptance of the
vanadium-modified materials.
Studies originally indicated that low impurity content vanadium-modified 2Cr-1Mo alloy
had similar susceptibility to stress relief cracking(3) as the conventional steel however, the
present day experience, as explained later in this paper, indicates that the vanadium-
modified steel appears to have a greater susceptibility. Other inherent disadvantages of this
material include limited availability of welding consumables, a need for higher preheat
temperatures, higher weld metal and heat affected zone (HAZ) hardnesses, a need for higher
PWHT temperatures, higher 54 Joule toughness transition temperatures, very low as-
deposited weld metal notch toughness and the need for intermediate stress relief (ISR) in
lieu of a dehydrogenation heat treatment (DHT) for restrained weld joints.
In January 2008, a European fabricator started detecting indications when carrying out its
own high sensitivity in-house ultrasonic examination following heating for re-rolling or ISR.
This high sensitivity ultrasonic testing (UT) procedure was utilized by the fabricator to
ensure that the vessel welds were free of indications that could potentially require repairs
INDEX çç Sommaire
when detected by non destructive examination (NDE) procedures required by Code and
contract specifications after final PWHT of the vessel. The indications were discovered with
70° transducer probes operating at frequencies of 2-4 MHz and calibrated to a 3 mm side
drilled hole (SDH).
The indications were transverse in the weld metal and could only be detected with the
probes traversing in one direction and were not apparent when the probes traversed in the
reverse direction. These indications led to more detailed UT examinations which confirmed
that the indications were micro cracks which had to have occurred during heating for re-
rolling or ISR as there were no indications in the weld metal prior to these heat treatment
steps. It was quickly determined that this was a new phenomenon that definitely did not
occur at this fabricator prior to December 2007/January 2008.
The indications were generally short in length ranging between 3 to 10 mm in length and
height, were located within the weld metal and not the heat affected zones (HAZ) and were
transverse to the longitudinal axis if the weld. At some locations within the welds there was
a single indication, but in most locations the indications were in clusters of many parallel
indications lined up in one given plane. The locations of the indications first appeared to be
close to the inner and outer surfaces of the weld between 10 to 40 mm from the surfaces.
However, more extensive examination and experience with reactors of other thicknesses
determined that the indications were dispersed at all levels within the welds. An interesting
and disheartening discovery was that these indications did not occur in heat treated
qualification or production test welds but only in actual production welds on vessel
components indicating that residual weld stress of a constrained weld joint may be a
contributing factor.
The welding in all cases was carried out by narrow gap submerged arc welding (SAW). The
indications were found in the circumferential welds of forged as well as rolled-plate shell
courses, longitudinal rolled-plate welds and nozzle forging welds but only in vanadium-
modified SAW welds. Soon after, other heavy wall fabricators in Europe and in Asia were
finding similar indications and similar problems were discovered in welds deposited with
consumables supplied by two different manufacturers. Altogether, some 25 vanadium-
modified heavy wall reactors, high pressure separators and heat exchangers ranging in wall
thickness of 100 mm to 300 mm appeared to be affected. These indications were only found
in pressure equipment fabricated using welding consumables from two European
manufacturers, whereas no problems were reported with consumables from a Japanese
manufacturer.
This very soon became a serious situation and welding fabrication came to a virtual
standstill as fabricators, welding consumable manufacturers and fabricator’s clients wanted
answers to what these indications were and what was their root cause. Boat samples were
removed and investigated and the first conclusive report based on optical microscopy
indicated that the indications were indeed cracks, inter granular in nature, propagating along
grain boundaries and that the cracking may be reheat cracking. Figures 1 and 2 respectively
indicate a photo micrograph picture of typical fine inter granular reheat cracks and a ground
section of a weld with evidence of dye penetrant bleed through indicating crack locations.
Other names for such cracking include stress relief cracking and stress relaxation cracking.
INDEX çç Sommaire
Figure (1) Typical Reheat Cracks In A Vanadium Modified Weldment
A number of independent laboratories were asked to carry out testing and electron
microscopy (SEM) on cracked weld samples and to compare the metallurgy with uncracked
samples welded with older consumables that had not exhibited the cracking phenomenon.
INDEX çç Sommaire
Research and welding laboratories that worked on samples provided by welding consumable
manufacturers and fabricators included The Welding Institute (TWI) in the United
Kingdom, ArcelorMittal Industeel, Air Liquide Welding, Biophy Research and a refiner’s
research center in France, AQM in Italy and the Belgium Welding Institute. Reports were
generated by these institutions in late February, March and April 2008 and from some of
these reports the following conclusions were reached:
• The transverse micro cracks were reheat cracks. No signs of hydrogen cracking were
detected.
• Chemical compositions of cracked welds and non cracked welds welded with older
consumable were essentially very similar and no significant differences were observed
between the approximately 30 alloying elements and the impurity elements that were
analyzed
• Cracking was determined to be entirely inter granular
• There appeared to be no evidence that indicated a clear involvement of hydrogen in the
cracking process within the welds
• Hot tensile Gleeble® tests carried out by ArcelorMittal Industeel indicated a decrease in
ductility in the critical temperature range of 620°C -650°C of the affected welding
consumables compared to those that did not cause cracking
• Suggestions were made to optimize welding parameters to give smaller weld beads, to
optimize batch selection of wire and flux to give lowest levels of Mn, Cr, Mo and V
(within specification limits) and lowest levels of impurities, to raise preheat and
minimum interpass temperatures slightly, to lower the maximum interpass temperature,
to raise the ISR temperature, to lower cooling rates and increase heating rate to cross
the sensitive range as quickly as possible. All of these steps were then tested on
production welds as described below.
European fabricators were also making tests of their own by trying to optimize welding,
preheat, ISR temperature and heat up and cool down parameters. One fabricator quickly
tested all these welding variables and showed that these steps usually did not prevent
cracking. They also did an extensive review for any changes in equipment, welders, ovens,
building facilities, etc,. from before January 2008 when the welds were crack free, to after
that time when the weld cracking began to occur. They confirmed that there had been no
changes in any of these factors.
Another fabricator carried out extensive testing and concluded that the reheat cracking
problem could be reduced but not completely eliminated by modifying these parameters.
Various fabricators set up ongoing fabrication plans which included welding, testing, ISR,
NDE, repairs if necessary by rewelding, etc., and proceeding with some manual welding
with shielded metal arc electrodes (SMAW) where such welding was feasible in order to try
to make progress on their fabrication completion schedules.
The decrease in ductility in the 620°C -650°C range reported by ArcelorMittal Industeel(4)
and indicated in Figure 3 generated a lot of interest and helped to explain the reheat cracking
phenomenon. It is known that thick-wall narrow gap welds generate high axial residual
stresses as the weld bead solidifies and for submerged arc welds the residual principal stress
intensity can be as high as 1.5 times the yield stress of the deposited weld metal (5). When the
weld material is undergoing an ISR and before the weld metal’s residual stress is relieved, it
loses some of its ductility in the 620-650°C range and this low ductility is then apparently
INDEX çç Sommaire
(4)
Figure (3) Typical Gleeble Test Comparison Results
sufficient to generate transverse cracks within the highly stressed weld metal.
Welding consumable manufacturers also suggested that high interpass temperatures could
also lead to reheat cracking, so emphasis was placed on reducing interpass temperatures
(even though the interpass temperatures that used were the same as had been used for the
successful welds pre-Jan. 2008). Heat treatment parameter modifications had initially been
suggested as a potential remedy by one of the welding laboratories and one fabricator
reportedly had been able to reduce reheat cracking in circumferential welds but not
eliminate them in nozzle welds.
After a few months, this fabricator however, appeared to have success in ranking flux
batches based upon a developed damage index matrix developed by looking microscopically
at fracture surfaces of welds deposited in such a condition that replicated production
conditions. He then determined a damage index based upon the presence of fracture inter
granular features and used this information in subsequently optimizing the welding process
parameters such as heat input, bevel geometry, wire diameter, current and welding process,
and modifying his ISR heat up and cool down parameters when required. Heat treatment
parameter modifications had earlier been suggested as a potential remedy by one of the
welding laboratories.
By March 2008, the main focus by at least one welding consumable manufacturer appeared
to be on flux cleanliness and one of their first steps was to produce a “super clean flux”. The
Gleeble® test was used as the basis the acceptance criteria for the new fluxes and the goal
was to produce welds with results similar to 2005 “good” welding consumables. On a urgent
basis, the ArcelorMittal Industeel laboratory developed a detailed test procedure necessary
for obtaining accurate and repeatable results(6). They also helped determine an acceptance
criteria for the high temperature Gleeble® tensile test which was also dependent on the
geometry of the tensile specimen, the heating rate, the specimen loading or strain rate and
INDEX çç Sommaire
the location of the Gleeble® test specimen within the deposited weldment. Without setting
these parameters, results can vary greatly by laboratory, equipment and test method.
During the course of time there had been six “super clean flux” batches that had been
developed by the welding consumable manufacturer each of which ended up in being unable
to consistently produce reheat cracking resistant welds. However by the end of July 2008,
with “Batch 7”, it appeared that one consumable manufacturer had been able to identify the
contaminant elements in the flux responsible for the reheat cracking phenomenon.
The ArcelorMittal Industeel working with consumables manufacturer determined that since
the K factor in the weld wire itself was always low, the primary differences in the
contaminant elements had to be coming from the flux. Using this information, the
consumables manufacturer developed a synthetic prototype flux in his laboratory which
produced very good Gleeble® test results. Batch #7 of “super clean flux” based upon the
laboratory prototype was produced in mid August and this batch was used in production
welding and resulted in welds free of reheat cracking.
During the time period since this cracking initiated, Fluor had many vanadium-modified
reactors under fabrication at a number of different fabricators, and hence was also proactive
in working with fabricators to develop comprehensive and reliable ultrasonic testing
procedures which focused on detecting the transverse reheat cracks throughout the full weld
thickness. Several factors make detection of these cracks difficult to accomplish:
1. The size of the cracks can be very small relative to the thickness of the material;
typically they are 4 mm wide by 4 mm high.
2. Because the cracks are oriented transversely to the weld manual shear wave scanning
must be performed on the top of the weld so the cap usually needs to be ground flush.
3. The typical curvature and the wall thickness of the vessel limit the depth to which the
manual shear wave primary detection angle of 70º can be used (Figure 5)
4. The submerged arc process can sometimes have co-planar slag inclusions that can
appear as reheat cracks on Time of Flight Diffraction (TOFD) scans.
INDEX çç Sommaire
Figure 4. The Developed Relationship Between ROA(%) And The K Factor(6)
In this procedure the TOFD technique supplemented by manual shear wave in “blind zones”
was to be used for initial weld flaw detection and this technique detects and rejects a high
percentage of cracking. The procedure requires scans with multiple probe setups to ensure
that the full weld thickness is thoroughly inspected.
Manual high sensitivity shear wave angle beam UT examination was to be used to
characterize relevant indications and determine their orientation and subsequent acceptance
or rejection according to established criteria. Both the TOFD and manual UT procedures
have now also been incorporated as Appendix A of API 934-A(7) to supplement API 934-A
fabrication and testing requirements.
Fluor was also instrumental in mandating hot tensile Gleeble® testing of all batches of
welding consumables prior to their use in production welding. This protocol or similar high
temperature tensile tests now appears to have been accepted at least for the interim by
European vessel fabricators as well as by the European consumables manufacturers.
Research to develop alternate test methods is currently being conducted by a number of
organizations to find a cheaper and/or more suitable characterization test for batches of
welding consumables to determine their susceptibility to reheat cracking.
Figure (5) Typical Curvature and Vessel Wall Thickness Limit The Depth That Manual Shear Wave
Primary Detection Angle of 70 Degrees Can Be used
INDEX çç Sommaire
It is interesting to note that the Japanese fabricators have not experienced the recent reheat
cracking problem in vanadium-modified welds following heat treatment. The reason appears
to be that the Japanese weld wire/flux formulation was quite different to the European
combinations. The Japanese filler metal supplier produces a bonded flux, while the
Europeans produce an agglomerated flux. It was also the Japanese who first developed
vanadium-modified consumables for the 3Cr-1Mo-V alloy in the late 1980’s using Kobe
Steel’s wire/flux combination. These consumables were formulated following intensive
testing to resist reheat cracking provide good weldability and were also used in the first
commercial vanadium-modified reactor that was fabricated and delivered to Canada by
Japan Steel Works in 1991.
The Japanese company had determined a strategy to prevent reheat cracking and used a
formulation of weld wire/flux combinations with special chemistry considerations. Gotoh et
al(8) determined that reheat cracking occurred when the difference between the strength of
the grain and the strength of the grain boundary became large. Based on the Japanese
company’s extensive research phase of welding of vanadium-modified materials, they found
two important chemistry considerations help their consumables to avoid transverse reheat
cracking in weld deposits.
1) The vanadium content in the weld metal must lie between 0.30 and 0.38%(8)
2) The oxygen content in the weld metal must be greater than 300 ppm.(9)
These fundamental differences in the approach to deposited weld metal chemistries help
explain the reason that fabricators using Japanese consumables have not experienced the
recent transverse reheat cracking problem in vanadium-modified Cr-Mo weldments.
However, the differences in the flux manufacturing methods between the Japanese and
European welding consumable suppliers meant that the special Japanese chemistry could
not be directly implemented by the European welding consumable manufacturers.
To summarize, at the time of the writing of this paper, confidentiality still the shrouds the
welding flux constituents that made up the basis of K factor and Batch No. 7 flux that
appeared to turn the tide in solving the reheat cracking problem in 2 Cr-1Mo-V welds
for one consumables manufacturer. Numerous welds have been completed, heat treated and
examined with highly sensitive ultrasonic weld testing procedures without finding evidence
of reheat cracks.
Apparently, it may be concluded that the reheat micro cracking issue in vanadium-modified
welds submerged arc welds has been solved by both European welding consumable
manufacturers following the modifications to and conducting subsequent characterizations
of their individual products. It is anticipated that in the near future the welding consumables
manufacturers will fully divulge their findings and provide a set of flux/wire cleanliness
parameters by which both weld wires and associated fluxes can be judged to be acceptable
for welds free of reheat cracking.
The toughness of the welds which had reheat cracking (in areas away from the cracks) was
shown to be acceptable after full PWHT by the production test plates and limited boat
samples testing. Although not tested on these specific welds, the toughness of all Cr-Mo
welds as-welded and after ISR are known to be low, with the vanadium-modified grade
being somewhat lower than then straight 2Cr-1Mo grades. However, there appeared to be
INDEX çç Sommaire
no evidence that the toughness played a role in the reheat cracking, or that toughness levels
had changed from the previous welds made before January 2008.
The techniques described in the Appendix A are focused on detecting transverse reheat
cracks occurring in the SAW weld deposits. In principle, TOFD procedures are used
augmented with manual UT to characterize flaws but in each case the procedures are
verified against specially made demonstration blocks to increase procedure sensitivity. The
default inspection mode is from the outside diameter (OD), but if the weld has not been
overlayed by stainless steel, the technique should be applied from the inside diameter (ID)
as well. This is due to the fact that often manual shear wave examination cannot reach the
ID as previously noted.
The procedure uses ultrasonic TOFD for initial detection, and this technique will detect and
often reject a high percentage of the cracking. The TOFD procedure should also comply
with ASME Section V, Non Destructive Examination, Art. 4 App. III 2007. This section
was extensively re-written for 2007 and contains many important provisions. TOFD was
chosen as it has several advantages of manual UT alone. Among them, manual UT is very
operator-dependent and there is no permanent record. Also, manual UT is prone to a
reduction in probability of detection (POD) as a result of operator fatigue.
Most important, for this application, are the limitations of the method to detect reheat cracks
at maximum depths and the sensitivity of the method to the orientation of the flaws.
However, manual UT is used to characterize reflectors (as there are often some indications
that are other types of non-injurious defects) and to scan the TOFD blind zones. On nozzle
INDEX çç Sommaire
welds, where TOFD often cannot be carried out, 100% manual UT or phased array UT is
necessary to inspect for weld metal reheat cracking.
Reheat cracking has been indicated in TOFD B-scans as intermittent co-planar (in the
through-wall direction) diffractors typically appearing in clusters as indicated In Figure 6.
These indications can then be investigated by manual shear wave angle beam UT techniques
to determine their primary orientation and shape. Flaws which are found to be planar and
transverse in their primary orientation should be considered reheat cracks.
Figure (6) A Typical TOFD Scan Indicating Reheat Cracks As A Series Of Co-planar Point Diffractors
One of the primary areas where the procedure exceeds ASME Code Section V, Article 4
requirements is in the calibration standard. Whereas, the Code would require calibration on
a SDH with a diameter which is a function of the wall thickness and would range from 6-10
mm for typical reactor wall thicknesses, this procedure requires a 3 mm SDH and electrical
discharge machine (EDM) notches resulting in the ability to detect much smaller
indications.
INDEX çç Sommaire
The demonstration block described in Appendix A of API 934-A and indicated in Figure 7
contains EDM notches machined from the side of the block at each 10% of thickness and
offset by 15% of thickness to allow clear interpretation of the diffracted signal. The notches
are 4 mm high and are welded closed to a width of 4 mm. This leaves a planar, transversely
oriented flaw of 4mm by 4mm x 0.5 mm size. Tests show that this has a very close signal
response to reheat cracks of the same size. They also represent about the limit of sensitivity
for current state-of-the-art TOFD equipment.
1 x Pro b e Wid th
Offse t Sca n Sca n
Welded Fill
1 x Pro b e Wid th
Ce n te re d Sca n
TOFD Pro b e
Se tu p
6 m m d e p th to
to p o f 4 m m EDM
n o tch
Side of Block
T
(TYP )
0 .1 T
0 .1 5 T ( TYP)
Figure (7) TOFD Sensitivity Demonstration Blocks For Transverse Reheat Cracks
Rejection Criteria
The following criteria apply to TOFD examination of thick wall sections when examining
welds for transverse reheat cracks.
ll. Clusters
If three indications (point diffractors) are observed in the beam same through
thickness plane (+/- 2.5 mm) and separated by 50 mm or less they shall be considered
reheat cracks unless manual ultrasonic examination can demonstrate that they are not
planar and not transverse.
Manual shear wave UT examination is also used for characterizing flaws indicated by
TOFD. Characterization methods based on EN 1713(11) are used to determine if the flaws
are reheat cracks i.e. they are planar and transversely oriented. This method requires
scanning the flaw from both the longitudinal and transverse directions and then comparing
the amplitudes obtained from each direction. Where the amplitude obtained from the
longitudinal direction is 9 dB greater than that obtained from the transverse direction the
flaw can be considered to be transversely oriented. The flaw shape can be determined by
investigating the flaw from one direction using two different probes with angles that have
a difference of at least 10°. Again, if the difference in amplitude is greater than 9 dB the
flaw can be considered planar.
The manual UT rejection criteria are based on distance amplitude corrected (DAC)
response. When the response is greater than 20% of DAC, the flaw is rejected, if the
response lies between 10-20% of DAC then further investigation and characterization are
required as noted above.
INDEX çç Sommaire
SUMMARY
The new vanadium-modified steels offer significantly higher design stress intensity values
especially as permitted by the new 2007 edition of the ASME Section VIII Division 2
Pressure Vessel Code with improved resistance to hydrogen attack, weld overlay
disbonding, temper embrittlement and provide lighter, thinner wall reactors.
Unfortunately, despite the many benefits of the vanadium-modified steels, these materials
can have and have presented problems during fabrication. It is important that fabricators
using these materials understand and be familiar with these problems and have previous
experience in fabricating with vanadium-modified steels.
The paper has also touched on some of the less desirable characteristics of the vanadium-
modified steels, mainly the need for higher preheat and PWHT temperatures, higher
hardnesses in the base metal, weld metal and HAZ, the need for ISR in lieu of DHT in
restrained weld joints and the extra general care and effort required during fabrication with
these steels.
The paper focused on SAW weld material’s susceptibility to transverse reheat cracking
outlining the recent incidents in Europe in some 25 heavy wall reactors being fabricated
by different fabricators. The paper has also outlined the tremendous research effort that
was undertaken by a number of different fabricators, welding laboratories and independent
laboratories in Europe in order to determine the root cause of the problem and develop a
subsequent consumables testing and characterization program. The paper also discussed
the high sensitivity ultrasonic techniques that needed to be developed and utilized to find
these transverse reheat microcracks in welded and heat treated SAW girth, longitudinal
and nozzle welds and the criteria used for flaw indication acceptance/rejection during
ultrasonic examination.
REFERENCES:
(1) Erwin W.T., Kerr J.G., “The Use of Quenched and Tempered 2Cr-1Mo Steel for
Thick Wall Reactor Vessels in Petroleum Refinery Processes”. Welding Research Council
Bulletin 275, N.Y., February 1982.
(2) Antalffy, L.P., Chaku P., “Metallurgical Design & Fabricators Aspects of Modern
Hydroprocessing Reactors” presented at The Materials Properties Council Second
International Conference on the Interaction of Steels with Hydrogen in Petroleum Industry
PVP Service. Vienna, Austria, Oct 20 1994.
(4) Chauvy C., “Reheat Cracking Of Weld Metal A542D Susceptibility Tests” A report
from ArcelorMittal Industeel, presented at the API 934 meeting in Salt Lake City, Utah,
page 9, September 2008.
INDEX çç Sommaire
(6) Chauvy C., Pillot S., “Prevention of Weld Metal Reheat Cracking during Cr-Mo-V
Heavy Reactor Fabrication”, American Society of Mechanical Engineering (ASME), PVP
Conference 2009, Prague, Czech Republic, Paper PVP 2009-78144.
(7) API Recommended Practice 934-A “Recommended Practice For Materials and
Fabrication of 2Cr-1Mo, 2Cr-1Mo-V, 3Cr-1Mo and 3Cr-1Mo-V Steel Heavy
Wall, Pressure Vessels for High Temperature, High Pressure Hydrogen Service”
American Petroleum Institute, Washington, D.C. 2010.
(8) Gotoh A., Nakagawa T., Hatano H., and Wada T., “The Effects Of Vanadium
Carbide Precipitation in 1-3%Cr-Mo Steel Weld Metals”, Int Institute of Welding
Document XI-708/99.
(9) U.S. Patent, Submerged Arc Welding Method for High Strength Cr-Mo Steel, No.
5,430,269, July 4, 1995.
(10) Code Case 2235-9 “Use Of Ultrasonic Examination In Lieu Of Radiography Section
I; Section VIII, Divisions 1 and 2; and Section XII”, Cases of ASME Boiler and Pressure
Vessel Code.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page83
RÉSUMÉ / ABSTRACT
98C
ÉTUDE DE SOUDABILITÉ DE JOINTS EN MATÉRIAUX DIFFÉRENTS UTILISÉS
DANS LA FABRICATION DE RÉCHAUFFEURS DE PRODUCTION D’ÉNERGIE
Pontevedra
[email protected] / [email protected] / 00 34 986 344 000
bIngeniería y Montajes Rías Bajas, Carretera Ourense-Pontevedra (N-541), km 80
RÉSUMÉ
Les réchauffeurs finals, qui sont implantés dans les centrales modernes à cycle combiné pour augmenter
la productivité, se composent d’un grand nombre de bobines fabriquées par pliage et soudage de tubes.
Au cours de leur durée de vie, ils souffrent de fortes contractions thermiques et dilatations. Par
conséquent, il est nécessaire de limiter le mouvement transversal par des compensateurs de dilatation.
Les tubes sont constitués d’aciers CrMo résistants au fluage et les compensateurs sont fabriqués en
acier inoxydable austénitique qui présente une excellente tenue à la corrosion. Les matériaux impliqués
dans les unions à réaliser présentent des propriétés très différentes qui sont même opposées aux
exigences du soudage. Ce document présente les résultats d’une étude de soudabilité impliquant
différentes technologies de soudage à l’arc et une large gamme de matières de base et combinaisons de
fils d’apport. Afin de caractériser les soudures, des essais destructifs et analyses métallographiques ont
été réalisés. En outre, les soudures GMAW, réalisées avec différents types de fils d’apport, ont été
soumises à un cycle de température dans un four dans le but d’accélérer les phénomènes de dégradation
qui ont lieu à haute température. Pour évaluer l’effet de ce processus, une étude approfondie de la
microstructure a été effectuée ainsi qu’une vérification de la variation des propriétés mécaniques.
ABSTRACT
Final superheaters, which are equipment implanted in modern combined cycle power plants to increase
productivity, consist of a large number of coils manufactured by bending and welding of tubes. During
its lifetime, these systems suffer strong thermal contractions and dilations. Therefore, it is necessary to
limit transverse movement by means of expansion compensators. The tubes are made in CrMo creep
resistance steels and compensators in excellent corrosion behaviour austenitic stainless steel. The
materials involved in such heterogeneous joint present very different properties and even have opposed
welding requirements. This paper shows the results of a weldability study involving different
technologies of arc welding and a wide range of base and filler material combinations. In order to
characterize the welds were performed destructive tests and metallographic analysis. Moreover, only
GMAW welds, made with different kinds of wire, were subjected to a temperature cycles in a furnace
with the aim of accelerating degradation phenomena that take place at high temperature. To evaluate
the effect of this process, it was carried out a comprehensive microstructural study and a check of the
variation of the mechanical properties.
AFIAP - ESOPE 2010 Programme
INDEX çç Sommaire
1. BACKGROUND
The energy requirements demanded by emerging and developed countries are increasingly
higher. Therefore, power plants must increase their efficiency and productivity in order to
attend them. In this sense, final superheater is one of the new equipment implanted in modern
combined cycle power plants to increase productivity. These systems heat the combustion
gases from the first stage of the cycle to produce steam which expands in a turbine generating
additional power [1].
The heaters consist of a large number of coils manufactured by bending and welding of tubes.
During its lifetime, these systems suffer strong thermal contractions and dilations. Therefore,
it is necessary to limit transverse movement by means of expansion compensators (see Figure
1). The tubes are made in CrMo creep resistance steels and compensators in excellent
corrosion behaviour austenitic stainless steel. This system is exposed to termomechanical
fatigue and oxidation.
The company “Ingeniería y Montajes Rías Bajas”, which belongs to the boilermaking
Galician sector, is specialized in the maintenance, repair and modification of final
superheaters. With the support of AIMEN Technology Centre, it is developing a project
focused on the study of weldability and high temperature behaviour of dissimilar metal welds
(SATELITE, "Soldadura a Alta Temperatura en Equipos Ligados a la Industria TÉrmica").
The materials, that take part in this joint, show very different properties (coefficients of
thermal expansion, ductility, electrical conductivity, etc.) and even present conflicting
welding requirements. On the one hand, CrMo steels have high hardenability and a strong
tendency to crack in cold. It is essential to take into account the processes of preheating and
post weld heat treatment (PWHT) in the welding procedure. On the other hand, austenitic
stainless steels present problems arising from the formation of carbides and brittle phases
when they are subjected during certain periods of time at temperatures typical of PWHT
needed in the case of CrMo steels. [2, 3]
The choice of filler material is also crucial. It must exhibit very specific characteristics to suit
INDEX çç Sommaire
the requirements of the welding process and the specific application. When operating
temperatures are relatively high (> 400 º C, as in the case study), using of nickel-based filler
instead of stainless steels materials is recommended for two main reasons. Firstly, it reduces
the migration of carbon from CrMo steel into the molten zone. Secondly, it minimizes stress
variables of temperature fluctuations, since its expansion coefficient has an intermediate value
between materials to be welded. Therefore, in principle, nickel-based filler materials are ideal
candidates for these applications. [2, 3, 4, 5, 6 and 7].
2. EXPERIMENTAL METHODOLOGY
2.1 Materials
In the tests, two grades of CrMo steel and one of stainless steel were used. The chemical
composition and mechanical properties of these materials are summarized in Tables 1 and 2.
Grade C Si Mn P S Ni Cr Mo Cu N
10CrMo9-10 0,08 - 0,14 0,50 0,40 - 0,80 0,020 0,010 2,00 - 2,50 0,90 - 1,10 0,30 0,012
13CrMo4-5 0,08 - 0,18 0,35 0,40 - 1,00 0,025 0,010 0,70 - 1,15 0,40 - 0,60 0,30 0,012
X6CrNi25-20 0,04 - 0,08 0,70 2,00 0,035 0,015 19,0 - 22,0 24,0 - 26,0 0,10
Table 1
Chemical composition of base materials
The thickness of 10CrMo9-10, 13CrMo4-5 and X6CrNi25-20 steel plates were 5.8 mm, 5.2
mm and 6.0 mm, respectively.
As filler materials, three different types of shielded electrodes for manual welding and three
different types of wire for robotic processes (GMAW and GTAW) were selected. Chemical
composition, mechanical properties at room temperature and identification data of filler
materials are shown in Tables 3, 4 and 5.
Designation C Si Mn P S Ni Cr Mo Nb Fe Cu Ti Al
ENiCrFe-3 0,04 0,40 7,5 Bal. 16,0 2,0 7,0
ENiCrFe-2 0,03 0,60 4,7 Bal. 19,0 1,5 1,9 4,0
EPRI P87 0,10 0,30 1,5 0,008 0,008 Bal. 9,0 2,0 1,0 38,0
ERNiCrMo-3 0,02 0,08 <0,1 <0,010 0,001 64,5 22,3 8,7 3,6 1 2,0 0,01 0,16 0,07
2,50- 18,0- 2,0-
ERNiCr-3 <0,10 <0,50 <0,030 <0,015 67,02 <3,0 <0,50 <0,75
3,50 22,0 3,0 1
ER309LSi <0,03 0,8 1,8 13,0 24,0 <0,3 Bal. <0,3
1 Nb+Ta / 2 Ni+Co
Table 3
Chemical composition (%) of filler materials
INDEX çç Sommaire
Designation Yield Strength (MPa) Tensile Strength (MPa) Strain (%)
ENiCrFe-3 > 390 > 620 > 30
ENiCrFe-2 400 650 40
EPRI P87 360 560 34
ERNiCrMo-3 500 780 40
ERNiCr-3 552 > 40
ER309LSi 440 600 41
Table 4
Mechanical properties of filler materials (room temperature)
2.2 Equipment
In the case of robotic processes, a Panasonic VR-006 master controller was used integrated
with a Panasonic YD-350GE2 GMAW power source and with a GTAW power source
Fronius MagicWave 3000. In the case of manual welding, an ESAB Tig2200i power source
was utilized.
The tensile tests were carried out in a MTS 250 kN universal testing machine. Metallographic
observations were performed in a Reichert metallographic microscope and microhardness
scans in an Emcotest durometer.
PWHT and temperature cycles were performed in a Termolab furnace with a resistance
heating system. An Almemo 2690-8 temperature recorder was used to record the temperature
evolution versus time.
2.3 Procedure
This section describes the steps taken to achieve two main purposes of this work:
• To study arc weldability of several combinations of base and filler materials, normally
used or potentially usable in the manufacture of the welded joints under consideration in
this work.
• To analyze the behavior of robotic GMAW welds subjected to temperature cycles in a
furnace in order to simulate their degradation in an accelerated manner.
Firstly, it is important to define the variables which were taken into account in the weldability
study. The identification of the samples is summarized in Table 6.
INDEX çç Sommaire
Base Material 1 Technology Base Material Filler Material Id
10CrMo9-10 M1
Robotic GMAW ERNiCrMo-3
13CrMo4-5 M2
ERNiCrMo-3 T1
10CrMo9-10
Robotic GTAW ERNiCr-3 T2
13CrMo4-5 ERNiCrMo-3 T3
X6CrNi25-20 ENiCrFe-3 S1
10CrMo9-10 ENiCrFe-2 S2
EPRI P87 S3
SMAW
ENiCrFe-3 S4
13CrMo4-5 ENiCrFe-2 S5
EPRI P87 S6
Table 6
Identification of the combinations taken into account in the weldability study
In order to simplify the weldability study, tests were not performed on real components but on
flat butt-welded specimens. The steps which were followed in the manufacture of the
specimens were: preheating (only CrMo steel), welding and PWHT. The shielding gas used in
the processes of gas-shielded arc was pure argon. GTAW electrode had a diameter of 3.2 mm
and the diameters of the shielded electrodes and wires were 2.5 mm and 1.0 mm, respectively.
The parameters, the welding conditions and methods of operation, were adequate to achieve
acceptable joints.
To characterize the welded joints, mechanical testing at room temperature (tensile test and
microhardness scan) and metallographic observations were performed. The dimensions of the
tensile specimens were according to ASME IX. The position of the microhardness tracks is
shown in Figure 2.
Specimens with different materials and conditions were made. Some of them were subjected
to extreme temperature cycles in a furnace. The remaining ones were stored to provide a
comparative basis. The identification of the specimens is summarized in Table 8.
INDEX çç Sommaire
Technology Base Material 1 Base Material 2 Filler Material PWHT Cycles Id
No A1
No
Yes A2
ER 309LSi
No B1
Yes
Yes B2
No C1
No
Yes C2
Robotic GMAW X6CrNi25-20 10CrMo9-10 ERNiCrMo-3
No D1
Yes
Yes D2
No E1
No
Yes E2
ERNiCr-3
No F1
Yes
Yes F2
Table 7
Identification of the combinations taken into account in the analyses of temperature cycles
Just as in the previous case, in order to simplify the analysis, tests were not performed on real
components but on flat butt-welded specimens. The procedure followed in the manufacture of
the specimens for GMAW was the same as in the weldability study.
Maximum and minimum temperatures were fixed in 400 ºC and 800 ºC, respectively. Each
cycle lasted 11 hours and the total length of the process was one week. The actual evolution
of temperature versus time is shown in Figure 3.
Figure 3: Temperature versus time inside the furnace during the temperature cycles
To characterize the welded joints, mechanical testing at room temperature (tensile test and
microhardness scan) and metallographic observations were performed in the same way that in
the weldability study.
INDEX çç Sommaire
3. RESULTS AND ANALYSIS
Due to the large amount of data that were achieved, this section of results and analysis will
focus only on those results that are more significant.
In all the specimens, the fracture during the tensile test occurred in the CrMo steel (base
material). Furthermore, the values achieved for tensile strength were higher than the minimum
required (see Table 2). Therefore, joint efficiencies were always greater than 100%. From the
point view of tensile test, there are no noticeable differences from the use of different
technologies or filler materials. Main results are summarized in Figure 4. The green line
indicates minimum tensile strength value for CrMo steels.
It should be noted in Figure 5 that GTA welds show a considerable decrease of the hardness
values in the molten zone but GMA and SMA welds do not. In addition, this drop is slightly
higher in the case of ERNiCr-3 filler. The heat input involved in GTAW was considerably
INDEX çç Sommaire
higher than in GMAW and SMAW and this is the major reason to explain this phenomenon.
Although metallographic observations did not show significant variations between the molten
zones achieved trough different techniques (it would be necessary to run other techniques),
the microstructures formed in GTAW specimens were softer.
The microstructural evolutions which could be observed were very similar regardless of the
technology. In Figures 7 - 9, it is reflected S3 configuration evolution whose whole
appearance is shown in Figure 6. This will serve as an example of the behaviour in all of the
specimens.
Base material consists of austenite with delta ferrite veins. There are no significant differences
between the HAZ and base material microstructure. In the melting line nearest zone, austenite
grain growth can be seen.
For all fillers, molten zone consists of intermetallic compounds in a matrix of austenite of
columnar dendritic morphology. In the interdendritic areas of the ENiCrFe-3 and EPRI P87
INDEX çç Sommaire
specimens, there are segregation and precipitation of second phases (possibly intermetallic).
The nature of the compounds may vary slightly from one specimen to another but its analyses
would require running other techniques.
It could be observed the following evolution close to the melting line: partial recrystallization
of the ferritic structure in least affected zone, formation of transformation products (acicular
ferrite, martensite and, above all, bainite) with increasing temperature and grain growth with
further increasing temperature. The base material consists of small carbides distributed in an
apparently ferrite matrix.
INDEX çç Sommaire
Figure 10: Temperature cycles analyses - Tensile test
According to Figure 11, the effect of extreme temperature cycles on the hardness values is
remarkable in three areas: molten zone, HAZ and base material of CrMo steel. In stainless
steel (both in the base material and in the HAZ), the dispersion of results is due to large
deformation and elastic recovery during the indent. This deformation has resulted in tracks
with strange morphology affecting the measurements. In the case of the HAZ and base
material of 10CrMo9-10, it should be noted a reduction in hardness values owe to the
dissolution of carbides and grain growth with respect to the basis state. The hardness values of
the closest area to the melting line are especially low and coincide with a decarburized area
evident in Figure 14. Despite these decreases, as it was said, fracture took place in material
base 10CrMo9-10. In the molten zone, it is noted that in the case of ERNiCrMo-3 filler,
hardness values are not only reduced but increased. This fact, certainly, is due to the nature
and intensity of the phenomena described in 3.2.3.
INDEX çç Sommaire
If the effect of tempering provided by the PWHT leads to soft and ductile structures and
reduces the risk of cracking in CrMo HAZ, the temperature cycles causes a deeper softening
of the zone.
"Basis"
The metallographic preparation of the specimens that have not been subjected to temperature
cycles show a similar evolution to that was described in 3.1.3. In the micrographs, it is not
possible to observe a tempering effect of the PWHT over the HAZ of 10CrMo9-10
confirming the results of the microhardness scan. Figure 13 presents micrographs of the
molten zone which will be the basis for comparison with results obtained after temperature
cycles. It could be seen that molten zone consists of intermetallic compounds in a matrix of
austenite of columnar dendritic morphology for the three different filler materials.
INDEX çç Sommaire
Figure 13: Molten zone of B1 (ER309LSi), D1 (ERNiCrMo-3) and F1 (ERNiCr-3)
"Temperature Cycles"
The specimens subjected to temperature cycles exhibit significant changes with regard to the
"basis". However, they have a very similar evolution among them. The general appearance of
the specimen D2 is showed in Figure 14 and its evolution is presented in Figures 15-17.
Firstly, great precipitation takes places in the grain boundary both in the base material and in
the HAZ of stainless steel (see Figure 15 and compare with Figure 7).
INDEX çç Sommaire
Figure 15: Base Material and HAZ - X6CrNi25-20 - D2
In the molten zone (see Figure 16), when comparing with the micrographs of Figure 13, it
should be noticed two phenomena: formation of a new phase in the interdendritic spaces and
precipitation of small particles in the matrix (probably, intermetallic compounds). Such
phenomena are common to the three filler materials with different intensities in each one. The
study of the composition and nature of these compounds would require running other
techniques.
The melting line between the CrMo steel and the molten zone (see Figure 17) is a key area in
dissimilar joints due to a possible occurrence of carbon migration phenomena. Two carbide
morphologies are observed at the interfaces reflected in Figure 17. Type I comprises a planar
array of globular carbides, as illustrated in ERNiCrMo-3 and ERNiCr-3 interfaces. Type II
consists of a diffuse array of smaller carbides. This is the case of the ER309LSi interface. It
had been said that Type I encourages interfacial cracking whereas Type II carbide inhibits [8].
INDEX çç Sommaire
Figure 17: Melting line of B2 (ER309LSi), D2 (ERNiCrMo-3) and F2 (ERNiCr-3)
Finally, regarding 10CrMo9-10, it is possible to observe significant changes. Both in the HAZ
and in the base material, it should be noticed phenomena of grain growth and dissolution of
carbides which justify the reduction in hardness values (see Figure 17). Moreover, abnormal
growth of ferrite grains takes place in the closest areas to the outer surfaces of the plates (top
and bottom) and in the HAZ closer to the melting line. Excessive decarburization of these
areas followed by a very aggressive recrystallization is proposed as a possible explanation for
this phenomenon.
Width measurements of the decarburized zone closer to the melting line (see Figure 19) were
carried out in the samples B2, D2 and F2, in order to look for connexions between this value
and the Cr content in the filler material.
INDEX çç Sommaire
Figure 19: Decarburized zone of B2 (ER309LSi), D2 (ERNiCrMo-3) and F2 (ERNiCr-3)
The width of the decarburized zone closer to the melting line (see Figure 18) is related to the
Cr content in the molten zone. The higher the Cr content (an element that has a great avidity
for C), the wider the decarburized zone. Hence, ERNiCr-3 filler exhibits the narrowest zone.
4. CONCLUSIONS
• In all the specimens, the fracture during the tensile test occurred in the CrMo steel (base
material). From the point view of tensile test, there are no noticeable differences from the
use of different technologies or materials.
• Despite of this fact, GTA welds show a considerable decrease of the hardness values in
the molten zone but GMA and SMA welds do not. Such phenomenon was owing to the
highest heat input involved in GTAW process and affected to greater extent ERNiCr-3
filler.
• A very similar microstructural evolution could be observed in all of the metallographic
preparations carried out.
Taking into account the tests performed in this weldability study, there are no significant
differences to stand out some configuration. Therefore, it is clear that it is necessary to carry
out specific test in order to find key points for characterize these kind of dissimilar joints.
INDEX çç Sommaire
• Formation of a new phase in the interdendritic spaces and precipitation of small particles
in the matrix (probably, intermetallic compounds) in the molten zone. In the case of
ERNiCrMo-3 filler, hardness values are not only reduced but increased because the nature
of these phenomena.
• Substantial reduction of the mechanical properties of 10CrMo9-10. Reduction in hardness
values due to the dissolution of carbides and grain growth with respect to the basis state.
• Abnormal growth of ferrite grains in the closest areas to the outer surfaces of the plates
(top and bottom) and in the HAZ closer to the melting line. Excessive decarburization of
these areas followed by a very aggressive recrystallization is proposed as a possible
explanation for this phenomenon.
5. BIBLIOGRAPHY
[4] “Welding Handbook. Materials and applications. Part 2”, AWS, 1998, 8º Ed.
[5] “Welding Creep Resisting Cr-Mo Steels” G. Hallowey, Metrode Group, 2009.
[6] “Alloy design for similar and dissimilar welding and their behaviours”, H. Heuser y C.
Jochum, Böhler Schweisstechnik Deustschland GmbH.
[7] “Evaluation of Filler Materials for Transition Weld Joints Between Grade 91 to Grade 22
Components”, K. Coleman (EPRI Project Manager), Technical Update, June 2005.
[8] “Dissimilar -Weld Failure Analysis and Development Program. Volume 1: Executive
summary.” EPRI CS-4552, Volume 1, November 1985.
6. AKNOWDLEGMENTS
It wishes to thank Electro Portugal Lda for providing its filler material for free.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page84
RÉSUMÉ / ABSTRACT
39C
PROGRAMME DE LOCALISATION EN CHINE DE LA FABRICATION
D’ÉQUIPEMENTS NUCLÉAIRES : ÉTENDUE ET PERSPECTIVES
RÉSUMÉ
D’ici à 2020, plus de la moitié des centrales électriques construites dans le monde le seront en Chine :
400 GW de centrales charbon, 100 à 150 GW de centrales hydrauliques, 40 GW de centrales éoliennes
et 70 GW de centrales nucléaires. Les objectifs du gouvernement chinois pour le programme nucléaire
sont de localiser 70 à 80% des fabrications d’équipements en Chine.
Mercredi
Cette présentation se veut un tour d’horizon du programme de fabrication des principaux composants
mécaniques des centrales nucléaires en Chine, en abordant :
ABSTRACT
From now to 2020, more than half of the power plants built in the world will be built in China: 400
GW of coal-fire, 100 to 150 GW of hydro-power, 40 GW of wind farms and 70 GW of nuclear power.
The objectives of the Chinese government for the nuclear program are to localise 70 to 80 % of the
equipments manufacturing in China.
This presentation is an overview of the manufacturing program of main mechanical components for
Chinese nuclear power plants, including:
¢
Main Chinese equipments suppliers
¢ European industrials contribution
¢
The influence of Chinese regulations
¢ The G3 program, which involves EPR and AP1000 technologies, oblige the Chinese
suppliers to comply both RCC-M and ASME requirements
INDEX çç Sommaire
INTRODUCTION
In Copenhagen Nov 2009, Chinese prime minister Wen Jiabao declared that China will reduce CO2
emissions per unit of GDP by 40-45% by 2020 from the 2005 level.
To reach the Chinese administration objective to “focus more on renewable energy and nuclear power”,
Zhang Huazhu, president of China Nuclear Energy Association (CNEA) said in April 2010 that,
“China will adjust its nuclear power development program, the target in 2020 could reach 70GW in
operation. We will keep a pace of launching around 8 units per year in the coming years.”
The installed capacity of nuclear power would then represent 5% of total power capacity, which is
estimated to 1500GW in 2020.
Chinese power market has developed rapidly since 2001, as represented on Figure 1. The installed
power capacity reached 874GW at the end of 2009, 10.3% increased compared to 2008. The increasing
rate dropped down slowly, from the highest growth rate 20.3% in 2006, down to 14.8% in 2007 and
11.1% in 2008. The ratio of non-fossil power in total power capacity increased from 22.1% in 2006 to
25.4% in 2009.
100% 1000
900
Total Installation (GW)
80% 800
700
Power Mix
60% 600
500
40% 400
300
20% 200
100
0% 0
2002 2003 2004 2005 2006 2007 2008 2009
Others 0.48 0.55 0.82 1.06 1.87 4.77 10.84 16.13
Nuclear 4.47 6.19 6.84 6.85 6.85 8.85 8.85 8.85
Hydraulic 86.07 94.90 105.24 117.42 128.57 145.26 171.52 196.79
Thermal 265.55 289.77 329.48 391.31 484.05 554.42 601.32 652.05
Total 356.57 391.41 442.38 516.63 621.34 713.30 792.53 873.82
No new NPP was put into operation since August 2007. 11 units of nuclear power reactors are now
under operation in China, representing around 9GW and 1.3% of total installed power capacity. The
total nuclear power generation is 69 TWh in 2009, around 2% of total power generation.
INDEX çç Sommaire
According to the report of National Energy Administration (NEA), no significant nuclear safety events
(≥class 2) occurred in Chinese operating NPPs. The general operation level has been improved: Unit 2
of Ling Ao keeps a record of 935 days of operation without non-planned shut down since its first
critical.
Until June 1st, 2010, 11 NPP projects (representing 30 units) got 30 Approved (23 construction) units
official approval from government. 23 units of them are currently by June 1, 2010 CNNC CGNPC CPI
under construction. The 1st concrete of Nuclear Island was AP1000 2(2) 2(1)
poured for 6 units in 2008 and for 9 units in 2009. The total EPR 2 (2)
capacity of the 30 approved units is about 32 GW. CPR1000 4(4) 16(11)
CNP650 4(3)
9 other NPP projects (representing 21 units) got authorization to
proceed preparation work from National Development
and Reform Commission (NDRC). 4 projects of the 9 are 23 units got authorization from NDRC
by June 1, 2010 CNNC CGNPC CPI Huaneng
expected to be launched in 2010 (Fangchenggang AP1000 4 2 4
CPR1000, Rongcheng HTGR – High Temperature Gas CPR1000 8 2
Reactor, Tianwan phase 3 CPR1000 and Fuqing unit 3-6 HTGR 1
CPR1000).
In 2009, China launched the construction of 9 new nuclear power reactors. 23 units are under
construction in total by June 2010. However, China needs more nuclear power to meet its fast
increasing power demand and to cope with the rising pressures of CO2 emission. According to experts,
by 2020, China’s nuclear power installed capacity could reach 70 GW to 75 GW, and additional
capacity of nuclear power plants under construction 30 GW to 45 GW.
Following the successful development in thermal power field, Chinese power equipment manufactures
showed much progress in nuclear power main equipment manufacturing based on the strong
government support and the prospect of the promising nuclear power market.
Three Generation 3 (G3) projects were launched in 2009, representing 4 AP1000 (Westinghouse
technology) and 2 EPR (Areva technology) units. In United Arabic Emirates, the success of Korean
nuclear industry has been perceived as a signal for Chinese to organize their industry in strong
cooperation and maintain competitiveness in their projects.
In 2010 and 2011, the new projects are most likely to be based on G2+ technology, as the CPR1000 –
an improved model from 900MW French technology used for Daya Bay and Ling Ao.
INDEX çç Sommaire
LOCALISATION OBJECTIVES
For the G2+ and G3 projects, Chinese government asked respectively 80 % and 70 % of localized
equipment manufacturing. Therefore, a certain number of resolutions have been undertaken, resulting
in the following issues:
- Subsidies are granted to Chinese suppliers of critical equipments in order to sustain their
development (forges, big components, I&C, safety classed pumps&valves...),
- Constraints are imposed to foreign suppliers in order to enhance the technology transfers,
- Joint Ventures with Chinese companies are often established with minority foreign investment,
but there are also some whole owned foreign enterprises (WOFE).
- An imposed ratio of localization has become a requirement for the tenders of foreign main
contractors.
The ambitious targets of the government for localization are however not detailed with explanations
regarding the subcontracted activities or raw material procurements to foreign suppliers. In the facts,
localization of G2+ program is expected to encounter the following results, according to China
Machinery Industry Federation (CMIF): 70 % for the units 1&2 of Hongyanhe (CPR1000, First
Operation Date scheduled in 2013), 75 % for Hongyanhe units 3&4 and Ningde project (CPR1000,
First Operation Date scheduled in 2014). For G3 program, all the main components will be supplied
from abroad for the first units of Taishan (EPR), Haiyang and Sanmen (AP1000), and let place to
progressive localization for the second units of these projects.
Apart from the official announcements, the Chinese industry has to cope with the actual lack of own
forging capacities, primary pump casing and steam generator tubes manufacturing qualifications, and
finally with the Chinese customers confidence, more inclined to reliable foreign technologies. This
situation leads, despite the strong willingness to localize and emancipate from the government, to
many contracts signed with European companies, and especially with the historical suppliers of French
nuclear program.
In order to accompany and secure the rapid development of his nuclear program, China established a
set of new laws and regulations. Two of these have a direct influence on localization:
- HAF601 : The Supervision and Management Regulations for Civilian Nuclear Safety Equipment
(manufactured in China)
- HAF604 : The Supervision and Management Regulations for Imported Civilian Nuclear Safety
Equipment
These regulations, under the control of Chinese Nuclear Safety Authority (NNSA), concern all
activities related to design, manufacturing, installation and non destructive testing on safety classed
equipments.
Now, the constructors have to obtain a manufacturing (and/or design) license for nuclear classed
equipments manufactured in China (HAF 601) or a license of importation (HAF604) for equipments
manufactured abroad. So far, there are 140 Chinese suppliers who have obtained NNSA HAF 601
(design 27% / manufacture 61% / installation 10% / NDT 2%) certificates, for only 100 foreign
suppliers getting HAF 604 certificates.
INDEX çç Sommaire
The main process of NNSA HAF 601 certification is as following:
- Submit documents (regulation, technical specification, mock-up manufacture plan, etc.) –
Witness of mock-up manufacture and tests – Evaluation and review – Internal meeting within
NNSA relevant units – Experts Committee Review – Decision of Equipment Department of
NNSA - Announcement to public – Approval and certificate delivery.
- Such certification process takes from several months to around 2 years, depending on the type of
equipment.
HAF601 is actually the oldest qualification required for the components designed and manufactured in
China. From 2004, 556 certifications were granted, with a peak in certifications in 2006, mainly
associated with the Ling Ao nuclear power plant project.
Three types of companies can get HAF601 certification: local companies, joint-venture companies,
and wholly-owned foreign enterprises. In the last two categories, mostly represented are the
components which are manufactured in China, but whose design has been transferred from foreign-
based headquarters to the JV or to the local branch.
Since January 2008, date of the promulgation of HAF 604, 100 foreign companies have obtained their
importation certification; among these 40% of French companies – see figure 2. This strong presence
of French companies is not only due to the French origins of the G2+ technology, but also to the
stimulation of partnerships within French associations, like the “Partenariat France-Chine Electricité”
(PFCE).
USA
12%
Germany
18%
The HAF604 market is expected to grow smoothly in China. As the regulation on nuclear industry is
increasing year after year, the NNSA will surely enlarge the list of equipment components requiring a
certification. While the local industry players are not as experienced as their European, American,
Japanese and Korean counterparts, aggressive investments are still being made in every field of the
nuclear power industry. An example of this would be the expansion of Class 1 (and 1E for electrical)
equipment qualifications in China over the last two years, and the successful partial localization of
production for the Ling Ao power plant.
INDEX çç Sommaire
MAIN SUPPLIERS OF NUCLEAR EQUIPMENTS
Equipment manufacturing localization is a key element of the Chinese huge nuclear program. Within
the past few months, several technical break-through of main component fabrication have been
considered as milestones in Chinese nuclear equipment localization program. However, the main
nuclear equipments manufacturers are still looking for stable reliability in following Chinese supplies:
- Forging pieces: that could meet the needs of CPR1000 projects, but not ready for mass production
of forging for G3, such as AP1000 and EPR. The qualification on G3 forging is still in progress.
- Steam generator tubes: the process of localization is on the tracks, notably with a 100% Chinese
JV (Baoyin) which is testing his first batch in June 2010.
- Reactor cooling pumps: no nuclear power primary pumps are realized locally except for the first
pump delivered by the DEC-AREVA JV in Dec 2009 to Ling Ao II project.
There are three main Nuclear Steam Supply System (NSSS) forging makers in China: China First
Heavy Industry (CFHI), China Erzhong heavy industry (CEZ) and Shanghai Heavy Machinery (SHM).
The imported forgings represent about 40 % for Reactor Pressure Vessel (RPV) and Steam Generators
(SG) and come mainly from Japan (JSW), Korea (Doosan) and France (ANP Creusot Forges), the rest
would be delivered by Chinese forgemasters. Until now, CFHI and CEZ have both performed the
qualification of the forging process (M140 qualification according to RCCM) for the RPV and SG of
G2+ projects.
The three forging makers have enlarged their facilities for making the pieces of G2+, EPR and AP1000,
and increased the biggest ingot to 600 tons and maximum diameter to 7 meters, which could meet the
needs of the main forged pieces.
The main figures of forging facilities of the three forging maker.
molten metal Max Ingo. /Max Forge Press force Size (m)
(ton) (ton) (ton) (wide x height)
CFHI 700 600/400 15000 7x8 (in Oct. 2006)
CEZ 700 600/400 16000 7x7.7 (in Oct. 2007)
SHM 720 600/400 16500 7.5x8 (in 2009)
The Inconel steam generator tubes are a critical point for the SG manufacturing. Until now, China
imported 100 % of the tubes: Valinox (France) and Sandvick (Sweden) are the main suppliers for G2+
and EPR, whereas Sumitomo (Japan) will supply for the AP1000 units.
The localization of the steam generator tubes manufacturing is on the track but will take time:
- Baosteel has set up a new factory in Jiangsu: Baoyin (the first tests for NNSA qualification are
scheduled in May-June 2010)
- Pan Steel has set up a factory in Sichuan : Panchang Steel
- Sanzhou planned the construction of a new factory to end in 2012
INDEX çç Sommaire
Main components manufacturing
Four suppliers are capable of fabricating main NSSS components (Reactor Pressure Vessel – RPV,
Steam Generator – SG, Pressurizer – PRZ): Dongfang Electric Corporation (DEC), Shanghai Electric
Corporation (SEC), Harbin Electric Corporation (HEC) and CFHI.
Hereunder are indicated the main progress for nuclear equipment manufacturing of these huge groups:
DEC:
· Obtained ASME N and NPT stamps in July 2009.
· So far, 6 SG and 1 RPV have been delivered for Ling Ao phase 2 (2008 and 2009).
SEC:
· Obtained ASME N and NPT stamps in 2008.
· So far, 1 RPV of Qinshan phase 2 (650MW unit) in 2001 and 3 SG of Qinshan phase 2-extension
(650MW unit) in 2009 have been delivered.
CFHI & HEC:
· These two are associated with each other for nuclear manufacturing, HEC focuses on SG, PRZ, and
CFHI focuses on RPV.
· CFHI realized the RPV of Pakistan Chashma 2 (300MW unit) in 2008 and Qinshan 2-extension in
2010.
· HEC has reference in Pakistan Chashma 2, but no references in 650MW class.
For the time being, the capacity of Chinese NSSS equipment suppliers are about to meet the needs of
domestic G2+ projects. But they are still in the starting stage of G3 equipment fabrication. In the
coming 3-5 years, it would be possible for them to meet the needs of G2+ and G3 domestic projects.
The manufacturing localization of primary pumps relies mainly on collaborations between Chinese and
foreign companies:
- DEC/AREVA: this JV (50/50) will supply most of the G2+ projects through the assembling of the
pumps
- HEC/Andritz: the technology transfer will enable the localized manufacturing, under a foreign
license, of primary pumps for G2+ projects Fuqing and Fangjiashan
- SEC/KSB: the JV (55/45) has been signed recently, but the manufacturing of primary pumps
doesn’t seem to be a priority
- Shenyang Pump/EMD (Curtiss Wright) was designated by NDRC as receiver of AP1000
technology transfer
For the G3, up to now all the Taishan EPR and AP1000 reactor cooling pump casing will be provided
from abroad.
INDEX çç Sommaire
CODES & STANDARDS
Apart from Chinese specific standards (Material, Mechanical testing, Non Destructive Tests…),
applied in common field of manufacturing, two main foreign codes are applied in China for design and
manufacturing of nuclear safety classed equipments:
- The French Code RCC-M (edited by AFCEN), used for all the Chinese Nuclear Power Plant of
CPR1000 type (Daya Bay, Ling Ao, Qinshan 2…), and now for Taishan EPR project. An
agreement has been signed between AFCEN and CGNPC to translate and publish a Chinese
translation of the RCC rules: the translation will be published within 2010.
- The American Code ASME, used for Qinshan 1&3 and now for AP1000 projects, is widely used
by the Chinese manufacturers for the construction of the huge Chinese coal-fire power plants. The
code is also well known (at least for the non nuclear sections), and ASME III already published a
Chinese official version in 2007.
In parallel with the regulation implementation, China established the target to build their own design
and construction code for nuclear equipments, aiming to issue it before the emergence of Chinese own
G3 technology. The process has already started with the nomination of working groups within the
Chinese industry.
Based on ASME and RCC requirements, the common part between them would be used directly; for
the distinguished parts, comparison and study would be performed before making the decision, so that
the final version would be between ASME and RCC, with the potential application of GB (Chinese)
standards. NNSA will endorse the final version.
CONCLUSION
The planning of the Chinese nuclear program seems tight in regard with the actual capacities of
Chinese forgemasters and the ambitious localization ratio requested by the government.
There are still some national officials and experts doubting in front of the huge nuclear power program,
due to concerns on the risks and potential bottlenecks of nuclear safety, construction quality, and
human resources, during the fast development. It cannot be denied that the lack of trained people, the
problematic maintenance of manufacturing tools and the bottleneck situations within the workshops
could lead to delays in the projects.
However, according to the current project development status, China could have around 40 GW of
NPPs in operation by 2015. If China keeps the pace of launching 6~8 new units each year before 2015,
even with some years of delay, the 2020 target of having 70 to 80 GW of NPPs in operation would be
reached.
The recent years have seen a remarkable development of the Chinese nuclear industry, with the
strengthening of the regulation and the attribution of many key components contracts to the Chinese
manufacturers. Even with the support of the government and the construction of high-tech impressive
workshops, the success of the “made in China” nuclear components remains a challenge, and leaves to
foreign qualified suppliers place for opportunities.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page85
RÉSUMÉ / ABSTRACT
63C
SIMULATION NUMÉRIQUE DU SOUDAGE :
LE PROJET MULTIPARTENAIRE MUSICA
5Institut de soudure, Espace Cormontaigne, 4, Bld. Henri Becquerel, 59970 Yutz, France
RÉSUMÉ
Le projet MUSICA a tenu ses promesses en débouchant sur des avancées très significatives.
Ce projet multipartenaires regroupant AREVA, le CEA, le CETIM, ESI Group et l’Institut de Soudure, sous la coordination du CEA, avait
un triple objectif : industriel en cherchant à remplacer progressivement en grande partie les essais de soudage par la simulation, scientifique
en intégrant des modèles physiques depuis le procédé de soudage à l’arc jusqu’aux transformations métallurgiques, et logiciel avec le besoin
d’un outil complet et opérationnel pour la simulation numérique du soudage.
L’idée force du consortium MUSICA aura été de créer un centre de compétences s’appuyant sur des savoir-faire polyvalents. Les principaux
résultats sont :
- la chaine logicielle MUSICA démontre la faisabilité de la réalisation d’outils accessible aux non experts pour la simulation du soudage
depuis le procédé jusqu’à ses conséquences (distorsions, contraintes résiduelles, impact sur la durée de vie),
- le gain de productivité grâce à l’amélioration continue des outils (ergonomie, robustesse, calculs distribués, maillage automatique, post
traitement dédié, mise en données simplifiées, assistant soudage multipasse…),
- la validation des outils logiciels développés par mise en application sur 3 cas tests industriels,
- deux guides méthodologiques renfermant les bonnes pratiques retenues au cours du projet pour mettre en œuvre les outils logiciels, pour
sélectionner les modèles procédés et matériaux adaptés aux applications, pour valider les calculs et réaliser des mesures.
Cette présentation est un focus sur les apports de MUSICA dans la simulation de la réalisation des assemblages, avec une démonstration des
fonctionnalités de la chaine d’outil logiciel MUSICA développé en réponse aux attentes industrielles.
ABSTRACT
Project MUSICA held its promises while leading to very significant advances.
This multipartners project, gathering together AREVA, the CEA, the CETIM, ESI Group and the Institut de Soudure, under the coordination
of the CEA, had a triple goal: industrial by trying to replace gradually a large part of welding tests by the simulation, scientific by integrating
physical models from the fusion of metal up to its solidification and numerical with the need of a full and operational tool for the numerical
simulation of welding.
The force of the consortium MUSICA will have been to create center of excellence based on general-purpose know-how. The main benefits are:
- MUSICA software chain, that demonstrates the feasibility of the realization of tools accessible to the non experts for the numerical welding
simulation, starting from the welding process to its consequences (distortions, residual stresses, design life),
- the gain of productivity according to the continuous improvement of tools (ergonomics, robustness, distributed calculations, automatic
meshing, automatic geometric tool, automatic post processing, multipass),
- the validation of the developed software tools by the simulation of 3 industrial tests provided by the partners,
- two guides containing the good practices selected during the project to operate software tools, to select the process models and the materials
models according to the applications, to validate the simulations and to perform measurements.
This paper focuses on the contributions of MUSICA for the welding simulation of assembly. It includes a demonstration of the features of
the chain of software tool MUSICA developed according to industrial expectations.
AFIAP - ESOPE 2010 Programme
INDEX çç Sommaire
SIMULATION NUMERIQUE DU SOUDAGE : LE PROJET MULTIPARTENAIRE
MUSICA
1
CEA, DEN, DM2S, SEMT, LTA, F-91191 Gif-sur-Yvette, France. [email protected]
2
CETIM (Centre Technique des Industries Mécaniques) – 52, Avenue Félix Louat, BP 80067,
60300 Senlis , France
3
ESI-group, Le Récamier, 70, rue Robert, 69458 Lyon Cedex 06
4
AREVA NP Paris, France
5
Institut de soudure, Espace Cormontaigne, 4, Bld. Henri Becquerel, 59970 Yutz, France
RESUME
Le projet MUSICA a tenu ses promesses en débouchant sur des avancées très
significatives.
Ce projet multipartenaire regroupant AREVA, le CEA, le CETIM, ESI Group et l'Institut de
Soudure, sous la coordination du CEA, avait un triple objectif : industriel en cherchant à
remplacer progressivement en grande partie les essais de soudage par la simulation,
scientifique en intégrant des modèles physiques depuis le procédé de soudage à l’arc
jusqu’aux transformations métallurgiques, et logiciel avec le besoin d'un outil complet et
opérationnel pour la simulation numérique du soudage.
L'idée force du consortium MUSICA aura été de créer un centre de compétences s'appuyant
sur des savoir-faire polyvalents. Les principaux résultats sont :
- la chaine logicielle MUSICA démontre la faisabilité de la réalisation d'outils
accessible aux non experts pour la simulation du soudage depuis le procédé jusqu'à ses
conséquences (distorsions, contraintes résiduelles, impact sur la durée de vie),
- le gain de productivité grâce à l'amélioration continue des outils (ergonomie,
robustesse, calculs distribués, maillage automatique, post traitement dédié, mise en
données simplifiées, assistant soudage multipasse…),
- la validation des outils logiciels développés par mise en application sur 3 cas tests
industriels,
- deux guides méthodologiques renfermant les bonnes pratiques retenues au cours du
projet pour mettre en œuvre les outils logiciels, pour sélectionner les modèles
procédés et matériaux adaptés aux applications, pour valider les calculs et réaliser des
mesures.
Cette présentation est un focus sur les apports de MUSICA dans la simulation de la réalisation
des assemblages, avec une démonstration des fonctionnalités de la chaine d'outil logiciel
MUSICA développé en réponse aux attentes industrielles.
ABSTRACT
Project MUSICA held its promises while leading to very significant advances.
INDEX çç Sommaire
This multipartners project, gathering together AREVA, the CEA, the CETIM, ESI Group and
the Institut de Soudure, under the coordination of the CEA, had a triple goal: industrial by
trying to replace gradually a large part of welding tests by the simulation, scientific by
integrating physical models from the fusion of metal up to its solidification and numerical
with the need of a full and operational tool for the numerical simulation of welding.
The force of the consortium MUSICA will have been to create center of excellence based on
general-purpose know-how. The main benefits are:
- MUSICA software chain, that demonstrates the feasibility of the realization of tools
accessible to the non experts for the numerical welding simulation, starting from the
welding process to its consequences (distortions, residual stresses, design life),
- the gain of productivity according to the continuous improvement of tools
(ergonomics, robustness, distributed calculations, automatic meshing, automatic
geometric tool, automatic post processing, multipass),
- the validation of the developed software tools by the simulation of 3 industrial tests
provided by the partners,
- two guides containing the good practices selected during the project to operate
software tools, to select the process models and the materials models according to the
applications, to validate the simulations and to perform measurements.
This paper focuses on the contributions of MUSICA for the welding simulation of assembly.
It includes a demonstration of the features of the chain of software tool MUSICA developed
according to industrial expectations.
.
INDEX çç Sommaire
PRESENTATION GENERALE
Démarche
afin d'aboutir en 3 ans à une première version d'outils logiciels industriels et opérationnels
permettant de simuler le soudage avec une bonne prédictibilité.
INDEX çç Sommaire
Avancées majeures et apports
INDEX çç Sommaire
Figure 1 : Schéma de principe de la chaine logicielle MUSICA [1].
La chaine logicielle MUSICA comporte principalement trois étapes, et trois logiciels associés.
L'un des intérêts majeurs de la simulation numérique du soudage est la prédiction des
contraintes résiduelles et des déformations résultants de l'opération de soudage. Ces effets
thermomécaniques sont pilotés par les évolutions de températures. Pour pouvoir réaliser une
bonne prédiction des effets thermomécaniques il est nécessaire de correctement simuler les
évolutions thermiques dans la pièce soudée. Il est considéré deux approches pour modéliser
l’apport de chaleur l'une : dite « multiphysique » (modélisation in extenso du procédé) et
l'autre qualifiée de « source de chaleur équivalente » (en fait : une fonction mathématique
spatio-temporelle dont la forme est postulée et les paramètres sont identifiés a posteriori sur
une expérience dédiée).
Par exemple, dans le cas du soudage à l'arc, l’approche « multiphysique » consiste à résoudre
un problème magnétohydrodynamique couplant l'électrode, le plasma d'arc, son interaction
avec le bain de soudage, le gaz de protection, le transfert de l’énergie dans ce bain, la
thermohydraulique du bain de soudage et la diffusion d'énergie dans le solide. Cette approche
reste très complexe car elle nécessite de décrire l’apport de matière, l’interaction arc /
matériau, les mouvements de bain,…etc. Néanmoins, un modèle 2D multiphysique et un
modèle 3D thermohydraulique sont déjà disponibles dans le logiciel WPROCESS développé
par le CEA sur la base du logiciel Cast3M [2].
INDEX çç Sommaire
WPROCESS est un logiciel métier destiné à la simulation du procédé de soudage, il
est basé sur le code éléments finis Cast3M, utilise des routines python et est intégré dans la
plateforme open source SALOME. WPROCESS a pour objet le traitement et l’analyse d’une
expérience de soudage, la détermination du chargement thermique (avec pour objectif
d’utiliser ce chargement en vue de mener des calculs thermomécaniques), la simulation
thermique du procédé de soudage.
Le traitement et l’analyse d’une expérience de soudage consiste en des outils permettant de
réaliser une conception optimale de l'expérience, l'analyse de sensibilité du modèle de
soudage, et de simplifier et limiter les actions utilisateurs. Le logiciel WPROCESS est
capable de sélectionner les résultats des thermocouples les plus pertinents pour l'analyse. De
plus, cet outil offre à l'utilisateur la possibilité d'effectuer une « conception optimale
d'expérience ». A partir d'une configuration donnée, le logiciel est capable de calculer la
position optimale de thermocouples qui seront à même de permettre une identification des
paramètres de la « source de chaleur équivalente ».
La détermination du chargement thermique est réalisée de deux manières :
1. indirecte et empirique, le chargement thermique est représenté par une source de chaleur
(c.a.d distribution spatiale de la densité volumique d’énergie apportée par le procédé de
soudage) dont les paramètres sont identifiés par méthode inverse en cherchant à minimiser
l’écart entre mesures et simulations, et l’utilisateur dispose de données expérimentales
(par exemple une macrographie donnant une idée des dimensions du bain fondu et/ou des
thermogrammes donnant les évolutions temporelles et les gradients de températures).
2. directe et a priori, l’utilisateur ne dispose pas de données expérimentales et dans ce cas la
source de chaleur est déterminée de deux façons :
a. par un module en fonction du type de procédé, de ses paramètres et de la largeur
estimée de la zone fondue,
b. par une simulation multiphysique du procédé qui calcule le flux de chaleur.
La simulation multiphysique des procédés de soudage est réalisée par une modélisation
magnétohydrodynamique du procédé et de son interaction avec la pièce en vue de mener un
calcul prédictif (soudabilité, forme précise de cordon, risque d’apparition de défauts, champ
thermique).
Les fonctionnalités de WPROCESS actuellement disponibles sont :
1. la détermination de chargement thermique pour les procédés de soudage à l’arc TIG et
MIG/MAG, les procédés laser et faisceaux d’électrons sur une géométrie de plaque et de
Té en assemblage bord à bord.
2. simulation thermohydraulique du bain fondu sur une géométrie de plaque en assemblage
bord à bord et un Té pour les procédés de soudage à l’arc TIG et MIG/MAG, les procédés
laser et faisceaux d’électrons une source de chaleur est utilisée pour modéliser l’apport
d’énergie (figure 2).
3. simulation multiphysique couplée 2D axisymétrique, de l’électrode, du plasma d’arc, et du
bain fondu pour l’application TIG spot sur une géométrie disque (figure 3).
4. La conception optimale d’expérience de soudage
5. L’analyse de sensibilités des modèles thermiques de simulation de soudage
A terme, le module multiphysique permettra à l’industriel de mener des calculs prédictifs
directement à partir des paramètres opératoires. Ce qui permettra de réduire les coûts et les
délais en substituant progressivement la simulation aux essais de soudage. Ce module
permettra aussi d’évaluer rapidement les niveaux optimisés des paramètres de fabrication et
de définir une stratégie de soudage multiobjectif : limiter et/ou déporter les contraintes
résiduelles et/ou limiter les distorsions et/ou augmenter la pénétration et/ou améliorer la forme
du raccordement…) ; ceci par exemple dans le cas de la définition de modes opératoires de
INDEX çç Sommaire
soudage ou la qualification de procédés sur des maquettes ou la définition de nouveaux
procédés.
Pour cela, l'utilisateur doit simplement sélectionner une géométrie de structure et ses
dimensions, un procédé, un matériau, les paramètres de soudage et renseigner un fichier
contenant des cycles de températures obtenus par thermocouples et/ou la forme de la zone
fondue tirée d’une macrographie. Le modèle aux éléments finis et la mise en données sont
automatiquement générés par WPROCESS. En sortie, on obtient les paramètres de la «
source thermique équivalente », une cartographie des températures ainsi qu'une comparaison
visuelle des cycles thermiques et des zones fondues expérimentaux et numériques.
La source thermique identifiée avec WPROCESS peut servir ensuite d'entrée à une simulation
de soudage classique d'une structure complète à l'aide du logiciel SYSWELD (contraintes
résiduelles et déformations) ou enrichir une base de données interne au logiciel SYSWELD.
Deux IHM ont été développés pour WPROCESS : l'une par ESI Group intégrée dans visual
environnement pour un chainage avec SYSWELD sous une même interface et l'autre par le
CEA (sous SALOME) pour exploiter le logiciel en "stand alone" et le chainer avec n'importe
quel code thermomécanique (Cast3M, Abaqus Sysweld, Aster,…).
Dans la suite WPROCESS sera incrémenté par un modèle 3D pour le soudage à l’arc avec
apport de matière (Thermohydraulique + MHD simplifiée pour l’arc) et par de nouvelles
configurations (chanfrein, multipasses,…).
Buse
Electrode
Plasma d’arc
Bain
Fondu
INDEX çç Sommaire
Figure 3 Simulation magnétohydrodynamique couplée électrode-arc-bain de fusion (TIG).
Les principes d'une simulation de soudage transitoire classique avec SYSWELD ont
déjà été largement présentés dans la littérature, notamment par Bergheau [3]. Ce logiciel aux
éléments finis, totalement dédié au domaine du soudage, permet de prédire les contraintes et
déformations résiduelles, tout en prenant en compte les phénomènes thermiques,
métallurgiques et mécaniques. SYSWELD dispose d'une large gamme de méthodes de
simulation dont certaines ont été améliorées dans le cadre du projet MUSICA. Ces techniques
plus ou moins automatisées, permettent de minimiser les temps de mise en données et de
calcul, comme par exemple les techniques de « macro-dépôt » pour le soudage multi-passe de
fortes épaisseurs, ou les méthodes « coque-solide » [4]. Malgré tout, ce type de calcul est
surtout utilisé pour simuler le soudage de pièces de taille relativement modeste. En effet,
malgré les progrès significatifs apparus ces dix dernières années que ce soit du côté des
développeurs de logiciels comme ESI-Group ou des fabricants de matériel informatique, il
n'est pas encore possible de réaliser une simulation de soudage d'une très grande structure en
utilisant une approche classique dans des temps industriellement compatibles. Par contre, la
méthode « local / global » qui a été automatisée et intégrée à la chaine logicielle MUSICA,
permet de prédire les distorsions résiduelles dues au soudage d'assemblages imposants
comportant un grand nombre de cordons de soudures.
Dans le cadre d'une approche « local / global », l'outil VLMA est dédié à la création
du modèle « local ». L’idée est de ne pas simuler l'ensemble de la structure soudée, mais
seulement une portion de celle-ci. Les calculs sont basés sur une approche transitoire
classique. Le calcul sur une partie réduite de la zone soudée se justifie par le fait que les
déformations plastiques se situent aux abords du joint, et ne dépendent que des conditions
thermiques, métallurgiques et mécaniques locales. Ces déformations plastiques seront ensuite
réinjectées dans un modèle global représentatif de l’ensemble de la structure via l'outil Pam-
Assembly.
L'un des avantages de l'outil VLMA provient de sa simplicité d'utilisation, ne nécessitant pas
de connaissances approfondies en calculs aux éléments finis. Comme pour l'outil
WPROCESS, l'utilisateur peut se contenter de choisir la géométrie qu'il souhaite traiter,
sélectionner un matériau, etc…La discrétisation de la pièce, l'ensemble de la mise en données
ainsi que les calculs sont automatiquement gérés par l'outil VLMA. Les calculs sont rapides,
puisque l'on ne simule qu'une portion de structure.
INDEX çç Sommaire
L'autre avantage de cet outil réside dans sa complémentarité avec le logiciel Pam-Assembly.
En effet, il existe un lien direct entre ces deux logiciels, permettant de mettre en œuvre une
simulation de type « local / global » de façon simple et automatisée. Ceci permet un gain de
temps considérable, car l'étape qui consiste à réinjecter les déformations plastiques calculées
sur le modèle « local » (à base d'éléments finis volumiques) dans le modèle « global » (à base
d'éléments finis de type coques) peut être fastidieuse à réaliser manuellement. De plus, tous
les modèles locaux réalisés via l'outil VLMA peuvent être insérés automatiquement dans une
base de données 'modèles locaux' interne à Pam-Assembly.
CAS TESTS
Les outils logiciels ont été développés autour de trois cas tests industriels proposés par les
partenaires.
Cas-test « AREVA »
INDEX çç Sommaire
Le calcul a demandé d’importants développements pour le maillage du joint soudé et a
nécessité le cofinancement par ESI et AREVA d’une version spéciale parallélisée de
SYSWELD. Néanmoins la modélisation n’a pas pu être réalisée en continu : c’est une
technique de dépôts « en bloc » de 4 cordons le long de la circonférence qui a été retenue
(figure 5). Ceci a sans doute pour effet de réduire les déformations cumulées.
La réalisation de la maquette a fait appel à des techniques performantes pour la mesure
des températures et les distorsions. Les températures ont été mesurées finement par la mise en
place de nombreux thermocouples à proximité du chanfrein par le LIMATB de l’Université
de Bretagne Sud. Ceci a permis de comprendre la répartition des flux de chaleur dans le
chanfrein étroit. Les mesures de déplacement de la maquette ont été réalisées par
photogrammétrie, technique utilisée par AREVA pour le montage de gros composants (figure
6).
INDEX çç Sommaire
Figure 6 Comparaison de la flèche mesurée et calculée en extrémité de tube.
Cas-test « CEA »
Cas-test « CETIM »
Le cas-test CETIM est un assemblage en Té, constitué d'une semelle et d'une âme
soudées avec 2 soudures d'angle. L'âme et la semelle sont en acier de nuance S355NL2,
soudées par procédé MAG robotisé avec produit d'apport G3Si1. Les deux passes de soudage
ont été réalisées en sens opposé. Des cas dérivés de celui-ci ont également été traités, avec
INDEX çç Sommaire
une modification du sens de soudage de la deuxième passe, des conditions de bridage plus
importantes ou une vitesse de soudage modifiée. Pour les besoins du projet MUSICA, plus
d'une dizaine de maquettes ont été réalisées au sein du site Assemblage et Matériaux de
l'Institut de Soudure à Yutz. L'intérêt des cas-tests est de pouvoir confronter résultats de
calculs et de mesures, afin de valider les outils développés dans le cadre du projet MUSICA.
Sur ce cas-test ont été effectuées des mesures de températures, de contraintes résiduelles ainsi
que des mesures de déplacements. La chaine logicielle MUSICA a été appliquée sur cet
assemblage en Té, afin de prédire numériquement les déformations et contraintes résiduelles
engendrées.
Différentes « sources thermiques numériques équivalentes » ont été testées. Certaines ont été
calibrées manuellement à l'aide du logiciel SYSWELD, d'autres ont été calibrées en utilisant
l'outil WPROCESS. Pour calibrer ces sources, l’utilisateur pouvait avoir accès à différents
niveaux d'informations, plus ou moins précises. Suivant le nombre et la pertinence des
informations ayant servi à la calibration, les cycles thermiques engendrés par les « sources
thermiques numériques équivalentes » sont très différents. Les résultats numériques les plus
proches des mesures expérimentales sont bien sûr obtenus avec les sources issues de
l'ensemble des informations disponibles.
Le cas-test en Té a été simulé par calculs « classiques » transitoires thermo-métallo-
mécaniques (figure 9) et en utilisant la méthode « local / global » (figure 10). Des
comparaisons entre les déformations mesurées, calculées par méthode « classique » transitoire
ou par méthode « local / global » ont montré des écarts inférieurs à 10% sur la flèche
maximale. Quant aux contraintes résiduelles obtenues par calculs « classiques » transitoires,
l'écart mesures / calculs varie entre 25 à 100 MPa, avec des profils similaires.
Pour réaliser l’analyse à la fatigue, nous avons établi un chainage logiciel entre SYSWELD
pour déterminer les contraintes résiduelles et ensuite des calculs éléments finis élasto-
plastique pour calculer les contraintes et déformations en enfin DesignLife pour évaluer la
durée de vie. Les méthodes utilisées pour le calcul de fatigue sont basées sur une approche
plane critique en contraintes / déformations normales. Différentes stratégies de calcul ont été
utilisées:
En définissant le cycle de chargement de fatigue dans SYSWELD: il s’agit d’un calcul
élastoplastique dont le premier pas est l’état de contraintes résiduelles. C’est
l’approche la plus précise mais la plus coûteuse en temps calculs.
En effectuant un calcul statique linéaire pour l’effort appliqué puis en superposant ce
résultat à celui des contraintes résiduelles avant éventuelle correction élastoplastique.
Cette méthode approchée a comme avantage la mise en œuvre facile et la plus rapide.
Les résultats des différentes simulations montrent l’amélioration de la tenue en fatigue lorsque
le rayon de raccordement au pied du cordon augmente. Ceci ne fait que confirmer
l’importance de la qualité de la soudure sur la résistance en fatigue des assemblages soudés.
Les figures 11, 12 et 13 présentent le cordon de soudure, la distribution des contraintes
résiduelles et l’évaluation de la durée de vie pour deux rayons de raccordement d’un
assemblage soudé (cas test Cetim).
INDEX çç Sommaire
Figure 13 Durée de vie pour deux rayons de raccordement.
GUIDES METHODOLOGIQUES
Ce guide, d’une centaine de pages, est composé de deux chapitres l’un décrivant les
méthodes d'instrumentation et l’exploitation d’une expérience de soudage et l’autre la
méthodologie de mise en œuvre des outils MUSICA.
INDEX çç Sommaire
Méthodologie de mise en œuvre des outils MUSICA.
La méthodologie complète de mise en œuvre des outils MUSICA est présentée et mise
en application sur les trois cas tests réalisés au cours du projet.
Les méthodes disponibles leurs intérêts et limites, le problème du calage de la source de
chaleur, le choix du modèle de comportement mécanique, la création des cordons de soudure,
la prise en compte des caractéristiques physiques et les points à considérer pour la génération
du maillage, discrétisation, et conditions aux limites, sont détaillés.
Des résultats comme l’influence des sources de chaleur, de la loi d’écrouissage sur les
contraintes et déformations résiduelles, du type de simulation (la méthode des macros dépôts
en énergie, la méthode des macros dépôts en température, la méthode « Local-Global », la
méthode « pas à pas ») sur les déformations et contraintes résiduelles qui ont été tirés des
résultats obtenus par les calculs des trois cas tests industriels proposés par les partenaires, sont
donnés.
Ce guide d’une cinquantaine de pages est composé de trois chapitres, les deux
premiers concernent la simulation thermométallurgique et thermomécanique des contraintes
résiduelles et distorsions de soudage et le troisième la simulation multiphysique du procédé de
soudage.
Simulation multiphysique
Cette partie concerne les différents couplages multiphysiques dans le bain et son
atmosphère environnante à prendre en compte dans une modélisation in extenso du procédé de
soudage à l’arc. Le procédé de soudage à l’arc et une revue des modèles physiques de la
littérature sont présentés. Les différentes zones caractéristiques de l’arc de soudage TIG
comme le la colonne d’arc les zones cathodiques et anodiques ; les modèles de cathode, de
plasma d’arc de bain de soudage de l’interface plasma/cathode et de l’interface plasma/anode
définissant l’état de l’art pour la modélisation du soudage à l’arc sont décrits. Les avancées
majeures dans la modélisation de l’arc ou du bain.
Pour terminer une revue, tirée d’expériences, de l’influence de quelques paramètres de
soudage, (l’intensité du courant, la hauteur d’arc et l’angle d’affûtage de l’électrode) sur
l’énergie transmise à la pièce, sur le rendement de l’arc et sur la géométrie du bain.
INDEX çç Sommaire
CONCLUSIONS ET PERSPECTIVES
BIBLIOGRAPHIE
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page86
100E Anomalies de pose de bouchons mécaniques dans les tubes des MORIN A. / COLONNA F.
générateurs de vapeur des réacteurs nucléaires d’électricité de LIMOUSIN S.
Mercredi
France (EDF) / Plug installation faults on EDF nuclear reactor steam ASN-DEP
generator tubes
69E Application de l’inspection basée sur criticité en Chine / Risk based XIAYI Q. / XIN M.
inspection implementation in China YING ZHI W.
JIANGSU SPECIAL
EQUIPEMENT SAFETY
SUPERVISION INSPECTION
INSTITUTE
TOPALIS P. / HAN O.
DNV
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page87
12E Gestion et suivi en service des équipements sous pression fixes / OSAGE D. A.
Life cycle management of pressurized fixed equipment THORNTON D. R.
HENRY P. A.
THE EQUITY ENGINEERING
GROUP, Inc
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page88
RÉSUMÉ / ABSTRACT
21/22E
EXEMPLE DE FISSURATIONS D’ÉQUIPEMENTS D’INSTALLATIONS
PÉTROCHIMIQUES
D. Goursaud / F. Clément
GPN
RÉSUMÉ
Cet article présente le retour d’expérience sur deux fissurations apparues sur un corps de chaudière
Mercredi
d’acide nitrique et sur un réacteur de méthanation. La présentation met en avant les différentes solutions
de réparations envisagées, les solutions retenues à la fois d’un point vu technique et réglementaires.
L’objet de cet article est de mettre en avant les avantages d’un service inspection retenu, notamment
par la mise en oeuvre de plans d’inspection basé sur la criticité.
INDEX çç Sommaire
ESOPE 2010 : Exemple de fissurations d'équipements d'installations pétrochimiques
L’atelier AN5 est un atelier de fabrication d’acide nitrique. Cette fabrication est réalisée
selon les trois étapes suivantes :
L’incident présenté est survenu, lors de la phase de démarrage de l’atelier, sur le corps
de chaudière de récupération. Ce corps est utilisé pour réaliser la phase de combustion
décrite ci-dessous.
Le corps de chaudière est revêtu intérieurement par du béton réfractaire d’une épaisseur
de 250 mm environ. Extérieurement un serpentin de réchauffage vapeur recouvre le
corps de chaudière, l’extérieur de la chaudière étant revêtu de laine de roche.
INDEX çç Sommaire
ESOPE 2010 : Exemple de fissurations d'équipements d'installations pétrochimiques
En 2003, lors du grand arrêt quinquennal, un contrôle par ressuage des zones réparées a
été réalisé. Aucune fissure n’avait été décelée.
INDEX çç Sommaire
ESOPE 2010 : Exemple de fissurations d'équipements d'installations pétrochimiques
o A l’issue de ce contrôle, plusieurs fissures sont décelées au niveau des
réparations réalisées en 1999.
o Toutes ces fissures sont situées dans la zone sud-ouest de la calandre.
• Des mesures d’épaisseur ont été réalisées, aucune diminution significative n’a été
relevée.
• Un sondage par ultrasons du type « recherche de défauts » a également été
réalisé au niveau des fissures, les fissures semblent s’amorcer de l’intérieur vers
l’extérieur.
Cet équipement est indispensable pour la marche de l’atelier, qui est le plus important
atelier d’acide nitrique de l’usine GPN de Mazingarbe.
Le corps est réalisé en acier faiblement allié au chrome/molybdène. Celui-ci est revêtu
intérieurement de béton réfractaire. Une réparation par l’intérieur, nécessite la dépose du
surchauffeur et du faisceau vaporisateur, la dépose du béton réfractaire. 3 solutions sont
étudiées en parallèle :
Cette solution consiste à découper les zones affectées par les diverses fissures et à les
remplacer. Une des difficultés est que la surface comporte 3 parties : 2 parties
cylindriques et une partie tronconique. Cette solution nécessite de sortir les équipements
INDEX çç Sommaire
ESOPE 2010 : Exemple de fissurations d'équipements d'installations pétrochimiques
internes de la calandre et le réfractaire de réaliser un traitement thermique, le délai pour
cette réparation est estimé à 4 mois.
L’inconvénient de cette solution est un résultat final non garanti dans le temps (risque
d’avoir d’autres problèmes sur la zone non changée). Un remplacement de la calandre
restera à prévoir.
Cette solution consiste à éliminer les fissures et à recharger pour retrouver un métal sain.
La réparation avec métal d’apport homogène, serait difficile à mettre en œuvre et
présente des risques lors de sa mise en œuvre :
- création de contraintes résiduelles de soudage qui nécessitera un traitement
thermique de détensionnement après soudage, quasi impossible à mettre en
œuvre sur site pour des problèmes de déformation et de maintien du corps de
chaudière,
- risque de fissuration à froid lors du soudage qui nécessiterait un préchauffage.
Une soudure hétérogène avec un métal d’apport plus doux (acier non allié) ne peut être
acceptée que s’il est avéré que le métal en question est compatible avec la température
de fonctionnement de l’équipement. Un calcul montre que la température de peau est de
l’ordre de 250°C (température de calcul prévue pour une nouvelle calandre de chaudière).
La solution de réparation par soudage avec un acier non allié est techniquement
envisageable.
Le risque lié à cette solution est que l’affouillement et la réparation par soudage des
fissures découvertes génèrent de nouvelles amorces de fissures non détectées.
L’intérêt de cette réparation est qu’il n’y aurait pas d’enlèvement du faisceau de la
chaudière et du briquetage et pas de traitement thermique.
Une réunion est organisée avec l’Organisme Habilité retenu, afin de présenter les trois
méthodes de réparation. A l’issue de cette réunion, la solution 3 est retenue sous réserve
de mise en place des conditions suivantes :
INDEX çç Sommaire
ESOPE 2010 : Exemple de fissurations d'équipements d'installations pétrochimiques
L’organisme habilité réalisera une réception limitée à la partie réparée, conformément au titre VI
de l’arrêté du 15 mars 2000 modifié.
Les 24 et 25 mars 2006 les travaux d’affouillement des fissures sont entrepris. Au cours
du meulage, il est rapidement constaté que les fissures ont tendance à s’étendre à
l’intérieur de la paroi. Ce phénomène semble confirmer le fait que les fissures s’amorcent
de l’intérieur vers l’extérieur.
Le 27 mars 2006 une nouvelle réunion est organisée avec l’Organisme Habilité, compte
tenu de l’étendue des meulages, il est décidé de créer une double enveloppe venant
couvrir les zones affouillées et rechargées. Cette double enveloppe aura pour objectif de
renforcer les zones réparées et de collecter une éventuelle nouvelle fuite.
Un calcul justificatif, de tenue à la pression, est réalisé par un bureau d’études sous-
traitant.
Pour permettre un approvisionnement rapide des matériaux, il est décidé de retenir un
acier d’usage courant (P265GH), des sondes de température seront mises en place afin
de s’assurer que la température de peau reste en dessous de celle validée par la note de
calcul.
Tous ces points sont validés à la fois par l’Organisme Habilité et par le réparateur.
INDEX çç Sommaire
ESOPE 2010 : Exemple de fissurations d'équipements d'installations pétrochimiques
o Soudage avec un métal d’apport « tendre » électrode de type SAFDRY 430
o Refroidissement lent
• Contrôle par ressuage après soudage
• Soudage des tôles pour fermeture de la double enveloppe
o Soudage avec métal d’apport homogène pour de l’acier P265GH
• Ressuage final après soudage
En 2007 le corps de chaudière est remplacé. Des prélèvements sont réalisés sur l’ancien
corps afin de faire réaliser une expertise et ainsi tenter de déterminer les modes de
dégradations ayant conduits à ces fissures. L’expertise est confiée à l’institut de soudure.
INDEX çç Sommaire
ESOPE 2010 : Exemple de fissurations d'équipements d'installations pétrochimiques
supports internes en cas d’absence de traitement de relaxation, contraintes de
service….)
- d’un agent corrosif dissous en phase aqueuse ou dans des condensats d’eau. Sur
les aciers non et faiblement alliés (cas de la paroi de la chaudière en acier ASTM
A387 grade A), les solutions susceptibles de générer des phénomènes de
corrosion sous contraintes sont principalement les solutions contenant des ions
nitrates (NO3-), carbonates (CO3 2-), phosphates (PO4 3-) ou des ions hydroxydes
(OH-) dans le cas des milieux caustiques.
La mise en évidence de ce type d’agents corrosifs est difficile car ils sont composés
principalement d’éléments légers (carbone, hydrogène, oxygène, azote). Les
microanalyses réalisées ont mis en évidence la présence de l’élément azote (en faible
quantité cependant). La détection de cet élément indique que les nitrates sont un agent
probable de la corrosion sous contraintes.
Les analyses par diffraction de rayons X effectuées sur les dépôts de corrosion récupérés
en surface interne de l’appareil n’ont cependant pas montré de présence de nitrate
d’ammonium (NH4NO3). Il est cependant possible que des composés azotés se trouvent
dans les dépôts sous une forme non parfaitement cristallisée (composés présents en tant
que phase amorphe). Par ailleurs, les éléments azotés recherchés peuvent être noyés et
masqués par les résidus du matériau réfractaire (alumine, silice, calcium, titane
notamment) qui ont très largement contaminé et pollué les dépôts de corrosion analysés
(diminuant d’autant la proportion et la concentration des éléments recherchés).
Les contrôles pour recherche de défauts sur les mêmes équipements des autres unités
de production d’acide nitrique, n’ont pas mis en évidence de défauts de ce type.
INDEX çç Sommaire
ESOPE 2010 : Exemple de fissurations d'équipements d'installations pétrochimiques
Ce cas reste un cas isolé dans la société. Sans que l’on puisse l’affirmer de façon
totalement sûre, il est vraisemblable que les défauts rencontrés soient le fruit de création
de nitrates dans la corps de chaudière lors d’incident sur les faisceaux vapeur, et que les
nitrates aient progressivement migrés à travers le réfractaire, pour polluer les parois
externes du corps de chaudière. Les nouvelles technologies de chaudières nitriques, ne
présentent plus ce type de problème car les parois ne sont plus protégées par des
réfractaires mais par des murs d’eau, ne permettant pas la migration de nitrates en cas
d’incident sur les faisceaux vapeurs internes vers les corps de chaudière. Les murs d’eau
(serpentins d’eau ou vapeur interne) permettent la protection thermique du corps de
chaudière sans utilisation de réfractaire.
2-1 Contexte :
Le bilan moyen de la réaction donne une production de chaleur de 16 thermies par tonne
de NH3 produit (dépend très fortement des taux de CO2 en sortie de la décarbonatation
et de CO en sortie des conversions). Une méthanation peut monter très vite en
température lors de dérèglement de la décarbonatation (ou des conversions). Les
sécurités de température associées à la méthanation doivent être fiables (tests
impératifs). Si les sécurités actuelles sont redondantes et fiables, dans le passé il ne peut
pas être exclu que des montées rapides en température aient eu lieu dans le réacteur
INDEX çç Sommaire
ESOPE 2010 : Exemple de fissurations d'équipements d'installations pétrochimiques
(l’unité date de la fin des années 60) et que ceci ait pu dégrader les matériaux du réacteur
de méthanation notamment par le risque de HTHA (High Temperature Hydrogen Attack
ou en Français attaque par l’hydrogène à haute température).
Des pertes d’intégrité brutale d’équipement peuvent survenir lorsque les fissures
réduisent suffisamment la résistance des assemblages soudés ou des matériaux.
L’HTHA est toujours précédé d’un temps d’incubation où on ne note pas de changement
détectable par les méthodes classiques d’inspection (contrôle) bien que le matériau ait
perdu localement de sa résistance mécanique.
Les tenues des aciers à l’HTHA sont fournies par les courbes de Nelson (API 941), mais
elles dépendent aussi fortement des qualités des opérations de fabrications notamment le
soudage et les traitements thermiques de détensionnement. La meilleure façon de
protéger un appareil contre le HTHA est la mise en place et le suivi de COCL, au moins
20°C en dessous des courbes de Nelson (attention à prendre la dernière édition !),
INDEX çç Sommaire
ESOPE 2010 : Exemple de fissurations d'équipements d'installations pétrochimiques
Soit par technique ultrasonore AUBT pour le métal de base, et US TOFD pour les
Soudures lorsque les parois internes ne sont pas accessibles, soit par magnétoscopie en
fluorescence humide ou par répliques lorsque les parois internes sont accessibles.
Il fut donc décidé par le service inspection de faire dès que possible un point zéro de
l’équipement sur le risque HTHA par US TOFD et AUBT.
PS 30 bar eff.
PE 45 bar eff.
Température calcul 515°C
Coeff. de soudure 0.9
Surépaisseur de corrosion 1,6 mm
Acier A387 Gr C
Constructeur CMP Arles
Volume intérieur "brut" 68 m3
Traitement thermique TTAS de 2 heures à 600/620
°C avec enregistrement des
Date d’épreuve d’origine août 1968 températures
INDEX çç Sommaire
ESOPE 2010 : Exemple de fissurations d'équipements d'installations pétrochimiques
Le plan suivant fourni une représention de l’équipement :
Erreur ! Des objets ne peuvent pas être créés à partir des codes de champs de mise en
forme.
Lors de la réalisation des contrôles TOFD, très rapidement les techniciens ont signalé
n’avoir rien trouvé concernant les défauts recherchés HTHA, car aucune des tôles, ni
soudures contrôlées n’avaient de signal rétrodiffusé.
Cependant deux défauts ont été détectés en TOFD, ce qui créa une certaine surprise !
Une analyse de l’acceptabilité du défaut suivant l’API 579 -1 (fitness for service) a
rapidement montré que le défaut était parfaitement stable dans toutes les conditions de
fonctionnement de l’équipement (épreuve, fonctionnement à basse température sous
pression, fonctionnement en pression maximum et température maximum) pour le plus
haut degré de sûreté fourni par l’API comme le montre le diagramme ci après :
INDEX çç Sommaire
GOURSAUD / CLEMENT GPN : EXEMPLE DE FISSURATIONS D'EQUIPEMENTS D'INSTALLATIONS PETROCHIMIQUES
14/16
la soudure longitudinale de la première virole a révélé une indication linéaire qui a été
constatée en peau interne de l’équipement dans la zone de liaison / zone à gros grains de
la ZAT.
Bien que les indications rencontrées aient pu être laissées en l’état avec un calcul du type
nocivité des défauts, il fut décidé par le SIR de décharger le catalyseur et de réaliser des
caractérisations internes par répliques et magnétoscopie, car l’origine du défaut rencontré
ne pouvant pas être expliqué sa cinématique était donc inconnue. Il était difficile de se
prononcer sur son évolution !
Les défauts rencontrés ont été caractérisés par répliques métallographiques. Les
analyses réalisées par l’Institut de Soudure révèlent des indications liées à la fabrication
d’origine, du type fissuration à froid, classiques des équipements fabriqués à cette
époque. Les bords très oxydés en fin de fissure indiquent clairement qu’il n’y a pas eu
d’évolution récente des défauts.
L’autre explication possible pouvait être un défaut du type fissure au réchauffage, lors
d’emballement dans le passé du réacteur. Cependant cette explication doit être écartée
car les répliques montrent que le métal fondu est sans défaut et le caractère linéaire des
fissures rencontrées est typique des fissurations à froid des ZAT des équipements
fabriqués à cette époque. De plus une fissuration au réchauffage aurait donné lieu à
plusieurs défauts de faible longueur et non de grands défauts linéaires.
INDEX çç Sommaire
GOURSAUD / CLEMENT GPN : EXEMPLE DE FISSURATIONS D'EQUIPEMENTS D'INSTALLATIONS PETROCHIMIQUES
15/16
La réparation par soudage fût rapidement abandonnée, en effet des précautions lourdes
auraient dû être mises en œuvre (dégazage de l’hydrogène avant soudure, préchauffage
et post chauffage, soudage dans une capacité préchauffée à plus de 150° C, traitement
thermique de détensionnement après soudage). Bien que cette piste ne fût pas exclue,
mais au vu de sa complexité, une autre solution a été privilégiée.
L’élimination des fissures a été jugée non notable suivant le Guide professionnel AQUAP
à savoir : Epaisseur résiduelle Epaisseur de Calcul et Epaisseur résiduelle 2/3 de
Epaisseur adoptée.
Les inspecteurs du SIR ont veillé à ce que les raccordements soient progressifs avec un
rayon supérieur à 10 mm afin d’éviter des concentrations de contraintes en service.
Pour compléter l’analyse et bien que l’épaisseur résiduelle soit supérieure à l’épaisseur
de calcul, un nouveau calcul de résistance de l’enveloppe sous pression a été réalisé,
pour estimer la diminution de marge par rapport à la situation d’origine.
De nouveau une surprise est apparue : les contraintes acceptables pour le matériau
n’étaient pas « correctes », dès l’origine. Elles ont été prises égales aux valeurs définies
dans l’arrêté de 1943 (limite d’élasticité à chaud, vérifiée sur produit), ne prenant pas en
compte les risques de fluage, ni le risque HTHA. De plus les contraintes maximales
tolérées par l’ASME (Volume 2 – Partie II), code de construction de l’équipement en
INDEX çç Sommaire
GOURSAUD / CLEMENT GPN : EXEMPLE DE FISSURATIONS D'EQUIPEMENTS D'INSTALLATIONS PETROCHIMIQUES
16/16
1967, pour ce matériau, ont été revues depuis 1967 à la baisse pour mieux prendre en
compte les risques de fluage et de HTHA (variations importantes principalement pour les
équipements travaillant au-delà de 450°C).
En conclusion je crois que ce retour d’expérience met en avant le progrès que constitue
la mise en œuvre de plan d’inspection ou de plan de contrôles, lorsque ceux-ci sont
basés sur des démarches « criticité » tel que celles définies par les guides professionnels
de l’UIC (DT32 ou DT 84). En effet la démarche criticité permet d’étendre la fonction
inspection à l’ensemble des partenaires du site et non pas aux seuls spécialistes. Mais
surtout le suivi des équipements sous pression par un Service Inspection Reconnu,
permet de suivre au plus près les équipements sous pression en prenant en compte
l’ensemble des éléments du procédé qui peuvent amener des dégradations de ceux-ci en
faisant travailler ensemble fabricant, ingénieurs procédé et spécialistes métallurgiques.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page89
RÉSUMÉ / ABSTRACT
100E
RÉSUMÉ
Lors des opérations de maintenance réalisées sur les générateurs de vapeur des réacteurs nucléaires,
EDF procède régulièrement au bouchage de certains tubes de générateurs de vapeur qui présentent des
défauts.
Les opérations de bouchage sont réalisées à l’aide de bouchons dit mécaniques fixés sur les parois des
tubes par l’intermédiaire de dents (cannelures) qui viennent s’y incruster. Ces interventions courantes
ont bénéficié d’un retour d’expérience satisfaisant en France jusqu’à ce que des anomalies remettent
en cause la maîtrise de l’opération.
Après un rappel du contexte et des circonstances de détection de ces anomalies de pose de bouchons,
les contrôles et les investigations engagés par EDF à la demande de l’ASN, et visant à comprendre les
origines de ces anomalies, sont présentés.
Les résultats des actions menées sont exposés ainsi que les évolutions du procédé de bouchage.
Enfin, sur la base de l’ensemble de ces éléments, l’ASN engagera l’examen du dossier dans la
perspective de prendre position sur le traitement générique des anomalies de pose des bouchons.
ABSTRACT
As part of the maintenance operations carried out on nuclear reactor steam generators (SG), EDF
plugs certain steam generator tubes in which defects are detected.
The plugging operations consist in blanking off the tubes using mechanical plugs fixed in the tubes by
means of ridges which grip the walls. This routine work has satisfactory experience feedback in France,
but however reveal faults which call into question the control of the operation.
After recalling the context and circumstances of detecting these plugs installation faults, monitoring
and investigations to understand the origins of these faults undertaken by EDF and following the request
of ASN, are presented.
The results of actions are exposed and the developments of the process of plugging operations.
Finally, on the basis of all these elements, ASN will conduct analysis of the record in view of taking a
position on the generic treatment of plugs installation faults.
INDEX çç Sommaire
INTRODUCTION
Depuis mai 2008, plusieurs anomalies de pose des bouchons obturant les tubes des générateurs de
vapeur ont été détectées par EDF. Ces anomalies, susceptibles d'altérer la résistance des appareils,
ont cependant été sans conséquence sur la sûreté des réacteurs.
L’objet de cet article est de présenter le contexte, les anomalies détectées et les contrôles mis en
place par EDF. Ces derniers s’inscrivent dans le cadre d’une demande de l’ASN de mener les
actions nécessaires à la compréhension des origines de ces anomalies et d'établir des critères de
contrôle permettant de vérifier la bonne tenue des bouchons.
Les résultats des actions menées ainsi que les évolutions envisagées du procédé des opérations de
bouchage des tubes de générateurs de vapeur sont exposés dans cet article.
Enfin, sur la base de l’ensemble de ces éléments, l’ASN engagera l’examen du dossier dans la
perspective de prendre position sur le traitement générique des anomalies de pose des bouchons.
CONTEXTE
Les générateurs de vapeur sont des échangeurs de chaleur entre l’eau du circuit primaire de l’eau
du circuit secondaire. Leur surface d’échange est constituée d’un faisceau tubulaire, composé de
3500 à 5600 tubes, selon le modèle, qui confine l’eau du circuit primaire et permet un échange de
chaleur en évitant tout contact entre les fluides primaire et secondaire (figure 1).
INDEX çç Sommaire
L’intégrité du faisceau tubulaire des générateurs de vapeur est un enjeu important pour la sûreté.
En effet, une dégradation du faisceau tubulaire peut générer une fuite du circuit primaire vers le
circuit secondaire. De plus, la rupture d’un tube du faisceau dans un scénario accidentel conduirait
à contourner l’enceinte du réacteur qui constitue la troisième barrière de confinement.
Les générateurs font donc l’objet d’un programme spécifique de surveillance en exploitation et de
maintenance.
Lors de ces opérations de maintenance, EDF procède régulièrement au bouchage de certains tubes
de générateurs de vapeur qui présentent des défauts (fissures, manques de matière, corrosion). Ces
opérations consistent à obturer les entrées et sorties des tubes à l’aide de bouchons dit mécaniques
fixés sur les parois des tubes par l’intermédiaire de dents (cannelures) qui viennent s’y incruster
(figures 2 et 3).
Le nombre de bouchons posés dans les générateurs de vapeur du parc nucléaire est d’environ
28 000 dont 21 500 bouchons mécaniques. La part restante est composée d’autres types de
bouchons (bouchons démontables, bouchons soudés).
Ces opérations de bouchage sont des interventions notables et sont soumises à un accord
générique de l’ASN pour le parc nucléaire d’EDF.
Ces interventions courantes ont bénéficié d'un retour d'expérience satisfaisant en France jusqu’à ce
que des anomalies remettent en cause la maîtrise de l'opération.
INDEX çç Sommaire
Page 3
DETECTION DES ANOMALIES
Ces anomalies, qui ont été détectées avant la mise en service des circuits, n’ont pas eu de
conséquence sur la sûreté des réacteurs. Toutefois, l’ éjection d’un bouchon, si elle a lieu pendant
le fonctionnement d’un réacteur, peut conduire à une rupture du tube concerné, comme cela s’est
produit en 1989 sur le réacteur n°1 de la centrale nucléaire de North Anna (Etats-Unis).
Suite à ces événements, il est apparu que le respect des critères de requalification de l'intervention
de bouchage (respect des paramètres de pose) figurant dans les accords génériques n'est pas
suffisant. EDF a donc précisé les éléments complémentaires à prendre en compte dans les données
d'intervention afin de garantir la pose correcte des bouchons et notamment la courbe de pose des
bouchons.
Une courbe de pose correcte est caractérisée par au moins un point d'inflexion avec changement
marqué de la pente traduisant l'incrustation des dents du bouchon dans le tube (figure 4).
0,6 Phase 3
0,5
Phase 2
Distance en pouces
0,4
Phase 1
0,3
0,2
Pose correcte
0,1
Courbe de Saint-Alban atypique
0
0 1000 2000 3000 4000 5000 6000
Pression en PSI
INDEX çç Sommaire
La phase 1 correspond au déplacement initial de la noix (noyau expanseur).
La phase 2 correspond à l’expansion de l’enveloppe du bouchon.
La phase 3 correspond à l’expansion de l’enveloppe et au sertissage du bouchon dans le tube.
Pour certaines courbes, la phase 3 peut être marquée par des ruptures de continuité de pentes
traduisant l'incrustation successive des dents du bouchon.
Enveloppe du bouchon
Noix
Dents du bouchon
A la demande de l’ASN, EDF a engagé, de juillet 2008 jusqu’à fin 2009, un programme de
vérification de la présence des bouchons mécaniques sur l’ensemble des générateurs de vapeur du
parc.
Ces contrôles ont eu pour objectif de vérifier par expertise télévisuelle (ETV) de la plaque à tube,
la présence, le non déplacement et l'absence de fuite au niveau de chaque bouchon.
Lorsqu’elles étaient disponibles, EDF a également examiné les courbes de pose des bouchons. En
effet, avant la détection de ces anomalies, les courbes de pose n’étaient pas systématiquement
conservées après la réalisation des opérations de bouchage.
Si les bouchons du lot de fabrication dont sont issus les bouchons en anomalie de pose sur Saint-
Alban ont été retirés, il s'avère que plusieurs générateurs de vapeur ont fait l'objet de bouchages à
l'aide de bouchons issus du lot de fabrication utilisé pour Penly 2. Parmi ces réacteurs, quatre ne
disposaient pas des courbes de pose.
L'ETV réalisée sur ces 4 réacteurs a permis de mettre en évidence un nouveau bouchon mal posé à
Paluel 3 en février 2009 et retrouvé, comme à Saint-Alban 2, à l’autre extrémité du tube.
Parallèlement, EDF a mis en œuvre des contrôles systématiques et approfondis après chaque
intervention de bouchage des tubes afin de s’assurer du sertissage suffisant des bouchons dans les
tubes. Ces données, et notamment les courbes de pose, doivent être examinées, validées et
transmises à l’ASN dans le cadre du bilan des travaux réalisés sur les circuits primaires et
secondaires principaux avant la remise en service du réacteur.
INDEX çç Sommaire
Compte tenu des lacunes mises en évidence dans le dossier de requalification, l’accord générique
de pose a été suspendu. Chaque opération de bouchage est soumise à un accord ponctuel pour un
réacteur après analyse de la synthèse de l’opération de bouchage précédente.
ETUDES ENGAGEES
Les contrôles réalisés par examen télévisuel ne permettant cependant pas d’exclure le risque d’un
déplacement éventuel ultérieur, l’ASN a demandé à EDF de mener les investigations nécessaires à
la compréhension des origines des déplacements des bouchons, à l’évaluation des risques de
déplacement et d’établir des critères de contrôle de pose des bouchons permettant de compléter les
actions mises en oeuvre de vérification de la présence de bouchons.
Les expertises métallurgiques réalisées sur les bouchons en anomalie de pose ont été associées à
celles obtenues sur des bouchons neufs et des bouchons posés sur cimblot (maquette).
Ces analyses mettent en évidence des traces de grippage et des défauts de sertissage, provoqués
par un frottement excessif au niveau de l’interface noyau expanseur / enveloppe du bouchon au
cours de son expansion (figures 6 et 7). EDF indique que ces frottements sont très probablement
liés à un manque de lubrification.
10mm
Figure 6 : zones de frottement
intense sur la noix Figure 7 : traces de grippage
dans la zone d’expansion
INDEX çç Sommaire
Lors de la fabrication d’un bouchon, le noyau expanseur et l’enveloppe sont lubrifiés. Le lubrifiant
est un produit intitulé NEOLUBE. On dit alors qu’un noyau lubrifié est « néolubé ».
L’origine de ces anomalies est probablement lié à la fabrication des bouchons.
Essais de pose
EDF a établi et mis en œuvre un programme d’essais de pose avec différentes configurations de
bouchons (lubrification ou non du bouchon, lubrification ou non de l’enveloppe, diamètre du tube
…). L’analyse des courbes de pose obtenues a été systématiquement effectuée après chaque essai.
Des paramètres plus fins ont été également étudiés, notamment le mode de néolubage du noyau
expanseur et le temps de séchage après néolubage.
Ces essais permettent de confirmer l’origine des anomalies à savoir un manque de néolubage de la
noix d’expansion. En effet, dans le cas des essais pour lesquels le noyau n’est pas néolubé, les
courbes de pose obtenues sont systématiquement atypiques.
EDF a également engagé des études par modélisation afin d’évaluer le comportement des
bouchons en service (figure 8).
Les calculs de déformée des perçages de la plaque tubulaire en zone centrale et en zone
périphérique ont été modélisés et réalisés pour les transitoires les plus pénalisants de 2ème catégorie
(transitoires de fonctionnement normal, incidents « courants » de fonctionnement) et épreuve
hydraulique.
La pression d’interface par dent entre la pose du bouchon et l’application du transitoire a été
également évaluée.
EDF conclut que la pression d’interface reste élevée dans toutes les situations, la plus pénalisante
étant néanmoins la situation d’épreuve hydraulique.
Figure 8 : maillage utilisé pour les études par modélisation du comportement du bouchon
INDEX çç Sommaire
Détermination des critères de pose de bouchons
Déplacement de la noix
La campagne de contrôle des bouchons par examen télévisuel (contrôle limité à la vérification de
la présence du bouchon, de l’altitude de pose du bouchon et de l’absence de traces de bore
traduisant la présence de fuites) entre juillet 2008 et fin 2009 complété par l’analyse de la courbe
de pose lorsque celle-ci était disponible a montré que 6 bouchons ont été détectés en anomalie de
pose sur le parc (tableau 10). Les bouchons mal posés sont issus des lots de fabrication dits de
“Penly 2” et de “St Alban 2”.
Ces lots de bouchons ont donc été mis à l’écart.
Les paramètres de pose (pression appliquée, course du noyau expanseur) de ces bouchons en
anomalie ont été contrôlés. Il a été constaté que ces bouchons, bien que respectant les critères de
bouchage, ont présenté de faibles courses du noyau expanseur.
Tableau 10 : bouchons détectés en anomalie de pose lors des contrôles par ETV
EDF a réalisé des contrôles complémentaires sur des bouchons issus du lot de fabrication de
« Penly 2 » et posés dans les générateurs de vapeur de différents réacteurs (pigeage de la position
des bouchons et mesure du diamètre du tube) (Figure 11).
INDEX çç Sommaire
Figure 11 : Position des bouchons contrôlés et issus du lot de « Penly 2 »
Les points points jaunes, rouges et les cercles bleus représentent les bouchons mal sertis détectés
respectivement à Saint-Alban 2, à Penly 2 et à Paluel 3.
Les 10 triangles roses représentent des bouchons présentant une courbe de pose atypique mais
correctement posés (contrôlés après essais de traction),
Le triangle noir représente un bouchon détecté à une mauvaise altitude mais correctement posé.
Les 11 triangles bleus représentent les bouchons issus des contrôles complémentaires et posés
dans les générateurs de vapeur du réacteur n ° 3 de Paluel.
Les six triangles verts représentent les bouchons issus du lot "Penly 2" posés à Flamanville et
correctement posés (contrôlés par pigeage de la noix).
L'ASN a remarqué que, selon ce graphe, les bouchons mal sertis sont situés dans une même zone
(cas de Saint-Alban, Penly et Paluel) correspondant à un déplacement de la noix inférieur à 0,5
pouces environ. Jusqu’à aujourd’hui, il n’a pas été constaté de bouchon mal serti présentant un
déplacement de la noix élevé (supérieur à 0,5 pouce ou 12,5 mm).
Parmi les bouchons posés dans les générateurs de vapeur du réacteur n ° 3 de Paluel et issus du lot
de « Penly 2 » , 18 bouchons ont été recensés ayant une course faible du noyau expanseur. Les
courbes de pose n’étant pas disponibles, l’ASN a demandé à EDF de traiter ces 18 bouchons
« douteux » (perçage du bouchon puis pose d’un nouveau bouchon à un altitude inférieure).
En conclusion, l’ASN a considéré que le déplacement de la noix devait être analysé comme un des
critères déterminants de la pose correcte des bouchons.
INDEX çç Sommaire
Conformité de la courbe de pose
Il a été constaté que le contrôle de la conformité de la courbe de pose des bouchons permettait
d’avoir une confiance raisonnable de l’absence de risque de déplacement du bouchon dans toutes
les situations de fonctionnement des réacteurs.
Toutefois, l’ASN a noté la difficulté de distinguer une courbe de pose correcte d’une courbe de
pose atypique. Les profils de courbes entre celles de référence et celles « avec le noyau expanseur
non néolubé » peuvent parfois être assez proches.
L’ASN a donc demandé à EDF de définir des critères de décisions lorsque des cas « douteux » se
présentent.
300
250
200
Pression en bar
150
100
Point de rupture
50
0
0 5 10 15 20 25
Déplacement en mm
Figure 9 : représentation d’une courbe de pose par 2 droites (cas d’un bouchon
correctement posé)
INDEX çç Sommaire
Base de données
EDF a également constitué une base de données à partir des rapports de fin de fabrication des
bouchons mécaniques et des rapports de fin d’intervention de pose des bouchons afin de
rechercher des éléments susceptibles d’expliquer la cause de la perte ou de déplacement des
bouchons.
EDF a précisé qu’il n’y avait pas eu de fabrication de bouchons entre 1991 et 2006, et qu’à la
reprise en 2006, il y avait eu changement d’opérateur chez le fabricant.
EDF indique que l’analyse de la base de données ne permet pas de mettre en évidence d’élément
susceptible d’expliquer la cause de la perte ou de déplacement des bouchons.
L’ASN note, à partir de l’examen du diagramme présentant les courses du noyau expanseur en
fonction de la date de fabrication, une diminution de la dispersion des valeurs des courses du
noyau à partir de la reprise des fabrications en 2006 ainsi que des valeurs de courses du noyau plus
faible globalement. L’ASN souhaite qu’il soit mis en regard cette constatation avec le fait que les
bouchons déplacés sont issus des lots fabriqués après 2006. L’analyse doit donc être prolongée
afin d’identifier si les évolutions observées sont liées à des évolutions des procédés ou des aléas de
fabrication.
Afin d’être à même de traiter les cas singuliers où le doute subsiste après analyse des paramètres
de pose ou suite aux contrôles réalisés par ETV, EDF a engagé le développement de procédés de
test de tenue des bouchons. Deux méthodes de tests non-destructifs sont en cours d’études et de
développement :
Les expertises, les essais et les études menées par EDF ont permis de comprendre l’origine des
anomalies de pose des bouchons et ont conduit à définir des évolutions en matière de fabrication et
de contrôle en service des bouchons.
Fabrication
Les expertises métallurgiques des bouchons en anomalie de pose et la campagne d’essais de pose
ont permis d’identifier l’origine des anomalies, à savoir l’absence ou le manque de lubrification
lors de la fabrication des bouchons.
EDF a donc préconisé des évolutions des spécifications de fabrication, à savoir :
• une évolution du procédé de lubrification par trempage,
• une attention particulière sur le respect des conditions d’utilisation et de conservation
du néolube,
• une habilitation des opérateurs en charge du martelage,
• une imposition du temps de séchage du néolube,
• un contrôle supplémentaire pour la bonne lubrification de la noix,
• une révision des étapes de surveillance.
INDEX çç Sommaire
EDF précise que ces évolutions pourront être introduites dès les prochaines fabrications de
bouchons.
Requalification
Après une intervention de bouchage de tubes de générateurs de vapeur, ce ne sont plus les tubes
mais les bouchons qui reprennent la pression du circuit primaire principal. Il y a donc
remplacement d’une partie résistante à la pression. L’ASN considère donc que cette l’intervention
de bouchage des tubes nécessite une requalification partielle dans un délai de 30 mois après la
pose des bouchons conformément aux dispositions de l’article 15.IV de l’arrêté du 10 novembre
1999 relatif à la surveillance de l’exploitation des CPP et CSP.
A la demande de l’ASN, EDF a transmis un programme de requalification partielle pour les
bouchons posés sur le parc par examen télévisuel (contrôle de la présence du bouchon). En cas de
doute, un examen télévisuel de caractérisation sera effectué puis éventuellement la mise en œuvre
d’essais non destructifs.
Ces dispositions de contrôle seront intégrées dans le programme de la requalification complète.
CONCLUSION
Deux années après la détection du bouchon absent à Saint-Alban, EDF a présenté les premières
conclusions stabilisées de l’affaire générique des anomalies de pose de bouchons mettant en
évidence l’origine des anomalies.
A ce jour, l’ASN est encore dans l’attente des derniers éléments des études, notamment l’analyse
de la base de données, les résultats de la phase de test du traitement automatique des courbes de
pose, les conclusions définitives des études par modélisation et les tests des outillages de tenue des
bouchons.
Tous ces éléments sont attendus au plus tard pour fin 2010 et seront intégrés dans le nouveau
dossier d’intervention de « bouchage mécanique des tubes de générateurs de vapeur » garantissant
la maîtrise de la pose des bouchons.
L’ASN examinera ce dossier et fera part à EDF de sa position sur la possibilité de délivrer un
nouvel accord générique ce qui permettra de clôturer l’affaire générique des anomalies de pose de
bouchons.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page90
RÉSUMÉ / ABSTRACT
69E
APPLICATION DE L’INSPECTION BASÉE SUR CRITICITÉ EN CHINE
Q. Xiayi, M. Xin*, W. Ying Zhi (Jiangsu Special Equipment Safety Supervision Inspection Insti-
tute, Rm. 610, Tianhe Plaza, No.216 Nor. Zhongshan Rd., Nanjing, P. R. China, 210003) 1
P. Topalis*, O. Han (DNV, Palace House, 3 Cathedral Street, London SE1 9DE, UK)
RÉSUMÉ
instituts d’inspection, comme CSEI, HGMRI et JSSEI ont pris l’initiative et ont commencé à travailler
sur la recherche et le développement d’une méthodologie IBC appropriée pour la Chine. Plusieurs
projets de recherche et projets pilotes qui ont été effectuées pendant dernières années ont conduit à
une acceptation progressive de la technique d’IBC dans le pays. Le résultat de ces projets est une
réglementation améliorée pour sécurité pour les équipements sous pression, qui sera bientôt
officiellement approuvée. La règlementation est accompagnée par des normes d’identification des
mécanismes de dommage pour les équipements sous pression et des guides pour la mise en œuvre de
l’IBC. D’autres documents « guides » pour l’IBC ont été rédigés et prennent en compte 5 unités de
raffinage typiques et 5 usines de produits chimiques typiques. Des études IBC pour presque 300 unités
de procédé ont déjà été effectuées en Chine. Ce document présente également la méthodologie IBC de
DNV et le logiciel ORBIT Onshore qui a été utilisé dans un nombre important d’études IBC en Chine.
ABSTRACT
This paper presents an overview of Risk-based Inspection (RBI) in China, including methodology review
and research, pilot project, comprehensive implementation, regulation and standard development. The
interest for RBI in China resulted from a need of improved safety, cost control and more effective
management of inspections. Since 2000, several inspection institutes, i.e. CSEI, HGMRI and JSSEI
took the initiative and started working on research and development of an appropriate RBI
implementation in China. Several research and pilot projects during the last few years led to a
progressive acceptance of the RBI technique in the country. The result of these projects is new improved
regulation on pressure vessel safety techniques, which will soon be formally approved. The regulation
is accompanied by standards on pressure equipment damage mechanism identification and RBI
implementation guidance. Additional RBI guidance documents have been drafted and account for 5
typical refinery units and 5 typical chemical plants. RBI studies for almost 300 process plant units have
already been conducted in China. The paper also presents the RBI methodology and software ORBIT
Onshore which has been used in a significant number of RBI studies in China.
INDEX çç Sommaire
APPLICATION DE L’INSPECTION BASEE SUR CRITICITE EN CHINE
Qian Xiayi, Ma Xin*, Wang Ying Zhi (Jiangsu Special Equipment Safety Supervision Inspection
Institute, Rm. 610, Tianhe Plaza, No.216 Nor. Zhongshan Rd., Nanjing, P. R. China, 210003) 1
Panos Topalis*, Oliver Han (DNV, Palace House, 3 Cathedral Street, London SE1 9DE, UK)
RESUME
ABSTRACT
This paper presents an overview of Risk-based Inspection (RBI) in China, including methodology
review and research, pilot project, comprehensive implementation, regulation and standard
development. The interest for RBI in China resulted from a need of improved safety, cost control and
more effective management of inspections. Since 2000, several inspection institutes, i.e. CSEI,
HGMRI and JSSEI took the initiative and started working on research and development of an
appropriate RBI implementation in China. Several research and pilot projects during the last few
years led to a progressive acceptance of the RBI technique in the country. The result of these
1
* Corresponding authors: [email protected], [email protected]
INDEX çç Sommaire
projects is new improved regulation on pressure vessel safety techniques, which will soon be
formally approved. The regulation is accompanied by standards on pressure equipment damage
mechanism identification and RBI implementation guidance. Additional RBI guidance documents
have been drafted and account for 5 typical refinery units and 5 typical chemical plants. RBI studies
for almost 300 process plant units have already been conducted in China. The paper also presents
the RBI methodology and software ORBIT Onshore which has been used in a significant number of
RBI studies in China.
INTRODUCTION
Risk-based Inspection (RBI) technology is a new trend in the inspection and management of
pressurized equipments in big petrochemical plants. Detailed analysis of the potential damage
mechanisms, failure modes, probability of failure (PoF), consequences of failure (CoF) and risk
levels of the pressure equipments identifies significant problems and weak links and a suitable
inspection program can be established. Equipment risks can then be reduced to a certain expected
level through the implementation of inspection methods and risk mitigation measures, to improve
safety, reduce the operating costs, and optimize the inspection strategy of the pressure equipments
[1], [2], [3], [4].
Through the enforcement of effective inspection procedures, RBI technology gives priority to
inspecting high risk equipment and reduces unnecessary maintenance and inspection of low risk
items. Therefore excessive, inadequate or ineffective testing and inspection can be avoided, and
repair and maintenance costs of pressure equipments can be reduced by 15%~40%. RBI reports are
the basis for optimizing inspection strategy of pressure vessels and piping in petrochemical plants,
and provide guidelines to determine inspection intervals and techniques [5], [6], [7].
In 1970s, the probabilistic risk analysis technique was proposed in the nuclear industry for handling
those extreme accidents with low probability and severe consequences, taking decisions on the basis
of risk and determining inspection priorities. In the energy area, Det Norske Veritas (DNV) first
employed the RBI technique [8], and developed two RBI software packages which can be used for
chemical/petrochemical equipments and offshore platforms, respectively.
In May 1993, the American Petroleum Institute (API) was supported jointly by 22 large energy
companies and developed practical methods for RBI in refineries/petrochemical companies. As a
result, API 581 Base Resource Document and API 580 Recommended Practice were produced in
2000 and 2002, respectively.
The RBI concept were introduced in China by some universities and research institutes in the later
90s. The Ministry of Science and Technology of the PRC and China Petroleum & Chemical
Corporation set a number of research projects to support RBI research. The RBI has gradually been
INDEX çç Sommaire
introduced in production in China, and some research has been done. Research on risk assessment
of pressurized equipments has been carried out in China for more than 10 years, and RBI
technology has been progressively introduced into petrochemical plants. The research of RBI
technique in China, has resulted in progressive growth of its application. In this study, the RBI
application experience in China is presented, and some application problems are discussed.
In the mid-1980s, some Chinese scholars did reliability research on pressure equipments. Prof. Dai
[9], [10], [11], [12] applied the fuzzy set theory to evaluate reliability techniques and the research
results were successfully applied in engineering practice. He proposed using the method of variance
analysis and experimental design to identify the primary parameters affecting reliability. The
reliability analysis was conducted by a Rackwitz-Skov type method. Verification results for a set of
pressure vessels were given. He also presented systematically the basic principles, methodologies
and engineering applications of the risk analysis technique for typical equipment types.
Zhang et al. [13] analyzed the factors affecting uncertainty for several risk analysis techniques in
the engineering area. The selection of measures preventing equipment failures on the basis of risk
analysis is discussed in the paper. Some examples were used to explain how to apply these risk
analysis techniques on engineering problems. Zhao et al. [14] did stress analysis, strength
assessment and risk analysis for defective pressure vessels with above-standard equipment
dimensions. Gu et al. [15] conducted risk analysis for key units such as a refinery hydro-cracker and
gave improvement advice. Zhang et al. [16] attempted to combine the online acoustic results from
storage tanks and a qualitative RBI technique in order to assess risk for storage tanks in a tank farm
which has inspection difficulties on shutdown or has other problems such as equipment stay-over
service, and the risk level of these examined oil tanks was reassessed. Zhou et al. [17] adopted the
API 581 risk-based inspection methodology, for the risk evaluation of SRT-VI tubular cracking
furnaces, which are subject to the simultaneous effect of high temperature and internal pressure and
are prone to creep. The quantitative risk results and risk-based inspection interval of furnace tubes
in the cracking furnaces were obtained and provided evidence of possible risk control and
management of the cracking furnaces. Yang et al. [18] performed quantitative or semi-quantitative
RBI analysis for equipments of petrochemical installations in China such as hydrogenation units,
catalytic cracking units and alkyl-benzene units. The major erosion damage mechanisms of
equipment and pipes in these units were identified, the risk level of equipment and pipes in these
units was assessed, the risk drivers were analyzed and proper reasonable risk mitigation measures
were proposed.
The advancement of RBI research in China helped to progressively expand the RBI application
range. Petrochemical plants like Maoming, Tianjin, Yangzi and Yanshan applied risk-based
inspection on trial on their equipment and good practical benefits were achieved. This helped in
gathering valuable experience for spreading the RBI technique. In 2006, the Chinese General
Administration of Quality Supervision, Inspection and Quarantine (AQSIQ) successively issued
documents No. 198 [19] and No. 70 and allowed the China Special Equipment Inspection and
INDEX çç Sommaire
Research Institute (CSEI), the Hefei General Machinery Research Institute (HGMRI) and the
Special Equipment Safety Supervision Inspection Institute of Jiangsu Province (JSSEI) to conduct
RBI trial applications for the main refinery equipment and the associated pipework in companies
with good management systems such as China Petrochemical Corporation and China Petroleum &
Chemical Corporation. In 2009, two safety regulations, TSG R0004-2009 and TSG D0001-2009
were published and formally approved [20], [21], and the RBI technique was first incorporated into
the safety regulation of special equipment in China. This provides the regulation basis for the RBI
application in China.
So far, the RBI research on nearly 300 petrochemical units has been completed by CSEI, GMRI and
JSSEI. The application range included pressure vessels, pipes, boilers and safety components. This
includes units such as atmospheric and vacuum distillation, catalytic cracking, cat-forming,
hydro-cracking, delayed coking, hydrogen production, ethene cracking, polythene, polypropylene,
ethanediol, aromatics, rubber, ammonia and urea.
INDEX çç Sommaire
SPECIFIC DEVELOPMENTS IN CHINA
CSEI is now in charge of preparing and establishing the RBI expert team in order to develop the
RBI techniques for pressurized equipment in China. This is also responsible for the special
equipment safety technique committee in AQSIQ. The expert team will provide the RBI technique
information for decision-making for special equipment safety supervision departments.
China petroleum & Chemical Corporation established a engineering risk research center in May
2004 in order to enhance the market competition, ensure long safe operation of the installations, and
promote the development of the RBI technique in China. The center undertook the commitment to
improve the equipment management, reduce risk and increase profit based on the RBI technique.
The engineering risk (ER) is a multi-discipline cross-domain technique and is developing fast. This
is considered as an emerging technology in the 21 century and is attracting the practitioner’s
attention. The ER technique can transform danger into safety, and thus ensure the operational safety
and economic and social benefits of operations. The RBI technique combines the probability of
failure and the consequence of failure and thus represents better the nature of production accidents.
To summarize the research results in the field of RBI technique, communicate the engineering
INDEX çç Sommaire
application experience of the RBI technique, discuss related regulations and standards with the RBI
technique, a national ER conference mainly focus on RBI technology has been held once a year
since 2006.
JSSEI has acquired the DNV RBI software ORBIT Onshore. The DNV RBI methodology is
presented in this section.
In essence, a risk assessment is performed for the installation equipment items for all active damage
mechanisms and risk-versus-time profiles are established for each item. Maximum acceptable risk
targets, possibly in combination with cost minimization methods are used to determine the
inspection dates and the most appropriate inspection techniques for each equipment item. If the risk
is not exceeded during the planning period, only regulatory compliance is recommended. If
inspection cannot reduce the risk, other measures (repair, replacement, fitness for service study) are
recommended.
Yes
INDEX çç Sommaire
RBI Modelling Functionality
As mentioned earlier, the RBI modules use risk assessment to optimize inspection planning for an
installation. It includes the following functionality:
• Risk Screening of the installation equipment to identify the highest contributors to risk
• Detailed Consequence analysis for the installation equipment with emphasis to loss of
containment accidents
• Detailed Likelihood of Failure assessment for the installed assets with emphasis on accidents
that can be prevented by inspection
• Risk summation for all active failure mechanisms
• Inspection planning on the basis of the risk results
• Additional integrity assessment for selected assets.
The DNV RBI software uses a qualitative risk assessment to screen the high risk equipment. The
features are tailored to cater for the special challenges of each discipline.
The Onshore RBI module has to handle a large variety of products. This is why the screening
approach uses a comprehensive database with 1700 chemicals with NFPA indices, which are used
for an estimate of the consequence extent (in addition to other data such as product category, the
material, and the operating & design conditions).
The Onshore RBI screening methodology offers the flexibility of three models for likelihood
assessment:
The damage mechanism method which allows expert judgement on damage mechanism
susceptibility
The statistical method mainly for moving part equipment
Remaining life calculations
A significant strength of the RBI screening analysis is that it can share data with the detailed
analysis (i.e. transition from the screening analysis to the detailed analysis can be done efficiently,
without data re-entry) and the screening analysis consequence and likelihood results can be
converted into detailed analysis results.
Most of the above features make our RBI screening approach suitable for almost all situations in the
oil and process industry. The use of comprehensive databases of chemicals and materials of
construction in the RBI Onshore module makes our approach superior to most of other screening
methodologies. The program has access to the ASME database with 1700 materials of
construction, but also special materials can be fully defined. The attached screenshot illustrates the
Qualitative Screening analysis for an onshore installation, using the damage mechanism method.
INDEX çç Sommaire
Figure 2: RBI Screening Analysis
Detailed consequence analysis is performed in the DNV RBI module (ORBIT Onshore). This
normally includes safety, economic and sometimes environmental risk. The RBI Onshore module
focuses on the calculation of effect areas (including toxic effect areas). For this purpose it uses the
lookup table module. This interpolates consequence results produced by the industry standard
Accident Consequence Analysis package PHAST 6, delivered by DNV Software. This is the default
consequence method and it is the most accurate. This method is usually combined with live use of
the PHAST discharge models. Alternatively, consequence results can be imported from the
screening analysis or from another program through the direct input facility. Consequence
calculations are performed for four hole sizes and take as input the chemical inventory, the chemical
phase, the type of scenario (pressurised or not), the vessel pressure and the chemical temperature.
There are currently about 100 chemicals with lookup tables. The lookup method can be extended to
any new chemical and the list is indeed continually expanding. The flexibility in defining new
chemicals in our software (including mixtures) is a unique strength of the package.
The figure below is an example of detailed consequence calculation. For a specific equipment item,
the program calculates automatically personnel fatality zones, toxicity zones, equipment damage
zones, outage days and the associated costs e.g. business interruption costs
INDEX çç Sommaire
Figure 3: Detailed Consequence Analysis
ORBIT Onshore calculates the likelihood of failure (LoF) for each asset and each active damage
mechanism as a function of time. In some of the modules, LoF is expressed as the product of a
generic failure frequency (GFF), a management factor (MF) and a damage factor (DF), DF being
calculated from the equipment item specific conditions and context.
INDEX çç Sommaire
Figure 4: Detailed Thinning Analysis
The RBI likelihood methodology has been subject to detailed review and has been approved by the
Dutch government for use in inspection planning by the process industry. The actual damage
mechanisms considered depend on the installation type, on the product, on the material of
construction and on the conditions.
This approach constitutes a full quantitative RBI and automatically produces an optimized
inspection plan. The Inspection plan is dependent on the conditions for the specific equipment item
and its inspection history (NDT method, findings and no-findings).
INDEX çç Sommaire
For some of the above mechanisms, there are models to estimate damage susceptibility and damage
rates such as corrosion rates. These damage rates or susceptibilities can then be used to calculate
Likelihood of Failure LoF.
Risk Assessment
Risk is often defined as the product of consequence and likelihood, but there is variety of risk
measure definitions depending on the country or industry context. A risk measure is typically
compared against a risk target which determines if the risk is acceptable and if inspection is needed.
The RBI Onshore module can calculate and display 7 different risk measures, and the user has the
flexibility to select one or more according to his context. The currently available risk measures
are:
Total economic cost
Business interruption cost
Equipment damage area
Safety effect area
Potential Loss of Life (PLL)
Likelihood of failure (LoF)
Damage Factor (DF)
Inspection Planning
Currently our RBI methodology can determine if there will be a need for inspection for each asset
during a future evaluation date or before this date. This is based on the risk variation with time for
the specific item. It also determines the inspection effectiveness, which is translated into a
specific inspection technique and coverage. The data from the Inspection Planning feature in the
RBI modules are sent to the ORBIT+IDS Inspection Database.
The attached figure illustrates the process of determining an inspection date and an inspection
effectiveness (translated to an inspection technique) on the basis of risk. A maximum acceptable
risk level is set. A future evaluation date is selected (it is typically the date of the 2nd turnaround)
and the risk is calculated as function of time. If the risk at the future evaluation date exceeds the
maximum acceptable level, an inspection is suggested. The intersection of the risk curve and the
maximum acceptable line sets the latest inspection date. The inspection may need to be included in
INDEX çç Sommaire
the next turnaround, if it cannot be done on-stream. An inspection effectiveness is selected so that,
after inspection, the risk does not exceed the maximum acceptable level at the future evaluation
date.
Risk / DF
Acceptable
Risk Level
Damage Factor (DF)
Fairly Effective
Highly Effective
Predicted Risk
Increase
Time to next
inspection
Now
st
Time
1 TA 2nd TA
Last Inspection date Max. insp. interval
Additional features to allow more detailed inspection management and are included in the RBI
Onshore module:
Optimization of multiple inspections (possibly over a longer period of time) mainly on the
basis of risk reduction or inspection/maintenance cost reduction
Risk Benefit Analysis. Quantifies the risk reductions achieved because of inspection for a
given inspection cost.
Cost Benefit Analysis. Quantifies the reductions in inspection/maintenance cost, when moving
from non-RBI methods to RBI.
The RBI modules include a number of standard reports as well as the capability of user-defined
custom reports. For instance the RBI onshore module includes:
Risk matrices for the Current Date, Future Date prior to inspection, Future date post
inspection
Executive summary with risks per equipment type and damage mechanism, risk variation
with time with / without inspection
Inspection plans showing what equipment items to inspect, for what type of damage, when
to inspect and what approach to use
INDEX çç Sommaire
Cost Benefit Analysis/ Risk benefit analysis
Equipment summary sheet
Detailed risk breakdown including business interruption, equipment damage and personnel
safety contributions
The attached screenshot presents an extract from the standard executive summary report of the RBI
Onshore module.
The user can also build his own reports by selecting a number of fields to display. Reports can be
exported into MS Word or PDF . Further more tabular reports can be exported into MS Excel.
What-if analysis is easily performed with the detailed quantitative RBI model.
INDEX çç Sommaire
DISCUSSION
Given several years of RBI experience of conducting the RBI in China, some issues should also be
highlighted:
For an RBI project in China, the related mandatory codes and safety regulations on pressure vessels
and piping should be taken into consideration. The target of RBI is to reduce operation and
maintenance costs and at the same time to guarantee the safety of pressurized equipments, and the
purpose of implementing safety supervision regulations on pressure equipments is also to assure the
safe operation of equipments in the chemical and petrochemical industry. It should be noted that
RBI technology itself is still in the process of continuous improvement and development. Detailed
implementation of the RBI results in China requires a large amount of work, including basic
material and chemical property data , in-service inspection and nondestructive testing technology,
supporting standards, codes & regulations.
It would not be appropriate to transfer thoughtlessly international engineering practices and RBI
regulations directly to pressure equipments in China, because significant differences exist in design,
manufacturing and service conditions of pressurized equipments in different nations, and the codes
and regulations on inspection requirements are accordingly different. Therefore, successful risk
assessment application in China requires experience gathering and trial applications. After RBI
analysis, the pressurized equipment risk should be within the acceptable level, and the pressurized
equipment with high PoF, especially those with high CoF, should be given special attention.
Meanwhile, the location, proportion, inspection interval and NDT methods for pressurized
equipments, obtained from RBI research, should be made compliant with the Chinese codes and
regulations as well as to the local company needs.
The RBI of outfits in China is mostly based on API 581 with the consideration of factors like
equipment weld ratio and erosion margin, but without the influence of material quality and
super-standard defects of equipments. Many petrochemical equipments in China were put into
production between 1970s and 1980s. Limiting to the metallurgy and manufacture situations in that
time, there would exist super-standard flaws in these equipments., and the quality and corrosion
resistance of steels produced in China was unstable, but the RBI systems of foreign countries did
not consider these two problems. These foreign RBI systems regard that there is no difference
among equipments if they are designed and fabricated based on the corresponding regulations.
INDEX çç Sommaire
Actually, the fabrication and material quality of equipments are essential tp the failure of
equipments. Therefore, these two problems should be considered and quantified properly in the
application of the RBI technique.
A suitable material and chemical database is the basis to conduct RBI study successfully. The
database of RBI software contains common chemicals and materials usually available in refineries
and chemical plants. A good database, which includes various sorts of material behaviors in
different media, the failure mechanism of typical pressure equipments, failure modes, performance
degeneration in high temperature, the corrosion feature of various media and so on, is pivotal to an
excellent RBI software. Considering the distinctive features of pressurized equipments in China, a
RBI database of materials and chemicals owing Chinese intellectual property is suggested to be
built gradually, and the database will be more complete and abundant with the accumulation of
experience on RBI technology.
ABBREVIATIONS
ACKNOWLEDGEMENTS
The authors are grateful to JSQTS (Jiangsu Quality & Technology Supervision Bureau) for its financial support
(Grant No. JG08026) for this study
INDEX çç Sommaire
REFERENCES
[1] Qian Xiayi, Miao Chunsheng, Ma Xin. Review And Prospect on Risk Assessment
Application in Pressure Vessels Inspection in Jiangsu. 9th International Conference on Engineering
Structural Integrity Assessment. Beijing, P. R. China. Oct.15-19, 2007
[2] American Petroleum Institute. Risk-Based Inspection Base Resource Document API
Publication 581. 2000. 5
[3] American Petroleum Institute. Risk-Based Inspection API Recommended Practice 580.
2002. 5
[4] Sun Xinwen. Prospect of Risk-based Inspection Application on Management of Special
Equipment in Petrochemical Plants. Petro-Chemical Equipment Technology. 2006, 27(3): 33-35 (in
Chinese)
[5] Aller, JE, Horowitz NC, Reynolds JT. et al. Risk Based Inspection for the Petrochemical
Industry. ASME Pressure Vessels Piping Div Publ PVP. 1995, 296: 399-406
[6] Reynolds, JT. Application of Risk-based Inspection Methodology in the Petroleum and
Petrochemical Industry. ASME Pressure Vessels Piping Div Publ PVP. 1996, 336: 125-134
[7] Zhang Pingsheng. Risk Management and Risk-based Inspection for Oil and Gas Pipelines.
OGST. 1999, 18(2): 13-18(in Chinese)
[8] Huang Xianbing, Li Yanyuan, Lan Zhenggui. Risk-based Inspection: An Emerging
Technology in Equipment Management. Purchasing, Store and Distribution. 2004, (3): 73-77(in
Chinese)
[9] Dai Shuhe. Reliability Predication of Pressure Vessel. Journal of Nanjing Chemical
College. 1986, 8(2): 1-18 (in Chinese)
[10] Dai Shuhe. Fuzzy Assessment of Process Equipment Reliability. Journal of Chemical
Industry and Engineering. 1993, 44(1): 103-110 (in Chinese)
[11] Dai Shuhe. Risk Analysis Techniques :Basis Principle and Methods. Chinese Pressure
Vessel Technology. 2002, 19(2): 1-9 (in Chinese)
[12] Dai Shuhe. Risk - Analysis Techniques :Applications in Engineering. Chinese Pressure
Vessel Technology. 2002,19(3): 1-6 (in Chinese)
[13] Zhang Peng, Duan Yonghong. An idea and method of the long distance transmission
pipeline risk technique. Natural Gas Industry. 1998, 18(5): 72-76 (in Chinese)
[14] Zhao Jianping, Shen Shiming. Risk Analysis on Pressure Vessels with Defects Exceeding
Geometric Limit. Chemical Equipments and Piping. 2000. 37(1): 21-23 (in Chinese)
[15] Gu Junming, Lu Peng, Wu Aogeng, et al. Application of Marsh. Inc. Risk-based
Inspection Technology in Nanjing Refinery. Petrochemical Safety Technology. 2000 (6): 1~5(in
Chinese)
[16] Zhang Ying, Dai Guang, Zhang Ying, et al. Applications of the Risk Base Inspection
(RBI) Technique in the Reserve Oil Tank Zones. Chemical Machinery. 2006, 33(3): 174-177(in
Chinese)
[ 17 ] Zhou Changyu, Zhao Tao. Risk-based inspection for cracking furnaces tubes.
Petrol-Chemical Equipment. 2006, 35(2): 73-76 (in Chinese)
[18] Yang Tiecheng , Chen Xuedong , Chen Sixiang, et al. Application of Quantitative RBI
Technology in Alkylbenzene Unit. Chinese Pressure Vessel Technology. 2006, 23(10): 41-44
INDEX çç Sommaire
[19] General Administration of Quality Supervision, Inspection and Quarantine of the People’s
Republic of China. On Trial-application of Risk-based Inspection (RBI) Technology in China.
Document No. 198, 2006. http://tzsbaqjcj.aqsiq.gov.cn/tzwj/zjwh/200610/t20061024_2276.htm (in
Chinese)
[20] General Administration of Quality Supervision, Inspection and Quarantine of the People’s
Republic of China. TSG R0004-2009 Supervision Regulation on Safety Technology for Stationary
Pressure Vessel. Xinhua Press. Beijing, 2009.
[21] General Administration of Quality Supervision, Inspection and Quarantine of the People’s
Republic of China. TSG D0001-2009 Pressure Pipe Safety Technology Supervision Regulation for
Industrial Pressure Pipe. Xinhua Press. Beijing, 2009.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page91
RÉSUMÉ / ABSTRACT
15E
ÉVALUATION DU DOMMAGE ET RÉPARATION DE RENFLEMENTS
DANS UN BALLON DE COKEFACTION
ABSTRACT
Coke drums are large refinery vessels that are exposed to severe mechanical and thermal loads during
their operation. Bulging is a common degradation mechanism that causes cracking, leakage, and fires
in these drums. In this presentation, a technique for assessing bulging severity is described along with
methods that can be used to repair severe bulges. These tools can be utilized to quantify the risk of
failure, plan maintenance outages, and extend the lives of coke drums.
INDEX çç Sommaire
ASSESSMENT AND REPAIR OF BULGES IN COKE DRUMS
ABSTRACT
Coke drums are large refinery vessels that are exposed to severe mechanical and thermal loads during
their operation. Bulging is a common degradation mechanism that causes cracking, leakage, and fires
in these drums. In this presentation, a technique for assessing bulging severity is described along with
methods that can be used to repair severe bulges. These tools can be utilized to quantify the risk of
failure, plan maintenance outages, and extend the lives of coke drums.
INTRODUCTION
A Delayed Coker Unit (DCU) is a refinery component that is used to fractionate heavy oil residues
into lighter fluids. The semi-batch process includes thermal cracking that produces solid coke and
lighter hydrocarbons.
Coke drums are the large vertical pressure vessels that are used for the batch side of the DCU process.
Their size varies from 15 to 30 feet in diameter and from 55 to 90 feet in height. Their metallurgy is
typically Carbon Steel, C-½Mo, or Cr-Mo alloys. They are usually designed per ASME’s Boiler and
Pressure Vessel Code, Section VIII Division 1.
The cracking process exposes these vessels to unusually harsh thermal and mechanical conditions
during their normal cyclic operation. Since these severe cyclic loads are not accounted for in the
design process most drums as they age develop permanent plastic deformations and experience
various failure mechanisms.
One of the most common failure modes and potentially the most serious one is the permanent plastic
deformation and growth of the shell which is manifested as radial growth, ovalization, and leaning
(tilting). Bulging or the non-uniform radial growth is the one that is more likely to cause cracks, leaks,
fires, operation delays, and potentially catastrophic accidents. An extreme case of bulging is shown in
Figure 1. The causes of radial growth and bulging are beyond the scope of this paper.
INDEX çç Sommaire
Figure 1: An extreme case of bulging
The advent of laser scanning in the 1990s gave refineries an insight into the complex geometric
patterns of a bulging drum. Figure 2 shows the laser scans of two sister drums that are unrolled into
three-dimensional surfaces. The three axes are azimuth, height, and radius. To demonstrate the
underlying features of the surfaces, the figure exaggerates radial growth in comparison to height and
azimuth coordinates. Despite the fact that these drums were identically designed, fabricated, and
operated, the locations and patterns of bulging are drastically different. Furthermore, while the drums
have similar bulging magnitudes, the drum on the left has no sign of cracking while the one on the
right has a bulging-induced through-wall crack.
Bulging accelerates the formation and propagation of cracks that can cause leaks and fires due to the
resulting increase in nominal stresses. Also, with some exceptions, the rate of bulging usually
accelerates over time.
INDEX çç Sommaire
To maintain the mechanical integrity and prolong the economic life of a drum, it is critical to:
1. Assess the severity of bulging in a given drum and its likelihood of cracking,
2. Identify and rank the specific areas of the drum that require further assessment,
4. Identify the areas that require repairs and perform engineered repairs in a timely manner.
Conventional pressure-loaded linear finite element models can’t be used to assess bulging severity in
coke drums because of the following reasons:
(1) Bulged geometries have significant plastic deformations and residual stresses that are not
accounted for in such simplified models.
(2) Loads applied on coke drums are significantly more complex than ordinary pressure vessel
loads. Field measurements and detailed finite element analyses of coke drums have shown that,
typically, transient stresses that develop in the shell can be an order of magnitude higher than
design stresses.
(3) The three-dimensional pattern of bulging makes the analysis too complex for a simplified
model to resolve.
To evaluate the fitness-for-service of a bulged drum, the API 579-1/ ASME FFS-1 Fitness-For-Service
Standard of 2007 [2] requires that a Level 3 Assessment be conducted that examines the likelihood of
plastic collapse, local failure, buckling, and fatigue failure. To do that, the following tasks need to be
conducted:
2. simulate how these bulges were formed to account for residual stresses and plastic
deformation in bulges using nonlinear model,
The above assessment procedure would require a significant effort that consists of long-term
monitoring and analysis which is both prohibitively expensive and technically challenging.
The Bulging Intensity Factor (BIF) is a waveform recognition technique that analyzes bulging patterns
to determine their severity and likelihood of cracking. To develop the technique, a database of scanned
drums with known bulging-related cracking histories was collected. Then, the geometric features
associated with crack locations were analyzed using image identification tools employed in such fields
as computer vision, character recognition, signal interpretation, and image processing. The technique
which relies on bi-axial frequency, magnitude, and curvature of bulges was found to be consistent in
identifying the geometric features associated with cracking history. The background and technical
details of this technique are described in [1].
INDEX çç Sommaire
The BIF is a non-dimensional parameter that is used to assess the level of bulging severity and,
therefore, the likelihood of bulging-related cracks in coke drums. As described above, the calculation
process is dependent on geometric patterns and independent of loading and the wall thickness of the
drum.
The outcome of the analysis is a matrix of BIF values at a grid of analysis points on the drum. Table 1
shows how BIF values are translated into severity grades for alloy drums. A BIF magnitude under
0.75 indicates a “low” grade of severity and likelihood of cracking. Magnitudes over 2 are associated
with a “severe” likelihood of cracking and typically represent the end of economic life for that portion
of the drum. Magnitudes between 0.75 and 2 are assigned grades of “medium”, “high”, and “very
high”. Negative and positive BIF values indicate external and internal initiation of cracks,
respectively.
Table 2 shows how the BIF-based severity grades can be used for making integrity management
decisions. The “low”, “medium”, “high”, “very high”, and “severe” grades of bulging severity are
related to “rare”, “seldom”, “occasional”, “repeated”, and “too frequent” patterns of bulging-related
cracking, respectively. These grades can also be used as guidance for recommended laser-scanning
intervals. As shown in the table, the time interval between scans varies from 1 to 3 years for BIF
magnitudes under 2. When the BIF magnitude is over 2, a partial or full replacement of the shell
should be considered.
REPAIR
As bulging increases in severity, cracks initiate at points of maximum stresses. If cracks are not
repaired or arrested, they propagate through the wall and cause leaks and possibly fires. Short-term
repairs involve gouging out cracks and welding the wall back up to original thickness. While this
process stops a particular crack from propagating, it has no impact on the rate of bulging growth.
Therefore, while short term crack repairs are necessary, they are insufficient. To extend the life of
bulged drums, a long-term repair that stops or slows down the rate of bulging is needed.
INDEX çç Sommaire
As shown in Figure 3, there are three main long-term repair techniques for bulges. Window and can
replacement are used to replace damaged material with new material. Weld overlay is a repair
technique that strengthens and improves the bulged section.
Window replacement is a traditional repair method for pressure vessels. It is relatively inexpensive
and fast to implement. However, due to the bi-axial difference in the elastic limit between the old
worked wall and the new plates and the inevitable geometric misalignments between the old
corrugated shell and the new straight plate, cracks usually initiate at window corners within one to two
years after installation.
Can replacement is more costly and time-consuming than window replacement. Its advantage over the
latter is that the elastic limit mismatch is uniaxial which is less severe than bi-axial mismatch. Because
of that, cracks tend to initiate two to five years after installation.
Weld overlay repair was traditionally used for relatively small areas. The advent of computerized
welding techniques has made possible the repair of large bulges using consistent weld depositing
process. The increase in wall thickness provides the original distorted wall with added resistance to
loads which reduces nominal stresses. Also, if properly designed, the weld overlay can employ
residual stresses and shrinkage patterns of the weld to slow down the rate of bulging growth and retard
crack initiation and propagation. This technique is relatively new and experience is relatively limited.
At the time of writing this paper, the oldest known weld overlay repair was about ten years old and it
was in excellent condition.
The best long-term repair option often depends on the desired life extension, the limit on unit
shutdown, and the availability of financial and technical resources. When the number of bulges and
frequency of cracks become too overwhelming, drum replacement becomes necessary.
SUNCOR’S EXPERIENCE
Suncor’s tar sands refinery in Fort McMurray Canada has one of the world’s largest upgrading coking
units. The plant contains six drums built in 1966, two in 1979, four in 2001, and eight new ones or
pending for a total of 20 drums. They range in diameter from 26 to 32 feet and in height from 66 to 94
feet.
INDEX çç Sommaire
Needless to say, the maintenance and integrity assurance of these drums is a top priority for refinery
management. As shown in Figure 4, a comprehensive integrity management approach has been
developed using laser scans, bulging severity analysis using the Bulging Intensity Factor (BIF), finite
element analysis (FEA), probabilistic crack propagation calculations, strain gage measurements, and
Acoustic Emission Testing (AET).
Drum Remaining
Cold Eyes Life
Review
The BIF is an essential part of Suncor’s integrity management process that is utilized for two main
purposes. First, it is used for initial screening of drums to determine their bulging severity and the
need for any further analysis or testing. Second, successive BIF analyses are used for making future
predictions regarding the cost of drums maintenance and the need for and timing of partial or full
replacement of the shell.
The two-dimensional BIF contour plot in Figure 5 and the three-dimensional maps shown in Figure 6
belong to one of Suncor’s drums that was experiencing a discernable bulge at a circumferential seam
that extended all the way around the drum. When successive scans of the drum were analyzed using
BIF, it was determined that this seam bulge (labeled B in the figures) was indeed a serious bulge at the
grade of “Very High” severity. However, the analysis also showed that its severity was relatively
INDEX çç Sommaire
stable over time. The smaller and more localized bulge (labeled A in the figure) that is located
between two seams was found to be more severe than Bulge B and was deteriorating faster. The
degradation model predicted that the localized bulge would reach the critical BIF magnitude of 2
between May 2005 and June 2006. Shortly after the analysis was completed, a through-wall crack
developed at this localized bulge in August 2005.
C A
INDEX çç Sommaire
In the above example, a long-term repair plan was analyzed and implemented for Bulge A using
engineering weld overlay. The options of applying the overlay on the inside, outside, and both sides
were analyzed and evaluated to assure the maximum effectiveness of the repair. A rigorous analysis of
the overlay was conducted using a three-dimensional finite element model that replicated the bulged
geometry of the drum based on the latest available scan. All stages of the repair process were analyzed
including preheat, welding, and post-heat stages for structural integrity, residual stresses, and
distortion. The transition between the overlaid area and the original wall was analyzed for alternative
taper geometries using solid axisymmetric models.
Figure 7: Representative plots from the finite element analysis of the overlay weld repairs
The above weld overlay has been in service for two years. After the repair, the severity of bulging
decreased as shown in Figure 8. No more cracking in Bulge A were found. Stable Bulge B is being
closely monitored. The drum continues to be routinely scanned and analyzed to track changes in
bulging patterns.
INDEX çç Sommaire
2.1 Severe
1.5
BIF
1.3
High 2009
1.1
0.9
Medium
0.7
Low
0.5
5000 5500 6000 6500 7000 7500 8000 8500 9000 9500
Cycles
As with the above example, BIF results are used by Suncor to track changes in severity of bulges in all
drums over time. By fitting a curve to BIF data points we can forecast when a particular drum will
likely reach the severity level “very high” and determine remaining life accordingly. So far, bulges
have failed before reaching a BIF magnitude of two.
BIF analyses have helped Suncor effectively manage the mechanical integrity of coke drums.
Rigorous inspection is conducted on bulges with higher severity grades. Lower inspection priority is
given to areas with lower bulging severity. Since drum areas with high severity are typically a small
percentage of total drum areas, this ranking scheme has helped optimize the allocation of inspection
and maintenance resources.
Suncor has found that the physical size of a bulge is not a good indicator of severity. The BIF number
was found to be a much more reliable indicator of when a drum will fail. Out of 6 drums that have
been operating since 1967, two have had cracks and four have never cracked. They are operated by the
same crews and under the same conditions and see the same product. The fact that some drums crack,
could easily lead plant personnel to panic and start a program of drum replacement which is very
costly. However, the BIF “measurement” has given us the insight and confidence that the drum life is
predictable and that we can plan accordingly.
In summary, Suncor uses the BIF along with other available tools to examine the structural integrity of
coke drums and make future predictions of inspection needs and projected life. Suncor’s experience
shows that the technique correlates well with actual cracking history.
ACKNOWLEDGMENTS
The authors recognize the valuable contributions of many engineers who have assisted in the analysis
and review of the material presented in this paper. Special thanks to Asad Ali, Vrajesh Shah, Charles
Stephens and Aaron Johnson of Suncor, and to Abdulhakeem Al-Mekhnaqi and Richard S. Boswell of
Stress Engineering Services, Inc.
INDEX çç Sommaire
REFERENCES
[1] Samman, M. and DuPlessis, P., “The Bulging Intensity Factor (BIF): A technique for assessing
the bulging severity of coke drums”, Paper RMC-07-100, Proceedings of the NPRA Reliability
and Maintenance Conference, San Antonio, 2007.
[2] API 579-1/ ASME FFS-1 Fitness-For-Service Standard, a joint publication by the American
Petroleum Institute and the American Society of Mechanical Engineers, 2007.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page92
RÉSUMÉ / ABSTRACT
12E
GESTION ET SUIVI EN SERVICE
DES ÉQUIPEMENTS SOUS PRESSION FIXES
D. A. Osage
D. R. Thornton
P. A. Henry
The Equity Engineering Group, Inc.
Shaker Heights, OH - USA
RÉSUMÉ
La plupart, voire tous les équipements sous pression resteront en service bien au delà de la durée de vie prévue lors de la
conception. Pour assurer des opérations en toute sécurité et fiabilité il est nécessaire d’adopter une procédure de gestion et
Mercredi
suivi en service (Life-Cycle Management - LCM). Dès la phase de conception la procédure LCM requiert l’identification
des mécanismes d’endommagement potentiels, la conception elle-même devant résister à ceux-ci ou les diminuer. Après la
mise en service, la procédure LCM se poursuit par la mise en place d’un plan d’inspection prescriptif ou basé sur une analyse
des risques. L’évaluation des résultats des inspections révèlera si les dommages constatés ont bien été anticipés (pris en
compte dans la conception) ou non. Si le dommage constaté a été anticipé et si l’on reste dans les limites de conception
(corrosion inférieure à la surépaisseur de corrosion, par exemple), l’équipement sera remis en service et un nouveau plan
d’inspection sera établi en se basant sur la durée de vie résiduelle (demi-vie ou autre approche suivant le type
d’endommagement). Si le dommage constaté n’a pas été anticipé, la procédure LCM requiert l’identification du mécanisme
d’endommagement et une évaluation de l’aptitude au service pour faciliter la prise de décision de conserver, détimbrer,
réparer ou remplacer le composant endommagé. Pour pouvoir mettre en place efficacement les procédures LCM des codes
ou normes couvrant tous les aspects de ces procédures doivent être disponibles. De plus l’intégration de ces règles doit être
coordonnée afin que les techniques d’analyses soient similaires en conception et en analyse en service. Une vue d’ensemble
des codes et normes API, ASME et d’autres sources est présentée, accompagnée d’une discussion sur les efforts en cours
pour développer les règles nécessaires à la mise en œuvre des procédures LCM.
ABSTRACT
Most, if not all, pressurized equipment items will continue to operate well beyond their intended design life. To
ensure that the equipment operates safely and reliably requires adoption of an equipment Life-Cycle Management
(LCM) process. At the equipment design stage the LCM process requires identification of potential damage
mechanisms and design to resist or mitigate the damage. After the equipment has been commissioned, the LCM
process continues with the use of prescriptive or Risk-Based inspection. An evaluation of the in-service inspection
results reveals whether any damage that may have occurred is anticipated (i.e., was considered in the initial
design) or unanticipated. When the damage is anticipated and within the design limits, the equipment is put
back into operation for a period of time that considers the equipment remaining life, with a new inspection at
the end of the operational period. If unanticipated damage is discovered the LCM process requires identification
of the damage mechanism and a subsequent Fitness-For-Service assessment to facilitate a decision to run as is,
rerate, repair, or replace the damaged components. To effectively implement the LCM approach, codes and
standards must exist that cover each aspect of the process. In addition, the technology integration in these codes
and standards must be coordinated so that similar analysis techniques are employed at the time of construction
and for in-service assessments, as required. An overview of API, ASME, and other codes and standards is
provided together with a discussion of the efforts to integrate technology to support the LCM process.
INDEX çç Sommaire
LIFE-CYCLE MANAGEMENT OF PRESSURIZED FIXED EQUIPMENT
David A. Osage
David R. Thornton
Phillip. A. Henry
The Equity Engineering Group, Inc.
Shaker Heights, OH - USA
ABSTRACT
Most, if not all, pressurized equipment items will continue to operate well beyond their intended
design life. To ensure that the equipment operates safely and reliably requires adoption of an
equipment Life-Cycle Management (LCM) process. At the equipment design stage the LCM
process requires identification of potential damage mechanisms and design to resist or mitigate
the damage. After the equipment has been commissioned, the LCM process continues with the
use of prescriptive or Risk-Based inspection. An evaluation of the in-service inspection results
reveals whether any damage that may have occurred is anticipated (i.e., was considered in the
initial design) or unanticipated. When the damage is anticipated and within the design limits, the
equipment is put back into operation for a period of time that considers the equipment remaining
life, with a new inspection at the end of the operational period. If unanticipated damage is
discovered the LCM process requires identification of the damage mechanism and a subsequent
Fitness-For-Service assessment to facilitate a decision to run as is, rerate, repair, or replace the
damaged components. To effectively implement the LCM approach, codes and standards must
exist that cover each aspect of the process. In addition, the technology integration in these codes
and standards must be coordinated so that similar analysis techniques are employed at the time of
construction and for in-service assessments, as required. An overview of API, ASME, and other
codes and standards is provided together with a discussion of the efforts to integrate technology
to support the LCM process.
RÉSUMÉ
La plupart, voire tous les équipements sous pression resteront en service bien au delà de la durée
de vie prévue lors de la conception. Pour assurer des opérations en toute sécurité et fiabilité il
est nécessaire d'adopter une procédure de gestion et suivi en service (Life-Cycle Management -
LCM). Dès la phase de conception la procédure LCM requiert l'identification des mécanismes
d'endommagement potentiels, la conception elle-même devant résister à ceux-ci ou les diminuer.
Après la mise en service, la procédure LCM se poursuit par la mise en place d'un plan
d'inspection prescriptif ou basé sur une analyse des risques. L'évaluation des résultats des
inspections révèlera si les dommages constatés ont bien été anticipés (pris en compte dans la
conception) ou non. Si le dommage constaté a été anticipé et si l'on reste dans les limites de
conception (corrosion inférieure à la surépaisseur de corrosion, par exemple), l'équipement sera
remis en service et un nouveau plan d'inspection sera établi en se basant sur la durée de vie
résiduelle (demi-vie ou autre approche suivant le type d'endommagement). Si le dommage
INDEX çç Sommaire
constaté n'a pas été anticipé, la procédure LCM requiert l'identification du mécanisme
d'endommagement et une évaluation de l'aptitude au service pour faciliter la prise de décision de
conserver, détimbrer, réparer ou remplacer le composant endommagé. Pour pouvoir mettre en
place efficacement les procédures LCM des codes ou normes couvrant tous les aspects de ces
procédures doivent être disponibles. De plus l'intégration de ces règles doit être coordonnée afin
que les techniques d'analyses soient similaires en conception et en analyse en service. Une vue
d'ensemble des codes et normes API, ASME et d'autres sources est présentée, accompagnée
d'une discussion sur les efforts en cours pour développer les règles nécessaires à la mise en
œuvre des procédures LCM.
INTRODUCTION
Owner-users of pressurized equipment including pressure vessels, piping, and tankage are
becoming increasingly interested in Life-Cycle Management (LCM) of equipment to enhance
reliability and availability. A key aspect of whole life management is the understanding and
consideration of potential damage mechanisms in the design process, and the identification of
active damage mechanisms when equipment is in-service. Welded components represent a
particularly difficult challenge in that the associated damage mechanics are strongly influenced
by materials of construction, environmental effects, and loading conditions. The Welding
Research Council (WRC) and American Petroleum Institute (API) have produced publications to
facilitate the damage mechanism determination process. In 2007, ASME released a new edition
of the Section VIII, Division 2 Boiler and Pressure Vessel Code that provides new rules for the
design of equipment based on the prevention of failure modes. These new rules include the use
of Examination Groups that are employed in some European standards including EN 13445,
permit ultrasonic examination in lieu of radiographic examination, and introduce the concept of
weld joint efficiencies and the associated partial examination requirements into Section VIII,
Division 2 for the first time. In addition, a new ASME Code Case to the 2007 Edition of Section
VIII, Division 2 that permits the use of 2.25Cr-1Mo-V for operating temperatures greater than
371°C (700°F) has been developed. Also in 2007, in conjunction with ASME, API released the
second edition of API 579 designated as API 579-1/ASME FFS-1 2007 Fitness-For-Service.
This new standard, which is based on the first edition of API 579, incorporates all planned
technical enhancements originally slated for the second edition, the key enhancements being the
inclusion of assessment procedures for pressurized components operating in wet H2S service and
for pressurized components operating in the creep range. These new standards together with API
and The National Board Inspection Codes and a new series of post-construction standards form
the basis of a LCM process for pressurized fixed equipment.
INDEX çç Sommaire
integration between construction codes and standards for inspection and Fitness-For-Service is a
necessary element for a successful LCM process for fixed equipment because similar
technologies are utilized for both design and in-service assessments.
A review of Figure 1 indicates that the necessary codes and standards are in place to implement
the LCM process. The codes and standards in this figure represent a LCM process that applies to
refineries and petrochemical plants that satisfies typical regulatory requirements in the U.S.
However, the framework of the LCM process that is presented is also applicable to other
industries and international locations by substituting the appropriate documents for damage
mechanism identification, new construction, in-service inspection, Fitness-For-Service, and
repair.
INDEX çç Sommaire
• Subsection C covers specific requirements applicable to the several classes of materials
used in pressure vessel construction. It deals with carbon and low alloy steels, nonferrous
metals, high alloy steels, cast iron, clad and lined material, cast ductile iron, ferritic steels
with properties enhanced by heat treatment, layered construction, and low temperature
materials, respectively. Allowable stress values for these classes of materials are
contained in ASME B&PV Codes, Section II, Part D.
• The mandatory appendices address specific subjects not covered elsewhere in VIII-1 , and
their requirements are mandatory when the subject covered is included in construction.
• The non-mandatory appendices provide information and suggested good practices.
VIII-1 is the most prevalent construction code for pressure vessels used in the refining and
petrochemical industry, as well as other industries. However, after years of modifications and
updates, VIII-1 suffers from poor organization, a narrative presentation of rules that has resulted
in an ambiguity of requirements, and an over abundance of appendices (i.e. 34 Mandatory
Appendices and 22 Non-Mandatory Appendices). These characteristics have made the code an
unmanageable document that is confusing to users because the content of many of these
appendices should be incorporated into the body of VIII-1 to facilitate use.
To address the short-comings with VIII-1, the ASME Boiler and Pressure Vessel Standards
Committee authorized a project in 1998 to rewrite the Section VIII, Division 2 pressure vessel
code to address user-friendliness, introduce new technologies and to produce more competitive
designs. This decision was made shortly after the design margin on specified minimum tensile
strength was lowered from 4.0 to 3.5 in Section I and Section VIII, Division 1. Instead of
revising the existing Section VIII, Division 2, the decision was made to perform a clean sheet
rewrite. By doing so it was felt that not only could the standard be modernized with regard to the
latest technical advances in pressure vessel construction, but it could be structured in such a way
as to make it more favorable to both users and the committees that maintain it.
ASME B&PV Code, Section VIII - Division 2
The scope of the AMSE B&PV Code, Section VIII - Division 2 Rules for Construction of
Pressure Vessels – Alternative Rules (VIII-2) [6] closely follows that of VIII-1. The
requirements of VIII-2, are contained in the nine Parts listed below. Each of these Parts and
related Annexes is composed of paragraphs that are identified by an alphanumeric numbering
system in accordance with the ISO Standard Template for the Preparation of Normative-Type
Documents.
• Part 1 – General Requirements: provides the scope of VIII-2 and establishes the extent of
coverage.
• Part 2 – Responsibilities and Duties: sets forth the responsibilities of the user and
Manufacturer, and the duties of the Inspector.
• Part 3 – Material Requirements: provides the permissible materials of construction,
applicable material specifications, and special requirements, physical properties,
allowable stresses, and design fatigue curves.
• Part 4 – Design-By-Rule Requirements: provides requirements for design of vessels and
components using rules.
• Part 5 – Design-By-Analysis Requirements: provides requirements for design of vessels
and components using analytical methods.
• Part 6 – Fabrication Requirements: provides requirements governing the fabrication of
vessels and parts of vessels.
• Part 7 – Examination and Inspection Requirements: provides requirements governing the
examination and inspection of vessels and parts of vessels.
INDEX çç Sommaire
• Part 8 – Pressure Testing Requirements: provides pressure testing requirements for
fabricated vessels.
• Part 9 – Pressure Vessel Overpressure Protection: provides rules for pressure relief
devices.
The organization within each part is as follows:
• Rules and requirements organized in paragraphs using the ISO numbering system
• Nomenclature
• Tables
• Figures
• Normative Annexes (mandatory)
• Informative Annexes (non-mandatory)
Mandatory and non-mandatory requirements are provided as Normative and Informative
Annexes, respectively, to the specific Part under consideration. The Normative Annexes address
specific subjects not covered elsewhere in this Division and their requirements are mandatory
when the subject covered is included in construction under VIII-2. The Informative Annexes
provide information and suggested good practices.
Unlike all of the other ASME BPV Standards, VIII-2 was published in single column format,
which facilitates use of the standard in electronic form. A detailed Table of Contents precedes
each Part, and each Part is numbered independently.
An overview of the enhancements made to VIII-2 is provided by Pastor, et al. [7]. A criteria and
companion document was prepared by Osage [8] covering all aspects of Section VIII, Division 2,
including development of new material toughness requirements, new Design-By-Rule
procedures, new Design-By-Analysis procedures, and description of changes in fabrication and
examination requirements. An example problem manual covering Parts 3, 4, 5, 6, 7, and 8 has
been prepared by Sowinski, et al. [9].
VIII-2 includes rules for the design of 2.25Cr-1Mo-V materials for temperatures up to 472°C
(900°F) in ASME Code Case 2605. The analysis method included in this code case includes
requirements for a detailed creep analysis using The Materials Properties Council (MPC) Project
Omega Method provided in Part 10 of API 579-1/ASME FFS-1. The method incorporates a
creep-fatigue interaction analysis developed by Prager et al. [10], [11] and places specific
limitations on predicted creep damage at welds and corresponding weld joint location
requirements. This code case removes the temperature limitation for this material that was set at
371°C (700°F) in previous editions of Section VIII, Division 2.
The standards committee that maintains the Section VIII codes, BPV VIII, recognizes that VIII-2
is the most technically advanced and best organized code and has started an initiative to place
common rules in Section VIII within VIII-2. These common rules will then be referenced from
VIII-1 and VIII-3. The only exception to this is that overpressure protection requirements will
reside in VIII-1 and be referenced by VIII-2 and VIII-3. In this context, Common Rules are
defined as those rules in Section VIII-1, VIII-2, and VIII-3 that are identical and difficult to
maintain because of complexity (i.e. either computationally or editorially complex) and/or
frequent updating due to the introduction of new technologies. Common Rules typically occur in
the Design-By-Rule (DBR) or Design-By-Analysis (DBA) parts of the code; but also exist in
material, fabrication, and examination requirements. Core rules for basic vessel design such as
wall thickness for shells and formed heads, nozzle design, etc. will be maintained in VIII-1 and
VIII-3 because these provide sufficient design requirements for many vessels. However, the
more complicated rules such as those for the design of heat exchanger tubesheets will reside in
VIII-2. The effort to resolve common rules will be done in a phased approach. User feedback
will be addressed during each phase to ensure that user friendliness is maintained in VIII-2, and
INDEX çç Sommaire
enhanced in VIII-1 and VIII-2. Also, as part of this initiative, separate example problem manuals
will be developed for VIII-1 and VIII-3 similar to that developed for VIII-2, and the example
problems will be removed from the body of these codes. When the Common Rules efforts is
completed, it is anticipated that VIII-1 and VIII-3 will be streamlined with references to VIII-2 as
required, and there will be separate example problems manuals for each of the codes that fully
illustrate the use of the construction rules in these codes.
ASME B&PV Code, Section VIII - Division 3
ASME B&PV Code Section VIII - Division 3 Rules for Construction of Pressure Vessels -
Alternative Rules for Construction of High Pressure Vessels (VIII-3) [12] provide rules for the
design, construction, inspection and overpressure protection of metallic pressure vessels with
design pressures generally above 10 ksi (70 MPa). However, it is not the intent of this division to
establish maximum pressure limits for VIII-1 and VIII-2, nor minimum pressure limits for VIII-
3. Specific pressure limitations for vessels constructed to the rules of this division may be
imposed elsewhere in this division for various types of fabrication. Pressure vessels within the
scope of VIII-3 are pressure containers for the retainment of fluids, gaseous or liquid, under
either internal or external pressure. The pressure may be generated by an external source, the
application of heat from a direct or indirect source, a process reaction or through any
combination thereof. VIII-3 is divided into eight Parts and mandatory and non-mandatory
appendices. The eight Parts cover general requirements; material requirements; design
requirements; fabrication requirements; pressure relief devices; examination requirements;
testing requirements; and marking, stamping, reports and records.
ASME B31.3
ASME B31.3 Process Piping (B31.3) [13] covers piping typically found in petroleum refineries;
chemical, pharmaceutical, textile, paper, semiconductor, and cryogenic plants; and related
processing plants and terminals. Construction requirements for materials and components,
design, fabrication, assembly, erection, examination, inspection, and testing of piping are
provided. B31.3 applies to piping for all fluids, including:
• Raw, intermediate and finished chemicals,
• Petroleum products,
• Gas, steam, air and water,
• Fluidized solids,
• Refrigerants, and
• Cryogenic fluids.
Also included within the scope of this code is piping that interconnects pieces or stages within a
packaged equipment assembly.
B31.3 excludes the following:
• Piping systems designed for internal gage pressures at or above zero but less than 105 kPa
(15 psig), provided the fluid handled is nonflammable, nontoxic and not damaging to
human tissues,
• Power boilers in accordance with BPV Code Section I and boiler external piping that is
required to conform to B31.1,
• Tubes, tube headers, crossovers and manifolds of fired heaters, which are internal to the
heater enclosure, and
• Pressure vessels, heat exchangers, pumps, compressors and other fluid handling or
processing equipment, including internal piping and connection for external piping.
INDEX çç Sommaire
API Std 530/ISO 13704
API Std 530/ISO 13704 Calculation of Heater-Tube Thickness in Petroleum Refineries (RP 530)
[14] covers the requirements and provides recommendations for the procedures and design
criteria used for calculating the required wall thickness of new tubes for petroleum refinery
heaters. These procedures are appropriate for designing tubes for service in both corrosive and
non-corrosive applications. The procedures have been developed specifically for the design of
refinery and related process fired heater tubes (direct-fired, heat-absorbing tubes within
enclosures). These design procedures in API Std 530 are not intended to be used for the design
of external piping.
API Std 620
API Std 620 Design and Construction of Large, Welded, Low-Pressure Storage Tanks (API 620)
[15] covers the design and construction of large, welded, low-pressure carbon steel above ground
storage tanks (including flat-bottom tanks) that have a single vertical axis of revolution. This
standard does not cover design procedures for tanks that have walls shaped in such a way that the
walls cannot be generated in their entirety by the rotation of a suitable contour around a single
vertical axis of revolution. The tanks described in this standard are designed for metal
temperatures not greater than 250°F and with pressures in their gas or vapor spaces not more than
15 psig.
The basic rules in API 620 provide for installation in areas where the lowest recorded 1-day
mean atmospheric temperature is –50°F or above. API 620 also covers stainless steel low
pressure storage tanks for refrigerated products at temperatures from +40°F to –60°F, and low-
pressure storage tanks for liquefied hydrocarbon gases at temperatures not lower than –270°F.
The rules in API 620 are applicable to tanks that are intended to (a) hold or store liquids with
gases or vapors above their surface or (b) hold or store gases or vapors alone.
Although the rules in this API Std 620 do not cover horizontal tanks, they are not intended to
preclude the application of appropriate portions to the designs and construction of horizontal
tanks designed in accordance with good engineering practice. The details for horizontal tanks
not covered by these rules shall be equally as safe as the design and construction details provided
for the tank shapes that are expressly covered by API Std 620.
API Std 650
API Std 650 Welded Tanks for Oil Storage (API 650) [16] covers construction requirements for
vertical, cylindrical, aboveground, closed and open top welded carbon or stainless steel storage
tanks in various sizes and capacities for internal pressures approximating atmospheric pressure
(internal pressures not exceeding the weight of the roof plates). However, special provisions are
provided for design of tanks for an internal pressure of up to 17.2 kPa (2.5 psig). Requirements
for materials, design, fabrication, erection, examination and testing are provided. API 650
applies to tanks whose entire bottom is uniformly supported and to tanks in non-refrigerated
service that have a maximum design temperature of 260°C (500°F) or less.
INDEX çç Sommaire
inspection programs for their equipment. As an alternative, these inspections codes, to varying
degrees, permit the use of the inspection planning documents listed below if permitted by the
regulatory body.
• API RP 580
• API RP 581
• ASME PCC-3
Owners and users of pressure equipment can use the above inspection codes, in combination with
other documents as listed below, to evaluate flaws and to plan for repairs and alterations to
pressure equipment if permitted by the regulatory body.
• API 579-1/ASME FFS-1
• ASME PCC-2
API 510
API 510 Pressure Vessel Inspection Code: Maintenance Inspection, Rating, Repair, and
Alteration (API 510) [17] covers the in-service inspection, repair, alteration, and rerating
activities for pressure vessels and the pressure-relieving devices protecting these vessels. This
inspection code applies to most refining and chemical process vessels that have been placed in
service. This includes:
• Vessels constructed in accordance with an applicable construction code;
• Vessels constructed without a construction code (i.e. a vessel not fabricated to a
recognized construction code and meeting no known recognized standard);
• Vessels constructed and approved as “jurisdictional special” based upon jurisdiction
acceptance of particular design, fabrication, inspection, testing, and installation;
• Nonstandard vessels (i.e. a vessel fabricated to a recognized construction code with a
missing nameplate).
Application of API 510 is restricted to owner/users that have access to the following technically
qualified individuals and organizations; an authorized inspection agency, a repair organization,
an engineer, an inspector and examiners.
API 570
API 570 Piping Inspection Code: In-Service Inspection, Rating, Repair, and Alteration of Piping
Systems (API 570) [18] covers the inspection, rating, repair, and alteration procedures for
metallic and FRP piping systems and their associated pressure relieving devices that have been
in-service. The intent of this code is to specify the in-service inspection and condition
monitoring program that is needed to determine the integrity of piping. That program should
provide reasonably accurate and timely assessments to determine if any changes in the condition
of piping could possibly compromise continued safe operation. API 570 was developed for the
petroleum refining and chemical process industries but may be used, where practical, for any
piping system. It is intended for use by organizations that maintain or have access to an
authorized inspection agency, a repair organization, and technically qualified piping engineers,
inspectors, and examiners. API 570 may be used, where practical, for any piping system. Piping
inspectors are to be certified as stated in this inspection code.
API Std 653
The scope of API Std 653 Tank Inspection, Repair, Alteration, and Reconstruction (API 653)
[19] covers minimum requirements for maintaining the integrity of tanks constructed to API 650
and its predecessor API 12C after they have been placed in service, and addresses inspection,
repair, alteration, relocation and reconstruction steel storage tanks. The scope is limited to the
tank foundation, bottom, shell courses, structure, roof, attached appurtenances and nozzles to the
INDEX çç Sommaire
face of the first flange, first threaded joint, or first welding-end connection. Many of the design,
welding, examination and material requirements of API 650 can be applied in the maintenance
inspection, rating, repair and alteration of in-service tanks. In the case of apparent conflicts
between the requirements of this standard and API 650 or its predecessor API 12C, API 653shall
govern for tanks that have been placed in service. This standard employs the principles of API
650; however, storage tank owner/operators, based on consideration of specific construction and
operating details, may apply this standard to any steel tank constructed in accordance with a tank
specification.
NB-23
NB SM-23 National Board Inspection Code (NBIC) [20] provides basic rules that are required in
many jurisdictions (see NB-370 for specific jurisdictional requirements) for installation, in-
service inspection, repair and alteration of boilers, pressure vessels, piping, and pressure-relief
devices. The three primary Parts of NB-23 are:
• Part 1 – Installation
• Part 2 – Inspection
• Part 3 – Repairs/Alterations
The NBIC provides administrative requirements for accreditation of repair organizations and
owner-user inspection organizations. Accredited organizations receive a Certificate of
Authorization and can apply an “R” stamp to pressure equipment that they have repaired as an
indication that the repairs conform to the requirements of the NBIC. The NBIC also provides:
• Installation requirements that impose requirements for new construction such as
accessibility for inspection,
• Precautions for conducting inspections (e.g., vessel entry requirements),
• General guidance on examination techniques and pressure testing,
• Descriptions of a limited number of damage mechanisms,
• Specific inspection requirements for boilers by boiler type,
• General requirements for inspection of pressure vessels and piping,
• Requirements for inspection and repair of overpressure protection devices,
• Methods for determining inspection intervals,
• General requirements for repairs and alterations and re-rating, and
• Repair, alteration and inspection of fiber-reinforced plastic equipment.
API RP 580
API Recommended Practice (RP) 580 Risk-Based Inspection (API 580) [21] provides users with
the basic elements for developing, implementing and maintaining a Risk-Based Inspection
program (RBI) program for fixed equipment and piping in the hydrocarbon and chemical process
industries. API 580 provides guidance to owners, operators, and designers of pressure-
containing equipment for developing and implementing an inspection program. These guidelines
include means for assessing an inspection program and its plan. The approach emphasizes safe
and reliable operation through risk-prioritized inspection. A spectrum of complementary risk
analysis approaches (qualitative through fully-quantitative) can be considered as part of the
inspection planning process. API 580 is intended to supplement API 510, API 570, API 653 and
NBIC. These API inspection codes and standards allow an owner/user latitude to plan an
inspection strategy and increase or decrease the code-designated inspection frequencies based on
the results of a RBI assessment.
API RP 581
API RP 581 Risk-Based Inspection Technology (API 581) [22] provides quantitative procedures
to establish an inspection program using Risk-Based methods for pressurized fixed equipment
INDEX çç Sommaire
including pressure vessel, piping, tankage, pressure relief devices, and heat exchanger tube
bundles. API 580 provides the minimum requirements and the corresponding basic elements for
developing, implementing and mainrtaining an RBI program to establish inspection plans for
fixed equipment in the refinery and petrochemical industry. However, specific details and
calculation methods are not provided in API RP 580. API 581 provides specific details and
calculation procedures based on a fully-quantitative risk analysis approach for establishment of
inspection plans for fixed equipment that are in compliance with all requirements and basic
elements defined in API RP 580.
The API RBI methodology is presented in a three Part volume.
• Part 1 – Inspection Planning Using API RBI Technology
• Part 2 – Determination of Probability of Failure in an API RBI Assessment
• Part 3 – Consequence Modeling in API RBI
The methods used to obtain an inspection plan are provided in Part 1 for fixed equipment
including pressure vessels, piping, atmospheric storage tanks, pressure relief devices, and heat
exchanger tube bundles. The pressure boundaries of rotating equipment may also be evaluated
using this Part. The probability of failure for fixed equipment is covered in Part 2. The
probability of failure depends on the component type and damage mechanisms present based on
the process fluid characteristics, design conditions, materials of construction, and the original
construction code. Part 3 provides methods for computing the consequence of failure. Two
methods are provided. The first method, or Level 1, is based on closed form solutions generated
for a limited set of reference fluids or fluid groups. The second method, Level 2, is a general,
more rigorous method that can be used for any fluid stream composition.
INDEX çç Sommaire
The first edition of API 579 Recommended Practice for Fitness-For-Service was published in
2000. API and ASME agreed to form a Fitness-For-Service Joint Committee (FFSJC) to produce
a co-branded API/ASME FFS second edition of this standard, designated as API 579-1/ASME
FFS-1 2007 Fitness-For-Service. API 579-1/ASME FFS-1 is based on the first edition of API
579, incorporates all planned technical enhancements originally slated for the second edition, and
also includes modifications to address the special needs of other industries such as the fossil
electric power industry, and the pulp and paper industry. An overview of API 579-1/ASME
FFS-1 2007 is provided by Osage [24].
Alternate FFS approaches are permitted for more advanced practitioners by using the Level 3
assessment procedures in API 579-1/ASME FFS-1. For example, the Level 3 assessment in Part
9 covering crack-like flaws provides references to Nuclear Electric R-6 [25], BS 7910 [26],
SAQ/FoU-Report 96/08 [27], WES 2805 [28], EPRI J-Integral Methodology [29], and A16 [30].
In addition, The European Community sponsored a project known as FITNET to review the
existing FFS procedures and develop an updated, unified and verified European FITNET FFS
Procedure [31], [32], [33] to cover structural integrity analysis to avoid failures due to fracture,
fatigue, creep and corrosion. It should be noted that the FITNET FFS Procedure is not a
standard; however, the procedures in this document may be used under the Level 3 assessment
procedures in API 579-1/ASME FFS-1, as applicable.
The current in-service inspection codes typically reference the ASME and API construction
codes to address mechanical integrity issues. While some of the technology is applicable, this is
not always the best approach because these codes are optimized for new construction. In-service
mechanical integrity issues are better addressed in the API 579-1/ASME FFS-1. API 579-
1/ASME FFS-1 was developed to: supplement the requirements of the in-service inspection
codes; ensure safety of plant personnel and the public while older equipment continues to
operate; provide technically sound FFS assessment procedures; ensure that different service
providers furnish consistent remaining life predictions; and help optimize maintenance and
operation of existing facilities to retain availability of older plants and enhance long-term
economic viability. The benefits of having a comprehensive FFS document that is tightly
integrated with the in-service inspection codes are:
• Ease of use in assessing flaws and damage mechanisms including jurisdictional
acceptance,
• Extended safe operation of damaged equipment based on industry accepted assessment
methods,
• Flexibility in developing tactics for repair and/or replacement of damaged equipment,
• New basis for inspection planning, and
• Turnaround support decision making with a goal to minimize turnaround scope and
length.
FFS and RBI are complementary technologies. The FFS and Risk-Based Inspection (RBI)
relationship depends upon the type of RBI implementation being performed. In a RBI
implementation using a qualitative evaluation, FFS assessment procedures can be used to alter
the risk-ranking of equipment based on the level of damage and the results of the assessment. In
a RBI implementation using a quantitative evaluation, the FFS assessment procedures provide a
damage model that can be used to establish a probability of failure as shown by Osage [24]. The
probably of failure results can be combined with a consequence of failure model to produce risk,
which can be utilized to develop an inspection plan as described by Osage and Henry [34].
Inspection planning involves determining the scope, method of inspection, and inspection
interval for a piece of equipment.
INDEX çç Sommaire
ASME POST CONSTRUCTION PUBLICATIONS
Overview
ASME saw the need to address in-service issues and in the mid 1990’s created the Post
Construction Committee (PCC). The charter of this committee is to develop and maintain
standards addressing common issues and technologies related to post construction activities, and
to work with other consensus committees in the development of separate, product specific, codes
and standards addressing issues encountered after initial construction for equipment and piping
covered by Pressure Technology Codes and Standards. Documents issued by the Post
Construction Committee to date include:
• ASME PCC-1 2000
• ASME PCC-2 2006
• ASME PCC-3 2007
• ASME PTB-3 2009
ASME PCC-1
ASME PCC-1 Guidelines For Pressure Boundary Bolted Flange Joint Assembly (PCC-1) [35]
provides bolted flange joint assembly guidelines for pressure-boundary flanged joints with ring-
type gaskets that are entirely within the circle enclosed by the bolt holes and with no contact
outside this circle.
ASME PCC-2
ASME PCC-2 Standard for the Repair of Pressure Equipment and Piping (PCC-2) [36] provides
methods for repair of equipment and piping (including pipelines) and piping components (such as
valves), boilers, pressure vessels (including heat exchangers), and storage tanks after they have
been placed into service. These repair methods include relevant design, fabrication, examination,
and testing practices and may be temporary or permanent, depending on the circumstances. The
methods provided in PCC-2 address the repair of components when repair is deemed necessary
based on appropriate inspection and flaw assessment. These inspection and flaw evaluation
methods are not covered in this document, but are covered in the in-service inspection codes and
API 579-1/ASME FFS-1. Only technical procedures and information are provided;
administrative or policy requirements are outside of the scope of this Standard. PCC-2 is divided
into five Parts.
• Part 1 covers the scope, organization, and intent and is applicable to all articles in this
Standard.
• Part 2 covers repair methods and techniques that include the use of welding, brazing,
soldering, or other methods involving metal deposit.
• Part 3 covers mechanical repairs, with or without sealant, such as bolted clamps or
fixtures and includes all repair methods not covered in Part 2 or Part 4.
• Part 4 covers repairs using nonmetallic means, such as nonmetallic liners and wraps, and
bonding (e.g., joining by epoxy), including bonding of metallic components.
• Part 5 covers examination and testing methods and techniques.
ASME PCC-3
ASME PCC-3 Inspection Planning Using Risk-Based Methods (PCC-3) [37] provides
information on using risk analysis to develop and plan an effective inspection strategy. ASME
PCC-3 is a derivative work of API 580 that was developed to cover pressurized equipment in
industries other than refining and petrochemical. Inspection planning is a systematic process that
begins with identification of facilities or equipment and culminates in an inspection plan. Both
the probability of failure and the consequence of failure should be evaluated by considering all
INDEX çç Sommaire
credible damage mechanisms that could be expected to affect the facilities or equipment. In
addition, failure scenarios based on each credible damage mechanism should be developed and
considered. The output of the inspection planning process conducted according to these
guidelines should be an inspection plan for each equipment item analyzed that includes:
• Inspection methods that should be used,
• Extent of inspection (percent of total area to be examined or specific locations),
• Inspection interval,
• Other risk mitigation activities, and
• The residual level of risk after inspection and other mitigation actions have been
implemented.
ASME PTB-3
The ASME Post Construction Committee recently published ASME PTB-3 Guide to Life-Cycle
Management of Pressure Equipment Integrity (PTB-3) [38]. This document provides a
“roadmap” to help users of pressure equipment and their designated agents, as well as
manufacturers, owners, regulators and other stakeholders, identify the codes, standards,
recommended practices, specifications and guidelines that apply to the Life-Cycle Management
of pressure equipment integrity. PTB-3 provides a summary of some of the more commonly
used documents (i.e. codes, standards, recommended practices (RPs), specifications and
guidelines) produced by organizations based in the United States that are used for LCM of
pressurized equipment. The following applications for pressure equipment are not specifically
included in the scope. However, the owner of these categories of equipment may use those
portions of this guide that are applicable:
• Upstream “non-process” equipment in the oil and gas industry (e.g., pressure equipment
used in oil and gas exploration and production such as Christmas trees, wellhead
equipment, flow lines, subsea equipment),
• Equipment in commercial nuclear power plants,
• Domestic plumbing and other domestic pressure equipment such as hot water heaters,
• Portable air receivers (air tanks) used by homeowners and contractors are excluded from
the scope but air receivers in industrial facilities are included,
• Liquefied natural gas (LNG) and liquefied petroleum gas (LPG) transport and storage
(API and ship classification societies),
• Pipelines, and
• Pressure equipment used in transport service.
In addition, PTB-3 includes only documents that are pertinent to maintaining equipment integrity
(e.g., pressure containment) through appropriate design, construction, inspection, maintenance,
alteration and repair. Standards related to areas such as identification schemes, plant or pipeline
operator qualification, etc. are outside of its scope. PTB-3 is a valuable reference that permits
users of pressure equipment and their designated agents to navigate the maze of documents
required to implement the LCM process. ASME plans to keep this document current and
hopefully will in future editions include other significant industry sectors such as upstream in the
oil and gas industry, and pipelines. In addition, it is hoped that cooperation between international
standard writing organizations could lead to an international version of PTB-3.
INDEX çç Sommaire
construction and in-service margins for application to components that are in-service and have
been evaluated using FFS technology.
New equipment is constructed to codes and standards that have design margins to ensure
structural integrity and safe operation for the specified design conditions. In-service margins
must be developed for FFS assessments to ensure structural integrity and safe operation of
existing equipment with damage. Knowledge of construction codes and standards and validation
studies are required to develop in-service margins.
Design rules in construction codes are typically based on: analytical derivations, test results, and
experience based on evidence of satisfactory performance. In addition, when design rules are
developed, ease of use is balanced with required accuracy.
• High accuracy – Effort and time to implement may be prohibitive.
• Ease of use – Over conservatism may make assessment rules useless or expense.
Design margins in construction codes for pressurized equipment are typically set to prevent
ductile rupture or plastic collapse. For example, in VIII-1 the design margin is used to determine
the allowable stress and for time-independent behavior this margin is the minimum of the
ultimate strength divided by 3.5 and the yield strength divided by 1.5; the ultimate and yield
strength at room temperature and at the design temperature are considered in the analysis. The
factors, such as 3.5 applied to the ultimate strength and 1.5 applied to the yield strength, vary
among the different pressure vessel codes. The other failure modes shown below that are
addressed in construction codes are typically addressed directly or indirectly by rules.
• Brittle Fracture
• Local Strain
• Structural Instability (elastic or plastic buckling)
• Fatigue
• Incremental Collapse (ratcheting)
• Creep Rupture
• Creep Buckling
• Creep-Fatigue Interaction
However, design margins also exist in the design rules or calculation procedures incorporated
into the construction codes. Many of these rules are made conservative for ease of application,
are based on test results with variability, or are service experience based. Since the rules are used
in conjunction with an allowable stress (that already contains a margin), the exact design margin
for a component is difficult to establish. Therefore, design margins may be considered as
follows:
• Explicit Design Margins – Margins that can be directly identified, e.g. margins placed on
material strength parameters, and
• Implicit Design Margins – Margins that result from design rule conservatism for ease of
application, and requirements imposed because of test results and service experience
FFS assessment procedures for in-service components cannot be developed unless both the
explicit and implicit design margins in design codes and standards are fully understood. Many of
the FFS assessment procedures are based on the latest analytical and modeling techniques
because this enables a more detailed assessment of the component so that both the implicit and
explicit design margins can be reduced using a technical basis while still ensuring structural
integrity and safe operation.
As an example of consistency in technology integration pertaining to margins, API 579 2000,
Appendix B, was used as the initial basis for creating VIII-2 2007, Part 5 Design-By-Analysis.
During the balloting process for API 579-1/ASME FFS-1 and VIII-2 2007, comments received
from both standards committees were addressed and resolved such that the resulting API 579-
INDEX çç Sommaire
1/ASME FFS-1 Annex B1 and VIII-2 2007, Part 5 are virtually identical. The main difference is
the accommodation of the design margins that are used in the different international codes.
Furthermore, under the common rules initiative previously discussed, the FFSJC has decided to
remove the common rules from Annex B1 and reference Part 5 directly. This will ensure
consistency in technology.
Another interesting example of consistency in technology integration where an understanding of
explicit and implicit design margins is being evaluated is in the development of new rules by the
Materials Properties Council (MPC) Joint Industry Project (JIP) on FFS. Under this JIP, new
FFS assessment rules are being developed for localized corrosion at nozzles. In order to develop
these rules, an evaluation of the new nozzle rules incorporated into VIII-2, Part 4, paragraph 4.5
is being performed. Once the technology and design margins are fully understood, FFS rules will
be developed to evaluate the local metal loss using the same methods as those used for design.
While the margin against burst will be reduced because of the localized metal lost, an acceptable
in-service margin can still be developed and applied if the vessel is subject to an inspection
program in accordance with the applicable in-service inspection codes.
Consistency in technology also applies to other areas of construction such as materials selection
and examination. As previously discussed, a thorough understanding of damage mechanisms is
required to select a material for construction. As another example, records retention of
examination results and delivery to the user is required by VIII-2 to support future in-service
inspections. In many cases, flaws are found during in-service examination, typically performed
using more sensitive examination methods, and it is important to know whether the flaw existed
at the time of examination or if it occurred in-service.
Members of API and ASME committees are working to develop consistency in technology for
construction and in-service codes and standards. The main effort thus far has been; identification
of damage mechanisms, harmonization of Design-By-Rule and Design-By-Analysis techniques
used in construction codes with the analytical methods used in FFS assessments, and
harmonization of fabrication techniques with those used for repair of in-service damage.
Additional harmonization is possible especially in the areas of construction record retention,
evaluation of material properties, fabrication and welding, and non-destructive examination.
CONCLUSION
A LCM process for fixed pressurized equipment has been defined for the refinery and
petrochemical industry. The fundamental requirements of this process are the understanding and
consideration of potential damage mechanisms in the design process, selection of appropriate
construction codes to ensure reliable designs, establishment of an in-service inspection program
to monitor both anticipated damage and to determine the presence of unanticipated damage
mechanisms, application of FFS technology if unanticipated damage is discovered, and
implementation of effective repair procedures, as required, to ensure mechanical integrity of in-
service equipment. ASME and API are committed to providing manufacturers and owner-users
of pressure equipment with the most up-to-date, consistent technology delivered in easy to use
codes and standards for both construction and in-service applications, and that address these
features to serve the international market in the LCM process. The framework of the whole-life
management process that is presented is also applicable to other industries and international
locations by substituting appropriate documents for damage mechanism identification, new
constriction, in-service inspection, Fitness-For-Service, and repair.
INDEX çç Sommaire
BIBLIOGRAPHY
[1] Dobis, J.D. and Bennet, D.C., Damage Mechanisms Affecting Fixed Equipment in the Pulp
and Paper Industry, WRC Bulletin 488, Welding Research Council, New York, N.Y.,
2004.
[2] Dobis, J.D., Cantwell, J.E., and Prager, M., Damage Mechanisms Affecting Fixed
Equipment in The Refining Industry, WRC Bulletin 489, Welding Research Council, New
York, N.Y., 2004.
[3] Dobis, J.D. and French, D.N., M., Damage Mechanisms Affecting Fixed Equipment in
Fossil Electric Power Industry, WRC Bulletin 490, Welding Research Council, New York,
N.Y., 2004.
[4] API, API 571 Damage Mechanisms Affecting Fixed Equipment in the Refining Industry,
American Petroleum Institute, Washington, D.C.
[5] ASME, ASME Boiler and Pressure Vessel (B&PV) Code, Section VIII - Division 1 Rules
for Construction of Pressure Vessels, ASME, New York, N.Y.
[6] AMSE, ASME B&PV Code, Section VIII - Division 2 Rules for Construction of Pressure
Vessels – Alternative Rules, ASME, New York, N.Y.
[7] Pastor, T.P. and Osage, D.A., “Modernization of Pressure Vessel Design Codes ASME
Section VIII, Division 2, 2007 Edition,” ESOPE Conference, Paris, France, 2007.
[8] Osage, D.A., ASME Section VIII – Division 2 Criteria and Commentary, ASME PTB-1-
2009, ASME, New York, N.Y.
[9] Sowinski, J. and Osage, D.A., ASME Section VIII – Division 2 Example Problems Manual,
ASME PTB-3-2010, ASME, New York, N.Y.
[10] Prager, M. and Osage, D.A., “Background To The Development of ASME B&PV Code
Case 2605 – Extension of 2.25Cr-1Mo-V Alloy into the Creep Range,” ESOPE
Conference, Paris, France, 2010.
[11] Prager, M., “Extend Low Chrome Steel Fatigue Rules,” ASME STP-PT-027, ASME, New
York, N.Y.
[12] AMSE, ASME B&PV Code Section VIII - Division 3 Rules for Construction of Pressure
Vessels - Alternative Rules for Construction of High Pressure Vessels, ASME, New York,
N.Y.
[13] ASME, ASME B31.3 Process Piping, ASME, New York, N.Y.
[14] API, API Std 530/ISO 13704, Calculation of Heater-Tube Thickness in Petroleum
Refineries, American Petroleum Institute, Washington, D.C.
[15] API, API Std 620 Design and Construction of Large, Welded, Low-Pressure Storage
Tanks, American Petroleum Institute, Washington, D.C.
[16] API, API Std 650 Welded Tanks for Oil Storage, American Petroleum Institute,
Washington, D.C.
[17] API, API 510 Pressure Vessel Inspection Code: Maintenance Inspection, Rerating, Repair
and Alteration, American Petroleum Institute, Washington, D.C.
[18] API, API 570 Piping Inspection Code: Inspection, Repair, Alteration, and Rerating of In-
Service Piping Systems, American Petroleum Institute, Washington, D.C.
[19] API, API 653 Tank Inspection, Repair, Alteration, and Reconstruction, American
Petroleum Institute, Washington, D.C.
[20] NBIC, NB-23 National Board Inspection Code, The National Board, Columbus Ohio, 2004.
[21] API, API RP 580 Recommended Practice for Risk-Based Inspection, American Petroleum
Institute, Washington, D.C.
INDEX çç Sommaire
[22] API, API RP 581 API Risk-Based Inspection Technology, American Petroleum Institute,
Washington, D.C.
[23] API, API 579-1/ASME FFS-1 2007 Fitness-For-Service, American Petroleum Institute,
Washington, D.C.
[24] Osage, D.A. “API 579-1/ASME FFS-1 2007 – A Joint API/ASME Fitness-For-Service
Standard for Pressurized Equipment”, ESOPE Conference, Paris, France, 2007.
[25] Nuclear Electric, Assessment of the Integrity of Structures Containing Defects, Nuclear
Electric R-6, Nuclear Electric, 1998.
[26] BSI, Guide on Methods for Assessing the Acceptability of Flaws in Structures, BS 7910,
British Standards Institute, 1999.
[27] SAQ/FoU, A Procedure for Safety Assessment of Components with Cracks – Handbook,
SAQ/FoU-Report 96/08, 1997.
[28] The Japan Welding Engineering Society, Method of Assessment for Flaws in Fusion
Welded Joints with Respect to Brittle Fracture and Fatigue Crack Growth, The Japan
Welding Engineering Society, WES 2805, 1997.
[29] Kumar, V., German, M.D., Shih, C.F., An Engineering Approach for Elastic-Plastic
Fracture Analysis, EPRI Report NP-1931, EPRI Palo Alto, CA, 1981.
[30] CEA, A16: A French Proposal for Fracture Assessment Methods at Elevated Temperature,
CEA Report DMT/97/394, September 19, 1997.
[31] Kocak, Mustafa, “Fitness-For-Service Analysis of Structures Using the FITNET Procedure:
An Overview”, proceedings of OMAE2005, 24th International Conference on Offshore
Mechanics and Arctic Engineering (OMAE 2005), June 12-17, 2005, Halkidiki, Greece.
[32] Kocak, M., Hadley, I., Szavai, S., Tkach, Y. and Taylor, FITNET Fitness-For-Service
(FFS) Procedure – Volume 1, Revision MK8, GKSS Research Centre, Greesthacht,
Germany, Jan 2008.
[33] Kocak, M., Hadley, I., Szavai, S., Tkach, Y. and Taylor, FITNET Fitness-For-Service
(FFS) Annex – Volume 2, Revision MK8, GKSS Research Centre, Greesthacht, Germany,
Jan 2008.
[34] Osage, D.A. and Henry, P.A. “Recent Developments and Technology Improvements in API
Risk-Based Inspection Planning Technology”, ESOPE Conference, Paris, France, 2007.
[35] ASME, ASME PCC-1 2000, Guidelines for Pressure Boundary Bolted Flange Joint
Assembly, ASME, New York, N.Y.
[36] ASME, ASME PCC-2 2006, Repair of Pressure Equipment and Piping Standard, ASME,
New York, N.Y.
[37] ASME, ASME PCC-3 2007, Inspection Planning Using Risk-Based Methods, ASME, New
York, N.Y.
[38] Sims, R., Guide to Life Cycle Management of Pressure Equipment Integrity, ASME PTB-3-
2009, ASME, New York, N.Y.
INDEX çç Sommaire
Specify Design Conditions and Identify
Damage Mechanisms (API 571,
WRC 488, WRC 489, WRC 490),
Select Materials of Construction
Construction Code
ASME VIII-1, VIII-2, VIII-3, B31.3
API 530, 620, 650
Commissioning
(Baseline Inspection)
In-Service Inspection
(Establish Inspection Interval)
Prescriptive (API 510/570/653,NBIC)
Risk-Based (API 580/581,PCC3)
Continue Service
Technology Integration
Anticipated
Damage
Inspection
Results
Unanticipated
Damage
Fitness-For-Service
API 579/ASME FFS-1
ASME PCC2
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page93
RÉSUMÉ / ABSTRACT
86E
RECHARGEMENT SUR SITE D’UNE COLONNE DE DISTILLATION AVEC
REVÊTEMENT ANTI-CORROSION
C. Narjoz
AQUILEX WELDING SERVICES B.V. Marconiweg 16 – 3225 LV – HELLEVOETSLUIS
PAYS-BAS / THE NETHERLANDS
RÉSUMÉ
Cas réel d’un rechargement sur site d’une colonne de distillation avec un revêtement en acier inoxydable
de type 317L. L’objectif de la présentation est de reprendre les étapes depuis l’étude de faisabilité à la
réalisation en passant pas l’étude des déformations par analyse par éléments finis, l’optimisation du
planning d’exécution et le suivi sur site durant l’arrêt programmé. Les aspects métallurgiques et
soudages seront développés.
ABSTRACT
Case study of weld overlay, with stainless steel layer 317L grade, on site inside a distillation column.
The objectives of the presentation are to describe all steps from feasibility study till execution. So this
will include distortion prediction using finite elements analysis, optimization of schedule and follow
up on site during the outage.
Metallurgical and welding aspects will be developed.
INDEX çç Sommaire
INEOS CASE STORY
Overlay DA101 column
by C.Narjoz, Technical Director Europe
DA101 June 2010 SHOW CASE
• INEOS
• AQUILEX Welding Services B.V.
• DA101 Challenges
• Project Information
• AWS Solution
• Comparison with replacement by windows
• Steps to success
– INEOS is the world’s third largest chemical company comprising 17
businesses each with major chemical company heritage. Production network
spans 64 manufacturing facilities in 14 countries throughout the world.
– u
• Founded in 1978: orbital welding for nuclear ind.
• 1985: non nuclear services
• 1996: First projects executed in Europe
• 1998: offices in Rotterdam
• 2002: WSI join Aquilex group
• 2004: creation of shop in Poland
• 2008: AWS move to Hellevoetsluis (NL) Corporate Headquarters
Atlanta Georgia
• 2009: SMS et WSI deviennent Aquilex®SRO
Key figures:
• Annual Sales 2009: WSI: 264M$ (SRO incl. AWS: 21,7M$)
• Human Resources: WSI: 210 persons + ~1100 welders
AWS: 28+71 persons + ~50 welders
• Equipments Resources: WSI: more than 300 welding robots
AWS: ~50 welding robots European Headquarters
• Over 300.000m² of overlay experience Hellevoetsluis Pays-Bas
DA101 Challenges
Main distillation column
– Reliability is critical as it feeds most of the rest of
the refinery
– Feed stock and working conditions have changed
• Options considered:
– Partial Replacement:
– Section replacement
– Windows replacement
– Unifuse® Weld Overlay
Project information
– Atmospheric distillation column build in 1971 according to ASME code
– Column dimensions: 7,8m/8,3m with total height of ~66m
– Column partially already cladded
– During June 2010 TU, both project conducted in parallel:
• Replacement by windows
• Weld overlay
Project Information
Scope of work:
- Internal overlay with Stainless steel 317L, ~55m²
- Welding of rings, and tray supports
- Internal overlay of piping, and flanges surfaces located in the
overlaid area
Distortion of vessel was a concern and have to remain within code criteria
=> FEA
The site is on the coast with regular wind coming from inland.
AWS Solution: Finite Element Analysis
FEA performed to:
Distortion Mitigation
Stress Analysis
FEA RESULT
As a result of the stress evaluation
outlined in this report, it is
concluded that no structural
compromise exist due to the
proposed internal overlay work. As
part of this effort, the predicted
shell distortion of the DA101
Distillation Column exhibits no out
of roundness issues and is
therefore satisfactory.
AWS Solution - Schedule
Week 19 Week 20 Week 21 Week 22 Week 23
S M T W T F S S M T W T F S S M T W T F S S M T W T F S S M T W T F S
Tasks
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 1 2 3 4 5 6 7 8 9 10 11 12
Bake-out test
Prejob set-up
Inside vessels activities
Extra activities
Replacement by indows
Prejob set-up
Windows replacement
From critical path to support the other paths to match global schedule:
-Inside tie-in on replacement by windows
-Heat exchanger repairs
-CUI repairs
AWS Solution - QA on site
Thickness examination FN Examination PMI Examination
Out-of-roundness/Distortion PT examination
AWS Solution – Flexibility to fix other corrosion issue
• Corrosion under insulation at the top of the column: CS build-up
Safety
No records, no major risk Heavy lifting, risky handling
Steps to success
• Early identification of the opportunity
• Several scenarios considered with client
• References and relationship to secure
client and build trust
• Solid FEA report
• Project management: detailed scheduling
RESULT
• PO awarded 6 months prior to outage
• Excellent execution
• Safety reward
RESULT
• Extension of the overlay during next outage (2015) but other
opportunities on coming unit outages
Steps to success: Technologies & knowhow
Tool
i
ngEngi
neer
ing Mat
eri
al
s&Wel
di
ng Pr
oject
Engi
neer
ing Oper
ati
onsSuppor
t
•Mechani
calSystems • Codes&St andards • Solut
ionDesign • Fl
eetMaint
enance
•Hi
ghDefini
ti
onVi deo • Weldi
ngPr ocesses • Appli
cati
onEng. • Mobi
li
zati
onSt agi
ng
•Cont
rol
s • Corr
osionCoatings • Fi
eldEngineeri
ng • Si
teTechnici
ans
•OEM Modif
icat
ions • Metal
lurgy • ProcessProcedur
es • Syst
em Tes t
ing
FEAModeling Mockup/Tr
aini
ngCenter 600+Procedures 180Aut
omati
cSystems
I
ntegr
ated3DCAD 300+Travel
erLibr
ary Temperbead,E/C,et
c. 75Semi-
AutoSyst
ems
Contr
olsSi
mulat
ion 1050+WelderCerti
fi
cat
ions Level
3NDECapabi l
i
ty 40RemoteVisi
onSets
99B Facteurs de selection d’aciers allies pour appareils à pression CRIPPS P.-R.
contenant de l’hydrogène et travaillant à heute temperature pour SNC-LAVALLIN CHEMICALS
hydrocarbure / Selection factors of alloy steels for pressure vessels AND PETROLEUM
in high temperature hydrocarbon services containing hydrogen
11B Fondements des règles du code case 2605 du B&PV Code de PRAGER M.
l’ASME – Extension de l’utilisation des alliages 2.25Cr1MoV dans THE MATERIALS PROPERTIES
le domaine du fluage / Background to the development of ASME COUNCIL, Inc.
B&PV Code Case 2605 – Extension of 2.25Cr1MoV alloy into OSAGE D.-A.
the creep range THE EQUITY ENGINEERING
GROUP, Inc.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page95
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page96
RÉSUMÉ / ABSTRACT
34B
O. Ancelet
(CEA Saclay – Bat. 607 PC116 – 91191 Gif-sur-Yvette – e-mail : [email protected])
RÉSUMÉ
L’acier Mod 9Cr-1Mo (T91) est un candidat pour la cuve et les structures internes des réacteurs à
caloporteur gaz (GCR). Afin de valider ce choix, il est nécessaire de vérifier si ce matériau est adapté
au milieu et aux opérations de ce type de réacteur, puis de vérifier si il est couvert par les codes de
Mercredi
ABSTRACT
Mod 9Cr-1Mo steel (T91) is a candidate material for pressure vessels and for some internal structures
of GCR (Gas Cooled Reactors). In order to validate this choice, it is necessary, firstly to verify that it
is able to withstand the planned environmental and operating conditions, and secondly to check if it is
covered by the existing design codes, concerning its procurement, fabrication, welding, examination
methods and mechanical design rules. A large R&D program on mod 9Cr-1Mo steel has been
undertaken at CEA in order to characterize the behavior of this material and of its welded junctions.
In this program, the role of the Laboratory for structural Integrity and Standards (LISN) is to develop
high temperature weld joint coefficient and defect assessment procedures under fatigue, creep and
creep-fatigue loadings. Concerning the GCR, complementary studies are conducted in order to validate
the existing methods (developed for the fast reactors) and to get new experimental data on Mod 9Cr-
1Mo steel. Moreover, if the geometry and the loadings of a standard CT specimen allow performing a
2D analysis, the case of industrial loadings appears much more complicated, notably because of surface
defects which propagate and present shapes that can be considered as half ellipse. Therefore, in the
frame of the defect assessment methods validation, the LISN realizes both standard tests on CT
specimens to determine the propagation laws and bending tests on large plates under high temperature
fatigue, creep or creep-fatigue loadings. These components present an initial semi elliptical surface
notch normal to the loading direction and its initiation and propagation are studied.
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
Comportement mécanique des matériaux pour les réacteurs à hautes temperatures : R&D en
support de l’analyse de nocivité des défauts et de l’intégrité des structures
O. Ancelet
(CEA Saclay – Bat. 607 PC116 – 91191 Gif-sur-Yvette – e-mail : [email protected])
ABSTRACT
Mod 9Cr-1Mo steel (T91) is a candidate material for pressure vessels and for some internal structures
of GCR (Gas Cooled Reactors). In order to validate this choice, it is necessary, firstly to verify that it is
able to withstand the planned environmental and operating conditions, and secondly to check if it is
covered by the existing design codes, concerning its procurement, fabrication, welding, examination
methods and mechanical design rules. A large R&D program on mod 9Cr-1Mo steel has been
undertaken at CEA in order to characterize the behavior of this material and of its welded junctions. In
this program, the role of the Laboratory for structural Integrity and Standards (LISN) is to develop
high temperature weld joint coefficient and defect assessment procedures under fatigue, creep and
creep-fatigue loadings. Concerning the GCR, complementary studies are conducted in order to validate
the existing methods (developed for the fast reactors) and to get new experimental data on Mod 9Cr-
1Mo steel. Moreover, if the geometry and the loadings of a standard CT specimen allow performing a
2D analysis, the case of industrial loadings appears much more complicated, notably because of
surface defects which propagate and present shapes that can be considered as half ellipse. Therefore, in
the frame of the defect assessment methods validation, the LISN realizes both standard tests on CT
specimens to determine the propagation laws and bending tests on large plates under high temperature
fatigue, creep or creep-fatigue loadings. These components present an initial semi elliptical surface
notch normal to the loading direction and its initiation and propagation are studied.
RESUME
L’acier Mod 9Cr-1Mo (T91) est un candidat pour la cuve et les structures internes des réacteurs à
caloporteur gaz (GCR). Afin de valider ce choix, il est nécessaire de vérifier si ce matériau est adapté
au milieu et aux opérations de ce type de réacteur, puis de vérifier si il est couvert par les codes de
construction en terme d’approvisionnement, de fabrication, de soudabilité, de contrôle et des règles de
dimensionnement. Un vaste programme de R&D sur l’acier Mod. 9Cr–1Mo a été entrepris par le CEA
afin de caractériser le comportement de ce matériau et de ses soudures. Dans ce programme, le rôle
du Laboratoire d’Intégrité des Structures et de Normalisation( LISN) est de développer les coefficients
de joint et les outils pour l’analyse de nocivité d’un défaut sous un chargement de fatigue, fluage et
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
fatigue-fluage. Concernant les réacteurs GCR, des études complémentaires ont été conduites pour
valider les méthodes existantes (développées à l’origine pour les réacteurs rapides) et de créer une
base expérimentale pour l’acier Mod.9Cr-1Mo. D’autre part, si la géométrie et le chargement des
éprouvettes standard CT permet d’effectuer des analyses ‘simples’ en 2D, le cas des chargements
industriels est plus complexe, notamment en raison d’une géométrie plus complexe des défauts (défaut
surfacique) qui se propagent sous forme semi-elliptique. C’est pourquoi, le LISN réalise des essais sur
éprouvettes standards CT permettant de déterminer les lois de propagation ainsi que des essais de
flexion sur de large plaque contenant un défaut semi-elliptique sous fatigue, fluage et fatigue-fluage à
haute température.
INTRODUCTION
In order to confirm the methods used for analysing the harmfulness at high temperatures of defects in
Type Z10CDVNb 9.1 steel and associated welds, SEMT/LISN ran a series of tests on quasi-
components referred to as PACRO specimens (PACRO is a French-language acronym standing for
“crack propagation in Type 9Cr-1Mo steel”). These were bending tests on large plates (450x350x30
mm) with defects subjected to fatigue and creep loading. Furthermore, to interpret the results of these
tests on structures, calibration tests of the fatigue and creep cracking laws were carried out on
laboratory specimens in parallel.
The purpose of this article is to first state the main Appendix A16 goals, then to describe the
determination of the fatigue and creep propagation laws for Type 9Cr at different temperature and
confirm its validity by tests on large plates with cracks.
To study defect propagation under fatigue and creep, a series of test were carried out
on compact tension specimens (CT specimens), which are shown in Figure 1. These
tests enabled determination of the propagation laws for fatigue and creep. The work
described here was carried out on steel containing 9% chromium at different
temperature between 20°C and 600°C.
The Paris law calibration tests were performed on standard CT specimens with a thickness B of 25 mm
and a width w of 50 mm. These had an initial notch of 24.5 mm in the base metal and 15 mm in the
welded metal and were pre-cracked at room temperature to 27.5 mm (a/w=0.55) in the base metal and
up to 20 mm in the welded metal (a/w=0.4).
The specimens were laterally grooved to obtain a net thickness Bn of 20 mm. For reference, the
purpose of these side grooves was to guarantee a straight crack front and plane strain.
The fatigue tests were carried out at 20°C, 450°C and 550°C. These were conducted at constant load
range ∆F, load ratio R=0.1 and frequency f=0.1Hz to a propagation of 35 mm. The length of pre-
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
cracking (ainitial) and the length of the final crack were measured afterwards on the basis of digital
pictures and numerical area measurement.
Crack progression was monitored with the compliance method. This consists of estimating the length
of a crack from the force/opening curve of the crack.
Opening opposite the loading line was expressed by the following equation [3]:
P
δ2 = f1 (a / w)
E*
1+ a/w 2
f1 (a / w) = ( ) × [2.1630 + 12.219.a / w − 20.065.(a / w) 2
1− a/ w
− 0.9925.(a / w)3 + 20.609.(a / w) 4 − 9.9314.(a / w)5 ] (1)
where P is the load per unit thickness,
E* = E / (1-ν2) where E is Young’s modulus (163 GPa) and ν is Poisson coefficient (0.3).
If the thickness of our specimens is taken to be (B.Bn)0.5, the following can then be written:
δ2 δ exp f 1 (a / w)
= = (2)
P Fexp E * B.Bn
Thus, for a recorded fatigue cycle, compliance (δexp / Fexp) was calculated and equation (2) was
solved for a/w. This gave the variation of crack length as a function of the number of cycles.
In a second stage, the rate of propagation was determined either by using the tangent to the points on
this curve or by calculating the average rate of the ith cycle saved as per equation (3).
da a − ai −1
( ) i = i +1 (3)
dN N i +1 − N i −1
Then variation of the stress intensity factor (∆K) was determined as per ASTM E399-90 with
equations (4) and (5):
∆F
∆K = . f 2 (a / W ) (4)
W . B.Bn
(2 + a / w)
f 2 (a / w) = (0.886 + 4.64(a / w)
(1 − a / w) 3 / 2
13.32(a / w) 2 + 14.72(a / w) 3 − 5.6(a / w) 4 ) (5)
Furthermore, to ensure that the specimen remained in the quasi-elastic domain throughout the test,
only the points that complied with the following equation were used:
4 K max 2
(W − a) ≥ ( ) (6)
π Rp 0.2
Figure 2 shows the results of the different fatigue tests carried out on 9Cr steel that made it possible to
propose the following Paris law at 550°C:
da
= 9.3.10 − 7 .∆K 1.83 at 550°C (7)
dN
(range of validity 7 < ∆K < 37 MPa.m0.5)
da at 450°C (8)
= 9.3.10 −8.∆K 2.33
dN
(range of validity 16 < ∆K < 29 MPa.m0.5)
da
= 1.2.10 −8.∆K 2.75 at 20°C (9)
dN
(range of validity 15 < ∆K < 38 MPa.m0.5)
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
1,E-3 da/dN in
mm/cycle
y = 9E-07x 1,83 1,E-2 da/dt in
550°C m m /cycle Weld
metal
1,E-3
1,E-4
Figure 2. Determination of the Paris Law for the base metal at Figure 3. Determination of the Paris law for the weld metal
550°C, 450°C and 20°C. at 550°C
Figure 3 shows the results of different tests carried out on 9Cr steel weld that made it possible to
propose the following Paris law:
da
= 2.4.10− 7.∆K 2.25 at 550°C (10)
dN
(range of validity 12 < ∆K < 60 MPa.m0.5)
It is to be noted that the difference between the two materials was small: when ∆K=10MPa.m0.5, the
fatigue propagation rate was twice as great in the base metal. The two curves then met when ∆K was
20MPa.m0.5 but when ∆K was 40MPa.m0.5 (the upper bound of the base metal Paris law validity
range), propagation rate da/dN of the deposited metal was faster than that observed in the base metal
by a factor of 1.4. At higher loading, the difference tended to increase. However, at such loading
levels, the behaviour of the material is not totally elastic. There was therefore a need to re-analyse the
results of the tests with allowance for the effect of plasticity.
The creep cracking law calibration tests were performed using standard CT specimens with a thickness
of 25 mm with properties similar to those used for fatigue. All the specimens were fitted with electrical
wires to measure the potential difference across the notch root. By establishing a ∆a = f(∆ddp)
calibration curve using the same tests, the potential difference readings could be used to monitor the
growth of the crack over time in each of the tests. Opening opposite the loading line was monitored
with a LVDT-type sensor. The creep tests on CT specimen were performed following the ASTM
standard [4].
Estimation of propagation rate da/dt
The fisrt stage of analysis consisted in determining the crack growth during the test. Crack length was
determined by potential difference measurement. As the initial and final lengths of the crack were
accurately known (due to post-cracking), a ∆a=f(∆ddp) (linear function) correlation was established
for each test.
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
It was noticed that, in each test, the potential difference signal increased from the outset, preventing
proper determination of the incubation time.
The propagation rate was then calculated, either by deriving smoothing a(t) or by calculating an
average instantaneous rate determined with the following equation:
da a − ai −1
( )i = i (11)
dt t i − t i −1
In all the analysis, we calculated a smoothing of the instantaneous rate. We generally found that the
variation of the propagation rate was linear.
Estimation of crack opening rate
The third stage of the analysis consisted of determination of the crack opening rate. This is an
important step because crack opening rate was used in calculating C*(t). This involved using as the
starting point the experimentally measured opening of the crack at the loading line. This comprised an
elastoplastic term due to the loading, a second term due to the growth of the crack through the
structure corresponding to the part associated with variation of the compliance, referred to as
structural, and a term due to actual creep, referred to as behaviour. It was then possible to write:
δ tot = δ loading + δ structure + δ creep (12)
where
δtot = opening measured during the test
δloading = elastoplastic opening measured after loading
δcreep = opening due to creep
δstructure = opening due to crack growth in the structure (structure).
For the purposes of our analysis, we opted to calculate the instantaneous average rate expression:
dδ tot δ tot (i ) − δ tot (i −1)
( )i = (13)
dt t i − t i −1
We then determined the part of the opening rate due to progress of the defect through the structure
with the following equation:
dδ structure da Bn 2 K I
2
= [ * + (n + 1) J p ] (14)
dt dt P E
where E* = E/(1-ν2) on the basis of the plane strain hypothesis
P a
KI = f 2 ( ) (15)
BBnW w
a
Where f 2 ( ) defined in equation (5)
w
It is important to state that, in all the tests, the base metal and the deposited metal Jp was null. This is
because, unlike what has been observed in the case of Type 316 austenitic steels, the elastic limit is
very high, compared to the level of loading applied in creep to obtain reasonable fracture times in the
laboratory (around 1,000 hours). This gives rise to a response in terms of opening to crack propagation
in the structure that remains elastic. Also, it was observed that this part of the opening associated with
progress of the crack through the structure (structure) was small compared to the part due to creep
(behaviour). The total rate in the experiment was thus virtually the same as the ‘behaviour’ rate.
Determination of C*(t)
The last stage of the analysis was to calculate C* with the following equation:
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
.
P δ creep n2 a
C* = [2 + 0.522(1 − )] (16)
B net (W − a) n 2 + 1 W
It needs to be stated that here we call C*(t) the value of C* calculated with this equation, using all the
instants of the test, without excluding the parts outside secondary creep. We kept “C*” for the
parameter calculated with this equation, for minimal values of opening rate δcreep. It will be noticed that
the instants at which the opening rate is at a minimum also correspond to the minimum values of C*(t).
1,E-2
Weld
1,E-03
550°C
1,E-3 y = 0,004x 0,601
Base metal
C* in N/mm.h C* in N/m m h
1,E-4 1,E-04
0,001 0,01 0,1 1 10 0,001 0,01 0,1
Figure 4. Determination of the creep propagation law for the base Figure 5. Determination of the creep propagation law for the weld
metal at 550°C and 600°C metal at 550°C
F. Curtit’s suggestion [5] for construction of the da/dt-C* correlation was used as a basis.
The latter argued in his PhD that, for each test, only the point corresponding to the lowest opening rate
should be adopted. The stage is then the onset of generalised secondary creep, and the structure term
is still low compared to the creep-related opening rate. A correlation was then established by power
law regression for a cloud of points, each of which corresponds to a test (Figure 4).
Analysis of creep propagation test results made it possible to propose the following creep propagation
law:
da
= 4.10 −3.C *0.6 at 550°C (17)
dt
(range of validity 0.003 < C* < 1N/mm.h)
da
= 6.1.10 −3.C *0.6 at 600°C (18)
dt
(range of validity 0.1 < C* < 10N/mm.h)
However it can be noticed that the lower limit of the range of validity would be too high to be tolerated
on operating plant and so some extrapolation should be made.
The Figure 5 presents the determination of the creep propagation law for the weld metal at 550°C.
Analysis of creep propagation test on welded metal results made it possible to propose the following
creep propagation law:
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
da (19)
= 1.24.10 − 2.C *0.6
dt
(range of validity 0.006 < C* < 0.1N/mm.h)
It can also be seen in the figure 5 that, for a given value of C*, the rate of creep propagation was 3.8
times faster in the deposited metal than the base metal.
A radiant furnace was used to raise the temperature to 550°C in a region extending 95 mm from either
side of the crack along the entire length of the specimen.
The instrumentation of the specimens and the test bed made it an easy matter monitor the load and the
displacement produced by the press, the plate deflection, the crack opening angle (CMOA) using two
laser beam layers separated by DN=50 mm, the arms rotation and the potential drop difference.
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
The plate tests were modelled with CAST3M [7]. The mesh was 3D with a semi-elliptical defect
(Figure 7) comprising 41,658 nodes and 9,379 Cu20 elements.
As only the central part of the plate was heated, only the hot part corresponding to a height of 250 mm
was modelled. Furthermore, for reasons of Fatigue test Fatigue test
symmetry, on one quarter of this region was Experiments
1 2
modelled (Figure 7). Plate in Plate with
Material
base metal weld joint
Table 1:Fatigue experiments on cracking plate
characteristics ∆F in kN 47.1 47.1
Arm of
350 350
The loading applied was a pure bending moment. lever in mm
Two fatigue tests were carried out at 550°C on the Bending
cracked plates (Table V): moment in 16.5 16.5
• the first test was carried out on a plate kN.m
made only of base metal. R 0.1 0.1
• The second test was carried out on a plate with a welded joint in the middle. The defect was
machined at the joint.
The features of the two tests are given in the following table Meshing of the initial crack
1.
To correctly model these tests, a model integrating allowance
for crack propagation during the test was devised. The Meshing of the cracked plate
Figure 8: Flow chart of the program of propagation Calculation of KI for the front of the crack
Figure 9 compares the experimental and numerical changes in the size of the defect as a function of the
number of cycles in each test. It can be seen that the numerical results are in excellent agreement with
the experimental data, as concerns both the depth of the defect (designated a) and its length at the
surface (designated 2c).
2.c
a 2.b
L
(a
45
20
a or c in )
c (experimental
a or c in c (experimental mm
40 mm data)
35
15
30 c c
25
a 10
a
20
15
5
10 a (experimental data)
5
a (experimental data)
Num ber of cycles
Num ber of cycles 0
0
0 2000 4000 6000 8000
0,E+0 1,E+4 2,E+4 3,E+4 4,E+4 5,E+4 6,E+4
(b (c
Fig. 9. Comparison of the experimental and numeric evoution of) the crack’s length during the fatigue experiments - (a) Definition
) of a
and c – (b) Fatigue test 1 – (c) Fatigue test 2.
In Figure 10, a picture of the fracture faces is shown, the different fronts corresponding to the number
of cycles of the different markings calculated numerically. It can also be seen that the agreement
between the experimental and numerical data is excellent.
This modelling confirms the Paris laws mentioned in §2.1 and, above all, its transposability to a
surface defect.
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
(a (b
) fatigue test – (a) fatigue test 1 – (b) Fatigue test 2
Fig. 10. Observation of the crack for the
)
Conclusion
In order to confirm the methods used for analysing the harmfulness at high temperatures of defects in
Type Z 10 CDVNb 9.1 steel and associated welds, CEA/LISN ran a series of tests on quasi-
components referred to as PACRO specimens. These were bending tests on large plates (450x350x30
mm) with defects subjected to fatigue, creep loading. To interpret the results of these tests on
structures, calibration tests of the fatigue and creep cracking laws were carried out on laboratory
specimens in parallel.
A comprehensive series of propagation tests in the fatigue and creep domains enabled determination of
the creep (da/dt- C*) and fatigue (da/dN-∆K) propagation laws for 9% chromium steel and its welds at
for different temperatures.
Two fatigue tests were made on large cracked plates. The defects used in these tests were surface ones
with the advantage of been close to the geometry or real ones. Complete modelling of these tests
showed the validity and the transferability to surface defects of the Paris laws determined on standard
CT specimens.
BIBLIOGRAPHIE
1. B. Drubay, S. Marie, S. Chapuliot, M.H. Lacire, B. Michel, H. Deschanel, ‘A16: Guide for
defect assessment at elevated temperature’, International Journal of Pressure Vessel Piping
2003;80:499-519.
2. RCC-MR 2007 "Design and Construction Rules for Mechanical Components of Nuclear
Installations" , 2007 AFCEN Code, Association Française pour les Règles de Conception et de
Construction des chaudières Electro-Nucléaires. www.afcen.com.
3. H. Tada, P. C. Paris, G.R. Irwin, “The stress analysis of cracks handbook”, third Edition –
ASME Press, 2000.
4. ASTM E 1457-98 “Standard test method for measurement of creep crack growth rates in
metals”
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page97
RÉSUMÉ / ABSTRACT
99B
FACTEURS DE SELECTION D’ACIERS ALLIES POUR DES APPAREILS A
PRESSION CONTENANT DE L’HYDROGENE ET TRAVAILLANT A HAUTE
TEMPERATURE POR HYDROCARBURES
P. R. Cripps
(SNC-Lavalin Inc., Chemicals and Petroleum,
605 5 Avenue S.W. Calgary, Alberta, Canada, T2P 3H5, [email protected])
RÉSUMÉ
ABSTRACT
This paper discusses the selection of materials resistant to hydrogen and hydrogen sulfide service in
high temperature applications. Classic selection factors such as fluid composition, design pressure,
etc are discussed as a foundation for further detailed material selection. Several API publications
discussing hydrogen service, hydrogen sulfide corrosion, Cr-Mo alloy steels and high temperature
embrittlement are discussed in their guidance of materials selection and materials specification. Cr-
Mo alloy steels are reviewed for materials selection and materials specification.
INDEX çç Sommaire
INTRODUCTION
INDEX çç Sommaire
PROCESS FLUID COMPOSITION
Process fluid composition(s) include hydrogen partial pressure under all operating
conditions. Unit processes utilizing hydrogen may be required to change feed stocks
or other inputs from time to time necessitating changes in hydrogen partial pressure
being handled in equipment. The related equipment is subjected to a range of
hydrogen concentrations over its lifetime. Process fluids compositions may also be
made more complex for materials selection if impurities such as salts (particularly
chlorides), cyanides or naphthenic acids are present in feeds. Chloride salts are
usually a key consideration in post reaction fluid product cooling and de-
pressurization. Of particular concern is the formation of ammonium chloride
precipitating from gas phase product upon heat exchanger (cooler) surfaces.
Cyanides are also a concern in product cooling as it can lead to acidic conditions
when water washing is used to remove chloride salts. Naphthenic acids are a major
concern if present in feeds in preheat and reaction sections of the unit process. As
the process fluid temperature rises to a threshold value, naphthenic acids may cause
severe corrosion particularly in higher velocity areas. Coping with the corrosive
effects of impurities can be a large factor in materials selection when weld overlay
metallurgy is being considered.
EQUIPMENT TYPES
Unit processes utilizing hydrogen are often complex with several processing steps in
treating hydrocarbons or compounds to achieve desired output fluids. Processing
may include preheating input fluids, several reaction steps, product cooling and
secondary treatment of products. Unit processing usually involves several
equipment types, as listed below.
These equipment types use a variety of material products: castings, forgings, tubing,
pipe, fittings, plates, sheets, weld overlay and powder metallurgy. Most equipment is
also fabricated by welding, so that welding products must also be considered.
Selecting an alloy for one equipment type involves specifying the properties of all the
different products used. Not every alloy is available in each product. For example,
pump casings are usually not available in chromium – molybdenum alloys, and in
lieu, often a stainless steel alloy is selected.
Not every alloy has the same properties in different products. Most manufacturing
processes for product forms can affect properties including corrosion resistance,
response to heat treatment, high temperature strength and so forth. So these are
INDEX çç Sommaire
additional factors to consider in material selection. For example a usual requirement
for high temperature piping is to specify seamless pipe only. Seam welded pipe is
considered to have a weak link: the weld and/or heat affected zones.
Many of the manufacturers of equipment employ heat treatment and other processes
to improve material properties. Manufacturers often employ solution annealing (e.g.
austenitic stainless steel castings following weld repairs) or stress relieving to reduce
hardness (e.g. butt welds in steels for sour service) or to remove fabrication stresses
(e.g. cold working from forming operations). These factors are considered for
material selection.
When selecting materials for different equipment, one has to consider fluid flow as
well. For example, inlet fluid velocity may be high entering a horizontal separator,
then fluid velocity decreases during the separation of phases and at fluid(s) outlets.
In contrast, the inlet velocity at pumps inlet is usually much less than the fluid velocity
at outlet. Fluid flow may have profound effects on corrosion resistance, on removal
of protective films and on causing erosion. Often relatively simple mechanical
arrangements or designs such as impingement plates, tangential nozzles, larger
diameter fittings or high alloy overlay at strategic locations may be necessary to
enhance materials’ performance.
DESIGN TEMPERATURE
Determining the MDMT for a unit is dependent upon the minimum ambient
temperature and the unit’s minimum temperatures during operations. Unit operations
consider normal operations, flare or purge conditions, normal shutdown and any
Joule-Thompson effect during any condition. If the minimum value for MDMT is the
minimum ambient temperature and the temperature is too low, then equipment
warming or heat tracing can be applied. Considering the MDMT temperature during
operations, if the value is too low then modifications to equipment operations may be
necessary.
Determining the MPT for individual equipment is dependent upon the MDMT,
equipment type, its thickness and material toughness. Unit designers or operators
may specify that MPT is MDMT plus a certain allowance that will ensure the
equipment materials’ toughness is at its highest. The allowance may be determined
through considering alloy type, thickness and its toughness variation with
temperature (especially near MDMT).
INDEX çç Sommaire
Determining the maximum operating temperature is from the units’ process design. If
this is not specifically known then a margin is added to the normal operating
temperature.
DESIGN PRESSURE
Design pressure is a material selection factor that may provide a safety margin in the
event of operation at a pressure beyond the maximum operating pressure.
Determining the design margin for design pressure may be from a code or standard
or it may be from consideration of a unit’s overall maximum operating pressure. Too
high a design pressure may lead to a material selection that is overly conservative
and perhaps costly or un-economic.
DESIGN LIFE
As a materials selection factor design life is determined from the annual corrosion
rate multiplied by the desired life time in years. While it seems a simple calculation,
selecting the annual corrosion rate for an alloy in a certain application can be a
challenge. If a corrosive environment is straight forward then published corrosion
data or past corrosion history can provide a fair corrosion rate to use. If the corrosive
environment is complex or new process conditions are creating corrosion conditions
not known before, then substantial research and/or judgment of published data and
of expert opinion may be required to determine an estimate of the anticipated
corrosion rate.
Design life calculation for high alloys such as stainless steels may have little
meaning. Stainless steels are usually highly resistant to general corrosion attack in
hydrocarbons and steam environments. Stainless steels are usually specified with a
zero corrosion allowance for general corrosion. Stainless steels corrode by localized
mechanisms: crevice, or pitting attack. Pitting rates and crevice corrosion rates can
be obtained for stainless steels. From these corrosion rates and a desired design
life, one can determine the required corrosion allowance for a particular stainless
steel alloy. Another approach is to select a stainless steel alloy that is resistant to
INDEX çç Sommaire
pitting and crevice corrosion: no corrosion allowance is required. This latter
approach is often suitable for weld overlay or cladding applications of stainless steel
and other corrosion resistant high alloys.
Equipment failures can lead to decreased unit and plant availability. To assess
equipment’s role in availability, equipment is ranked in criticality: its importance to a
unit’s production or the plant. Materials for highly critical equipment are scrutinized to
assess their capability of long term use without degradation. Measures are taken to
ensure materials for equipment are selected and specified in accordance with their
criticality ratings.
Materials selection for hydrogen and hydrogen in combination with hydrogen sulfide
considers the API standards, 941 (2) and 939-C (3).
Figure 1 from API 941(4) indicates the safe operating limits of carbon steel and alloy
steels as a series of boundary lines on a temperature versus hydrogen partial
pressure graph. Under the boundary line for each metal the safe operating
temperature and pressure combinations are found. For a particular application
usually the maximum design temperature and the maximum design hydrogen partial
pressure are used to determine the minimum alloy recommended from API 941
INDEX çç Sommaire
Figure 1 (4). Some plant or unit owners may designate an additional safety margin
upon the design pressure or design temperature or both.
Reviewing API 941, Figure 1(4) for carbon steel, the highest allowable temperature
for hydrogen service is about 215 C. This temperature is insufficient to cause
damage to carbon steel from creep, temper embrittlement or high temperature
damage. Usually, for hydrogen service carbon steel is stress relieved and the
hardnesses of welds are limited to 200 BHN. As a precaution about hydrogen attack,
it can occur in liquid hydrocarbon streams too as indicated in API 941 (5).
API 941, Figure 1(4), alloy steels may be used up to the temperature range 450 C to
500 C. While standard alloys complying with ASTM or ASME specifications may be
used, some users have found after prolonged use of these alloys conforming to these
standard specifications, the alloys have become embrittled and suffered damage.
Additional requirements to chemical composition and other modifications of material
properties are used to limit temper embrittlement, hydrogen damage and creep. API
941, Figure 1(4), suggests the primary choices for economical materials selection in
high temperature hydrogen service are: carbon steel, 1Cr–Mo, 1Cr–Mo, 2Cr–
1Mo, 2Cr–1Mo–V and 3Cr–1Mo. Note that 3Cr-1Mo-V and 3Cr-1Mo-Cb are also
available.
The austenitic stainless steels are safe for hydrogen service at all pressures and
temperatures in API 941, Figure 1(4). Hydrogen diffuses through these stainless
steels at a rate about 1/100 of the rate that hydrogen diffuses through carbon steel or
alloy steel (6). Note credit should not be taken for applying an austenitic stainless
steel barrier (cladding or weld overlay) in selecting the substrate carbon steel or alloy
steel (6).
API 941 (2) provides the effects of hydrogen only, on carbon steel and alloy steels.
There may be other factors in process streams which will affect an alloy’s
performance (7). These factors are: other corrosives, hydrogen sulfide, creep,
temper embrittlement, high temperature damage mechanisms, interaction between
hydrogen and stress, and, interaction between high temperature hydrogen attack and
creep (7). API standards 939-C (3), 934-A (8) and 934-C (9) are applicable in these
conditions.
API 939-C (3) reviews two corrosion mechanisms: H2 free sulfidation corrosion and
H2/H2S corrosion and. From API 939-C Section 6.3 (10), H2/H2S attacks carbon steel
and most alloy steels above about 232 C (450 F) and the attack is usually so high
that carbon steels and alloy steels are not selected. Austenitic stainless steels can
provide acceptable corrosion rates. The type of hydrocarbon is important: gas oil is
more corrosive than naphtha. Corrosion rate prediction curves are provided as
INDEX çç Sommaire
Couper-Gorman hydrogen/hydrogen sulfide corrosion graphs for carbon steel, low
alloy steels, 12% Cr stainless steel and 18% Cr stainless steel (11). These graphs
indicate the low corrosion rates for stainless steels in both gas oil and naphtha.
In conclusion, austenitic stainless steels (e.g. types 304, 321 or 347) are used to
provide protection from H2/H2S corrosion. Type 321 and weld overlay type 347 SS
are often the preferred choice as they are also resistant to polythionic acid stress
corrosion cracking (PSCC) (12). Polythionic acids can form on piping or equipment
internal surfaces during shutdowns. The H2S in the process fluid causes iron sulfide
to form during operation. When cooling occurs at shutdown condensed fluids
combine with the iron sulfide to form polythionic acids. Un-stabilized stainless steel
type 304 is highly susceptible to PSCC (12).
API 934-A AND API 934-C REDUCING TEMPER EMBRITTLEMENT AND HIGH
TEMPERATURE DAMAGE
API 934-A (8) specifies the requirements for 2Cr and 3Cr alloy steels in high
temperature, high pressure hydrogen service. It includes equipment with an internal
weld overlay for corrosion resistance (e.g. H2S). API 934-A (13) specifies the
following requirements in additional to ASME standards (for components of 2Cr
and 3Cr alloy steels).
1. Steel shall be vacuum degassed.
2. Steel composition shall meet: J factor is 100 maximum. J = (Si+Mn) X (P+Sn) X
104
3. Steel composition shall meet the following maximum levels (wt %): Cu=0.20 and
Ni=0.30 (0.25 for the enhanced steels).
These additional requirements provide relief from temper embrittlement and high
temperature damage.
API 934-A (13) specifies other requirements to the ASME standards as follows.
1. Steel shall be normalized and tempered or quenched and tempered.
2. All specimens for mechanical properties shall come from the T location.
Normalizing and tempering and quench and tempering provide the alloy steel with
higher and more consistent properties through its thickness. The normalizing or
quenched heat treatment produces high strength and high hardness. Tempering
reduces the strength but it is primarily done to reduce hardness for H2 and H2S
service. The difference between normalizing and quenching is cooling rate. A
thicker component requires a greater cooling rate (quenching) to obtain high strength
in the middle of the component (aka T location).
3. Welding consumable composition is to meet: X bar is 15 ppm maximum. X bar =
(10P + 5Sb + 4Sn + As)/100. The reduced impurities reduce weld temper
embrittlement.
INDEX çç Sommaire
4. Step cooling tests are required for weld metal (fusion zone) and heat affected
zone. This test establishes a view of the embrittlement which is possible. The
acceptable level of embrittlement permitted is specified.
API 934-C (9) applies to both 1Cr and 1Cr alloys. API 934-C (14) specifies these
additional requirements to the ASME standards (for components of 1Cr and 1Cr
alloy steels).
1. Steel shall be vacuum degassed.
2. Steel shall be made to fine grain practice.
3. Steel composition shall meet: X bar is 15 ppm maximum.
4. Steel composition shall meet the following maximum levels (wt %): C=0.15,
P=.007, S=0.007, Cu=0.20 and Ni=0.30.
These additional requirements provide relief from temper embrittlement and high
temperature damage.
API 934-C (14) specifies other additional requirements to the ASME standards as
follows.
1. Steel shall be normalized and tempered or quenched and tempered.
2. For thicknesses greater than 50 mm, quench and tempering is recommended.
3. Each specimen for mechanical property testing shall be sampled from the T
location.
4. Tensile testing coupons represent the maximum post weld heat treatment.
5. Impact testing coupons to represent both minimum and maximum post weld heat
treatments that the equipment is to receive during fabrication.
The maximum post weld heat treatment may be defined as the sum of the heat
treatments the steel will be subjected to through its service life including production
and fabrication. After a component is initially heat treated as normalized or
quenched, any further heat treatments will cause it to lose strength, lower hardness
and lose toughness. Further heat treatments may include the following.
1. Tempering after normalizing or quenching.
2. Tempering after being hot formed into usable plate, forging, etc.
3. Post weld heat treatment of welded sub-assemblies.
4. Post weld heat treatment to repair welded fabrication error.
5. Post weld heat treatment of final fabrication.
6. Post weld heat treatment of eventual field repair (during operation).
Heat treatments 1 and 2 may be combined. Step 3 may not be required if the
component is simple, some are not. Step 4 may not be required but experience
counts. Steps 5 and 6 are required. So perhaps a minimum of 4 post weld heat
treatments are required. In some instances, 4 post weld heat treatments and one
tempering will be required.
To measure the effect of post weld heat treatments the Larson-Miller parameter is
calculated. The parameter uses temperature of post weld heat treatment or
tempering and the duration (hours) the equipment is exposed to. The LMP of each
tempering or post weld heat treatment cycle is determined separately and the sum of
INDEX çç Sommaire
all LMP values is determined. The calculated LMP is compared to the available LMP
the alloy steel provides. The calculated LMP has to be less than the available LMP,
otherwise the steel will be over-tempered and it will not have sufficient strength
and/or sufficient toughness.
In API 934-C (14), applications of alloys 1Cr-1/2Mo and 1Cr-1/2Mo are limited to
the following:
1. A maximum temperature of 441°C.
2. Thicknesses up to 100 mm, beyond which strength and/or toughness may not be
sufficient.
From Figure 1 of API 941 (4), the upper boundary lines for 1Cr-1/2Mo and 1Cr-
1/2Mo alloy steels are respectively, 482°C and 510°C. Hence, alloys 1Cr-1/2Mo and
11/4Cr-1/2Mo should not be used in these higher temperature regions of Figure 1,
API 941(4). In API 934-C (14), it is noted that at the heavier thicknesses, nozzle
forgings of 1 Cr – Mo may have insufficient strength or toughness and that 2
Cr – 1 Mo forgings may be necessary. There are limitations of the applicability of
1Cr-1/2Mo and 1Cr-1/2Mo in H2 service or H2/H2S service. Determining the
Larson-Miller parameter will be a key factor in selecting 1Cr-1/2Mo and 1Cr-1/2Mo
successfully.
The use of Cr-Mo alloy steels is widespread in power generation and petrochemical
industries. A principal feature of the 2Cr and higher Cr alloys is that their
metallurgy allows components of high strength and high toughness to be made with
high thicknesses. Their metallurgy has been much studied and their high
temperature strength properties have been enhanced in the alloys: 2Cr-V, 3Cr-V
and 3Cr-Cb. The increase in allowable stresses and temperatures that the enhanced
alloys permit is provided in Table 1 (15). These allowable stresses are for forgings
conforming to ASME SA182 of the grades indicated. Alloy 2 CrMoV has highest
allowable stress through the temperature range, followed by alloys 3 CrMoV and 3
CrMoCb. The increased allowable stresses translate into less equipment weight for
large size and/or high thickness vessels.
From Table 1, the enhanced grades are almost equal in strength. The standard
grades 2 Cr and 3 Cr are also similar in strength except at higher temperature
where 2 Cr is stronger. The enhanced alloys are clearly stronger.
INDEX çç Sommaire
In reviewing API 941, Figure 1 (4), the curves for 2Cr-V and 3Cr-V are identical but
the curve for 2 Cr provides less temperature resistance compared to 3Cr alloy. In
summary the selection of Cr-Mo alloy steels is as follows.
1. 2 Cr and 3 Cr offer similar performance except at higher temperatures. 3Cr is
more resistant to hydrogen, but 2 Cr offers higher strength.
2. 2 Cr-V and 3 Cr-V offer similar performance especially at higher temperatures.
3Cr-Cb is similar to 3 Cr-V.
Selection of the proper alloy is dependent upon equipment factors and “practical”
fabrication factors such as the following.
1. Equipment type, fabricator capability and specialization
2. MDMT and MPT required
3. Proposed dimensions
4. High temperature and very high pressure: forgings in lieu of plate
5. Proposed weights and dimensions of vessel cans and the constraints of
transportation logistics
6. Alloy availability
7. Fabrication facility availability
For example, the fabrication of equipment and piping of these Cr-Mo alloys requires
extensive know-how in heat treatment and welding. Their high strength metallurgy
demands that the fabricator can determine proper pre-heating, intermediate stress
relief, dehydrogenation, post weld heat treatment and tempering heat treatments.
ACKNOWLEGEMENTS
The author wishes to acknowledge the support his employer, SNC-Lavalin Inc, has
provided.
INDEX çç Sommaire
LIST OF REFERENCES
(1) Hilkes, J. and Gross, V, Welding CrMo steels for Power Generation and
Petrochemical Applications, IIW Conference, Singapore, 12/18 July 2009, page 1.
(2) American Petroleum Institute, Recommended Practice 941, Steels for Hydrogen
Service at Elevated Temperatures and Pressures in Petroleum Refineries and
Petrochemical Plants, Seventh Edition, August 2008
(3) American Petroleum Institute, Recommended Practice 939-C Guidelines for
Avoiding Sulfidation (Sulfidic) Corrosion Failures in Oil Refineries, First Edition, May
2009
(4) American Petroleum Institute, Recommended Practice 941, Steels for Hydrogen
Service at Elevated Temperatures and Pressures in Petroleum Refineries and
Petrochemical Plants, Seventh Edition, August 2008, page 3.
(5) American Petroleum Institute, Recommended Practice 941, Steels for Hydrogen
Service at Elevated Temperatures and Pressures in Petroleum Refineries and
Petrochemical Plants, Seventh Edition, August 2008, Section 3.3, page 2.
(6) American Petroleum Institute, Recommended Practice 941, Steels for Hydrogen
Service at Elevated Temperatures and Pressures in Petroleum Refineries and
Petrochemical Plants, Seventh Edition, August 2008, Section 5.5, page 10.
(7) American Petroleum Institute, Recommended Practice 941, Steels for Hydrogen
Service at Elevated Temperatures and Pressures in Petroleum Refineries and
Petrochemical Plants, Seventh Edition, August 2008, Section 3.2, page 2.
(8) American Petroleum Institute, Recommended Practice API 934-A, Materials and
Fabrication of 2Cr-1Mo, 2Cr-1Mo-V, 3Cr-1Mo and 3Cr-1Mo-V Steel Heavy
Wall Pressure Vessels for High-temperature, High-pressure Hydrogen Service,
Second Edition, May 2008
(9) American Petroleum Institute, API 934-C Materials and Fabrication of 1Cr-Mo
Steel Heavy Wall Pressure Vessels For High Pressure Hydrogen /service Operating
at or Below 825 F (441 C), First Edition, May 2008
(10) American Petroleum Institute, Recommended Practice 939-C Guidelines for
Avoiding Sulfidation (Sulfidic) Corrosion Failures in Oil Refineries, First Edition, May
2009, Section 6.3, page 9
(11) American Petroleum Institute, Recommended Practice 939-C Guidelines for
Avoiding Sulfidation (Sulfidic) Corrosion Failures in Oil Refineries, First Edition, May
2009, Annex B, pages 19-30.
(12) American Society for Metals International, Metals Handbook, Volume 13,
Corrosion, pages 1266 and 1267.
(13) American Petroleum Institute, Recommended Practice API 934-A, Materials
and Fabrication of 2Cr-1Mo, 2Cr-1Mo-V, 3Cr-1Mo and 3Cr-1Mo-V Steel
Heavy Wall Pressure Vessels for High-temperature, High-pressure Hydrogen
Service, Second Edition, May 2008, pages 4-9.
(14) American Petroleum Institute, API 934-C Materials and Fabrication of 1Cr-
Mo Steel Heavy Wall Pressure Vessels For High Pressure Hydrogen /service
Operating at or Below 825 F (441 C), First Edition, May 2008, pages 4-7.
(15) American Society of Mechanical Engineers, Boiler and Pressure Vessel Code,
Section II, Materials Part D, 2007 with 2009 Addenda, pages 38-41.
(16) Wells, W., Private Communication, Dillinger Hutte Canada, May 31, 2010.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page98
RÉSUMÉ / ABSTRACT
11B
FONDEMENTS DES RÈGLES DU CODE CASE 2605 DU B&PV CODE DE
L’ASME - EXTENSION DE L’UTILISATION DES ALLIAGES 2.25CR-1MO-V
DANS LE DOMAINE DU FLUAGE
RÉSUMÉ
De nouvelles règles, ainsi qu’un Code case, ont été mises au point dans la section VIII Division 2 pour l’utilisation
de matériaux à haute limite d’élasticité à des températures dans le domaine du fluage ou proche. Dans ces
domaines de contraintes, taux de déformation, temps et températures, les durées de vie sont notablement
Mercredi
raccourcies en raison de l’interaction entre fatigue et fluage qui se répercute sur la microstructure et le
comportement des matériaux. L’interaction fatigue-fluage est particulièrement importante pour les aciers à haute
limite d’élasticité qui subissent des déformations à haute température incluant les phases de démarrage et arrêt.
Dans les nouvelles règles de la section Division 2, les contraintes admissibles pour les aciers faiblement alliés à
haute limite d’élasticité dans le domaine du fluage sont notablement supérieures à celles pour lesquelles il serait
possible d’omettre les vérifications de tenue à la fatigue en se basant sur une expérience de plusieurs années
d’usage sûr. Il en et de même pour tous les nouveaux aciers ou pour applications pour lesquelles le retour
d’expérience est insuffisant. Le Code Case 2065 fournit la marche à suivre pour les aciers à haute limite
d’élasticité en chrome-molybdène quand l’analyse à la fatigue ne peut être omise. Ce Code Case est basé sur un
nouveau modèle qui rend compte du fait que les déformations dans les aciers ferritiques à haute limite d’élasticité
accélèrent les phénomènes d’adoucissement à haute température et diminuent la résistance à la traction, la dureté
et la résistance au fluage. Ce modèle nécessite un calcul de durée de vie au fluage seul dans les zones de
concentration de contrainte et de ne pas positionner de cordon de soudure dans les zones à fort endommagement.
ABSTRACT
New rules and a Code Case under ASME Section VIII, Division 2 rules permit high strength materials to be
used to temperatures approaching and into their respective creep ranges. Very significant reductions in life
occur in those regimes of stress, strain-rate, time and temperature where the damage mechanisms due to creep
and fatigue interact to influence the microstructure and behavior. Creep-fatigue is of special concern especially
when high strength alloys may undergo time-dependent straining including the effects of start-up and shut down
procedures. Under the new ASME Section VIII, Division 2 rules, as the respective creep ranges of high strength
low alloys steels are approached, allowable stresses are significantly higher than those with which there is
service experience that would permit exempting design details from fatigue analysis based on documented “years
of relevant experience”. The same limitation exists for the any new alloys and or applications for which there
is no relevant service experience. ASME Code Case 2605 provides a template for designing with high strength
chrome-moly steels where such exemption is not possible. The Code Case incorporates a novel model that
recognizes that straining of high strength ferritic steels accelerates softening at elevated temperatures lowering
the tensile strength, hardness and creep resistance. It requires calculation of creep life at discontinuities without
cycling and relocation of welds away from locations with high damage rates.
INDEX çç Sommaire
BACKGROUND TO THE DEVELOPMENT OF ASME B&PV CODE CASE
2605 – EXTENSION OF 2.25CR-1MO-V ALLOY INTO THE CREEP
RANGE
M.Prager
The Materials Properties Council, Inc.
Shaker Heights, OH - USA
D.A. Osage
The Equity Engineering Group, Inc.
Shaker Heights, OH - USA
ABSTRACT
New rules and a Code Case under ASME Section VIII, Division 2 rules permit high strength
materials to be used to temperatures approaching and into their respective creep ranges. Very
significant reductions in life occur in those regimes of stress, strain-rate, time and temperature
where the damage mechanisms due to creep and fatigue interact to influence the microstructure
and behavior. Creep-fatigue is of special concern especially when high strength alloys may
undergo time-dependent straining including the effects of start-up and shut down procedures.
Under the new ASME Section VIII, Division 2 rules, as the respective creep ranges of high
strength low alloys steels are approached, allowable stresses are significantly higher than those
with which there is service experience that would permit exempting design details from fatigue
analysis based on documented “years of relevant experience”. The same limitation exists for the
any new alloys and or applications for which there is no relevant service experience. ASME
Code Case 2605 provides a template for designing with high strength chrome-moly steels where
such exemption is not possible. The Code Case incorporates a novel model that recognizes that
straining of high strength ferritic steels accelerates softening at elevated temperatures lowering
the tensile strength, hardness and creep resistance. It requires calculation of creep life at
discontinuities without cycling and relocation of welds away from locations with high damage
rates.
RÉSUMÉ
De nouvelles règles, ainsi qu'un Code case, ont été mises au point dans la section VIII Division 2
pour l'utilisation de matériaux à haute limite d'élasticité à des températures dans le domaine du
fluage ou proche. Dans ces domaines de contraintes, taux de déformation, temps et températures,
les durées de vie sont notablement raccourcies en raison de l'interaction entre fatigue et fluage
qui se répercute sur la microstructure et le comportement des matériaux. L'interaction fatigue-
fluage est particulièrement importante pour les aciers à haute limite d'élasticité qui subissent des
INDEX çç Sommaire
déformations à haute température incluant les phases de démarrage et arrêt. Dans les nouvelles
règles de la section Division 2, les contraintes admissibles pour les aciers faiblement alliés à
haute limite d'élasticité dans le domaine du fluage sont notablement supérieures à celles pour
lesquelles il serait possible d'omettre les vérifications de tenue à la fatigue en se basant sur une
expérience de plusieurs années d'usage sûr. Il en et de même pour tous les nouveaux aciers ou
pour applications pour lesquelles le retour d'expérience est insuffisant. Le Code Case 2065
fournit la marche à suivre pour les aciers à haute limite d'élasticité en chrome-molybdène quand
l'analyse à la fatigue ne peut être omise. Ce Code Case est basé sur un nouveau modèle qui rend
compte du fait que les déformations dans les aciers ferritiques à haute limite d'élasticité
accélèrent les phénomènes d'adoucissement à haute température et diminuent la résistance à la
traction, la dureté et la résistance au fluage. Ce modèle nécessite un calcul de durée de vie au
fluage seul dans les zones de concentration de contrainte et de ne pas positionner de cordon de
soudure dans les zones à fort endommagement.
INTRODUCTION
ASME B&PV Code, Section VIII - Division 2 Rules for Construction of Pressure Vessels –
Alternative Rules (VIII-2) [1] includes rules for the design of 2.25CR-1Mo-V materials for
temperatures up to 472°C (900°F) in ASME Code Case 2605. The analysis method in this code
case includes requirements for a detailed creep analysis using The Materials Properties Council
(MPC) Project Omega Method provided in Part 10 of API 579-1/ASME FFS-1 [2]. The new
method reported here incorporates a creep-fatigue interaction analysis developed by Prager [3]
and places specific limitations on predicted creep damage at welds and corresponding weld joint
location requirements. This code case removes the temperature limitation under previous
editions of ASME VIII-2 where cyclic service was limited to 371°C (700°F) for this material and
other low alloy steels. The intention is to extend the approach to other important high strength
steels.
Overview
The requirements of ASME Code Case 2605 can be summarized as follows. The first step for
application of the code case is to provide a loading histogram. Based on this histogram and the
stated design conditions, the design, i.e. wall thickness and configuration, is based on Design-By-
Rule (DBR) in Part 4 or Design-By-Analysis (BDA) of Part 5 using the allowable stress for
Section II, Part D, Table 5A. An inelastic analysis including creep is then performed to evaluate
ratcheting and to determine the creep life absent fatigue. A supplemental check to determine
creep damage is performed as part of the creep analysis, and a requirement to locate welds away
from areas of creep damage stipulated. Finally, the permissible number of cycles, the creep-life
with fatigue, and design life for the component is determined.
The code case is based on Materials Properties Council (MPC) Project Omega technology
developed by Prager, see references [5] through [12]. It should be noted that details of the
approach as described herein were developed specifically for pressure vessel applications as
covered by this code case. The Omega Method has been used extensively for over 20 years in
the refining and petrochemical industry for remaining life assessment of high temperature
components. It has been incorporated into API 579-1/ASME FFS-1 and has been validated by
international investigators and by successful industrial application.
The extension of this method by MPC to address the creep-fatigue problem was accomplished
using a creep/cyclic strain softening model and available test data for similar creep strength
INDEX çç Sommaire
enhanced ferritic alloys for validation. The model and associated material data included in this
code case were selected to produce a conservative design. Equations for creep life in the code
case are consistent with time-dependent properties used to establish allowable stresses for VIII-2.
Future testing was recommended to refine the model to remove conservatism in the design
procedure. Current work under an MPC group sponsored program is directed at that objective.
The procedures included in the code case explicitly consider the effects of stress relaxation,
stress/strain redistribution, and the associated creep damage.
• Standard pressure parts that comply with the ASME product standards listed in Part 1,
Table 1.1 may be used. In addition, Part 4 paragraph 4.1.11 may be used for closure
components or piping connections at the pressure temperature rating listed in the ASME
standard.
• The skirt attachment detail shall be in accordance with Part 4, paragraph 4.2, Figure 4.2.4,
Detail (e). Alternatively, Part 4, paragraph 4.2, Figure 4.2.4, Detail (f) may be used if the
weldment and adjacent base material of all weld joints is located a minimum distance of
25 mm (1 in.), measured from the weld bevel, from regions where the creep damage, Dc ,
exceeds 0.50. The need for this requirement is discussed below.
• Integral reinforcement must be used for nozzles and conical transitions. This requirement
is typical for heavy wall construction and vessel operating at higher temperatures, see
WRC [4].
• Flange designed in accordance with Part 4, paragraph 4.16 may be used for closure
components or piping connections. The allowable stress used for flange design is
required to be in accordance with Section II, Part D, Table 1.A, the Section VIII, Division
1 allowable stress. The lower Division 1 allowable stress as compared with the Division
INDEX çç Sommaire
2 allowable stress is used to provide a thicker component thereby avoiding the need for a
detailed inelastic analysis of a flanged joint and directionally reducing the likelihood of
flange leakage.
• The rules for external pressure and compressive stress design in Part 4, paragraph 4.4 may
be used if the strain rate computed using Equation (2) based on the membrane stress for
the most severe combination of applied loads that results in compressive stress satisfy
Equation (1). Note that satisfaction of this equation implies minimal creep at the imposed
stress and temperature condition being evaluated. Therefore, the DBR in paragraph 4.4
can be used for external pressure and compressive stress design.
3 (10 )
−8
εc ≤ (1)
hr
INDEX çç Sommaire
through (13). The coefficients for these equations are provided in Table 1. The inelastic analysis
and creep model equations and nomenclature are based on API 579-1/ASME FFS-1, Part 10.
εoc
εc = (2)
1 − Dc
⎧⎪ ⎛ A1 + A2 Sl + A3 Sl2 + A4 Sl3 ⎞ ⎫⎪
log10 ε oc = − ⎨ Ao + ⎜
⎟⎟ ⎬ (3)
⎜ Tref + T
⎩⎪ ⎝ ⎠ ⎭⎪
Sl = log10 (σ e ) (4)
1 ⎡ 2 0.5
( σ 1 − σ 2 ) + (σ 1 − σ 3 ) + (σ 2 − σ 3 ) ⎤
2 2
σe = (5)
2⎣ ⎦
t
Dc = ∫ D c dt ≤ 1.0 (6)
0
D c = Ω mεoc (7)
Ω m = Ωδn +1 (8)
⎛ B + B2 Sl + B3 Sl2 + B4 Sl3 ⎞
log10 Ω = Bo + ⎜ 1 ⎟⎟ (10)
⎜ Tref + T
⎝ ⎠
⎛ A + 2 A3 Sl + 3 A4 Sl2 ⎞
n = −⎜ 2 ⎟⎟ (11)
⎜ Tref + T
⎝ ⎠
⎛ 3p ⎞
δ =β⎜ − 1.0 ⎟ (12)
⎝ σe ⎠
σ1 + σ 2 + σ 3
p= (13)
3
Primary creep is not included in the model described above. Primary creep at the effective
stresses in question here would be extremely small and primary creep does not contribute to life
shortening creep damage in a significant way. Examination of creep test results for the material
in the code case indicates that even above design stress levels, primary strain would be very
small. By the same token, at high stresses, the Omega model, while not treating primary behavior
explicitly does predict accelerated creep rates and assigns damage to the relaxation associated
with this creep. This treatment of relaxation is more conservative than letting primary creep relax
without damage. The Omega coefficients provided in the code case do represent minimum
material properties. It is recognized that high principal stresses occur at discontinuities. For those
multiaxial situations the effective stresses are reduced and significant primary creep does not
occur.
INDEX çç Sommaire
Inelastic analysis is required to be performed, for selected locations experiencing the most
extreme conditions of stress and temperature, to determine the creep life absent fatigue, Lcaf .
Sufficient locations are required to be selected for analysis to ensure that the most critical
conditions have been considered. The creep life absent fatigue is defined as the time in which the
inelastic analysis produces an accumulated creep damage such that 0.95 ≤ Dc < 1.0 or 1,000,000
hours, whichever is reached first. A range is given for Dc to account for numerical accuracy in
the inelastic analysis. For design purposes, and to account for uncertainties in the analysis, a
lower value of Dc may be used, at the discretion of the designer, to establish a conservative
value for Lcaf .
Strains at local discontinuities are accounted for in the inelastic analysis, the creep damage
calculation, and the creep-fatigue calculation. The shakedown requirement essentially ensures
small plastic strains during fatigue, and this requirement is imposed at a point.
Based on the time, temperature, and load history and the results of the inelastic analysis, the
following criteria are required to be satisfied.
• The creep damage at all locations shall be such that Dc < 1.0 . In addition, the weldment
and adjacent base material of all weld joints must be located a minimum distance of 25
mm (1 in.), measured from the weld bevel, from regions where the creep damage, Dc ,
exceeds 0.50.
• The equivalent total accumulated inelastic strain may not exceed the values shown below
in the table. The strain limits in this table for a weld and HAZ were included in the code
case for compatibility with Section III, Subsection NH. The complex behavior of a
weldment would require modeling and analysis far beyond what is practical for a code
case. The solution to this problem is to move the weld away from the region of
significant damage. The code case requires the strain check but also requires the creep
damage check described above because for high strength materials a check on total
accumulated strain accumulation is not sufficient to preclude problems with weldments.
• Protection against local failure shall be determined in accordance with Part 5, paragraph
5.3.3. This analysis need not consider the effects of creep.
INDEX çç Sommaire
Figure 1 and Table 2 shall be used in the screening assessment, and the value of S as shall
be based on (10) 4 cycles. In addition, in STEP 3 and in STEP 4 of paragraph 5.5.2.4, the
alternating equivalent stress amplitude based on the primary plus secondary plus peak
stress determined from ratcheting analysis described above may be used instead of C1S .
The permissible number of cycles, N , in Equation (14) is determined in STEP 3 of
paragraph 5.5.2.4. The creep life with fatigue, Lcwf , is determined using Equation (14)
or Equation (15), as applicable. The equivalent plastic strain amplitude, Δε peq , in
Equations (14) and (15) may be determined from Table 3 as a function of the alternating
equivalent stress amplitude.
⎛ β cf ⋅ Δε peq ⋅ N ⎞
Lcwf = Lcaf ⋅ ⎜ ⎟ Δε peq > 0 (14)
⎜ exp ⎡ β cf ⋅ Δε peq ⋅ N ⎤ − 1 ⎟
⎝ ⎣ ⎦ ⎠
• If Option 2 above is used, then the fatigue analysis is be performed in accordance with
Part 5, paragraph 5.5.4, except that the fatigue curves in Figure 1 and Table 2 shall be
used to determine the accumulated fatigue damage. The accumulated fatigue damage
shall satisfy the requirements of Part 5, paragraph 5.5.4. The creep life with fatigue, Lcwf ,
is determined using Equation (16) or Equation (17), as applicable. The equivalent plastic
strain range for the k th loading condition or cycle in Equations (16) and (17) is
determined directly from the strain-based fatigue analysis results.
⎛ k ⎞
⎜ β cf ∑ Δε peq , k
⋅ ⎟
Lcwf = Lcaf ⋅⎜ i =1 ⎟ Δε peq > 0 (16)
⎜ ⎡ k
⎤ ⎟
⎜ exp ⎢ β cf ⋅ ∑ Δε peq , k ⎥ − 1 ⎟
⎝ ⎣ i =1 ⎦ ⎠
In Option 1, the use of C1S in screening criteria removes conservatism; the actual stress
calculated for the geometry and loading is used in lieu of an approximate stress. Use of S as at
(10) 4 cycles is based on maximum cycle life provided in the fatigue curves of the code case.
The development of the creep-fatigue curves, including hold-time and strain softening effects is
described by Prager in reference [3]. In that work he points out that relatively high strength
alloys such as 2.25Cr-1Mo-V (22V) and modified 9 Cr- 1Mo-V-Cb-N (91) achieve their superior
properties through accelerated cooling of hardenable alloy steel compositions from high
(normalizing) temperatures, transformation of the microstructure to martensite or bainite
followed by tempering. For these materials the specified minimum ambient temperature yield
and tensile strengths are 60 (420) and 85 (590) ksi (MPa), respectively. Typical values of
strengths in finished pressure vessels are likely to be about 70 (490) ksi yield and 92 (640) ksi
(MPa) tensile. For the ranges of room temperature strengths usually expected the time-dependent
stress-rupture and creep properties increase linearly with tensile or yield strength.
However, elevated temperature straining of the alloys under consideration during creep exposure
or cyclic stressing lowers the tensile strength by altering the optimal microstructure from that
INDEX çç Sommaire
obtained by proper heat treatment and thereby reduces creep life. This “strain softening”
behavior is well known and has been reported for decades in studies of 1Cr-1Mo-V turbine rotor
steels and, more recently, in studies of the modified 9Cr-1Mo-V alloy that is widely used in
power piping and similar applications.
Nondestructive Testing
Nondestructive examination of all welds shall be performed in accordance with Examination
Group 1a, see Part 7, Tables 7.1 and 7.2. The supplemental examination for cyclic service in
paragraph 7.4.7 shall be performed.
CREEP-FATIGUE INTERACTION
Pressure vessel service might be simulated by tests to demonstrate how a relatively small number
of cyclic loads would shorten creep life. The severe effect of fatigue damage on creep life known
for high strength alloys might be attributed in large part to a high level of strain softening that
occurs which reduces the tensile strength and thereby degrades the creep properties. Data
reported from creep-fatigue tests seldom if ever include post mortem information on the material
properties or microstructural changes due to cycling. However, acceleration of softening
associated with creep straining is well known from interrupted stress-rupture tests of the high
strength ferritic alloys. What is needed is modeling and quantification of the effects.
To start to develop a model for creep fatigue interaction an expression for creep strain rate that
could be modified to describe the increase of strain and strain rate with time (creep damage) and
cycling was needed. It is immaterial what model is chosen for the strain computation as long as
it includes an explicit term for strain rate that can be accelerated by cycling or other damage
related to fatigue. It is also important that the function be amenable to integration to obtain stress
rupture life.
Starting with the creep strain rate, εc , defined as follows:
dε c
εc = (18)
dt
INDEX çç Sommaire
For usual applications in which primary creep due to steady loading at allowable stresses is
small, we can chose to express tertiary creep behavior as follows:
Note that in the model here the notion of “secondary” creep is not applicable. The secondary
behavior observed in laboratory testing may be viewed as the sum of decreasing primary and
increasing tertiary components which together add up to an approximately constant “minimum”.
For the strain-softening, fatigue-damaging materials of interest we may describe the acceleration
of creep rate due to cycling with the following relation
ˆ
βcf ε p Nt
εc = εcoeΩεc ⋅ e (20)
It should be apparent that the product N̂t indicates the number of cycles accumulated in the time
t which can be a variable of integration to obtain total life.
The reduction of stress rupture life is obtained by separating variables and integrating the above
relation to give.
(1 − e ) = ( e
−Ωε f ˆ
β cf ε p Nt
)
−1
(21)
εco Ω β cf ε p Nˆ
(1 − e ) = t −Ωε f
(22)
εco Ω r
t =
(e ˆ
β cf ε p Nt f
−1 ) (23)
β cf ε p Nˆ
r
Examination of limited data revealed that β cf = 2 was a reasonable value to correlate failure
times from the longer hold time tests.
Equation (25) is used to establish the design curves in Figures 1 and 1M and in Tables 2 and 2M
for several calculated rupture lives. It should be noted that the stress amplitudes are
pseudoeleastic stresses calculated for minimum material properties based on the values in Table
3.
INDEX çç Sommaire
CONCLUSION
A code case for creep –fatigue design was developed that combines state of the art creep life
assessment by finite element analysis with appropriate conservative material properties and a
physical model that recognizes the metallurgical changes fostered by plastic straining of a
modern high strength steel. Lower properties of welds are dealt with by removing welds from
critical areas identified in the analysis.
While only limited data were available, conservatism was introduced through the design
assumptions and use of minimum material properties. Work continues to quantify the relation
between plastic strains encountered on start up and shut down and the expected diminution of
creep strength for 2.25Cr-1Mo -0.25V and similar steels.
Nomenclature
A0 → A4 material coefficients for the MPC Project Omega strain-rate-parameter, see Table 1.
B0 → B4 material coefficients for the MPC Project Omega Omega-parameter, see Table 1.
β MPC Project Omega multiaxial creep parameter equal to 1/3.
β cf creep strain reduction factor per unit strain per cycle, equal to 2.0.
Dc creep damage.
D c creep damage rate.
Δε peq equivalent plastic strain amplitude based on the alternating equivalent stress
amplitude determined in the fatigue screening analysis.
Δε peq ,k equivalent plastic strain amplitude for the k th loading condition or cycle.
δ MPC Project Omega parameter.
εc accumulated creep strain.
εp plastic strain per cycle.
εf fracture strain.
εc creep strain rate.
εco initial creep strain rate at the start of the time period being evaluated based on the
stress state and temperature.
Lcaf creep life absent fatigue.
Lcwf creep life with fatigue.
n MPC Project Omega parameter.
N permissible number of cycles.
N̂ cycle frequency or 1 divided by the hold time.
p hydrostatic stress.
σe effective stress.
σ1 principal stress in the 1-direction.
σ2 principal stress in the 2-direction.
σ3 principal stress in the 3-direction.
t time in hours.
tf creep life with fatigue.
T temperature.
INDEX çç Sommaire
Tref 460 using the coefficients in Table 1 for units of ksi and o F , 273 using the
coefficients in Table 1M for units of MPa and oC .
Ω uniaxial Omega damage parameter, a coefficient characterizing the rate strain rate
increases with creep strain absent fatigue
Ωm multiaxial Omega damage parameter.
Ωn adjusted uniaxial Omega damage parameter.
BIBLIOGRAPHY
[1] AMSE, ASME B&PV Code, Section VIII - Division 2 Rules for Construction of Pressure
Vessels – Alternative Rules, ASME, New York, N.Y.
[2] API, API 579-1/ASME FFS-1 2007 Fitness-For-Service, American Petroleum Institute,
Washington, D.C.
[3] Prager, M., “Extend Low Chrome Steel Fatigue Rules,” ASME STP-PT-027, ASME, New
York, N.Y.
[4] Carucci, V.A., Chao, R.C., and Stelling, D.J., Recommendations for Design of Vessels for
Elevated Temperature Service, WRC Bulletin 470, The Welding research council, New
York, N.Y.
[5] Prager, M., “Development of the MPC Project Omega Method for Life Assessment in the
Creep Range,” PVP-Vol. 288, ASME, 1994, pp. 401-421.
[6] Prager, M., “Damage Evaluation and Remaining Life Assessment in High Temperature
Structural Components by the Omega Method,” Proceedings of 7th Workshop on the Ultra
Steel, Ultra Steel: Requirements from New Design of Constructions, June 24 and 25, 2003,
pages 150-158.
[7] Prager, M., “The Omega Method – An Effective Method for Life and Damage Prediction in
Creep Tests and Service,” Oikawa (eds.), Strength of Materials, Japan Institute of Metals,
1994, pp. 571-574.
[8] Prager, M., “Proposed Implementation of Criteria for Assignment of Allowable Stresses in
the Creep Range,” ASME Journal of Pressure Vessel Technology, May, 1996, Vol. 335, pp.
273-293.
[9] Prager, M., “Generation of Isochronous Creep, Tubing Life and Crack Growth Curves
Using the MPC Omega Method, Structural Integrity,” NDE, Risk and Material
Performance for Petroleum, process and Power, PVP-Vol. 336, ASME, 1996, pp. 303-322.
[10] Prager, M. and Ibarra, S., “Approaches to Long Term Life prediction of Furnace and Boiler
Tubes,” Fitness For Adverse Environments in Petroleum and Power Equipment, PVP-Vol.
359, ASME, 1997, pp. 339-352.
[11] Prager, M. and Masuyama, F., “Examination of the Effects of Materials Variables in
Advanced Alloys by Studying Creep Deformation Behavior,” Strength of Materials,
Oikawa et al. ed., ICSMA 10 Sendai, The Japan Institute for Metals, 1994, pp 575-578.
[12] Prager, M., and Osage, D., “Special Topics in Elevated Temperature Life Applications
Including Assessment Rules for API 579,”, PVP-Vol. 411, Service Experience and Fitness-
For-Service in Power and Petroleum Processing, ASME, 2000, pages 91-104.
INDEX çç Sommaire
Table 1 – MPC Project Omega Creep Data
INDEX çç Sommaire
Table 2 – Design Fatigue Curve for 2.25Cr-1Mo-V Steel for Temperatures Not Exceeding
850ºF, Design Cycles as a Function of Stress Amplitude and Creep Life Absent Fatigue
Stress Creep Life Absent Fatigue (hrs)
Amplitude
300,000 400,000 500,000 600,000 700,000 800,000 900,000 1,000,000
(ksi)
577.3 15 18 21 24 26 29 31 32
440.6 17 21 25 28 31 34 37 39
339.2 20 25 29 33 37 40 44 47
263.9 22 28 34 39 43 48 52 56
207.9 26 33 39 45 51 56 61 66
166.2 29 37 45 52 59 65 71 77
135.2 33 42 51 60 68 75 83 89
112.0 37 48 58 68 78 87 95 104
94.6 41 54 66 77 88 99 109 119
81.5 46 61 74 88 100 113 125 136
71.6 52 68 84 99 114 128 142 155
64.1 58 76 94 111 128 145 161 176
58.3 64 85 105 125 144 163 181 199
53.9 72 95 117 139 161 183 204 224
50.3 80 105 131 156 180 205 228 252
47.6 88 117 145 174 201 229 256 282
45.3 98 130 162 193 224 255 285 315
43.4 109 144 180 215 249 284 318 352
41.9 121 160 199 238 277 316 354 392
40.6 133 177 221 265 308 351 394 436
39.4 148 197 245 294 342 390 437 485
38.4 164 218 272 325 379 432 485 538
37.4 181 241 301 360 420 479 538 597
36.6 200 267 333 399 465 531 596 662
35.8 222 295 369 442 515 588 661 734
35.0 245 327 408 489 570 651 732 813
34.3 271 361 451 541 631 721 810 900
33.6 300 400 499 599 698 798 897 996
33.0 332 442 553 663 773 883 993 1102
32.3 367 489 611 733 855 977 1098 1220
31.7 406 541 676 811 946 1080 1215 1350
31.1 449 598 747 897 1046 1195 1344 1493
30.5 496 661 827 992 1157 1322 1487 1651
30.0 549 731 914 1097 1279 1462 1644 1826
29.4 607 809 1011 1213 1415 1616 1818 2020
28.9 671 894 1118 1341 1564 1787 2011 2234
28.3 742 989 1236 1483 1730 1976 2223 2470
27.8 820 1093 1366 1639 1912 2185 2458 2731
27.3 906 1209 1511 1813 2115 2416 2718 3020
26.8 1002 1336 1670 2004 2338 2672 3006 3339
26.3 1108 1477 1847 2216 2585 2954 3323 3692
INDEX çç Sommaire
Table 2 – Design Fatigue Curve for 2.25Cr-1Mo-V Steel for Temperatures Not Exceeding
850ºF, Design Cycles as a Function of Stress Amplitude and Creep Life Absent Fatigue
Stress Creep Life Absent Fatigue (hrs)
Amplitude
300,000 400,000 500,000 600,000 700,000 800,000 900,000 1,000,000
(ksi)
25.8 1225 1633 2042 2450 2858 3266 3674 4082
25.3 1354 1806 2257 2709 3160 3611 4063 4514
24.8 1497 1997 2496 2995 3494 3993 4492 4991
24.4 1656 2207 2759 3311 3863 4414 4966 5518
23.9 1830 2440 3050 3661 4271 4881 5491 6100
23.5 2024 2698 3373 4047 4722 5396 6070 6745
23.1 2237 2983 3729 4474 5220 5966 6711 7457
22.6 2473 3298 4122 4947 5771 6596 7420 8244
22.2 2735 3646 4558 5469 6381 7292 8204 9115
21.8 3023 4031 5039 6047 7054 8062 9070 10078
21.4 3343 4457 5571 6685 7799 8913 10028 11142
21.0 3696 4927 6159 7391 8623 9855 11086 12318
20.6 4086 5448 6809 8171 9533 10895 12257 13619
20.2 4517 6023 7528 9034 10540 12045 13551 15057
19.9 4994 6659 8323 9988 11652 13317 14982 16646
19.5 5521 7362 9202 11042 12883 14723 16564 18404
19.1 6104 8139 10174 12208 14243 16278 18312 20347
18.8 6749 8998 11248 13497 15747 17996 20246 22495
18.4 7461 9948 12435 14922 17409 19896 22383 24870
18.1 8249 10998 13748 16498 19247 21997 24747 27496
17.8 9120 12160 15200 18240 21279 24319 27359 30399
17.4 10083 13443 16804 20165 23526 26887 30248 33609
17.1 11147 14863 18579 22294 26010 29726 33441 37157
16.8 12324 16432 20540 24648 28756 32864 36972 41080
16.5 13625 18167 22709 27250 31792 36334 40875 45417
16.2 15064 20085 25106 30127 35149 40170 45191 50212
15.9 16654 22205 27757 33308 38860 44411 49962 55514
15.6 18412 24550 30687 36825 42962 49100 55237 61375
15.3 20356 27142 33927 40713 47498 54284 61069 67854
Notes:
1. Interpolation of design cycles as a function of design life is permitted.
2. Extrapolation of design cycles as a function of stress amplitude is not permitted.
3. The reference temperature and Young’s Modulus for these fatigue curves are 850ºF and 26,106 ksi,
respectively.
INDEX çç Sommaire
Table 2M – Design Fatigue Curve for 2.25Cr-1Mo-V Steel for Temperatures Not Exceeding
454ºC, Design Cycles as a Function of Stress Amplitude and Creep Life Absent Fatigue
Stress Creep Life Absent Fatigue (hrs)
Amplitude
300,000 400,000 500,000 600,000 700,000 800,000 900,000 1,000,000
(MPa)
3982 15 18 21 24 26 29 31 32
3039 17 21 25 28 31 34 37 39
2339 20 25 29 33 37 40 44 47
1820 22 28 34 39 43 48 52 56
1434 26 33 39 45 51 56 61 66
1147 29 37 45 52 59 65 71 77
932 33 42 51 60 68 75 83 89
772 37 48 58 68 78 87 95 104
652 41 54 66 77 88 99 109 119
562 46 61 74 88 100 113 125 136
494 52 68 84 99 114 128 142 155
442 58 76 94 111 128 145 161 176
402 64 85 105 125 144 163 181 199
371 72 95 117 139 161 183 204 224
347 80 105 131 156 180 205 228 252
328 88 117 145 174 201 229 256 282
312 98 130 162 193 224 255 285 315
300 109 144 180 215 249 284 318 352
289 121 160 199 238 277 316 354 392
280 133 177 221 265 308 351 394 436
272 148 197 245 294 342 390 437 485
265 164 218 272 325 379 432 485 538
258 181 241 301 360 420 479 538 597
252 200 267 333 399 465 531 596 662
247 222 295 369 442 515 588 661 734
242 245 327 408 489 570 651 732 813
237 271 361 451 541 631 721 810 900
232 300 400 499 599 698 798 897 996
227 332 442 553 663 773 883 993 1102
223 367 489 611 733 855 977 1098 1220
219 406 541 676 811 946 1080 1215 1350
215 449 598 747 897 1046 1195 1344 1493
211 496 661 827 992 1157 1322 1487 1651
207 549 731 914 1097 1279 1462 1644 1826
203 607 809 1011 1213 1415 1616 1818 2020
199 671 894 1118 1341 1564 1787 2011 2234
195 742 989 1236 1483 1730 1976 2223 2470
192 820 1093 1366 1639 1912 2185 2458 2731
188 906 1209 1511 1813 2115 2416 2718 3020
185 1002 1336 1670 2004 2338 2672 3006 3339
181 1108 1477 1847 2216 2585 2954 3323 3692
INDEX çç Sommaire
Table 2M – Design Fatigue Curve for 2.25Cr-1Mo-V Steel for Temperatures Not Exceeding
454ºC, Design Cycles as a Function of Stress Amplitude and Creep Life Absent Fatigue
Stress Creep Life Absent Fatigue (hrs)
Amplitude
300,000 400,000 500,000 600,000 700,000 800,000 900,000 1,000,000
(MPa)
178 1225 1633 2042 2450 2858 3266 3674 4082
175 1354 1806 2257 2709 3160 3611 4063 4514
171 1497 1997 2496 2995 3494 3993 4492 4991
168 1656 2207 2759 3311 3863 4414 4966 5518
165 1830 2440 3050 3661 4271 4881 5491 6100
162 2024 2698 3373 4047 4722 5396 6070 6745
159 2237 2983 3729 4474 5220 5966 6711 7457
156 2473 3298 4122 4947 5771 6596 7420 8244
153 2735 3646 4558 5469 6381 7292 8204 9115
150 3023 4031 5039 6047 7054 8062 9070 10078
148 3343 4457 5571 6685 7799 8913 10028 11142
145 3696 4927 6159 7391 8623 9855 11086 12318
142 4086 5448 6809 8171 9533 10895 12257 13619
140 4517 6023 7528 9034 10540 12045 13551 15057
137 4994 6659 8323 9988 11652 13317 14982 16646
135 5521 7362 9202 11042 12883 14723 16564 18404
132 6104 8139 10174 12208 14243 16278 18312 20347
130 6749 8998 11248 13497 15747 17996 20246 22495
127 7461 9948 12435 14922 17409 19896 22383 24870
125 8249 10998 13748 16498 19247 21997 24747 27496
123 9120 12160 15200 18240 21279 24319 27359 30399
120 10083 13443 16804 20165 23526 26887 30248 33609
118 11147 14863 18579 22294 26010 29726 33441 37157
116 12324 16432 20540 24648 28756 32864 36972 41080
114 13625 18167 22709 27250 31792 36334 40875 45417
112 15064 20085 25106 30127 35149 40170 45191 50212
110 16654 22205 27757 33308 38860 44411 49962 55514
108 18412 24550 30687 36825 42962 49100 55237 61375
106 20356 27142 33927 40713 47498 54284 61069 67854
Notes:
1. Interpolation of design cycles as a function of design life is permitted.
2. Extrapolation of design cycles as a function of stress amplitude is not permitted.
3. The reference temperature and Young’s Modulus for these fatigue curves are 454ºC and 180,000 MPa,
respectively.
INDEX çç Sommaire
Table 3 – Stress Amplitude versus Plastic Strain
Alternating Equivalent Stress Amplitude
Equivalent Plastic Strain Amplitude
(ksi) (MPa)
577.5 3982 0.02000000
440.7 3039 0.01480000
339.2 2339 0.01095200
263.9 1820 0.00810400
207.9 1434 0.00599700
166.3 1147 0.00443800
135.2 932 0.00328400
112.0 772 0.00243000
94.6 652 0.00179800
81.5 562 0.00133100
71.7 494 0.00098500
64.1 442 0.00072900
58.3 402 0.00053900
53.9 371 0.00039900
50.4 347 0.00029500
47.6 328 0.00021900
45.3 312 0.00016200
43.5 300 0.00012000
41.9 289 0.00008860
40.6 280 0.00006550
39.4 272 0.00004850
38.4 265 0.00003590
37.4 258 0.00002660
36.6 252 0.00001960
35.8 247 0.00001450
35.0 242 0.00001080
34.3 237 0.00000796
33.6 232 0.00000589
33.0 227 0.00000436
32.3 223 0.00000323
31.7 219 0.00000239
31.1 215 0.00000177
30.5 211 0.00000131
30.0 207 0.00000097
Notes:
1. The plastic strain below 207 MPa (30 ksi) shall be taken as 0.0.
2. Interpolation of data values is permitted.
3. Extrapolation of data values is not permitted.
INDEX çç Sommaire
Notes:
1. Interpolation of design cycles as a function of design life is permitted.
2. Extrapolation of design cycles as a function of stress amplitude is not permitted.
3. The reference temperature and Young’s Modulus for these fatigue curves are 850ºF and 26,106
ksi, respectively.
Figure 1 – Design Fatigue Curve for 2.25Cr-1Mo-V Steel for Temperatures Not Exceeding
850ºF, Design Cycles as a Function of Stress Amplitude and Creep Life Absent Fatigue
INDEX çç Sommaire
Notes:
1. Interpolation of design cycles as a function of design life is permitted.
2. Extrapolation of design cycles as a function of stress amplitude is not permitted.
3. The reference temperature and Young’s Modulus for these fatigue curves are 454ºC and 180,000
MPa, respectively.
Figure 1M – Design Fatigue Curve for 2.25Cr-1Mo-V Steel for Temperatures Not
Exceeding 454ºC, Design Cycles as a Function of Stress Amplitude and Creep Life Absent
Fatigue
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page99
RÉSUMÉ / ABSTRACT
37A
UNE MÉTHODE POUR MODÉLISER LA DÉCHIRURE DUCTILE DE
L’APPARITION DE LA PLASTICITÉ JUSQU’AUX GRANDES
PROPAGATIONS
RÉSUMÉ
Les modèles continus pour la rupture ductile permettent la prédiction de la déchirure ductile grâce à
leur formulation qui prend en compte l’effet de la triaxialité des contraintes. Cependant les problèmes
numériques comme la localisation et la non convergence empêchent de mener des calculs pour de
grandes propagations.
Ce papier présente une méthode pour introduire une fissure dans une modélisation continue (modèle
de Rousselier) au moyen d’un modèle cohésif cohérent. L’introduction des zones cohésives est faite
dans le cadre de la Méthode des Eléments Finis Etendue (XFEM), ainsi le modèle continu reste actif
tant qu’il est le plus apte pour prédire le comportement, puis il est relayé par le modèle cohésif.
La solution proposée résout les difficultés numériques associées à l’approche locale de la mécanique
de la rupture. Les performances de la méthode sont illustrées sur un essai de traction sur éprouvette CT
en comparant les prédictions obtenues à celles données par la modélisation utilisant le modèle local de
Rousselier uniquement.
ABSTRACT
Continuum models for ductile fracture allow the accurate modelling of ductile tearing on 3D real
structures thanks to their stress triaxiality dependent formulation. Nevertheless their use induces several
numerical problems as localization and non convergence that hinder large crack propagation
prediction.
This paper presents a method to switch from the continuum model, the Rousselier model in this case,
to a real crack through a consistent cohesive zone. The cohesive discontinuity is introduced in the
framework of the eXtended Finite Element Method, so that cohesive zone segments can be introduced
during the problem calculation. This permits to use the continuum model as long as it is the most
appropriate to predict the material behaviour with regards to the stress triaxiality and then to introduce
cohesive zone segments.
The proposed solution solves numerical difficulties associated with the continuous vision of fracture
without accuracy loss. The performance of the method is illustrated through modelling of ductile tearing
on a CT specimen with a comparison of prediction with the conventional modelization.
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
INTRODUCTION
Large scale components can often sustain large ductile tearing propagation before the maximum
endurable loading conditions are reached. In these components, ductile tearing modelling using a local
approach model as the Rousselier model [1] is in general difficult due to numerical problems that often
appear. These difficulties are usually induced by large distortions of failed elements in the vicinity of
the current crack front, where the use of element elimination techniques is deficient because the
removal of failed elements does not ensure the energy conservation around the crack front and induces
an error which grows continuously as the crack progresses in the calculation.
One solution to the distortion problem is to switch during the calculation from a diffuse damage model
to a real crack. The development of the eXtended Finite Element Method (XFEM) [2] provides an
innovative and a practical solution to our problem. Enriched elements allow avoiding an explicit
representation of the crack in the mesh. The crack is taken into account by additional degrees of
freedom (d.o.f.) in the element formulation corresponding to gap functions.
The goal of the present work is to introduce a real crack before complete failure of the elements during
a ductile tearing analysis with the Rousselier model. The discontinuity is introduced through
enrichment of the displacement approximation with gap d.o.f in the concerned elements. As we face
here ductile fracture, the cohesive zone approach is chosen to bridge the local approach and the real
crack in the elements where the critical void volume fraction is near to be reached. The cohesive zone
then opens progressively up to a real crack (when the resistance of the cohesive zone falls to zero). For
that matter, a method was previously developped in order to identify a cohesive law that induces the
same dissipation as the Rousselier model in this remaining failure step [3].
This paper describes how the consistent cohesive zone model can be introduced in the modelization in
order to make large propagation analyses. First the Rousselier model and its consistent cohesive law
are presented, then the implementation of the cohesive zone model in the XFEM framework of
Cast3M is described and finally we will analyse some results.
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
It is not possible to have an arbitrary location of the crack tip with this approximation. But it is in
accordance with the following theoretical assumption: The existence of a process zone of non
negligible size leads to the disappearance of the crack tip singularity. Another issue of using this
discontinuous approximation is that the integration scheme has to be of higher order in order to situate
the discontinuity accurately. Therefore, the propagation zone is meshed with 16 Gauss points linear
quadrangles, denoted XFEM elements and depicted on figure 3.
A cohesive zone model permits to model the damage and/or plastic deformation localization process
on a double surface that can be seen as a crack whose lips opening depends on a specific constitutive
law that gives the softening behavior before appearance of a real crack. The modelizaton of cohesive
zones in the XFEM framework corresponds to this description. Gauss points are added on the
discontinuity surface, as shown in figure 4, in order to compute the traction corresponding to the
cohesive zone opening ahead of the real crack tip (real crack is meant in opposition to the stiffness loss
induced by damage, now the real crack lips are free of loading).
It is to remark that these Gauss points are related with the enrichment d.o.f. so that their addition does
not induce the addition of new d.o.f. in comparison to a classical XFEM macrocrack.
In this work, the cohesive tractions are hence applied as an external loading acting onto the crack lips.
Basically, this can be seen as the addition of nonlinear springs between the crack lips whose stiffness is
associated with the previously identified cohesive law. Moreover, this strategy permits to introduce the
cohesive zone during the calculation without any modification of the global stiffness matrix in
comparison with XFEM crack with free lips. The equilibrium of the cohesive zone as well as the bulk
structure is then verified using a standard iterative process.
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
APPLICATION
Material definition
An application was done with the VOCALIST [8] pipe ferritic steel, for which it was not possible to
model large crack propagation with the classical formulation of the Rousselier model. The tensile
curve of the studied material is given in figure 6 and its other characteristics in table 1. The parameters
of the Rousselier model for this material are given in table 2, where λc corresponds to the mesh size to
be used with 4 Gauss point quadratic quadrangles. In order to obtain a similar stiffness for the XFEM
linear element mesh as for the classical quadratic one, the XFEM element size λXFEM is 75µm. In
addition to that, using 16 Gauss points linear elements tends to accelerate the growth of the porosity
through higher stress gradient in the element. In order to compensate this process, the critical porosity
for this mesh was taken to be fcXFEM=5%.
was done on a characterization test for ductile tearing. Indeed, tensile tests on CT specimens are used
to determine characteristics for the global approach but also to determine the propagation characteristic
length of the Rousselier model.
The specimen was a CT20 with lateral grooves (each groove 10% of the initial thickness) in order to
favour a plane strain state. The calculations were made under 2D-plane strain hypothesis with a large
strain and large displacement formulation. The resulting mesh is depicted in figure 8. The internal
radius of the test zone is 70µm and the external radius is taken to be 3 times the element size, that is
l=240µm.
On figure 9 and 10 can be seen that the coupled modelization gives mechanical results that recover the
ones of the standard modelization using only of the local Rousselier model. This shows that the
coupled method preserves a good conversion of the mechanical energy. The use of a ring of internal
radius being of the order of the material characteristic length permits to model crack initiation
accurately. Moreover, it permits to reduce the influence of the mesh size on the numerical results.
The coupled method allows for modelling a much longer crack growth than the classical formulation.
Besides, the calculation could be led further if the propagation zone were meshed larger. Indeed,
introducing a real discontinuity whose propagation is driven by void growth and coalescence and by
the cohesive zone model suppresses the numerical difficulties associated to distorted elements: the
Jacobian remains physical on both sides of the crack.
The difference between test and numerical results for the load-CMOD (Crack Mouth Opening
Displacement) curve is due to the plane strain hypothesis, which does not permit to take into account
the initially curved crack front that results from the fatigue pre-cracking. In order to model this
correctly, it would be necessary to make 3D calculations.
The possibility of modelling that large amount of crack growth shows that the Rousselier parameters
that were used are not able to predict large crack propagation accurately. This is due to the inability to
identify it on a sufficient amount of propagation. This is an important practical result for the
identification of the material properties of this model. In order to obtain more accuracy, the Rousselier
model parameters could be newly identified using the coupled method.
CONCLUSION
The phenomenological basis and the independence of local approach models with regards to the scale
and the geometry of specimens make them very appealing. However, they are difficult to apply to the
modelling of large crack growth due to numerical instabilities.
This paper proposes a coupled approach that permits to combine a local approach model for ductile
fracture with a cohesive zone model. This combination is possible thanks to the XFEM framework
through which a discontinuity can be meshed independently of the structural mesh. Cohesive laws are
added onto the discontinuous surface to model the progressive material failure. The cohesive tractions
are applied as dual components of the additional d.o.f. and its behaviour is calculated at the cohesive
zone Gauss points.
Applications on a CT specimen for a ferritic steel show the ability of the coupled method to model
large crack propagation. In addition to that, this method indicates that the Rousselier model parameters
were identified on a too small amount of crack growth in order to be accurate for large crack
propagation. The use of the coupled approach could permit to reidentify the Rousselier model
parameters in order to predict accurately large crack growth.
These results permit to consider large crack growth prediction and more challenging 3D crack
propagation prediction on large scale components as the VOCALIST pipe invoked in the introduction.
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
NOTATIONS
σ1 Rousselier model parameter σm Average stress
σ vM Equivalent von Mises stress σ Stress tensor
ρ Initial density ρ0 Current density
D Rousselier model parameter Dmp Average plastic strain rate
f Void volume fraction ; porosity p Equivalent plastic strain
F Dissipation potential H Heaviside function
R(p) Strain hardening curve of the material
bj Enrichment degree of freedom vector Ni Finite element standard shape
at node j function at node i
Displacement jump between the
t zc Cohesive stress [u ]
discontinuity lips
ui Standard degree of freedom vector at u
h
Approximated displacement field
node i
REFERENCES
[1] Rousselier, G. Ductile fracture models and their potential in local approach of fracture. Nuclear
Engineering and Design 105, 97-111. 1987.
[2] Moës, N.; Dolbow, J. & Belytschko, T. A finite element method for crack growth without
remeshing, International Journal For Numerical Methods In Engineering 46(1), 131-150. 1999.
[3] Simatos, A., Marie, S., A. Combescure & F. Cazes. Modelling ductile tearing from diffuse
plasticity to crack propagation. Proceeding of the PVP. 2010.
[4] Rousselier, G. Dissipation in porous metal plasticity and ductile fracture. Journal of the
Mechanics and Physics of Solids 49(8), 1727-1746. 2001.
[5] Comi, C. & Mariani, S. Extended finite element simulation of quasi-brittle fracture in
functionally graded materials, Computer Methods in Applied Mechanics and Engineering
196(41-44), 4013 - 4026. 2007.
[6] Wells, G. N. & Sluys, L. J. A new method for modelling cohesive cracks using finite elements,
International Journal For Numerical Methods In Engineering 50(12), 2667--2682. 2001.
[7] Prabel, B.; Marie, S. & Combescure, A. Using the X-FEM method to model the dynamic
propagation and arrest of cleavage cracks in ferritic steel, Engineering Fracture Mechanics
75(10), 2984--3009. 2008.
[8] Marie, S. Analytical expression of the thermal stresses in a vessel or pipe with cladding
submitted to any thermal transient, International Journal of Pressure Vessels and Piping, Vol. 81,
pp.303-312. 2004.
TABLES
FIGURES
Figure 4: Cohesive zone discretization Figure 5: Offset method for the cohesive law
adaptation
INDEX çç Sommaire
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901
800
400
200
0
0 10 20 30
Displacement jump (µm)
Figure 6: Tensile Curve of the TU52B ferritic steel Figure 7: Cohesive law
5
CMOD (mm)
2
Test
1
Standard Calculation
Cohesive Zone Calculation
0
0 1 2 3 4 5 6 7 8
Crack Growth (mm)
25
20
Force (kN)
15
10
5 Test
Standard Calculation
0
Cohesive Zone Calculation
0 1 2 3 4 5 6 7
CMOD (mm)
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page100
RÉSUMÉ / ABSTRACT
35A
DÉVELOPPEMENT D’UN MODULE DE TEST DE COUVERTURE
TRITIGÈNE POUR ITER
RÉSUMÉ
Dans un réacteur à fusion thermonucléaire, les couvertures tritigènes sont des composants de première
paroi faisant face au plasma; ils ont trois fonctions : la récupération de chaleur, la production de Tritium,
et la protection radiologique. Dans le cadre des études pour le développement d’un concept de
Mercredi
couverture au Plomb-Lithium refroidi à l’Hélium (« Helium Cooled Lithium Lead » – HCLL), le CEA
Saclay, sous contrat avec l’entreprise commune européenne Fusion for Energy, a en charge la conception
d’un module de test de couverture (« Test Blanket Module » - TBM) pour ITER basé sur le concept
HCLL. Ce module présente : un caloporteur Hélium à 8 MPa, circulant au travers de canaux creusés
dans l’épaisseur des plaques réfrigérées constituant sa structure principale ; un eutectique Plomb-
Lithium tritigène circulant dans le volume clos par la structure ; une température de structure
s’échelonnant de 300 à 550°C ; un matériau constitutif spécifique de type acier ferritique-martensitique
à basse activation (Eurofer). Ses caractéristiques fonctionnelles et sa conception en font un équipement
sous pression nucléaire « multi-compartiments ». Cet article dresse un aperçu de la conception de ce
composant complexe et des analyses associées, pour lesquelles les règles appliquées sont basées sur le
code RCC-MR 2007, complétées par des critères spécifiquement développés pour le réacteur ITER
(« ITER Structural Design Criteria ») afin de prendre en compte les effets induits par l’irradiation de
neutrons de 14 MeV ; les besoins en développements en terme de règles de conception et de fabrication
sont aussi abordés.
ABSTRACT
In a thermonuclear fusion power plant, the breeding blankets are components facing the plasma playing
three roles: heat recovery, tritium breeding, and radiological protection. In the frame of the studies for
the development of the Helium Cooled Lithium Lead (HCLL) breeding blanket concept, the CEA Saclay,
under contract with the European Joint Undertaking “Fusion for Energy”, has in charge the design of
a Test Blanket Module (TBM) for ITER, based on the HCLL concept. This module features 8 MPa
helium coolant circulating through channels in the plates constituting its box structure, liquid PbLi
eutectic circulating in the enclosed volume, a structure temperature ranging from 300 to 550°C, and a
specific Reduced Activation Ferritic Martensitic steel (Eurofer) as constitutive material. According to
its functional features and its design, the TBM is a “multi-chamber” nuclear pressure equipment. This
paper provides information about the design of this complex component and the associated on-going
analyses, for which design rules are based on RCC-MR 2007 code, completed with ITER Structural
Design Criteria to take into account 14 MeV neutron irradiation effects; moreover, developments needs
in terms of design rules and manufacturing are also discussed.
INDEX çç Sommaire
1. INTRODUCTION
In a Tokamak thermonuclear fusion power plant, the breeding blankets are the components
facing the plasma playing three roles: heat recovery, Tritium breeding, and radiological
protection. Several concept studies are on-going for the selection of a candidate technology of
breeding blankets to be used for a demonstration power plant, named DEMO. On the way to
DEMO, the experimental reactor ITER, aimed at gaining the data necessary to the design and
operation of the first electricity-producing fusion power plant, will provide the ability to test
key technologies, such as concepts of breeding blankets. Dedicated ITER ports will allow
testing several concepts, through devices called Test Blanket Modules (TBM). The CEA
Saclay, responsible for the development of the Helium Cooled Lithium Lead (HCLL)
breeding blanket concept, has in charge, under contract with the European Joint Undertaking
“Fusion for Energy”, and through a consortium agreement (TBM-CA) the design of a TBM
based on the HCLL concept.
This paper provides information about the design of this complex component and the
associated on-going analyses, for which design rules are based on RCC-MR 2007 code,
completed with ITER Structural Design Criteria to take into account 14 MeV neutron
irradiation effects; moreover, developments needs in terms of design rules and manufacturing
are also discussed.
2. GLOSSARY
INDEX çç Sommaire
plasma radiation and conversion of neutrons kinetic energy; active cooling is ensured
by channels embedded in constitutive plates;
a liquid metal breeder (PbLi eutectic) as both neutron multiplier and breeder material,
where Tritium is produced through neutron capture by Li atoms; the PbLi is slowly
circulating in the internal volume of the module;
a radiological shielding of the Vacuum Vessel (VV) and other surrounding components
of the reactor; this function is ensured passively by the module thickness and content.
In order to fulfil these objectives, the TBM must exhibit, in particular, functional and
structural features similar to the corresponding DEMO blanket concept; in the case of the
HCLL concept:
a Eurofer97 steel “compact” box, with DEMO geometrical layout principles (cooling
channels, stiffeners, manifolds…);
an active cooling of the structure by circulating 8 MPa Helium with 300°C/500°C
inlet/outlet temperatures, the maximum structure temperature being limited to 550°C;
a PbLi eutectic circulation in the internal volume;
similar manufacturing technologies for the assembly of the structure.
The integration to the ITER reactor completes the specifications and requirements that define
the TBM. In particular:
the external dimensions of the module are dictated by the allocated space in the ITER
VV;
the loading conditions, and in particular the heat fluxes and power depositions, are
ITER-specific;
the supporting structure is ITER-specific.
It has to be mentioned that the TBM is a part of a complete system, the Test Blanket System
(TBS), which includes all the TBM-associated sub-systems, such as, in the case of the HCLL
concept, a helium loop, a PbLi loop, a purification system, a tritium extraction system, as well
as monitoring and control systems. The TBS requires a tight collaboration with ITER
Organisation for its integration to the Tokamak and in the buildings.
Several versions of TBM are foreseen, for progressive qualification and different test
objectives according to the various ITER test phases. However, these versions share the same
design for the “withstanding” structure, i.e. playing a mechanical role, called the “box”.
INDEX çç Sommaire
The grid defines an array of internal cells for the breeder units (BUs). During operation, each
cell is filled with slowly flowing Pb-15.7Li. The liquid metal heat removal is ensured by
parallel horizontal cooling plates (CPs), connected to a BU back plate providing the insert
rigidity. CPs are cooled by He flowing in 8 parallel channels forming three U-turns in a way
that He enters and exits from the rear where it is collected in two manifolds located behind the
BU back plate.
Figure : Overview of the HCLL TBM Box, and details of internal structures
The cooling/heat extraction of the structure is insured by pressurised Helium distributed and
organised by the back manifold, and circulating in the thickness of the plates constituting the
“front” structure. The flow scheme principles and main characteristics are described in Figure
. The quantitative values provided are the result of the design process, developed in section
Design studies. It has to be mentioned that an important part of the cooling power has to be
diverted through a by-pass pipe after the First Wall, as a result of the different balance
between the heat flux from the plasma and the neutronic power deposition in ITER when
compared to DEMO.
INDEX çç Sommaire
FW
10 channels
2 passes/channel INLET
CP 1.3 kg/s
8 W-shaped 300°C
channels/CP
2 CP/BU
16 BU
SP(+Caps in //)
3 W-shaped
channels/hSP BY-PASS
14 hSP 0.975 kg/s
8x3 W-shaped 389°C
channels/vSP or
Cap
OUTLET
0.325 kg/s
#500°C
P
R
The tritium breeding function is ensured by liquid PbLi circulating in the inner volume
defined by the structure, according to the scheme given in Figure .
Figure : Principle of the PbLi circulation in the HCLL TBM, and PbLi volume
INDEX çç Sommaire
Figure : Detailed view of the TBS inside the Equatorial Port
The environmental conditions, combined with the various ITER and TBS operating
conditions, lead to the definition of numerous loading conditions, corresponding to the
combinations of elementary load cases, which are used for the design studies of the TBM.
Two of them have been assessed in the present studies. They are synthesized in Table .
These loading conditions are important for the definition and assessment of the design
parameters and of the structure, as they rely on:
a conservative assumption of the nominal operating condition, where the heat flux and
heat deposition are maximised;
a maximum allowable pressure PS, in the sense of the PED directive Directive
97/23/EC of the European Parliament and of the Council of 29 May 1997 on the
approximation of the laws of the Member States concerning pressure equipment, OJ L
181, 9.7.1997, p. 1, that has been calculated on the basis of usual margins to be taken
for a rupture disk, which is the safety device chosen to be installed on the Helium
Cooling System (HCS);
an accidental scenario, assuming an internal leakage leading to a communication of
the pressurised He compartment with the liquid metal compartment; the pressure
considered in this case is corresponding to the PS value; thus, a double failure is
assumed: a failure of an internal component, and a failure of the Pressure Control
System of the HCS.
5. Material
The candidate material for the HCLL TBM is the Eurofer97 RAFM steel. RAFM steels are
derived from the conventional modified 9Cr-1Mo steel, but with the high activation elements
(Mo, Nb, Ni, Cu and N) eliminated or replaced by their equivalent low activation elements
(e.g. W,V and Ta) (D. S. Gelles, Development of martensitic steels for high neutron damage
applications. Journal of Nuclear Materials 239 (1996) 99-106., E.E. Bloom, The challenge of
developing structural materials for fusion power systems. Journal of Nuclear Materials 258-
263 (1998) 7-17.). Their main advantages over austenitic stainless steels are related to the
excellent dimensional stability (low swelling) under neutron irradiation. In addition, they
present lower coefficients of thermal expansion and higher coefficients of thermal
conductivity at high temperatures: at 500°C RAFM steels have a surface heat capability factor
about 2.5 times higher than that of austenitic steels (F. Tavassoli, Present limits and
INDEX çç Sommaire
improvements of structural materials for fusion reactors. Journal of Nuclear Mater. 302 (2000)
73-88., S. J. Zinkle, N. M. Ghoniem, Operating temperature windows for fusion reactor
structural materials. Fus. Eng. & Des. 51-52 (2000) 55-71.). In the past years, development of
RAFM steels has been the object of extensive cooperation between the different ITER parties;
Eurofer97 is one of the variants developed in the European Union within the framework of the
EFDA (T. Muroga, M. Gasparotto, S.J. Zinkle, Overview of materials research for fusion
reactors. Fusion Engineering and Design 61-62 (2002) 13-25.-B. van der Schaaf et al., The
development of EUROFER reduced activation steel. Fusion Engineering and Design 69
(2003) 197-203.). Given the lack of data concerning this type of steel in present-day design
codes, a characterisation program has been launched in the past years to address the short-
term needs related to TBM design activities (R. Lindau, A. Möslang, M. Schirra, Thermal and
mechanical behaviour of the reduced-activation-ferritic-martensitic steel EUROFER. Fusion
Engineering and Design 61-62 (2002) 659-664., F. Tavassoli et al., Material design data for
reduced activation martensitic steel type Eurofer. Journal of Nuclear Materials 329-333 (2004)
257-262.). The data produced in the framework of this programme have been collected in F.
Tavassoli, Fusion Demo Interim Structural Design Criteria (DISDC)/Appendix A Material
Design Limit Data/A3.S18E Eurofer Steel, EFDA TASK TW4-TTMS-005-D01. CEA Report
DMN/DIR/NT/2004-02/A, December 2005.. While the database of Eurofer97 mechanical
properties is not yet up to par with that of more commonly used austenitic steels, the recent
availability of additional data concerning fatigue, creep and irradiation induced changes in
material properties have brought the characterization of Eurofer97 at least to the level of other
9Cr steels included in nuclear design codes.
1. PED/ESPN classification
Since ITER will be built in France, TBMs must fulfil French regulations on pressure vessel
equipments, specifically in its nuclear extension, defined by the “Arrêté du 12 décembre 2005
relatif aux équipements sous pression nucléaires” (ESPN order) Arrêté du 12 décembre 2005
relatif aux équipements sous pression nucléaires, NOR: INDI0506414A, Rev. 22/1/2009..
According to ESPN, Nuclear Pressure Equipments (NPE) are classified into three levels, from
N1 to N3, in relation to the significance of the radioactive emissions possibly resulting from
their failure, and in 5 categories according to the pressure hazard: Category 0, I, II, III and IV.
A preliminary assessment of the PED/ESPN classification of the TBM has been performed,
the main results being summarized in Table .
According to the guideline 2/19 of the PED, a TBM is a multi chamber pressure equipment
with common boundaries. They constitute chambers of the same vessel.
Technical requirements and conformity assessment procedure shall be applied as follows:
Helium box shall be designed according to the technical requirements for category IV
of ESP decree;
PbLi box shall be designed according to the technical requirements of category I and
level N2 of French nuclear order.
INDEX çç Sommaire
The conformity assessment procedure that applies to the whole TBM is based on the
highest category of the chambers. So it is one of those for nuclear level N2 and hazard
category IV equipment: B+F or G.
5. DESIGN STUDIES
Design analyses have been supported by numerical models used to quantify:
the expected performances of the box from the thermal-hydraulic point of view
(temperature levels, heat extraction capability), including reachable DEMO-relevancy
level;
the accordance of the geometry to design criteria, and mainly related to thermo-
mechanical behaviour of the box.
The computational models that have been used for these analyses were developed according
to the following guidelines:
They were generated directly from CAD models of the box geometry.
The suitable extent has been determined to be a “quarter” of the box, delimited by the
radial-poloidal (vertical) and radial-toroidal (horizontal) mid-planes.
INDEX çç Sommaire
Finite Elements (FE) method was used for the computations. The FE mesh (see Figure
) is based on 3D tetrahedron elements, using a linear formulation for thermal analyses,
and a quadratic formulation for mechanical analyses. On all surfaces where heat
exchange with coolant takes place, the FE model has been coupled with an explicit 1D
finite difference model of the helium flow.
Design analyses have been performed according to the following scheme:
- A thermal analysis determining the thermal field corresponding to the assessed
situation. The volumetric power deposition in the structure has required a former
neutronic analysis, which is not discussed here. The coupling of hydraulic and thermal
phenomena has allowed obtaining the temperature field in the structures and the
helium temperature evolution along the cooling channels.
- A linear mechanical analysis, using, in addition to the mechanical loads of the
assessed situation, the corresponding thermal field previously computed.
- In both calculations (thermal and mechanical), steady state analyses have been
performed at first, followed by transient time-dependant calculations simulating the
pulsed operation of plasma in ITER.
- A post computation of the results used to determine the accordance to the design
criteria of the various damage types.
1. Thermal studies
The goal of the thermal-hydraulic studies is to reach a set of operating parameters (cooling
parameters) that ensure, for each of the assessed operating conditions, the respect of the
imposed temperature limits. These limits are fixed the same as DEMO ones in order to
guarantee DEMO relevancy. Other requirements are furthermore imposed by the specificities
INDEX çç Sommaire
of the ITER/TBS environment. The thermal design process takes into account the following
constraints:
Eurofer minimum and maximum allowable temperatures are fixed at 300°C and 550°C
respectively.
He inlet and outlet temperatures are fixed at 300°C and 500°C respectively.
PbLi/Eurofer interface temperature is fixed at 550°C.
The same cooling scheme of DEMO shall be used to cool the subcomponents. Since in
TBM there is a strong unbalance between the power deposition in the First Wall and in
the Breeding Zone (when compared to DEMO), an important part of the He mass flow
has to be diverted with a by-pass before entering the BZ.
The upper limit of the global He mass flow is fixed to 1.4 kg/s.
The geometry of the components (and of the cooling channels) being fixed at this stage of the
analysis, the results of the calculations give the required He mass flow in the FW and the
fraction that has to be diverted before the BZ. Because of heat recuperation effects between
the FW and the BZ, this has to be the result of an iterative process. The results presented are
the final step of this process.
MANIFOLD
1.3 kg/s
1.3 kg/s Stage 1
Tin=300°C
Tout= 302°C
FW
Tin=302°C
Tout= 387°C
1.3 kg/s
SG CAPS
Tin=389°C Tin=389°C
Tout=477°C Tout=450°C
TBM outlet
Tavg=417°C
INDEX çç Sommaire
Figure : Temperatures and He mass flows in the cooling system of the TBM.
Transient analysis
The main concern in the study of the design load case LC1 is the hypothesis that plasma
pulses are sufficiently long to attain stationary temperature conditions (and that these
conditions are the most demanding for the mechanical withstanding of the box structure). This
assumption is certainly not true as ITER operating mode is based on plasma pulses, the
typical pulse for the D-T phase being 30 sec ramp-up, 400s plateau, 60s ramp-down, total
duration 1800s. A simulation of this thermal transient has been performed, with the objective
to assess time-dependant mechanical criteria.
The results (Figure ) show that:
Apart from the FW, stationary conditions are never reached. Helium outlet temperature
at the end of the plasma pulse is ~435°C.
There is an inversion of the thermal gradient between the FW and the BZ at the
beginning and at the end of the plasma pulse (both in average and maximum
temperatures).
Three time instants are significant for the thermo-mechanical analysis:
- t1=60s: the instant when the FW reaches nearly stationary conditions.
- t2=430s: the end of the plasma pulse, when temperatures in TBM structures are the
highest.
- t3=600s: the cooling phase, when the inversion of the temperature differential between
the First Wall and the Breeding Zone is maximal.
INDEX çç Sommaire
-
500
Tmax FW Tmax SG
450
T (°C)
400
350
300
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800
t (s)
Figure : Evolution of maximum temperature in TBM sub components during a typical ITER plasma pulse.
INDEX çç Sommaire
noted that additional damage modes related to neutron irradiation (irradiation induced creep
and swelling) are not a concern for Eurofer97 (at least in ITER).
The possibility/necessity to address these damage modes is depending however on several
factors: Eurofer97 physical and mechanical properties, availability of the data required by the
codes, applicability/degree of conservatism of the rules contained in the codes with respect to
Eurofer97. The set of performed analyses have offered the possibility to identify the present
situation with regard to these issues (see §Accordance of the design rules / integrity criteria,
with the Eurofer97).
Figure : LC1, primary + secondary equivalent Von Mises stress intensity (MPa) on deformed model
(x100 amplification).
INDEX çç Sommaire
Criteria for ratcheting are exceeded in several zones of the FW, while the maximum number
of cycles that the module can withstand according to criteria given by the codes is about 1000,
while 7500 cycles are foreseen for TBM components. Given the fatigue-softening behaviour
of Eurofer97 (see §Accordance of the design rules / integrity criteria, with the Eurofer97),
elasto-plastic analyses are not expected to improve these results and the design of the TBM
will probably have to be revised.
Figure : Deformation pattern of the FW at selected instants of the plasma pulse (x100 amplification).
Analysis of LC3
The analysis performed for LC3 corresponds to the initial conditions in case of a LOCA. The
stationary temperature field is assumed for evaluating stress limits. The evolution of the
temperature distribution after the local depends on the time needed to shut down the plasma.
If the shutdown is (nearly) immediate, this case can be considered conservative. The primary
Von Mises equivalent stress intensity and the corresponding deformation of the structure are
shown in Figure . Results show that the most critical component is in this case the Stiffening
Grid, whose function is precisely to insure the integrity of the box in case of internal
pressurization. Peak stresses appear at local junctions between the SG and the FW, but criteria
are also exceeded in correspondence of the He channels where the thickness of the plates is
reduced.
INDEX çç Sommaire
Figure : LC3, primary equivalent Von Mises stress intensity (MPa) on deformed model (x100 amplification).
6. MANUFACTURING
The manufacturing of the TBM is a challenge. The complexity of the structure, which requires
compactness, ability to withstand to severe conditions (temperature, high pressure, corrosion,
neutron fluence...), high level of confidence, tightness to Helium gas, as well as its specific
material, involves an extensive development of manufacturing techniques. Such a work has
started several years ago, in the frame of EFDA programs. It is still on-going, in the frame of
the European TBM project handled by F4E. This section highlights some of the ways that are
explored to achieve the manufacturing of the HCLL TBM, as well as their status.
1. Sub-components manufacturing
The manufacturing of the cooled components, i.e. plates with embedded meandering channels
(§Main structure description) has a dedicated development. The fabrication process relies on
Hot Isostatic Pressing (HIP), a technique that enables diffusion welding of complex parts. HIP
is used primarily to remove porosities and other kind of closed defects from materials. This is
achieved through the application of gas at high pressure and high temperature. Gaps and pores
at the interface between two plates can be removed, thereby leading to diffusion welding,
provided that the gas is prevented from filling the interface (the periphery of the interface
must be sealed).
Typical HIP parameters for Eurofer97 are 1000-1100°C, 200-1200bar. Several HIP techniques
have been investigated Y. Poitevin et al., Development of welding technologies for the
manufacturing of European Tritium Breeder Blanket Modules, ; the candidate ones depend on
the component, according to its dimensional characteristics. One can distinguish:
The First Wall
The Side Caps and the Stiffening Plates
The Cooling Plates
One technique is based on the use of tubes embedded between plates. During HIP, tubes are
open at their ends so they do not collapse. Rectangular tubes and thin plates can be used.
Alternatively, thin round tubes associated to grooved plates can be chosen (Figure ). In that
INDEX çç Sommaire
case, either the tubes are formed before HIP (using hydraulic expansion), either expansion is
achieved in situ thanks to the HIP gas pressure. These techniques are not applicable to the
Side Caps, SP and CP because these components exhibit small channels with sharp U-turns,
but their application to the FW deserves attention. Looking in more detail, there are many
fabrication options. First of all, the structure can be bent before HIP (i.e. tubes and plates are
bent individually and then diffusion welded) or after HIP. The feasibility of bending after HIP
depends on the FW bending radius and the dimensional accuracy requirements. Bending
before HIP is preferred. In the case of rectangular tubes, accurate bending is difficult. This is
why thin round tubes are preferred.
Figure : 3 channels FW mock up made using grooved plates and thin tubes. HIP joint as red line.
Another technique consists in diffusion welding of two symmetric grooved plates with the
help of a two steps process. In the first step, a low pressure HIP cycle is applied onto the pair
of plates. Channels are encapsulated, no gas enters the channels. A low pressure is required in
order to avoid channel deformation. After this first step, the plates are tightly bonded but the
joint mechanical properties are not optimum. So the channels are opened by drilling and a
second high pressure HIP cycle is applied. This process is applicable mainly to the side caps,
the stiffening plates and the first wall (bending after HIP). One difficulty of this process is
that, since the minimum available pressure in a HIP is about 20MPa, the first HIP cycle may
in some instances lead to unacceptable deformation. This is why the success of this process is
design dependent. Another issue is that an area not sealed by the first HIP cycle will not be
sealed during the second HIP cycle.
Last, a process has been developed that avoids the drawbacks of the two steps process while
being suitable to all sub components. It consists in sealing grooves by welding strips onto
their top and then adding a cover using HIP. During HIP, the channels do not deform because
it is possible to let the gas fill the grooves. Furthermore, the welds are austenitised and
homogenized during HIP. A CP mock up has been manufactured; it withstood successfully
thermo-mechanical cycling E. Rigal et al., Manufacturing a HCLL cooling plate mock up,
Fusion Engineering and Design 83 (2008) 1268-1272. Figure shows this mock up which
features 16 parallel channels 4x4.5mm. For testing purposes, flanges have been integrated to
the mock up.
INDEX çç Sommaire
Figure : Cooling plate mock up. Top: principle of the process, bottom left: view of the mock up, bottom right:
detailed view.
2. Structure assembly
The TBM structure has to be designed and assembled in conformity with the requirements
imposed by its PED/ESPN classification. In addition to the use of design rules presumably
compatible, the assembly techniques have to be fully qualified. The specificities of the TBM
in this area are the following:
a new material, which has not been standardised;
a compact structure, with limited access;
multi-compartments volumes, providing limited inspection possibilities;
tight welded junctions;
up to 30 mm thickness for fully penetrated joints.
The fusion welding processes are envisaged for the assembling of SP and the integration of
SG, BU and the BP into the box. The choice of welding process depends of numerous
parameters:
compatibility with the material properties (chemical, microstructure, mechanical…)
accessibility of the welding tool
reliability and automatization (reproducibility) of process
distortion induced by the welding operation
Non Destructive Test of weld
mechanical properties of weldment
background (R&D, industrial)
compatibility with the manufacturing code
compatibility of process with the industrials facilities and cost
An important quantity of work has been produced in past studies within the frame of EFDA
programs. They have addressed punctual issues but are not necessarily relevant of the fully
integral manufacturing process, and in particular of the cumulated effects (heat treatments,
induced distortions…) which are the main objectives of the further developments discussed
here.
GTAW and laser YAG processes have been investigated on Eurofer material. The two welding
processes could be applied and are complementary. GTAW is a low productivity welding
process and permits to weld from low to high thickness. The narrow gap welding could be
used to take into account the channel environment and minimize the distortion. The automatic
INDEX çç Sommaire
and manual configurations are used for respectively nominal condition and repairing or
nominal conditions. For low thickness, YAG laser permit to realize narrow weld in full
penetration and produce less distortion than GTAW process.
If the two processes could be both applied for the same weld, they should be discriminated in
terms of metallurgical weldability, weld defects requirements, welding tool accessibility,
reliability of the welding conditions, repair, industrial capability…
The YAG laser process seems more flexible and adapted to the TBM geometry. By adapting
the welding energy and sequence, sound welds are obtained in both cases.
In the case of high thickness structure to weld, the electron beam has been investigated. Due
to the presence of delta ferrite in the weld metal, which could degrade the mechanical
strength, this welding process is not selected for the TBM manufacturing.
Generally, a pre and post heating on 9Cr steel are recommended to avoid cold cracking. The
exemption of this preconisation is adapted following the welded thickness and could be
assessed via technologies tests. Because of the temperature range to meet, the sensitivity of
laser tools in a hot and confined environment has to be assessed. In both cases, PWHT is
necessary to temper the weak martensite in the weld zone, reinforce the Heat Affected Zone in
view to obtain mechanical characteristics comparable to base material (ductility, toughness…)
and reduce the residual stresses. The effect of the PWHT on each Eurofer weld (metallurgical
and mechanical) has to be measured: the whole production of the TBM (in starting from the
box to the back plates) could need to do a PWHT after each production step by fusion
welding. In order to be able to evaluate the feasibility of all the welding sequences without
decreasing too much the mechanical characteristics of each weld in reason of the cumulated
time in PWHT, it is necessary to define the typical PWHT ranges (adjusted with each welded
thickness) for each specific geometry of the TBM subcomponents. This work could be carried
out in performing computing of heat treatments and experimental validations on
representative geometries, with usual acceptance criteria (tensile test, Charpy test…).
For subcomponent manufacturing and structure assembly, the development of each welding
(squeeze, tacking, welding, PWHT, repairing…) procedure implies the production of specific
clamping and welding tools. To optimize the quality of welds and minimize the repairing
operation, it should be taken into account the specific inspection tools to characterize the
welding operation during welding (observation of the welding pool, joint tracking...). For the
same reasons (quality, reproducibility, welding of heated workpieces), mechanised and
automatic welding tools should be promoted.
Due to the important welded length and the complex geometry, it is important to assess and to
integrate dimensional control for each production step. The assessment of the distortions
induced by each welding step has to be characterized; straightening methods have also to be
developed. Experimental and numerical developments could be performed for helping in the
definition of the chamfer geometry and of the welding sequence.
The quality of the welded structure should be assessed using, metallurgical, mechanical
characterizations and non destructive inspection. The using of standardized mechanical
samples is very often not possible. Then intermediate mock-ups have to be developed to take
into account the specific geometry of the TBM and the specific welding procedures developed
(fusion welding + HIP by example).
Repairing procedures and non destructive tests have to be developed and qualified for each
weld. The developed methodology for the NDE inspection must be adapted to the design of
sections, the geometry of the welds and in taking into account the material microstructure and
constraints of the geometrical environment. To help to the inspectability developments
required for the TBM, the preparation procedure of the surfaces, the NDE tools and their
integration in the HCLL TBM structure could be assessed via experimental and numerical
developments.
INDEX çç Sommaire
7. CHALLENGES IN THE PRESENT DESIGN
The present status of the design studies raises challenging issues in several areas.
INDEX çç Sommaire
cycles some little strain-hardening may take place) and then stabilizes at a
value which decreases slowly as a function of the number of cycles, dropping
sharply just before failure of the specimen. Fatigue design curves at several
temperatures are available.
Effects of irradiation on fatigue More experimental tests are needed in order to obtain the fatigue design
behaviour curve in irradiated conditions. For temperatures above 350°C however,
unirradiated values can be used.
Table : main characteristics of the Eurofer97 mechanical properties
The limited ductility of Eurofer97 is the main issue for its use as structural material in nuclear
components. It should be recalled that ESPN defines minimum allowable values for the
properties related to the ductility of the materials used in the construction of nuclear pressure
equipments. These limits are not present in SDC-IC, but are instead considered in RCC-MR.
Of even more concern are the changes in material properties (i.e. embrittlement and loss of
strain-hardening) following neutron irradiation. The uniform elongation of Eurofer97 is
already lower than 2% starting from 400°C even in unirradiated conditions. At 300°C it falls
to 0.2% already at 0.7 dpa. Those values are comparable to those of irradiated austenitic steels
above 10 dpa. Eurofer97 meets the requirements defined by the regulations at BOL but not in
the irradiated conditions foreseen in TBMs. In particular at 3 dpa the DBTT is shifted above
0°C (Table ).
ESPN/RCC-MR Eurofer97 at BOL Eurofer97 at 3 dpa
Limit
t (RT) >14% 22% ~14% (300°C)
Toughness(0°C) >40 J 220 J ~200 J (300°C)
~20 J (RT)
Sy/Su (RT) <0.85 0.83 0.92 (300°C)
Table : Minimum requirements on the ductility of the materials as defined by ESPN/RCC-MX and comparison
with Eurofer97 properties in unirradiated/irradiated conditions.
It should be noted that ESPN does not state that the limits defined in Table should be verified
under irradiated conditions, but only that the effects of irradiations should be taken into
account in the design of the component. SDC-IC defines in fact additional design rules when
irradiation is relevant. Moreover, 300°C is the most critical irradiation temperature for
Eurofer97, while irradiation does not seem to affect the ductility and strain-hardening
capability of the material above 400°C (operating temperature of TBM components are in the
range 350°C-550°C). Nonetheless, the ductility properties of Eurofer97 at BOL are already
sufficiently low that the use of more stringent design rules appears to be justified. However,
limits defined by the codes (mainly for austenitic steels) will have to be checked and adapted
to Eurofer97 mechanical properties.
Beside the specificities of the mechanical properties previously mentioned, neutron irradiation
causes additional creep and swelling of the material. Both SDC-IC and RCC-MX introduce
additional criteria to take into account these phenomena, either in the form of additional
design rules or by increasing the safety factors. However, irradiation-induced creep and
swelling can be neglected in case of Eurofer97, since ferritic/martensitic steels show a good
resistance to these phenomena. According to F. Tavassoli, Fusion Demo Interim Structural
Design Criteria (DISDC)/Appendix A Material Design Limit Data/A3.S18E Eurofer Steel,
EFDA TASK TW4-TTMS-005-D01. CEA Report DMN/DIR/NT/2004-02/A, December
2005. swelling is negligible at least up to 60 dpa, while creep at a stress of 150 MPa is
negligible up to 5 dpa. Both values are above the level of dpa expected in TBM.
INDEX çç Sommaire
Local junctions between Peak stresses are observed Geometrical singularitiesAn optimisation of the
components (mainly in several junction areas, leading to peak stresses.geometry of local
Stiffening Grid and Back under LC1 and LC3 junctions is necessary to
Manifold) loading conditions. remove sharp singularities.
First Wall (bent area) Progressive deformation Strong differential thermal An overall reduction of the
criteria are not fulfilled expansion due to thermal plates’ thickness could be
under loading conditions gradient in sub- envisaged to reduce
LC1. Allowed number of components and between thermal gradients, but
cycles is lower than sub-components, with implies (possibly) major
expected. secondary stresses modifications of the
enhanced by different design. Margins are
stiffness between already very low.
components.
Stiffening Grid Immediate plastic collapse Lack of thickness of the Further assessment on the
and stability criteria are Stiffening Plates, or accidental scenario
not fulfilled under loading channels height too high quantitative data.
conditions LC3 (accidental with regard to plates’ Modification of channels’
internal pressure). thickness. Accidental cross-section. Re-design of
pressure value has been the singular zones
increased by 20% since (junctions).
previous design studies
(see Uncertainties on
the design loads
(pressure, thermal)),
and module radial depth
increased.
First Wall Under transient situation Due to different thermal The operating conditions
(plasma pulse) the stress time constants between should be carefully
intensity range is very First Wall and Breeding considered to assess the
significant, and current Zone, alternate states of possibilities to limit the
criteria cannot be fulfilled secondary stresses occur. amplitude of variation of
(ratcheting and fatigue). thermal states.
Table : Review of the mechanical integrity issues in the present HCLL TBM design
The next priority of the design process is to address these issues, by means of dedicated
models and studies, combined with the assessment of possible margins in the requirements
and in the operating scenarios.
8. CONCLUSION
This paper draws an overview of the on-going activities related to the design of a breeding
blanket mock-up (Test Blanket Module – TBM) to be installed in thermonuclear fusion
experimental reactor ITER. The complexity of the structure, the severe loading conditions it is
submitted to, its classification as nuclear pressure equipment, make it a challenging test object
to design. Present geometry, reference design rules, computational studies, discussion on
material, manufacturing R&D, and issues have been presented.
INDEX çç Sommaire
Future developments will focus on:
the improvement of the mechanical behaviour and of the compliance with design rules
through studies for the optimisation of the design;
further implication in the implementation of design rules adapted to ferritic-
martensitic material, shared by several high temperature applications;
qualification of manufacturing processes.
Aknowledgements
This work, supported by the European Communities under the contract of Association
between EURATOM and the TBM-CA, was carried out within the framework of Fusion for
Energy TBMs development. The views and opinions expressed herein do not necessarily
reflect those of the European Commission.
References
[1] D. S. Gelles, Development of martensitic steels for high neutron damage applications.
Journal of Nuclear Materials 239 (1996) 99-106.
[2] E.E. Bloom, The challenge of developing structural materials for fusion power
systems. Journal of Nuclear Materials 258-263 (1998) 7-17.
[3] F. Tavassoli, Present limits and improvements of structural materials for fusion
reactors. Journal of Nuclear Mater. 302 (2000) 73-88.
[4] S. J. Zinkle, N. M. Ghoniem, Operating temperature windows for fusion reactor
structural materials. Fus. Eng. & Des. 51-52 (2000) 55-71.
[5] T. Muroga, M. Gasparotto, S.J. Zinkle, Overview of materials research for fusion
reactors. Fusion Engineering and Design 61-62 (2002) 13-25.
[6] N. Baluc et al., Status of reduced activation ferritic/martensitic steel development.
Journal of Nuclear Materials 367-370 (2007) 33-41.
[7] B. van der Schaaf et al., The development of EUROFER reduced activation steel.
Fusion Engineering and Design 69 (2003) 197-203.
[8] R. Lindau, A. Möslang, M. Schirra, Thermal and mechanical behaviour of the
reduced-activation-ferritic-martensitic steel EUROFER. Fusion Engineering and
Design 61-62 (2002) 659-664.
[9] F. Tavassoli et al., Material design data for reduced activation martensitic steel type
Eurofer. Journal of Nuclear Materials 329-333 (2004) 257-262.
[10] F. Ta v a s s o l i , Fusion Demo Interim Structural Design Criteria
(DISDC)/Appendix A Material Design Limit Data/A3.S18E Eurofer Steel, EFDA
TASK TW4-TTMS-005-D01. CEA Report DMN/DIR/NT/2004-02/A, December 2005.
[11] Directive 97/23/EC of the European Parliament and of the Council of 29 May
1997 on the approximation of the laws of the Member States concerning pressure
equipment, OJ L 181, 9.7.1997, p. 1
[12] Arrêté du 12 décembre 2005 relatif aux équipements sous pression nucléaires,
NOR: INDI0506414A, Rev. 22/1/2009.
[13] Design and Construction for Mechanical Components of FBR nuclear islands
(RCC-MR), AFCEN, 2007 edition.
[14] Règles de conception et de construction des matériels mécaniques des
réacteurs expérimentaux de leurs auxiliaires et des dispositifs d’irradiation (RCC-
MX). CEA Edition 2008.
[15] Y. Poitevin et al., Development of welding technologies for the manufacturing
of European Tritium Breeder Blanket Modules, To be published in Journal of Nuclear
Materials, Proceedings of ICFRM-14 (Sapporo, Japan), under review
[16] E. Rigal et al., Manufacturing a HCLL cooling plate mock up, Fusion
Engineering and Design 83 (2008) 1268-1272
INDEX çç Sommaire
[17] P. Fernandez et al., Creep strength of reduced activation ferrit/martensitic steel
Eurofer97. Fusion Engineering and Design 75-79 (2005) 1003-1008.
[18] G. Yu, N. Nita, N. Baluc. Thermal creep behaviour of the EUROFER 97 RAFM
steel and two European ODS Eurofer97 steels. Fusion Engineering and Design (75-
79) 2005 1037-1041.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page101
RÉSUMÉ / ABSTRACT
16A
FATIGUE DES ASSEMBLAGES SOUDÉS – UNE ÉTUDE COMPARATIVE
A. Al-Mekhnaqi, Ph.D.
Senior Analyst, Stress Engineering Services, Inc.
13800 Westfair East Drive, Houston, Texas 77041, USA
[email protected]
ABSTRACT
This paper compares fatigue analysis results obtained by the Structural Stress Method used in the 2007
ASME Boiler and Pressure Vessel Code, Section VIII, Division 2 to ones obtained by the conventional
2004 ASME Section VIII, Division 2 method. Two case studies were described – an undefined weld
high-cycle fatigue example and a defined-profile weld low-cycle fatigue example. For these test cases,
the following was found: In the undefined weld high-cycle fatigue example, fatigue life calculated by
the 2007 ASME SSM (-3σ) was about four times that calculated by the conventional 2004 ASME
Method using nominal stress and FSRF of 4, and when linearized stresses (from stress integration
values) were used instead of nodal forces, fatigue lives were up to 17% higher. The defined-profile weld
low-cycle fatigue example showed that fatigue life calculated by the 2007 method (-3σ) was less than
one eighth that calculated by the conventional 2004 ASME method using peak stress.
INDEX çç Sommaire
FATIGUE ANALYSIS OF WELDED COMPONENTS - A COMPARATIVE
STUDY
ABSTRACT
This paper compares fatigue analysis results obtained by the Structural Stress Method used in the 2007
ASME Boiler and Pressure Vessel Code, Section VIII, Division 2 to ones obtained by the conventional
2004 ASME Section VIII, Division 2 method. Two case studies were described – an undefined weld
high-cycle fatigue example and a defined-profile weld low-cycle fatigue example. For these test cases,
the following was found: In the undefined weld high-cycle fatigue example, fatigue life calculated by
the 2007 ASME SSM (-3σ) was about four times that calculated by the conventional 2004 ASME
Method using nominal stress and FSRF of 4, and when linearized stresses (from stress integration
values) were used instead of nodal forces, fatigue lives were up to 17% higher. The defined-profile
weld low-cycle fatigue example showed that fatigue life calculated by the 2007 method (-3σ) was less
than one eighth that calculated by the conventional 2004 ASME method using peak stress.
INTRODUCTION
The Structural Stress Method (SSM) is a technique for analyzing fatigue of welded structures that was
developed to overcome shortcomings of the classic notch-stress approach such as mesh-dependency
and reliance on smooth-specimen test results. In 2007, the method was incorporated into the ASME
Boiler and Pressure Vessel Code (BPVC), Section VIII, Division 2 and the API 579-1/ ASME FFS-1
Fitness-For-Service Standard. This paper compares results obtained by SSM to ones by the 2004
ASME Section VIII, Division 2 method.
Using ASME terminology, a stress profile across a section can be linearized into membrane (M),
bending (B), and peak (F) stresses. The Notch Stress Method (NSM) is the classical fatigue
assessment technique that is based on the total stress at the location of interest including peak stress
(F). The Equivalent stress used in NSM is the max stress intensity value of the stress difference
between any two points in time at a given location.
Prior to 2007, ASME BPVC used a methodology for assessing fatigue life that is based on fatigue test
data from smooth bar specimens. The same approach was used for welded and unwelded structures.
The stress used for the assessment was either the total stress that included all of the above stress
components, or nominal membrane and bending (M+B) stresses adjusted using experimentally-
INDEX çç Sommaire
derived multipliers such as Fatigue Strength Reduction Factors (FSRF). This methodology suffered
from three main shortcomings:
1. The estimation of notch stresses in non-machined welded structures is unreliable. Peak stresses
obtained by numerical techniques, such as finite element analysis, are dependent on the
fineness of the mesh and the assumed geometry of the weld. On the other hand, the
experimentally-derived multipliers that are used with nominal stresses are crude estimates that
often result in too conservative answers.
2. Fatigue curves are based on smooth-specimen test results which behave differently from
welded components. Fatigue of welded components is dominated by crack propagation while
smooth-specimen contains both crack initiation and propagation.
The SSM was developed by Battelle through a joint industry project in the late 1990s. The objective
was to develop a mesh-insensitive technique that produces robust and reliable results that correlate
well with fatigue test data of welded components.
To analyze fatigue damage in welded joints, the SSM uses an equivalent stress range (called structural
stress) that is calculated from nodal forces obtained from linear elastic stress analysis. This structural
stress is obtained from membrane and bending stresses normal to a hypothetical crack plane. Since
peak stresses are not included in the assessment process, the impact of mesh fineness on results is
relatively insignificant.
The curve that relates linear elastic stress to fatigue cycles (or S-N curve) used by this methodology
was obtained by adjusting nominal stress values of available welded fatigue test data to equivalent
structural stress values. Different joint geometries, loadings, and plate thicknesses were included in the
process. As a result, previously scattered fatigue data collapsed into one master curve for all weld
types.
Compared to the pre-2007 fatigue methods of the ASME Code, this approach has the following
advantages:
The procedure adopted by the ASME Sec. VIII Division 2 and API 579-1/ ASME FFS-1 in 2007 is a
more conservative version of the original technique developed by Battelle. This modified version was
used to conduct the comparison in this paper.
Details of the technique can be found either in the ASME Code or API/ASME Standard. [1,2] The
following is a brief overview of the typical steps needed to conduct an assessment:
• Transform nodal forces (and moments) into a local coordinate at weld toe
INDEX çç Sommaire
• Obtain line forces and line moments from nodal forces and moments
• Obtain bending and membrane stresses from line forces and line moments
• Determine the elastically calculated structural stress range and strain range
• Determine the local nonlinear structural stress and strain ranges by solving Neuber’s Rule and
hysteresis loop curve
• Determine the equivalent structural stress range with adjustments to account for the ratio of
bending stress range to total stress range, effective thickness, and the ratio of minimum to
maximum stress.
• Determine fatigue life using adjustments for weld finish, fluid environment, loading frequency,
temperature, and material variables such as grain size and chemical composition. Life can be
calculated per the desired statistical basis of survival.
CASE STUDIES
The following two case studies compare the results of the above two techniques for undefined and
defined weld profiles. In these examples, results of the fatigue evaluation procedures of the 2007
edition of ASME Section VIII Div. 2 [3] are compared to those of the 2004 edition of ASME Section
VIII Div. 2. [1]
¾ Welded neck flange nozzle 1.375 inch thick in the spherical head of a pressure vessel made of
A204, Gr C-1/2Mo with 0.375 inch fillet weld is modeled using second order axisymmetric solid
elements, Figure 1.
INDEX çç Sommaire
(a) Model geometry
INDEX çç Sommaire
Page 4
4 50 0 0 0
400000 1 element
3 50 0 0 0
8 elements
16 elements
300000
Life (Cycles)
2 50 0 0 0
200000
150 0 0 0
10 0 0 0 0
50 0 0 0
0
No tc h S tre s s F S R F =4 Mean .-2σ .-3σ
(a) The 2004 ASME conventional method versus 2007 ASME SSM method
1200000
400000
200000
0
-3 σ -2 σ -σ Mean σ 2σ 3σ
(b) The 2007 ASME SSM using nodal forces versus linearized stresses
INDEX çç Sommaire
Case 2: Defined Weld Profile
¾ The skirt of a coke drum was modeled using fine second order axisymmetric solid elements, Figure 3.
Gussets away from the attachment weld were modeled using plane stress elements.
¾ During each operating cycle, the process applies internal & hydrostatic pressure as well as severe thermal
transients apply on the drum.
900
800
700
600
Life (Cycles)
500
400
300
200
100
0
-3 σ -2 σ -σ Mean
INDEX çç Sommaire
CONCLUSIONS
For the two case studies described in this paper, the following conclusions were made:
a. Fatigue design life calculated by the 2007 ASME SSM (-3σ) was about four times that
calculated by the conventional 2004 ASME Method using nominal stress and FSRF = 4.
b. The 2007 ASME SSM was practically insensitive to mesh density. The conventional 2004
ASME method was highly sensitive to mesh size.
c. For the 2007 ASME SSM, when linearized stresses (from stress integration points) were used
instead of nodal forces, fatigue lives were up to 17% higher for the coarsest mesh (one element
through thickness). For finer meshes, the difference was below 8%.
2. The defined-profile weld low-cycle fatigue example showed that fatigue life calculated by the
ASME SSM (-3σ) was less than one eighth that calculated by the conventional 2004 ASME
method using total stresses.
REFERENCES
[1] ASME Boiler and Pressure Vessel Code, Section VIII, Division 2, the American Society of
Mechanical Engineers, 2007.
[2] API 579-1/ ASME FFS-1 Fitness-For-Service Standard, a joint publication by the American
Petroleum Institute and the American Society of Mechanical Engineers, 2007.
[3] ASME Boiler and Pressure Vessel Code, Section VIII, Division 2, the American Society of
Mechanical Engineers, 2004.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page102
70G Investigation par la méthode des éléments finis sur des appareils à AKKUS N. / GENC G.
pression composites renforcés par des fibres sous chargement de TOPTAS E.
choc / FEM investigation of the fiber reinforced composite pressure MARMARA UNIVERSITY
Mercredi
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page103
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page104
RÉSUMÉ / ABSTRACT
60G
LA CONCEPTION DES TUYAUTERIES EN COMPOSITE
B. Pitrou - SNCT
H. Mallard, L.Boulay - CETIM
RÉSUMÉ
La conception de réseaux de tuyauterie réalisés avec des tubes en matériau composite a été principalement étudiée
dans le cas des Centrales Nucléaires suite à l’expérience acquise sur la corrosion des tuyauteries d’eau brute due
à l’eau de la rivière. Les solutions inox ou acier revêtues néoprène ne pouvaient satisfaire les contraintes chimiques.
Le domaine de fonctionnement imposé vont de -30°C à 120°C, de 8 à 15 bars, des diamètres de 50 à 700 mm.
Cela correspond au domaine des solutions composites à base de résine époxy et de renfort en verre E.
La conception de tels réseaux a posé immédiatement des problèmes de critères de dimensionnement, des
problèmes dans l’utilisation du code CASE 155-2. La quantification des marges calculées et la prise en compte
de l’anisotropie ont toujours été particulièrement soulignées.
Ceci est l’origine directe des études en cours au CETIM dans le cadre d’un groupe de travail spécifique GT4 de
Mercredi
la Commission Chaudronnerie. Elles étaient à rapprocher des préoccupations mentionnées dans les réseaux de
recherche et d’innovation technologiques (R2IT). «L’objectif est de favoriser une recherche pertinente permettant
de lever les verrous technologiques pour le développement en commun de produits et services basés sur de
nouvelles technologies.» Une enquête partielle auprès de 20 entreprises, en particulier ayant un type d’activité
d’Ingénierie et utilisant l’acier et le composite, réalisée au CETIM en juin 2001 indique que les tuyauteries sont
toujours ou partiellement soumises à une pression de fonctionnement significative. Ces entreprises pensent que
ces matériaux sont sous-employés à cause de leur coût et du manque de normalisation. Le marché qu’elles
estiment oscille entre stagnation et augmentation. Les applications potentielles sont la Chimie, les réseaux d’eau,
le Pétrole-Gaz, l’Énergie et l’Environnement. Les matériaux sont en général qualifiés en interne, avec le
fournisseur de matériau. Ces entreprises recherchent un code de construction de type CODETI®.
Cet article présente la démarche de dimensionnement des tuyauteries en matériau composite telle qu’elle sera
définie dans la division 4 du CODETI®.
ABSTRACT
Composite piping design has mainly been studied in the nuclear industry with experience gained on the corrosion
of piping due to river water. Stainless steel or neoprene coating could not resolve the chemical constraints.
Operation conditions range from -30°C to 120°C, 8 to 15 bars in pressure, 50 to 700mm in diameter. It
corresponds to the scope of E-glass reinforced epoxy composites.
The design of such networks has led to immediate problems of design criteria, problems in the use of code CASE
155-2. The quantification of safety factors and the consideration of anisotropy have always been particularly
emphasized.
This is the direct cause of ongoing studies at CETIM within a specific working group WG4 of the Pressure
equipment Commission, relating to the concerns expressed in the research and innovation networks (R2IT). «The
goal is to promote relevant research to remove the technological barriers for the co-development of products
and services based on new technologies.» A partial survey among 20 companies, particularly with an
Engineering activity and using steel and composite, has been carried out by CETIM in June 2001. It indicates
that the pipes are all or partly under significative service pressure. These companies believe that these materials
were underemployed because of their cost and lack of standardization. The market oscillates between stagnation
and growth. Potential applications are chemistry, water supply network, Off-shore, Energy and Environment.
Materials are qualified in-house at the material supplier’s premises. These companies require a design code
like CODETI ®.
This paper introduces the general design of composite pipes as it will be defined in the Division 4 of CODETI®.
INDEX çç Sommaire
INTRODUCTION
Le grand nombre de combinaisons résine/fibre/procédé possibles rend la normalisation
des matériaux composites difficile. La démarche consiste à utiliser plus systématiquement les
modèles d’homogénéisation éprouvés pour estimer à partir des caractéristiques des
constituants de base, des combinaisons et orientations de couches, les caractéristiques du
matériau composite tout en assurant une cohérence essais/modèles/contrôle.
Un verrou technique important actuellement pour les pièces chaudronnées en composite
est le surdimensionnement. L'utilisation de coefficients de sécurité élevés empêche toute
forme d'optimisation.
L’analyse mécanique des structures en composites, demandant la manipulation d’un
grand nombre de variables, exige un lien étroit entre la fabrication, le dimensionnement, les
essais de caractérisation, les contrôles, la fractographie. En parallèle des essais "in vivo", des
essais "in silico" sont un complément important. La mécanique des composites se traite par
une "multi-approche": multi-matériaux, multi-physique, multi-échelles.
ACIER/COMPOSITE
Pour les structures en acier, de façon courante, et certes de façon un peu simplifiée, l'analyse
mécanique au niveau de la conception ou bien lors d’analyse de défaillances s’effectue à
partir des données suivantes:
o une géométrie définie par des lignes, surfaces, volumes,
o un comportement matériau élastique linéaire, ou élasto-plastique, comportement
résumable par quelques données (masse volumique, module d’Young, limites en
résistance, dilatations thermiques...), soit un nombre de données de quelques unités
dont il existe des valeurs normalisées,
o des conditions aux limites,
o des chargements.
Pour le composite, l'analyse mécanique doit prendre en compte les éléments suivants
o la géométrie est la base pour la dépose des couches, elle se complexifie donc
immédiatement par la séquence d’empilement, par les orientations associées, par les
épaisseurs constitutives. Toutes choses égales par ailleurs, la donnée géométrique
"épaisseur acier" , est remplacée par une vingtaine de données épaisseurs, orientations.
Notons aussi que la dépose des couches est liée à la géométrie, la dépose d’un tissu sur
une surface courbe voit ce dernier se conformer à cette surface, modifiant ainsi
localement la donnée initiale,
o le comportement matériau est une donnée à entrer au niveau stratifié, comportement à
l’échelle du cm et au niveau des couches à cause de l’hétérogénéité intrinsèque qui
oblige à avoir cette donnée de comportement sur une échelle de l’ordre du mm.
L’anisotropie vient ajouter des paramètres directionnels au comportement et dans le
domaine linéaire, le nombre de données acier est ici multiplié par 3 environ. Il faut y
rajouter des problèmes spécifiques comme les propriétés transverses faibles, les
dilatations dues à l’humidité. Les matériaux composite ne font pas l'objet de normes
harmonisées, la caractérisation est donc un complément à prévoir à la conception. La
dispersion des propriétés est un facteur qui entraîne une caractérisation sur un nombre
d’éprouvettes de l’ordre de 3 fois plus,
o les conditions aux limites sont proches des données acier, le temps est une composante
ajoutant une viscosité ou dérive des propriétés dans le temps plus fréquemment
INDEX çç Sommaire
rencontrée pour les polymères utilisés plus près de leur température limite que les
aciers.
o dans le cas du composite, les chargements se combinent, l’analyse en résistance est
dépendante des types et chemins de chargement. Un comportement élastique isotrope
peut être lié à une résistance limite anisotrope. Les raisonnements d’analyse obligent
ainsi à revoir chaque cas de chargement de façon indépendante. Une marge par rapport
à la rupture peut être grande, mais un changement de chemin de chargement de
quelques degrés peut entraîner une marge de plusieurs ordres de grandeur inférieurs,
o en post-processeur d’analyse, il est difficile de résumer par un chiffre une structure
composite, la marge par rapport à la rupture remplacera les contraintes de Von Mises
et dépendra du chemin de chargement, la communication des résultats sous forme
graphique est lourde, les logiciels ne s’étant pas encore appropriés ces types
d’analyse...
Ce paragraphe rappelle les complexités des composites dues au grand nombre de variables et
de leurs associations possibles. Le niveau technique d’analyse mécanique reste pourtant du
même degré que pour celui de l’acier, la quantité d’informations est grande et elle demande
des efforts particuliers.
MATERIAU
Les résines utilisées pour la fabrication des tuyauteries composites sont des résines d'usage
général telles que les polyesters, vinylesters ou époxydes. Les données par défaut utilisées
dans les modèles sont listées ci-dessous, pour une température de 23°C. Elles sont facilement
accessibles en fourniture.
masse volumique 1200 kg/m3
module d'Young 3200 MPa
coefficient de Poisson 0.37
amortissement en traction 8%
amortissement en cisaillement 8.70%
limite en traction 74 MPa
limite en compression 126 MPa
limite en cisiaillement 43 MPa
limite en flexion 104 MPa
0.000068
dilatation thermique
m/m/°C
dilatation hygrométrique (en m/m/(kg
0.39
d'eau absorbée/kg de matière)
conductivité thermique 0.2 W/(m.°C)
capacité calorifique massique 1050 J/kg/°C
Les fibres le plus couramment utilisées pour les tuyauteries composites sont les fibres de verre
(E,C...). Les caractéristiques par défaut utilisées dans les modèles sont listées ci-dessous.
INDEX çç Sommaire
masse volumique 2540 kg/m3
module d'Young 73000 MPa
coefficient de Poisson 0.22
amortissement en traction 0.006
amortissement en cisaillement 0.09
limite en traction 1990 MPa
limite en compression 1200 MPa
limite en cisaillement 1200 MPa
dilatation thermique 5.3e-6 m/m/°C
dilatation hygrométrique (en m/m/(kg
0
d'eau absorbée/kg de matière)
conductivité thermique 1.1 W/(m.°C)
capacité calorifique massique 840 J/kg/°C
INDEX çç Sommaire
limite élastique sens 1 traction X 45 MPa
limite élastique sens 1 compression X' 58 MPa
limite élastique sens 2 traction Y 87 MPa
limite élastique sens 2 compression Y' 79 MPa
limite élastique sens 12 cisaillement S 94 MPa
direction longitudinale 1
direction radiale 3
direction cironférentielle 2
Le choix des matériaux s’appuie sur un ensemble de propriétés spécifiques aux matériaux
composites. Ces propriétés générées par calcul s’obtiennent couramment en fourniture et
fabrication. Elles permettent d’assurer les marges de sécurité requises. La garantie de ces
propriétés de base est nécessaire pour valider la réalisation d’une tuyauterie en matériaux
composites.
INDEX çç Sommaire
FABRICATION
Les tuyauteries composites sont principalement fabriquées par enroulement filamentaire.
Elles présentent soit des nappes superposées, soit des croisements de mèches soit des fils
continus (Roving) assimilés à un empilement de couches.
mandrin
fibres
résine
MODE DE RUPTURE
Sous l'effet des sollicitations auxquelles elle est soumise, pour un chemin de mise en
charge connu, la défaillance d'une tuyauterie composite peut survenir suivant l'un des modes
ci-après :
o micro-fissurations, perlage ou micro-fuites au travers de la paroi, sans dégradation
apparente pouvant résulter d’un dépassement des contraintes limites dans le plan
des couches ou d’un excès de cisaillement transverse local dû à une discontinuité,
o rupture de type fragile ou ruptures élastiques successives de type fragile par
morceaux, - instabilité élastique, locale de paroi, ou générale de structure,
o fluage ou fatigue statique,
o fatigue,
o diminution d'épaisseur travaillante par la reprise d’humidité,
o érosion, usure ,...,
o corrosion, vieillissement,
INDEX çç Sommaire
Pour la rupture d’un empilement de couche, il faut distinguer les premières fissurations
de couches, les premières ruptures de fibres et la rupture ou ruine finale. La rupture de la
matrice ou de l’interface fibre-matrice est la source de ces premières ruptures. La rupture de
première couche est la limite prise en compte quand on n’admet pas de perlage ou de micro-
fissurations si un liner est disposé. Le perlage surviendra en fait au delà de la rupture de
première couche dès qu’un chemin continu de microfissures se crée dans l’épaisseur. La
première rupture de couche se détecte principalement par émission acoustique. Le critère dit
de Tsai-Wu qui sert à calculer cette première rupture est conservatif.
Micro-fuites
Bulles en surface
INDEX çç Sommaire
COEFFICIENTS DE SECURITE
Un verrou technique important existant actuellement pour les pièces chaudronnées en
composite est le surdimensionnement. Dans les codes existants, des marges de sécurité sont
actuellement imposées jusqu’à une valeur de 6 en statique, voire 9 en durabilité. Ils sont issus
d’analyse mécanique calquée sur l’acier avec des propriétés isotropes, des critères de rupture
de type Von-Mises. Ils sont conservatifs en apparence mais ne correspondent pas à une
analyse prenant en compte les spécificités des composites. Ces valeurs empêchent toute
recherche d’optimisation et toute comparaison avec d’autres matériaux.
Les coefficients de sécurités utilisés doivent permettre de garantir le même niveau de
fiabilité que ceux utilisés pour les tuyauteries métalliques.
Le coefficient de sécurité à utiliser dans les calculs est lié à la durée de vie souhaitée.
En situation normale de service, l’objectif est de garantir un coefficient de sécurité minimal de
1.5 pendant toute la durée de vie de l’installation.
Alors, pour garantir un coefficient de 1.5 sur toute la durée de vie, il convient d’utiliser un
coefficient de sécurité initial intégrant ces phénomènes.
Lorsqu’aucune valeur de résistance à long terme n’est disponible, le coefficient de sécurité
est déterminé à partir de la courbe suivante.
D’après cette courbe, l’utilisation d’un coefficient de sécurité de 3 permet de garantir une
marge résiduelle de 1.5 au bout d’une durée de service supérieure à 20 ans.
INDEX çç Sommaire
Caractéristiques mécaniques Temps court
Situation Temps long
à utiliser dans les calculs : (voir Note 1)
Caractéristiques mécaniques 3
1.5
Situation initiales (Voir Note 2)
Normale Caractéristiques mécaniques
- 1.5
mesurés sur un temps long
Essai de
Caractéristiques mécaniques /
résistance 1.1
initiales (Voir Note 3)
initiale
Caractéristiques mécaniques
Exceptionnelle 1.5
initiales
ou 1.1
occasionnelle Caractéristiques mécaniques
1.1
évoluant dans le temps
Note 1 : Dans ce cas d’utilisation, le Fabricant doit obtenir du Producteur la garantie de durée de
vie du matériau par des essais de résistance à long terme (HDB).
Note 2 : Pour conserver un coefficient de 1.5 au court du temps il convient d’appliquer un
coefficient lu sur la courbe en fonction du temps d’application de la charge.
Note 3 : Pour l’essai de résistance, il n’est pas nécessaire de vérifier le dimensionnement en
conditions de fluage.
INDEX çç Sommaire
ANALYSE LIMITE - CRITERE DE RUINE
Le dimensionnement des tuyauteries composites s'effectue en découplant l'analyse en
contraintes planes de l'analyse en contraintes transverses (efforts tranchants) suivant deux
critères indépendants.
L'analyse en contraintes transverse est réalisée avec un critère de contrainte maximale (en
cisaillement et en contrainte normal). Elle est réalisée dans les zones de discontinuités
(assemblages, supportages,...) où des contraintes transverses peuvent se développer.
contraintes transverses à la
paroi du tube
L'analyse dans le plan de la paroi du tube est réalisée en utilisant les enveloppes de
premières ruptures du composite établies à partir du critère de Tsaï-Wu. Dans ce cas nous
nous intéressons aux contraintes de membrane 1, 2 et 12 se développant dans la paroi du
tube sous l'effet du chargement appliqué. L'admissibilité des contraintes revient à vérifier que
les points de fonctionnement en contraintes longitudinales et contraintes circonférentielles
sont situés à l'intérieur de l'enveloppe de perlage pour un niveau de torsion donné.
INDEX çç Sommaire
éclatement d'un tube sous un état de contraintes planes
Les enveloppes de perlage sont illustrées sur la figure suivante pour trois températures pour
un tube enroulé à +/-55° avec un taux massique de fibre de verre de 70%
23°C 100°C
-35°C
INDEX çç Sommaire
– Chemin de chargement pression/effet de fond - flexion acceptable (zone en
compression) : 1 = 0 (ou l = 0).
Cette méthode permet de garder une réserve en contrainte longitudinale pour une
éventuelle mise en flexion globale de la ligne, exemple supportage de la partie aérienne..
2
2
1
1
1
DEMARCHE DE DIMENSIONNEMENT
Analyse matériau
Les données matériaux nécessaires à l'analyse de la tuyauterie sont déterminées en
fonction du type de matériau choisi devant répondre aux contraintes de tenue mécanique, de
température, de compatibilité chimique...
L'enveloppe de perlage s'en déduit en fonction de l'angle d'enroulement, de la fraction
massique de fibre et de la température.
La température oriente le choix de la résine. La limite d'utilisation des résines est donnée
par leur température de transition vitreuse, température au delà de la quelle les propriétés
élastiques chutent. En situation normale, la température de service doit être inférieure de 20°C
à cette température de transition vitreuse. La reprise d'humidité fait chuter de 20 à 30°C la
température de transition vitreuse, par conséquent, en situation de reprise totale d’humidité la
température de service doit être inférieure de 40°C à la température de transition vitreuse de la
résine.
Le fluide véhiculé oriente aussi le choix de la résine et l'éventuelle couche de protection
chimique à ajouter.
dimensionnement analytique
Il s'agit de pré-dimensionner les épaisseurs de tuyauteries et les assemblages par rapport
au niveau de pression et à la température du fluide. Ces calculs permettent de justifier le choix
des tubes et des accessoires en fonction des conditions d'utilisation et donne une réserve par
défaut pour la flexion.
INDEX çç Sommaire
• Détermination du coefficient de sécurité
• Dimensionnement à la pression intérieure:
o Détermination de l'épaisseur minimale du tube droit en fonction du
niveau de pression et de la température. L'épaisseur de paroi minimale
ne doit pas être inférieure à l’épaisseur calculée à partir de l’une des
formules suivantes :
PDi PDm
e= e=
2 f ac P 2 f ac
Analyse du réseau
Le dimensionnement en pression avec effet de fond est un chemin particulier de
chargement pour les tubes à enroulement à 55°, tout écart par rapport à ce chemin par des
effets de flexion, des blocages axiaux, fait chuter les propriétés limites apparentes du tuyau.
Pour appréhender ces phénomènes, le réseau est analysé avec des outils numériques (éléments
finis par exemple). Ce type d'analyse permet de prendre en compte la géométrie du réseau, les
efforts de pression générés par les changements de directions, les dilatations thermiques, les
dilatations hygrométriques, le poids et la rigidité des terres, le séisme,...
Les efforts et moments sont calculés en chaque point du réseau. Ils permettent de
déterminer le niveau de contrainte local dans la tuyauterie ainsi que les efforts de réaction
dans les supports.
Les propriétés mécaniques des matériaux étant connues, le calcul du réseau est abordable
avec des modèles de type poutres. Les analyses de portions en coque ou en 3 D sont
inévitables au niveau des coudes ou tés en particulier lors d’avaries ou pour éviter des
surdimensionnements.
Dans le cas d’un réseau de tuyauteries considéré comme un ensemble de poutres reliées
entre elles, il est possible d’utiliser les programmes de calcul classiques qui, en général ne
prévoient pas le cas de poutres orthotropes ou poutres ayant une rigidité en torsion
indépendante de la rigidité axiale. Les données d’entrée sont limitées aux module d’Young et
coefficient de Poisson, des équivalences pour obtenir un comportement adéquat sont
possibles.
Dans le cas des tubes composites on va tout d'abord entrer :
o le module longitudinal réel E1.
INDEX çç Sommaire
o le diamètre extérieur et l'épaisseur qui permettent de calculer les propriétés de la
section du tube.
o Dans certains codes le module de cisaillement est calculé à partir du coefficient de
poisson. . Il convient donc d’utiliser un « faux » coefficient de Poisson permettant
G12
de retrouver le module de cisaillement réel du tube composite. = 1
2 E1
INDEX çç Sommaire
CONTROLE
Les méthodes de contrôles non destructifs adaptées aux tuyauteries composites sont :
• contrôles visuels : ils sont destinés à rechercher les défauts de surface (défauts de
fabrication, traces d'impacts, dégradations dues au vieillissement,...) et constituent une
exigence minimale de contrôle.
• contrôles par mesure de dureté Barcol : destinés à détecter les états de sous-
polymérisation des tubes, accessoires en matériaux composites.
• contrôle par mesure DSC: pour détecter des états de sous-polymérisation pour les
accessoires stratifiés en matériaux composites, les assemblages collés ou assemblages
frettés et de vérifier la conformité de la tuyauterie à son domaine d’utilisation en
température.
• contrôle par radiographie : pour détecter les manques de colle en contrôle
radiographique par rayons X (limité à 120kV), sur des collages de tuyauteries
composites, dans la mesure où la colle a été préparée en vue de ce contrôle,.
• contrôle par ultrasons : Les principaux défauts détectables par méthode ultrasonore
sont les surfaces non collées, les délaminages, les vides et les variations d'épaisseur.
CONCLUSION
La rédaction de codes de construction par le SNCT est une de ses principales missions. Elle
consiste à examiner l'état de l'art de la profession pour définir le type d’ouvrage à réaliser.
Dans le cadre des études réalisées pour la Commission Chaudronnerie et Tuyauterie
Industrielle, le Cetim en collaboration avec les industriels et le SNCT réalise des études pour
répondre aux besoins de la profession et de faire évoluer les codes de constructions.
Dans le domaine de la tuyauterie, se retrouvent :
- Le "CODETI® Division 1 - Code de construction des tuyauteries industrielles (version
française)
- Le "CODETI® Division 2 - Code de construction des canalisations de transport
(version française)
A ces deux divisions, le SNCT compte ajouter une division 4 applicable aux matériaux
composites.
Un soutien technique permanent existe. Le syndicat participe activement aux instances
techniques de la Profession. Le SNCT et le Cetim participent aux instances élaborant les
réglementations et aux groupes de travail développant les normes européennes et
internationales. Les experts du SNCT et du Cetim dispensent des formations aux codes de
construction.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page105
RÉSUMÉ / ABSTRACT
70G
INVESTIGATION PAR LA MÉTHODE DES ÉLÉMENTS FINIS SUR DES
APPAREILS À PRESSION COMPOSITES RENFORCÉS PAR DES FIBRES
SOUS CHARGEMENT DE CHOC
ABSTRACT
The effect of impact loading on FRP composite vessel with high internal pressure and has been inves-
tigated by FEM simulation in the present study. Composite vessel which has an Aluminum liner and
wound by carbon fiber has been simulated by using 3-D model. MARC-Mentat commercial code has
been selected as a computational tool. The geometry of the vessel has been generated using Mentat
pre-post interface software. The material properties of the Aluminum A-6061-O liner of and 12K carbon
fiber + epoxy composite material layers have been included in the simulations as an isotropic and or-
thographic, respectively. An impactor with a diameter of 10 mm has been employed for dynamic loading
of the vessels. The maximum displacement of the composite part which was just under impactor has
been divided into several steps during the simulations to check the progress of the impactor damage.
The simulation results with respect to the plastic deformation of Al liner + carbon epoxy composite
was compared to that of obtained from experiment. The details of the damage caused by impactor have
been analyzed using simulation results. The results of the simulations showed that some zones under
the impactor damage would face compression deformations rather that tensional deformation, which
is generally believed.
INDEX çç Sommaire
FEM INVESTIGATION OF THE FIBER REINFORCED COMPOSITE
PRESSURE TUBE UNDER IMPACT LOADING
ABSTRACT
The effect of impact loading on Carbon Fiber Reinforced Plastic (CFRP) composite tube without
and with high internal pressure has been investigated by FEM simulation in the present study.
Composite tube which has an Aluminum cylinder and wound by carbon fiber was simulated by
using 3-D model. MARC-Mentat commercial code has been selected as a computational tool. The
geometry of the cylinder has been generated using Mentat pre-post interface software. The material
properties of the Aluminum 6061T liner of and TORAY T-700SC + Epoxy composite layers have
been included in the simulations as an isotropic and orthographic, respectively. The dome region of
the vessel has not been included in the simulations. The maximum displacement of the composite
part which was just under impactor has been divided into several steps during the simulations to
check the progress of the impactor damage. The model of the simulation has been created by
considering the real experiment which has been conducted by other researcher. The results of the
simulations show that some zones under the impactor damage would face compressions stresses
rather that tensional deformation, which is generally believed.
1. Introduction
Resistance to high velocity impact is an important requirement for high performance structural
materials. Even though, polymer matrix composites are characterized by high specific stiffness and
high specific strength, they are susceptible to impact loading. For the effective use of such materials
in structural applications, their behavior under high velocity impact should be clearly understood.
For this reason this topic is the interest of researchers.
Naik et al. [1] presented investigation on the ballistic impact behavior of two-dimensional woven
fabric. Tan et al. [2] presented an FE model of woven fabric that reflects the orthotropic properties
of the fabric, the viscoelastic nature of the yarns, the crimping of the yarns, the sliding contact
between yarns and yarn breakage using an assembly of viscoelastic bar elements. Duan et al. [3]
presented a finite element analysis to study the influence of friction during ballistic impact of a rigid
sphere onto a square fabric panel that was firmly clamped along its four edges and projectile-fabric
friction and yarn–yarn friction were investigated. Fawaz et al. [4] presented an effective
methodology for the optimum design of two-component armours using Florences model and a new
hybrid solving technique. Übeyli et al. [5] presented that the ballistic performance of steel against
7.62 mm armor piercing projectiles. Lopresto et al. [6] investigated that low-velocity impact tests
were carried out on stitched carbon fiber-reinforced plastic laminates of various thicknesses, whose
behavior was studied with reference to the overall force–displacement curve, first failure load,
penetration, indentation and damage extent. Duan et al. [7] studied a 3-D finite element analysis
model using LS-DYNA to simulate the transverse impact of a rigid right circular cylinder onto a
square patch of plain-woven Kevlar fabric. Mamalis et al. [8] studied that in the simulation works
the LS-DYNA 3-D explicit finite element code is used to investigate the compressive properties and
INDEX çç Sommaire
crushing response of square carbon fiber reinforced plastic tubes subjected to static axial
compression and impact testing. Abdullah et al. [9] investigated the high velocity impact response
of a range of polypropylene-based fiber–metal laminate (FML) structures. Vaidya et al. [10] studied
that the vibration response of composite sandwich plates composed of laminate face sheets and
aluminum foam core was also studied under a free–free boundary condition and the vibration
response (natural frequency and damping ratio) is reported as a function of impact to the sandwich
plate. Caprino et al. [11] studied that four stitched graphite/epoxy laminates of different thicknesses
were subjected to high-velocity impact tests. Two steel spheres, 12.7 and 20 mm in diameter, were
used as bullets during the tests, carried out at two different speeds (65 and 129 m/s).
As seen above, most of the studies concentrated on low speed ballistic impact and quasi-static
loading both of which is far from real situations and damages occurs under high velocity impact
loading of cylinders. Some studies which are conducted under high speed used laminates instead of
cylinders. A FE model has been proposed in the present study in which the speed of impactor can
vary from low speed to high speed loading in Al + CFRP composite tube. The FE model has been
compared to experiment in terms of Al and composite deformation which is happened after
impactor made an indentation on tube. This was to understand the reliability of FE model.
The nodes at the one end of the tube were fixed along X axis and the nodes located at the other end
of the tube were left free along X axis deformations. All the nodes which are located at both end of
tube were fixed in Y and Z direction to make the boundary conditions are similar to that of
experiments. The boundary conditions and the rigid ball which is simulating the impactor are shown
in Fig. 2 (a). The deformed shape of tube after impact loading in simulation is also given in Fig. 2
(b).
INDEX çç Sommaire
Impactor
(a)
CFRP tube (b)
Fig. 2 (a) The boundary conditions and (b) the deformed shape in FEM simulation
It is very difficult to obtain different data during impact testing because of high speed of impactor
and short testing time. High speed testing also needs more sophisticated and expensive equipment
to obtain experimental data. Thus it is very useful to make simulation to see the stress and strain
variation of composite tube once the reliability of the FE model is understood.
0
0 5 10 15
Displacement (mm)
5 mm
-1
-2
-3
-4
-5
-6
Fig. 3 Comparison of experiments with simulation in term of indentation depth. (a) The deformed
shape after impactor hit composite tube [14] (b) The displacement vs step number in simulation.
INDEX çç Sommaire
3
3. Results of FEM Simulation
The effect of the surface damage on the strength of a filament wound an (FW) gas cylinder which
subsequently causes loss of stiffness and strength resulting from accidental impact in fiber
composite structures remains of much concern. Although a number of previous investigators have
delineated the major features of the problem, important details remain uncertain. Although a
number of studies have been carried out on impact of flat composite plates, relatively little work has
been done on tubular geometries such as pressure vessels despite their usage in applications.
FEA provides a good approach to the analysis of composite cylinders under high internal pressure
to understand how the stress distribution of cylinder changes. Fig. 4 shows displacement along tube
longitudinal axis after indentation. All the displacements along longitudinal axis of tube with no
internal pressure (continues-solid line) are always under minus direction. This means all the
cylinder body is more affected when it is empty. But the ends of the cylinder are in plus area when
cylinder has 50 MPa internal pressure. Comparing the length of the curves along longitudinal
direction indicates that the cylinder with 50 MPa internal pressure has more longitudinal
deformation. This is important how to approach and propose analytical solutions when composite
cylinder under impact loading is considered. As it is well known the compression strength is
considerable different than tensional strength in terms of fiber reinforced composites. Comparison
of curves of CFRP and Al reveals that Aluminum tube has the same behavior like CFRP. This
indicates that aluminum behaves like same as CFRP since the displacement are similar to each
other. But, it should be considered that Al mechanical properties are quite different than CFRP, thus
attentions should be paid when Al liners of composite high pressure vessels are subjected to fatigue
life calculations.
Von Mises stress distribution of Al cylinder in terms of no internal pressure and with 50 MPa
internal pressure is given in Fig. 5. The stress difference between tube end and the middle of tube
where impactor deformations occurred is higher in low internal pressure and low in higher internal
pressure. But the end of the tube has almost the same stress as impactor affected area. This should
be carefully considered while the tube is designed.
Generally, designing the cylinders under internal pressure needs only tensional mechanical
properties. Fig. 6 shows that compression and tension occurs at the same time during impact
loading of cylinder. Considering that tensional and compression mechanical properties are quite
different in orthographic materials This figure also reveals that the deformation immediately under
the impactor is bending + tension in the longitudinal direction but become compression in the
circumferential direction. Further away from the impactor towards to the end of cylinder, tensional
stress become active in longitudinal and circumferential direction.
Stress (MPa)
80
60
40
20
0
2 25 47 70 93 115 138 161 183
Tube Length (mm)
FRP
500
Axial Stress (MPa)
0
0 23 45 68 90 113 136 158 181 203
-500
-1000
-1500
Length (m m )
FRP
3000
Circumferential Stress
2000
1000
(MPa)
0
-1000 0 23 45 68 90 113 136 158 181 203
-2000
-3000
Length (mm)
Circumferential stress distribution of CFRP composite material with 50 MPa internal pressure and
without internal pressure is given in Fig. 7. This figure shows that compression stress occurs where
the cylinder body was hit by impactor. It has been explained in reference [13] that some fiber
breakage would occur depending on the speed of the impactor. This may cause stress redistribution
of tube drastically and may result severe plastic deformation in Al tube central part. The effect of
autofrettaged to make the Al tube fatigue life longer may become ineffective too.
INDEX çç Sommaire
4. Concluding remarks
A FE model to simulate the effect of impact loading on Al + CFRP composite cylinder with high
internal pressure has been investigated in the present study. The reliability of the simulation model
has been controlled by comparing the real experiment which has been conducted by other
researcher the results of simulations revealed that some of the cylinder parts which is generally
believed and designed to work under tensional stresses may under go compression stress under
impact loading. Further studies will be conducted to understand the details of Al + CFRP composite
cylinder under impact loading especially to predict the fatigue life of damaged high pressure
vessels.
5. References
[1] Naik, N.K, Shrirao, P., Reddy, B.C.K., “Ballistic impact behavior of woven fabric
composites: Formulation”, International Journal of Impact Engineering 32 (2006) 1521–1552
[2] Tan, V.B.C., Ching, T.W., “Computational simulation of fabric armour subjected to ballistic
impacts”, International Journal of Impact Engineering 32 (2006) 1737–1751
[3] Duan, Y., Keefe, M., Bogetti, T.A., Cheeseman, B.A.,Powers, B., “A numerical investigation
of the influence of friction on energy absorption by a high-strength fabric subjected to
ballistic impact”, International Journal of Impact Engineering 32 (2006) 1299–1312
[4] Fawaz, Z., Behdinan, K., Xu, Y., “Optimum design of two-component composite armours
against high-speed impact”, Composite Structures 73 (2006) 253–262
[5] Übeyli, M., Yıldırım, R.O., Ögel, B., “On the comparison of the ballistic performance of steel
and laminated composite armors”, Materials and Design 28 (2007) 1257–1262
[6] Lopresto, V., Melito, V., Leone, C., Caprino, G., “Effect of stitches on the impact behaviour
of graphite/epoxy composites”, Composites Science and Technology 66 (2006) 206–214
[7] Duan, Y., Keefe, M., Bogetti, T.A., Powers, B., “Finite element modeling of transverse
impact on a ballistic fabric”, International Journal of Mechanical Sciences 48 (2006) 33–43
[8] Mamalis , A.G., Manolakos, D.E., Ioannidis, M.B., Papapostolou, D.P., “The static and
dynamic axial collapse of CFRP square tubes: Finite element modelling”, Composite
Structures 74 (2006) 213–225
[9] Abdullah, M.R., Cantwell, W.J., “The impact resistance of polypropylene-based fibre–metal
laminates”, Composites Science and Technology 66 (2006) 1682–1693
[10] Vaidya, U.K., Pillay, S., Bartus, S., Ulven, C.A., Grow, D.T., Mathew, B., “Impact and post-
impact vibration response of protective metal foam composite sandwich plates”, Materials
Science and Engineering A 428 (2006) 59–66
[11] Caprino, G., Lopresto, V., Santoro, D., “Ballistic impact behaviour of stitched graphite/epoxy
laminates”, Composites Science and Technology 67 (2007) 325–335
[12] MARC User’s Guide, Volume A – Theory and User Information, MARC Analysis Research
Corporation, 2007
[13] MARC User’s Guide, Volume B-Element Library, MARC Analysis Research Corporation,
2007, 33- 43
[14] Takekusa T., Kobayashi S., Akkus N., Wakayama S., Takehana T., “Evaluation of Burst
Strength of FW-FRP Composite Vessel after Impact” Nippon Kikai Gakkai Kikai Zairyo,
Zairyo Kako Gijutsu Koenkai Koen Ronbunshu Vol. 8th; No; 75-76(2000)
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page106
RÉSUMÉ / ABSTRACT
52G
COMPORTEMENT SOUS PRESSION HYDROSTATIQUE EXTERNE DE
TUBES HYBRIDES ACIER-COMPOSITE
RÉSUMÉ
Afin de promouvoir l’utilisation de matériaux composites pour les applications grande profondeur de
Mercredi
l’industrie offshore, des tubes hybrides acier-composite ont été fabriqués et testés pour vérifier la tenue
de telles structures sous chargement externe de pression hydrostatique. L’idée de base est d’utiliser le
renfort en matériau composite pour accroître le coefficient de sécurité de manière à atteindre une limite
de sécurité satisfaisante tout en réduisant significativement le poids de la structure.
L’intérêt de telles structures sous pression externe n’a pas été clairement démontré et un mode de ruine
inattendu a été constaté. Une analyse élément fini, a posteriori, a permis d’identifier le mode de ruine
en flambement observé, toutefois la pression de rupture n’a pu être prédite de manière satisfaisante.
En parallèle à cette étude, l’utilisation de capteurs fibre optique à Réseau de Bragg noyés dans la
structure a été investiguée pour le suivi du comportement en service.
ABSTRACT
In order to promote the use of composite materials for deep offshore applications, steel-composite tubes
have been manufactured and tested to examine the behavior of such structures under hydrostatic
pressure. The objective was to use a composite material layer to increase the safety factor of a metallic
tube in order to reach a satisfactory safety limit and to significantly reduce the weight of the ensemble.
The interest of such structures under hydrostatic compression has not been clearly demonstrated, and
an unexpected failure mode has been observed. Post collapse FE modeling has predicted the buckling
mode of the structure but at this time the collapse pressure cannot be calculated accurately. In parallel
with this study, the use of Fiber Bragg grating sensors to monitor deep underwater structures has been
investigated. The possibility to use embedded sensors has been verified, and a modified interface has
been developed.
INDEX çç Sommaire
INTRODUCTION
STRUCTURES D’ESSAI
INDEX çç Sommaire
Le chargement hydrostatique (avec effet de fond) générant des contraintes circonférentielles
doubles des contraintes axiales, il est proposé de renforcer circonférentiellement l’acier par de
l’enroulement filamentaire de fibres imprégnées de résine époxyde (procédé humide) dans la
partie centrale de la structure (structure type II). L’épaisseur de composite appliqué sur le tube
acier a arbitrairement été choisie égale à l’épaisseur initiale de la partie centrale de la structure
en acier.
Pour cette étude, l’utilisation de la fibre de carbone a été privilégiée. Il doit cependant être
mentionné qu’une démarche de dimensionnement optimisée peut conduire à un choix
différent en termes d’épaisseur de composite et de type de fibre.
Le renfort en composite a été réalisé par enroulement filamentaire au laboratoire LGMT de
l’Université de Toulouse (Figure 2).
Dans un but de comparaison, une structure en acier de même épaisseur que la structure
hybride a également été testée (Type III).
Type I II III
Diamètre interne (mm) 124 124 124
Diamètre externe au centre (mm) 134 150 150
Epaisseur d’acier (mm) 5 5 13
Longueur totale (mm) 520 520 520
Epaisseur de composite (mm) 0 8 0
Poids (kg) 12.5 14.2 23.0
Nombre d’essais 1 5 1
Préalablement aux essais, une analyse aux éléments finis des différentes structures a été
réalisée. Les paramètres du modèle sont reportés table 2.
Software Adina®
Type d’éléments 3D volumique
20 nœuds (paraboliques) pour les zones sans contact
4 nœuds pour les zones de contact (si nécessaire)
Chargement Pression externe avec effet de fond
Conditions limites Fermetures infiniment rigides
Les prédictions conduisent à une ruine de la structure par flambement avec plastification de
l’acier. Les résultats sont reportés table 3.
Flambement mode 3.
Type I Pression de collapse
5 mm d’acier calculée : 400 bar.
Flambement puis
plastification de l’acier.
Flambement mode 2.
Type II Pression de collapse
5 mm d’acier calculée : 610 bar.
8 mm de composite Plastification de l’acier
puis flambement.
Flambement mode 2.
Pression de collapse
Type III calculée : 1000 bar.
13 mm d’acier Plastification de l’acier
puis flambement.
FBG
30
m
m
optical fiber .
INDEX çç Sommaire
Figure 4 : Spectre de réflexion du capteur
CONDITIONS D’ESSAIS
Les essais ont été réalisés dans un caisson hyperbare de capacité 1000 bar (figure 5). Sept
tubes ont été testés : 1 de type I, 1 de type III et 5 de type II. Les structures d’essais sont
instrumentées par 12 jauges résistives collées en surface. Le multi-capteur à fibre gravée a été
implanté au niveau de la section centrale sur 3 éprouvettes type II (tubes hybrides). Une
connexion sous-marine spécifique a été développée afin d’assurer le bon comportement de la
fibre optique vis-à-vis de la pression hydrostatique.
La vitesse de chargement retenue pour ces tubes a été fixée à 10 bar/min. Pour certains essais,
des cycles de charges-décharges ont été appliqués par pas de 50 bar pour investiguer
l’éventuelle apparition d’endommagement.
INDEX çç Sommaire
COMPORTEMENTS DES TUBES
Mode 3 Mode 3
Type I
5 mm d’acier 400 bar 419 bar
Mode 2 “Mode 1”
Type II
5 mm d’acier 610 bar 431- 517 -583 - 410
8 mm de composite – 414 bar
Mode 2 Mode 2
Type III
13 mm d’acier 1000 bar 1040 bar
Il est notable que pour les structures type I et type III (tubes en acier), les résultats de
modélisation sont en accord de manière satisfaisante avec les résultats d’essais. Le modèle
décrit correctement le mode de flambement et la pression de collapse. Le mode de
flambement est initié dès le début du chargement et peut être détecté par les jauges de
déformation.
Pour les essais sur les tubes hydrides, les enseignements suivants peuvent être tirés :
1 - Tous les collapses sont obtenus avec le même mode de flambement, mode à un lobe. Ce
mode n’a a priori jamais été relevé sur des structures métalliques ainsi que sur des tubes
tout composite lors d’essais sous pression externe. Il semble toutefois que des tubes
flexibles présentent le même type de mode de flambement.
2 - Une très large variation de résultats est notée sur les pressions de collapse (410 à 583 bar)
bien que les tubes aient été fabriqués de la même façon. Aucune explication n’a à ce jour
été trouvée pour justifier cette dispersion sur les pressions de collapse (préparation de
surface, condition de mise en œuvre et de cuisson…).
3 - Un délaminage du renforcement composite sur le tube acier a été identifié vers 300 bar sur
un cas (pression de collapse 431 bar). Pour les 4 autres essais, les jauges de déformation
n’ont pas permis d’identifier un éventuel transfert de charge du renfort composite sur la
structure acier avant le collapse. La structure délaminée atteint une pression de collapse
proche de la pression relevée sur les tubes type I.
INDEX çç Sommaire
Une modification des conditions aux limites du modèle éléments finis a permis de simuler le
mode de flambement (mode à un lobe) relevé lors des essais. Les conditions limites mises en
œuvre sont alors les suivantes : délaminage initial de la partie composite et pression
hydrostatique appliquée sur la demi-surface délaminée (figure 6).
Un des buts de cette étude était d’évaluer le potentiel d’utilisation des capteurs à fibre optique
noyés dans le composite pour suivre le comportement de structures soumises à pression
hydrostatique. Les résultats obtenus par le multi-capteur à fibre gravée implanté dans le tube
ayant collapsé à 517 bar sont reportés figure 7. Ces résultats montrent qu’il est possible de
suivre les déformations avec un tel capteur jusqu’à des pressions supérieures à 500 bar. Les
résultats relevés sont en accord avec ceux obtenus avec les jauges de déformation résistives.
Cependant, des difficultés et des pertes de précision ont été observées à partir de 200 bar.
0
strain (µdef)
-500
-1000
-1500
-2000
-2500
0 50 100 150 200 250 300 350 400 450 500
pressure (bar)
INDEX çç Sommaire
L’analyse du suivi de l’amplitude du signal réfléchi révèle, avec l’accroissement de la
pression, une importante perte (-10 dB) sur les 5 réseaux (figure 8). Cela rend délicat la
détermination de la fréquence du pic et explique la perte de précision relevée sur l’analyse des
déformations à partir de 200 bar. De manière surprenante, un accroissement de l’amplitude du
signal est noté pour une pression supérieure à 450 bar. A ce stade, cette augmentation n’est
pas clairement expliquée mais pourrait être attribuée à une modification locale des contraintes
autour de la fibre induite par le chargement hydrostatique. L’amplitude du signal issu de la
fibre optique est fortement affectée par la présence de micro-courbures qui peuvent être
induites lors de la fabrication et qui pourraient être amplifiées par l’application de la pression
externe.
Ceci peut conduire à une limitation d’utilisation de capteurs à fibre optique gravée noyés dans
une structure composite pour des pressions supérieures à 200 bar sans mise en œuvre de
précaution particulière supplémentaire.
-46
amplitude
-48
-50
-52
-54
-56
-58
0 50 100 150 200 250 300 350 400 450 500
pressure(bar)
CONCLUSIONS
Une étude sur le comportement de tube hybride acier-composite pour application sous-marine
a été conduite. Cette étude révèle, à ce stade, des difficultés quant à la prédiction du
comportement de ce type de structure sous chargement de pression hydrostatique. Le bénéfice
d’un renforcement d’une structure acier par une couche supplémentaire de matériau
composite n’a pas été démontré. Un gain potentiel de performance supérieure à 50 % en terme
de pression de collapse peut être envisagé mais une grande dispersion de résultats a été
relevée sans que l’on puisse fournir une justification à cette dispersion. Un mode de
flambement inattendu a été relevé et il semble que la qualité de l’interface composite/acier
conditionne fortement le comportement de structure hybride sous ce cas de chargement.
Les possibilités d’utilisation de capteurs à fibre optique à réseau de Bragg gravé, directement
implantés dans la structure composite lors de la fabrication, ont été investiguées pour suivre le
comportement d’une structure sous-marine. L’intérêt de l’utilisation a été vérifié jusqu’à une
pression de 200 bar. Pour des pressions supérieures, des précautions particulières
d’implantation doivent être étudiées pour garantir la fiabilité d’un tel système de mesure.
INDEX çç Sommaire
BIBLIOGRAPHIE
[1] Chauchot, P., et al, Composite Pressure Vessels: Towards a Composite Subsea
Separator Proc. CMOO4, 2005
[2] Bigourdan, B., et al., “Composite materials for subsea oil separation”, Proc. DOT,
Marseille, 2004
[3] Collombet, F., et al, Entire Life Time Monitoring of Filament Wound Composite
Cylinders Using Bragg Grating Sensors (parts I, II and III) Applied Composite
Materials, 2009
[4] Martin, R., (Editor), “Ageing of composites”, Woodhead publishing, 2008
[5] Ferdinand, P., "Capteurs à fibres optiques à réseaux de Bragg ", Techniques de
l’Ingénieur, traité Mesures et Contrôle, R 6 735-1, 1999
[6] Mulle, M., et al, "Instrumented Technological Specimen (ITS): Curing process and
mechanical characterization using embedded fibre Bragg gratings", Proc. ECCM 11,
2004
[7] Botsis, J., et al, "Embedded fiber bragg grating sensors for internal strain measurements
in polymeric materials", Optics an lasers in Engineering, 43, 491-510, 2005
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page107
RÉSUMÉ / ABSTRACT
6G
RÉALISATION D’APPAREILS À PRESSION EN COMPOSITES
THERMOPLASTIQUES PAR ENROULEMENT FILAMENTAIRE
D. Mastain
(Cetim – TECHNOCAMPUS EMC2 – Z.I du Chaffault - 44340 Bouguenais
E-mail : [email protected])
RÉSUMÉ
Les matériaux composites connaissent un essor important dans de nombreux secteurs industriels.
Fortement utilisés dans l’aéronautique et le nautisme depuis plusieurs années, ils pénètrent d’autres
secteurs d’activité comme celui des appareils à pression.
Sous l’effet de la pression réglementaire environnementale, les matériaux composites thermoplastiques,
pour réaliser de la pièce structurelle, gagnent du terrain sur les composites thermodurcissables
traditionnellement utilisés.
Le procédé de mise en œuvre par enroulement filamentaire est particulièrement adapté aux structures
telles qu’appareils à pression
Au sein du Technocampus EMC2, le CETIM a développé cette technologie pour transformer les
matériaux composites thermoplastiques. Le principe de transformation de ces matériaux étant différent
de celui des composites thermodurcissables, il a fallu intégrer et maîtriser un système de chauffage
permettant de fondre la matrice thermoplastique puis de la reconsolider instantanément au contact du
mandrin ou du liner.
Nous présenterons le développement réalisé sur le procédé et les résultats obtenus sur les matériaux
transformés pour des applications appareil à pression. Elle montrera que cette technologie appliquée
aux composites thermoplastiques permet de gagner en temps de cycle de fabrication et d’intégrer la
notion de recyclage, tout en respectant les exigences de qualité et de tenue mécanique des structures
fabriquées.
ABSTRACT
The use of composite materials is growing in various industrial sectors. Mainly used in the aeronautic
and the boat industry for several years, they are getting in other sectors such as gas pressure vessels.
Because of environmental regulations, thermoplastic composites materials dedicated to manufacture
structural parts are taking ground on traditional thermoset composites.
Filament winding process is particularly adapted to components such as pressure vessels.
At Technocampus EMC², CETIM developed a technology to process thermoplastic composites. The
principle of processing these materials is different of the one of thermosets. It has been necessary to
adapt and control a heating arrangement to melt the thermoplastic resin and then to solidify it when it
in contact with the mandrel.
We will present the development carried out on the process and the results obtained with the processed
materials for pressure vessels applications. It will demonstrate that that technology applied to
thermoplastic composites saves manufacturing cycle times and can be recycled but can nevertheless
comply with the quality and structural requirements.
AFIAP - ESOPE 2010 Programme
INDEX çç Sommaire
REALISATION D’APPAREIL A PRESSION EN COMPOSITES
THERMOPLASTIQUES PAR ENROULEMENT FILAMENTAIRE
Auteur : D.MASTAIN
(Cetim – TECHNOCAMPUS EMC2 – Z.I du Chaffault - 44340 Bouguenais
E-mail : [email protected])
RESUME
Nous présenterons le développement réalisé sur le procédé et les résultats obtenus sur les
matériaux transformés pour des applications appareil à pression. Elle montrera que cette
technologie appliquée aux composites thermoplastiques permet de gagner en temps de cycle
de fabrication et d’intégrer la notion de recyclage, tout en respectant les exigences de qualité
et de tenue mécanique des structures fabriquées.
ABSTRACT
The use of composite materials is growing in various industrial sectors. Mainly used in the
aeronautic and the boat industry for several years, they are getting in other sectors such as
gas pressure vessels.
Because of environmental regulations, thermoplastic composites materials dedicated to
manufacture structural parts are taking ground on traditional thermoset composites.
Filament winding process is particularly adapted to components such as pressure vessels.
INDEX çç Sommaire
At Technocampus EMC, CETIM developed a technology to process thermoplastic
composites. The principle of processing these materials is different of the one of thermosets. It
has been necessary to adapt and control a heating arrangement to melt the thermoplastic
resin and then to solidify it when it in contact with the mandrel.
We will present the development carried out on the process and the results obtained with the
processed materials for pressure vessels applications. It will demonstrate that that technology
applied to thermoplastic composites saves manufacturing cycle times and can be recycled but
can nevertheless comply with the quality and structural requirements.
INTRODUCTION
INDEX çç Sommaire
coquilles
Four chauffant IR vue entrée de roving
Tube Vortex
(refroidisseur
du four)
Galets montés sur
cartouche chauffante
Souffleur
Entrée des roving dans four IR d’air
chaud
Sortie des roving Tête
d’enroulement
Galets montés sur
cartouche chauffante
Coffret électrique
INDEX çç Sommaire
MONTAGE EXPERIMENTAL
Pendant nos essais nous avons enregistré le temps d’enroulement et la vitesse de défilement
en fonction des températures en entrée et en sortie du four.
Aux extrémités du four, nous avons placé 2 thermocouples en entrée et en sortie afin de noter
les températures de l’air à l’intérieur du four.
Thermocouple T1 (entrée
roving)
INDEX çç Sommaire
RESULTATS
Afin de tester mécaniquement les tubes réalisés, nous enroulons les fibres de façons radiales.
Nous effectuons plusieurs passages dans le but de vérifier la cohésion entre les différentes
couches déposées. Les essais présentés ont été réalisé avec de la fibre Twintex.
Zone de calibrage
1ère couche d’enroulement Polypropylène en fusion
Nous réalisons ensuite des essais de traction type NOL et des coupes microscopique afin de
vérifier l’absence de porosités.
Après plusieurs essais d’enroulement, nous obtenons des valeurs de traction proche des
450Mpa avec moins de 1% de porosité. Cela confirme nos paramètres de transformation.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page108
RÉSUMÉ / ABSTRACT
73G
RÉSERVOIRS DE STOCKAGE EN MATÉRIAUX PLASTIQUES : CAS
VÉCUS DE RUPTURE BRUTALE – RETOUR D’EXPERIENCE DE
CONTRÔLES PAR ÉMISSION ACOUSTIQUE
RÉSUMÉ
Nous avons eu, ces dernières années, deux effondrements de réservoirs de stockage en matériaux plas-
tiques (liner thermoplastique renforcé par un stratifié verre résine). Nous décrivons les deux incidents
Mercredi
ainsi que les expertises menées pour comprendre l’origine des problèmes. Par ailleurs, un plan d’action
complet a été lancé (enquête dans le groupe, analyse des risques site par site, révision des standards in-
ternes). L’un des points clés de la fiabilité de fonctionnement de ces équipements est l’existence et la
pertinence de contrôles non destructifs adaptés. L’émission acoustique est probablement le contrôle le
plus complet et le plus fiable pour estimer l’intégrité mécanique des équipements et nous donnons un
retour d’expérience sur le contrôle d’un certain nombre de réservoirs.
ABSTRACT
In the past years, two storage tanks in plastic materials fall down, full of products. They were made in
GRP (Glass Reinforced Plastic i.e. thermoplastic liner reinforced by thermoset resin and glass fibers).
We describe the failure mechanism which and the materials investigations made for understanding the
problems.
An action plan was launched in the company (inventory of plastic equipments in the group, risk analysis
plant by plant, revision of the internal technical standards).
One of the key points regarding reliability of plastic equipment is the existence and pertinence of non
destructive techniques adapted to plastics. Acoustic Emission is certainly, from our point of view, the
most complete and reliable technique for mechanical integrity estimation. We give experiences of
Acoustic Emission control concerning plastic equipment.
INDEX çç Sommaire
I. INTRODUCTION
Les stockages et autres récipients contenant des fluides très corrosifs (acides tels HCl,
H2SO4, ou alcalins comme la Javel, …) sont, pour beaucoup, depuis quelque 25/30 ans,
réalisés en matériaux plastiques car ils sont les mieux adaptés. Nous avons très peu
d’équipement réalisés en matériaux thermoplastiques massifs (PE, PP, PVC …). La plupart
du temps, il y a une couche de résistance mécanique en SVR (Stratifié Verre Résine). La
résistance à la corrosion est alors assurée soit par un liner interne en thermoplastique, soit par
une couche de résine (on parle alors de SVR massif).
Le retour d’expérience pour ce type d’équipements est encore limité (quelques dizaines
d'années) et est sans commune mesure avec celui des équipements en matériaux métalliques.
Ceci vient en particulier de la spécificité des matières plastiques qui ont des comportements
différents des métalliques. De plus, la connaissance intrinsèque de ces matériaux est
nettement moins avancée que celle des matériaux métalliques, notamment en ce qui concerne
le vieillissement dans le temps, les lois de comportement mécanique ou les phénomènes de
fatigue.
Enfin, la construction de ces équipements est encore largement manuelle, ce qui crée une
importante dispersion des caractéristiques mécaniques.
Cette situation explique un certain nombre d’incidents survenus sur des équipements en
matériaux plastiques.
Cette présentation montre dans un premier temps comment, suite à des incidents notables, il a
été mis en place au sein du groupe Arkema un plan d’actions afin d’améliorer les pratiques
sur ces équipements.
Il sera détaillé dans la suite de l’exposé les principes du contrôle par émission acoustique sur
les équipements en SVR et le retour d’expérience sur les contrôles réalisés.
En mars 2007, un bac de stockage d’acide chlorhydrique s’est rompu brutalement, libérant
l’acide dans la cuvette de rétention. Il n’y a pas eu de blessés, ni de pollution significative.
L’annexe 1 donne une vue d’ensemble du bac après la rupture (figure 1) ainsi que quelques
éléments de l’expertise.
Le trou d’homme a été projeté brutalement sur le côté du bac qui s’est donc vidé très
rapidement du fait du diamètre de l’ouverture. Lors d’une vidange si rapide, le trou d’évent
n’est pas assez gros pour compenser la pression et le bac passe au vide. Les équipements SVR
de ce type n’ont quasiment pas de résistance au vide et la virole du bac se rompt brutalement
INDEX çç Sommaire
au niveau du changement d’épaisseur. Elle n’est pas tombée car elle était fixée à une
passerelle au niveau du toit.
L’expertise s’est portée principalement sur toute la zone du trou d’homme. La cassure
observée est particulière et se situe à la liaison virole / tubulure de trou d’homme sur une
demi-circonférence et à la liaison bride / tubulure sur la demi circonférence opposée (figure 2
annexe 1).
Les observations visuelles, par coupe macrographique ou au microscope électronique à
balayage (MEB) ont permis de constater notamment:
• des faciès de rupture sur le liner caractéristiques d’un phénomène de fissuration
progressive par fatigue,
• une transformation lente du liner (plastification) probablement due aux traces
d’organiques présents dans l’HCl dans toute son épaisseur. La figure 3 Annexe 1
montre la migration d’HCl dans le PPH au niveau du piquage incriminé. Elle montre
également très bien le décollement du PPH au niveau de son entoilage et la diffusion,
en retour, de produit ayant stagné entre le liner PPH et la couche de résistance
mécanique en SVR.
• une forte décohésion de la stratification de la virole dans les zones où l’acide a pu
pénétrer, ainsi qu’une attaque prononcée de la résine et plus légère concernant les
fibres de verre.
L’analyse fractographique indique que le liner a été fissuré après environ 2000 cycles, soit
approximativement 10 ans d’utilisation suivant les relevés d’exploitation. Cela indique que
l’acide est entré en contact avec la résine depuis plusieurs années. Cette observation est tout à
fait cohérente avec la figure 3 de l’annexe 1 sur laquelle il est possible d’estimer le temps
depuis lequel. L’acide est passé dans le SVR. En effet, l’épaisseur de la migration en retour
correspond à environ 1/5 de l’épaisseur de migration depuis l’intérieur du bac. Sur une durée
d’exploitation de 10 à 15 ans, cela signifie qu’il y a 2 à 3 ans que l’acide est dans le SVR.
La rupture de ce bac a pour origine une fissuration progressive par fatigue mécanique
cyclique au niveau de la liaison trou d’homme / virole. Les cycles sont induits par les
remplissages / vidanges.
L’amorce de fissuration a eu lieu dans cette zone qui est probablement l’une des plus
contraintes du réservoir,. Au bout d’un certain nombre d’années, le liner s’est fissuré et
l’acide a attaqué le SVR.
Il est probable qu’une lente évolution du liner en polypropylène, par diffusion des impuretés
organiques contenues dans l’acide, a conduit à favoriser l’amorçage et la propagation de la
fissuration. Lorsque le liner est fissuré dans toute son épaisseur, l’acide entre en contact avec
la résine.
L’attaque de la résine isophtalique est assez importante, ce qui est connu pour les résines
isophtaliques en milieu acide. Les fibres de verre sont légèrement attaquées par l’acide. Il y a
donc une lente destruction du SVR qui assure la résistance mécanique sur le pourtour du trou
d’homme, au fur et à mesure de l’avancée de la fissure de fatigue du liner.
De surcroît, le phénomène de fatigue auquel est aussi soumis le SVR, a accéléré son
endommagement.
Lorsque la destruction du SVR a été suffisante, le trou d’homme a rompu brutalement, bien
entendu dans une situation où le bac était plein, conduisant à une contrainte hydrostatique
élevée. Le débit de fuite était tel que le bac est passé au vide. Il s’en est suivi une rupture
brutale de la virole qui ne peut résister à cette mise en dépression.
INDEX çç Sommaire
La forme particulière de la rupture (deux demi circonférences) est expliquée sur la figure 4 de
l’annexe 1
II.4 Recommandations
En septembre 2006, un bac de stockage d’acide sulfurique a commencé par fuir pour
finalement se rompre et chuter brutalement.
La cuvette de rétention a été cassée dans la chute. Il n’y a eu aucun blessé, ni aucune pollution
à l’extérieur du site.
Il faut noter que ce bac était récent (durée de vie < 5 ans).
L’annexe 2 donne une vue d’ensemble du bac après la rupture (figures 1 et 2) ainsi que
quelques éléments de l’expertise.
Une petite fuite est apparue au niveau de la patte support de la passerelle pour se transformer
rapidement en un jet de 15 m passant au-dessus de la cuvette de rétention. Cette passerelle est,
en effet, fixée sur le sommet de la cuvette et sur la virole du bac.
L’observation visuelle a montré que la fuite s’est agrandie suivant une déchirure verticale de
la virole au droit de la patte support.
Il y a eu ensuite déformation du haut du bac avec torsion de la structure, puis rupture de la
virole au niveau de la base du bac, au ras de la virole inférieure renforcée.
Il faut noter que l’évent a été examiné et n’était pas bouché le jour de l’incident.
Hormis l’observation visuelle réalisée sur site, des pièces comprenant la zone de fuite avec la
déchirure, des soudures du PVDF, ou une partie de virole saine ont été expertisées par notre
laboratoire.
INDEX çç Sommaire
Les principales observations sont :
Analyse du faciès de rupture au droit de la fuite
L’expertise, sur coupe macrographique ou au Microscope Electronique à Balayage
(MEB), montre que l’origine de la fuite est une fissuration progressive du PVDF par
fatigue mécanique en pleine plaque, amorcée côté acide exactement au droit de la patte
support d’une des poutres de la passerelle.
Le pas d’avancement unitaire de la fissure est relié aux cycles de remplissage/vidange
du bac.
Etat du PVDF et des soudures
Le PVDF est parfaitement sain ; aucune attaque chimique d’aucune sorte n’a été
décelée.
Les soudures dans leur ensemble sont correctes. En particulier, la soudure la plus
proche de la zone de fuite, qui a été contrôlée par ressuage fluorescent, ne présente pas
de fissures.
Etat du SVR
Dans les parties saines, la stratification du bac a été contrôlée. Les essais réalisés
(dureté barcol, mesure du taux de fibres, détermination de la Tg de la résine) montrent
une polymérisation correcte de la résine et une stratification conforme au cahier de
construction.
Dans les zones où l’acide est passé, la résine est nettement attaquée et les fibres de
verres sont corrodées, voir rompues. Il en résulte un délaminage et une baisse très
importante des caractéristiques mécaniques du SVR.
Un essai de corrosion par immersion a montré que cette attaque a une cinétique
relativement lente.
III.4 Recommandations
INDEX çç Sommaire
• bac de forme horizontale (minimise les effets de la pression hydrostatique)
• utilisation d’une résine vinylester,
• surépaisseur de construction (augmentation de la pression de calcul)
• gelcoat externe non pigmenté ; la présence d’un gel-coat opaque gêne la détection
visuelle précoce de dégradations.
• passerelle et tous accessoires indépendants.
Suite aux incidents précédents, un plan d’actions a été mis en place, avec pour but de collecter
un retour d’expérience sur le comportement des réservoirs en matériaux plastiques, et de
définir les suites à donner sur les appareils en service et les appareils futurs.
Ce plan d’actions concerne aussi bien les réservoirs de stockage que les capacités procédés.
Il consiste en les étapes suivantes :
• Recensement des équipements,
• Analyse des inventaires,
• Information et sensibilisation des services inspection,
• Définition d’un plan d’actions site par site à court et moyen terme,
• Révision des standards,
• Définition de bonnes pratiques,
• Consolidation du retour d’expérience.
Sur l’ensemble du groupe, 38 sites dans le monde sont concernés et cela représente environ
700 équipements d’une capacité de quelques litres à 350m3.
Les actions définies à court terme ont été principalement de faire un certain nombre de visites
externes et/ou internes des appareils jugés potentiellement les plus « à risques ».
Au besoin, ces visites sont complétées par des contrôles complémentaires, notamment par
émission acoustique, pour vérifier l’intégrité des réservoirs « à risque » les plus importants.
Les actions définies à moyen terme sont notamment de mettre en place, là où ce n’était pas le
cas, des plans d’inspections basés sur la criticité et tenant compte du retour d’expérience
accumulé.
Le retour d’expérience acquis au travers de ce plan d’actions a permis de définir des bonnes
pratiques. Le but final est d’adapter notre comportement aux spécificités de ces matériaux et
d’améliorer nos pratiques concernant la conception, la fabrication et le suivi en service de ces
équipements.
Les principaux enseignements techniques sont issus, non seulement des incidents récents,
mais aussi de l’enquête générale réalisée dans le groupes et sont les suivants :
• Il est préférable d’utiliser du SVR massif chaque fois que possible plutôt qu’un
thermoplastique fretté SVR. L’écart de comportement mécanique et thermique entre
un thermoplastique et un SVR est important et peut être source de dégradations. Par
ailleurs, l’attaque éventuelle de la couche de résine anticorrosion d’un SVR massif est
généralement uniforme, ce qui permet une réfection. L’utilisation d’un liner
thermoplastique conduit, le plus souvent, à une dégradation très localisée du SVR et
donc à un risque accru pour l’intégrité mécanique de l’appareil.
• Dans le cas d’un équipement en matériau thermoplastique renforcé avec un SVR, il est
préférable de réaliser la couche de résistance mécanique à l’aide d’une résine ayant
une résistance élevée à la corrosion (résine vinylester)
• Sauf cas particulier, il ne faut pas pigmenter ou peindre la couche de finition d’un
équipement SVR ou thermoplastique renforcé. Cela facilite grandement l’inspection
visuelle externe et l’anticipation de problèmes éventuels.
INDEX çç Sommaire
Il faut noter qu’une des difficultés rencontrées est qu’à ce jour, la gamme des contrôles non
destructifs, notamment pour le suivi en service, reste encore peu développée pour les
matériaux plastiques dans nos applications industrielles.
Dans ce cadre, une des techniques de contrôle testée et donnant des résultats satisfaisants, est
le contrôle par émission acoustique.
Cette technique de contrôle fait l’objet du développement ci-après.
Il est vrai que l’inspection des appareils en plastique demande des connaissances particulières
car ces matériaux ont des comportements différents de ceux des métalliques. Même si ces
propriétés sont relativement bien connues, la principale difficulté est de raisonner avec une
« culture plastique » et de ne pas faire de similitude systématiquement avec les
« métalliques » devant un problème. La nature essentiellement stratifiée des SVR conduit à
sous estimer certains endommagements mais le coefficient de sécurité très élevé (8 à 10 en
général) conduit à sous estimer les durées de vie résiduelles.
L’enjeu industriel pour nous est de connaître l’état exact d’un équipement, pour savoir si on
peut le maintenir en service ou non et pour combien de temps. Il est très difficile de répondre
à ces questions avec les techniques traditionnelles d’inspection. Les cas présentés montrent
que, même si un liner est fissuré, l’appareil n’est pas forcément en danger immédiatement
mais il est toujours difficile de se prononcer.
L’émission acoustique est probablement la seule technique permettant cette approche
d’intégrité mécanique globale et d’estimation de durée de vie résiduelle.
La pratique de l’émission acoustique sur les équipements en SVR n’est pas récente
puisqu’elle date du milieu des années 80 mais elle est assez peu utilisée jusqu’à présent dans
le groupe.
Les deux exemples décrits ci-dessous vont permettre de comprendre l’intérêt de cette
technique.
Nous n’entrerons pas dans le détail de la technique car ce n’est pas le sujet ici mais nous
pouvons en faire une description rapide.
Dans les matériaux, de nombreux événements peuvent générer des ondes élastiques qui se
propagent ensuite dans la matière. Dans les matériaux SVR, il peut s’agir de fissures dans la
résine, de décohésion résine/fibre, de ruptures de fibres.
Les capteurs d’émission acoustique ont une telle sensibilité qu’ils peuvent détecter ces
événements individuels mais ils peuvent aussi détecter de nombreux autres phénomènes tels
que le bruit généré par l’écoulement d’un liquide sur une paroi, l’impact de particules sur
cette paroi, le bruit généré par une petite fuite …
Pour bien utiliser l’émission acoustique, il faut connaître les modes de sollicitation mécanique
d’un appareil donné et, lors du contrôle, il faudra exercer les bonnes sollicitations. On dispose
alors un certain nombre de capteurs sur l’appareil, suivant un plan d’implantation précis.
L’objectif est de pouvoir localiser, par triangulation, les sources d’émission acoustique.
Ensuite, il faut solliciter mécaniquement l’appareil. Dans la plupart des cas, il s’agit d’une
montée en pression qui, pour les équipements en plastiques, s’effectue généralement par un
INDEX çç Sommaire
remplissage, éventuellement complété par une légère pression pneumatique en fonction du
niveau de sollicitation souhaité.
Aujourd’hui la méthodologie la plus répandue pour ces contrôles par émission acoustiques sur
les équipements en plastique est basée sur le code CARP ou la norme ASTM E1067
(Standard Practice for Acoustic Emission Examination of Fiberglass Reinforced Plastic Resin
Tanks/Vessels).
Cette méthodologie repose sur l’enregistrement des signaux des capteurs durant tout le cycle
de montée en pression, suivie d’un traitement des données permettant de localiser les sources
émissives. Le code Carp présente l’avantage de fournir une méthodologie rigoureuse pour
conduire le contrôle par émission acoustique des équipements en SVR mais l’inconvénient
d’avoir des critères d’acceptation trop sévères. Il a, en effet, été conçu à l’origine, pour la
réception d’équipements neufs construits suivant code ASME avec des épaisseurs
importantes. Il est possible que les équipements construits en Europe ne passent pas les
critères du code Carp, sans pour autant présenter de risques de fonctionnement. C’est la raison
pour laquelle, il existe aujourd’hui des échelles de sévérité de défaut qui ont prouvé leur
efficacité et qui permettent un diagnostique plus fin de l’intégrité mécanique d’un équipement
et fournissent une aide à la décision précieuse quant au maintient en service ou non d’un
appareil.
Il s’agit donc d’une méthode globale (détection de l’ensemble des défauts de la structure)
avec une indication sur la sévérité de ces défauts. Elle est donc tout à fait adaptée à nos
objectifs de sécurité et de fiabilité et elle constitue une aide à la décision précieuse.
Certains points de la méthode sont importants :
• Cette technique ne détecte pas tous les défauts mais seulement ceux qui sont évolutifs
lors de la sollicitation mécanique imposée. Les autres défauts qui n’évoluent pas sont
donc considérés comme non dangereux. C’est la raison pour laquelle il faut bien
connaître le type de sollicitation mécanique à appliquer.
• La matière a un certain effet mémoire des contraintes qu’elle a subies. Il faut donc
appliquer une contrainte supérieure à ce que l’équipement a vu dans les mois ou
l’année qui précède le contrôle.
• Le cycle de montée en contrainte est important et doit absolument contenir des paliers
d’écoute à contrainte constante. On comprend bien que si un défaut continue à émettre
alors que la contrainte est constante, c’est relativement grave pour l’appareil.
Nous donnons deux exemples qui montrent l’étendue des possibilités d’utilisation de la
technique.
Nous avons, sur une usine, un certain nombre de bacs verticaux de 200 m3 contenant de l’HCl
procédé qui contient des traces d’organiques ainsi que des traces d’HF (< 100 ppm). Nous
savons que l’HF est une petite molécule qui diffuse à la longue, dans les plastiques. On peut
donc avoir des craintes pour la tenue mécanique à long terme car si l’HF rencontre les fibres
de verre, la vitesse de corrosion est élevée.
Il a donc été décidé de réaliser un contrôle par émission acoustique sur le plus vieux de ces
bacs (21 ans de fonctionnement) pour avoir une idée de durée de vie résiduelle. L’usine a
décidé, avant le contrôle, de remplacer ce bac.
Description du bac
INDEX çç Sommaire
• réservoir vertical à fond plat de 200 m3 en SVR massif
• année de construction 1987
• résine Atlac 4010AD, épaisseur anti-corrosion 6,3 mm
• fluide : HCl 30% + traces HF, temperature de service 30°C
Résultat du contrôle EA
L’annexe III donne le résultat du contrôle par EA du bac. On observe que le bac est considéré
comme sain et que seules les zones des capteurs 1 et 9 ont repéré des défauts, classés en
catégorie B (non critiques) et nécessitant une évaluation ultérieure à moyen terme.
La technique nous aurait donc permis de maintenir en service ce bac, en demandant un
contrôle ultérieur par EA d’ici 2 à 3 ans. On voit donc le grand intérêt de cette approche, qui
permet d’avoir une vision globale de l’appareil, sans l’ouvrir (la mise en pression se fait avec
le fluide procédé) et permettant une aide à la décision pour le maintien en service ou non.
L’usine ayant décidé de remplacer le bac, nous avons pu découper la virole, guidés par les
indications EA, dans la zone du capteur 1.
INDEX çç Sommaire
L’une de nos usines possède 4 bacs de stockage de brome liquide qui ont environ 24 ans
d’âge. La question s’est posée de connaitre l’état de ces bacs pour le maintien en exploitation..
L’EA a été utilisée de deux manières différentes.
Dans un premier temps, nous avons réalisé une simple écoute « corrosion » sur les quatre
bacs. En effet, des travaux réalisés dans notre laboratoire montrent qu’il est tout à fait possible
de suivre une corrosion en cours de formation dans une structure SVR. Bien entendu, dans ce
cas, le fluide de remplissage est le brome lui-même car il faut que la corrosion soit active au
moment de l’écoute.
Dans un deuxième temps, nous avons réalisé un contrôle d’intégrité mécanique sur deux des
quatre bacs. Ce contrôle a été réalisé en suivant la procédure du code Carp. La sollicitation
mécanique a été réalisée par remplissage à l’eau complétée par une mise en pression
pneumatique pour compenser la différence importante de densité entre l’eau et le Brome
liquide.
Enfin, l’usine a décidé de détruire le bac qui sera jugé le plus critique afin de corréler les
résultats du contrôle EA avec les défauts éventuels observés lors de l’expertise destructive.
Description des bacs
• bacs horizontaux en PVDF de 5 mm d’épaisseur fretté SVR, volume utile 3 m3
• année de construction : 1984
• résine orthophtalique
• fluide : brome liquide de densité 3,3, température ambiante.
INDEX çç Sommaire
brome liquide parvient à cheminer à travers une microfissure, il y a attaque de
l’entoilage et de la résine.
Nous observons donc une cohérence parfaite entre l’ensemble des tests d’émission acoustique
et les investigations destructives. La très faible émissivité du bac B lors de l’écoute
« corrosion » est due à une migration de brome liquide à travers des microfissures dans une
soudure du PVDF. L’activité acoustique correspondait donc à une corrosion de la résine en
cours de formation. Cependant, l’activité est très faible car la quantité de brome liquide
capable de se frayer un chemin à travers les microfissures est très faible. Comme la résine de
la couche de résistance mécanique n’est pas attaquée, l’intégrité mécanique du bac est très
bonne.
VI. CONCLUSIONS
Nous avons décrit deux incidents sur des bacs de stockage en matériaux plastiques qui ont
amené Arkema à réagir rapidement pour mener un certain nombre d’actions dans le groupe.
Ces actions sont :
• une analyse de fond des causes des incidents
• un recensement complet des équipements en plastique dans le groupe (environ 700)
• une analyse de criticité sur chaque équipement avec un programme de contrôle
personnalisé par site
• un retour d’expérience de ces plans de contrôle
• une modification de nos standards techniques pour prendre en compte le retour
d’expérience.
Les deux incidents présentés ont certaines similitudes et l’expérience acquise nous permet de
dégager les tendances suivantes :
• Limiter autant que possible les constructions avec liner thermoplastique comme
barrière anticorrosion. La différence de comportement mécanique est telle entre le
liner et le SVR, qu’il y a régulièrement fissuration des soudures dans des points
singuliers,
• Limiter si possible l’utilisation de bacs de grande hauteur par rapport au diamètre.
Cela ne concerne pas les colonnes qui contiennent peu de liquide et dont les modes de
ruine potentiels sont très différents de ceux présentés,
• Utiliser une résine vinylester, à résistance chimique élevée, même pour la couche
structurelle SVR en cas de liner thermoplastique,
• Ne pas pigmenter la couche de finition externe.
L’émission acoustique est une technique très performante pour rechercher les défauts et les
classer en termes de sévérité. Elle permet donc d’avoir une bonne vision de l’intégrité
mécanique d’un équipement. Elle permet également de fiabiliser nos unités de manière
efficace.
INDEX çç Sommaire
ANNEXE I - RUPTURE D’UN BAC D’ACIDE CHLOHYDRIQUE
Trou d’homme
haut
bas
Figure 2 : Pièces découpées pour expertise autour du trou d’homme. Position particulière de la
cassure du trou d’homme (demi circonférence à la liaison tubulure / virole et l’autre demi-
circonférence à la liaison bride / tubulure)
INDEX çç Sommaire
Tubulure de trou
d’homme
Soudure
Entoilage au dos du
PPH
Liner de la
virole
Figure 3 : Vue en coupe du PPH dans la zone de rupture. Les points remarquables sont une
migration pleine épaisseur dans le PPH , décollement au niveau de l’entoilage et migration en
retour
Fissuration due à un
report de la déformation
à la liaison
Amorçage de la
fissuration du liner,
déstructuration du Augmentation de
SVR l’amplitude de
déformation
INDEX çç Sommaire
ANNEXE II - RUPTURE D’UN BAC D’ACIDE SULFURIQUE
Cuvette de rétention
INDEX çç Sommaire
Position de la Zones de
jonction des deux rupture du
viroles lors de la bac
construction du bac Cassure
longitudinale
Pattes de fixation de
la passerelle
fuite
Position du haut du
muret de la cuvette
de rétention
Echelle à crinoline
Poutres passerelle
Poutres passerelle
Figure 4 : Schéma montrant les sollicitations mécaniques dues aux remplissages / vidanges. Seule la patte à
droite a conduit à une fissuration par fatigue, la déformée à gauche étant limitée pas l’échelle.
INDEX çç Sommaire
ANNEXE III – Résultat du contrôle par émission acoustique bac HCl
Figure 1 : Schéma d’implantation des capteurs et résultats obtenus en échelle de sévérité de défauts
INDEX çç Sommaire
ANNEXE III – Prélèvements en partie basse (zone capteur 1) et partie haute à
l’aplomb l’un de l’autre
INDEX çç Sommaire
ANNEXE IV : Contrôle par EA de bacs de Brome liquide
A B
C D
Figure 1 : résultat de l’écoute « Corrosion » des différents bacs. Vue développée des bacs.
A B
Figure 2 ; résultat du contrôle intégrité mécanique des bacs A et B
INDEX çç Sommaire
Figure 4 : vue au dos du PVDF au droit de la Figure 5 : Vue en coupe caractéristique de
soudure avec décollement. Dissolution de l’ensemble du bac (excepté zone figure 4)
l’entoilage. montrant l’absence de migration de brome dans
le SVR ou d’accumulation de brome entre
PVDF et SVR
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page109
RÉSUMÉ / ABSTRACT
88G
DE LA NÉCESSITÉ DU CONTRÔLE NON DESTRUCTIF POUR LA
QUALITÉ DES ÉQUIPEMENTS EN COMPOSITE
RÉSUMÉ
Les méthodes CND, dont l’utilisation est largement répandue et normalisée pour les matériaux métalliques,
voient pour certaines d’entre elles leur champ d’application s’étendre aux composites. Comme souvent, c’est le
secteur aéronautique qui est leader dans ce domaine, ou l’application des contrôles par ultrasons est maintenant
systématique sur des produits de structure, souvent dans le cadre d’une mise en œuvre automatisée. Certaines
de ces méthodes de contrôle non destructif peuvent s’appliquer aux équipements sous pression en composite.
Ainsi, les méthodes non destructives de contrôle permettent d’aider les industriels à garantir la qualité des
produits, dans le cadre des contrôles en cours ou en fin de fabrication. Ces méthodes permettent aussi de maintenir
les composants en service en condition opérationnelle, lors d’opérations de suivi , ou pour prévenir une possible
défaillance des composants.
La communication présentera :
¢
Le besoin : via des exemple de défaillance
¢ Des exemples de défauts recherchés
¢ Les spécificités des produits composites pour le contrôle
¢ Les méthodes de contrôle classiques et les méthodes avancées applicables aux composites,
notamment les ultrasons multiéléments, la tomographie (scanner X), la thermographie stimulée
¢ Les potentialités des méthodes non destructives pour le contrôle pendant la fabrication
¢ L’intégration de capteurs de contrôle dans le produit (health monitoring)
La présentation conclura sur la potentialité des méthodes non destructives pour garantir la qualité des composants
composites des appareils à pression.
Mots-clé : contrôle non destructif ; composites ;ultrasons, multiéléments ;ressuage ; contrôle visuel ;
thermographie ; thermographie impulsionnelle ;tomographie ; cartographie ; contrôle en fabrication ; contrôle
intégré, health monitoring ; S.H.M.
INDEX çç Sommaire
DEX
mmaire
DEX
mmaire
DEX
mmaire
DEX
mmaire
DEX
mmaire
DEX
mmaire
DEX
mmaire
DEX
mmaire
DEX
mmaire
DEX
mmaire
DEX
mmaire
DEX
mmaire
DEX
mmaire
DEX
mmaire
DEX
mmaire
DEX
mmaire
DEX
mmaire
DEX
mmaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page110
13D Procédés innovants pour la réparation des centrales nucléaires / CARPREAU J.-M.
Innovative processes for reparing power water reactor JARDIN N.
EDF R&D
85D Contrôle des soudures des piquages des instrumentations (BMI) / BURAT O.
Inspection of PWR Bottom Mounted Instrumentation (BMI) JWELD KERNIN Y.
AREVA INTERCONTROLE
GLASS S.W.
AREVA
CASULA O.
CEA
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page111
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page112
RÉSUMÉ / ABSTRACT
13D
PROCÉDES INNOVANTS
POUR LA RÉPARATION DES CENTRALES NUCLÉAIRES
RÉSUMÉ
EDF réalise des études, d’ingénierie ou R&D, d’anticipation de réparation in-situ des principales zones
stratégiques des gros composants afin d’optimiser la maintenance des réacteurs nucléaires. EDF R&D
étudie, en amont des dossiers d’ingénierie, les procédés « innovants » (MAG-CMT, usinage LASER,
FSW, etc.). Les premiers résultats obtenus en soudage MIG/MAG CMT temperbead robotisé sont
Mercredi
ABSTRACT
EDF studied methods of repairing the main strategic areas of major components to optimize the
maintenance of nuclear reactors. EDF R&D studies, the processes of «innovative» (MAGCMT, Laser
Machining, FSW, etc..). The first results obtained in MIG / MAG robotic CMT temperbead are
encouraging. They offer significant opportunities to improve productivity while simplifying the
management of trajectories filling compared to TIG. The removals of matter by laser (laser pulse
average power 50 W) have achieved a removal of material acceptable industrially. The flow field
reached with non-optimized parameters is 3.65
mm3/min. The preliminary study of FSW welding has achieved the patching of a notch with a depth of
0.8 mm in an austenitic stainless steel plate type 18-10. However, before an industrial application, it
seems essential to achieve a more detailed study demonstrating the processes controlling the impact of
metallurgical and mechanical process.
INDEX çç Sommaire
1-INTRODUCTION
EDF réalise en amont des visites de maintenance des études, d’ingénierie ou R&D,
d’anticipation de réparation in-situ des principales zones stratégiques des gros composants
afin d’optimiser la maintenance des réacteurs nucléaires. C’est dans ce cadre qu’EDF R&D
étudie, en amont des dossiers d’ingénierie, les procédés « innovants » qui pourraient être
utilisés tant pour le parc en exploitation (Réacteur à eau pressurisé) que pour les réacteurs du
futur (réacteur de quatrième génération ou GENIV). Cette contribution sera particulièrement
centrée sur l’étude de trois procédés innovants (cf.§2), le soudage MAG-CMT (Métal Active
Gaz ; Cold Metal Transfet), l’usinage laser et le FSW (soudage par friction malaxage).
2-GLOSSAIRE ET ABREVATIONS
L’objectif du paragraphe ci-dessous est de définir les termes principaux utilisés dans l’article
afin de faciliter sa compréhension.
Procédé innovant = c’est un procédé existant mais non utilisé classiquement dans le domaine
nucléaire de manière générique ou pour une utilisation particulière et dont le développement
pourrait, soit améliorer la qualité d’une opération, soit permettre une opération difficilement
réalisable compte tenu des contraintes d’intervention. A titre d’exemple, les contraintes pour
la mise en œuvre des procédés sur les réacteurs de quatrième génération sont une température
d’intervention à « froid » de 180°C avec des travaux à réaliser sous atmosphère d’argon ou
d’azote avec la présence de sodium ou de vapeurs de sodium.
Actuellement, le remplissage d’un affouillement lors d’une réparation d’un gros composant
est réalisé par soudage TIG Temperbead (Tungstène Inert Gaz). Le soudage TIG est un
procédé qui donne des soudures de haute qualité mais présente des limites dont une faible
productivité. D’autres procédés, par exemple le soudage MAG pourraient venir en
substitution du soudage TIG afin de réaliser un rechargement lors d’une réparation. Le
soudage MAG présente l’avantage d’avoir un taux de dépôt plus élevé que le soudage TIG,
tout en gardant un apport de chaleur pouvant être relativement modéré (fonction des modes de
transfert). De plus, il pourrait être moins contraignant vis à vis des problématiques
d’accessibilité et de gestion des trajectoires (automatisation de l’intervention).
INDEX çç Sommaire
Après avoir présenté la problématique de réparation des gros composants REP (effet
Temperbead), le procédé MAG sera décrit. Ensuite, un essai de rechargement MAG-CMT
(Cold Métal Transfert) robotisé sera présenté afin d’illustrer l’effet Temperbead sur un cas
concret et de comparer les résultats par rapport au TIG.
Les gros composants REP sont en aciers faiblement alliés (type 16 MND5) revêtus par de
l’acier inoxydable austénitique. En fabrication, les composants subissent des traitements
thermiques après soudage qui permettent de rétablir les caractéristiques mécaniques de l’acier.
Ils subissent par ailleurs des pré et postchauffages en vu de se prémunir du risque de
fissuration à froid. Lors d’une réparation in situ, la réalisation de ces traitements est souvent
délicate à mettre en œuvre. Pour s’en affranchir, il est commode d’utiliser le procédé dit
temperbead qui permet d’assurer un revenu en ZAT lors du soudage des couches successives.
Les derniers travaux de l’international Institut of welding [3] recensent 6 modes de base de
transfert du métal lors d’un soudage MIG/MAG. Dont le moins énergétique est le mode court-
circuit (mode A). Ce mode de transfert est obtenu pour un arc court, c’est à dire une faible
distance entre la buse et la pièce, pour une tension de soudage comprise entre 14V et 20V et
pour une intensité de soudage inférieure à 200A. Notons que pendant les courts-circuits, il n’y
a pas apport de chaleur. Le principal avantage du transfert en mode court-circuit est de
transmettre une faible énergie au bain de fusion (avantage pour le Temperbead) et devrait
donc conduire à de faibles contraintes résiduelles. De plus, l’arc est court donc moins sensible
au soufflage magnétique.
Les deux principaux défauts de ce mode de transfert sont :
1. le taux de projection élevé (court-circuit) ;
2. le risque de manque de fusion pendant la phase d’extinction de l’arc (manque
d’apport d’énergie - surtout vrai en soudage manuel).
En effet, si le courant de court-circuit est excessif les forces qui conduisent à la formation de
la goutte vont être violentes et conduire à la formation de gouttelettes. Une bobine d’induction
est introduite dans le circuit afin de limiter les variations de courant et par extension les
éclaboussures.
Les industriels ont développés, suite aux progrès de l’électronique, des optimisations de ce
modes de transfert qui exploitent des technologies particulières au niveau du poste de soudage
et conduisent, à des modes de transfert spécifiques afin de corriger les faiblesses du MAG
(exemple : Cold Métal Transfert chez Fronius, etc. ).
INDEX çç Sommaire
Les avantages de ce mode de transfert sont la faible température du bain et le peu
d’éclaboussure.
Le mode de transfert en court-circuit forcé semble donc adapté pour les trois premières passes
en raison du relativement faible apport de chaleur (faible impact métallurgique). Pour la phase
de remplissage de la fouille, un mode de type pulvérisation axiale pulsé pourrait être envisagé
(soudage plus chaud et plus productif). Cette opération est réalisable avec le même poste
simplement en programmant des lois de synergies spécifiques après les trois couches
Temperbead.
L’objectif des essais est de déterminer la dureté sous cordon après un cordon puis après deux
et trois cordons afin d’évaluer l’impact des cycles thermiques. Une comparaison sera réalisée
avec le soudage TIG Temperbead (procédé maîtrisé).
Le dépôt Temperbead est réalisé sur un bloc massif en acier 20MND5. Les passes de la
première couche du rechargement Temperbead sont réalisées avec un apport ER 309 L. Les
passes des deuxième et troisième couches du rechargement Temperbead sont réalisées avec
un apport ER 308L.
Le gaz de protection est un gaz de désignation commerciale Arcal 12. Pour des applications
de réparation sur les REP, un mélange à base d’argon peut être envisagé. Des études seraient à
réaliser afin de déterminer si de faibles additions d’hélium, de gaz carbonique ou d’oxygène
ne seraient pas à même d’améliorer le rapport productivité/qualité.
Tous les essais ont été conduits sans préchauffage et avec une température interpasse
comprise entre 20 et 80°C. La position de soudage utilisée est la position corniche. L’essai de
soudage a été réalisé avec un poste de soudage FRONIUS TPS 4000 permettant de travailler
avec le mode CMT.
L’énergie moyenne déposée lors du soudage (U*I/Vitesse de soudage), calculée à partir des
valeurs moyennes indiquées par le générateur, est de 2,8 kJ /cm. En soudage TIG temperbead,
les énergies utilisées lors d’un essai comparable sont de 13 kJ/cm.
Les essais ont été réalisés au centre de recherche de l’Institut de soudure (Yutz).
INDEX çç Sommaire
Dans un premier temps un
cordon en position corniche a
été réalisé. Les examens
macrographiques montrent
une profondeur de pénétration
variant entre 1,8 et 1,9 mm,
une largeur de cordon variant
entre 8,6 et 8,9 mm et une
dilution variant entre 27 et
35%.
Figure n°1 : dureté sous un cordon
La dureté sous un cordon est élevée. En limite de ZAT, la valeur peut atteindre HV10 = 455
pour une valeur en métal de base de HV10= 238. Pour un tel niveau de dureté, la limite
élastique est supérieure à 950 MPa.
Ensuite des rechargements avec plusieurs cordons ont été réalisés afin d’évaluer l’impact des
cycles thermiques du soudage sur la dureté sous cordon
dureté
Variation de la dureté en fonction de nombre de couche (effet de revenu)
500
475
450
425
400
375
350
325
300
275
250
225
200
0 1 2 3
nombre de couche
Dans la zone sous trois cordons, la dureté maximale est de HV10=386 avec une valeur
moyenne de HV10=355 (écart type 20). Une analyse plus fine des résultats montre que dans la
zone médiane de la ZAT, la dureté moyenne est HV10=350 (écart type 20). A l’interface métal
de base ZAT, cette valeur est légèrement plus élevée HV10=366 (écart type 20). Une
affectation thermique plus faible dans cette zone permet d’expliquer cette différence. A
l’interface du métal fondu et de la ZAT, la dureté (HV10=356 et HV10=342) est comparable à
celle de la zone médiane.
En conclusion, la dureté sous cordon est de HV10=366 après trois couches. Un effet de revenu
INDEX çç Sommaire
par les cycles thermique des deuxièmes et troisièmes cordons est bien observé. La dureté,
après trois cordons, diminue de HV10= 455 (une mesure) à HV10=366 (écart type 20). Après le
deuxième cordon l’effet temperbead semble déjà se manifester HV10=371 (écart type 26).
Lors d’essais de soudage TIG temperbead, en position corniche à une énergie de 13kJ/ cm, sur
un matériau comparable et avec des produits d’apport identiques, la dureté moyenne en zone
médiane de la ZAT est de HV10= 298 (écart type 17).
Les premiers résultats obtenus en soudage MIG/MAG CMT temperbead robotisé sont
encourageants. Ils offrent des perspectives d’amélioration significatives de la productivité tout
en simplifiant la gestion des trajectoires de remplissage par rapport au procédé TIG. En
affinant les paramètres opératoires et la composition du gaz de protection des améliorations
des caractéristiques mécaniques pourraient être envisagées. Toutefois, en amont d’une
application industrielle, il semble indispensable de réaliser des études plus détaillées
démontrant la maîtrise de la qualité de la soudure (absence de manque de fusion) dans le
cadre d’une productivité améliorée.
4- USINAGE LASER
Les principaux avantages qui ont conduit à analyser les potentialités de l’enlèvement de
matière par usinage laser pour les opérations de réparation des réacteurs (principalement
GENIV) sont :
1- l’usinage peut être réalisé sans reprise d’effort ce qui va conduire à une simplification des
porteurs et par extension des contraintes associées sur la conception,
2- l’usinage est réalisé à distance,
3- l’usinage peut être réalisé sur de petites surfaces ce qui permet de limiter les effets
thermomécaniques associés au procédé.
Après avoir rappelé les principes de base de l’usinage laser, les avantages et les inconvénients
des divers modes d’usinage laser seront analysés afin de sélectionner celui qui est le mieux
adapté à la réparation des réacteurs nucléaires (principalement de quatrième génération).
Enfin dans la dernière partie nous présenterons des résultats d’essais d’usinage laser.
INDEX çç Sommaire
4-1 Principes physiques de procédé
L’usinage laser consiste à enlever de la matière par échauffement de la surface à l’aide d’un
faisceau laser. En fonction de la puissance surfacique (W/cm2) du laser, la matière passe de
l’état solide à l’état liquide (liquéfaction) puis à l’état gazeux (vaporisation). Au final, c’est
l’évacuation de cette matière (liquide ou gaz) qui conduit à la formation d’un affouillement.
Le premier mode exploite plus la puissance moyenne d’un laser (régime thermique établit).
Les deux dernières requièrent un laser impulsionel et exploitent les puissances crêtes des
lasers (régime transitoire).
L’usinage par Keyhole est réalisé à des puissances surfaciques de l’ordre Ps >105 W/cm2
avec une tache focale sur la pièce de l’ordre du mm2. Il y a formation d’un keyhole associé à
la présence d’une phase liquide sur le front d’avancé (processus identique au soudage).
L’usinage est réalisé par application d’un jet de gaz neutre (argon ou azote) à haute pression
(ordre de 15 à 20 bars) qui évacue la phase liquide par la face arrière. Cette méthode est bien
adaptée pour réaliser du perçage ou de la découpe (évacuation de la matière par la face
arrière). Actuellement, EDF étudie, dans le cadre du démantèlement de Super-Phenix la
possibilité d’éliminer les rétentions de sodium par découpe laser.
Par exemple, la référence [1] présente un perçage percusionel avec une pression surfacique de
l’ordre de 10 MW/cm2, associé à des faisceaux focalisés (400 µm) et de temps d’impulsion
d’une µs. La projection par effet de piston des gouttelettes de métal en fusion peut conduire à
INDEX çç Sommaire
l’apparition de scories. Des stratégies d’amorçage ont été développées afin de minimiser la
projection de scories. L’augmentation de la vitesse de perçage par augmentation de la
puissance (et donc de la température) conduit à une diminution de l’épaisseur de la couche
liquide. Ensuite, le régime devient un régime d’usinage par vaporisation. L’usinage par
percussion ne sera pas retenu car il conduit à des projections liquides qui en se solidifiant
donnent des scories.
L’usinage par vaporisation est réalisé pour des puissances surfaciques plus élevées, par
exemple de l’ordre Ps > 17 MW/cm2, obtenues par des faisceaux focalisés (<400µm) et des
temps crête encore plus faibles (<µs). Le laser fait vaporiser la matière et la phase liquide est
négligeable (régime ablatif). Cette phase sera d’autant plus négligeable que la durée de
l’impulsion est faible ce qui minimise la diffusion thermique et donc l’effet thermique sur le
matériau. Par exemple, les lasers utilisés classiquement en gravage pourraient convenir. Leurs
caractéristiques sont :
- temps de l’impulsion inférieure à 150 ns (peu de chaleur diffusée dans la pièce),
- puissance du laser 100W (moyenne),
- puissance crête 10 KW (soit une puissance surfacique de 8 MW pour fibre de diamètre
400µm).
Des effets mécaniques peuvent apparaître pour des temps crêtes inférieurs à la nanoseconde et
des énergies supérieures à 106 MW/cm2 (exemple, en choc laser t =10-12s P= 1012 MW/cm2).
L’interaction laser-matière conduit à l’apparition d’un plasma très localisé et très dense. La
détente du plasma conduit à l’apparition d’une onde de choc
Des essais de faisabilité ont été réalisés afin de confirmer cette première analyse
Le laser utilisé pour ces essais est un laser IPG type YLP-1/120/50/50. Ce laser est un laser
fibré, dopé aux ions ytterbium Yb3+. Les caractéristiques du laser sont une longueur d’onde de
1064 nm, une puissance moyenne de 50 W et une durée de pulse de 120 ns pour une
fréquence de fonctionnement entre 50 et 100 kHz.
INDEX çç Sommaire
Le scanner utilisé est un scanner SCANLAB modèle
Hurryscan 10, refroidi par eau, montée avec une lentille
focale 163 mm offrant un champ de travail de 100 x 100
mm.
Figure n° 4 : Installation
La stratégie d’usinage par laser repose sur le fait de décomposer le volume en fines strates.
Chaque strate sera successivement usinée en 2D puis décalée en profondeur pour obtenir une
géométrie 3D. De cette manière, le faisceau laser usine une couche puis le point focale est
descendu de l’épaisseur de la couche pour usiner la suivant et ainsi de suite.
La maquette a ensuite été découpée afin de faire une analyse plus précise de l’usinage. Les
parois de l’usinage présentent un agglomérat de dépôt noir. Mais contrairement au dépôt à la
surface du matériau qui se présente sous forme de poudre, ce dépôt-ci est solide et bien
adhérent à la paroi. Plusieurs tentatives pour enlever ce dépôt ont été testées : nettoyage par
solvant, brossage à la brosse synthétique et décapage laser. Aucune de ces tentatives n’a été
fructueuse. Le fond de la fouille présente un aspect propre et brillant. C’est la seule surface
qui a subi une passe de finition laser. En effet, avec le montage utilisé, il n’est possible
d’utiliser le faisceau que dans un plan horizontal.
Cette première série d’essais a été réalisée sans protection gazeuse ce qui peut expliquer le
couleur du dépôt. Des essais avec une protection argon sont à envisager.
5- FRICTION MALAXAGE
Le soudage FSW est un procédé qui pourrait s’utiliser sous eau (Réacteur REP) ou sous
sodium (GEN IV) pour la « réparation » des gros composants d’une centrale nucléaire.
L’utilisation de ce procédé conduira à une évolution des stratégies de réparation. En effet,
actuellement les stratégies de réparation conduisent à une élimination du défaut par
affouillement suivi d’un rechargement par soudage. Le FSW conduira plus à une
« élimination » des défauts par malaxage.
Après avoir rappelé les principes du soudage FSW et les essais réalisés, une analyse des
perspectives d’applications sur GENIV sera réalisée.
Le principe du soudage par friction malaxage est relativement simple. La tête de soudage
échauffe la matière afin de faciliter son écoulement puis ensuite, malaxe cette matière afin de
créer un joint soudé. Le soudage par friction malaxage présente la caractéristique principale
de ne pas conduire à la fusion de la matière, c’est un soudage à l’état solide, à une température
inférieure à la température de fusion limitant ainsi les conséquences d’un cycle thermique
brusque et de grande amplitude (transformation de phase, contraintes résiduelles, etc.).
La tête de soudage est composée d’un épaulement (shoulder) auquel est fixé un pion (pin ou
probe) sous sa face inférieure. Le diamètre de l’épaulement est de 2 à 2,5 fois supérieur au
diamètre du pion est mis en rotation rapide.
INDEX çç Sommaire
Pour le soudage de matériaux plus « durs » tels que les
aciers et les aciers inoxydables, le titane, les outils de
soudage sont soumis à des sollicitations mécaniques
très élevées à haute température. Dans ce cas, le choix
de la matière outil conditionne la réussite du procédé.
Le tungstène-rhénium a été utilisé avec succès pour le
soudage des aciers et du titane (tenue à l’abrasion
insuffisante pour la fabrication de pièces en grande
série). Des céramiques peuvent également être
employées ainsi que le PCBN (nitrure de bore cubique
Figure n°7 : Pion et épaulement polycristallin) qui possède une tenue à l’abrasion
élevée mais une certaine fragilité au choc.
Sur une coupe micrographique d’un joint FSW, quatre zones sont observées :
1. le métal de base qui ne subit aucune déformation ni échauffement,
2. la zone affectée thermiquement (ZAT) où aucune déformation par le passage de l’outil
n’est observée,
3. la zone affectée thermo-mécaniquement (ZATM) où les effets de la déformation se
font
ressentir
4. la zone malaxée ou noyau
Le joint soudé FSW n’est pas symétrique. Cette dissymétrie provient principalement des flux
qui diffèrent de part et d’autre du plan de joint. Ainsi, on définit un côté avançant (Advancing
side), côté suivant lequel les vecteurs avance et rotation de l’outil sont dans le même sens et
un coté reculant (retreating side), côté suivant lequel les vecteurs sont de sens opposés.
Autour de l’outil, les flux de matière vont du côté avançant vers le côté reculant.
Le flux de matière est complexe. Il peut être décomposé en plusieurs zones. La première sous
l’épaulement où le mouvement résulte du frottement de l’épaulement sur la surface de la pièce
(flow arm zone). C’est une zone complètement recristallisée à la surface du cordon, zone
forgée par le frottement de l’épaulement. Cette zone est appelée flow arm. La deuxième zone
(zone malaxée ou noyaux) est celle où la matière est travaillée par la cinématique du pion. La
troisième zone (Zone Affectée Thermomécaniquement) est à l’interface des deux premières et
résulte d’un mouvement combiné. Enfin, une zone est localisée sous le pion où la matière est
entrainée par le pion. Ceci explique pourquoi les pions sont toujours d’une profondeur plus
INDEX çç Sommaire
faible que le joint à souder. La combinaison de ces mouvements permet de déplacer la matière
d’avant en arrière.
Les essais mécaniques montrent que la zone la plus faible mécaniquement est la zone de
liaison entre la ZAT et la zone affectée thermomécaniquement.
Des réparations ont été réalisées par FSW (nommé dans ce cas Friction Stir Repairing, FSR).
Deux stratégies ont été adaptées soit sans apport de matière pour les défauts de faible
compacité soit avec apport de matière (avec des inserts vissés).
Des essais de faisabilité ont été réalisés afin d’analyser la capacité du procédé FSW à
reboucher une petite fissure.
L’objectif de cette étude préliminaire était de reboucher par FSW robotisé une fissure de
faible profondeur sur une maquette en acier inoxydable austénitique afin de mieux cerner le
potentiel du procédé et par extension de définir les axes de progrès afin d’établir
l’applicabilité de ce procédé.
La tête de soudage est constituée d’une électrobroche, d’un porte outil et d’un outil. L’effort
vertical maximal développé est de 15 kN. Le porte outil a été équipé d’ailettes permettant
d’augmenter la surface de contact avec l’air, et ainsi d’améliorer son refroidissement dans le
but de ne pas propager la chaleur issue de l’opération de soudage vers l’électrobroche.
La tête de soudage est mise en œuvre par un robot 6 axes KUKA KR 500 (Version MT, tête
de soudage CYTEC). La vitesse maximale d’avance est de 4000 mm/min. La vitesse de
rotation maximale de la broche de 2500 tr/min.
Les essais ont été réalisés au centre de recherche de l’Institut de Soudure (Yutz).
La présence d’une fissure a été simulée par la réalisation d’une entaille de largeur 0,15mm, de
profondeur 0,8mm et de longueur 30mm. Le cordon FSW est réalisé en pleine tôle sur
l’entaille.
INDEX çç Sommaire
Figure n°9 : macrographie des entailles
La coupe métallographique de l’essai représentée sur la figure n°10, montre que l’entaille a
été totalement rebouchée et remalaxée lors du passage de l’outil FSW (l’outil y est représenté
schématiquement en blanc). Au niveau du défaut, seule une trace bleutée subsiste. Cette trace
pourrait correspondre à la présence d’éléments (oxyde, etc.) dans l’entaille.
De plus, ce cordon présente une zone hétérogène de couleur jaune. Cette zone se trouve dans
le noyau du cordon soudé (zone malaxée directement par l’outil et recristallisée
dynamiquement). Des mesures de dureté ont été réalisées dans cette zone. Les points 1 et 2 se
trouvent dans la zone affectée thermo-mécaniquement de la soudure, les points 3 et 4 dans la
zone hétérogène, et les points 5 et 6 dans la zone malaxée. Le tableau ci-dessous présente les
résultats obtenus :
Dureté mesurée
Point de mesure Zone malaxée
(HV)
1 203 ZATM
2 187 ZATM
3 304 Zone hétérogène
INDEX çç Sommaire
Le métal de base possède une dureté de 181 HV. Il apparaît donc que la zone contenant les
points 3 et 4 possèdent une plus grande dureté que le reste de l’échantillon. Cette zone
pourrait être composée d’oxydes pouvant provenir de la surface de la pièce.
La surface des coupons avant soudage, est décapée en utilisant un disque rotatif à lamelles,
grain 80 avant de reprendre le programme d’essais à effort de soudage variable.
Les examens macroscopiques et les micrographies de l’essai représenté sur la figure n°11
démontrent que le décapage de la surface des coupons avant soudage permet d’améliorer
l’aspect de surface du cordon qui présente dès lors un aspect brillant. La micrographie montre
l’absence de zone hétérogène significative dans le cordon. Une filiation de dureté a été
réalisée sur le cordon 11. Le profil de dureté est relativement plat, avec des variations peu
prononcées par rapport à la dureté du matériau de base (dureté maximale mesurée dans le
cordon : 191 HV2, dureté minimale relevée dans le cordon : 168 HV2). Le métal de base
possède une dureté de 181 HV2.
Enfin, les comparaisons des micrographiques représenté sur la figure n°11 montrent que
l’effort de soudage semble pouvoir être abaissé à 7kN.
Figure n°11 : coupes micrographies d’un essai avec un effort de soudage de 9 kN (figure
supérieure) et 7 kN (figure inférieure).
L’étude préliminaire a permis de réaliser le rebouchage d’une entaille d’une profondeur de 0,8
mm dans une plaque d’acier inoxydable austénitique de type 18-10. L’étude a permis de
démontrer qu’un effort de forgeage de 10kN était très certainement supérieure à l’effort
minimum à appliquer afin d’obtenir un forgeage correcte de la matière. Des essais à 7 kN
semblent convenir (Gain de 30%). A ce niveau d’effort il faut vérifier la pertinence du
chauffage de la matière par friction.
INDEX çç Sommaire
6- CONCLUSION
Cette présentation à permis de montrer qu’au-delà des procédés parfaitement maîtrisés comme
le soudage TIG et le fraisage grande vitesse, il existait d’autres procédés comme le soudage
MAG-CMT, l’usinage laser et le soudage par friction malaxage qui pourraient améliorer la
qualité des réparations sous réserve de réaliser les études permettant de maîtriser les principes
physiques mis en jeu et les paramètres influents. Les outils de modélisation et de simulation
sont un excellent moyen pour réaliser ces études et capitaliser l’expérience acquise.
7- BIBLIOGRAPHIE
[1] M. Schneider, « perçage profond par laser analyse des processus», thèse 2006,
Université de Paris 6.
[3] : D Iordachescu “Steps toward a new classification of metal transfer in gas metal arc
welding” Journal of materials processing technologie, 2007.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page113
RÉSUMÉ / ABSTRACT
41/81D
LES SERVICES INSPECTION À EDF
M. Bethmont et F. Morin
RÉSUMÉ
EDF a mis en place des services inspection sur les sites de production d’électricité (centrales thermiques
et nucléaires) dès les années 2000 en vue de leur reconnaissance pour appliquer des plans d’inspection
aux équipements sous pression.
Les services inspection sont des équipes indépendantes au sein de chaque établissement en relation
étroite avec les Services Opérationnels. Les modalités mises en place pour assurer l’autorité,
l’indépendance et la compétence des SIR sont commentées. Une caractéristique majeure des plans
d’inspection est le contrôle des « zones sensibles » identifiées par le retour d’expérience. La méthode
d’EDF qui permet de déterminer les périodicités de contrôle de ces zones est une démarche qui associe
l’analyse du risque de défaillance à la gravité de la défaillance.
Un retour d’expérience (REX) des dix années des services inspection est décrit. Des exemples de
préconisations d’arrêts d’exploitation d’équipements sous pression et d’actions préventives sont
présentées. De plus un REX sur l’organisation, le maintien des compétences, l’indépendance, etc. est
présenté avec un commentaire sur les évolutions envisagées.
ABSTRACT
From the year 2000, Inspection Services (IS) have been set up on EDF power plant stations with the
view to be « recognized » by French Safety Authority and to apply “inspection program” to control
pressure vessel equipment.
The organisation has been defined with a few people in each IS so that there is a strong relationship
which is created between IS and O&M Services. The conditions to maintain the authority, the
independence and the competence of IS are described. A specificity of the inspection program is the
periodical control of “sensitive zones” identified from the feed back experience. The principles to
evaluate the periodicity of these controls are based on a method mixing risk failure analysis and gravity
analysis.
The IS experience is analyse after these 10 years. Examples are given to show how IS have had to order
to stop operation of pressure vessels or to recommend preventive actions.
The feed back experience about IS organisation, competences, independence, etc. is presented with a
discussion on the possible evolution.
INDEX çç Sommaire
1 – INTRODUCTION
Pour sa production d’électricité, EDF exploite 20 000 Equipements sous pression (ESP)1
répartis sur les 57 tranches des centrales nucléaires2 et les 23 tranches des centrales
thermiques (charbon ou fioul).
En tant qu’exploitant d’un parc important d’ESP, EDF s’est naturellement interrogé dès les
années 2000 sur la possibilité de mettre en place l’organisation des services inspection, telle
que l’Administration le proposait, pour pouvoir appliquer des plans d’inspection (PI) adaptés
aux modes d’exploitation des tranches.
Sur la base du retour d’expérience des industriels, en particulier les industries pétrolières et
chimistes, il est apparu que les services inspection, dans la mesure où ils pouvaient être
« reconnus » par l’Administration, apporteraient des gains tant du point de vue de la sécurité,
que de l’exploitation (disponibilité) et de la maintenance des installations des centrales
nucléaires qui fonctionnent en « base » et des centrales thermiques dont la mission est
d’assurer les pointes de production.
Ainsi les enjeux des Services Inspection Reconnus (SIR) pouvaient se résumer par :
Sur la base de cette analyse, la décision a été prise par les Divisions de Production Nucléaire
(DPN) et Production Thermique (DPIT) de mettre en place et de faire reconnaître
systématiquement des services inspection sur chaque site conformément à la réglementation
(DMTP 32-510).
L’objet de cette présentation est de décrire les principales caractéristiques des SIR EDF (cf.
chap. 2 et 3) et de commenter leur retour d’expérience après ces dix années (cf. chap. 4).
1
Equipements réglementés selon l’arrêté du 15 mars 2000.
2
Ne sont pas compris les équipements sous pression « nucléaires » (ESPN)
INDEX çç Sommaire
2 – LES SIR à EDF
Sous réserve de respecter les prescriptions de la DMTP 32-510, les SIR peuvent avoir des
organisations différentes selon les établissements. L’option d’EDF a été de constituer des SIR
de dimension relativement restreinte au sein de chaque établissement en relation étroite avec
les Services Opérationnels. Aussi les SIR ne prennent pas en charge l’ensemble des activités
associées aux ESP, ils en confient la réalisation aux Services Opérationnels. Les activités
confiées et les activités propres du service inspection sont précisément définies, contrôlées et
font l’objet de supervision.
Par ailleurs, les SIR peuvent s’appuyer sur des Services Centraux nationaux d’EDF qui ont en
charge :
- l’animation du réseau des SIR et l’appui auprès du management des sites ;
- les relations avec l’Administration ;
- l’appui pour l’expertise et le retour d’expérience (ingénierie nationale) ;
- les analyses de traitement d’écart par des calculs de non-nocivité des défauts et des
dossiers de calculs de durée de vie (calculs mécaniques, modes de dégradations,
métallurgie, CND, process, …) ;
- le pilotage du processus de révision des « guides spécifiques »3
- la conduite des audits internes des SIR (avant reconnaissance et renouvellement…) ;
- la qualification des inspecteurs et des responsables SIR par une Commission Nationale
de Reconnaissance des Compétences (CNRC) : Entité transverse et indépendante des
Unités de Production.
Une attention particulière a été portée dès le début pour s’assurer que la compétence des
Services Centraux n’empièterait pas sur la responsabilité des SIR.
Conformément aux exigences de la DMTP 32-510, il est attendu de la part des SIR qu’ils
assurent complètement leurs missions grâce à leur autorité, leur indépendance de jugement et
leur compétence technique.
3
Un guide spécifique est un rapport technique national dressant la liste des zones sensibles pour chaque ESP
compte tenu du REX, que les SIR utilisent pour élaborer les plans d’inspection
INDEX çç Sommaire
Direction
SIR
De plus, les SIR doivent avoir l’autorité nécessaire, notamment pour les autorisations de
remise en service d’un équipement soumis à sa surveillance et par l’émission de
préconisations (à statut réglementaire) en tant que de besoin. Cette autorité attendue est
formellement contractualisée entre le RSI et le chef d’établissement.
La compétence des inspecteurs et des responsables des SIR est un élément clé indispensable à
la base des deux autres critères. C’est pourquoi les exigences pour la qualification des
inspecteurs sont fortes. Tout d’abord les inspecteurs doivent avoir une expérience et un
parcours professionnel suffisants. En outre, ils reçoivent une formation adaptée en fonction de
leur profil. Cette formation est réalisée à travers des stages ou en suivant la formation
spécifique ESP récemment mise en place par EDF avec l’Ecole des Mines de Douai et à
travers un compagnonnage dûment tracé.
La CNRC a été mise en place au niveau national pour pouvoir qualifier quatre types de
compétences :
- Inspecteur Niveau 1,
- Inspecteur Niveau 2,
- Responsable de Service Inspection (RSI)
- Inspecteur - Suppléant de RSI
Sur la base de cette qualification, le chef d’établissement habilite le RSI et les inspecteurs (en
fonction du respect de critères propres à chaque établissement).
Sous réserve que les inspecteurs et les responsables n’aient pas cessé leur activité durant plus
de six mois alors la CNRC peut reconduire leur qualification tous les trois ans, sur la base
d’un dossier justificatif du maintien de leur compétence.
INDEX çç Sommaire
Plans d’inspection
Les plans d’inspection (PI) sont élaborés par les SIR conformément au « guide professionnel
EDF » qui a été approuvé en 2004 par la Section Permanente Générale (SPG) de la
Commission Centrale des Appareils à Pression » (CCAP)4.
Une caractéristique majeure des PI est le contrôle des « zones sensibles » caractérisées par le
retour d’expérience (REX) interne ou externe à EDF. La périodicité de ces contrôles est
évaluée selon une analyse spécifique développée et mise en œuvre à EDF (cf. chap. 3). Les
inspections, proprement dites, sont maintenues avec une périodicité maximale de 6 ans (au
lieu de 18 ou 40 mois) et les requalifications à 12 ans (au lieu de 10 ans).
Ces plans d’inspection, précisent aussi les modalités de contrôle des paramètres d’exploitation
et en particulier ceux définis en tant que Condition Opératoire Critique Limite (COCL). Les
paramètres influant directement sur les modes de dégradation qui sont objet de COCL, sont
par exemple, la température de fonctionnement, les débits de vapeur, le nombre d’écarts de
température, etc. (cf. tableau 1).
Ainsi, les plans d’inspection couvrent-ils non seulement les contrôles à l’arrêt mais aussi un
suivi programmé des ESP en exploitation.
Fluage Température
Tuyauteries vapeur saturée et eau Corrosion/érosion PH
surchauffée dont le taux de Température
chrome < 1%
(Toutes tranches)
Tableau 1 : Exemples de COCL (centrales thermiques)
4
Une première version avait été approuvée pour les centrales thermiques en avril 2002 et la seconde version,
dont le domaine est élargi aux sites nucléaires, a été approuvée en 2004
INDEX çç Sommaire
3 – LES PLANS D’INSPECTION : LA METHODE EDF
Les plans d’inspection sont élaborés par chaque SIR sur chacun des sites de façon
indépendante permettant de prendre en compte toutes les spécificités locales c’est-à-dire le
retour d’expérience des évènements spécifiques des sites mais aussi, les conditions
d’exploitation des installations. Par contre la démarche pour chacun des SIR est équivalente
en commençant par le partage d’expérience national.
En effet, le retour d’expérience sur les défaillances des ESP de l’ensemble des sites est
répertorié au niveau national et partagé pour être pris en compte sous la responsabilité de
chaque SIR. Ce REX « national » permet de définir des zones sensibles (ZS) c’est-à-dire des
zones pour lesquelles des dégradations ont été détectées sur le parc des centrales exploitées à
EDF.
La méthode d’EDF qui permet de déterminer les périodicités de contrôle des ZS repose sur
une démarche pragmatique, auditable et permettant à chaque inspecteur de l’expliquer, étape
par étape. Ce n’est pas une méthode d’analyse de criticité au sens strict du terme mais une
méthode d’évaluation de la périodicité des contrôles des ZS. Les principes partent des
constats suivants :
3. En revanche, le REX démontre que des zones les moins sensibles peuvent n’être
contrôlées que lors des requalifications, voire après. Par précaution toute zone sera
contrôlée a minima lors des requalifications à une périodicité portée à 12 ans ;
4. Pratiquement, les périodicités des zones sensibles seront fixées entre 18 mois (1,5 an)
et 12 ans selon un « indice» caractéristique de la sensibilité de la zone. Il a été décidé
que l’on se limitera à 4 périodicités pour des raisons de commodité ; à savoir : 1,5 ; 3 ;
6 et 12 ans ;
5. La démarche pour déterminer l’indice de la périodicité des contrôles de ces zones a été
voulue simple et pragmatique. Elle est décrite dans un logigramme qui croise le
« risque de défaillance » avec la « gravité de la défaillance » ce qui est une démarche
INDEX çç Sommaire
d’analyse de risque. Elle aboutit à déterminer un indice dit de périodicité compris
entre 1 et 4 (cf. figure 1).
Ainsi l’indice de périodicité est évalué entre 1 et 4 et la périodicité des contrôles en est
déduite d’après le tableau de la figure 2.
On notera que cette démarche n’a pas comme objectif de diminuer la « criticité » de zones
sensibles, en préconisant des modes d’exploitation ou des activités de maintenance
spécifiques. Mais il s’agit de s’assurer que les périodicités de contrôles sont adaptées aux
risques d’endommagement des zones sensibles. Elle se justifie dans la mesure où les
équipements ainsi contrôlés sont en grande majorité sous pression d’eau ou de vapeur c’est-à-
INDEX çç Sommaire
dire en milieu peu agressif. Et de façon générale ils ne sont pas remplacés pendant la durée de
vie de l’installation.
Enfin, initialement cette méthode avait été construite pour les équipements des centrales
thermiques. Par la suite, elle a été étendue pour les équipements conventionnels des centrales
nucléaires. Mais dans ce cas, deux modifications ont été apportées. Les périodicités
spécifiques aux ESP des centrales nucléaires ont été limitées à des périodicités
calendaires soit : 2, 4, 6 ou 12 ans pour tenir compte des périodes d’arrêts des installations. De
plus compte tenu du REX plus récent de ces ESP, des zones sensibles « d’avis d’expert » ont
été définies. Il s’agit de zones d’endommagement potentiel encore non détecté mais qui de
l’avis des experts pourrait se produire.
Figure 3 : Schéma des zones sensibles d’un transformateur de vapeur (centrale nucléaire)
4 – RETOUR D’EXPERIENCE
Le risque pression c’est-à-dire le risque d’une défaillance d’un ESP peut être « maîtrisé » soit
par des mesures immédiates dès l’apparition du risque de la défaillance (fuite ou rupture)
voire à la défaillance elle-même soit par des mesures préventives empêchant ou limitant le
risque d’apparition des défauts à l’origine de ces défaillances. Sur ces deux aspects les SIR
interviennent directement.
INDEX çç Sommaire
Concernant les mesures immédiates, plusieurs exemples peuvent être cités. Des
préconisations d’arrêt d’exploitation des ESP ont été formulées sur plusieurs sites lorsque
ceux-ci présentent des risques de fuite ou de rupture ou dans le cas de non conformité des
inspections ou requalifications. Dans quelques cas, ces préconisations ont déjà conduit à
l’arrêt de l’exploitation des tranches. Bien entendu, lors de ces situations, qui sont
exceptionnelles, le SIR est un acteur majeur pour valider et contrôler le traitement de l’écart
afin que celui-ci soit résolu dans les meilleurs délais.
Un bilan montre que les SIR ont été amenés à prescrire plus fréquemment des arrêts d’ESP
moins contraignants mais également liés à des problèmes de sécurité ; par exemple, en cas de
dépassements de date de requalification, de dossiers de calculs erronés (exemple d’un dossier
de calcul suite à des affouillements de vannes d’admission de vapeur), de non-conformité
d’équipement (cas d’un groupe froid car sans accessoire de sécurité), etc.
Mais c’est aussi par l’anticipation que le rôle du SIR est majeur et les exemples sont
nombreux.
Par exemple, la détection d’une fuite sur une tuyauterie, à l’entrée d’un condenseur de vapeur
d’un site nucléaire, a conduit le SIR à exiger l’arrêt de la ligne pour expertise et réparation
provisoire. Sans l’intervention du SIR, cette fuite aurait été « collectée », comme l’Exploitant
l’envisageait, et aurait conduit selon toute vraisemblance à la rupture de la tuyauterie qui
aurait fait perdre le vide du condenseur et finalement aurait conduit à l’arrêt de la tranche sur
une durée plus longue.
On peut citer aussi, l’anticipation d’un SIR d’un site thermique qui a préconisé par prévention
des contrôles de tuyauteries sous pression à hautes températures sur la tranche voisine de la
tranche où ces mêmes tuyauteries ont présenté des fuites. Le résultat s’est révélé pertinent
puisque les mêmes fissurations ont été détectées et réparées lors d’un arrêt de tranche
programmé. De façon semblable, sur un site nucléaire, une rupture brutale d’un compensateur
entre une tuyauterie et une calandre s’est produite et a entraîné trois semaines
d’indisponibilité de la tranche. Or l’expertise réalisée a conclu à la présence d’une
dégradation liée à des phénomènes de fatigue et de corrosion sous contrainte à cause d’un
montage non conforme des équipements (mise sous contraintes excessives). Compte tenu de
ces résultats, le SIR a préconisé la dépose d’un compensateur voisin, lors d’un arrêt
programmé, dont l’expertise (cf. figure 4) a également mis en évidence des fissures de
fatigue qui à terme auraient également conduit à la rupture. La préconisation du SIR a permis
d’éviter une indisponibilité de la tranche de l’ordre de trois semaines.
INDEX çç Sommaire
Cas du compensateur de dilatation du 2 GSS 001 ZZ
Le SIR a demandé lors de l’AT 2 de 2007 le remplacement et l’expertise du compensateur
GSS voisin de celui rompu en juillet 05 sur le 2 GSS 001 ZZ
l’expertise a montré la présence de fissures de fatigue
La stabilité des fissures sous l’effet d’un transitoire de pression n’était pas garantie…
R
Peut être un accident et un arrêt fortuit évité
1 CNPE St Laurent des Eaux - Revue de direction du Service Inspection - Ch.Duquenoy le 26/11/07
Sur les installations secondaires (ou conventionnelles) des centrales nucléaires, un risque
important d’endommagement est la corrosion–érosion. Aussi les mesures d’épaisseur des
zones sensibles définies dans les plans d’inspection ont-elles mis en évidence des pertes
d’épaisseur non suspectées jusqu’à présent. Sans ces contrôles, des fuites ou des ruptures
auraient pu survenir en cours de cycle de fonctionnement pouvant entraîner des
indisponibilités importantes pour la réparation des tuyauteries. Sur plusieurs sites nucléaires
des sous épaisseurs ont été détectées sur les tubes de départs des lignes de purges
d’équipements sécheur-surchauffeur de vapeur ou encore sur des lignes de sous-tirage de
vapeur alimentant des réchauffeurs d’eau (cf. figure 5). Dans tous les cas, ces éléments ont été
remplacés lors d’un arrêt programmé des tranches.
INDEX çç Sommaire
Enfin, on peut citer des gestes beaucoup plus courants qui forment à eux seuls un réel renfort
vis-à-vis de la maîtrise du risque pression ; à savoir :
- la tenue à jour de la liste des ESP, y compris les extincteurs, ARI… et les équipements
en location (tous les SIR peuvent citer des ESP qui avaient été « oubliés ») ;
- la surveillance en fonctionnement (paramètres STE) ;
- la mise en place de rondes thématiques ESP et des analyses d’incidents (type coup de
bélier, dépassement température…) par l’Exploitant ;
- la mise sous surveillance volontaire d’équipements non soumis à la réglementation
mais dont le REX le justifie (ex : des tuyauteries de vidange de 40 mm de diamètre du
site thermique de Martigues).
- Et enfin, l’apport de la qualité et de la rigueur de la démarche (conformément à la
circulaire DM-TP 32510) ;
Les organisations des SIR de EDF des parcs thermiques et nucléaires sont semblables à la
différence près des effectifs (effet de taille). Un principe majeur est que les SIR ne prennent
pas en charge les activités d’ingénierie relatives aux ESP. Ils doivent se limiter à leur mission
première, celle des inspections et du contrôle. En conséquence il y a beaucoup d’interfaces
avec les services opérationnels. Cette organisation fait ses preuves d’efficacité mais elle
nécessite de « sensibiliser » régulièrement les équipes opérationnelles ; ce que font
régulièrement les SIR.
Si les SIR sont organisés avec des effectifs relativement réduits, cela a nécessité la mise en
place d’astreintes communes entre sites voisins ou de protocoles sur des plaques
géographiques permettent d’assurer une permanence de compétences d’inspecteurs à tout
moment.
En outre, pour certaines centrales thermiques les ressources et les charges de travail ne
justifient pas la présence permanente d’un responsable. Ainsi sur plusieurs sites
géographiquement proches, il a été possible de « mutualiser » la responsabilité des SIR (un
responsable unique pour 2 SIR). Le bilan est positif car chaque site dispose d’un inspecteur de
niveau 2 également habilité suppléant responsable capable d’assurer les activités nécessaires
en cas d’urgence. Toutes les missions des SIR peuvent être assumées y compris en cas
d’évènements nécessitant des interventions rapides.
Quant aux appuis des services centraux tels que définis ci-dessus. Leur rôle a été conforté et
des adaptations mineures ont été apportées.
INDEX çç Sommaire
Compétence, indépendance et l’autorité
Concernant l’indépendance et l’autorité des SIR, un retour d’expérience objectif peut être
fait sur la base de l’ensemble des constats formulés lors des audits internes EDF mais aussi
lors des audits de reconnaissance ou de renouvellement des reconnaissances des SIR par
l’Administration (DRIRE ou DREAL).
Un bilan exhaustif des constats DRIRE a été dressé, par exemple, sur les années 2008 et 2009
pour le parc thermique (présenté à l’Administration fin 2009). On a pu montrer qu’aucun
constat ne mettait en évidence le manque d’indépendance des SIR. Le rattachement sur
l’organigramme des SIR au chef d’établissement est certes un élément nécessaire mais aurait
pu ne pas être suffisant.
Cette indépendance doit rester un objectif permanent ; il n’est jamais définitivement acquis.
C’est pourquoi, les services centraux veillent à ce que l’indépendance des SIR soit un objectif
partagé par les chefs d’établissement. Des démarches sont réalisées par les pilotes nationaux
auprès des Directions de site pour rappeler le rôle des différents acteurs. Ces messages
peuvent aussi être portés lors de rencontres nationales de chefs d’établissement. Enfin, les
réunions des réseaux des responsables des SIR, organisés trois fois par an environ à DPN
comme à DPIT, prévoient systématiquement une analyse des évènements majeurs et du
comportement des SIR. L’ensemble de ces démarches contribue vraisemblablement au
maintien de l’indépendance et de l’autorité des SIR.
A ce jour, l’autorité des SIR apparaît pour les Directions des sites beaucoup plus comme un
appui réel, grâce à cette indépendance de jugement, que comme une contrainte. Ce qui était
une crainte pour certains lors de la mise en place des SIR.
La gestion des compétences est un élément majeur indispensable car il s’agit de compétences
« rares » en interne comme à l’externe d’EDF. Les compétences exigées des inspecteurs sont
vastes (métallurgie, mécanique, technologie, CND, maintenance, process, codes et normes,
réglementation…) et nécessitent des parcours professionnels spécifiques.
C’est pourquoi, afin d’assurer le maintien à niveau des compétences des SIR dans la durée,
chaque site doit assurer sa propre Gestion Prévisionnelle des Effectifs et des Compétences
(GPEC) avec l’appui, si besoin, des services centraux. Le retour d’expérience sur ces dix
années montre qu’une bonne anticipation des mouvements et la mise en place de
« pépinières » par chaque site (qui peuvent être communes à plusieurs métiers voisins tels que
la maintenance) ont de façon générale permis d’anticiper suffisamment tôt le renouvellement
des inspecteurs.
La formation de l’Ecole des Mines de Douai, mise à disposition des SIR depuis 2007, est une
formation en alternance de 13 semaines qui fait preuve d’efficacité. Ce cursus permet une
formation « accélérée » sur l’ensemble des domaines couverts par les inspecteurs. Cette
formule est surtout bien adaptée à des personnes qui viennent de services autres que ceux de
la maintenance. Ce cursus de formation sera encore développé en recherchant des formations
modulables.
INDEX çç Sommaire
La vérification des compétences des inspecteurs et des responsables des SIR est confiée à la
CNRC dont la mission est de mettre en place des jurys de qualification capables d’une analyse
objective et homogène (entre eux). Des évaluations sur ces questions sont régulièrement
effectuées. A ce jour aucun dysfonctionnement n’a été mis en évidence. La Visite de
Surveillance effectuée par l’Administration en 2009 confirme cette analyse.
Ainsi, il apparaît que le système mis en place de qualification des inspecteurs et des
responsables sur la base d’un dossier de compétence validé par la présence du candidat lors
d’un jury est efficace et pérenne.
Conformément à la réglementation décrite dans le guide professionnel d’EDF, cité plus haut,
il est prévu de se ré-interroger régulièrement sur la nécessité de réviser ce guide et, en
particulier, sur la méthode d’élaboration des plans d’inspection. Hormis quelques aspects
réglementaires, concernant la méthode décrite ici, il apparaît que la méthode présente une
robustesse suffisante et ne nécessite pas de modification dans ses principes.
Lors de réunions annuelles de bilans des inspections, organisées au niveau national, des zones
sensibles ont été modifiées ou ajoutées dans les listes antérieures. Ces révisions se font donc
systématiquement.
5 – CONCLUSION - PERSPECTIVES
Compte tenu de cette expérience, les SIR ont maintenant pris toute leur place au sein des
établissements de production d’électricité d’EDF. Ils répondent effectivement aux enjeux
attendus.
A ce jour, l’ensemble des sites de production dispose de SIR. Des adaptations des
organisations ont permis d’optimiser les ressources. Le maintien des compétences reste un
objectif permanent car la multiplicité des domaines de compétence attendue des inspecteurs,
rend ce métier motivant mais nécessite une longue expérience.
L’efficacité du système qualité mis en place par l’organisation des SIR est démontrée. C’est
ainsi que cette organisation a pu être étendue, sur un site thermique, aux autres domaines
comme la sécurité (levage – manutention, incendie, ATEX…) qui font l’objet de
réglementation. Le responsable SIR porte la responsabilité de l’ensemble de ces activités ; un
bilan sera dressé sur cette expérience a priori très positive.
Enfin, pour conforter l’autorité des SIR, le domaine d’expertise des inspecteurs peut être
étendu. Ainsi les premiers SIR qui ont reconduit leur reconnaissance après les trois premières
années d’expérience ont demandé et obtenu la possibilité d’effectuer par eux-mêmes les
requalifications des tuyauteries. Dans cette même logique les délégations ont été demandées
INDEX çç Sommaire
et obtenues pour quelques sites afin que les inspecteurs deviennent « experts » pour les
contrôles après interventions notables. Cette autorisation apporte en outre des gains
importants en terme de disponibilité pour l’exploitation5. Elle sera visée par tous les sites
nucléaires et thermiques.
5
Cette autorisation nécessite les compétences et le contrôle de l’Organe Inspection Utilisateur (OIU) d’EDF.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page114
RÉSUMÉ / ABSTRACT
85D
INSPECTION DES SOUDURES EN J DES PÉNÉTRATIONS DE FOND DE
CUVE DES RÉACTEURS À EAU PRESSURISÉE
INSPECTION OF PWR BOTTOM MOUNTED INSTRUMENTATION (BMI) J-WELD
O. Burat – [email protected] - M. Piriou – [email protected]
Y. Kernin – [email protected] - M. Debroise – [email protected]
AREVA, INTERCONTROLE – NETEC, 4, rue T. Dumorey, 71109 Chalon sur Saône
S. W. Glass - [email protected] - AREVA, NETEC, 10 rue J. Récamier, 69456 Lyon – France
O. Casula – [email protected] - CEA Saclay, LIST, Bâtiment 611, 91191 Gif sur Yvette – France
RÉSUMÉ
AREVA réalise des inspections des Pénétrations de Fond de Cuve (PFC) des Réacteurs à Eau sous Pression, depuis plus de 15 ans.
Des démonstrations ont été réalisées par l’EPRI en 2004 et une qualification RSEM a été obtenue en 2005. Cette inspection par
ultrasons permet de détecter et caractériser des indications de fissures dans le métal de base et des indications à l’interface avec la
soudure. Un examen complémentaire de la soudure par contrôle visuel ou par Courants de Foucault peut être envisagé, mais les
contrôles testés jusqu’alors manque de fiabilité, fiables pouvant rater des indications ou au contraire détecter de fausses inclusions ou
soufflures pouvant conduire à des réparations injustifiées.
Mercredi
ABSTRACT
Reactor Pressure Vessel (RPV) Bottom Mounted Instrumentation (BMI) primary nozzle ultrasonic (UT) inspections have been per-
formed by AREVA for more than 15 years. EPRI demonstrations were performed in 2004 and an RSEM qualification was obtained
in 2005, This UT inspection detects and characterizes cracks in the base metal of the penetration and its interface with the weld. A
complementary Visual Examination of the weld or an Eddy Current examination may also be performed but these tests are understood
to be marginal with a possibility to miss indications or to mistake inclusions or weld voids for in-service cracks and therefore could
force an unnecessary repair.
Difficulties associated with examination of this weld include:
¢
Unpredictable ultrasonic behaviour in the weld material
¢
wavy and complex OD weld surface profile
¢
Restrictive accessibility around the weld surface
AREVA’s Non Destructive Examination Technical Center (NETEC), has recently developed several techniques to improve surface
and volumetric inspection of this weld including:
¢
A TOFD probe to detect and characterize cracks within the weld bulk from the simpler ID surface of the nozzle.
¢
A flexible UT Phased Array probe to accommodate the wavy and complex weld surface, in partnership with CEA,
¢
Evaluation of flexible ET multi-coils and reduced surface ET probes for improved surface inspection, in partnership with CEA,
¢
Evaluation of Infrared Photo-Thermal Camera examination, from the weld surface.
This paper discusses the design and the performances of these specific probes and NDT techniques, and their possible applications
to other similar components.
Keywords : RPV, Bottom Mounted Instrumentation, Inspection, UT, ET, Infra-red Photo Thermal Camera
INDEX çç Sommaire
INSPECTION DES SOUDURES EN J DES PENETRATIONS DE
FOND DE CUVE DES REACTEURS A EAU PRESSURISEE
O. Burat – [email protected]
M. Piriou– [email protected]
Y. Kernin – [email protected]
M. Debroise – [email protected]
AREVA, INTERCONTROLE – NETEC, 4, rue T. Dumorey, 71109 Chalon sur Saône
S. W. Glass- [email protected]
AREVA, NETEC, 10 rue J. Récamier, 69456 Lyon – France
O. Casula – [email protected]
CEA Saclay, LIST, Bâtiment 611, 91191 Gif sur Yvette – France
RESUME
AREVA réalise des inspections des Pénétrations de Fond de Cuve (PFC) des Réacteurs à Eau
sous Pression, depuis plus de 15 ans. Des démonstrations ont été réalisées par l’EPRI en 2004
et une qualification RSEM a été obtenue en 2005. Cette inspection par ultrasons permet de
détecter et caractériser des indications de fissures dans le métal de base et des indications à
l’interface avec la soudure. Un examen complémentaire de la soudure par contrôle visuel ou
par Courants de Foucault peut être envisagé, mais les contrôles testés jusqu’alors manque de
fiabilité, fiables pouvant rater des indications ou au contraire détecter de fausses inclusions ou
soufflures pouvant conduire à des réparations injustifiées.
Les difficultés associées à l’examen de cette zone de soudure sont :
• Un comportement imprévisible des ultrasons propagés dans la soudure,
• Un profil complexe et irrégulier de la surface externe de la soudure,
• Une accessibilité restreinte à la surface de la soudure.
NETEC, le Centre Technique des Examens Non Destructifs d’AREVA, a récemment
développé plusieurs techniques en vue d’améliorer l’inspection de la surface et du volume de
cette soudure :
Une sonde TOFD pour détecter et caractériser les fissures dans le volume de la soudure
par accès interne à la pénétration,
Une sonde UT Phased Array conformable, en partenariat avec le CEA, pouvant
s’adapter à la surface externe complexe et évolutive de la soudure.
INDEX çç Sommaire
L’évaluation de sondes ET multi-bobines, conformables (en partenariat avec le CEA),
ainsi que des sondes de surface réduite, permettant d’améliorer le contrôle à partir de la
surface de la soudure,
L’évaluation de la Caméra Photo-Thermique à Infrarouge, de la surface de la soudure.
Ce papier présente la conception et les performances de ces techniques de contrôles spécifiques
et des sondes associées, ainsi que leur possible application à des composants similaires.
ABSTRACT
Reactor Pressure Vessel (RPV) Bottom Mounted Instrumentation (BMI) primary nozzle
ultrasonic (UT) inspections have been performed by AREVA for more than 15 years. EPRI
demonstrations were performed in 2004 and an RSEM qualification was obtained in 2005,
This UT inspection detects and characterizes cracks in the base metal of the penetration and its
interface with the weld. A complementary Visual Examination of the weld or an Eddy Current
examination may also be performed but these tests are understood to be marginal with a
possibility to miss indications or to mistake inclusions or weld voids for in-service cracks and
therefore could force an unnecessary repair.
Difficulties associated with examination of this weld include:
• Unpredictable ultrasonic behaviour in the weld material
• wavy and complex OD weld surface profile
• Restrictive accessibility around the weld surface
AREVA’s Non Destructive Examination Technical Center (NETEC), has recently developed
several techniques to improve surface and volumetric inspection of this weld including:
A TOFD probe to detect and characterize cracks within the weld bulk from the simpler
ID surface of the nozzle.
A flexible UT Phased Array probe to accommodate the wavy and complex weld surface,
in partnership with CEA,
Evaluation of flexible ET multi-coils and reduced surface ET probes for improved
surface inspection, in partnership with CEA,
Evaluation of Infrared Photo-Thermal Camera examination, from the weld surface.
This paper discusses the design and the performances of these specific probes and NDT
techniques, and their possible applications to other similar components.
Keywords : RPV, Bottom Mounted Instrumentation, Inspection, UT, ET, Infra-red Photo Thermal
Camera
BACKGROUND
The Bottom Mounted Instrumentation (BMI) Nozzles (see figure 1) of most pressure Water
Reactors in the US and western Europe are made from Inconel 600 tubing material and are
installed with an alloy 82/182 J-Weld. Since the eighties, these materials are known to be
susceptible to stress-corrosion-cracking (SCC). An inspection program of the nozzle and weld
interface has been underway in France for more than 15 years. An RSEM qualification was
INDEX çç Sommaire
successfully performed in France in 2005 and EPRI demonstrations were performed in 2004.
This UT volumetric inspection to date however only addresses the base material of the nozzle
and the interface between the tube and the weld material but has not been successfully
extended to the full volume of the J-Groove weld.
Crack initiation is equally likely to be in the J-weld itself and a destructive analysis of one of
the leaking BMI Nozzles attributed crack initiation to a weld defect.
Within this frame of the state of the industry’s capabilities, AREVA decided to develop and
evaluate the performances of several innovative techniques for inspection of the BMI weld.
The design and the performances of these specific probes and NDE associated techniques are
discussed in this paper.
INDEX çç Sommaire
qualified in France, Belgium, and the US for the base nozzle material examination. This probe
(figure 2) has an axial UT TOFD beam with the pitch-catch focal cross-point near the base-
nozzle OD. For the weld examination, AREVA’s initial effort was simply to extend this focal
depth into the weld.
INDEX çç Sommaire
Nevertheless, improvement is necessary to permit clear detection of all reflectors, particularly
those at 90 degrees to the upper and lower weld axis. Anticipated improvements currently
planned include angle beam adaptation and frequency tuning.
FLEXIBLE UT PHASED ARRAY PROBE EVALUATION FROM THE WELD
SURFACE
A second UT approach to inspect the weld from its surface external to the nozzle uses a
flexible UT Phased Array. This probe, developed by CEA (Commissariat à l’Energie
Atomique) is designed to compensate the wavy and complex weld surface. This probe was
evaluated to detect reflectors of the BMI mock-up shown in figure 4. The probe consists of
small independent piezo-composite rectangular elements melded in a polymer matrix. At each
pulse during the acquisition, the delay laws are adapted to the target focal point taking the
known position of each active element on the complex weld surface into account. The UT array
is used in a TOFD configuration.
Various Eddy current approaches may be considered for this inspection challenge. ET must
have close surface contact and can only hope to find surface breaking or near-surface cracks.
Commercially available surface probes for this measurement include the ZETEC +Point
CLIP
+Point
Page 5/10
Figure 7: ET surface probes and BMI J-weld mock-up used for their evaluation
INDEX çç Sommaire
pencil probe and a ZETEC flexible array “mouse” probe. AREVA has also developed a CLIP
probe (2) that seemed suitable to apply for the BMI J-weld representative mockup.
Except for restrictive dimensions of the probes housings which prevented access to all EDM
notches, the acquisition results show that the CLIP probe:
• Could detect all EDM notches (except where geometry precluded good contact)
• Had high circumferential and radial resolution,
• Could detect simultaneously circumferential and radial cracks because of its two 90°
separated receivers,
• Should offer a better discrimination between cracks and crevices or rough weld
profiles.
To take advantage of the high circumferential and radial resolution of the CLIP probe, the
inspection mesh must be very tight to fully characterize surface cracks in the J-weld geometry.
micro-bobbin
coils
Carrier in
silicon for
flexibility
Amplifiers
ET flexible array probe approach has been also considered to detect surface defects. Array
probes offer the advantage of single axis scanning, thus simplifying and shortening the
inspection time. Tests have been performed on BMI J-weld mockup used before with two
kinds of array probes.
The first one (figure 8) was developed by the CEA (Comissariat à l’Energie Atomique) for the
detection of surface and sub-surface cracks, for the aeronautical industry. The probe consists of
micro-bobbin coils etched on a Kapton flexible thin film. Coils are arranged in 4 staggered
rows of 8 elements. Each element is composed of one transmitting coil and one receiving coil
close together in order to detect short cracks ( 0.4 mm long).
The other probe (figure 9 – see next page) is marketed by ZETEC specifically for BMI
(Bottom Mounted Instrumentation) and CRDM (Control Rod Drive Mechanism) weld
inspection. It is made up of 2 rows of 12 +Point coils. Each element could be used in
impedance mode (for 0° & 90° oriented cracks detection), T/R mode (for 45° & 135° oriented
cracks detection) and absolute mode (for lift-off visualization). +Point coils are placed in a gel
filled pouch shaped to fit the minimum curvature radius found in J-weld geometry
(approximately 10 mm).
INDEX çç Sommaire
Tests results (figures 10a and 10b) showed that although not developed specifically for
BMI/CRDM J-weld inspection, the CEA flexible array probe gave the better laboratory results.
All notches are detected with large SNR (Signal to Noise Ratio) whatever the orientation and
the localization of the notch. Furthermore, CEA technology offers a higher resolution (see C-
scan below) and good compliance to fit the geometry. AREVA and CEA believe an improved
CEA concept with a specific probe design for J-weld application can be developed.
Coils
5 mm long radial
notch
carbon
weld
steel
MANIPULATOR CONCEPT
During AREVA’s South Texas campaign to fully characterize the leaking nozzles, a delivery
robot was quickly developed and deployed with the ZETEC flexible array probe. Building on
this experience, a new manipulator concept has been conceived for either flexible or pencil, UT
or ET contact sensors.
The sensor is located in a cylindrical housing. This cylindrical housing allows the probe to scan
the full geometry range from the smallest to largest radius of curvature with good control of the
sensor’s planer position. This inspection head is mounted on a crown wheel centered on the
BMI tube. A motion controlled motor initiates the index axis in the circumferential direction.
INDEX çç Sommaire
INFRA-RED PHOTO THERMAL CAMERA EXAMINATION
Fiber optic
cable Crack
Displacement
Laser line
EXCITATION Two-pole signal from a crack
Displacement
The creation of a transverse heat flux in the material allows imperfections to be detected via the
thermal barrier that they induce. When the excitation/detection (E/D) beams approaches a
surface defect or underlying defect, the IR signal increases because heat diffusion ceases in the
direction of the defect. When the E/D system passes over the imperfection, the detected signal
fades because the imperfection retards the laser-generated heat spreading through the detection
zone. Continuous measurement of the IR signal during the scanning and the reconstitution of
the information in a 2D representation enable a full image of the examination surface where the
imperfections are displayed.
INDEX çç Sommaire
Good detection of the EDM notches within the mock-up J-welds were realized (Figure 12)
independent of the orientation, however field implementation is recognized to be a challenge as
the IR method is not suitable for use through water.
Figure 12 : Outer BMI mock-up with EDM notches in J-Weld (L 5mm, D 2mm, W 0.1mm)
oburat 5/5/10 16:33
Mis en forme: Police :Times Roman, 12
CONCLUSIONS AND PERSPECTIVES oburat 10/3/10 10:36
Supprimé: 13
Inspection of the primary BMI nozzle including the tube-weld interface is well managed via oburat 10/3/10 9:43
various ID UT techniques. Although extending these techniques to the weld examination are Supprimé:
desirable, success for 100% coverage weld examination from tube ID can be questioned. OD
techniques using either UT or ET are generally effective however the extreme geometries of
the weld must be managed with very flexible sensors and a flexible manipulator that can
deliver the sensors in a well controlled manner. This is a difficult examination and a complete
industrial solution is currently unavailable.
The industry is not currently demanding this pressure-boundary weld examination. Utilities
and regulators are satisfied that the chances for a serious safety incident are low and the current
program of nozzle examinations and bare-metal examinations provide adequate protection. If
another leak is discovered, the industry’s position will most likely change. Weld examination
will be a particular focus if subsequent leaks are determined to initiate from the weld as in the
South Texas case. AREVA’s preparations will enable a rapid response when or if the industry
starts to demand this examination.
REFERENCES
1. Meier,R., Fischer,E., Burat,O., & Glass, S.W., Feasibility Assessment and Development
of Phased Array BMI Nozzle and J-Weld Inspection; 5th International Conference on
INDEX çç Sommaire
NDE in Relation to Structural Integrity for Nuclear and Pressurized Components, San
Diego Ca, May 2006.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page115
RÉSUMÉ / ABSTRACT
68D
CONTRÔLE EN FABRICATION DES SOUDURES DE TUYAUTERIES :
PRÉSENTATION DU GUIDE DES ALTERNATIVES À LA GAMMAGRAPHIE À
L’IR 192
RÉSUMÉ
Face à la nécessité prochaine de justifier l’emploi de la gammagraphie à l’iridium 192, les industriels
doivent dès à présent se préparer à la mise en oeuvre de techniques alternatives de contrôle des soudures
de leurs tuyauteries. L’Institut de Soudure, fort de sa longue expérience dans le domaine des contrôles
non destructifs, a analysé les stratégies possibles de substitution, déterminer les techniques candidates,
leurs domaines d’application et leur efficacité, dans le cadre de l’étude associative Alter-X regroupant
plusieurs grands industriels ( EDF, GRTgaz,TOTAL, TECHNIP).
Le guide a été réalisé à partir des résultats de l’étude Alter-X achevée en 2008, et est bâti autour d’essais
concrets. Il constitue un outil précieux pour les industriels qui souhaitent anticiper un changement
technologique qui s’annonce délicat.
Il est constitué :
¢ d’une recommandation générale qui aide à définir la meilleure stratégie de substitution,
¢ de 3 guides de mise en œuvre pour les techniques : TOFD, ultrasons multi-éléments
et radionumérique.
ABSTRACT
Faced with the needs to justify the near future use of Iridium 192 gammagraphy, the various industrial sectors
concerned must as from now prepare themselves to set up alternative techniques for testing their piping welds.
Institut de Soudure, confident of their long experience in the field of NDT, analysed the possible alternative
strategies, identifying candidate techniques, their applications and their efficiency in the frame of a joint industry
project called ALTER-X sponsored by: EDF, GRTgaz, Total and Technip.
From the results of this study completed in 2008, a guideline was issued taking into account trials carried out
on representative industrial mock-ups. It is a very valuable tool for all parties wishing to anticipate a coming
technological change, involving many difficulties.
It consists of:
¢ a general recommendation helping to define the best strategy for replacing Ir 192 gammagraphy,
¢ 3 recommended practices to apply on piping welds : TOFD, phased array and ionising radiation by means
of CR & radioscopic equipment.
INDEX çç Sommaire
CONTROLE EN FABRICATION DES SOUDURES DE TUYAUTERIES :
PRESENTATION DU GUIDE DES ALTERNATIVES A LA
GAMMAGRAPHIE A L’IR 192
RÉSUMÉ
Le guide a été réalisé à partir des résultats de l’étude Alter-X achevée en 2008, et est bâti
autour d’essais concrets. Il constitue un outil précieux pour les industriels qui souhaitent
anticiper un changement technologique qui s’annonce délicat.
Il est constitué :
- d’une recommandation générale qui aide à définir la meilleure stratégie de
substitution,
- de 3 guides de mise en œuvre pour les techniques : TOFD, ultrasons multi-éléments et
radionumérique.
ABSTRACT
Faced with the needs to justify the near future use of Iridium 192 gammagraphy, the various
industrial sectors concerned must as from now prepare themselves to set up alternative
techniques for testing their piping welds.
Institut de Soudure, confident of their long experience in the field of NDT, analysed the
possible alternative strategies, identifying candidate techniques, their applications and their
INDEX çç Sommaire
efficiency in the frame of a joint industry project called ALTER-X sponsored by: EDF,
GRTgaz, Total and Technip.
From the results of this study completed in 2008, a guideline was issued taking into account
trials carried out on representative industrial mock-ups. It is a very valuable tool for all
parties wishing to anticipate a coming technological change, involving many difficulties.
It consists of:
- a general recommendation helping to define the best strategy for replacing Ir 192
gammagraphy,
- 3 recommended practices to apply on piping welds : TOFD, phased array and ionising
radiation by means of CR & radioscopic equipment.
INTRODUCTION
Il n’est pas possible de remplacer la gammagraphie à l’Iridium par une seule technique CND.
Cependant des solutions alternatives sont envisageables dans certains cas. Leur mise en œuvre
est toutefois tributaire du secteur industriel, du type de composant à inspecter, de la nature des
matériaux, du type de soudure (bout à bout, piquage …), de l’orientation et de la position des
défauts à détecter et de l’environnement dans lequel se trouve le composant à inspecter.
Quelles sont donc les alternatives possibles à l’Iridium 192 pour le contrôle des soudures de
tuyauterie ?
INDEX çç Sommaire
• Réaliser l’état de l'art des moyens de contrôle non destructif applicables en
particulier aux soudures de tuyauteries avec leurs avantages / inconvénients et
limites.
• Évaluer les possibilités d'application des techniques les plus prometteuses comme
par exemple : la radiographie avec tubes à rayons X de dernière génération, la
radiographie numérique avec la dernière génération de plaque phosphore, écran C-
MOS ou silicium amorphe, avec l’emploi du sélénium 75, TOFD, « phased array »,
traducteurs ultrasonores spéciaux... Le choix des techniques à expérimenter a été
effectué d’un commun accord avec les participants, à l’issue de l’étude
bibliographique et des essais exploratoires.
• Définir les domaines d'application et les limites de ces techniques en fonction des
caractéristiques géométriques du composant à inspecter : diamètre, épaisseur,
proximité de la soudure dans le cas d’un té ou d’un coude ou de variation
géométrique (désalignement, délardage …). L’évaluation des techniques a été
réalisée au moyen de contrôles réalisés sur maquettes et d’un nombre limité
d’essais sur site.
• Rédiger des recommandations de contrôle pour les techniques CND les plus
prometteuses.
L’état de l’art a été réalisé de façon aussi exhaustive que possible et a porté sur toutes les
méthodes : rayonnements ionisants, méthodes électromagnétiques, méthodes ultrasonores.
Les résultats d’études telles que l’étude des performances comparées des ultrasons
automatiques (AUT), et de la radiographie pour le contrôle des soudures de pipe de
l’Université de Gand ainsi que les études KINT et TOFDPROOF concernant notamment les
probabilités de détection comparées du TOFD et des autres techniques CND
conventionnelles, ont été inclus dans cette revue.
L’état de l’art a permis de recenser 520 références dont 180 se sont révélées particulièrement
pertinentes et à partir desquelles un rapport de synthèse a été rédigé.
INDEX çç Sommaire
• Le niveau de qualification requis de l’opérateur
• Les domaines d’application (matériau, épaisseur, géométrie…)
• Les avantages et limitations
• Leur mise en œuvre
• Les performances industrielles par rapport à la radiographie Ir-192
A l’issue de l’état de l’art, les techniques jugées les plus intéressantes dans le cadre du projet
Alter-X pour constituer une alternative à la gammagraphie Ir-192 sont les suivantes : le Time
Of Flight Diffraction (TOFD), les ultrasons avec multiéléments, les techniques de
radionumérique exploitant les plaques images au phosphore et les techniques de radioscopie.
Des sources de rayonnement alternatives à l’Ir 192 (Se 75, RX) ont également été retenues
pour l’évaluation.
Les figures 1 et 2 illustrent les bonnes qualités d’image susceptibles d’être obtenues en
radionumérique ou en radioscopie pour le contrôle de soudures.
Figure 1 :
Radionumérique
plaques
phosphores - acier
carbone épaisseur
4 mm (avec RX
100 kV)
Figure 2 :
Radionumérique
plaques
phosphores - acier
carbone épaisseur
4 mm (avec RX
100 kV)
INDEX çç Sommaire
PRESENTATION DU GUIDE DES ALTERNATIVES
A LA GAMMAGRAPHIE IR 192
Généralités
Le guide, édité sur 2 colonnes, comporte 134 pages et de nombreuses illustrations. Il est
constitué d’une recommandation générale et de 3 recommandations concernant la mise en
œuvre des techniques alternatives.
Figure 3 :
Extrait d’un
diagramme
présentant les
avantages,
performances,
frein à l’emploi
du Se 75 comme
technique
alternative,
indications
technico
économique
Le lecteur est guidé dans ses choix à partir des éléments suivants :
• contexte réglementaire et normatif
• génération ou non d’une image
• accès et conditions de surface
• dégagement par rapport au bord du cordon de soudure
• distance intertuyauteries
• polyvalence vis-à-vis des couples diamètres/épaisseur à contrôler
• polyvalence vis-à-vis de la gamme de matériau à contrôler
• polyvalence vis-à-vis du type de géométrie des composants
INDEX çç Sommaire
• besoin de personnels qualifiés
• bénéfices en termes de radioprotection
• critères économiques
A titre d’exemple le tableau suivant illustre les capacités des techniques recensées vis-à-vis de
la géométrie.
tube/té Non sauf (1) Oui Possible (4) (5) Oui Oui Oui
Partiel car
tube/coude généralement Oui Possible (4) (5) Oui Oui Oui
impossible dans
intrados (2)
A valider au cas
tube/réduction par cas (technique Oui Possible 4) (5) Oui Oui Oui
spéciale) (2)
Piquage posé
droit Non Oui (8) Possible(6) Oui Possible (10) Possible (11)
Piquage posé
pénétrant Non Oui (8) Possible (6) Oui Possible (10) Possible (11)
Piquage posé
oblique Non Oui (8) Partiel (7) Oui (9) Possible (10) Possible (11)
Piquage
pénétrant oblique Non Oui (8) Partiel (7) Oui (9) Possible (10) Possible (11)
INDEX çç Sommaire
• Recommandations pour le choix des paramètres d’examen et conditions
d’interprétation
• Préparation avant contrôle
• Méthodologie pour définition des offsets scans
• Illustration des problèmes d’accessibilité
• Images types et radiogramme correspondant
• Présentation des principaux critères d’acceptation des codes et normes
La figure 5 illustre une difficulté de mise en œuvre susceptible de survenir en présence d’une
réduction.
Figure 5 :
Exemple de difficulté de
mise en œuvre
La figure suivante présente une comparaison des résultats obtenus sur une soudure de coude
comportant un manque de pénétration
INDEX çç Sommaire
Figure 6 : Figure 7 : Figure 8 :
Schéma d’une soudure Image TOFD correspondante radiogramme correspondant
tube/coude
A titre d’exemple la figure 9 présente les résultats obtenus sur des soudures d’un élément de
« skid » 6’’/8’’
INDEX çç Sommaire
Figure 9 : Résultats obtenus en ultrasons multiéléments
La figure suivante présente le principe d’un bloc de qualification lorsque la soudure n’est
accessible que d’un seul côté. Ce bloc comporte notamment une entaille réalisée avec l’angle
le plus défavorable par rapport à la préparation de la soudure.
INDEX çç Sommaire
Figure 10 : Bloc de qualification – cas d’une soudure accessible d’un seul côté
INDEX çç Sommaire
Pour obtenir des résultats reproductibles qui
permettent le contrôle de la stabilité à long
terme d’un système, il est nécessaire de
définir une disposition type pour effectuer les
mesurages. Elle doit être formalisée par écrit
de façon détaillée afin de pouvoir être
correctement reproduite par la suite. Ce
document doit contenir les données relatives
à la source de rayonnement X ou gamma
(type de tube ou de source, tension, courant,
filtration, …) ainsi que les collimateurs
éventuels. Les indicateurs d’essais doivent si
possible être placés côté source sur l’élément
contrôlé et selon les recommandations de la
figure ci-contre
Les essais doivent être effectués sans rien
changer au matériel ni aux réglages utilisés
dans des conditions normales de
fonctionnement.
En l’absence de classement d’un système
numérique, on peut également s’appuyer sur Figure 12 : Positionnement recommandé des
une telle disposition pour évaluer les IQI
capacités d’un système
CONCLUSIONS
Les techniques de CND alternatives expérimentées dans le cadre du projet, ont été
sélectionnées pour leur capacité à se substituer à la gammagraphie à l’Iridium 192, leur
maturité technique et leur possibilité d’industrialisation. La gammagraphie à l’Iridium 192 ne
peut être remplacée par une seule technique. Il peut être nécessaire d’associer différentes
techniques de CND entre elles. L’association optimale doit cependant être effectuée secteur
par secteur et pour une application donnée.
INDEX çç Sommaire
- l’emploi du sélénium 75 associé à des équipements facilitant la gestion de la
radioprotection tel les atténuateurs comme le γ−Prox, du fait de sa polyvalence et de
l’existence de personnels qualifiés,
- l’emploi de solutions mixtes exploitant majoritairement le TOFD et/ou des ultrasons
multiéléments pour le contrôle des soudures bout à bout et limitant l’emploi des
techniques RT aux soudures non contrôlables par imagerie ultrasonore. A moyen terme,
on peut cependant espérer que les progrès des technologies ultrasonores et
l’augmentation du nombre de personnels qualifiés permettra de réduire la part
d’utilisation des RT dans une fabrication.
Les avantages et inconvénients de ces 2 solutions sont résumés dans le tableau suivant.
REFERENCES
INDEX çç Sommaire
[3] RD12. RT21 – Projet ALTER-X – Recommandation concernant l’application de la
méthode TOFD pour le contrôle en fabrication des soudures de tuyauteries en acier au
carbone
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page116
RÉSUMÉ / ABSTRACT
7/8D
EXEMPLE D’ÉVALUATION DE CONFORMITÉ MENÉE PAR L’OIU D’EDF
POUR L’EPR FLAMANVILLE 3
RÉSUMÉ
La directive européenne DESP97/23/CE déclinée en droit français par le Décret 99-1046 et l’Arrêté
Ministériel du 12/12/05 (ESPN) prévoit que les équipements sous pression et les ensembles dans les-
quels ces équipements sont inclus fassent l’objet d’une évaluation de conformité avant leur mise sur le
Mercredi
marché. La plupart de ces évaluations sont réalisées par des organismes notifiés à la Commission Eu-
ropéenne et par dérogation et sous certaines conditions par des organes d’inspections des utilisateurs
(OIU)
L’évaluation de conformité d’un ensemble est une évaluation globale qui comprend différentes étapes
: évaluation de chacun des équipements constituant l’ensemble, évaluation de l’intégration de ces équi-
pements entre eux et évaluation de la protection de l’ensemble contre le dépassement des limites ad-
missibles. Dans le cadre de la construction de l’EPR de Flamanville 3, l’OIU (Organe d’Inspection des
Utilisateurs) d’EDF évalue la conformité des plusieurs équipements et ensembles conventionnels et
nucléaires. A travers un exemple d’évaluation de tuyauteries et d’un ensemble nucléaire, l’OIU d’EDF
présente son organisation pour vérifier sa conformité à la réglementation des ESPN.
ABSTRACT
The DESP 97/23/EC has been converted in France by the French Decree 99-1046 and the Ministerial
order of December 12th 2005 (ESPN). This regulation specifies that a conformity assessment has to
be made for pressure equipments and assemblies containing these equipments before their placing of
the market. Most of these conformity assessments are performed by notified bodies and by way of de-
rogation, and under specific conditions by User Inspectorates.
The conformity assessment of an assembly is a global process which includes several steps: the as-
sessment of each item of pressure equipment making up the assembly, the assessment of the integration
of the various components of the assembly and the assessment of the protection of an assembly against
exceeding the permissible operating limits. In the frame of the EPR new project of Flamanville 3, EDF
User Inspectorates (OIU) is responsible for the conformity assessment of several nuclear and non nu-
clear equipments and assemblies. Through one example, EDF User Inspectorates present its organi-
zation to perform the assessment conformity of the assembly.
INDEX çç Sommaire
INTRODUCTION
La Directive Européenne DESP97/23/CE [1] déclinée en droit français par le Décret 99-1046
[2] et l’Arrêté Ministériel du 12/12/05 (ESPN) prévoit que les équipements sous pression et
les ensembles dans lesquels ces équipements sont inclus fassent l’objet d’une évaluation de
conformité avant leur mise sur le marché. La plupart de ces évaluations sont réalisées par des
organismes notifiés à la Commission Européenne et par dérogation et sous certaines
conditions par des organes d’inspections des utilisateurs (OIU).
Après un rappel de différentes définitions issues de la réglementation européennes
(ensembles, installations, OIU…), l’objet de cet article est de présenter dans un premier
temps le découpage en ensembles et installations pour le projet de construction neuve EPR
Flamanville 3 et dans un deuxième temps, l’organisation mise en place par l’OIU d’EDF pour
réaliser l’évaluation de conformité d’un ensemble nucléaire.
GENERALITES - DEFINITIONS
Dans le contexte réglementaire français relatif à la mise sur le marché des équipements sous
pression (Décret 99-1046 et Arrêté ESPN du 12/12/05), le groupe EDF a choisi de se munir
d’un organe d’inspection des utilisateurs (OIU). Cet organisme peut réaliser, au même titre
qu’un organisme notifié, l’évaluation de conformité d’équipements sous pression ou
d’ensembles. Cependant, certaines restrictions lui sont imposées :
- les équipements ou ensembles doivent être destinés à être montés sur des sites
appartenant au groupe EDF,
- les seuls modules d’évaluation utilisables sont A1, C1, F et G, (voir [3])
- le poinçon apposé sur les équipements est spécifique et déposé au Journal Officiel
(voir figure 1),
OIU
Figure 1 : poinçon OIU
L’OIU d’EDF fête en 2010 ses 10 années d’existence. Depuis son habilitation en 2000, et son
acceptation par l’administration nucléaire en 2006 pour réaliser des évaluations de conformité
d’équipements sous pression nucléaire selon l’arrêté ESPN du 12/12/05 [4]. Jusqu’à ce jour,
les évaluations de conformité ont essentiellement été menées selon le module G, qui consiste
à un examen de la conception, de la fabrication jusqu’à l’épreuve hydraulique. Les premières
évaluations concernaient des accessoires sous pression de type vannes, compensateurs de
dilatation. Par la suite, des récipients de type échangeurs, réchauffeurs ont été traités.
Aujourd’hui, l’OIU d’EDF intervient essentiellement sur la construction de la nouvelle
centrale nucléaire EPR de Flamanville 3, en évaluant des équipements individuels et des
ensembles. Environ 80 dossiers sont à ce jour en cours d’instruction.
INDEX çç Sommaire
Ensembles et installations
INDEX çç Sommaire
L’OIU assure l’évaluation de conformité de différents équipements qui seront installés dans le
bâtiment réacteur, dans les locaux des auxiliaires « Balance Nuclear Island » et dans la salle
des machines. Il réalisera également l’évaluation des ensembles suivants :
L’évaluation de conformité des ensembles est réalisée selon le module G. Elle consiste à
examiner la conception, à inspecter les opérations d’assemblage des ESP(N) entre eux et à
réaliser les vérifications finales (dossier réglementaire, examen des ensembles et épreuves
hydrauliques).
D’une façon générale, et bien entendu à l’exception des tuyauteries, tous les ESP(N)
composant les ensembles évalués par l’OIU sont livrés sur le site avec une déclaration de
conformité.
Seuls les assemblages (permanents et non permanents) entre ESP ainsi que les supports de
ces équipements font l’objet d’une évaluation sur site (évaluation d ‘ensemble).
De même, pour des raisons pratiques, le Groupe EDF en accord avec les Fabricant, a décidé
de finaliser l’évaluation des ESP tuyauteries lors du montage sur site et lors des
vérifications finales des ensembles correspondants.
Selon les catégories de risque pression et le niveau de sûreté, les ESP(N) Tuyauteries
pourront être évalués par l’OIU d’une façon individuelle en module G (conception,
fabrication, vérification finale) ou par échantillonnage en module A1 (vérification finale).
INDEX çç Sommaire
Dans le cadre d’une évaluation selon le module G, la figure [3] décrit plus précisément les
principales actions d’inspections réalisées.
Dans le cas d’un module G, des inspections de chaque ESP tuyauterie sont organisées dans les
usines (préfabrication) afin de vérifier la conformité réglementaire des opérations de
fabrication (formage, cintrage, soudage) et les contrôles associés. C’est aussi le cas pour les
accessoires de tuyauterie obtenu par formage et soudage et les tubes
INDEX çç Sommaire
La figure 4 présente les inspections réalisées en usines ou sur site selon les modules
d’évaluation G ou A1.
Pour la préfabrication et les montage, les fréquences d’application des guides d’inspection
dans chaque usine et sur site sont précisées dans le tableau ci-dessous.
Inspection en usine et Ensembles (module G) Tuyauteries évaluées Tuyauteries évaluées en
sur site en module G module A1
1/25 ESP(N)
CND Radio (tir)
Les ESP(N) appartiennent au même ensemble
INDEX çç Sommaire
Inspection en usine et Ensembles (module G) Tuyauteries évaluées Tuyauteries évaluées en
sur site en module G module A1
Bibliographie
[2] Décret « Equipements sous pression » n° 99 1046 du 13/12/99 modifié. Son titre II
transpose la directive 97/23/CE. Son contenu est pratiquement identique, à l’exception de
quelques différences relatives au champ d’application et à la terminologie.
[4] Arrêté du 12/12/05 relatif aux équipements sous pression nucléaires (ESPN)
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page117
RÉSUMÉ / ABSTRACT
95D
LA REPRÉSENTATION NUMÉRIQUE PARAMÉTRIQUE D’ÉQUIPEMENTS
SOUS PRESSION POUR CERTIFICATION DE SÛRETÉ
P. Pittiglio (ISPESL DIPIA - Via Fontana Candida 1, 00040 Monte Porzio Catone, Rome, Italy – email: [email protected])
P.-A. Bragatto (ISPESL DIPIA - Via Fontana Candida 1, 00040 Monte Porzio Catone, Rome, Italy – email: [email protected])
A. Tonti (ISPESL DCC - Via Alessandria 220, 00198 Rome, Italy – email: [email protected])
F.Ferrone (ISPESL DCC - Via Alessandria 220, 00198 Rome, Italy – email: [email protected])
ABSTRACT
For over 80 years, ISPESL carries out certification of pressure equipment. Over time we developed a
collection of technical regulations for testing the stability of pressure vessels, which has been conti-
nuously updated to follow the technological developments and then to harmonize it with European
standards. The collection of technical standards takes in compact form the knowledge accumulated in
decades of activity. For this is especially popular in Italian industries manufacturing mechanical com-
ponents and plant components, for which is an essential resource.
After having worked for decades on an exclusive basis, as public body, the Institute since about seven
years works in competition, but maintains a competitive advantage, especially because it has a wealth
of knowledge and experience.
Before the new approach, the certification work has been governed by an established protocol. The
manufacturers submitted paper drawings and key data to the public body; then the documentation were
reviewed by this body. While over the decades this procedure remained the same, the mechanical en-
gineering design has changed, particularly in the last decade. It was changed step by step: first from
paper representation to digital representation, than 2D and 3D by CAD products. It was relatively sim-
ple to get to the digital representation of parameterized three-dimensional objects with complex CAD
products including CATIA © of Dassault Systems.
The targets were:
- to increase the effectiveness of certification tools for exploiting digital representation of mechanical
objects
- to create 3D digital representations of the equipment to be certified, based only on design data used
to verify pressure and temperature stability
- to have a tool accepted by the certification body and the designer, making possible to dynamically
assess the effect on the safety of the vessel, while changing the geometrical and technological parame-
ters (material properties)
- to let more available the knowledge contained in the Code by the parameterization, connecting the
stability verification tool with a more powerful and more explicit graphical representation.
These targets were achieved by developing an ad hoc software that accepts as input the design data of
a single pressure part, filling the stability calculation form. The software produces the matrix of para-
meters that is used by the CATIA software © for to create a 3D item. The items are than assembled as
separate parts. For openings and branches applied to the shell and bottoms, the possibilities of digital
modelling of CATIA © (union, intersection, extrusion etc..) were exploited.
The developed software was made available to ISPESL certification personnel. The user interface fol-
lows that of the form already used for paper-based procedure. Catia © and its possibility of calculation
remains as a «black box». The certification personnel has not to learn difficult instruments, but only
has the advantages of a more powerful representation. The software is made available through the in-
tranet and it’s accessible from about forty ISPESL departments in Italy (more than one hundred people
can access this tool).
INDEX çç Sommaire
The parametric digital representation of pressure equipment for safety
certification
Paolo Pittiglio
(ISPESL DIPIA - Via Fontana Candida 1, 00040 Monte Porzio Catone, Rome, Italy – email:
[email protected])
Andrea Tonti
(ISPESL DCC - Via Alessandria 220, 00198 Rome, Italy – email: [email protected])
Fabrizio Ferrone
(ISPESL DCC - Via Alessandria 220, 00198 Rome, Italy – email: [email protected])
Abstract
For over 80 years, ISPESL carries out certification of pressure equipment. Over time we developed
a collection of technical regulations for testing the stability of pressure vessels, which has been
continuously updated to follow the technological developments and then to harmonize it with
European standards. The collection of technical standards takes in compact form the knowledge
accumulated in decades of activity. For this is especially popular in Italian industries manufacturing
mechanical components and plant components, for which is an essential resource.
After having worked for decades on an exclusive basis, as public body, the Institute since about
seven years works in competition, but maintains a competitive advantage, especially because it has
a wealth of knowledge and experience.
Before the new approach, the certification work has been governed by an established protocol. The
manufacturers submitted paper drawings and key data to the public body; then the documentation
were reviewed by this body. While over the decades this procedure remained the same, the
mechanical engineering design has changed, particularly in the last decade. It was changed step by
step: first from paper representation to digital representation, than 2D and 3D by CAD products. It
was relatively simple to get to the digital representation of parameterized three-dimensional objects
with complex CAD products including CATIA © of Dassault Systems.
The targets were:
- to increase the effectiveness of certification tools for exploiting digital representation of
mechanical objects
- to create 3D digital representations of the equipment to be certified, based only on design
data used to verify pressure and temperature stability
- to have a tool accepted by the certification body and the designer, making possible to
dynamically assess the effect on the safety of the vessel, while changing the geometrical and
technological parameters (material properties)
- to let more available the knowledge contained in the Code by the parameterization,
connecting the stability verification tool with a more powerful and more explicit graphical
representation.
These targets were achieved by developing an ad hoc software that accepts as input the design data
of a single pressure part, filling the stability calculation form. The software produces the matrix of
INDEX çç Sommaire
parameters that is used by the CATIA software © for to create a 3D item. The items are than
assembled as separate parts. For openings and branches applied to the shell and bottoms, the
possibilities of digital modelling of CATIA © (union, intersection, extrusion etc..) were exploited.
The developed software was made available to ISPESL certification personnel. The user interface
follows that of the form already used for paper-based procedure. Catia © and its possibility of
calculation remains as a "black box". The certification personnel has not to learn difficult
instruments, but only has the advantages of a more powerful representation. The software is made
available through the intranet and it's accessible from about forty ISPESL departments in Italy
(more than one hundred people can access this tool).
1 Introduction
The pressure equipment are key components throughout the process industry, particularly in the
areas of energy, oil & gas, basic chemicals, fine chemicals and pharmaceuticals, petrochemicals).
The verification of safety criteria in the design, construction and operation of pressure equipment
has always played an essential role in the prevention of accidents at work and activities of
certification, although with different names or modes, was carried out continuously by ISPESL for
eighty years. The codes for the verification of the pressure vessels (fired or unfired) VSG [1] and
VSR [2] are a concentration of huge knowledge gathered over many years of activity in the field
of industrial components. Pressure equipment, as work equipment, are required to comply with
specific laws and regulations transposing EU directives. In establishments where is enforced Seveso
Directive, safety of pressure components is also a necessary condition for the prevention of major
accidents. In the field of Seveso Directive the certification of pressure equipment is always aligned
with technological developments. In the field of pressure equipment have occurred many
innovations in materials, construction techniques and methods of operation which have been
included in the activity of certification. A major technological change in recent years has had a
significant impact on the sector: it is computer-aided design (CAD, computer aided design). Started
as a simple computer support to mechanical design, CAD has subsequently evolved following the
three-dimensional geometric modeling of the parties, the parametric representation of the
components, the definition of features associated with the geometry. Today CAD is a
comprehensive design, which also define the geometric and functional rules which the components
must comply.
It should be emphasized that technical regulations are the essence and synthesis of knowledge and
experience. The ability to computerize the technical rules in a fair and efficient design means faster,
higher quality and cost control due to unforeseen design errors. For this reason, the CAD systems
that can define and manage rule sets are called Knowledge Based CAD KB-CAD. Knowledege
Based Engineering KBE, is another term used in all phases of a project. Users, in fact, not only
design the piece but also make explicit the rules through the knowledge and expertise on which
basis the design activities. In this case, the digital representation of the components includes, in
addition to geometry, report rules and then becomes a true representation of knowledge. KBE tools
have many applications in mechanical design, automotive and aerospace and they are in continuous
development, as documented in the recent review [3].
Since CAD systems are now more or less developed, a resource common to all manufacturers of
pressure vessels, it is also reasonable to think the certification activities should use the opportunity
to have digital "smart" representation of components under pressure. The same ability to manage
rules, can be conveniently exploited by inserting their national rules, like ISPESL VSR and VSG,
harmonized with European legislation and used by all the Italian industry. This paper describes a
newly developed application that exploits the potential of CAD to make more effective the work of
equipment certification. Born as a prototype in a research program about risk of a major accident
(Bragatto & al. 2005), the solution presented in this paper was subsequently engineered and
INDEX çç Sommaire
integrated with the system in use in the Department of Certification of ISPESL, ready for
operational use.
The design of plant and equipments must meet a set of criteria or minimum safety requirements,
under which the facilities are approved by the competent authorities. Among the components
subject to special criteria or safety requirements electrical systems, automatic protection against
electric risk, containers and tanks, piping, pressure equipments are included.
The requirements may be related to single equipment in the installation, their connections,
processes, substances, materials, security systems, etc.. For the designer the minimum safety criteria
may leave a wide margin of freedom, since there are explicit instructions for building and
manufacturing, but simply establish conditions under which the project can be considered safe. In
most cases, the safety policy occurs only after the first design phase. In order to guide the work of
the designer in all his steps, and avoid that the project should be reconsidered after, it was verified
that the requirements of safety are defined in different application areas of notified bodies, which
define even at the level of detail, the geometrical characteristic of the components to achieve, in
order to comply with all minimum requirements.
Technical regulations, created for testing, are detailed in order to be directly used at design stage.
Even if they put constraints to the freedom of the designer, they allow him to have a faster and more
efficient design stage, with direct achievement of safety. Achieving the safety requirements without
having concrete rules it is possible, but it requires a much greater efforts.
The work of verification and approval must be undertaken by the authorities responsible for risk of
major accidents, and it can be facilitated by the presence of strictly codified rules rather than by
general principles.
In the framework of Italian Legislation on Major Accident Hazard, the technical rule on the criteria
of analysis and evaluation of safety reports (DPCM 31/03/89) considers the adoption of design
standards for tanks and vessels is an important element in assessing the completeness and adequacy
of the contents of safety reports.
From a logical point of view it is clear that the rules arising from the criteria, but it has not said that
this is obvious. In many cases, the rules are simply imposed without motivating them in detail . This
applies equally to corporate standards or industry standards as to national standards. The general
criteria are as presented in explicit form only when interchange situations arise with areas where
different rules apply formally, even if arising from the same general criteria. In this case the criteria
may erroneously appear to the user as additional conditions to the technical regulations.
The distinction between rules and standards it is far from academic. Technical rules defined and
tested, in fact, imply, unlike the basic criteria, the ability to "automate" the observance of the rules,
so that is not even possible to design something dissimilar to them. In this case the designer would
be forced to draw components and systems according to defined rules and safety would become an
inherent feature of the drawing. To achieve this it is necessary that the management of rules passes
the responsibility of the individual projects to systems that already automate almost all phases of
design cycle.
The set of safety standards is the VSR code, developed in Italy in the early seventies by
manufacturers and government agencies, implementing the legislation on the verification of
pressure vessels. VSR code is similar to other codes developed by the major industrialized countries
like the UK BS5500, ASME VIII Div 1 in USA and AD- MERKBLATT in Germany. For 30 years,
ISPESL has been mandated to manage and update the code VSR in Italy. Facilitating tools have
INDEX çç Sommaire
been developed, during the years, such as forms to be filled in, including the complete data set for
the verification. Automated computation of individual formulae had been developed in the past.
In 1997, the European Commission published Directive PED 97/23/EC to enable the free exchange
of pressure equipment throughout the EU. It imposes no national code, but only defines the general
criteria, the essential safety requirements which must be verified. The VSR has been aligned with
the PED, and is now widely used in Italy for the design and certification of pressure equipment.
The certification work is divided into two distinct categories: certification according to national
legislation, relating to pressure equipment already approved under the national legislation and today
the subject of repair and / or minor changes; PED certification activities for pressure equipments
completely new, or subject to considerable modifications.
The distinction is crucial, because EU rules are different from national ones. Inspections to be
carried out under national legislation follow the technical rules contained in the Collections ISPESL
VSR and VSG, while those carried out according to PED follow Collections amended or CTI
Recommendations for the use of VSR / VSG with PED .
In order to respect the distinction mentioned above, the application VSR / VSG lets you choose the
following verification procedures:
1. VSR for pressure equipment
2. VSG for steam generators
3. VSR PED pressure equipment conform to PED
4. VSG PED for steam generator in accordance with the PED.
At the organizational level should be stated that the technician carrying out audits, normally
operates with the support of the paperwork submitted by the applicant. It is not easy to operate with
computerized procedures transcribing data obtained only on paper. Once entered this data,
managing technical and administrative practice becomes completely automated and carried out by
computer procedures (in addition to VSR / VSG, the administrative part is handled by another
application called SIDPI).
In order to assure for the unfired vessel the containment of internal and external pressure under
design condition, VSR rules define minimal thickness of cylindrical or spherical shells, of domed or
conical heads and covers, of flat ends and plates and state conditions on all vessel details, such as
stiffening rings, openings, flanges, jackets, nozzles, valves and tube-sheets. In VSR code,
conditions are expressed as functions of geometrical parameters, such as diameter, length, width,
distance, angle, areas and volume; operation parameters such as peak pressure, test pressure, design
pressure, operation temperature; material feature such as elasticity modulus, Poisson coefficient and
corrosion allowance.
Each individual rule in VSR has as identifier, a string of digits, separated by commas. A
hierarchical classification of the member parts underlies the VSR and VSG rules. Each digit in the
string represents a class or subclass of the member part taxonomy. Each rule may contain many
formulas. VSR has some hundred “rules” and some thousand formulas. The VSG code for the fired
vessels, for internal organization and size, is analogous to the VSR code.
VSR and VSG codes are the results of decades of research and experience of thousands of
engineers, designer and inspectors, in the Italian process industries. VSR and VSG are a highly
valuable knowledge asset, which now is threatened by the fast personnel turnover. A further threat
come from the time to market, which is becoming shorter and shorter. A work a cycle usually
involve designers which propose new equipment, certifiers which recommend modifications to
improve safety and again designer, which have to implement recommendations, respecting
customer requirements. As time to market is very short both certifier and designer are highly
pressed with potential negative effects on work accuracy.
Therefore it is essential to protect such an asset and, mainly, make it as usable as possible.
Information Technology may offer opportunity to make VSR and VSG codes more usable.
INDEX çç Sommaire
3 Objectives
The objective of the research presented in this article is to establish a system design knowledge-
based (KB) is to support engineers in the certification, the system must verify compliance with the
rules of construction pressure equipment. With the instrument described in this article, once the
model built on pressure equipment in question can be verified to satisfy the VSR. Since the model
of equipment, as described in the set of rules or VSR VSG, geometry and functional constraints can
be verified through the form of KBE CAD system. Verification is carried out in a semi-automatic,
or is able to highlight possible non-compliance to the user, who may decide to implement corrective
action.
The general objectives of an application that exploits the digital representation of the mechanical
parts include:
- Increase the effectiveness of certification tools using digital representation of mechanical objects.
It allows to put together all the verification on individual part and to have an overall verification.
- Create 3D digital representations of equipment to be certified, based only on information provided
with the documentation required in the procedure.
- Having a tool shared between CA and designer, making it possible in particular to assess the
dynamic effect on the safety of the vessel to vary the geometrical parameters and technological
(material).
- To make more accessible the knowledge contained in the body of rules connecting it to a graphical
representation more attractive and more powerful (thanks to parameterization)
- Implement dynamic models with simulation of degradation mechanisms
In this field you enter the control product life cycle of equipment that allow simulations on
production systems more complex. The possibilities include the design of any complex assembly,
design and free-form surfaces, managing the various stages production, assembly, disassembly, etc..
The system may also be coupled with finite element solvers, completing therefore the overall
framework of the monitoring arrangements set out in the computer codes, including the VSR.
Concerning the multi code (simultaneous application of multiple codes), it is still something to
experience, but increasingly required. Often we are asked to design according to international codes
and check according to the national standard (VSR). To this end should not be forgotten that the
VSR and VSG are verification codes and design.
4. Methods
Certification activity may be improved by using the same sophisticated KBE systems used in design
activity. By exploiting as a interfering engine CATIAV5, a powerful KBE system provided by
Dassault Systemes, an application has been developed, aiming to support the certification activity,
according to the VSR and VSG safety codes. The application, named RP-V2 starts from the
documents provided by the manufactures, in order to obtain the certification of conformity to the
according to the VSR and VSG safety codes. Documents are supposed to be provided by the
manufacturers, according to the usual protocol, as enforced for a long time. Basically the protocol
require the scrutiny of each individual part of the vessel (ends, shields and opens), considering the
formulas of the VSR or VSG code.
5.1 Generality
From these data, the certifier, by means of RP-V2, is enabled to organized all data required to build
the complete 3D model of the assembled vessel and send them to the KB-CAD system CATIA
V5©. From a point of view of information technology, RP-V2 is a client of CATIA V5©, which
may provide different clients with computation and rules management services. All benefits of a
complete 3D solid representation are available in the KB-CAD system, including the automatic
generation of drawing tables, featuring canonical, axonometric views and sections. A number of
computations may be made by the KB-CAD system, including. weight, centers, surface and
automated meshing. Furthermore it is possible to have the model in independent format. In such a
way, even without a CAD systems, a credible 3D model may be shared and discussed between
different stakeholders (designers, manufacturers, certifiers, installers). An effective visualization is
not the only advantage of a 3D digital representation of the vessel. The effects of vessel safety due
to any change of material or of geometrical parameters, may be evaluated in real time and discussed
between designer and certifier.
The user interface of RP-V2 is very simple and easy to learn for a user. Of course the user is
supposed to have an expertise in safety codes, but he is not required to have any competencies in
INDEX çç Sommaire
CAD systems. The User interface has been developed, having in mind the workflow of equipment
certification, as shown in figure 1.
Create 3D
model
The first step is to introduce the identification data, the operation condition and the material
parameters. The next step is to analyze each part, organized according to the VSR -VSG taxonomy,
mentioned in the paragraph 2.1. The input of the required parameters is supported by an
iconography. For each item in VSR – VSG taxonomy, RP-V2 provides a small 3D picture, where
the geometric parameters are impressively represented. In Figure from 2-4 there are a few
screenshots, which show the input interface for a few VSR rules.
Fig 2 Input of the parameters Fig 3 Input of the parameters Fig 4 Input of the parameters
for an end for a shell for a open
In the next step, the member parts have to be organized in a logical sequence, following the basic
axial symmetry of the vessel. Ends and flanges have to be oriented respect to the vessel axis. Opens
have to be positioned, by indicating their cylindrical coordinates respect to the vessel origin. The
graphical user interface of RP-V2 has two keys, “>” and “<”, to move the parts forward and
backward along the vessel axis, the parts are represented by very simple icons, including “(“ and
“)” for curve heads, “[“ and “]” for flat ends, “=” for shells; “>]” and “[< “ for flanges, “⊥” for opens.
The screenshot in figure 5 may give a feeling of this procedure.
INDEX çç Sommaire
Figure 5. Ordering the part list
When the parts has been ordered, RP-V2 is able to verify the closure of the vessel. At this step the
geometrical coherence of all parts is automatically checked. The screenshot in figure 6 show this
work phase.
5.3 Exporting RP V2
INDEX çç Sommaire
If the defined member parts compose a closed assembly, they are sent to the server, where the
CATIA-V5© engine generates automatically the 3D model and the drawing with the canonical and
axonometric views of the vessel, as shown in figure 7.
A model in VRML format may be produced and sent back to the RP-V2 client. As said before the
VRML format is independent on CAD systems. VRML format, which is a standard file format for
representing 3-dimensional (3D) interactive vector graphics, designed for the Web. Many VRML
free viewer, featuring basic 3D visualization capabilities (rotate, zoom, pan) may be found in
internet. In figure 8 the VRML model is shown.
INDEX çç Sommaire
Of course it is always possible to use CATIA V5© in an interactive way and exploit all its
remarkable computation capabilities. In figure 9 a CATIA screenshot is shown where a single
member part has been highlighted. In the left side of the figure the CATIA data structure is shown.
Conclusions
RP-V2 make more usable the huge knowledge asset contained in the VSR and VSG standard codes.
It connects the rules and the formulas in the code to a digital 3D representation, which allows to
have realistic pictures. The 3D parametric representation is also a powerful tool for shortening the
approval cycle. Any potential changes in member parts and in material may be quickly simulated
and discussed via web by the certifier and the designer. Furthermore it is possible to verify in the
same digital model, other technical requirements, not included in VSR and VSG codes; such as
welding, material and manufacturing procedures.
At now RP-V2 is in test phase and soon will be given to some hundred certifiers, in the forty local
ISPESL offices in Italy.
References
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page118
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page119
M. J. GOELLNER MEEDDM/BSEI
M. M. HAHN FORUM DES ORGANISMES NOTIFIÉS
M. A. HARRIS ORGALIME
M. F. RIEFFEL ACM
M. Y. GUYOT TECHNIP
M. S. CAMERON DOOSAN BABCOCK ENERGY
M. P. DAOUST TRACTEBEL ENGINEERING – GDF SUEZ
M. A. GINGELL INDUSTEEL
Thème/Subject :
La table ronde organisée dans le cadre des Journées ESOPE 2010 sera dédiée au marché des
équipements sous pression. Les différents aspects stratégiques ayant un impact sur l’harmonisa
tion, le développement et la compétitivité, pour la fabrication et l’exploitation des équipements
sous pression seront abordés par les membres de la table ronde et au travers des questions de
l’auditoire
The panel organized in the context of the ESOPE 2010 Symposium will be dedicated to the pres
sure equipment market. The different strategic aspects having an impact on harmonization, de
velopment and competitiveness for the design, fabrication and service of pressure equipment will
be covered by the panel members. They will have the opportunity to express their own views and
answer to the questions of the audience, and to those expressed during the discussion.
Durée/Duration:
Les deux heures prévues pour la table ronde, correspondent aux interventions des membres et
aux réponses aux questions de l’auditoire.The two hours panel include the presentation of panel mem
bers view and the answer of the questions of the audience.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page120
RÉSUMÉ / ABSTRACT
107SP
PLAN DE MODERNISATION DES INSTALLATIONS INDUSTRIELLES :
PRÉVENIR LES RISQUES LIÉS AU VIEILLISSEMENT
I. Griffe
(MEEDDM/DGPR/BSEI – Arche Nord – 92055 La Défense –
e-mail : [email protected])
RÉSUMÉ
Ouverture d’un bac dans un dépôt pétrolier en 2007, fuites de canalisations en 2008 : plusieurs incidents
et accidents survenus ces dernières années dans les installations industrielles françaises ont pointé du
doigt la problématique du vieillissement des installations, de leur maintenance et de leur surveillance.
Ce constat a conduit le ministère du développement durable à lancer fin 2008 un plan pour la maîtrise
du vieillissement dans les installations industrielles.
A l’issue des concertations menées en 2009, le ministre de l’écologie, de l’énergie, du développement
durable et de la mer a arrêté un plan d’une quarantaine de mesures qui concernent tous les secteurs
industriels avec, de façon plus spécifique, l’ensemble de la filière du pétrole et de ses produits dérivés
ainsi que le secteur de la chimie. Un volet particulier couvre la problématique des équipements sous
pression (récipients et tuyauteries).
Jeudi
ABSTRACT
Opening of a tank in an oil depot in 2007, leaking pipes in 2008 : several incidents and accidents in
recent years in the french industrial plants have pointed the problem of ageing risks, maintenance and
monitoring.
This led the Ministry of sustainable development to launch in the end of 2008 a plan to prevent the
ageing risks in industrial plants .
Following the consultations held in 2009, the minister of ecology, energy, sustainable development and
sea adopted a plan of about forty measures which concern all industrial sectors with, more specifically,
the entire sector of petroleum and its derivatives as well as the chemical sector. A particular aspect
concerns the issue of pressure equipments (vessels and piping).
INDEX çç Sommaire
LE CONTEXTE
Ouverture d’un bac dans un dépôt pétrolier en 2007 ayant pollué sur près de 40 kilomètres des
berges de la Garonne, fuites de canalisations en 2008 ayant nécessité trois mois de ravail de
750 personnes pour dépolluer les berges de la Loire et les côtes atlantiques, rupture de
canalisation à l’été 2009 souillant plusieurs hectares du parc régional de la Crau : plusieurs
incidents et accidents survenus ces dernières années dans les installations ndustrielles
françaises ont pointé du doigt la problématique du vieillissement des installations, de leur
maintenance et de leur surveillance.
Ce constat a conduit le ministère du développement durable à lancer fin 2008 un plan pour la
maîtrise du vieillissement dans les installations industrielles. Tous les secteurs industriels sont
concernés avec de façon plus spécifique l’ensemble de la filière du pétrole et de ses produits
dérivés ainsi que le secteur de la chimie.
L’ORGANISATION DE LA DEMARCHE
La démarche a été menée dans un esprit de concertation élargie. Le ministère a entamé dès
janvier 2009 des rencontres avec un certain nombre d’acteurs : fédérations professionnelles et
experts (notamment établissements publics compétents tels que l’INERIS, Institut national de
l’Environnement industriel et des risques, et l’IRSN, Institut de radioprotection et de sûreté
nucléaire) de façon à réfléchir à la mise en place de cette démarche.
À partir du mois de février, six groupes de travail, mettant à contribution près de 130
personnes, experts et gestionnaires (spécialistes du monde industriel, experts techniques,
représentants de l’administration pour la plupart) se sont réunis régulièrement pour fonder les
bases du plan vieillissement.
LE CONTENU DU PLAN
Le plan arrêté en janvier 2010 par Jean-Louis Borloo et Chantal Jouanno contient 38 mesures,
regroupées en six thématiques :
- en thématique transversale, une action concernant la modernisation des systèmes de
gestion de la sécurité pour mieux prendre en compte la problématique du
vieillissement
- une thématique génie civil, qui reprend 4 actions
- une thématique bacs de stockage qui reprend 14 actions
- une thématique canalisations de transport qui reprend 8 actions
- une thématique capacités et piping (tuyauteries sur site industriel) qui reprend 7
actions
- une thématique instrumentation de sécurité qui reprend 4 actions.
INDEX çç Sommaire
ACTIONS THEMATIQUES - CAPACITES ET PIPING (TUYAUTERIES SUR SITE)
Situation actuelle
Ces actions traitent des questions d’équipements sous pression (ESP), qu’ils soient ou non
soumis à surveillance réglementaire.
La réglementation relative aux équipements sous pression est ancienne et détaillée. Des
guides techniques ont été élaborés dans certains secteurs d’activité. Les modes de dégradation
des équipements y sont normalement détaillés.
Néanmoins, il est constaté chaque année un certain nombre d’incidents ou d’accidents mettant
en cause des équipements sous pression.
Deux catégories d’équipements peuvent être distinguées : les récipients dont le suivi apparaît
correct dans la majorité des cas et les tuyauteries dont le suivi est souvent insuffisant.
Pour le suivi des équipements sous pression, les sites les plus importants disposent de services
d’inspection, reconnus (SIR) ou non par l’administration, qui regroupent des inspecteurs
certifiés formés en propre par les exploitants. La criticité des équipements est établie selon
une méthode dite RBI (Risk Based Inspection) et des plans d’inspection détaillant le suivi à
réaliser sont rédigés. En dehors du cas d’équipements identifiés comme consommables, l’âge
des équipements est un critère peu ou pas utilisé.
Les industriels indiquent que le système du service d’inspection n’est pas généralisable mais
qu’un « garant » est néanmoins présent sur tous les sites (des secteurs de la chimie et de la
pétrochimie) afin d’assurer le suivi des équipements sous pression ; cette appellation recouvre
néanmoins des situations très hétérogènes selon la taille des sites et la formation du personnel.
Les techniques de contrôle sont maîtrisées et doivent être adaptées aux modes de dégradation
des équipements sous pression (radiographie…). L’émergence de techniques alternatives est
rendue difficile par la nécessaire phase d’industrialisation mais permettrait d’améliorer la
contrôlabilité de certains équipements difficilement accessibles.
Un retour d’expérience est organisé par la profession lors de journées dédiées (journées
GEMER organisées par l’UFIP (Union Française des Industries Pétrolières), journées
techniques de l’AFIAP (Association Française des Ingénieurs en Appareils à Pression),
journées «inspection UIC » (Union des Industries Chimiques)….
INDEX çç Sommaire
Les deux actions suivantes sont d’ores et déjà en cours de mise en oeuvre. Elles doivent
permettre d’améliorer la connaissance des équipements et de leurs modes de dégradation. Il
s’agit :
Les autres actions proposées ont été engagées et seront finalisées durant l’année 2010.
INDEX çç Sommaire
- activités à hauts risques (par exemple, en référence aux rubriques de la nomenclature
des installations classées),
- mécanismes d'endommagement (fissuration par fatigue, fluage, par exemple),
- équipements qui présentent des risques importants (pression-volume-température,
criticité élevée pour les SIR, autres critères pour les non SIR…).
Une disposition réglementaire sera prise pour instaurer ce mécanisme (un article sera ajouté
dans l’arrêté ministériel du 15 mars 2000 relatif au suivi en service des équipements sous
pression).
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page121
RÉSUMÉ / ABSTRACT
106SP
DES ÉQUIPEMENTS NUCLÉAIRES SURS ET COMPÉTITIFS POUR UN DÉVELOPPEMENT
DURABLE : ANTICIPATION, STRATÉGIE INDUSTRIELLE, MÉTHODES ET OUTILS
RÉSUMÉ
La demande en énergie devrait doubler d’ici à 2030. Cela amène de plus en plus de pays à repenser leur stratégie de mix-énergétique et à
opter pour l’introduction ou le renouvellement d’une génération de puissance d’origine nucléaire. Cette solution est à la fois durable,
respectueuse de l’environnement, fiable et économique.
Toutefois, la renaissance du nucléaire est quelque peu ralentie par les incertitudes du marché liées aux problématiques de capacités de
production électrique nécessaires pour soutenir le développement et de calendrier de déploiement.
Satisfaire les besoins de partenaires historiques ainsi que de nouveaux acteurs sur la scène internationale, prendre en considération les
contraintes de leurs cadres législatifs et réglementaires sont autant de défis auxquels le groupe AREVA doit répondre.
Afin de fournir les meilleures solutions à ses clients actuels et futurs, la stratégie d’AREVA repose sur trois principaux piliers :
1. Harmonisation, Standardisation et Flexibilité : Soutenir les organisations et les initiatives internationales (AIEA, WENRA,
WNA, MDEP …) visant à améliorer le niveau de sûreté global en particulier par la compréhension et l’harmonisation des
réglementations et des pratiques ; comprendre les différences et s’adapter aux réglementations spécifiques ; tels sont les enjeux
d’aujourd’hui. L’harmonisation est un préalable à la standardisation ainsi qu’à la flexibilité et permettra d’atténuer les
incertitudes et de réduire les coûts et les délais des programmes de centrales nucléaires
2. Qualité et Maîtrise des processus : Développer et maintenir des codes de conception et de construction de haute qualité, intégrer
les résultats de programmes de recherche et de développement ; construire des partenariats stratégiques ; maintenir et développer
des outils industriels adaptés pour optimiser et maîtriser nos processus de fabrication avec des niveaux de standards de qualité
élevés
3. Ressources Humaines et Expertise : Développer l’expertise de nos équipes et de nos partenaires en préservant, partageant et
développant les connaissances et l’expérience via des programmes de formation spécifiques ; responsabiliser les équipes et
continuer à accroître leur culture en matière de sûreté, remettre en question certaines réflexions et tirer des enseignements des
expériences passées afin d’être toujours plus performant
et sur une politique de développement de réacteurs de génération III+ (réacteurs EPR™, ATMEA1™ et KERENA™), associant performances
opérationnelles et niveau de sûreté, pour une exploitation sure dans le long terme.
ABSTRACT
The worldwide nuclear renaissance expected now for a few years is still surrounded by uncertainties regarding manufacturing capacity
(new reactors and new equipment for life extension) and construction and implementation schedules; historical and new stakeholders emerg-
ing on the world scene and their various legal and regulatory frameworks represent as many challenges to be addressed by AREVA.
To provide the best sustainable, environmentally responsible, safe, reliable, timely and economical solutions to its current and future Clients,
AREVA has implemented a development program based on three main axes:
1. Harmonization, standardization and flexibility : Providing support to international organizations and initiatives (IAEA,
WENRA, MDEP …) aiming at increasing safety worldwide through harmonized regulations and practices ; understanding
our environment, issues and trends and being flexible enough to adapt to specific regulations ;
2. Quality and mastering of processes : Developing and maintaining high level design and construction codes, integrating R&D
results as well as experience feedback ; maintaining and developing adapted industrial facilities to optimize and control our
fabrication process with effective quality standards ;
3. Human resources and expertise : Developing expertise and safety culture by preserving, sharing and developing knowledge
and experience through education and training programs for our staff and our partners ; developing awareness of and res-
ponsibility for safety issues in their work
and on a reactor model development policy aiming to offer generation III+ reactor (EPR™, ATMEA1™ and KERENA™ reactors) with
unprecedented operational performance and safety level, for long-term safe operation.
AFIAP - ESOPE 2010 Programme
INDEX çç Sommaire
SAFE AND COMPETITIVE NUCLEAR EQUIPMENT
FOR A SUSTAINABLE DEVELOPMENT: ANTICIPATION, INDUSTRIAL
STRATEGY, METHODS AND TOOLS
Alexis Marincic - Executive Vice President Products & Technology business unit
Etienne Touzain, Franck Lignini
AREVA – 1, Place Jean Millier – 92084 Paris la Défense Cedex
email : [email protected]
RESUME
La demande en énergie devrait doubler d’ici à 2030. Cela amène de plus en plus de pays à repenser
leur stratégie de mix-énergétique et à opter pour l’introduction ou le renouvellement d’une génération
de puissance d’origine nucléaire. Cette solution est à la fois durable, respectueuse de l’environnement,
fiable et économique.
Toutefois, la renaissance du nucléaire est quelque peu ralentie par les incertitudes du marché liées aux
problématiques de capacités de production électrique nécessaires pour soutenir le développement et de
calendrier de déploiement.
Satisfaire les besoins de partenaires historiques ainsi que de nouveaux acteurs sur la scène
internationale, prendre en considération les contraintes de leurs cadres législatifs et réglementaires sont
autant de défis auxquels le groupe AREVA doit répondre.
Afin de fournir les meilleures solutions à ses clients actuels et futurs, la stratégie d’AREVA repose sur
trois principaux piliers :
1. Harmonisation, Standardisation et Flexibilité : Soutenir les organisations et les initiatives
internationales (AIEA, WENRA, WNA, MDEP …) visant à améliorer le niveau de sûreté
global en particulier par la compréhension et l’harmonisation des réglementations et des
pratiques ; comprendre les différences et s’adapter aux réglementations spécifiques ; tels sont
les enjeux d’aujourd’hui. L’harmonisation est un préalable à la standardisation ainsi qu’à la
flexibilité et permettra d’atténuer les incertitudes et de réduire les coûts et les délais des
programmes de centrales nucléaires
2. Qualité et Maîtrise des processus : Développer et maintenir des codes de conception et de
construction de haute qualité, intégrer les résultats de programmes de recherche et de
développement ; construire des partenariats stratégiques ; maintenir et développer des outils
industriels adaptés pour optimiser et maîtriser nos processus de fabrication avec des niveaux de
standards de qualité élevés
3. Ressources Humaines et Expertise : Développer l’expertise de nos équipes et de nos
partenaires en préservant, partageant et développant les connaissances et l’expérience via des
programmes de formation spécifiques ; responsabiliser les équipes et continuer à accroître leur
culture en matière de sûreté, remettre en question certaines réflexions et tirer des
enseignements des expériences passées afin d’être toujours plus performant
INDEX çç Sommaire
et sur une politique de développement de réacteurs de génération III+ (réacteurs EPR™, ATMEA1™
et KERENA™), associant performances opérationnelles et niveau de sûreté, pour une exploitation sure
dans le long terme.
ABSTRACT
The worldwide nuclear renaissance expected now for a few years is still surrounded by uncertainties
regarding manufacturing capacity (new reactors and new equipment for life extension) and
construction and implementation schedules; historical and new stakeholders emerging on the world
scene and their various legal and regulatory frameworks represent as many challenges to be
addressed by AREVA.
To provide the best sustainable, environmentally responsible, safe, reliable, timely and economical
solutions to its current and future Clients, AREVA has implemented a development program based on
three main axes:
1. Harmonization, standardization and flexibility : Providing support to international
organizations and initiatives (IAEA, WENRA, MDEP …) aiming at increasing safety
worldwide through harmonized regulations and practices ; understanding our environment,
issues and trends and being flexible enough to adapt to specific regulations ;
2. Quality and mastering of processes : Developing and maintaining high level design and
construction codes, integrating R&D results as well as experience feedback ; maintaining and
developing adapted industrial facilities to optimize and control our fabrication process with
effective quality standards ;
3. Human resources and expertise : Developing expertise and safety culture by preserving,
sharing and developing knowledge and experience through education and training programs
for our staff and our partners ; developing awareness of and responsibility for safety issues in
their work
and on a reactor model development policy aiming to offer generation III+ reactor (EPR™,
ATMEA1™ et KERENA™ reactors) with unprecedented operational performance and safety level,
for long-term safe operation.
INDEX çç Sommaire
Numerous studies have been recently performed to build projections on the electricity demand and on
the nuclear energy generation capacities in the short, medium (commonly up to 2030) and longer term.
The available projections show that while some countries will expand existing nuclear power
programs, a significant number of countries are considering launching new nuclear power programs.
Their expectations in terms of nuclear generation capacity (number and size of units) and the
associated deployment calendar are subject to uncertainties which will require flexibility and reactivity
from vendors to anticipate on and to establish the required production capacity for nuclear equipment.
To effectively contribute to the expansion of carbon free electricity generation, AREVA is aiming at
building 1/3 of the worldwide new nuclear generation capacity in accessible markets by 2030. One of
the key success factors to reach this ambitious objective is the design standardization of the nuclear
reactors. This standardization would allow utilities to order AREVA reactors without major design
adaptations needed to comply with specific national regulations. This would in turn allow to fully
leverage fleet and series effects and to lower both project risks and operational costs (e.g. lower
engineering costs on the “standard” design scope, scale effects on the supply chain…).
AREVA’s Generation III+ reactors, based on
evolutionary designs, benefit from maximum
standardization to simplify licensing in any
country. The overall performance achieved
thanks to AREVA’s uninterrupted
engineering, manufacturing and supply chain,
the inclusion of leading I&C technology
allows a maximized plant availability, a high
plant efficiency and low operation and
maintenance costs. The reactor portfolio
designs display an outstanding safety level
such as resistance to large commercial airplane
crash, reduced doses to the workers in situ,
and the assurance of no unacceptable impact
beyond the power plant site limits, whatever
the situation inside the reactor. In addition,
these reactors authorize the use of several
types of fuel (Uranium, MOX, …) leading to a
sustainable natural resource management.
AREVA is currently building four Generation III+ 1650 MWe class EPRTM reactors in Finland, France
and China. AREVA and MHI are associated to develop ATMEA1TM, a Generation III+ 1100 MWe
class Pressurized Water Reactor adapted to countries with relatively small electrical grids. AREVA is
also developing KERENATM, a Generation III+ 1250 MWe class boiling water reactor in close
partnership with E.on.
For the longer term, AREVA is actively involved in the development of Generation IV concepts such
as sodium cooled fast reactors, allowing more efficient use of uranium and offering interesting
perspectives for the transmutation of highly radioactive and long live waste. AREVA is collaborating
with EDF and with the French CEA on the development of the ASTRID prototype, a sodium fast
reactor industrial demonstrator, planned to be built in France in the 2020s prior to industrial
deployment in the following decades.
INDEX çç Sommaire
AREVA’s industrial strategy to bolster the sustainability of nuclear power relies on the 3 following
main areas:
1. Harmonization, standardization and flexibility
2. Quality and industrial process
3. Human resources and expertise
INDEX çç Sommaire
AREVA is encouraging all Standard Following an audit performed by the American Society of Mechanical
Development Organizations to Engineers (ASME) in December 2009, the Reactors & Services Business
develop their relationships and work Group in France (the legal entity AREVA NP SAS France) obtained a 3-year
on harmonization of their ASME certification on 22 January 2010.
This certification, along with the
requirements and actively supports
ones already valid in Germany
international initiatives through the and the United States, means
involvement of its experts in various AREVA is recognized for the
initiatives and networks design and manufacturing of
(participation in the drafting and nuclear reactor components in
accordance with the ASME
review of IAEA Safety Standards,
standards as well as for their on-
WNA CORDEL, ISO and AFNOR site installation. It entitles
technical committees…) as well as AREVA to work with countries
in codification and standardization recognizing American
bodies’ activities. regulations.
INDEX çç Sommaire
3 QUALITY AND INDUSTRIAL PROCESS
For AREVA, the renewed interest in nuclear energy will concern two main market segments :
deployment of new nuclear power plants and replacement of equipment of existing nuclear power
plants.
A unique set of manufacturing capacities
AREVA’s Equipment Business Unit provides forged and mechanical components for the nuclear
island, designed and manufactured at industrial sites around the world:
• Heavy components: steam generators, pressurizers, vessels, heads, internal vessel equipment,
and support structures. These components are at the core of nuclear steam production
• Mobile components: primary reactor pumps cooling the main reactor circuit, and control rod
drive mechanisms enabling precise reactor core control
• Large forged, molded and machined parts used to manufacture heavy components for the
nuclear island and in the refining industries (e.g. petrochemicals)
Improvement of production capacity with a
continuously increasing quality is a real
industrial challenge for any industry, but it
is even more important for the nuclear
industry considering the potential
consequences on safety of failures.
Considering the cost of non conformances,
mastering of processes and compliance
with the highest quality standards must be
achieved at the first attempt. The flawless
production of components weighing up to
several hundred tons and requiring a
precision of a few microns is achieved by
AREVA thanks to controlled procurement
and manufacturing processes, continuous
quality control, traceability as well as
rigorous management of repairs should non EPR PRESSURIZER. CLADDING OF SHELL SECTION. AREVA, CHALON SAINT MARCEL
MANUFACTURING PLANT (FRANCE) – Copyright Quatrain René
conformances be detected during the
manufacturing process.
AREVA’s components are designed and constructed following the rules of proven design and
construction codes such as RCC-M developed by AFCEN in the 1970s. RCC-M is regularly up-dated
and improved to integrate actual practices, latest industrial developments and operational experience in
order to maintain it as an effective tool allowing achievement of the required quality at the right cost.
For example, following the performance of tests by AREVA Technical Center, AFCEN decided to
develop a non mandatory annex to RCC-M which should render use of Time of Flight Diffraction
(TOFD) technique a possible alternate solution to radiographic testing.
In addition to code improvement, AREVA is pursuing a policy for continuous innovation to maintain
and improve competitiveness of our products, exploring wherever possible the use of new materials,
modelling and digital simulation applications, use or development of even more reliable and more
precise manufacturing systems, development of increasingly rigorous and efficient equipment
inspection techniques. Better understanding material properties and damaging mechanisms (e.g.
environmental fatigue), improving efficiency of destructive and non-destructive examination
techniques are objectives allowing to maintain or increase safety margins and quality while reducing
uncertainties.
INDEX çç Sommaire
An international presence
Efforts invested in the development of processes and quality must be combined with efforts to increase
production capacity in order to provide dynamic and quality response to market needs. Extending the
Equipment business unit over several industrial sites internationally allows it to secure reliable
component procurement on behalf of AREVA’s customers. To be closer to its Clients and achieve
better industrial flexibility, AREVA’s strategy consists in developing international industrial
localisation through construction of new factories or partnerships with or acquisition of companies
comprised of recognized skills in the manufacture of pressure equipment.
In France:
• The Saint-Marcel plant, near Chalon-sur-Saône, manufactures heavy nuclear components,
• The JSPM plant, in northern France, produces mobile components for nuclear power plants,
• The casting and machining workshops of Creusot Forge and Creusot Mécaniques, located in Le
Creusot, Saône-et-Loire, produce the large forged and cast parts used in the manufacture of the
heavy components,
• The SOMANU subsidiary, located in Maubeuge maintains nuclear equipment in a hot
workshop.
International:
• AREVA DONGFANG (ADJV), in Deyang, China: this joint venture, formed with DongFang
Electrical Machinery (DFEM) in 2005, manufactures JSPM-designed reactor coolant pumps
for the Chinese market.
• AREVA Newport News LLC, in Newport News, United States (Virginia): this joint venture
was created by AREVA and Northrop Grumman Shipbuilding to build a new world-class
industrial site for manufacturing heavy components for the American market as well as for
other projects. This new plant is in construction phase.
With the same logic for anticipation,
AREVA secured the procurement of
NOZZLE SUPPORT RING OF THE REACTOR
critical components such as very large VESSEL FOR THE EPR REACTOR; JSW
MANUFACTURING PLANT -JAPAN - Copyright
forgings for heavy primary JAPAN STEEL WORKS
INDEX çç Sommaire
Among them are
INDEX çç Sommaire
implementation of this scheme ensuring reliable and appropriate quality level. The objective is to
satisfy the expectations and requirements of any safety authority overseeing the subsequent project.
AREVA is also confident that benefits for safety worldwide derived from such practice should
encourage implementation of a harmonized international regulatory framework.
expertise networks within the group, with a particular focus on integration of new recruits as well as of
young talents through internship, PhD or through “block-release” training where successive periods of
school courses and work periods in our premises are repeated.
Several hundred of experienced engineers are identified as “expert” of different levels, based on their
experience, their renown and their contribution to the business. Their missions include training young
recruits as well as participation to research and development.
Partnerships with education sectors were also established, resulting in the creation of professional
curricula and specific training schools, in particular for welders or forge-masters.
To ensure delivering safe, reliable and competitive nuclear equipments, AREVA also implemented a
safety culture development policy, considering individual as well as organizational factors. 3 levels (as
defined by IAEA) and 10 elements sustaining safety culture were defined in a guide applicable to all
production and construction sites. Safety culture is also a factor taken into consideration when
selecting and qualifying suppliers and sub-contractors.
AREVA’s policy for safety culture and for knowledge & skill preservation and development is not
limited to production facilities and Technical Centers. Several companies, suppliers or sub-contractors
are directly concerned with the program developed by AREVA. A policy for deployment of a new
skilled workforce was implemented and deployed with our partners.
The volume of AREVA’s procurements and the rigorous quality requirements associated to the
delivered products participate in the preservation and the development of capabilities and skills at our
suppliers’ and more generally benefits to the whole pressure equipment manufacturing sector (forging,
casting, boiler-making, metallurgy) by allowing specialized companies to maintain a significant
volume of quality and skilled activities.
INDEX çç Sommaire
5 CONCLUSIONS
In order to provide the safest, most reliable, timely and economical solutions to its current and future
Clients, AREVA has implemented a comprehensive development program. This program is based on
technologies which ensure the best combination of industrial performances and safety level, supported
by a rigorous quality and expertise management program, and relying on a unique set of engineering
skills and manufacturing capacities.
To offer wholly sustainable and environmentally responsible power generation solutions to its
customers, AREVA developed an integrated and unique model to complement its nuclear plant
technologies with an offer of products and services across the whole nuclear cycle, from uranium
mining all the way to recycling.
6 BIBLIOGRAPHY
[1] International Atomic Energy Agency, Common User Considerations by Developing Countries
for Future Nuclear Energy Systems : Reports of Stage 1, NP-T-2.1, IAEA Vienna (2009).
[2] International Nuclear Safety Group, Stakeholder Involvement in Nuclear Issues, INSAG-20,
IAEA Vienna (2006).
[3] International Atomic Energy Agency, Fundamental Safety Principles, IAEA Safety Standards
Series No. SF-1, IAEA Vienna (2006).
[4] World Nuclear Association, International Standardization of Nuclear Reactor Designs, A
proposal by the World Nuclear Association’s Working Group on Cooperation in Reactor Design
Evaluation and Licencing (January 2010).
[5] Western European Nuclear Regulators’ Association, Safety Objectives for New Power Reactors,
Study by WENRA Reactor Harmonization Working Group (December 2009).
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page122
RÉSUMÉ / ABSTRACT
108SP
LE MAINTIEN DES COMPÉTENCES POUR LE SUIVI EN SERVICE DE
L’INTÉGRITÉ DES ÉQUIPEMENTS
Ph. Prudhon
(UIC – F 92800 Paris La Défense cedex - e-mail : [email protected])
RÉSUMÉ
Les clés du succès du maintien de l’intégrité des équipements nécessitent plusieurs conditions : un ni-
veau de formation des salariés élevé, chaque acteur (producteur maintenance…) est impliqué dans le
processus, un système de reconnaissance et d’habilitation est mis en place, la traçabilité est assurée,
une culture du retour d’expérience développée… De plus, le système de management, mis en place,
favorise la détection des anomalies, leur résolution et la mesure de l’efficacité des actions.
L’UIC collecte des incidents auprès de ses adhérents et diffuse les leçons tirées pour l’ensemble de la
profession, de plus des guides professionnels sont rédigés et diffusés. L’UIC publie un nouveau guide
technique et révise un à deux guides par an avec une collection de 90 guides.
ABSTRACT
Jeudi
The key of success of the preservation of the skills for the follow-up in service of the integrity of equip-
ment require several conditions: a high level of training of workers, each actor (production, mainte-
nance...) is involved in the process, certification procedure is applied, traceability is mastered, a culture
of analysis of incidents is applied... In these conditions, the safety management favours the anomalies
detection, their solution and the measure of efficiency.
The UIC has gathered incidents to its members and spread the learning to the whole chemical industry
moreover; technical guidelines are written and spread. The UIC publishes a new guideline once a year
and one/two guides a year are reviewed among a collection of 90 guides
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page123
During the Symposium, posters will be presented inside the exhibition, all attendees will can
meet with authors involved.
18P
ZHAO Z. / BOURGES Ph. / PILLOT S. INDUSTEEL CREUSOT
CHAUVY C. / COUDREUSE L. INDUSTEEL LOIRE
Tôles fortes en Nuance 91 pour application aux réacteurs nucléaires de la nouvelle génération / Thick
Plates in Grade 91 for new generation Nuclear Reactors
36P
JASON L. CEA, DEN, DM2S, SEMT, LM2S, CEA SACLAY
MALO J.Y. CEA, DEN DER, CEA CADARACHE
PLANCQ D. GALLE C. CEA, DEN, DER, SESI, CEA CADARACHE
GHAVAMIAN S. CEA, DEN, DER, SESI, CEA CADARACHE NECS
Caissons en béton précontraint comme solutions alternatives au circuit primaire de référence des GFR.
Etude du comportement mécanique / Mechanical behaviour of prestressed reinforced concrete vessels
as an option to replace GFR reference primary circuit
38P
CORPACE F. CEA DEN/DANS/DM2S/SEMT/LTA ICMCB, CNRS, UPR9048
MONNIER A. CEA DEN/DANS/DM2S/SEMT/LTA
MANAUD J.P. ICMCB, CNRS, UPR9048
POULON QUINTIN A. ICMCB, CNRS, UPR9048
Evaluation de l’influence des paramètres opératoires du soudage par résistance de gaines combustibles
en acier ODS pour les RNRNA / Evaluation of the influances of the resistance upset welding process
parameters applied to the claddings made in ODS steels for SFR
49P
BOND S. – VP OF TECHNOLOGY / FLEXITALLIC
POTTIER P. / LABRE B. – SIEM SUPRANITE
Introduction à un concept innovant de joint spirale / Introduction to a new concept of spiral wound
gasket
59P
FAIDY C. EDF SEPTEN
Un nouveau challenge pour les exploitants : les conditions de fonctionnement et critères complémen
taires en liaison avec la réglementation ESPESPN / A new challenge for utilities : Operating conditions
and complementary requirements in connection with ESP – ESPN regulation
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page124
67P
BLETTNER A. / CHAUVEAU D. INSTITUT DE SOUDURE INDUSTRIE
Retour d’expérience concernant l’application de techniques de gammagraphie de proximité /
Feed back of experience on closerange gammagraphy technique applications
79P
ROUSSEAU M. / J.C. BOYERE / J.F. ADAM – AFS
La part de la France en qualification soudage dans le monde / France participation in welding
qualification all over the world
80P
ROGUIN P. – ESSA
ESSA Ecole Supérieure du Soudage et de ses Applications
Présentation des travaux des étudiants ayant reçu une bourse AFIAP 20092010 / Presentation of
students having received an award for the year 20092010
92P
VITTECOQ D. – EYNARD ROBIN
Etanchéité haute performance / High performance sealing
93P
J.P. LONGIN APAVE
P. KADZIOLA CERTIFER
POSTERS
96P
BRIOTTET L. / SICARDY O. / CHU I. / LEIBOLD J.M. / BEDEL L. CEA/DRT/LITEN
BOURGUIGNON D. / DEPREZ B. CEA/DEN/DER
Développement, qualification et réalisation de jonctions Zircaloy/acier austénitique par soudage
diffusion pour le réacteur de recherche cabri / Development, qualification and production of junctions
Zircaloy/austenitic steel by weldingdiffusion for the research cabri
97P
BOURGUIGNON D. / DEPREZ B. / PONT T. CEA/DEN/DER
LEBLANC Y. AREVA TA
Réalisation d’une enceinte sous pression en zircaloy pour le réacteur de recherche Cabri / Reazilation
of a Zircaloy pressure vessel for the research reactor Cabri
102P
WEYN A. / VAST G. / RUTZE E. – VINCOTTE
Evaluation de la conformité de l’enceinte à vide d’ITER aux exigences de l’arrêté du 12 décembre 2005
relatif aux ESPN / Conformity Assessment of Iter Vacuum Vessel versus Essential Safety Requirements
of124
French Order dated December 12, 2005 concerning Nuclear Pressure Equipment
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page125
109P
PITROU B. SNCT
LOTH C. SOCIETE LJC
GARE M. PONTICELLI
DURAND S. SOCIETE MECA
Etudes de calcul pour la validation des assemblages à brides des réseaux SEC et SRU de l’EPR
FLAMANVILLE / Studies for the validation of the design of flanges used for the EPR Flamanville SEC
and SRU pipes network
110P
ZIMMER S. – INSTITUT DE SOUDURE INDUTRIE
Le soudage par Friction malaxage et les équipements sous pression / Friction Stir Welding and Pressure
equipment
113P
DAHMENE F. CETIM
Evaluation de la conformité de l’enceinte à vide d’ITER aux exigences de l’arrêté du 12 décembre 2005
relatif aux ESPN / Conformity Assessment of Iter Vacuum Vessel versus Essential Safety Requirements
of French Order dated December 12, 2005 concerning Nuclear Pressure Equipment
115P
NOTIN A. / BESSON J. / GOURGUES A.F. LABORATOIRE ROBERVAL – UNIVERSITE DE TECHNOLOGIE
DE COMPIEGNE
Evaluation à moindre coût de la fiabilité des structures sollicitées en fatigue / Assessing cost structural
relaibility sought fatigue
117P
HILKES J. / GROSS V. – BÖHLER SCHWEISSTECHNIK DEUTSCHLAND GMBH
LEGAIT P.-A. – GROUPE BÖHLER SOUDAGE FRANCE SAS
Soudage d’aciers au CrMo pour la production d’énergie, le raffinage de pétrole et la pétrochimie —
passé, présent, futur / Welding CrMo steels for power generation and petrochemical applications —
past, present, future
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page126
RÉSUMÉ / ABSTRACT
18P
TÔLES FORTES EN NUANCE 91 POUR APPLICATION AUX RÉACTEURS
NUCLÉAIRES DE LA NOUVELLE GÉNÉRATION
RÉSUMÉ
Le devenir énergétique de la planète se voit menacé par la raréfaction des combustibles fossiles. Il est
pourtant capital de répondre aux besoins croissants avec une énergie sûre, économique et respectueuse
de l’environnement. L’énergie nucléaire propose une solution attractive. Dans ce cadre, des acteurs
majeurs comme la France ou les Etats-Unis envisagent les réacteurs dits de génération IV. Ceux-ci
comprennent entre autres les Réacteurs à Neutrons Rapides (RNR) au sodium ou au gaz, et les réacteurs
à haute température HTR. Au sein d’un RNR, la température moyenne des circuits primaires est de
550°C et le système HTR fonctionne en moyenne à 450°C. Ceci impose des donc nuances fortement
alliées au chrome-molybdène et en particulier le grade 91 (A387 Gr91 / EN10028-2 X10CrMoVNb9-
1), initialement conçue d’ailleurs pour ces applications.
Afin de répondre aux futurs besoin du secteur nucléaire, INDUSTEEL a développé une version modifiée
POSTERS
ABSTRACT
The future of global energy supply is under threat from the depletion of fossil fuels. It is therefore essential
to satisfy growing needs with a safe and cost-effective energy in keeping with environmental conservation.
Nuclear energy seems to be the most suitable way to meet this growing demand for energy safely and
economically. The main nuclear energy users namely France and United States are involved in a research
program to fourth-generation reactor technology. Since 2000 to nowadays, France decided to focus its
efforts on the Sodium Fast Reactor and Gas Fast Reactor technologies, whereas the US selected the
technology of High Temperature Reactor. These two technologies work at high temperatures between
450°C to 550°C in average. Extensive studies have been launched to assess steel grades able to meet
these new conditions. One of the candidate grades is Vanadium modified 9% Chromium (EN 10028-2
X11CrMoVNb 9-1 - ASTM/ASME A387 Gr 91).
To meet the future needs in nuclear energy, INDUSTEEL improved its conventional 9Cr1MoVNb steel,
initially dedicated to thermal power plants. Preliminary studies revealed the feasibility of thick plates in
this enhanced grade. INDUSTEEL fabricated a 200 mm thick plate via electric furnace process. Plates
have then been hot rolled and heat-treated. Tests performed on both base metal and weldment proved the
excellent mechanical properties of the grade, particularly regarding toughness property. This work
demonstrated the industrial feasibility of thick plate in grade 91 for nuclear applications.
INDEX çç Sommaire
Grade 91 Thick Plates For New Generation Nuclear Reactors
RESUME
Afin de répondre aux futurs besoins du secteur nucléaire, INDUSTEEL a développé une
version modifiée de sa nuance 91 actuellement destinée principalement aux centrales
thermiques. Des travaux préliminaires ont démontré la faisabilité industrielle de cette nuance
modifiée nucléaire sous forme de tôles fortes. Après élaboration au four électrique d'un lingot,
préforgeage et laminage à chaud, INDUSTEEL a traité des tôles d'épaisseur jusqu'à 210 mm.
Les essais de caractérisation réalisés ont mis en évidence d’excellentes propriétés mécaniques
et en particulier en ténacité tant à l’initiation qu’en arrêt de fissure. Ceci garantit la capacité
industrielle à fabriquer des tôles pour des ensembles chaudronnées de grandes dimensions
répondant aux impositions sévères de l’industrie nucléaire
ABSTRACT
The future of global energy supplies is under threat from the depletion of fossil fuels. It is
therefore essential to satisfy growing needs with a safe and cost-effective energy in keeping
with environmental conservation. Nuclear energy is one of the most suitable ways to meet this
growing demand for energy safely and economically. The main nuclear energy users namely
France and United-States are devoted to a research program to fourth-generation reactor
INDEX çç Sommaire
technology. France decided to focus their efforts on the Sodium Fast Reactor and Gas Fast
Reactor technologies, whereas US selected the technology of High Temperature Reactor.
These two technologies work at high temperatures between 450°C to 550°C in average.
Extensive studies have been launched to assess steel grades that could meet these new
conditions. One of the grades of candidate is the 9% of Chromium steels (X10CrMoVNb9-1or
A387 Gr91).
To meet the future needs in nuclear energy, INDUSTEEL improved its current Vanadium
enhanced 9% of Chromium steel (A387 Gr91), initially dedicated to thermal power plants.
Preliminary studies revealed the feasibility of thick plates in this improved grade.
INDUSTEEL fabricated up to 210 mm thick plate via electric furnace processing. Plates have
been then preforged, hot rolled and heat-treated. Test performed on both base metal and
weldment proved the excellent mechanical properties of this improved grade, particularly
regarding toughness property. This work demonstrated the industrial feasibility of thick plate
made of grade 91 for nuclear applications.
INTRODUCTION
Taking into account the expected increase in energy demand worldwide and the growing
awareness about global warming, climate change issues and sustainable development, nuclear
energy will be necessary to meet future global energy demand. That is why the next
generation nuclear energy systems (Generation IV) will provide sustainable energy in
compliancy with clean air objectives. Furthermore, these new reactors will also provide long-
term availability of systems and effective fuel utilization for worldwide energy production [1].
There are many challenges to overcome before Generation IV reactors become commercially
viable. Some of the most important challenges concern the identification and development of
materials and alloys that can safely operate under the corresponding environmental conditions
present in Gen IV reactors. Indeed, this new generation of reactors will work at higher
temperature than current Gen II and Gen III reactors. Extensive studies have been launched
worldwide to assess steel grades able to fulfil these new operating conditions. One of the best
candidates is the Vanadium-modified 9% Chromium steel (EN 10028-2 X11CrMoVNb9-1,
ASTM/ASME SA387 grade 91 class 2) defined in the early 1980’s by ORNL for nuclear
purposes.
Thick plates, 140 mm and 210 mm thick, were hot rolled from 82 metric tons ingot. Welded
joints have been prepared. Tests performed on both base material and welded zones exhibit
excellent mechanical properties of the grade, especially regarding toughness property. This
contribution is a review of the work done as well as the result obtained on the plates and
weldment.
The production route followed by Industeel to produce this material is one conventionally
used for high quality products such as conventional nuclear grades. It started by a rigorous
selection of the scraps to lower the impurity level. This first stage allows at the same time a
strong mastering of the chemical content in disturbing elements (i.e. with high activation
capacities).
INDEX çç Sommaire
This step is immediately followed by a dephosphoration stage in the electric furnace to
achieve very low phosphorus content (below 30ppm at this stage to achieve a final 80ppm
max taking into account phosphorus uptake coming from the fero additions). Desulfurization
stage is done in the refining ladle. Extra low sulphur content (below 20ppm) is then obtained
with this process. The chemical balance of alloying elements is also done in the refining ladle.
At the end of the refining process, the heat is transferred to the degassing unit to remove
hydrogen. For nuclear applications, the RH unit permits to reach extremely low hydrogen
content (below 1ppm) and then to avoid the risk of crack appearance in the ingot and in the
plates after rolling.
BASE MATERIAL
Chemical Composition
The chemistry of the ingot and then of the plates was set up to fulfill the requirements of both
ASTM/ASME and ISO standards. Table 1 summarizes the requirements of the standards or
defined on the basis of Industeel’s experience and the results of the chemical analysis
performed on the 210 mm thick plate. 3 measurements were done one at mid thickness and at
quarter thickness of the plate to evaluate segregations. As demonstrated by these
measurements, segregation level is very low. It is well known that, when carbide forming
elements (Cr, Mo, V, Nb…) are high these steels may be more susceptible to reheat cracking
during PWHT, so a special attention was paid in the balance of these elements and of
INDEX çç Sommaire
impurities (S, P). Furthermore, these elements were also balanced, keeping in mind the
toughness optimization of the grade.
From thermal plant experiences, P91 could be submitted to type IV cracking issues (soft zone
delayed creep cracking). Statistical assessment of this population showed that very low N:Al
ratio can lead to very quick failure due to strain accumulation in the HAZ. From those studies
[1], a critical level of N:Al ratio has been defined. In general, the target is above 3. Most
recently, EPRI tends to define a target at 4 with no value below 2.5.
Fully tempered martensitic microstructure was obtained after heat treatments. Figures 3 to 5
show the typical microstructure obtained on the 210mm thick plate respectively close to the
skin, at quarter and mid-thickness.
Micrographic examinations performed on the plates confirm the absence of major segregation
as demonstrated before by the chemical analyses. These examinations also demonstrate the
absence of delta ferrite in the products, even at mid thickness, thanks to the balance of
chemistry. The homogeneity of the product is also confirmed by through thickness hardness
measurements performed concurrently and represented in figure 6. HV10 values are in every
case in the range 188-205 Vickers.
INDEX çç Sommaire
Figure 3: Microstructure of sub-surface Figure 4: Microstructure of quarter-thickness
300
275
250
HV10
225
200
175
150
0 20 40 60 80 100 120 140 160 180 200
d (mm)
Mechanical properties were evaluated for each plate in as-delivered conditions and after heat
treatment corresponding to PWHT (Post Wed Heat Treatment). The latter treatment
corresponds to As-Delivered treatment followed by a supplementary tempering at 750°C for
12 hours.
Tensile tests at mid thickness were performed on 210mm thick plate from room temperature
up to 800°C by step of 50°C. The results are summarized in figure 7 and compared to EN
INDEX çç Sommaire
10028-2 X11CrMoVNb9-1 minimum strength values. It appears clearly that even after long
tempering, tensile properties remains compliant with the requirement of this standard. The
results are also in accordance with ASTM/ASME SA387 grade 91 minimum strength values
(YS > 415MPa and UTS in the range 585-760MPa).
Furthermore, homogeneity of tensile values was checked through the thickness of the plates.
Figure 8 gives an overview of the results obtained at room temperature, 500 and 590°C. The
latter two temperatures correspond to the lowest and highest temperatures that should face
Vanadium-modified 9%Cr during service (including thermal excursions). As expected,
regarding microstructure and hardness measurements, the tensile properties are very
homogeneous too.
700
YSmin - NF EN 100028-2 X11CrMoVNb9-1
YS 210mm - As delivered
600 YS 210mm - As delivered + PWHT
UTS 210mm - As delivered
UTS 210mm - As delivered + PWHT
500
YS/UTS ( MPa)
400
300
200
100
0
0 100 200 300 400 500 600 700 800 900
Temperature (°C)
A387 Grade 91 - Base material thickness 140 & 210mm after PWHT
500
YS mini = 435MPa Sub surface 140mm
1/4 thickness 140mm
450
1/2 thickness 140mm
Sub surface 210mm
400
YS mini = 316 MPa 1/4 thickness 210mm
1/2 thickness 210mm
350
300
YS (MPa)
250
200
150
100
50
0
RT 500°C 590°C
Testing temperature
Figure 8: Homogeneity of Yield Strength through thickness after PWHT (750°C - 12 hrs).
INDEX çç Sommaire
Charpy V-Notch tests results
All results obtained on 140 and 210mm thick plates after PWHT are summarized in table 2
hereafter. The tested materials exhibit in every case very low transition temperatures (Fracture
Appearance Transition Temperature – FATT). FATT is always below -24°C, corresponding
to impact energy above 100J. CVN values at room temperature are in every case above 175J
(minimum value measured at mid-thickness in transverse direction of the 140mm thick plate).
This insures a ductile behaviour of the material even at room temperature. This point will be
assessed more deeply in the following sections by fracture mechanics analyses.
250 250
Mid thickness longitudinal Mid thickness transverse
Quarter Thickness longitudinal Quarter Thickness transverse
Sub surface longitudinal Sub surface transverse
200 200
Im pact Energy (J)
150 150
100 100
50 50
0 0
-120 -100 -80 -60 -40 -20 0 20 40 -120 -100 -80 -60 -40 -20 0 20 40
Figure 9: CVN properties after PWHT Figure 10: CVN properties after PWHT
(750°C - 12 hrs) in longitudinal orientation. (750°C - 12 hrs) in transverse orientation.
INDEX çç Sommaire
Kv Kv
Kv (J)
(J) (J)
Code at KV
at - at
+20°C (J)
Thickness 20°C 0°C FATT (°C) Tk 28J (°C) Tk 41J (°C) Tk 54J (°C)
Upper
NF EN 10028-2 > 27J > 34J > 40J
Shelf
moyenne > 40 J > 60 J
RCC-MR > 72 J
individuelle > 28 J > 42 J
longitudinal 179 211 223 -46 -79 -72 -66 223
Sub surface
transverse 156 194 213 -39 -88 -75 -61 213
longitudinal 149 198 216 -34 -75 -67 -59 216
140 mm 1/4 thickness
transverse 153 198 209 -36 -75 -67 -59 209
longitudinal 164 192 206 -41 -81 -72 -64 206
1/2 thickness
transverse 103 158 178 -24 -78 -67 -57 178
longitudinal 176 210 230 -46 -84 -70 -76 230
Sub surface
transverse 176 210 224 -46 -75 -60 -76 224
longitudinal 176 210 221 -37 -103 -88 -76 221
210 mm 1/4 thickness
transverse 158 197 212 -35 -95 -79 -68 212
longitudinal 149 195 221 -31 -103 -88 -59 221
1/2 thickness
transverse 113 160 180 -25 -101 -86 -49 227
Table 2 : CVN requirements and obtained values on both 140mm and 210mm thick plates
ADVANCED MECHANICAL TESTS ON BASE MATERIAL
Fracture mechanics tests were performed on the plates to assess both quasi-static properties
and crack arrest behaviour of the material. The creep behaviour of the material has also been
investigated.
Figure 11 represents the static toughness mastercurves obtained on the plates at quarter and
mid-thickness in transverse orientation (T-L orientation as defined by ASTM E399 [2]).
Fracture mechanics results are in total accordance with the CVN values and reference
temperatures (To) estimated by Wallin’s rule in ASTM E1921 [3,4] standard and given by
Equation 1:
T0 = TK28 J + C (1)
250.0
200.0
150.0 Pf 1%
100.0
50.0
0.0
-80 -60 -40 -20 0 20 40 60 80
T-T0 (°C)
INDEX çç Sommaire
A typical Instrumented CVN curve is given by figure 12. Several specimens must be broken
at various temperatures in the brittle to ductile transition. A measurement of the evolution of
the force at arrest of crack propagation (symbolized by the red circle in figure 12) with
temperature permits to determine TFa4kN or TFa4,12kN (i.e. the temperature to get the arrest
force equals to respectively 4kN and 4,12kN).
25000 250
20000 200
15000 150
Force (N)
Energie (J)
10000 100
Deflexion / Force
Fa = 4000N Deflexion / Actual impact Energy
5000 50
Deflexion / Appearante Energy
0 0
0 0.005 0.01 0.015 0.02 0.025
Déflexion (mm)
Figure 12: Typical Force-Displacement curve obtained with instrumented Charpy test.
Then correlations given by Equations 3 and 4 can be applied to estimate NDTT. The results
obtained on the 140 mm thick plates are summarized in table 3 hereafter.
From these measurements and estimates, it appears that NDTT transition is in the classical
range of those obtained in the case of SA508/SA533 thick products.
Cross analyses of both NDTT temperatures and CVN results leads to the estimation of RTNDT
(Reference Temperature for Nil-Ductility Transition). The RTNDT is defined as being greater
or equal to the NDTT; at RTNDT plus 33°C, three CVN specimen results must exhibits a
minimum of 68J of absorbed energy and 0,89 mm of lateral expansion. In many cases (i.e.
with conventional nuclear grades), RTNDT is driven by CVN values, but in this case, as it is
generally the case with Cr-Mo steel grade, RTNDT is driven by NDT and then RTNDT is equal
to NDT.
From this conclusion, a confrontation of fracture mechanics data with KIR curve can be done.
The equation of KIR curve is given by Equation 4.
INDEX çç Sommaire
220
K IR = min
36,5 + 3,1 exp[0,036.(T − RTNDT + 55,5)]
(4)
Figure 13 represents the fracture mechanics values obtained on both plates confronted with
KIR curve. From this figure, it can be concluded that KIR curve is very conservative in the
case of quasi-static fracture toughness. Then, designed based on KIR curve are really safe.
Further investigation based on dynamic behaviour is necessary. Nevertheless, the safety
margin highlighted by this figure allows a certain level of confidence on that particular
subject.
ASTM/ASME SA387
350 grade 91 Nuclear
140mm and 210mm
300
KJc25 (MPa.m 1/2 )
250
200
KIR
150
100
50
0
-100 -80 -60 -40 -20 0 20 40 60 80
T-RTNDT (°C)
Figure 13: Quasi static fracture mechanics data superimposed on KIR curve.
Creep Properties
One other main points of the bidbook was to insure good creep properties of the produced
material. Then creep tests have been launched to verify this point. As it is generally difficult
to conciliate toughness requirements with creep resistance, it has been chosen from the very
beginning to target a creep resistance between ECCC mean curve and ECCC mean curve
minus 20%. The specimens have been taken in the 210 mm thick plate at quarter thickness in
the transverse orientation. Some tests are still running, as summarized in table 4. This set of
tests consists in a reduced preliminary creep program and further tests will be launched soon
to complete the database.
Temperature Stress
Larson Miller Time to rupture
El% RoA%
Parameter
(°C) (MPa) (h)
(C=36)
550 224 132 32,34 27 86
550 198 806 32,92 34 89
600 153 218 34,05 38 93
600 131 1890 34,66 27 86
600 123 3401 34,92 27 87
600 109 8110 (Running) 35,34 (running) 2,6 (running) -
600 100 8110 (Running) 35,66 (running) 2,1 (running) -
Table 4: Creep test results obtained in the 210 mm thick plate (1/4 th, transverse orientation).
INDEX çç Sommaire
Creep results obtained up to now are in accordance with creep-optimized grade 91 as referred
by ECCC (Figure 14) and proved that pretty good toughness resistance can be achieved
without strong degradation of creep performances.
Stress (MPa)
100
Predich
CECA
10
29.5 30 30.5 31 31.5 32 32.5 33 33.5 34 34.5 35 35.5 36
PLM (C=36)
The obtained creep elongation curves are illustrated in Figure 15. Ductility level of the
material is in accordance with the results commonly obtained on grade 91. At the onset of
specimen fracture, no elongation below to 27% has been measured.
50
550°C - 198MPa
45
550°C - 224MPa
600°C - 153MPa
600°C - 123MPa
600°C - 131MPa
40
35
Elongation (%)
30
25
20
600°C - 109MPa
15
600°C - 100MPa
10
0
0 1000 2000 3000 4000 5000 6000 7000 8000
Time (Hrs)
A full thickness (210 mm thick plate) Submerged Arc Welding (SAW) narrow gap joint has
been carried out using a commercially available conventional grade 91 filler product. It means
that this filler material is not optimized for toughness purpose but for creep resistance. That is
the reason why all attention was focused on the Heat Affected Zone (HAZ).
INDEX çç Sommaire
Realization of the welded joint
Filler material has been ordered following the requirements given in Table 5. The welding
conditions as well as the welding sequence are summarized in Table 6.
Figure 16 presents the macrograph of the welded joint made on the 210 mm thick plate. Over
180 beads were deposited to complete the joint (62 layers with 3 beads per layer)
Upper
skin
¼ th
210 mm
½ th
¾ th
Root
skin
Figure 16: Macrograph of the 210mm thick SAW Narrow Gap welded joint.
INDEX çç Sommaire
During the welding process, no particular difficulties have been noticed. Although 9% of
chromium steels are known for hot-cracking sensitivity, no problem was encountered in that
case. The welded joint has been finally controlled by X-ray radiography as it is usually done
in nuclear industry.
Mechanical properties of the welded joints have been evaluated after the Post Welded Heat
Treatment (PWHT) corresponding to a tempering treatment performed at 750°C for 12 hours.
Figure 17 presents the hardness evolution through the welded joint in as-welded conditions
and after PWHT. The hardness measurements show clearly the the PWHT performed at
750°C for 12 hours is sufficient to obtain hardness values below to 250 HV10 throughout.
330
280
230
180
0 20 40 60 80 100 120 140 160 180 200
d(mm)
Cross weld tensile tests have been carried out at room temperature and at service temperature
(500°C). These results (Table 7) exhibit ultimate tensile strength (UTS) values in accordance
with the requirements of the standard (in the range 585 - 760 MPa). For each specimen,
rupture occurs in the sub-critical heat affected zone (SC-HAZ), corresponding to the over-
tempered zone of base material due to the welding sequences.
In addition, complete CVN transition curves have been plotted for each zone of the welded
joint using Instrumented Charpy tests. From these experiments, transition temperatures have
been estimated as well as NDT. Actual measurements of these values are planed as a next step
of the study. The results are summarized in tables 8 and 9. Table 8 presents all the
measurements done using Charpy V-Notch tests while table 9 gives the estimation of To and
NDTT.
INDEX çç Sommaire
Kv (J) Kv (J) Kv (J)
Code
at -20°C at 0°C at +20°C KV (J)
Thickness NF EN 10028-2 > 27J > 34J > 40J FATT (°C) Tk 28J (°C) Tk 41J (°C) Tk 54J (°C) Upper
moyenne > 40 J > 60 J Shelf
RCC-MR > 72 J
individuelle > 28 J > 42 J
Fusion Line -73°C -66°C -61°C
208J 236 242J -58°C 240J
210 mm (FL)
¼ Fusion Line +1mm -59°C -102°C -93°C -85°C
224J 242 250J 250J
thickness (FL+1mm)
of the Fusion Line +3mm -88°C -75°C -65°C
150J 193J 218J -40°C 250J
SAW NG (FL+3mm)
joint Fusion Line +5mm -85°C -75°C -67°C
154J 194J 220J -40°C 250J
(FL+5mm)
Table 8: CVN requirements and obtained values on 210mm thick welded joint
250
Fusion Line
Fusion Line + 1mm
Fusion Line + 3mm
200
Fusion Line + 5mm
150
100
50
0
-120 -100 -80 -60 -40 -20 0 20 40
Temperature (°C)
CONCLUSIONS
This paper presents the investigations conducted by Industeel on modified 9Cr1Mo (EN
10028-2 X11CrMoVNb 9-1 - ASTM/ASME SA387 grade 91) in the shape of very thick
plates devoted to the nuclear market.
To meet the future needs in nuclear energy, Industeel improved its conventional 9Cr1MoVNb
steel devoted to the fabrication of steam lines in thermal power plants. Preliminary studies
revealed the feasibility of thick plates in this enhanced grade. Then Industeel launched the
fabrication of one 140mm and one 210mm thick plates via electric furnace process. Plates
have then been preforged, hot rolled and heat-treated. Tests performed on both base metal and
welded zones proved the excellent mechanical properties of the grade, especially regarding
toughness property.
This work demonstrates the industrial feasibility of very thick plates made of grade 91 for
nuclear applications with very good toughness without degradation of high temperature creep
resistance.
Nevertheless, developments are still running to propose filler materials able to reach
properties close to those of base material. Further investigations are also continuing on high
temperature fracture mechanics, creep and ageing.
INDEX çç Sommaire
REFERENCES
[2] ASTM standards, ASTM E399, Standard Test Method for Linear-Elastic Plane-Strain
Fracture Toughness KIc of Metallic Materials
[4] ASTM standards, ASTM E1921, Standard Test Method for Determination of Reference
Temperature, To, for Ferritic Steels in the Transition Range
[5] K. WALLIN, Descriptive potential of Charpy-V fracture arrest parameter with respect to
crack arrest KIa, VTT-MET B-221, VTT Espoo, Finland, January 1993.
[6] K. WALLIN, Application of the master curve method to crack initiation and crack arrest,
Proceedings of ASME Pressure Vessels and Piping Conference, PVP-Volume 393, Boston,
Massachussetts, USA, August 1-5, 1999.
[7] S.K. ISKANDER et al., Use of forces from instrumented Charpy V-notch testing to
determine crack arrest toughness, ASTM STP 1325, ASTM, 1997
INDEX çç Sommaire
Research and Development
Grade P91 Thick Plates For New Generation Nuclear Reactors
Tôles fortes en nuance P91 pour application aux réacteurs nucléaires de nouvelle
génération
Z. ZHAO1, S. PILLOT1, S. CORRE1, C. CHAUVY2, L. COUDREUSE2, P. TOUSSAINT3
1 Industeel France - Site du Creusot, ArcelorMittal Group, France,
[email protected]/[email protected]/[email protected]
2 Industeel France - Site de la Loire, ArcelorMittal Group, France, [email protected] / [email protected]
3 Industeel Belgium, ArcelorMittal Group, Belgium, [email protected]
Abstract
To meet the future needs in nuclear energy, Industeel improved its conventional 9Cr1MoVNb steel originally devoted to the fabrication of steam lines in thermal power
plants. A complete characterizations program was performed on thick plates, 140 and 210mm thick, its were preforged and hot rolled. This contribution is a review of the
work done as well as the results obtained on the plates.
Résumé
Afin de répondre aux futurs besoin du secteur nucléaire, INDUSTEEL a développé une version améliorée de sa nuance 91 actuellement destinée principalement aux
centrales thermiques. Un programme de caractérisation a été mené sur les tôles fortes de 140 et 210 mm d’épais, qui ont été préforgées et laminées à chaud. Ce poster
présente les travaux réalisés et les résultats obtenus dans cette étude.
METALLURGICAL AND CONVENTIONAL MECHANICAL PROPERTIES ADVANCED MECHANICAL PROPERTIES OF BASE MATERIAL
OF BASE MATERIAL
Chemical Composition Static Fracture Toughness Mastercurve & Nil-Ductility Transition Temperature (NDTT)
C Si Mn P S Al Cr Ni Mo V Nb N Static toughness master curve A387 Grade 91 •Static fracture toughness mastercurves
400.0 obtained on the plate at quarter and mid-
¼ th 0.102 0.226 0.403 0.007 0.002 0.009 8.640 0.134 0.982 0.204 0.080 0.046 140mm - 1/2 thickness To=-110°C Pf 99% thickness in transverse orientation.
350.0 140mm - 1/4 thickness To=-115°C
½ th 0.114 0.230 0.407 0.007 0.002 0.009 8.580 0.136 0.978 0.204 0.080 0.050 210mm - 1/2 thickness To=-109°C
210mm - 1/4 thickness To=-112°C •The transition temperature T0 is below -
300.0
• Very low segregation of alloying elements and very low impurities level (Sulfure and Phorphorus) to meet
Validity range of ASTM E1921
Pf 50% 100°C even on the 210 mm thick plate at mid-
nuclear specifications requirements thickness.
KJc25 (MPa.m 1/2 )
250.0
Microstructures 150.0 Pf 1%
Measured
•Fully tempered martensitic Location
100.0
NDTT (°
(°C)
microstructure has been obtained on
50.0 Sub-
Sub-surface -35
the 210 mm thick plate
0.0
¼ thickness -25
•Absence of delta ferrite and major -80 -60 -40 -20 0 20 40 60 80
•Hardness measurement through From these measurements, it appears that NDTT transition is in the classical range of those obtained in the
thickness very homogeneous (~190- case of SA508/SA533 thick products
210HV10)
Mid-thickness
Creep Properties
Quarter thickness
Base Material - A387 grade 91 210mm thick
Tensile Tests 1000
700
YSmin - NF EN 10028-2 grade 91
YS - P91 200mm - As delivered
600
YS - P91 200mm - As delivered + PWHT ECCC mean curve •Creep results obtained up to now are in
UTS - P91 200mm - As delivered
•Tensile tests performed on 210 mm thick accordance with creep optimized grade 91 ad
500 plate at mid-thickness, from room
Stress (MPa)
100
400
condition and after PWHT (750°C-12hrs). •This proves that good toughness properties can
be achieved without inappropriate degradation
300 •Even after a long tempering heat treatment Predich
ECCC mean curve minus
CECA of creep properties.
750°C–12 hrs, tensile properties are in 20%
200 210mm thick plate
accordance with the standard EN 10028-2 210mm thick plate (running tests)
100
(X11CrMoVNb9-1)
10
29.5 30 30.5 31 31.5 32 32.5 33 33.5 34 34.5 35 35.5 36
0 PLM (C=36)
0 100 200 300 400 500 600 700 800 900
Temperature (°C)
RCC-MR > 72 J
individuals > 28 J > 42 J ½ th
(FL)
150 150
Fusion Line +1mm
224 242 250 -59 -102 -85
(FL+1mm)
Fusion Line +3mm
100 100
150 193 218 -40 -88 -65
(FL+3mm)
Fusion Line +5mm
154 194 220 -40 -85 -67
50 50 (FL+5mm)
•CVN results are obtained at quarter thickness of the welded joint after PWHT performed at 750°C for 12
0 0
-120 -100 -80 -60 -40 -20 0 20 40 -120 -100 -80 -60 -40 -20 0 20 40 hrs.
Temperature (°C) Temperature (°C)
•Very good CVN results have been obtained even in the Inter-Critical Heat Affected Zone (FL + 3mm). In
CVN transition curves after PWHT (750°C – 12hrs) in longitudinal and transverse orientations on 210 mm that case, TK28J is below -85°C, insuring very high toughness to this zone.
thick plate.
çç
Conclusions & Perspectives
INDEX
•This work demonstrates the industrial feasibility of very thick plates made of grade 91 for nuclear applications with very promising toughness properties without Sommaire
degradation of high temperature creep resistance. Furthermore, this work has als shown, that toughness properties of HAZ retain a very high toughness level.
•Developments are still running to propose filler materials better able to meet toughness properties close to those of base material. Further investigations are also going
ahead on high temperature fracture mechanics, creep and ageing properties.
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page127
RÉSUMÉ / ABSTRACT
36P
CAISSONS EN BÉTON PRÉCONTRAINT COMME SOLUTIONS ALTERNATIVES
AU CIRCUIT PRIMAIRE DE RÉFÉRENCE DES GFR. ETUDE DU
COMPORTEMENT MÉCANIQUE.
RÉSUMÉ
Le concept de réacteur à neutrons rapides refroidis au gaz (GFR) est considéré par le Commissariat à
l’Energie Atomique comme un concept prometteur, en utilisant de l’hélium comme caloporteur. L’étude
présentée ici consiste à revisiter la solution de référence proposée par le CEA en 2007, en considérant
un caisson en béton précontraint dans lequel seraient inclus tous les composants du circuit primaire.
Deux configurations sont étudiées (cylindre monocavité et cylindre multi-cavités) et soumises à des
chargements de pression et de température. Une étude thermomécanique simplifiée montre la nécessité
d’inclure de la précontrainte dans la structure si les contraintes de traction, sources d’éventuelle
fissuration, veulent être évitées. A partir des configurations initiales, le système de précontrainte
(type, section et tension des câbles) est défini. Une simulation thermomécanique en trois dimensions
est alors effectuée afin de valider le principe du concept et montre que les configurations initiales ne
parviennent pas à éviter totalement les contraintes de traction. Une étude paramétrique est alors proposée
intégrant les variables comme l’épaisseur ou la température interne du béton pour proposer une solution
optimisée et souligner la nécessité d’études complémentaires.
ABSTRACT
The Gas cooled Fast Reactor (GFR) is considered by the French Commissariat à l’Energie Atomique
as a promising concept, using Helium as coolant. The study presented in this paper consists in revisiting
the reference solution proposed at the end of 2007 by assessing a reinforced prestressed concrete vessel
to include every component of the primary circuit. Two configurations are considered (single cavity
or multiple cavities) which are submitted to pressure and temperature loadings. A simplified thermo-
mechanical study shows the necessity to include prestress if tensile stresses want to be avoided, from
which cracks may develop. From the initial configurations, the prestressing system (type, section and
tension) is defined. A three dimensional thermo-mechanical simulation is then performed and shows
that initial configurations do not manage to avoid tensile stresses. A parametric study is then proposed
considering variables like the concrete thickness or the internal concrete temperature to improve the
design, propose an optimized solution and underline the need for further studies.
INDEX çç Sommaire
Mechanical behavior of prestressed concrete vessels as on option to replace GFR reference primary
circuit
Résumé : Le concept de réacteur à neutrons rapides refroidis au gaz (GFR) est considéré par le Commissariat à
l’Energie Atomique (CEA) comme un concept prometteur, combinant les avantages du spectre rapide et de haute
température en utilisant de l’hélium comme fluide caloporteur. L’étude présentée ici consiste à revisiter la solution de
référence du circuit primaire en considérant un caisson en béton précontraint dans lequel serait inclus l’ensemble des
composants. Une étude de faisabilité est réalisée aboutissant à la proposition de deux concepts de caisson.
Abstract : The Gas cooled Fast Reactor (GFR) is considered by the French Commissariat à l’Energie Atomique as a
promising concept, combining the benefits of fast spectrum and high temperature, using Helium as coolant. The
current study consists in revisiting the reference solution for the primary circuit by assessing a reinforced prestressed
concrete vessel incorporating all components. Two configurations are proposed and analyzed.
Steel reference solution
Type of prestressing
tendons
each cavity contains one component the only cavity contains all components
(total diameter = 32 m) (total diameter = 37 m)
rib anchorage
duct (20 cm
diameter) Maximum Optimization
stress
anchorage Spacing of
about 5 cm distribution position of
prestressing
Maximum principal stress (MPa) tendons
Zoom on the anchorage
INDEX çç
Sommaire
Conclusions: Even if the mechanical feasibility has not yet been totally achieved due to some residual tensile zones, it seems nevertheless possible to
come to a solution with some additional studies on the design of the prestressed vessels. The study also underlines the necessity to take care of the
mechanical behavior on singular zones (like openings) that may be responsible for some stress localization.
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page128
RÉSUMÉ / ABSTRACT
38P
EVALUATION DE L’INFLUENCE DES PARAMÈTRES OPÉRATOIRES DU
SOUDAGE PAR RÉSISTANCE DE GAINES COMBUSTIBLES EN ACIER ODS
POUR LES RNR-NA
b
CNRS, Université Bordeaux 1, ICMCB, UPR9048, 33608 PESSAC Cedex, France
RÉSUMÉ
Les aciers ODS (Oxide Dispersion Strengthened) sont des matériaux candidats pour la réalisation du
gainage combustible des réacteurs nucléaires de Génération IV de type RNR-Na (Réacteur à Neutrons
Rapides refroidi au sodium).
Les propriétés mécaniques à haute température de ces matériaux sont assurées par une dispersion
d’oxydes nanométriques qui peut être modifiée lors du passage par l’état liquide de l’acier. Il est donc
préconisé de réaliser la soudure des bouchons du gainage par des procédés de soudage en phase solide.
Le soudage par résistance est actuellement à l’étude en tant que procédé de référence.
Dans le cadre de cette étude, une double approche simulation numérique - expérimentation a été mise
en place. Les premiers résultats numériques permettent d’identifier les phénomènes thermomécaniques
POSTERS
ABSTRACT
ODS steels (Oxide dispersion strengthened steels) are candidate materials for fuel cladding of SFR
(Sodium Fast Reactor). In these materials, the mechanical properties at high temperature are increased
by a dispersion of nano-sized oxide that can be damaged by fusion of the material. Therefore a solid
state bounding is recommended in order to weld the clad and the plug. This study is focused on the
resistance upset welding
Numerical simulation has been coupled with experiments. The simulation allows to identify identified
the predominant thermo-mechanical phenomena and shows a major heating-up in the clad part located
out of the electrode due to a high current density. The comparison between experiments and simulations
allows pointing out experimental measurements that are linked to the welding thermo-mechanical
phenomena. Studying the influences of the process parameters on these experimental measurements
linked to the thermo-mechanical phenomena, allows the optimisation of the process parameters and
an increase of the weld quality.
INDEX çç Sommaire
INTRODUCTION
En 2002, le Forum international Génération IV a retenu six concepts de réacteur. Une des
technologies étudiées en France est la technologie des Réacteurs à Neutrons Rapides avec
caloporteur sodium (RNR-Na). Le design du cœur pour ce type de réacteurs limite les
déformations admissibles et les aciers de type "austénitique avancé" préalablement utilisés sur
ce concept de réacteurs ne répondent plus au cahier des charges. Les aciers ODS (Oxide
Dispersion Strengthened) possèdent une bonne tenue au fluage thermique et au fluage sous
irradiation et constituent donc une famille de matériaux potentielle pour la réalisation du
gainage combustible de ces réacteurs. Leurs bonnes propriétés mécaniques à haute
température sont assurées par une dispersion homogène d'oxydes nanométriques au sein du
matériau. Le passage de ces aciers en phase liquide peut entraîner le regroupement des
particules d’oxyde, causant une dégradation des propriétés mécaniques [1, 2]. Par conséquent,
les procédés de soudage en phase liquide peuvent provoquer une modification de la
nanostructure et les procédés d'assemblage en phase solide sont donc préconisés.
Du fait de précédentes études [3, 4, 5], le soudage par résistance en bout est le procédé de
référence considéré pour le soudage du bouchon du gainage combustible en acier ODS.
Dans cette étude, une double approche simulations numériques – essais instrumentés est
utilisée. Après avoir rappelé le principe du soudage par résistance, le modèle numérique est
présenté. La simulation permet d'identifier les phénomènes thermomécaniques pilotant le
soudage par résistance du gainage combustible. Les résultats de la simulation numérique sont
ensuite comparés ceux à des essais instrumentés. Des points de mesures remarquables sont
alors identifiés pour leur représentativité des phénomènes thermomécaniques. Une étude
d'influence des paramètres opératoires sur ces points remarquables permet alors leur
optimisation.
La première étape est l'étape d'accostage. Les pièces à souder sont amenées en contact et
maintenue en place par une force (Fs).
La seconde étape est l'étape de soudage. Le courant de soudage est imposé dans les pièces à
souder par l'intermédiaire d'électrodes. L'effet Joule permet une création de chaleur
dépendante des résistances traversées par le courant. Les résistances sont de deux types :
résistances volumiques des pièces et résistances au niveau des interfaces du fait de l'existence
d'une résistance de contact électrique (RCE). L'intensité du courant est nommée Is et le temps
de passage du courant ts.
La dernière étape est l'étape de forgeage. La force Fs est maintenue après la fin du passage du
courant afin de forger les pièces.
Les paramètres opératoires de ce procédé sont la force de soudage Fs, l'intensité du courant de
soudage Is et le temps de passage du courant ts.
INDEX çç Sommaire
MATERIAUX D'ETUDE
L'acier ODS étudié est la nuance PM2000 KKL6 (grains millimétrique). Ce matériau est un
acier ODS inoxydable ferritique à 20%m de chrome. Il est renforcé par une dispersion
d'oxydes d'yttrium nanométriques (Y2O3). Le KANTAHL APM est utilisé dans cette étude
comme acier de test. Sa composition est proche de celle du PM2000 mais il ne présente pas de
renfort par dispersion d'oxydes.
INSTALLATION EXPERIMENTALE
Afin de réaliser des essais expérimentaux instrumentés, une installation d'essais de soudage
par résistance du gainage combustible a été développée. Sur cette installation, le courant
utilisé est un courant redressé lissé imposé à travers les pièces par l'intermédiaire d'électrodes
en CuCrZr entourant les pièces. La force est appliquée à l'aide d'un vérin pneumatique.
Un système de mesures permettant l'acquisition à haute fréquence d'échantillonnage des
données est utilisé. Le courant traversant les pièces est mesuré à l'aide d'une boucle de
Rogowski. Le déplacement des pièces est mesuré sans contact à l'aide d'un capteur à courant
de Foucault. Un capteur piézoélectrique permet d'enregistrer la force en fonction du temps.
Des fils pour la mesure de tension ont été installés sur les électrodes.
INDEX çç Sommaire
A partir de 5 ms (pour ces paramètres Bouchon
opératoires), les températures les plus
élevées se situent dans la gaine. En
effet, le passage du bouchon massif à
Lg
la gaine étroite génère une
macro-constriction du courant. La
densité du courant dans la gaine est
donc plus élevée que dans le bouchon
Electrode
et l'effet Joule y est plus important. La
partie de la gaine dépassant de
l'électrode s'échauffe puis commence à Gaine
s'affaisser. 2 ms 7 ms 13 ms
Les phénomènes thermiques sont donc Figure 4 : Distribution des températures en cours de soudage obtenue
principalement pilotés par les par simulation numérique
(Fs, Is, ts) = (2200 N, 16000 A, 13 ms)
changements de section
macroscopique. La RCE n'a
d'influence que sur le début du procédé. La longueur de gaine dépassant de l'électrode (Lg) est
donc un paramètre important puisque la génération de chaleur est principalement localisée
dans cette zone.
Par conséquent, la longueur de gaine dépassant de l'électrode est considérée comme un
paramètre opératoire à optimiser.
9,E-04 0
Les grandeurs comparées et
présentées sur la Figure 5 sont la 8,E-04
Aff.
-0,05
2,E-04 -0,35
En début de procédé, on observe
1,E-04 (1800 N; 12000 A; 10 ms ) -0,4
une chute de la résistance
0,E+00 -0,45
comparable entre la simulation et 0 0,002 0,004 0,006 0,008 0,01 0,012
l'expérimentation. Cette chute peut Rélec - exp Rélec - simu Dépla - exp Dépla - simu Temps ( s )
s'expliquer par l'observation Figure 5 : Comparaison entre simulation et essai sur Kanthal APM
réalisée à l'aide de la simulation
concernant le chauffage des interfaces et la diminution des RCE entre les pièces mais aussi
entre les pièces et les électrodes. Le chauffage des interfaces est donc bien évalué par la
simulation et un premier point de mesure remarquable peut être identifié (Rinit), représentatif
du chauffage des deux types d'interfaces. Sur la mesure de résistance, une seconde chute plus
faible apparait (t = 8 ms) suite à l'affaissement des pièces. Cette chute observée sur la
simulation et lors des essais correspond à la diminution de la longueur de gaine dépassant de
l'électrode, principale source de résistance volumique. On peut alors identifier un point de
mesure remarquable (Rfin) représentatif de la résistance de la pièce après affaissement. La
INDEX çç Sommaire
différence entre Rinit et Rfin appelée ∆R est une grandeur remarquable. Il représente la chute de
résistance dû à l'affaissement.
Sur la courbe de déplacement, on observe une dilatation des pièces entre 4 et 6 ms puis un
affaissement. La dilatation est due au chauffage des volumes (et notamment de la gaine) qui
finissent par s'affaisser sous l'effet de la force. Le temps au début de l'affaissement (taff) peut
donc être considéré comme un point de mesure remarquable représentant le temps nécessaire
pour atteindre une température dans la gaine permettant son affaissement à la force
considérée. Ce point est par ailleurs correctement évalué par la simulation montrant le bon
accord des résultats de la simulation sur le chauffage des volumes. L'affaissement de la pièce
est plus rapide sur la simulation que sur les mesures expérimentales. Cependant,
l'affaissement final des pièces est proche confirmant un accord correct de la simulation
numérique mécanique. L'affaissement final (Aff.) est un point remarquable.
L'énergie électrique dissipée dans le système est calculée et deux points remarquables peuvent
être identifiés. L'énergie à l'affaissement (Eaff) représente l'énergie apportée entre t = 0 et t =
taff et correspond à l'énergie nécessaire au chauffage de la gaine jusqu'à une température
permettant son affaissement. Le second point est l'énergie finale Efin qui représente l'énergie
électrique totale dissipée dans le système durant le soudage.
La Figure 6 présente la déformée finale
simulée et la géométrie finale d'une soudure Bouchon
réalisée sur KANTHAL APM et observée par
coupe macrographique. On observe sur les 2200 N
deux images, une déformation principalement 14000A
localisée dans la gaine. Des différences au 13 ms
niveau de la géométrie du bourrelet de
soudure interne et externe sont observées.
Cela peut s'expliquer par le fait que le
glissement du contact sur l'extérieur observé Gaine
sur la macrographie n'est pas représenté dans
la simulation (pour des améliorer la robustesse Figure 6 : Comparaison de la déformée simulée et observée
de la simulation). Le bourrelet interne simulé sur une coupe macrographique sur KANTHAL APM
est légèrement plus important.
La simulation est en bon accord avec les mesures en cours de soudage. L'étude de l'influence
des paramètres opératoires sur les points remarquables identifiés va permettre l'optimisation
du procédé.
INDEX çç Sommaire
valeur du facteur et une flèche vers le bas correspond à une diminution de la valeur de la
réponse lors de l'augmentation de la valeur du facteur; le calcul de cette grandeur peut être
trouvé dans [7]). On notera que les contributions des interactions et les résidus ont été calculés
mais ne sont pas présentés ici et seront considérés comme négligeables.
Les résultats montrent que la force (Fs) a peu d'influence sur l'ensemble des mesures.
Lg et Is ont une influence importante sur la résistance initiale respectivement du fait d'une
résistance des parties massives plus importante et d'un échauffement des interfaces accru.
L'influence de Lg est en revanche très faible sur Rfin. Cela traduit le fait que la partie de gaine
dépassant des électrodes s'est en grande partie affaissée pendant le soudage comme observé
dans les simulations et les coupes macrographiques.
On observe que Lg est le paramètre le plus influent sur Eaff alors que Is et ts ont peu
d'influence. Cela montre que la géométrie de la pièce, et donc les macro-constrictions de
courant, pilotent l'échauffement et l'affaissement.
Is et Lg sont les paramètres les plus influents sur taff. On peut donc supposer que plus Lg est
élevé, plus les températures augmentent rapidement dans la partie de la gaine dépassant des
électrodes.
L'affaissement final est principalement influencé par Lg confirmant que la zone affaissée est
bien la partie de la gaine dépassant de l'électrode.
L'étude de l'influence des paramètres opératoires sur les points de mesures remarquables qui
sont représentatifs des phénomènes thermomécaniques du soudage permettent de réaliser une
optimisation.
OPTIMISATION
L'objectif étant le soudage des pièces en acier ODS, il est nécessaire de limiter les zones pour
lesquelles des températures élevées, pouvant entrainer une modification de la répartition des
oxydes, sont atteintes. C'est le cas pour la partie de la gaine dépassant des électrodes. Une
diminution de Lg est donc préconisée puisque l'augmentation de ce paramètre tend à
augmenter la zone de température élevée. De même, Is augmente le chauffage de la gaine
puisque son influence est importante sur taff contrairement à ts. Il est donc préconisé d'utiliser
des valeurs basse de Is.
Des valeurs basses de Is et de Lg sont donc préconisées. Des soudures de bonne compacité ont
été obtenues pour ces paramètres.
CONCLUSIONS ET PERSPECTIVES
La simulation numérique et les mesures en cours de soudage ont permis de mettre en avant
l'importance de la géométrie des pièces soudées du fait de macro-constrictions de courant
importantes. Ces macro-constrictions sont responsables d'une augmentation de la génération
de chaleur par effet Joule. La longueur de gaine dépassant de l'électrode a alors pu être
identifiée comme un nouveau paramètre opératoire.
Des points de mesures remarquables, représentatifs des phénomènes thermomécaniques, ont
été identifiés grâce à la simulation et aux essais instrumentés. L'étude de l'influence des
paramètres opératoires sur les mesures en cours de soudage et plus particulièrement sur des
points de mesures remarquables a montré l'importance de ce nouveau paramètre. L'étude de
l'influence des paramètres opératoires sur les phénomènes thermomécaniques du soudage par
résistance appliqué au gainage combustible en acier ODS a permis une optimisation des
paramètres opératoires.
INDEX çç Sommaire
Des analyses microstructurales pour étudier l'influence des paramètres opératoires sur la
répartition d'oxyde après soudage sont en cours. De plus la tenue mécanique des pièces
soudées devra être quantifiée.
Ce travail est réalisé dans le cadre du programme de recherche R&D "Matériaux pour les
RNR-Na" cofinancé par le CEA, AREVA NP et EDF.
[1] Lemmen, H.J.K. et al., Laser beam welding of an Oxide Dispersion Strengthened super
alloy. Journal of Materials Science, 2007. 42(13): p. 5286-5295.
[2] Shoemaker, L.E. Joining techniques for a ferritic oxide dispersion strengthened alloy. in
Advances in welding science and technology. Proceeding, Gattingburg, TN, USA, 18-22 May
1986.
[3] de Burbure, S., Resistance welding of pressurized capsules for in-pile creep experiments.
Welding Journal, 1978. 57(11): p. 23-30
[4] Seki, M. et al., Pressurized resistance welding technology development in 9Cr-ODS
martensitic steels. Journal of Nuclear Materials, 2004. 329: p. 1534-1538.
[5] Zirker, L.R., et al., Fabrication of oxide dispersion strengthened ferritic clad fuel pins.
International Conference on Fast Reactors and Related Fuel Cycles, Kyoto, Japan, 28-31 October
1991
[6] Sibilia, G., Modélisation du soudage par point: Influence des conditions interfaciales du
procédé. 2003, Ecole polytechnique de Nantes
[7] Linder, R., Les plans d'expériences: un outil indispensable à l'expérimentateur. 2005:
Presses des Ponts.
INDEX çç Sommaire
Resistance upset welding of ODS steel fuel cladding for the
4th generation of nuclear power plants
Fabien Corpacea,b, Arnaud Monniera, Jean-Pierre Manaudb, Angéline Poulon Quintinb
a CEA, DEN, DANS, DM2S, SEMT, LTA, F-91191 Gif-sur-Yvette, France
b CNRS, Université de Bordeaux, ICMCB, 87 avenue du Dr. A. Schweitzer, Pessac, F-33608, France
Plug (ODS)
Nanometre-sized oxides, high density
Good mechanical
properties at high
temperature
Understanding the physical phenomena of resistance welding apply to cladding in order to control the process parameters and
optimised them to maintain the metallurgical structure guaranteeing goods mechanical properties at high temperature
Local comparison between
Instrumented experiments experiments and simulations Simulation by Finite Element Method
Tested materials: stainless ferritic steel 20wt% Cr: 1E-03 4,1
Coupled thermal - electrical computation and coupled thermal - mechanical
Kanthal APM (not strengthened); PM2000 KKL6 (commercial ODS PLANSEE, large grains) 9E-04 4,02
computation sequenced. Software : SYSWELD (ESI Group)
1,E-03 0,05 8E-04 1 3,94
7E-04 3,86
Plug Plug
9,E-04 -0,05
2
Displacement ( mm )
Rcl amp ( Ohm )
8,E-04
tcol -0,15
6E-04
4 3 3,78
5E-04 3,7
7,E-04 -0,25
6,E-04 -0,35
5
3E-04 3,54
Cs
RClamp ( Ohm )
5,E-04 -0,45
Rinit 2E-04 3,46
4,E-04 -0,55 (1800N; 12000A; 10ms, 0,8mm)
1E-04 3,38
3,E-04
Rfin -0,65
0E+00 3,3
0 0,002 0,004 0,006 0,008 0,01 0,012
2,E-04 -0,75
Rclamp - exp Rclamp - simu Displa - exp Displa - simu Time ( s )
(1800N; 16000A; 15ms)
1,E-04 -0,85
Clamp
Clamp
1
0,E+00 -0,95
0 0,002 0,004 0,006 0,008
Rclamp
0,01
Displacement
0,012 0,014 0,016
Time ( s )
0,018 0,02
2 Collapse kinetic
3 Final collapse Clad
● Realisation of an experimental design in order to calculate the variance 4 ECRs drop kinetic Clad
contribution of the process parameters on the measurements during the process 5 Final resistance 2 ms 7 ms 13 ms
Simulation by FEM of the process – Temperature in pieces
Rinit Rfin ∆R Eaff Efin tcol Col.
● High temperature in the Clad part out of the clamping system
Force (F) ↓ 1% 0% ↓ 1% ↓ 5% 0% ↓ 5% ↓ 1%
The Clad stick-out (Cs) is a
Intensity (I) ↓ 43 % ↓ 66 % ↑ 3% ↓ 2% ↑ 50 % ↓ 29 % ↑ 34 % ● Low influence of the electrical contact resistance (ECR) between
process parameter to be pieces on the maximum temperature reached inside the pieces
Time (t) ↓ 1% ↓ 25 % ↑ 26 % ↑ 10 % ↑ 45 % ↓ 2% ↑ 3%
optimised.
Cs ↑ 49 % ↑ 2% ↑ 50 % ↓ 43 % ↑ 4% ↓ 24 % ↑ 54 % ● High influence of the Clad stick-out (Cs : length of the clad out of the
Results on Kanthal APM; Rinit , Rfin, tcol, Col. : See graph; ∆R: Rfin-Rinit ; Eaff : Energy at collapse beginning; Efin : final energy
The welding force and the clamping system) on the maximum temperature reached inside the pieces
I and Cs are the process parameters with the highest ECR have little influence. Thermal phenomena are mainly driven by section
influence on the measurements. F has little influence. changes and current constrictions in the pieces.
Kanthal APM
The final energy is
3 mostly influenced by Useful responses from
the current intensity
and the welding time. experiments, simulations
2
Correlation between the structure and the measurements (Kanthal APM)
and structures have been Clad
identified to optimise the
Increasing the current intensity and/or the welding The device is optimised based on geometrical
time lead to an increase of the welded interface state. process. considerations using simulations.
Plug Plug
Plug
Clad Clad
Clad
(F, I, t) = (1800 N, 14000 A, 15 ms) (F, I, t) = (2200 N, 18000 A, 15 ms) TEM – Base material (PM2000)
Metallurgical structure
ESOPE - European Symposium On Pressure Equipment - 28, 29 et 30 September 2010 This work is part of a research program for materials suitable for SFR financed by CEA, AREVA NP and EDF.
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page129
RÉSUMÉ / ABSTRACT
49P
INTRODUCTION A UN CONCEPT INNOVANT DE JOINT SPIRALE
Dr S. Bond – VP of technology
(FLEXITALLIC – Scandinavia Mill Hunsworth Lane – Cleckheaton BD19 4LN- Angleterre
E-mail: [email protected])
P. Pottier
(SIEM SUPRANITE – 1, rue Villaret de Joyeuse – 75017 Paris - France
E-mail : [email protected])
B. Labre
(SIEM SUPRANITE – 1, rue Villaret de Joyeuse – 75017 Paris - France
E-mail : [email protected])
RÉSUMÉ
Fort d’un héritage d’innovation continue dans le domaine de l’étanchéité remontant à l’invention du
joint spiralé, FLEXITALLIC et SIEM SUPRANITE souhaitent présenter un nouveau concept de joint
semi-métallique issu du savoir faire dans les joints spiralés et striés revêtus.
Basé sur un feuillard métallique de profil spécifique formé à la manière d’un joint spiralé, ce nouveau
type de joint a démontré dans le cadre d’essais de caractérisation, une étanchéité supérieure à celle des
joints spiralés ainsi qu’un couple compressibilité / reprise élastique plus élevé que celui des joints striés
revêtus (type Kammprofile). Ce nouveau joint semi-métallique est disponible aussi bien avec du gra-
phite expansé qu’avec notre qualité spécifique pour les applications hautes températures en vermiculite
expansée.
ABSTRACT
FLEXITALLIC and SIEM SUPRANITE have a long history of innovation in the gasket industry, dating
back to the invention of the spiral wound gasket. True to this heritage we would like to present a new
semi-metallic concept that draws on the strengths of FLEXITALLIC and SIEM SUPRANITE in spiral
wound gaskets and Kammprofile gaskets.
This new gasket concept is based on a proprietary shaped wire that is wound into a gasket. In standard
lab tests this new gasket shows enhanced sealing compared to a spiral wound gasket, and compression
and recovery superior to a Kammprofile gasket. This semi-metallic gasket is available with graphite
sealing material or with our proprietary high temperature sealing material THERMICULITE®.
INDEX çç Sommaire
INTRODUCTION :
Les joints semi-métalliques actuellement les plus utilisés dans l’industrie sont les joints
spiralés et joints striés revêtus, chacun pour ses propres caractéristiques.
Les joints striés revêtus offrent un niveau d’étanchéité élevé ainsi qu’une bonne rigidité. Ils
peuvent s’accommoder de portées de joints étroites. Ils sont ainsi largement préconisés en
remplacement des joints métalloplastiques sur les échangeurs de chaleur. Leur faible
restitution élastique est cependant problématique lorsqu’ils sont utilisés sur des process
industriels présentant d’important cyclages en pression et/ou température.
Les joints spiralés sont reconnus pour leur fiabilité et leur excellente reprise élastique.
Cependant, ils nécessitent un effort de serrage supérieur aux joints striés revêtus et doivent
être utilisés avec des anneaux limiteurs de compression. Dans certains cas et notamment pour
des portées de joints étroites, leur dimensionnement est impossible.
Constitué d’un feuillard métallique de forme spécifique formé à la manière d’un joint spiralé
(voir figures 1 et 2 ci-dessous), il présente des surfaces d’étanchéité comparables à celles des
joints striés (figure 3).
Ce nouveau joint se comporte comme un joint strié revêtu tout en offrant une reprise élastique
similaire à celle d’un joint spiralé.
Les résultats des différents essais de caractérisation présentés par la suite désignent ce
nouveau concept de joint comme la solution à bon nombre de problème d’étanchéité.
INDEX çç Sommaire
Figure 3 : Surface d’étanchéité. Figure 4 : Points de soudure :
Une fois le feuillard métallique mis en forme, les surfaces La technologie de soudure utilisée est
d’étanchéité sont similaires à celle d’un joint strié. adaptée au feuillard métallique.
INDEX çç Sommaire
Fig. 6 & 7 : Analyse aux éléments finis de deux profils différents
Le Nouveau joint possède un niveau d’étanchéité supérieur aux joint spiralés quelque soit la
contrainte appliquée. Le comportement au délestage est meilleur que celui d’un joint strié
revêtu.
INDEX çç Sommaire
ESSAI DE COMPRESSION
Cet essai de compression, réalisé sur un joint de 4’’ à l’aide une presse d’une capacité de 300
tonnes, a pour objectif de déterminer la contrainte maximale admissible par le Nouveau joint.
Les résultats de cet essai démontrent que le joint n’a pas atteint sa contrainte maximale
admissible (destruction). En effet, aucun signe de flambage ou de plastification de la structure
n’a été observé. La courbe issue de cet essai ne présente pas de point d’inflexion indiquant un
effondrement de la structure (figure 9).
La résistance mécanique de ce Nouveau joint est comparable à celle des joints striés revêtus.
Figure 9 : Essai de compression
INDEX çç Sommaire
Figure 10 : Essai de cyclages entre brides comparé d’un joint strié revêtu graphite et du nouveau joint.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page130
RÉSUMÉ / ABSTRACT
59P
UN NOUVEAU CHALLENGE POUR LES EXPLOÎTANTS : LES CONDITIONS DE
FONCTIONNEMENT ET CRITÈRES COMPLÉMENTAIRES EN LIAISON AVEC
LA RÉGLEMENTATION ESP – ESPN
C. Faidy
EDF-SEPTEN
[email protected]
Pour la conception et la fabrication des équipements sous pression l’Exploitant doit fournir
formellement au Fabricant l’ensemble des conditions de fonctionnement prévues pour l’équipement
considéré.
Dans le cas des Equipements sous pression nucléaires l’exigence est très précise et permettra au
Fabricant de définir ses marges de conception, de réaliser son analyse de risque, son dimensionnement
et l’analyse de comportement de l’équipement.
De plus dans le cas de l’utilisation des normes harmonisées, l’Exploitant responsable de la sûreté de
l’installation, devra compléter ces exigences d’objectifs de marges et d’analyse complémentaires des
conditions accidentelles de l’installations. On trouvera la prise en compte de systèmes dédiés à la gestion
POSTERS
des accidents, les agressions internes-externes (dont le feu ou le séisme) et leur combinaison aux
situations de fonctionnement, les situations post-accidentelles…
L’objet de ce papier est un résumé des principales données à fournir par l’Exploitant au Fabricant dans
le contexte réglementaire français.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page131
RÉSUMÉ / ABSTRACT
67P
RETOUR D’EXPÉRIENCE CONCERNANT L’APPLICATION DE TECHNIQUES
DE GAMMAGRAPHIE DE PROXIMITÉ
A. Blettner
(Institut de Soudure Industrie – 90 rue des Vanesses – 93420 Villepinte – e-mail :
[email protected])
RÉSUMÉ
La réglementation applicable lors de l’examen d’une pièce ou d’un objet à l’aide d’un appareil mobile émetteur de rayonnements
ionisants impose la délimitation d’une zone d’opération matérialisée par un balisage. À la limite de celui-ci, le débit d’équivalent
de dose moyen évalué sur la durée totale de l’opération doit rester inférieur à 0,0025 mSv/h.
Ce périmètre de sécurité, dont l’accès est limité aux travailleurs devant nécessairement être présents (Catégorie A), peut atteindre
des valeurs importantes.
Afin de réduire cette zone, Institut de Soudure Industrie a développé un dispositif adapté au concept de gammagraphie de
proximité et capable de fonctionner avec les gammagraphes autorisés sur le territoire français. Ce dispositif appelé “γ-Prox”
permet de renforcer la protection des opérateurs par une réduction des débits de rayonnements ionisants, tout en diminuant la
distance de balisage.
À titre d’exemple, un examen radiographique effectué à l’aide d’une source radioactive d’iridium 192 d’activité courante de
0,93 TBq (25 Curies) et dont les temps cumulés d’exposition seraient de 2 heures pour une durée d’opération de 6 heures impose
la délimitation d’une zone d’opération de rayon égal à environ 40 mètres (valeur obtenue à l’aide d’un collimateur en uranium
appauvri 1/250e, mesurée sur le coté de celui-ci – atténuation pratique d’environ 1/100e).
Dans les mêmes conditions opératoires de mise en œuvre (temps d’opération, temps d’exposition), mais en utilisant une source
de sélénium 75 avec le γ-Prox, la délimitation de la zone serait de 8 mètres environ (valeur obtenue avec le système de collimation
optimisé, conditions de mesures identiques à celles ci-dessus).
ABSTRACT
The regulations applicable during the examination of a part or object using a portable ionising radiation emitting apparatus
demand cordoning off of the operating zone with boundary markers. At the edge of the exclusion zone the average equivalent
dose rate evaluated over the total duration of the operation, must remain less than 0.0025 mSv/h.
This safety perimeter, whose access is restricted to personnel whose presence is essential (Category A), can reach major
proportions.
In order to reduce this zone, Institut de Soudure Industrie has developed a device which is an adaptation of the Close-Range
Gammagraphy concept, which is capable of operating with the gammagraphic apparatus approved for use in France. This
device, known as “γ-Prox” provides enhanced protection for operators by reducing the ionising radiation dose rate whilst
reducing the scanning exclusion zone.
For example an X-ray examination using a radioactive Iridium 192 source of 0.93 TBq (25 Curies) standard activity and for
which the cumulative exposure times would be 2 hours for a total operation duration of 6 hours involves cordoning off an
operating zone whose radius is equal to about 40 metres (this value is obtained using a 1/250th depleted uranium collimator,
measured on its side – practical attenuation approximately 1/100th).
Under the same operating conditions (operating and exposure times), but using Selenium 75 source with γ-Prox, the radius of
the exclusion zone is reduced to about 8 metres (this value is obtained using the optimised collimation system, and identical
measurement conditions as those used above).
INDEX çç Sommaire
RETOUR D’EXPÉRIENCE CONCERNANT L’APPLICATION DE
TECHNIQUE DE GAMMAGRAPHIE DE PROXIMITÉ
FEED BACK OF EXPERIENCE ON CLOSE-RANGE GAMMAGRAPHY
TECHNIQUE APPLICATIONS
A. Blettner
(Institut de Soudure Industrie – 90 rue des Vanesses – 93420 Villepinte – e-mail :
[email protected])
RÉSUMÉ
La réglementation applicable lors de l’examen d’une pièce ou d’un objet à l’aide d’un
appareil mobile émetteur de rayonnements ionisants impose la délimitation d’une zone
d’opération matérialisée par un balisage. À la limite de celui-ci, le débit d’équivalent de dose
moyen évalué sur la durée totale de l’opération doit rester inférieur à 0,0025 mSv/h.
Ce périmètre de sécurité, dont l’accès est limité aux travailleurs devant nécessairement être
présents (Catégorie A), peut atteindre des valeurs importantes.
Afin de réduire cette zone, Institut de Soudure Industrie a développé un dispositif adapté au
concept de gammagraphie de proximité et capable de fonctionner avec les gammagraphes
autorisés sur le territoire français. Ce dispositif appelé “γγ-Prox” permet de renforcer la
protection des opérateurs par une réduction des débits de rayonnements ionisants, tout en
diminuant la distance de balisage.
À titre d’exemple, un examen radiographique effectué à l’aide d’une source radioactive
d’iridium 192 d’activité courante de 0,93 TBq (25 Curies) et dont les temps cumulés
d’exposition seraient de 2 heures pour une durée d’opération de 6 heures impose la
délimitation d’une zone d’opération de rayon égal à environ 40 mètres (valeur obtenue à l'aide
d'un collimateur en uranium appauvri 1/250e, mesurée sur le coté de celui-ci – atténuation
pratique d'environ 1/100e).
Dans les mêmes conditions opératoires de mise en œuvre (temps d’opération, temps
d’exposition), mais en utilisant une source de sélénium 75 avec le γ-Prox, la délimitation de
la zone serait de 8 mètres environ (valeur obtenue avec le système de collimation optimisé,
conditions de mesures identiques à celles ci-dessus).
ABSTRACT
The regulations applicable during the examination of a part or object using a portable
ionising radiation emitting apparatus demand cordoning off of the operating zone with
boundary markers. At the edge of the exclusion zone the average equivalent dose rate
evaluated over the total duration of the operation, must remain less than 0.0025 mSv/h.
This safety perimeter, whose access is restricted to personnel whose presence is essential
(Category A), can reach major proportions.
In order to reduce this zone, Institut de Soudure Industrie has developed a device which is an
adaptation of the Close-Range Gammagraphy concept, which is capable of operating with the
INDEX çç Sommaire
gammagraphic apparatus approved for use in France. This device, known as “γ-Prox”
provides enhanced protection for operators by reducing the ionising radiation dose rate
whilst reducing the scanning exclusion zone.
For example an X-ray examination using a radioactive Iridium 192 source of 0.93 TBq (25
Curies) standard activity and for which the cumulative exposure times would be 2 hours for a
total operation duration of 6 hours involves cordoning off an operating zone whose radius is
equal to about 40 metres (this value is obtained using a 1/250th depleted uranium collimator,
measured on its side – practical attenuation approximately 1/100th).
Under the same operating conditions (operating and exposure times), but using Selenium 75
source with γ-Prox, the radius of the exclusion zone is reduced to about 8 metres (this value is
obtained using the optimised collimation system, and identical measurement conditions as
those used above).
INTRODUCTION
Des mesures de débit de dose ont été réalisées dans des conditions opératoires identiques avec
un embout d’irradiation (photo 1) - irradiation sur 360° sans atténuation du faisceau, un
collimateur (photo 2) - collimateur en uranium appauvri atténuation 1/250e en face arrière et
d’angle de 120° et le γ-Prox (photo 3).
Les mesures sont réalisées en 3 points cardinaux par rapport au centre du tube, soit
respectivement dans l’axe du faisceau de rayonnement (position A – cf. photo 4),
perpendiculairement à l’axe du faisceau (position B – cf. schéma 1) et à l’opposé de la
direction du faisceau de rayonnement (position C – cf. schéma 1).
INDEX çç Sommaire
C
µSv/h
µSv/h
1,2
1,2
Sonde γ
Tube 10’’ A S
Source
Ir 192
1,2
µSv/h
B
Les valeurs de débit de rayonnement ainsi obtenues sont reportées dans le tableau 1, ainsi que
le résultat des calculs du coefficient d’atténuation, avec, comme référence, l’embout
d’irradiation dont on considère un facteur d’atténuation (équivalent à 2 mm d’acier
inoxydable) très voisin de l’unité.
Point de mesure
A B C
Débit de dose à Embout d’irradiation. 36 106 200
700 mm (mSv/h) Collimateur 120° 29,4 7,4 3,3
γ-Prox 20,3 1,16 1,12
Coefficient Embout d’irradiation. 1 1 1
d’atténuation Collimateur 120° 1,22 14,3 60,6
γ-Prox 1,77 91,4 178,6
Des mesures réalisées avec une source de sélénium 75 donnent des résultats encore plus
éloquents par rapport à un collimateur 120° (efficacité de 1,2 dans l’axe, 21 en face latéral et
20 en face arrière).
INDEX çç Sommaire
EXEMPLES D’APPLICATION DU γ-PROX
Les valeurs mesurées sur site par une tierce partie ont démontré l’efficacité du γ-Prox. En
effet le débit d’équivalent de dose moyen mesuré à 10 mètres était de 0,0016mSv/h, soit
largement inférieur à la valeur maximale autorisée de 0,0025 mSv/h par l’arrêté zonage.
Le système γ-Prox a été mis en œuvre pour la première fois en 2009 sur un chantier Institut de
Soudure Industrie en zone urbaine à forte densité de population (photos 7, 8 et 9), afin
d’assurer le contrôle par gammagraphie de soudures circulaires de canalisations de transport
de gaz.
Photo 9 : Le γ-Prox
en situation de contrôle
sur une canalisation 8’’
INDEX çç Sommaire
Sans la mise en œuvre du γ-Prox, il aurait été nécessaire de faire évacuer le public susceptible
de se trouver dans la zone d’opération lors des travaux de contrôle. Par exemple en zone
urbaine, évacuation des résidents et fermeture des commerces, en zone industrielle,
évacuation des salariés, fermeture pendant plusieurs heures des voies à la circulation.
Les valeurs mesurées sur site ont démontré l’efficacité du γ-Prox. En effet les débits
d’équivalent de dose moyen obtenus ont permis d’établir des distances de balisage de
quelques mètres (< 5 m) et de l’ordre du mètre dans le cas d’opération de contrôle en fouille,
et ce en respect de l’application de l’arrêté zonage.
Ces calculs sont établis sur une durée de l’intervention de 6 heures dont 60 minutes
d’expositions (émission de rayonnements ionisants lors de l’irradiation des films) et sur la
mise en œuvre d’écran atténuant de 2 mm d’épaisseur de plomb (correspond à l’écran de
blocage utilisé couramment lors des contrôles par gammagraphie pour s’affranchir du
rétrodiffusé pouvant impacter la qualité d’image des radiogrammes).
INDEX çç Sommaire
CONCLUSIONS
Le γ-Prox permet donc à la fois de réduire les risques d’irradiation des opérateurs effectuant
les contrôles par radiographie et des tiers susceptibles de se trouver dans le domaine public et
aussi présente de nombreux avantages :
- simplicité de mise en œuvre sans recours à des moyens métrologiques sophistiqués,
- réduction importante des débits équivalent de dose moyen,
- adaptabilité aux gammagraphes du commerce autorisés sur le territoire français sans
modification des accessoires.
Le γ-Prox version contact a fait l’objet d’une validation sur site (raffinerie) pour le contrôle de
tuyauterie entre 4’’ et 10’’. Les caractéristiques de focalisation du système γ-Prox et d’énergie
du radioélément sélénium 75 permettent de réduire de manière significative les débits
d’équivalents de dose moyens. Dans ce cas d’application, les valeurs de débits de dose
obtenues ont permis de réduire le périmètre de balisage d'un facteur 5 à 10, tout en respectant
les contraintes imposées par l’Arrêté du 15 mai 2006 telles que requises à la périphérie de la
zone balisée (le débit d’équivalent de dose moyen évalué sur la durée totale de l’opération
doit être inférieur à 0,0025 mSv/h). La réduction de ce périmètre de balisage permet
d’envisager des travaux en coactivité à proximité.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page132
RÉSUMÉ / ABSTRACT
79P
LA PART DE LA FRANCE EN QUALIFICATION SOUDAGE DANS LE
MONDE
RÉSUMÉ
A l’heure de la mondialisation, l’AFS fait le point sur l’attribution des diplômes de qualification
internationale du personnel en soudage.
La comparaison est faite avec les autres pays qui participent au système de formation et qualification
de l’Institut International de la Soudure (IIW).
ABSTRACT
POSTERS
In a period of world wide globalisation, AFS takes stock on awarding of international qualification
diplomas concerning welding personal.
Comparison is made with another countries participation in training and qualification system of
International Institute of Welding (IIW).
INDEX çç Sommaire
1. INTRODUCTION
L'Association Française du Soudage (AFS) a été créée en 1991 par deux membres fondateurs,
l'Institut de Soudure et la Société Française des Ingénieurs, Techniciens et Scientifiques en
Soudage (SIS) avec la participation de constructeurs et organisations professionnelles.
Le but est de mettre à disposition des entreprises, soit dans le cadre des exigences
contractuelles exprimées par leur client, soit pour améliorer la qualité de leurs prestations, un
système de reconnaissance des connaissances des personnes en charge de l'activité soudage
selon des critères fixés tant au niveau Européen qu'International.
Le présent exposé est destiné à comparer ce qui a été réalisé en France par rapport aux autres
pays délivrant des diplômes similaires.
2. ORGANISATION DE L’AFS
L’AFS est gérée par un Conseil d’Administration dans lequel les places sont attribuées aux
membres selon leur catégorie.
Il existe les catégories suivantes :
- membres fondateurs disposant de trois sièges chacun
- membres constructeurs disposant de deux sièges
- membres donneurs d’ordres disposant de deux sièges
- membres d’organismes d’inspection disposant de deux sièges
- autres membres disposant de deux sièges
L’AFS dispose d’un Secrétariat Permanent dont le titulaire est également le Responsable
Qualité.
Les moyens opérationnels de l’AFS sont représentés par deux comités exécutifs :
- CEITS (Comité d’Enregistrement des Ingénieurs et Techniciens en Soudage) en charge de
l’attribution des diplômes et attestations concernant les coordonnateurs en soudage ;
- CC (Comité de Certification) en charge des certifications des agents d’inspection en
soudage de la certification des personnels en soudage ainsi que de la certification IIW
qualité en soudage des constructeurs selon ISO 3834.
L’agrément de l’AFS par l’IIW est réévalué tous les cinq ans lors d’un audit de
renouvellement, un audit de suivi ayant lieu entre chaque audit de renouvellement.
L'AFS est l'organisme français agréé par la Fédération Européenne du Soudage, Assemblage
et Coupage (EWF) au niveau européen depuis 1993 et par l'Institut International de la
Soudure depuis 1998. A ce titre, l'AFS est autorisée à émettre des diplômes Internationaux
d'Ingénieur, Technologue, Spécialiste, Praticien en Soudage, Agents d’Inspection en Soudage
INDEX çç Sommaire
et soudeurs ainsi que les diplômes Européen de Spécialiste Collage du diplôme Européen de
colleur. L’année 2008 a vu le troisième renouvellement de l’agrément EWF et le deuxième de
l’agrément IIW. Un audit de surveillance a eu lieu en 2008. Ces diplômes font suite à un
programme de formation conforme à celui défini au niveau international et dispensé dans un
des centres agréés par l’AFS selon des règles internationales.
Les cours sont diffusés par les ATB (centre de formation autorisé). En France il existe 19
ATB répartis dans 7 centres. A titre d’exemple il existe 371 ATB en Allemagne.
Le nombre de diplômes IIW établi depuis l’origine a connu un pic en 2001 (période de forte
transformation des diplômes européens en diplômes internationaux) et a repris depuis cette
date un rythme de croissance comme l’indique le tableau suivant :
Si on regarde par catégorie, on constate que les diplômes les plus attribués sont ceux de IWE
et IWS, suivis assez loin derrière par IWT et IWI. Les diplômes de soudeurs ont fortement
augmentés depuis 2007. Les diplômes IWP sont peu demandés.
INDEX çç Sommaire
3.1 Diplôme d’Ingénieur en Soudage International (IWE ou EWE)
Les premiers diplômes ont été attribués en France à la promotion de l’ESSA sortie en 1992.
L’agrément officiel de la Fédération Européenne du Soudage (EWF) n’a été opérationnel
qu’en 1993.
La voie normale pour l’attribution du diplôme est de suivre une formation dans un centre
autorisé par l’AFS (ATB) suivie de la réussite à un examen conduit par un Jury dont le
Président est désigné par l’AFS.
Avec la reprise de l’attribution par l’Institut International de la soudure, il a existé une période
transitoire d’attribution du diplôme sur dossier qui s’est prolongée jusque fin 2001.
En France, trois centres sont autorisés à réaliser la formation et à faire passer les examens.
INDEX çç Sommaire
3.2 Diplôme de Technologue en Soudage International (IWT ou EWT)
En France, six centres sont autorisés à réaliser la formation et à faire passer les examens.
INDEX çç Sommaire
3.3 Diplôme de Spécialiste en Soudage International (IWS ou EWS)
Comme pour les deux diplômes précédents le premier agrément remonte à 1993.
La période transitoire est terminée depuis fin 2001.
Il existe également une route alternative.
En France, cinq centres sont autorisés à réaliser la formation et à faire passer les examens.
INDEX çç Sommaire
3.4 Diplôme de Praticien en Soudage Européen ou International (IWPou
EWP)
INDEX çç Sommaire
3.5 Diplôme de Personnel d’Inspection en Soudage International (IWIP)
Pour attribuer ces diplômes, l’agrément de l’AFS par l’IIW a été prononcé le 01/01/2004.
La période transitoire pour l’attribution du diplôme sur dossier a été possible jusqu’au
30/06/2007.
En France, un seul centre est autorisé à réaliser la formation et à faire passer les examens.
INDEX çç Sommaire
3.6 Diplôme de Soudeur International
En France 3 centres sont agréés pour former et faire passer les examens selon le guide IIW
mais aucun diplôme n’a encore été attribué.
Le nombre de diplômes attribués par pays est le suivant :
INDEX çç Sommaire
Pour les inspecteurs, seule l’Australie a émis 10 diplômes dans un autre pays au titre de la
période transitoire et le Portugal 2 après formation.
Pour les soudeurs, aucun diplôme n’a été remis dans un autre pays.
Il s’agit des certificats pour les ingénieurs, technologues, spécialistes et praticiens en soudage.
A ce jour 6 diplômes ont été délivrés.
INDEX çç Sommaire
4. CERTIFICATION D’AGENTS D’INSPECTION EN SOUDAGE
De telles certifications existent dans d’autres pays tels que l’Italie, le Royaume Uni,
l’Espagne et les Etats-Unis. Elles n’ont pas été reprises à ce jour dans le système
international.
Ce système est purement français. A notre connaissance il n’existe pas dans d’autres pays.
Certains pays, par contre, continuent à délivrer des diplômes nationaux d’ingénieurs,
technologues, spécialistes ou praticiens en soudage.
Selon l’EN ISO 14731, la nomination du Coordonnateur en Soudage est réalisée par le chef
d’entreprise sur deux bases :
- les connaissances en soudage de la personne qui peuvent être Complètes (C), Spécifiques
(S) ou Basiques (B) ;
- sa compétence professionnelle.
Afin de permettre aux chefs d’entreprises d’habiliter les coordonnateurs en soudage (lorsque
les personnes à habiliter ne possèdent pas les diplômes énumérés au point 3) l’AFS a mis en
place un système d’attestation des connaissances en fonction des activités de l’entreprise dans
laquelle ils travaillent et des fonctions qu’ils exercent.
Cette disposition peut être utile quand le chef d’entreprise qui habilite le coordonnateur n’a
pas de connaissances en soudage pour éviter toute contestation.
INDEX çç Sommaire
6. CERTIFICATION QUALITE EN SOUDAGE DES ENTREPRISES SELON
EN ISO 3834
Cette certification qui existait au sein de la Fédération Européenne du Soudage (EWF) a été
reprise en 2007 par l’IIW. L’AFS est maintenant agrée par l’IIW pour assurer cette
certification.
Selon le système européen, le nombre d’entreprises certifiées par pays est de :
✾✾✾✾✾✾✾✾✾
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page133
RÉSUMÉ / ABSTRACT
80P
PRESENTATION DES ETUDIANTS ESSA AYANT RECU
UNE BOURSE AFIAP 2009-2010
ESSA
Espace Cormontaigne
4, boulevard Henri Becquerel
59 970 YUTZ-THIONVILLE
M. Camus
Travaux d’expertise a l’institut de soudure industrie / Work welding expertise at the Institut de Soudure
Industrie
C. Coly
Aciers faiblement alliés revêtus par alliages base nickel. Amélioration de la rectitude de la ligne de fu-
sion pour faciliter les contrôles non destructifs / Low alloy steels coated with Ni alloys. Improvement
of fusion line straightnes to make CND easier
C. Jardin - A. Pruvost
Etude de la réparabilité par soudage de corps de vannes en acier au chrome –molybdène utilisées dans
l’industrie / Repair welding of Cr – Mo low alloy steels valves used in Bugey nuclear power plant
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page134
RÉSUMÉ / ABSTRACT
92P
ETANCHEITE HAUTE PERFORMANCE
HIGH PERFORMANCE SEALING
D. Vittecoq
(EYNARD ROBIN -1, allée de la Grange - 69191 Saint Fons -
[email protected])
RÉSUMÉ
Aujourd’hui, dans le domaine de l’étanchéité industrielle, les exigences de sécurité de plus en plus sé-
vères, le respect de l’environnement et les niveaux d’étanchéité requis imposent plus que jamais de
s’appuyer sur les meilleures expertises. Le Groupe EYNARD ROBIN, spécialiste des solutions d’étan-
chéité depuis plus d’un siècle, partenaire de l’industrie et des constructeurs (OEM), vous propose trois
solutions en étanchéité haute performance :
1 KLINGER® Quantum, joint souple « calandré hydrogéné », joint unique et universel, une innovation
technologique en étanchéité statique depuis l’invention de la KLINGERit
2 EYRTOR, joint torique métallique aux caractéristiques de hautes performances, développé pour
des conditions extrêmes et les applications aux exigences les plus élevées
3. GRAPHITOR®, joint contact métal/métal avec une âme métallique rigide monobloc
et bagues d’étanchéité en graphite, capable d’assurer une étanchéité optimale
Ces trois joints offrent des performances exceptionnelles en pression et température (du vide à 1000
POSTERS
bar et de -250°C à +750°C) et des taux de fuite de l’ordre de 10-8 à 10-10 selon le TA-Luft VDI 2440
(suivant le type de joint).
ABSTRACT
Today, in the sealing industry, people and environment safeties demanded by the market are getting
more and more severe. Moreover, sealing required levels are imposing more than ever to be supported
by the best practices and expertise. The EYNARD ROBIN Group, specialized in Sealing Solutions for
more than one century, in partnership with Industries and Constructors (OEM’s), is providing 3 high
performance sealing solutions:
1 KLINGER® Quantum, a soft gasket, calendered and hydrogenated. This unique and universal gasket
is a technological innovation in static sealing since the KLINGERit invention.
2 EYRTOR, metallic ORing with high performance features has been developed for extreme
conditions and for the most demanding applications
3. GRAPHITOR® , a metal/metal contact gasket with a metallic, rigid and single-piece core and
graphite sealing ring(s) is able to meet an optimal sealing level.
These 3 gaskets withstand exceptional sealing conditions in pressure and temperature (from vacuum
to 1000 bar and from -250°C to +750°C) and they also reach leakage rates between 10-8 à 10-10 according
to the TA-Luft VDI 2440 (depending on the gasket type).
INDEX çç Sommaire
JOINT SOUPLE KLINGER®Quantum
Depuis l’interdiction de l’amiante, de nombreux utilisateurs ont recherché un matériau de
joint possédant les mêmes propriétés vis à vis des hautes températures que la KLINGERit.
En tant que premier fabricant de matériau de joints statiques, KLINGER fut le pionnier dans
le développement des joints sans amiante.
L’évolution ultime fut le lancement du KLINGER®top-sil-ML1, qui bien que représentant un
grand progrès, n’atteignait toujours pas le but final.
Avec le développement du KLINGER®Quantum, KLINGER inaugure une nouvelle ère dans
le monde de l’étanchéité.
INDEX çç Sommaire
Des propriétés exceptionnelles
Le test de cambrage 3-points est souvent utilisé comme méthode d’évaluation de la flexibilité
des matériaux renforcés fibres. Des tests spéciaux, sur des éprouvettes conditionnées
fournissent une indication sur la fragilité et par conséquent sur le vieillissement des
élastomères utilisés.
Avant le test, les spécimens sont d’abord conditionnés, puis testés.
Les résultats des essais sur les échantillons artificiellement vieillis fournissent des
informations sur la résistance au vieillissement des différents concepts de matériaux ; en
particulier dans les applications vapeur où l’on rencontre fréquemment des «coups de bélier»
qui endommagent les joints.
Un joint plus flexible qui peut surmonter des dilatations considérables sans rupture est un
facteur décisif pour l’obtention d’un joint plus fiable. Dans ce test KLINGER®Quantum
montre sa position unique et remarquable comparée à tous les matériaux renforcés fibres
disponibles.
La flexibilité de KLINGER®Quantum sous hautes températures est 8 fois supérieure à celle
des matériaux de joints traditionnels. Tous les aspects négatifs des joints plats comme la
fragilité, la formation de fissures, et une augmentation des fuites peuvent être réduits de façon
significative par l’emploi de KLINGER®Quantum. La manipulation du matériau est similaire
à celle des matériaux déjà connus, et de ce fait familièrement simple.
Dans ce test l’échantillon est déformé en son milieu entre deux points d’appui avec une
vitesse constante, jusqu’à la rupture ou jusqu’à ce qu’une valeur spécifiée ait été atteinte.
Pour ce test de flexibilité les échantillons des matériaux renforcés de fibres de référence et de
KLINGER®Quantum ont été préalablement chauffés à 200°C et 250°C pendant 48 heures.
Les résultats de ces tests sur ces éprouvettes artificiellement vieillies fournissent des
informations sur la résistance au vieillissement des deux matériaux et mettent en évidence la
haute performance du KLINGER®Quantum
INDEX çç Sommaire
Le comportement unique de KLINGER®Quantum est encore plus remarquable dans les tests à
long terme. Pour ceux-ci un matériau renforcé de fibres de référence et le
KLINGER®Quantum ont été à nouveau traités à une température de 200°C pendant 300 et
600 heures.
Après 600 heures à une température de 200°C KLINGER®Quantum montre une flexibilité qui
est 8 fois supérieure à celle des matériaux renforcés de fibres connus.
En particulier dans les applications vapeur il y a de forts «coups de bélier» qui endommagent
le matériau de joint. Un matériau plus flexible qui admet des déformations plus grandes sans
rupture est un facteur qui contribue à un assemblage plus fiable et plus sûr.
Test de cambrage sous hautes température et Test de cambrage après 300 et 600 heures
INDEX çç Sommaire
Caractéristiques du KLINGER®Quantum
Zones d’application
Zone 1 – Compatible : Le contrôle technique est facultatif. Vérifier simplement la
compatibilité chimique.
Zone 2 - Compatibilité sous réserve : Le contrôle technique est recommandé.
Zone 3 – Non-compatible sans contrôle : Le contrôle technique est indispensable, prendre
contact avec notre service technique.
INDEX çç Sommaire
Méthode KLINGER de test de compression à chaud et à froid
INDEX çç Sommaire
Facteur QSmax suivant EN13555
QSmax est la pression d’assise maximale qui doit être appliquée sur le joint à la température
indiquée sans destruction du joint. La détermination de QSmax pour un matériau en feuille,
peut conduire à une surestimation de la capacité de ce matériau et il est important que toutes
les valeurs de QSmax pour les feuilles soient vérifiées par un test pour PQR à la même
température et pression d’assise que pour la valeur QSmax.
Cette caractéristique du joint est déterminée par la reprise d’épaisseur du joint entre la
pression d’assise initiale et la décharge à un tiers de la pression d’assise initiale.
Cette détermination est faite lors des cycles de décharges du test QSmax.
La valeur de EG change avec la pression d’assise sur le joint.
Une valeur de EG élevée signifie une faible reprise élastique, une valeur basse représente une
plus grande reprise élastique et est, de ce fait, la preuve d’une moindre fragilité du matériau.
INDEX çç Sommaire
Facteur QSmin suivant EN 13555 (étanchéité sous hautes températures)
QSmin exprime la pression d’assise minimale requise sur le joint dans les conditions de
service de telle façon que la classe d’étanchéité L soit maintenue pour la pression intérieure
donnée. Pendant le test le joint doit subir les cycles de compression et décharge et le taux de
fuite sera mesuré pour une pression d’assise définie dans la norme et pour une pression
interne de 40 bar. L’hélium est utilisé pour ce test.
Graphique 1 Graphique 2
Un assemblage boulonné restera étanche tant que la pression d’assise en service est supérieure
à la pression d’assise minimale requise pour maintenir les niveaux d’étanchéité, mais
inférieure à la pression d’assise maximale admissible. Mais les exigences de plus en plus
sévères pour ces assemblages (par exemple classe d’étanchéité 0.01) nécessitent l’application
de contraintes plus élevées.
Dans les conditions de charge cyclées nous recommandons une pression d’assise minimale de
30 Mpa et que le joint soit aussi fin que possible.
INDEX çç Sommaire
Agréments / Certificats du KLINGER®Quantum
INDEX çç Sommaire
Comparaison des coûts entre des joints classiques et le Joint KLINGER®Quantum
INDEX çç Sommaire
JOINTS TORIQUES METALLIQUES EYRTOR
Les joints toriques métalliques EYRTOR sont conçus pour une utilisation dans des conditions
de services extrêmes et dans les secteurs les plus exigeants comme l’aérospatiale, le nucléaire,
l’automobile, l’industrie ou le médical. Fabriqués à partir de tubes, bandes et fils en acier
inoxydable ou en alliage haute température, ces joints d’étanchéité haute performance
résistent à des conditions sévères de température et de pression, ainsi qu’aux environnements
corrosifs.
A : Début de la compression
B : Effort minimal pour étancher : charge d’appui à plus de 90% de sa valeur maximale.
L’épaisseur du joint sous cet effort correspond à la profondeur maximum de la gorge.
C : Contact métal/métal : charge d’appui maximale et écrasement maximum du joint.
L’épaisseur du joint correspond à la profondeur minimum de la gorge.
D : Reprise élastique : réaction du joint supérieure à la charge résiduelle (décompression)
E : Reprise élastique utile : reprise élastique du joint à 20% de la charge maximale (Lmax)
F : Reprise élastique totale du joint d’étanchéité
INDEX çç Sommaire
Tests démontrant les zones de contrainte lors de la compression du joint EYRTOR
INDEX çç Sommaire
Profils
INDEX çç Sommaire
Guide de sélection
Evaluation :
Internes
C Statique Externes
Axiales
Internes
O Statique Externes
Internes
Y Statique NR Externes
Internes
OV Statique Externes
Internes
CS Statique NR Externes
Internes
OS Statique Externes
Internes
OG Statique Externes
CO Dynamique NR Axiales
INDEX çç Sommaire
Composition
• Du joint :
o Alliage de nickel X-750,
o Alliage de nickel 718,
o Inox 321,
o Alliage de nickel 600,
o Inox 304,
o Inox 316,
o Hastelloy,
o Inox 302
• Du revêtement :
o Argent (Ag) : malléable, il offre une bonne résistance à la température et à la
corrosion pour une grande variété d’application. Températures maximales
d’utilisation allant de - 650°C (non oxydant) à + 430°C (oxydant)
o Or (Au) : malléable, il présente une excellente résistance aux produits
chimiques et à l’oxydation, Température maximale d’utilisation : + 930°C
o Cuivre (Cu) : assez malléable, il est économique et supporte jusqu’à + 930°C
o Nickel (Ni) : dur, utilisé à la place de l’argent dans les environnements chauds
et oxydants, il endure jusqu’à + 1200°C
o Plomb (Pb) : de nature extrêmement malléable, il est excellent pour la
cryogénie et peut être utilisé pour les faibles contraintes (70 MPa max) et
température maximales : + 200°C
o Téflon (PTFE) : extrêmement malléable, il est utilisé pour les faibles
contraintes (80 MPa max). Chimiquement inerte il supporte + 230°C
maximum
o Non revêtu : sa température maximale dépend de la matière de base
Caractéristiques techniques
Applications
Les joints métalliques EYRTOR ont été conçus pour tous pour tous les marchés, y compris les
secteurs aérospatial, nucléaire, automobile, industriel et médical entre autres.
INDEX çç Sommaire
JOINTS GRAPHITOR®
Les joints GRAPHITOR®, une exclusivité du Groupe Eynard Robin, sont composés d’une
âme métallique extrêmement rigide et de bagues en graphite.
Les bagues graphite sont insérées dans des gorges usinées sur chaque face du joint, elles
assurent l’étanchéité. L’âme métallique sert de limiteur d’écrasement à ces bagues graphite.
Grâce au contact métal/métal, la partie active du joint ne peut pas subir de sur-contrainte lors
des variations de pression et/ou de température.
Le profil des gorges possède un design particulier de forme trapézoïdale dont le rôle est de
maintenir la bague sans adjonction de colle.
Les bagues graphite ont une forme spécifique qui favorise le contact métal/métal après
serrage. La densité de la bague est calculée en fonction des conditions de service.
Les joints GRAPHITOR®, s’appuient sur une conception robuste avec une âme métallique
monobloc qui leur confère une grande rigidité pour une meilleure résistance aux conditions
sévères de température et de pression. Ils se déclinent en 2 versions pour couvrir un maximum
d’application en étanchéité statique :
• GRAPHITOR GBM® IMPERATOR, avec bagues en graphite expansé matricé
• GRAPHITOR GFE® IMPERATOR, avec bagues en graphite flexible expansé.
INDEX çç Sommaire
Profils
Type Schéma Description Type de bride
Joint
GRAPHITOR Tous types de
GBM® GBM® OU brides y compris
GFE® GFE® simple emboîtement
piste superposée simple et double
Joint
GRAPHITOR Brides plates ou
Version GBM DP® GBM® OU à faces surélevées
GBM® GFE DP® GFE® doubles
et pistes superposées
GFE® Joint
GRAPHITOR
GBM DPF® GBM® OU Brides plates ou
GFE DPF® GFE® doubles à faces surélevées
pistes superposées
avec détection de
fuites
Joint Tous types de
GRAPHITOR brides y compris
GBM UN® GBM® OU celles n’acceptant
GFE UN® GFE® simple que les joints
piste décalée de faible épaisseur
Composition
• Ame métallique :
o En standard : Inox 316 L, Monel 400, Alloy C276 (autres nuances sur
demande)
o Application nucléaire : Inox 316 L version RCCM
• Bagues Graphite
o En standard : Graphite pur spécial Haute température - pureté 98%
o Application nucléaire : Graphite PMUC – pureté 99,5%
Applications
Agréments et certificats
Les joints GRAPHITOR GBM® et GFE® avec graphite de qualité nucléaire (pureté 99, 5% et
bas taux de soufre) sont référencés GRAPHITOR® PMUC n° 17 182, fabrication
IMPERATOR Industries
INDEX çç Sommaire
Conclusion
• GRAPHITOR® : véritable joint statique plat à contact métal/métal avec âme monobloc
et bagues d’étanchéité en graphite pour très hautes performances et exigences sévères.
Pour toute installation à risque chimique, zones dangereuses et applications
nucléaires…
INDEX çç Sommaire
TABLEAU DES ABBREVIATIONS
INDEX çç Sommaire
BIBLIOGRAPHIE
INDEX çç Sommaire
Annexe 1
INDEX çç Sommaire
Annexe 2
INDEX çç Sommaire
INDEX çç Sommaire
Annexe 3
INDEX çç Sommaire
INDEX çç Sommaire
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page135
RÉSUMÉ / ABSTRACT
93P
CERTIFICATION QUALITÉ EN SOUDAGE :
APPLICATION AU DOMAINE FERROVIAIRE
RÉSUMÉ
ABSTRACT
INDEX çç Sommaire
Certification qualité en soudage :
Application au domaine ferroviaire.
- les ateliers sont soumis quels que soient leur taille, les procédés de soudage
employés et le métal de base traité (aciers, alliages d’aluminium…)
- seuls sont dispensés les ateliers fabriquant uniquement des pièces simples
dont la défaillance n’a pas de conséquence sur la sécurité telles que sièges,
grilles de ventilation, panneaux de signalisation, marches, mains courantes,
pare-soleil de wagons…
Il s’agit aussi bien de véhicules à passagers que des wagons de fret ou transportant
des matières dangereuses circulant sur le réseau ferré national. Les réseaux ferrés
ayant les mêmes conditions d’exploitation (partenaires par exemple) sont aussi
concernés.
INDEX çç Sommaire
- Partie 5 : Vérifications, contrôles et documentations.
4 niveaux de certification sont définis pour la certification des constructeurs. Ils sont
fonction
• d’une part du tableau ci-dessous
• d’autre part du rôle dans la sécurité du composant ou du sous-ensemble dont
l’assemblage soudé fait partie
INDEX çç Sommaire
De la catégorie de contrainte et de sécurité des équipements aux types de
contrôles
Déroulement de la certification
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page136
RÉSUMÉ / ABSTRACT
96P
DÉVELOPPEMENT, QUALIFICATION ET RÉALISATION DE JONCTIONS
ZIRCALOY/ACIER AUSTÉNITIQUE PAR SOUDAGE-DIFFUSION POUR LE
RÉACTEUR DE RECHERCHE CABRI
RÉSUMÉ
Le réacteur de recherche Cabri est implanté sur le centre d’études nucléaires de Cadarache. Ce réacteur
est exploité par le CEA/DER et dédié à la réalisation d’essais de sûreté de l’IRSN. Pour la réalisation
du programme international CIP piloté par l’IRSN sous un accord OCDE, une nouvelle enceinte
expérimentale en alliage de zirconium a été implantée. Des jonctions hétérogènes tubulaires entre un
acier inoxydable et cet alliage de zirconium sont nécessaires pour établir la liaison avec la boucle
permettant l’établissement des conditions thermo hydrauliques.
Le CEA/LITEN et le CEA/DER ont mis au point de telles jonctions hétérogènes par soudage-diffusion
assistée par compaction isostatique à chaud. La démarche suivie pour développer et qualifier ces joints
POSTERS
est présentée en insistant sur les contraintes spécifiques liées au contexte nucléaire. Le choix des
intercalaires ainsi que des conditions de température et de pression lors de la CIC reposent sur des
considérations métallurgiques, mécaniques et de tenue à la corrosion. L’optimisation de la géométrie
de joint a été obtenue par des calculs aux éléments finis donnant l’évolution des contraintes après
soudage, après usinage final de la jonction et en service. Des moyens spécifiques de caractérisations
CND ont été développés. Les jonctions optimisées ont finalement été soumises à des tests représentatifs
des sollicitations en service.
ABSTRACT
The Cabri research reactor is located at the Cadarache nuclear research center. This reactor is operated
by CEA/DER and devoted to IRSN safety programmes. For the need of the Cabri International Program
(CIP) carried out by IRSN under an OECD agreement, a new experimental vessel made of zirconium
alloy has been implemented. Tubular junctions between stainless steel and zirconium alloy are required
to liaise with the loop allowing the establishment of thermal hydraulic conditions.
CEA/LITEN and CEA/DER have developed such heterogeneous welded joints-diffusion assisted by Hot
Isostatic Compaction (HIP). The approach taken to develop and qualify these joints is presented with
emphasis on the specific constraints linked to the nuclear context. The choice of infill and the HIP
conditions of temperature and pressure are based on metallurgical, mechanical and corrosion resistance
considerations. The optimization of the joint geometry was obtained by finite element calculations
giving the evolution of stress after welding, after final machining and under operating conditions.
Specific methods of NDT characterizations have been developed. The optimized junctions were finally
subjected to representative tests of in service conditions in a loop.
INDEX çç Sommaire
ABREVIATIONS :
INDEX çç Sommaire
CONTEXTE
Le réacteur de recherche CABRI, implanté sur le centre du CEA à Cadarache, est utilisé pour
des études de sûreté liées au comportement des crayons combustibles (gainage et
combustible) des réacteurs nucléaires en situation accidentelle de type RIA (Reactivity
Initiated Accident). Initialement conçue pour les combustibles des réacteurs à neutrons
rapides refroidis au sodium, l’installation est actuellement significativement modifiée pour
l’adapter aux études des combustibles des réacteurs à eau dans le cadre d’un programme
international, le programme CABRI Boucle à eau, piloté par l’IRSN et placé sous l’égide de
l’OCDE.
Les essais réalisés sur ce réacteur consistent à soumettre l’échantillon de crayon combustible
testé à un transitoire rapide de puissance représentatif de la situation accidentelle
correspondant à l’éjection d’une barre de commande.
INDEX çç Sommaire
- La seconde constituée des différents organes hydrauliques (pompe primaire,
pressuriseur, échangeurs,..) placés dans un caisson de confinement et de canalisations
sous double enveloppe les reliant à la cellule en pile au travers d’un filtre recueillant
les particules du combustible d’essai susceptibles d’être éjectées dans la boucle.
Afin d’obtenir le meilleur couplage possible entre le cœur nourricier et le crayon d’essai, le
matériau retenu pour l’enceinte EP et le tube de sécurité est le Zircaloy 4. En effet, cet alliage
présente la particularité d’être relativement transparent aux neutrons, ce qui améliore le
couplage, tout en présentant des caractéristiques mécaniques élevées nécessaires pour le
dimensionnement des structures. Le choix de ce matériau pour la réalisation de la cellule en
pile impose le développement de jonctions hétérogènes zircaloy/inox permettant une liaison
avec le reste des circuits réalisés en acier austénitique. En effet, le soudage par des procédés
classiques de ces deux matériaux est impossible. De telles jonctions étaient réalisées par le
passé par soudage eutectique. Elle reposait cependant sur un savoir-faire très précis entraînant
des problèmes de reproductibilité. Le soudage par explosion est aussi une solution possible
pour l'assemblage de ce couple de matériau. Il n'est cependant pas compatible avec les
dimensions souhaitées pour ces jonctions tubulaires. Ce procédé étant plus approprié à
l'assemblage de plaques.
INDEX çç Sommaire
L'étude présente ici le développement d’une jonction hétérogène zircaloy/inox par soudage-
diffusion. Dans un premier temps, la mise au point de ces jonctions a concerné les besoins des
dispositifs d'essais sous la responsabilité de l'IRSN. Les résultats obtenus ont permis
d'effectuer le transfert technologique pour les besoins de l'enceinte CABRI.
Le soudage-diffusion
Le soudage-diffusion est un procédé de soudage en phase solide dans lequel les matériaux
sont portés à haute température et haute pression. La pression permet d'assurer un contact
intime des surfaces à souder et la température favorise les mécanismes de diffusion. Pour
l'application considérée, les pièces, une fois préparées de manière appropriée, sont placées à
l'intérieur d'un conteneur permettant de réaliser le vide (figure 2). L'ensemble est alors
introduit dans une enceinte de Compaction Isostatique à Chaud (CIC) qui permettra d'imposer
le cycle de température et pression préconisé.
Parmi les particularités de ce procédé de soudage, on peut noter : l'absence de ZAT (Zone
Affectée Thermiquement) remplacée par une Zone Affectée par la Diffusion d'épaisseur
inférieure au mm, l'assemblage de géométries complexes et de couple de matériaux difficiles
à souder par d'autres techniques, une homogénéité en température de l'ensemble des pièces
(imposant la recherche d'un compromis sur les traitements thermiques des différents
matériaux si nécessaire), une bonne reproductibilité … Pour la jonction considérée, les blocs
de matière préalablement préparés sont soudés conformément à la figure 2 suivant un cycle de
CIC approprié. La géométrie tubulaire finale est alors usinée dans un deuxième temps.
Matériau 1
Matériau 2
Dimensionnement de la jonction
Les jonctions hétérogènes à réaliser sont destinées à être montées sur la boucle du réacteur
CABRI et assurer leur fonction pendant toute la durée de vie du réacteur : soit environ 350
cycles de montées à 155 bars et à une température de 300°C pour une durée cumulée de 6
mois à cette température. Ces jonctions étant des organes participant à la sûreté de la boucle
(première barrière du circuit primaire), un accrochage mécanique est demandé en supplément
de la jonction soudée. Les règles de détermination des épaisseurs tirées du RCCM [5] pour
des tuyaux droits ont alors permis d'obtenir le design présenté sur la figure 3 et nommé dans
la suite jonction Z.
Des simulations thermo mécaniques par éléments finis (code CAST3M) ont alors été
conduites afin d'estimer les contraintes exercées sur le joint et ainsi d'optimiser la forme finale
du joint ainsi que le choix du matériau extérieur. Dans un premier temps, le calcul de
refroidissement du conteneur contenant les pièces soudées depuis la température de CIC
jusqu'à la température ambiante est réalisé (un état de contrainte nulle est supposé en fin de
palier haute température de la CIC). Les contraintes obtenues le long du joint sont alors
uniquement dues au différentiel de coefficients de dilatation entre les matériaux. L'usinage
théorique (i.e. sans ajout de contraintes supplémentaires) de la jonction hétérogène est ensuite
simulé par enlèvement de matière et rééquilibrage des contraintes. Finalement, les
sollicitations nominales de chargement de la jonction sont prises en compte et ajoutées aux
contraintes résiduelles obtenues à l'ambiante.
C D
A B
La figure 4 présente les conditions limites utilisées pour ces différentes étapes de calcul.
D'autre part, si le comportement des matériaux de base est raisonnablement connu pour
INDEX çç Sommaire
permettre ces simulations, celui des intercalaires d'une épaisseur de 50 m ne l'est pas et est
difficilement estimable du fait de la diffusion des éléments. Les calculs réalisés n'ont donc pas
pris en compte la présence de ces intercalaires. Les contraintes normales s'appliquant sur la
jonction proviennent des contractions relatives des pièces massives de part et d'autre et
peuvent donc néanmoins être utiles à l'optimisation de la jonction. Sur la figure 5 sont portées
les contraintes normales le long du joint au cours des différentes étapes dans le cas où l'inox
est à l'extérieur. Il apparaît que la partie du joint débouchant vers l'extérieur (zone CD sur
figure 4) est constamment en traction. En revanche les zones AB et BC sont en compression
après usinage et passent légèrement en traction à 300°C, 155 bars.
Refroidissement
Usinage
Pression + Température
A B
C D
Figure 5 : Contrainte normale (MPa) le long de l'interface – 316L extérieur (les lettres
font référence à la figure 4)
Les résultats numériques montrent d'autre part que les contraintes normales au joint en
condition d'essais sont principalement dues au chargement thermique et très peu à la pression.
Le cyclage thermique lié aux différentes campagnes d'essai sera donc un point critique à
prendre en compte pour l’évaluation du bon comportement des jonctions en service.
Comme nous l'avons présenté en introduction, la jonction tubulaire hétérogène 316L / Zy-4
est destinée à être introduite dans une boucle d'un réacteur nucléaire expérimental. Un suivi
très précis de son développement, des étapes de fabrication mises en place et de sa
caractérisation est donc nécessaire. Cela entraîne, au fur et à mesure de l'avancement du
projet, la rédaction de procédures spécifiques qui permettront d'assurer la qualité de la
jonction, la reproductibilité du procédé et la conservation du savoir-faire ainsi développé. En
particulier, le nettoyage des différentes pièces ainsi que la mise en place des intercalaires
doivent être particulièrement soignés afin d'éviter des zones de non recouvrement entraînant
la formation d'intermétalliques fragilisant la jonction.
CARACTERISATION DE LA JONCTION
Analyses métallurgiques
INDEX çç Sommaire
La figure 6 présente une coupe de la jonction réalisée après CIC associée à un profil de
diffusion obtenu par EDS. L'image MEB permet de vérifier la propreté du joint. Les distances
de diffusion sont au maximum de l'ordre de 20 m de part et d'autre de chaque interface. A
l'interface Nb / Zy-4, une zone de Zy-4 enrichie en Nb est présente. Elle correspond à
l'augmentation de dureté observée sur la filiation de microdureté (figure 7). L'absence de
phase intermétallique est confirmée par le profil de diffusion ainsi que par les résultats de
microdureté.
Zry-4
Nb
53m
45m Cu
316 L
INDEX çç Sommaire
Comportement à la corrosion
Afin de déterminer la tenue d'une telle jonction dans l'environnement d’une boucle en eau
sous pression, des essais de corrosion à 300°C et 90 bars dans de l'eau déminéralisée traitée
avec de la lithine et de l’hydrazine ont été réalisés. Des échantillons ont été prélevés à
différentes étapes jusqu'à 6 mois maximum (correspondant à la durée de vie de
fonctionnement estimée). Ces essais de corrosion ont permis de montrer que pour la durée
concernée, il n’y avait pas d’évolution significative du joint soudé. Ces essais ont toutefois
mis en évidence la formation d'une couche de ZrO2 pénétrant à l'interface Zy-4 / Nb et la
formation d'une microcrevasse à l'interface 316L / Cu. Après 6 mois (figure 8), la couche
d'oxyde sur le Zy-4 atteint environ 5 m et pénètre d'environ 10 m le long de l'interface avec
le Nb. La microcrevasse observée entre le Cu et le 316L atteint quant à elle, une profondeur
de 30 m. Excepté ces deux singularité, l'analyse MEB des différentes interfaces ne montrent
pas de différence entre le cœur et les zones débouchantes.
Zr Cu
ZrO2
316 L microcrevasse
Nb
Essais mécaniques
Afin de contrôler la tenue mécanique des jonctions Z, des éprouvettes de traction ont été
prélevées sur le joint plan externe de la jonction Z et testées à 20°C et à 345°C. On observe à
l'ambiante une rupture soit dans le cuivre, soit dans le Zircaloy loin du joint. Dans ce cas, les
valeurs de contrainte à rupture sont en général plus faibles que pour une rupture au joint
(figure 9 a). Lorsque la rupture a lieu dans le cuivre, l'analyse des faciès de rupture ainsi que
des coupes transversales montrent clairement que la fissure se propage dans le cuivre et non le
long des interfaces initiales. Des essais de résilience à l'ambiante sur des éprouvettes de type
KCU (essai considéré comme discriminant vis à vis de la bonne qualité de la jonction),
prélevée sur le joint interne de la jonction Z, confirment ce type de la propagation : on
observe ainsi un amorçage de la fissure dans le cuivre puis une propagation dans le niobium
(figure 9 b). En ce qui concerne des jonctions obtenues par soudage-diffusion, la confirmation
que la rupture n'est pas liée à une ouverture des interfaces initiales est un gage de bonne
qualité.
INDEX çç Sommaire
(a) (b)
Figure 9 : (a) Traction 20°C (rupture dans le Zy-4) – (b) Résilience 20°C (rupture mixte
Cu et Nb)
L'ensemble des essais de traction réalisés confirme que le joint présente une tenue mécanique
assez reproductible (que l'on peut grossièrement estimer à environ 480 MPa pour la géométrie
d'éprouvette utilisée). En revanche, il apparaît que les propriétés du Zircaloy-4 traité CIC sont
plus hétérogènes que celles du Zircaloy-4 n'ayant pas subi ce traitement.
D'autre part, des essais de fatigue à 20°C et à 345°C sous vide ont permis d'évaluer d'une part
la tenue du zircaloy-4 traité CIC et d'autre part celle du joint. Les sollicitations imposées ont
été déterminées à l'aide des calculs thermomécaniques simulant le passage de la température
ambiante à la température d'emploi de ces jonctions. Les résultats obtenus à 345°C ont pu être
comparés à ceux du zircaloy (figure 10) présentés dans le RCC-MX [6] (règles de conception
et de construction des matériels mécaniques des réacteurs expérimentaux, de leurs auxiliaires
et des dispositifs d'irradiation). Sur cette figure, les points rouges sont les points
expérimentaux obtenus sur les éprouvettes avec joint. Celles-ci présentent un nombre de
cycles à rupture plus faible que le Zircaloy-4 (courbe bleue). Mais l'évolution de la courbe t
imposée en fonction du nombre de cycles à rupture suit approximativement celle du Zircaloy-
4. Notons que si ces essais de fatigue sur éprouvettes hétérogènes apportent de précieuses
informations sur le plan qualitatif, ils doivent être interprétés avec précaution du fait de la
répartition hétérogène, et difficilement accessible, de déformation dans les différents
matériaux.
Pour la géométrie d'éprouvettes utilisées, on remarque que les résultats obtenus avec les
éprouvettes bimétalliques correspondent aux valeurs minimum pour le Zircaloy-4 (courbe
noire). L'analyse des éprouvettes rompues indique que la rupture de la jonction est toujours
INDEX çç Sommaire
localisée dans le cuivre et sans décohésion des interfaces. Ce résultat confirme la bonne tenue
mécanique des jonctions soudé-diffusées obtenues, au regard du fonctionnement envisagé
associé à un nombre de cycles très limité.
Parmi ces contrôles, certains ont fait l’objet de mise au point pour être adaptés à la géométrie
particulière du joint ou de la pièce. C’est le cas du ressuage interne mis en œuvre dans un
diamètre intérieur de 21 mm, des ultrasons et du test en pression. Ces mises au point ont été
réalisées par l’APAVE (agence de TASSIN) en collaboration avec BABB Co pour le
ressuage, General Electric Inspection Technologies pour les ultrasons et CERCA pour le test
en pression.
La mise au point des ultrasons a nécessité la réalisation de trous à fond plat étalons de 0.5 mm
à 2 mm localisés précisément dans les différentes zones du joint à contrôler : couronne
externe, surface tronc conique, couronne interne et joints débouchant.
Le tronc conique est analysé par un traducteur de fréquence 15 MHz travaillant en émission -
réception d’ondes longitudinales (figure 11). Il est équipé d’une semelle en plexiglas adaptée
à la géométrie de la jonction.
La couronne interne est vérifiée par une méthode dite delta. Elle consiste à faire vibrer toute la
largeur de la couronne interne par un traducteur d’angle émettant des ondes transversales et à
détecter les ondes longitudinales diffractées en direction de la surface au droit de la couronne
(figure 11). La fréquence utilisée est 5 MHz. Le dispositif est équipé de 2 capteurs séparés (un
pour l’émission et l’autre pour la réception) montés sur une semelle mise en forme au rayon
de courbure de la jonction.
La couronne externe est elle aussi contrôlée par la méthode delta. Cependant, l’exploration de
la totalité de la couronne nécessite le choix de 2 positions différentes d’émetteur : une pour le
centre de la couronne et une pour la partie la plus intérieure de la couronne (figure 11).
Finalement, la partie débouchante de la couronne externe est contrôlée par un examen en
ondes rampantes. Un traducteur 5 MHz à émission et réception séparées génère des ondes de
surface qui se réfléchissent sur le joint débouchant (figure 11).
INDEX çç Sommaire
Figure 11 : Schéma de principe des contrôles par US développés
Les mesures réalisées sur les trous étalons montrent que l'on peut détecter des défauts dont le
pouvoir réflecteur équivaut à des trous à fond plat de diamètre compris entre 1 et 2 mm selon
la zone analysée. Ces développements ont permis de concevoir les procédures de contrôles
appropriées au contrôle des jonctions de fabrication.
Finalement, il est crucial de pouvoir tester quelques jonctions dans des conditions proches de
la réalité. En effet, les essais mécaniques réalisés sur des éprouvettes usinées dans les
jonctions hétérogènes ne permettent pas de reproduire les états de contraintes obtenus en
cours d'essai. Pour cela, un dispositif permettant de soumettre la jonction hétérogène à des
cycles de pression et température correspondant aux conditions d'utilisation du réacteur
CABRI a été réalisé par l'IRSN au CEA Cadarache (figure 12). Sur ce banc, des
thermocouples permettent de suivre en continu la température proche des interfaces au cours
des cycles de température imposés. D'autre part, l'intérieur de la jonction est baigné dans un
fluide identique à celui de CABRI. Cet essai est donc aussi représentatif en terme de corrosion
de la jonction.
INDEX çç Sommaire
Le programme de qualification a été déroulé sur trois jonctions hétérogènes issues :
- de la phase de faisabilité,
- de la phase de QMOS,
- de la phase de production.
Ce programme de cyclages a consisté pour chacune des trois jonctions en :
- une première phase représentative du premier programme d’essais prévu (CIP)
consistant à reproduire :
o 3 épreuves hydrauliques,
o 50 cycles à 155 bars et 300 °C,
o 10 cycles à 185 bars et 320 °C.
- une deuxième phase représentative de la durée de vie des jonctions hétérogènes
consistant à enchaîner trois séquences de :
o 25 fois : 3 cycles à 155 bars – 300 °C + 1 cycle à 185 bars – 320 °C.
A la fin de chacune des séquences d’essais, chacune des jonctions a subit les mêmes CND que
ceux réalisés lors de la recette, soit :
- un contrôle visuel,
- un contrôle par ressuage fluorescent interne et externe,
- un contrôle par ultrasons,
- un test d’étanchéité hélium.
En parallèle de ce programme de qualification en cyclages, on peut également noter la
réalisation d’une épreuve hydraulique à très haute pression. Afin de démontrer la bonne tenue
mécanique de la liaison, un essai à plus de 2000 bars a été mis en œuvre et n’a conduit ni à la
destruction de la jonction, ni à sa perte d’étanchéité, ni même à des déformations rémanentes.
CONCLUSIONS
INDEX çç Sommaire
- les contraintes liées à une utilisation dans le milieu nucléaire ont imposé un suivi
qualité supplémentaire lors de la phase de mise au point des jonctions. Des procédures
précises ont du être rédigées afin de décrire précisément chaque étape du procédé et de
garantir une bonne reproductibilité,
- de plus, des développements de contrôles non destructifs spécifiquement adaptés à la
géométrie des jonctions ont été entrepris. Ils ont permis de définir les procédures
nécessaires aux contrôles des jonctions de production.
Remerciements
Les auteurs remercient l'IRSN pour le financement d'études participant à la mise au point de
ces jonctions hétérogènes zircaloy/inox par soudage-diffusion.
Références :
[1] Shaaban H.I., Hammad F.H., Baron J.L., Diffusion-bonding of stainless steel to
zircaloy-4 in presence of an iron intermediate layer, Journal of Nuclear Materials,
78, pp 431-433, 1978.
[2] Huet M., Hourcade M., Hocheid B., Soudage diffusion zircaloy-4 / acier Z2CN18-
10 avec interposition de métaux intermédiaires, Mémoires et Etudes Scientifiques
Revue de Métallurgie, pp 313-325, juin 1988.
[3] Kale G.B., Bhanumurthy K., Ratnakala K.C., Khera S.K., Solid state bonding of
zircaloy-2 with stainless steel, Journal of Nuclear Materials, 138, pp 73-80, 1986.
[4] Bhanumurthy K., Krishnan J., Kale G.B., Fotedar R.K., Biswas A.R., Arya R.N.,
Tubular transition joint between zircaloy-2 and stainless steel, Journal of materials
processing technology, 54, pp322-325, 1995.
[5] RCC-M, volume B, paragraphe B 3641.
[6] RCC-MX, CEA
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page137
RÉSUMÉ / ABSTRACT
97P
RÉALISATION D’UNE ENCEINTE SOUS PRESSION EN ZIRCALOY POUR
LE RÉACTEUR DE RECHERCHE CABRI
RÉSUMÉ
Le réacteur de recherche Cabri est implanté sur le centre d’études nucléaires de Cadarache. Ce réacteur
est exploité par le CEA/DER et dédié à la réalisation d’essais de sûreté de l’IRSN. Dans le cadre du
programme international CIP piloté par l’IRSN sous un accord OCDE, une boucle expérimentale
permettant de recréer les conditions thermo hydrauliques équivalentes à celles d’un réacteur REP a été
réalisée. L’enceinte expérimentale destinée à recevoir les échantillons traverse le cœur nourricier du
réacteur. Pour des raisons de transparences neutroniques, cette enceinte et son confinement ont été
réalisés en alliage de zirconium, ceci permettant d’avoir un meilleur couplage neutronique entre le cœur
cabri et l’échantillon testé.
Le code RCC-MX (Règles de Conception et de Construction des Matériels mécaniques des Réacteurs
Expérimentaux) dont le développement a été lancé par le CEA en 1998 n’étant pas disponible lors du
lancement de la phase de réalisation, cette enceinte dimensionnée pour résister à une pression de 190
bars et 355°C a été réalisée suivant un niveau de qualité équivalent aux règles du niveau 1 du RCC-M.
L’établissement de ce référentiel (le zircaloy n’étant pas référencé dans le RCC-M), sa justification et
sa mise en application sont présentés en insistant sur chacune des phases de la réalisation :
approvisionnement, caractérisation mécanique, fabrication, soudage et contrôles.
ABSTRACT
The Cabri research reactor is located at the Cadarache nuclear research center. This reactor is operated
by CEA/DER and devoted to IRSN safety programmes. In the frame of the Cabri International Program
(CIP) carried out by IRSN under an OECD agreement, a new experimental loop to recreate the thermo
hydraulic conditions equivalent to those of a pressurized water reactor was implemented. The
experimental vessel linked to the loop and receiving samples goes through the core of the reactor. For
reasons of neutron transparency, this vessel and its containment have been made of zirconium alloy,
this allows a better neutronic coupling between the core and the sample tested.
The code RCC-MX (Rules of Design and Construction for Mechanical Materials of Experimental
Reactor), whose development has been launched by the CEA in 1998, was not available when launching
the implementation phase, so the vessel design (to resist to a pressure of 190 bar and 355 °C) was
performed according to a quality level equivalent to the rules of level 1 of the RCC-M.
The establishment of these rules (the zircaloy is not referenced in the RCC-M), its justification and its
implementation are presented with emphasis on all phases of production: supply, mechanical
characterization, fabrication, welding and controls.
INDEX çç Sommaire
ABREVIATIONS
ASTM : American society for testing and material
CEA : Commissariat à l’énergie atomique et aux énergies alternatives
CIC : Compression isostatique à chaud
CIP : Cabri international program
CND : Contrôle non destructif
DEN : Direction de l’énergie nucléaire
DER : Département d’études des réacteurs
ED : Essais destructifs
EP : Enceinte EP (premier confinement)
FE : Faisceau d’électrons
IRSN : Institut de radioprotection et de sûreté nucléaire
OCDE : Organisation de coopération et de développement économiques
PCS : Probe center separation (distance entre les points d’émergence des faisceaux ultra-sons)
PE : Pulse écho
QMOS : Qualification de mode opératoire de soudage
QS : Qualification de soudeur
RCC-M : Règles de conception et de construction des matériels mécaniques des réacteurs
REP
RCC-MX : Règles de conception et de construction des matériels mécaniques des réacteurs
expérimentaux
REP : Réacteur à eau pressurisée
RIA : Reactivity initiated accident
RX : Radiographie par rayons X
TIG : Tungsten inert gas
TOFD : Time of flight diffraction
TS : Tube de sécurité (deuxième confinement)
US : Ultrasons
INDEX çç Sommaire
INTRODUCTION
Le réacteur de recherche CABRI, implanté sur le centre du CEA à Cadarache, est utilisé pour
des études de sûreté liées au comportement des crayons combustibles (gainage et
combustible) des réacteurs nucléaires en situation accidentelle de type RIA (Reactivity
Initiated Accident). Initialement conçue pour les combustibles des réacteurs à neutrons
rapides refroidis au sodium, l’installation est actuellement significativement modifiée pour
l’adapter aux études des combustibles des réacteurs à eau dans le cadre d’un programme
international, le programme CABRI Boucle à eau, piloté par l’IRSN et placé sous l’égide de
l’OCDE.
Les essais réalisés sur ce réacteur consistent à soumettre l’échantillon de crayon combustible
testé à un transitoire rapide de puissance représentatif de la situation accidentelle
correspondant à l’éjection d’une barre de commande.
Les travaux présentés ci-après concernent principalement les composants de cette boucle
contenant le crayon expérimental et traversant le cœur nourricier. Pour des raisons présentées
dans le deuxième paragraphe, ces composants ont du être réalisés en alliage de zirconium.
Le code RCC-MX (Règles de Conception et de Construction des Matériels mécaniques des
Réacteurs Expérimentaux) dont le développement a été lancé par le CEA en 1998 n’étant pas
disponible lors du lancement de la phase de réalisation, cette enceinte dimensionnée pour
résister à une pression de 190 bars et 355 °C et son enceinte de confinement ont été réalisées
suivant un niveau de qualité équivalent aux règles du niveau 1 du RCC-M (Règles de
Conception des Matériels Mécaniques applicables au réacteur de puissance REP).
INDEX çç Sommaire
L’établissement de ce référentiel spécifique (le zircaloy n’étant pas référencé dans le RCC-
M), sa justification et sa mise en application sont donc présentés pour chacune des phases de
la réalisation : approvisionnement, caractérisation mécanique, fabrication, soudage et
contrôles.
Ce document s’intéresse aux deux composants que sont : l’enceinte EP et son Tube de
Sécurité, tous deux réalisés en alliage de zirconium.
INDEX çç Sommaire
Le principe d’un essai étant de coupler neutroniquement le cœur nourricier du réacteur Cabri
avec l’échantillon de combustible testé, il est indispensable de limiter au strict minimum les
écrans neutroniques entre ces deux éléments. Il a donc été retenu le Zircaloy 4 comme
matériau pour les structures de la cellule en pile. En effet, cet alliage présente la particularité
d’être relativement transparent aux neutrons, ce qui améliore le couplage, tout en présentant
des caractéristiques mécaniques élevées nécessaires pour le dimensionnement des structures
aux différents chargements dont la pression. D’autre part, les alliages de zirconium sont
classiquement utilisés comme matériau de gainage de combustible et plus épisodiquement
comme matériau de structures dans des applications spécifiques liées à l’industrie nucléaire
ou chimique, ce qui permet d’accéder une base documentaire traitant de certaines
problématiques.
L’enceinte EP (figure 4) est un cylindre de 8980 mm de long reposant par sa bride supérieure
sur le haut du tube de sécurité. Cette disposition, enceinte suspendue, permet la libre dilatation
des structures lors des mises en chauffe.
Elle présente successivement (du haut vers le bas) des diamètres intérieurs de 145, 140, 138,
98, 96, 92, 85, 82 et 71 mm. Son épaisseur varie entre 22,5 et 15,5 mm. Différents piquages y
sont reliés dont les principaux sont :
- au niveau - 8857 mm : entrée Eau Pressurisée (4 tubes de diamètre 11,176 mm),
- au niveau - 5314 mm : sortie Eau Pressurisée (1 tube de diamètre 31,75 mm),
- au niveau – 3851 et - 4220 mm : piquages gaz (tube de diamètre 11,176 mm).
Au niveau de sa partie centrale (de - 3875 à - 4571 mm), un échangeur à air a été intégré afin
de pouvoir, si nécessaire, assurer un refroidissement de la zone des joints du dispositif
expérimental.
Aux niveaux - 7680 et - 6240 mm, 2 bagues extérieures assurent le centrage de la partie basse
de l’enceinte EP dans le Tube de Sécurité, avec un jeu de 0,5 mm au rayon.
Celui-ci est constitué de 2 tubes reliés au niveau - 5200 mm par un cône.
Les tubes inférieur et supérieur présentent respectivement un diamètre intérieur de 154 mm /
380 mm et une épaisseur de 7 mm / 9 mm.
INDEX çç Sommaire
T19 + T20 T18
T4
T16
T15
T14
T13
T12
T11
T10
T9
T8
T7
T6
T5
T1 / 2 / 21 / 3
ZIRCALOY – APPROVISIONNEMENTS
La nuance retenue est l’alliage de zirconium Zircaloy 4. Le choix de cette nuance s’explique
par ses bonnes caractéristiques mécaniques et sa bonne tenue à la corrosion. De plus, cette
nuance est régulièrement produite comme matériau de gainage de combustible, ce qui en
facilite l’approvisionnement. En complément de ces éléments, cette nuance a déjà été utilisée
par le passé au CEA pour la réalisation de dispositifs expérimentaux, ce qui a permis de
remonter aux dossiers de caractérisation du matériau.
D’une manière générale, la composition chimique retenue est celle du grade R60804 de
l’ASTM. Toutefois, concernant les barres destinées à la zone sous flux neutronique et en vue
d’améliorer les propriétés de résistance mécanique et de tenue au fluage, la spécification de la
composition chimique a été ajustée de la manière suivante :
- la valeur cible pour la teneur en étain était de 1,7%, ce qui correspond à la borne haute
de l’ASTM,
- de même pour la valeur cible de la teneur en oxygène fixée à 0,16%, ce qui est
également la borne haute,
- une spécification pour la teneur en souffre (de 10 à 50 ppm) a également été fixée
alors qu’elle n’existe pas dans l’ASTM.
Une autre particularité de ces barres destinées à la zone sous flux vient de leur état
métallurgique de livraison. Alors que les produits en zircaloy 4 sont couramment livrés dans
un état dit recristallisé, ces dernières ont été livrées dans un état ‘’bêta’’ résultant d’une
trempe réalisée après les opérations de forgeage. Cette trempe en phase finale permet
d’effacer l’anisotropie créée par l’orientation des grains lors des opérations de forge. Il en
résulte que les propriétés mécaniques obtenues sont beaucoup plus homogènes dans les deux
sens (long et travers) alors que cet alliage présente couramment une différence marquée.
Pour répondre au besoin de la fabrication, quatre types de produits ont été approvisionnés :
- des barres de diamètres allant de 140 mm à 230 mm,
- des largets parallélépipédiques de 540 X 540 X 100 mm,
- des tôles de largeurs de 445 mm (limite fixée par les capacités du fournisseur),
d’épaisseurs de 7 à 12mm et de longueurs pouvant aller jusqu’à 3 mètres,
- des tubes de diamètres externes allant de 11 mm à 32 mm.
INDEX çç Sommaire
L’approvisionnement des barres, tôles et largets n’a pas posé de problèmes particuliers. Ces
types de produits font partie de produits (ou demi-produits) classiquement fournis et
contrôlés. Ces contrôles de recette et les essais complémentaires ont par ailleurs permis de
montrer une homogénéité satisfaisante des produits. Le lancement de l’approvisionnement des
tubes a demandé nettement plus d’investissement. Les dimensions particulières des petits
tubes notamment (diamètre 11 mm en épaisseur 1,5 mm) ne faisant pas partie des productions
courantes, la spécification technique et plus particulièrement la spécification particulière de
recette ont nécessité beaucoup d’échanges avec le producteur.
Zone
prolongateur
Zone sous pression Zr4 _ Zr4 _ _
E1 E2 E4 E6 E8 E10 E14
ZIRCALOY – CARACTERISATIONS
L’établissement de la base de données s’étoffant au fil des collectes de données et des essais
matériau, le projet n’a pas pu disposer dès le départ d’une vision d’ensemble des propriétés
matériau. Au gré des besoins de données pour le dimensionnement, six jeux de données
spécifiques Cabri ont été déterminés en fonction des types de produits et de leur état
métallurgique. Dans le cadre de ce document, il n’a pas été retenu de présenter ces jeux de
données de manière isolée, mais plutôt de situer les valeurs minimales du RCC-MX 2008 par
rapport à la base de données. Il est retenu comme base de référence dans le code, un matériau
en alliage de zirconium ASTM R60804 de classe 1 (1000 à 1500 ppm) à l’état recristallisé.
Les résultats présentés sont en partie issus de ce type de produits mais couvrent également des
structures à l’état traité . Il n’est pas fait de distinctions entre les différents types de produits
(barre, tôle, larget ou tube), ni des sens de prélèvements. Bien que ces éléments influencent
les propriétés mécaniques du zircaloy, elles ne sont pas précisées dans l’annexe X3 du code
qui donne des valeurs minimales.
INDEX çç Sommaire
Propriétés de traction
Un peu plus de 700 essais de traction ont été enregistrés dont plus de la moitié ont été réalisé
sur des produits réalisés pour le projet Cabri. Quinze coulées différentes de zircaloy 4 ont
ainsi pu être testées. Concernant la résistance à la rupture, l’ensemble des points (tous
produits confondus) est enveloppé par la courbe minimale du RCC-MX (figure 5). Dans le
domaine de 20°C à 200°C, on peut noter une marge de 30 à 40 MPa par rapport à cette courbe
minimale. Cette observation reste valable si on ne compare que des produits de type
recristallisé. Pour ce qui est de la limite élastique, les conclusions sont du même ordre :
l’ensemble des points est enveloppé par la courbe minimale et toujours dans le domaine de
20°C à 200°C, une marge existe (un peu plus faible toutefois), de l’ordre de 20 MPa (figure
6). Un dépouillement complémentaire doit être réalisé concernant les valeurs d’allongement
pour lesquelles les prescriptions du RCC-MX semblent être en limite des propriétés du
matériau.
700
600
100
0
0 200 400 600 800 1000
Température (°C)
INDEX çç Sommaire
Limite élastique Rp02
700
600
500
1 - Isabelle 2007
Rp02 (MPa)
2 - Cabri 2006
0
0 200 400 600 800 1000
Température (°C)
Un zoom particulier est fait sur l’approvisionnement spécifique destiné à la zone sous flux.
Les évolutions apportées à la composition chimique et à la gamme de fabrication ont permis
d’augmenter ses propriétés mécaniques. Il est constaté sur la limité élastique à 20° et 300° et
sur la résistance à la rupture à 20° une augmentation des valeurs minimales mesurées de 50
MPa. L’augmentation de la résistance à la rupture à 300°C est plus limité (+25 MPa sur la
valeur minimale).
Propriétés de ténacité
Tout comme pour les essais de traction, un important programme de caractérisation en
ténacité du zircaloy 4 a été réalisé. Ce programme a conduit à réaliser plus de deux cents
essais dont la moitié a été réalisée à la température ambiante. Ces essais ont concernés les
différents types de produits et les différentes coulées mais également un certain nombre de
soudures réalisées par un des procédés suivants : faisceau d’électrons, TIG ou Plasma.
On peut constater (figure 7) que quelque soit le produit ou la coulée et que l’on s’intéresse au
métal de base ou aux soudures, les valeurs de ténacités minimales mesurées sont toujours
faibles et comprises entre 30 et 40 KJ/m2. A contrario, les valeurs maximales présentent une
énorme dispersion pouvant aller jusqu’à un facteur 4.
INDEX çç Sommaire
Zirc aloy rec ris tallis é - Barre - Diamètre 124mm -
Zircaloy 4 - Ténacité - 20°C Coulée 265346 - Code SRMA A 36 - "Barre
Phébus "
Zirc aloy bêta - Barre - Diamètre 229 mm - Coulée
272794 - Code SRMA A 37
100
KJ/M2
Zircaloy 4
180
160
140
120
Ténacité (kJ/m2)
100
Zircaloy 4
80
60
40
20
0
0 50 100 150 200 250 300 350
Tem pérature (°C)
INDEX çç Sommaire
Propriétés de fatigue et de fluage
Concernant ces propriétés, les données utilisées par le projet sont issues d’essais de
caractérisations en fatigue et en fluage réalisés initialement en support à des études de
dimensionnent de l’ancienne boucle Cabri – sodium et réalisés sur du zircaloy 4 recristallisé.
Des essais complémentaires destinés à s’assurer que les valeurs utilisées couvraient bien les
produits Cabri, et notamment les produits bêta, ont été réalisés. Ainsi des essais de fluage aux
températures de 300, 350, 375 et 400 °C ont été mise en œuvre avec des niveaux de
contraintes conduisant à des ruptures dans une gamme de temps de quelques centaines à
quelques milliers d’heures (cette gamme étant adaptée à la durée de fonctionnement en
température des structures).
Concernant la fatigue, différents essais ont permis :
- d’établir une courbe de fatigue à 400 °C,
- d’établir une loi de Paris à 300 °C,
- de vérifier que les propriétés utilisées étaient bien applicables aux joints soudés.
Lors de chacune de ces campagnes de caractérisations, les propriétés ont systématiquement
été comparées aux valeurs du RCC-MX et non révélés aucune incohérence.
ENCEINTE EP – FABRICATION
Adaptation du référentiel de fabrication
Avant toutes opérations de fabrication ou de qualification des procédés spéciaux le RCC-M,
applicable au projet, a nécessité plusieurs adaptations nécessaires à la mise en œuvre du
Zircaloy 4.
Les adaptations du référentiel, qui ont été nécessaires à ces fabrications, portent sur les sujets
suivants :
- Essais de dureté : définition d’une dureté maximale admissible à 260 HV10,
- Essais de pliage : modification du diamètre du mandrin de pliage à 10 X l’épaisseur,
- Essais de corrosion : définition des essais de corrosion suivant ASTM G2,
- Contrôle radiographique : utilisation d’IQI à trou en Zirconium conforme à l’ASTM E1025
en adéquation avec les épaisseurs radiographiées en X,
- Contrôle par ressuage des chanfreins : application d’un contrôle visuel renforcé en
remplacement des ressuages pour éviter les problématiques liées aux pollutions,
- Qualification de soudage :
o définition des modalités de QMOS,
o application de la norme NF EN ISO 9606-5 pour les qualifications de soudeur (QS),
o définition des modalités de recette des métaux d’apport en Zircaloy 4. Les fils utilisés
lors d’une QMOS sont qualifiés sur la base des résultats de la QMOS. Seule est alors
réalisée en complément une analyse chimique hors dilution suivant ASTM E146
conformément aux valeurs spécifiées pour la nuance UNS R60804 de l’ASTM B350.
- Qualification de formage : exigence d’une qualification du mode opératoire de formage quel
que soit l’allongement et définition des modalités de qualification de formage,
- Mise en propreté : définition des modalités de mise en propreté, dont les modalités de
décapage du Zircaloy 4.
INDEX çç Sommaire
Formage
La tôle de fermeture de l’échangeur air de l’EP ( 222 x ép. 4 mm) a été qualifiée et réalisée
par formage à la presse plieuse.
L’ensemble des formages et des cintrages réalisés sur l’enceinte EP et le Tube de Sécurité ont
été réalisés à froid. Les formages n’ont pas subi de traitement thermique de détentionnement.
La principale problématique rencontrée a concerné l’adaptation du diamètre du mandrin de
formage afin d’éviter de dépasser les capacités de déformation des produits.
Les CND réalisés sur les formages de qualification sont
identiques aux contrôles des formages de production. Ils
comportent les contrôles suivants :
- Avant formage : Visuel et ressuage des tôles ou des
tubes,
- Après formage et soudage : Visuel et ressuage des
zones cintrées extrados et intrados (lorsque
possible), et contrôle géométrique (gabarit) des
formages,
Le Tube de Sécurité est constitué d’une partie inférieure usinée dans la masse, comportant un
forage borgne sur trois mètres. Cette conception a permis de s’affranchir de toutes soudures
sous flux.
L’EP est constituée de tronçons usinés dans des ronds pleins de Zircaloy 4 de nuance pour
la zone supérieure et de nuance pour la zone inférieure (fonctionnant aux conditions REP).
L’ensemble de l’EP a fait l’objet de ré-usinages externe et interne pour garantir sa rectitude.
Les différents piquages ont été réalisés sur pièce finie par fraisage.
INDEX çç Sommaire
Soudage
Procédés de soudage
Les procédés de soudage retenus pour la réalisation des soudures des deux composants sont :
- Pour l’enceinte sous pression :
o Soudage par faisceau d’électrons (FE) sur latte support provisoire du corps,
o Soudage TIG double flux avec métal d’apport Zr4 bout à bout de la tôle
d’échangeur,
o Soudage TIG avec métal d’apport Zr4 bout à bout et emmanché soudé des
tuyauteries,
- Pour le Tube de Sécurité :
o Soudage Plasma avec métal d’apport Zr4 des soudures longitudinales,
o Soudage Plasma ou TIG avec métal d’apport Zr4 des soudures circulaires,
Avant soudage :
Après soudage :
INDEX çç Sommaire
Le soudage Plasma
Le soudage multipasses Plasma, avec métal d’apport Zircaloy 4, a été retenu pour les
soudures des composants chaudronnés (tôles envirolées du TS). Malgré la qualité du formage,
les plus petits défauts d’accostages (ovalisation, écartement) sont incompatibles avec le
procédé FE. Comme pour tous les procédés, la qualité de la protection du bain fondu des
pollutions gazeuses est primordiale. Dans le cadre du soudage plasma (comme par ailleurs
pour le soudage TIG), cette protection est assurée par la mise en place de chambre inertée. La
qualité de l’inertage est garantie par mesure d’oxygène.
Le soudage TIG
Le soudage TIG, avec métal d’apport
Zircaloy 4, a été retenu pour les soudures
bout à bout et de piquage (type emmanché
soudé) des tuyauteries sur le corps de l’EP
ainsi que pour le soudage de la tôle de
l’échangeur.
INDEX çç Sommaire
Les soudures des piquages du pied et de la tôle de l’échangeur de l’EP ont fait l’objet de
maquettage de validation des conditions de soudage (accessibilité, inertage, …).
INDEX çç Sommaire
Pour se prémunir d’un éventuel effet d’entaille de type manque de fusion, les CND
des soudures FE de l’EP ont été complétés des contrôles
suivants :
o contrôle de surface télévisuel (camera x 24) envers,
o contrôle par méthode ultrasonore de la compacité
orientée sur un défaut de type manque de fusion du plan
de joint,
Figure 9 : Illustration d’un
manque de fusion du plant de
Recette des équipements joint
Les équipements ont fait l’objet lors de leur recette des
contrôles suivants :
- dimensionnel, incluant un essai d’insertion de l’EP
dans le TS,
- épreuve hydraulique,
o 40 bars à température ambiante pour le TS,
o 285 bars à T > 50°C pour l’EP,
- test d’étanchéité hélium (après épreuve).
INDEX çç Sommaire
- d’effectuer une inspection télévisuelle de l’intérieur de l’enceinte EP.
Le point zéro de l’enceinte EP a été réalisé en décembre 2009. Un cycle vidéo, un cycle
PULSE ECHO et 2 cycles TOFD (l’un pour les diamètres 71 à 92, l’autre pour les diamètres
145) se sont déroulés successivement suivant le mode opératoire et les procédures validées en
INDEX çç Sommaire
phase de développement Les mesures ultra sonores et les acquisitions ont été réalisées en
présence d’un contrôleur COFREND 3, de manière à valider en temps réel les acquisitions.
L’inspection visuelle de l’enceinte EP n’a pas fait apparaître de « singularité » de la surface
interne de l’enceinte EP.
Le contrôle par la méthode TOFD des soudures et de la zone cœur n’a pas détecté de défauts
longitudinaux et transversaux situés en surface externe ou interne de l’enceinte EP.
Les mesures d’épaisseur et de diamètre par la méthode PE sont cohérentes avec les
informations issues de la fabrication de l’enceinte EP. Cette étape constitue le point de
référence pour le suivi en service de l’équipement.
RETOUR D’EXPERIENCE
Le projet Cabri a été lancé à une époque où aucun code de conception et de réalisation
nucléaire couvrant le zircaloy n’était disponible. Devant les problèmes concrets de conception
et de réalisation, le projet Cabri a développé un référentiel adapté au zircaloy en partant du
RCC-M. La rédaction du code RCC-MX se déroulant en parallèle, la communication entre les
équipes a permis de confronter l’expérience industrielle à la codification et d’intégrer une
partie du retour d’expérience dans celui-ci. Une deuxième étape reste à mener pour compléter
la capitalisation de l’acquis. Concernant les sujets techniques, on pourra principalement noter
l’intégration du REX industriel lié aux procédés de soudage, de formage et de contrôle.
Concernant les données matériaux, la capitalisation des essais réalisés en soutien au projet
confirme le bien fondé des valeurs codifiées. Concernant les propriétés de traction, il est
constaté qu’il existe une marge non négligeable entre les valeurs de résistance à la traction les
plus faibles mesurées et la courbe minimale dans le domaine de 20°C à 200°C. Dans ce
domaine, cette propriété étant utilisée pour déterminer la contrainte admissible, un gain
potentiel existe.
CONCLUSION
La réalisation d’équipements sous pression en zircaloy pour le projet Cabri a nécessité de
nombreux investissements financiers, techniques et humains pour garantir le haut niveau de
qualité requis pour un équipement nucléaire.
L’établissement d’un référentiel spécifique pour le projet a permis de capitaliser ce travail.
Aujourd’hui, l’intégration de cette expérience industrielle dans le code RCC-MX est
partiellement achevée.
En parallèle de cette réalisation industrielle, on peut noter le volume important d’essais de
caractérisation mécaniques du zircaloy qui a permis la validation des données retenues pour le
dimensionnement. Ce travail important permet aujourd’hui de conforter les courbes
minimales retenues dans le code par rapport à une population statistique importante.
En remerciant toutes les personnes qui ont permis cette réalisation, tant dans les unités de
l’IRSN et du CEA que dans les entités d’AREVA et chez les autres industriels largement
impliqués.
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page138
RÉSUMÉ / ABSTRACT
102P
EVALUATION DE LA CONFORMITÉ DE L’ENCEINTE À VIDE D’ITER AUX
EXIGENCES DE L’ARRÊTÉ DU 12 DÉCEMBRE 2005 RELATIF AUX ESPN
L’enceinte à vide d’ITER est un équipement sous pression nucléaire à plusieurs compartiments dont la fonction
principale est le confinement du plasma dans une étanchéité spécifiée. Les composants principaux de l’enceinte
à vide sont les neuf secteurs, formant ensemble un tore, les portes et les supports. L’enceinte est une structure
de type caisson. Entre les deux parois se trouvent le fluide de refroidissement et les écrans de protection. Les
raidisseurs poloïdaux et toroïdaux renforcent la structure de l’enceinte.
Sur base des valeurs de la pression, du volume et de la radioactivité du fluide de refroidissement, l’enceinte à
vide est classée comme un équipement sous pression nucléaire de niveau N2 selon l’arrêté français du 12 Dé-
cembre 2005 (ESPN) et catégorie IV selon le décret No. 99-1046 du 13 Décembre 1999. Ce dernier est la trans-
position de la Directive européenne d’Equipements Sous Pression (DESP) dans le droit français.
Ceci implique que l’enceinte à vide doit répondre aux exigences essentielles de sécurité (EES) du DESP et de
l’Arrêté ESPN et aux exigences de radioprotection de l’Arrêté ESPN. Un organisme agréé et mandaté par l’Au-
torité de Sûreté Nucléaire doit évaluer la conformité de l’enceinte à vide à ces exigences. Cette évaluation
concerne toutes les phases du projet : conception, fabrication des secteurs et des portes dans les ateliers, assem-
blage et essais sur site.
POSTERS
ITER Organization (IO) est le Fabricant de l’enceinte à vide selon les termes de la législation. IO a choisi le mo-
dule d’évaluation G de l’annexe 2 du DESP pour démontrer la conformité avec les EES et les exigences de ra-
dioprotection. Le Code de Conception et de Construction français RCC-MR éd.2007 a été sélectionné par IO
pour la conception et la construction de l’enceinte à vide. Après une procédure de sélection, le rôle de l’organisme
agréé a été confié à AIB-Vinçotte International.
La fabrication de l’enceinte à vide est un projet unique et particulier. Plusieurs points requièrent une attention
particulière dès la phase de conception pour assurer que l’enceinte à vide répondra finalement aux exigences de
la législation :
- définition des responsabilités et des interfaces entre les différents participants au projet
- figer une conception qui permettra de réaliser toutes les expérimentations de fusion envisagées
- réaliser une analyse de risques tenant compte des incertitudes liées aux expérimentations de fusion
- définition des exigences de radioprotection
- identification des exigences résultant du programme d’inspection en service
- classification des composants de l’enceinte à vide selon leur localisation et leur fonction
- spécification des propriétés des matériaux et conformité des matériaux avec les exigences du DESP et
de l’Arrêté ESPN
- préparation des spécifications d’approvisionnement des matières
- justification de la tenue structurelle tenant compte de la complexité de l’équipement et de la présence des
charges électromagnétiques
- respect des tolérances de fabrication
- évaluation du code RCC-MR et des procédures de fabrication et des essais alternatifs vis-à-vis des EES
et des exigences de radioprotection
Il est important que tous les participants au projet aient une bonne connaissance de la législation applicable et
se rendent compte de l’impact de leurs décisions sur la conformité finale de l’enceinte à vide aux exigences es-
sentielles de sécurité et les exigences de radioprotection.
INDEX çç Sommaire
INTRODUCTION
L’enceinte à vide d’ITER (VV : Vacuum Vessel) est un équipement sous pression nucléaire à
plusieurs compartiments dont la fonction la plus importante est le confinement du plasma dans une
étanchéité spécifiée. Les composants principaux du VV sont les neuf secteurs, formant ensemble le
réservoir principal (tore), les portes et les supports. La figure 1 donne une vue générale du VV.
L’enceinte est une structure à doubles parois du type caisson. Entre les deux parois se trouvent le
fluide de refroidissement et les écrans de protection. Les raidisseurs poloïdaux et toroïdaux
renforcent la structure de l’enceinte.
Les modules de couverture sont attachés à la paroi interne, côté plasma, du VV. Leur fonction
est, entre autres, de fournir une protection contre les neutrons à haute énergie résultant des réactions
de fusion. Une partie des modules de couverture seront utilisés dans les phases ultérieures du projet
pour tester des matériaux générant du deutérium et du tritium pendant le procédé de fusion.
La figure 2 est une section transversale poloïdale du VV montrant le réservoir principal, les
portes et la structure des supports.
Les conditions de pression et de température de l’eau de refroidissement dans les différentes
parties du VV, conduisent à le classifier comme un ensemble de plusieurs équipements sous
pression nucléaires catégorie IV niveaux N2 et N3 selon la réglementation [1], [2], [3]. Dès lors, les
exigences essentielles de sécurité et les exigences de radioprotection de la réglementation doivent
être satisfaites. Un organisme notifié et agréé doit vérifier la conformité du VV avec ces exigences.
INDEX çç Sommaire
NOMENCLATURE
Le tableau 1 donne les dimensions principales et les conditions de fonctionnement. Les parties
de l’enveloppe sous pression sont en acier austénitique 316L(N)-IG qui est très proche du X2CrNi-
Mo17-12-2 à l’azote contrôlé spécifié dans le RCC-MR [5], mais avec des exigences
supplémentaires.
La figure 3 présente l’arrangement global d’un secteur 40° du VV. Les portes supérieures,
équatoriales et inférieures permettent l’accès aux composants internes pendant l’entretien en
service. Ils fournissent également le passage pour les systèmes d’alimentation et du pompage, les
diagnostiques, les équipements pour le chauffage du plasma, les modules de couverture et les
éléments du divertor. Les supports du VV se trouvent en dessous des portes inférieures.
INDEX çç Sommaire
La paroi du VV est une structure du type ‘caisson’ constitué d’une paroi interne jointe à la paroi
externe par des raidisseurs soudés. Les supports des modules de couverture sont soudés aux parois
interne/externe et contribuent également à la résistance structurelle du VV.
Dimensions
- Tore : diamètre externe 19.4 m
- Tore : diamètre interne 6.5 m
- Tore : hauteur 11.3 m
- Epaisseur de paroi 60 mm
Température d’entrée de l’eau
(secteurs) :
- Fonctionnement normal 100°C
- Chauffage périodique de l’enceinte 200°C pour VV
Pression d’entrée de l’eau (secteurs) :
- Fonctionnement normal 1.1 MPa abs
- Chauffage périodique de l’enceinte 2.4 MPa abs
Pression dans la chambre plasma en 0 MPa abs
situation de fonctionnement (vacuum)
Volume d’eau de refroidissement 200 m3
dans les secteurs (9 secteurs)
Fluide de refroidissement Eau
Masse totale du réservoir assemblé ± 5000 ton
Masse totale en situation de ± 9000 ton
fonctionnement
Matière 316L(N)-IG
ASPECTS REGLEMENTAIRES
Comme le VV est un équipement sous pression contenant un fluide radioactif, les législations
françaises [1], [2] et [3] s’appliquent. Les deux derniers sont l’implémentation de la Directive
Européenne d’Equipements Sous Pression en France. Le Guide [4], établi par l’ Autorité de Sûreté
Nucléaire française, donne les lignes directrices à suivre par les Organismes Notifiés et Agréés dans
leur évaluation de la conformité des Equipements Sous Pression Nucléaires (ESPN).
Tenant compte des valeurs de pression, volume et radioactivité du fluide de refroidissement, le
VV est classifié, selon [1], comme un ensemble des équipements sous pression nucléaires catégorie
IV de niveaux N2 et N3. Le réservoir principal et la plus grande partie des portes forment un
équipement multichambre niveau N2. Des parties limitées des portes supérieures et équatoriales
sont de niveau N3. Ceci nécessite que le VV doit satisfaire les EES de [2] plus les EES et ERP de
[1]. Un Organisme Notifié et Agréé par l’Autorité de Sûreté Nucléaire française doit évaluer la
conformité du VV avec ces exigences.
ITER Organization, le fabricant du VV selon les termes des réglementations [1], [2] et [3], a
opté pour le module d’évaluation G pour démontrer la conformité du VV avec les EES et ERP. Ceci
requiert que l’Organisme Notifié et Agréé réalise des inspections dans toutes les phases du projet.
Après une procédure de sélection, ITER Organization a confié le rôle de l’Organisme Notifié et
Agréé à AIB-Vinçotte International. Le scope de l’évaluation comprend l’approbation de la
INDEX çç Sommaire
Figure 3 : Arrangement général du VV
conception, les inspections chez les constructeurs dans les usines (Europe, Corée du Sud, Fédération
Russe) et les inspections pendant l’assemblage et les épreuves sur le site de Cadarache en France.
INDEX çç Sommaire
Figer une conception permettant l’exécution de toutes les expérimentations de fusion envisagées:
Le Tokamak ITER est toujours un réacteur expérimental et doit permettre d’exécuter une grande
variété d’expérimentations de fusion. Puisque la science de fusion est constamment en
développement, les conditions de conception sont continuellement remises en cause par les
scientifiques. Par conséquent, c’est un grand défi de figer une conception du VV, bien avant que le
premier plasma se produise, qui réponde à l’objectif de fournir une démonstration intégrale de la
faisabilité scientifique et technologique de la production d’énergie par fusion.
Avant le démarrage réel de la fabrication, tous les paramètres de conception doivent être figés
sans induire des restrictions sur le programme des expérimentations de fusion envisagées.
Analyse de risque tenant compte des incertitudes liées aux expérimentations de fusion :
Les réglementations exigent de réaliser une analyse de risque pour identifier toutes les risques
liées au VV et à son utilisation. Parmi les plus grands risques se trouve la défaillance sous les
charges dynamiques électromagnétiques pouvant se produire pendant les expérimentations de
fusion : disruption du plasma, déplacement vertical du plasma et déchargement rapide des aimants.
Comme d’autres réacteurs de la grandeur d’ITER Tokamak n’existent pas, des incertitudes sur
la magnitude de ces charges électromagnétiques doivent être considérées dans le cahier des charges.
Des précautions additionnelles doivent être prises en opération comme le démarrage progressif et
l’installation de l’instrumentation pour le monitoring des charges électromagnétiques.
INDEX çç Sommaire
Justification de la tenue structurelle tenant compte de la complexité de l’équipement et de la
présence des charges électromagnétiques dynamiques :
L’enveloppe du VV a une structure du type caisson avec des parois renforcées par des
raidisseurs et des boîtiers de support flexible. Une analyse linéaire élastique des contraintes de telles
structures, caractérisées par la présence de beaucoup de discontinuités, conduit à de très hautes
contraintes locales et l’interprétation correcte des résultats est difficile pour les cas de charges
primaires. Par contre, des analyses nonlinéaires telles que les analyses charges limites, permettent
de déterminer les marges de sécurité présentes pour les charges primaires.
Le poids, la pression d’essai hydraulique et les charges électromagnétiques induisent les
contraintes les plus hautes dans la structure. Les charges électromagnétiques sont extrapolées des
charges observées lors des expérimentations dans d’autres réacteurs de fusion tels que JET au
Royaume-Uni.
Les charges électromagnétiques sont dynamiques, peuvent agir dans différentes directions et
sont généralement nonaxisymmétriques. En combinaison avec la grandeur et la complexité de la
géométrie du VV, requérant des analyses nonlinéaires, l’établissement du modèle de calcul est un
vrai challenge.
L’évaluation de la variation totale des intensités de contraintes primaires plus secondaires ainsi
que l’analyse de fatigue demandent une sélection judicieuse des combinaisons des cas de charges.
Les neuf supports en dessous des portes inférieures doivent être conçus pour le poids total du
VV y compris les composants internes (± 9000 tonnes en conditions de fonctionnement) plus les
charges électromagnétiques. Des goussets bien conçus doivent contribuer au transfert des charges
entre le VV et les portes inférieures.
Le cahier des charges prises en compte dans l’analyse des contraintes, doit être consistant avec le
contenu du rapport de sûreté.
Evaluation du code RCC-MR, tel qu’appliqué pour le VV, pour sa conformité avec les EES et
ERP :
Le code nucléaire français RCC-MR, [5], a été choisi par ITER Organization comme code de
construction pour le VV. Le fabricant doit démontrer que, en appliquant le RCC-MR pour le VV,
les EES et ERP sont satisfaites. L’appendice A18 du RCC-MR inclut des lignes directrices pour
réaliser cette démonstration.
CONCLUSION
Il est important que tous les participants au projet possèdent une bonne connaissance de la
réglementation en vigueur et qu’ils se rendent compte de l’impact de leurs décisions sur la
conformité finale du Vacuum Vessel aux exigences essentielles de sécurité et aux exigences de
radioprotection applicables.
RECONNAISSANCES
Les auteurs expriment leur reconnaissance à ITER Organization d’avoir placé leur confiance à
AIB-Vinçotte International pour le mandat de l’Organisme Notifié et Agréé du Vacuum Vessel.
Les auteurs expriment en particulier leur reconnaissance à K. Ioki, V. Barabash, B. Giraud, C.
Jun et C. Bachman de ITER Organization pour le climat positif dans lequel l’évaluation de
conformité a lieu.
Nous remercions ITER Organization pour avoir donné son accord pour l’utilisation des figures
incluses dans ce texte.
REFERENCES
[1] Arrêté du 12 décembre 2005 relatif aux équipements sous pression nucléaires.
[2] Décret n° 99-1046 du 13 décembre 1999 relatif aux équipements sous pression.
[4] Guide de l’ASN n°8 du 31 mars 2009 pour l’évaluation de la conformité des équipements sous
pression nucléaires.
[5] RCC-MR 2007, Règles de conception et de construction des matériels mécaniques des
installations nucléaires applicables aux structures à haute température.
INDEX çç Sommaire
EVALUATION DE LA CONFORMITE DE L’ENCEINTE A VIDE D’ITER AUX
EXIGENCES DE L’ARRETE DU 12 DECEMBRE 2005 RELATIF AUX ESPN.
RESUME ABSTRACT
L’enceinte à vide d’ITER (VV: Vacuum Vessel) est The ITER Vacuum Vessel (VV) is a nuclear pressure
un équipement sous pression nucléaire selon l’arrêté equipment according to the French Order of December
français du 12 décembre 2005. Ceci implique que le 12th 2005. Therefore the VV manufacturer must
fabricant du VV doit démontrer que ce dernier répond aux demonstrate that the applicable essential safety
exigences essentielles de sécurité et de radioprotection requirements and radioprotection requirements are
applicables. Un Organisme Notifié et Agréé par l’Autorité satisfied. An Agreed Notified Body by the French Nuclear
de Sûreté Nucléaire doit évaluer la conformité du VV à ces Safety Authorities must evaluate the conformity of the
çç
exigences. Cette évaluation concerne toutes les phases VV with these requirements. This evaluation concerns all
INDEX Sommaire
du projet: conception, fabrication des secteurs et des project phases : design, manufacturing of the sectors and
portes dans les ateliers, assemblage et essais sur site. the ports in the workshops and the testing and assembling
on site.
La construction du VV est un projet unique et plusieurs
points requièrent une attention particulière dès la phase The building of the VV is a unique project and several
de conception pour assurer que l’enceinte à vide répondra aspects require a particular attention from the start in
finalement aux exigences de la législation. order to assure that the VV will finally comply with the
legal requirements.
Il est important que tous les participants au projet
aient une bonne connaissance de la législation applicable It’s important that all participants to the project
et se rendent compte de l’impact de leurs décisions sur have a good understanding of the applicable legislation
la conformité finale de l’enceinte à vide aux exigences and are aware of the impact of their decisions on the
essentielles de sécurité (EES) et aux exigences de final conformity of the VV with the essential safety
radioprotection (ERP). requirements and radioprotection requirements.
L’enceinte à vide d’ITER (VV: Vacuum Vessel) est un • Définition précise des responsabilités des différents
équipement sous pression nucléaire à plusieurs compartiments participants au projet. Clear definition of the responsibilities of the
dont la fonction la plus importante est le confinement du plasma different participants to the project.
Il est important que toutes les parties soient conscientes que ITER
dans une étanchéité spécifiée. Les composants principaux du VV
Organization est le fabricant du VV selon la terminologie de la législation et
sont les neuf secteurs, formant ensemble le réservoir principal est en tant que tel le responsable pour que les exigences de la réglementation
(tore), les portes et les supports. L’enceinte est une structure à soient respectées.
doubles parois du type caisson. Entre les deux parois se trouvent • Figer une conception permettant l’exécution de toutes
le fluide de refroidissement et les écrans de protection. Les les expérimentations de fusion envisagées. Fixation of a
raidisseurs poloïdaux et toroïdaux renforcent la structure de design allowing to perform all planned fusion experiments.
l’enceinte. C’est un grand défi de figer une conception du VV, bien avant que le premier
plasma se produise, qui réponde à l’objectif de fournir une démonstration
Les conditions de pression et de température de l’eau de intégrale de la faisabilité scientifique et technologique de la production
d’énergie par fusion.
refroidissement dans les différentes parties du VV, conduisent
à le classifier comme un ensemble de plusieurs équipements • Analyse de risque tenant compte des incertitudes liées
sous pression nucléaires catégorie IV niveaux N2 et N3 selon la aux expérimentations de fusion. Hazard analysis taking into
réglementation. Dès lors, les exigences essentielles de sécurité account the uncertainties about the fusion experiments.
Comme d’autres réacteurs de la grandeur d’ITER Tokamak n’existent pas, des
et les exigences de radioprotection de la réglementation doivent incertitudes sur la magnitude de ces charges électromagnétiques doivent
être satisfaites. Un Organisme Notifié et Agréé doit vérifier la être considérées dans le cahier des charges.
conformité du VV avec ces exigences.
• Exigences pour l’inspection en service. Requirements for in-
service inspection.
La géométrie particulière du VV, la présence des structures autours du VV
CARACTÉRISTIQUES PRINCIPALES DU VV dans le Tokamak et l’activation des matériaux requièrent le développement
des techniques spéciales pour l’inspection en service.
VV MAIN CHARACTERISTICS
• Classification des éléments du VV. Classification of the VV parts.
Dimensions : Tous les éléments doivent être classifiés selon leur fonction en éléments
• Tore: diamètre externe 19.4 m (principaux) résistant à la pression, éléments contribuant à la résistance à
• Tore: diamètre interne 6.5 m la pression, et d’autres éléments. En fonction de cette classification des
• Tore: hauteur 11.3 m éléments, les réglementations et le Code de Construction RCC-MR imposent
• Epaisseur de paroi 60 mm des exigences distinctes aux propriétés des matières, au type du certificat
des matières, à la catégorisation des contraintes, etc.
Température d’entrée de l’eau (secteurs) :
• Fonctionnement normal 100ºC • Spécification des matières. Specification of the material properties.
• Chauffage périodique de l’enceinte 200ºC pour VV Des Evaluations Particulières de Matériaux doivent être établies pour
démontrer que les matériaux répondent aux EES et ERP applicables.
Pression d’entrée de l’eau (secteurs) :
• Fonctionnement normal 1.1 MPa abs • Justification de la tenue structurelle tenant compte
• Chauffage périodique de l’enceinte 2.4 MPa abs
de la complexité de l’équipement et de la présence des
Pression dans la chambre plasma en situation de 0 MPa abs (vacuum) charges électromagnétiques dynamiques. Justification of the
fonctionnement structural strength taking into account the complexity of the equipment and
Volume d’eau de refroidissement dans les secteurs 200 m³ (9 secteurs) the presence of dynamic electromagnetic loadings.
Fluide de refroidissement Eau Les charges électromagnétiques sont dynamiques, peuvent agir dans
différentes directions et sont généralement nonaxisymmétriques. En
Masse totale du réservoir assemblé ± 5000 ton combinaison avec la grandeur et la complexité de la géométrie du VV,
Masse totale en situation de fonctionnement ± 9000 ton requérant des analyses nonlinéaires, l’établissement du modèle de calcul est
un vrai challenge.
Matière 316L(N)-IG
• Respect des tolérances de fabrication sévères. Respect of
severe fabrication tolerances.
Les tolérances de fabrication sont petites, en particulier si on les compare
ASPECTS RÉGLEMENTAIRES aux dimensions globales du VV. Exemple: les tolérances de surface de la paroi
côté plasma d’un secteur 40˚ vis-à-vis de la géométrie de référence sont ±
REGULATORY ASPECTS 10 mm après fabrication en usine.
Tenant compte des valeurs de pression, volume et radioactivité • Evaluation du code RCC-MR, tel qu’appliqué pour le VV,
pour sa conformité avec les EES et ERP. Assessment of the
du fluide de refroidissement, le VV est classifié comme un ensemble
RCC-MR Code, as applied for the VV, for its compliance with the ESR and RPR.
des équipements sous pression nucléaires catégorie IV de niveaux Le fabricant doit démontrer que, en appliquant le RCC-MR pour le VV, les EES
N2 et N3. Un Organisme Notifié et Agréé par l’Autorité de Sûreté et ERP sont satisfaites.
Nucléaire française doit évaluer la conformité du VV avec ces • Evaluation des procédures de fabrication et d’essais
exigences. ITER Organization a confié le rôle de l’Organisme Notifié alternatives. Assessment of alternative manufacturing and testing
et Agréé à AIB-Vinçotte International sa. Le scope de l’évaluation procedures.
comprend l’approbation de la conception, les inspections chez les Le fabricant doit démontrer que ces alternatives fournissent au moins le
même niveau de qualité et de sécurité que les procédures habituelles afin de
constructeurs dans les usines (Europe, Corée du Sud, Fédération pouvoir conclure que les EES et les ERP sont satisfaites.
Russe) et les inspections pendant l’assemblage et les essais sur le
site de Cadarache en France. All figures are courtesy of the ITER Organization.
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page139
RÉSUMÉ / ABSTRACT
109P
ÉTUDES DE CALCUL POUR LA VALIDATION DES ASSEMBLAGES À
BRIDES DES RÉSEAUX SEC ET SRU DE L’EPR FLAMANVILLE
STUDIES FOR THE VALIDATION OF THE DESIGN OF FLANGES USED FOR
THE EPR FLAMANVILLE SEC AND SRU PIPES NETWORK
B. Pitrou : Président du TC267 (EN13480), Président du Comité de direction CODETI,
Président des Commissions miroir française:
UNM 710 - Norme EN 13480 et UNM 70 Norme EN 1591 - [email protected]
C. Loth : Directeur, société LJC 4-6, rue de Bièvre - BP46 - F-92163 Antony, Membre Commission
EN TC74 WG8 –WG10 – UNM 70 (Brides et Joints), http://www.ljc.fr –[email protected]
M. Gare : PONTICELLI 5 place des Alpes 75013 PARIS- [email protected]
S. Durand : Ingénieur Calcul et gérant société MECA 5
place du commerce 44000 NANTES - http://www.calcul-meca.fr - [email protected]
RÉSUMÉ
Ce poster présente les études réalisées pour la validation des assemblages à brides DN700 et DN300 des
réseaux SEC et SRU de l’EPR FLAMANVILLE.
Dès le départ, l’approche des calculs a associé conjointement résistance et étanchéité en tenant compte de
la dispersion liée au moyen de serrage. Les choix technologiques retenus après une première étude des
assemblages suivant la norme EN1591 ont positionné ces assemblages dans un domaine de validation
situé hors des codes de construction. En effet, le type de joint le plus approprié au fonctionnement de ceux-
ci est de type Full Face non couvert par la méthode de calcul EN 1591. De plus la présence d’un revêtement
en élastomère des brides modifiait le contact entre le joint et la bride. Le comportement mécanique du
matériau du revêtement pouvait s’avérer non linéaire et fortement dépendant de la température.
Le calcul des assemblages a été réalisé avec le logiciel ADINA par la méthode des éléments finis en statique
non linéaire. Le modèle de calcul intégrait les spécificités des assemblages (contact, matériaux élastomères,
non linéarités, dépendance en température). Les niveaux de contraintes dans les brides, les boulons et le
revêtement pour les situations les plus critiques de type séisme ont nécessité l’utilisation de critères de
justification adaptés au modèle de calcul. Les critères utilisés ont été justifiés mécaniquement pour permettre
de juger de l’acceptabilité du dimensionnement. Des essais sous pression ont été réalisés et ont confirmé
les résultats en terme de tenue mécanique et étanchéité.
ABSTRACT
This poster shows the studies performed for the validation of the design of DN700 and DN300 flanges
used for the EPR FLAMANVILLE SEC and SRU pipes network.
From the beginning of the studies, the calculations were based on both strength and fluid sealing
approach, always taking into account the tightening deviation. After a first preliminary study with
EN1591 standard, technologic choices have been proposed. These choices make the bolted joint out of
standards used for the design of flange assembly. In fact, the most suitable gasket for the assembly is a
Full Face type no covered by design standard like EN1591. Moreover, there is an elastomeric coating
on the flanges. This coating modifies the contact between the flange and the gasket. This coating also
shows a non linear behavior very sensitive to temperature.
Bolted Joints calculations have been performed with ADINA software using Finite Element Analysis
with static non linear assumptions. Calculation model was representative of flange specifications
(contact, elastomeric material, non linearity, temperature). Due to stress levels in the flanges, the bolts
and the coating for seismic conditions, it was necessary to use specific criterion adapted to the
calculation model. Used criterion were justified based on a mechanical analysis, in order to show the
compliance of the design. Pressure tests have been done. These tests have confirmed the results of the
studies with respect of mechanical strength and fluid sealing.
AFIAP - ESOPE 2010 Programme 139
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page140
RÉSUMÉ / ABSTRACT
110P
LE SOUDAGE PAR FRICTION MALAXAGE ET LES ÉQUIPEMENTS SOUS
PRESSION
RÉSUMÉ
Les résultats présentés correspondent à une étude menée par l’Institut de Soudure au profit du CETIM.
L’objectif est de démontrer que le procédé de soudage par friction malaxage (FSW) est utilisable sur
des enveloppes soumises à pression avec un coefficient de joint au moins équivalent à celui des procédés
à l’arc. La soudabilité opératoire de la nuance d’alliage d’aluminium EN AW 5754-H111 en épaisseur
5 et 8mm, utilisée dans les appareils à pression, a été étudiée. Après analyse des performances des
cordons soudés, il apparaît que ces nuances sont soudables par le procédé de soudage par Friction
Malaxage sans perte de caractéristiques mécaniques. La réalisation de soudure d’enveloppe sous
pression peut donc être envisagée par ce procédé. De plus, l’étude a montré que selon les applications,
un robot poly-articulé pourrait être utilisé, rendant cette technologie accessible à des coûts
d’investissement réduit.
POSTERS
ABSTRACT
The results presented correspond to a study realized by the Institute de Soudure for the CETIM. The
objective is to demonstrate that the Friction Stir Welding (FSW) can be used on pressurized envelopes
thanks to a joint coefficient at least equal to the one obtained by arc welding processes. The weldability
of aluminium alloy used in pressure vessels, EN AW 5754-H111 with a thickness of 5 and 8 mm, has
been studied. After analysing the FSW joint performance, it appears that this alloy can be welded by
FSW without loss of mechanical properties. Therefore, pressure vessels can be welded by FSW.
Moreover, the study showed that depending on the application, a serial kinematics robot could be used,
making this technology available by reducing the investment costs.
140
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page141
RÉSUMÉ / ABSTRACT
113P
HAUTE SURVEILLANCE DE LA TEMPERATURE DES APPAREILS A
PRESSION PAR TECHNIQUE D’EMISSION ACOUSTIQUE
F. Dahmene [email protected]
A. Laksimi [email protected]
Laboratoire Roberval, UTC-CNRS,UMR 6253, BP 20529, 60205 Compiègne, France
S. Hariri [email protected]
Laboratoire TPC&IM, Ecole des Mines de Douai, B.P. 838, 59508 Douai Cedex, France
C. Hervé [email protected] ;
L. Jaubert , [email protected] ;
M. Cherfaoui
Département Contrôles et essais industriels, CETIM, BP80067, 60304 Senlis, France.
RÉSUMÉ
L’objectif de cette étude est le développement du contrôle non destructif par émission acoustique (E.A.)
des équipements sous pression (ESP) fonctionnant en température. L’exploitation de l’émission acous-
tique dans ce travail a permis de mettre en évidence le comportement mécanique en terme d’endom-
magement et de propagation de fissure, en température, pour les différents matériaux étudiés (acier
304L et P265GH). L’utilisation conjointe de la thermographie infrarouge, de la corrélation d’image
ainsi que l’émission acoustique contribuent à l’identification des signatures acoustiques des différents
phénomènes d’endommagement en température dans les éprouvettes CT.
Les mesures expérimentales des vitesses de propagation des ondes acoustique ainsi que leurs atténua-
tions en fonction de la température, ont permis de valider les vitesses calculées analytiquement. Ces
dernières sont très utiles pour la localisation d’un défaut dans un ESP fonctionnant à haute température,
dans le cas où ils sont inaccessibles expérimentalement.
Le calcul par éléments finis de l’intégrale de Rice J, a permis d’évaluer la sur-contrainte qu’il faudra
appliquer lors du contrôle des ESP à température ambiante afin de compenser les effets thermiques de
l’essai en température.
Ce travail a montré qu’il est possible d’effectuer des contrôles à température ambiante sur les équipe-
ments sous pression travaillant à haute température en ajoutant une légère surpression. Ceci est d’autant
plus intéressant que le contrôle par émission acoustique de ce type de structure peut s’effectuer en évi-
tant l’utilisation de capteurs à haute température qui rendrait cette opération très coûteuse et moins sen-
sible.
MOTS CLES
Contrôle non destructif ; Equipement sous pression ; Emission acoustique
Endommagement ; Nocivités des défauts ; Intégrale J – Calcul numérique
AFIAP - ESOPE 2010 Programme 141
INDEX çç Sommaire
High temperature pressure vessels monitoring by
acoustic emission technique
Laboratoire Roberval Fethi Dahmene12 , Abdelouahed Laksimi1 , Said Hariri2 , Catherine Hervé3, Lionel Jaubert3, Mohamed Cherfaoui3
Unité de recherche en mécanique
1 : LaboratoireRoberval, UTC-CNRS,UMR 6253, BP 20529, 60205 Compiègne, France
2 : Laboratoire
TPC&IM, Ecole des Mines de Douai, B.P. 838, 59508 Douai Cedex, France
3: Département Contrôles et essais industriels, CETIM, BP80067, 60304 Senlis, France.
OBJECTIVES
ü Development of an experimental protocol to study high temperature wave propagation and crack propagation.
ü Effect of waveguide and temperature on acoustic wave propagation (velocity and attenuation).
ü Identification of damage mechanisms during crack propagation at high temperature (250°C for P265GH, 450°C for 304L) and it’s acoustic response.
ü Evaluation of the harmfulness level of a crack in a pressure equipment during an overload to substitute temperature effect.
ü Material selection (P265GH, 304L). ü Waveguide effect on elastic wave transmission at room temperature: 200kHz on plate
Amplitude [mV]
Tensile specimens (Compact tensile pre-cracked by fatigue).
Time [µs]
200kHz on waveguide (800mm)
Amplitude [mV]
Waveguide deforms the wave form but it doesn’t affect its frequency.
Time [µs]
Amplitude [dB]
Waveguide length doesn’t affect signal magnitude. Time [s]
ü Waveguide design (shape, length) ü High temperature effect on wave propagation velocity and attenuation:
ü Furnace design and realization for high temperature wave propagation tests.
Velocity decrease by 14.77% (13.56%) at 600°C for signal frequency 200kHz in a 304L (P265GH) material.
Temperature doesn’t affect signal magnitude in the case of 304L material. An attenuation of about 8 dB is
observed at 600°C for P265GH material.
ü P265GH (tensile tests were performed from 20°C to 250°C) ü Evaluation of the harmfulness level of an inner elliptical crack in a cylindrical pressure vessel during an overload
AE from plastic deformation Fast fracture or a warm-up.
Infrared thermography
observation
Figure1: Acoustic emission activity at 150°C (P265GH).
Crack mesh Meshing of half cylinder with
The presence of serrations in load versus load line displacement plots at 150°C and 200°C, indicates that Dynamic elliptical crack
Strain Aging «DSA» occurs in this temperature regime. Fracture mechanism consists of alternate crack tip plastic flow and
fast fracture.
By increasing the temperature to 350 ° C, the stress intensity factor also increases for an applied pressure of
10MPa (P265GH).
ü 304L (tensile tests were performed from 20°C to 450°C)
AE from plastic deformation
Fast fracture
DP
Damage beginning
Crack tip plasticity
Contacts :
Fethi Dahmene: [email protected]
Abdelouahed Laksimi: [email protected]
Saïd Hariri: [email protected]
Catherine Hervé: [email protected]
Lionel Jaubert: [email protected]
INDEX çç Sommaire
1
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page142
RÉSUMÉ / ABSTRACT
115P
ÉVALUATION A MOINDRE COUT DE LA FIABILITE DES STRUCTURES
SOLLICITEES EN FATIGUE
A. Notin, H. Jaffal
CETIM – Pôle ICS
52 avenue Félix Louat - B.P.80067,
F60304 Senlis Cedex
[email protected]
N. Gayton
Institut Français de Mécanique Avancée
Laboratoire de Mécanique et Ingénierie
B.P.10448, F-63000 Clermont-Ferrand
[email protected]
RÉSUMÉ
Ce travail s'insère dans le contexte général de l'estimation de la fiabilité des structures sollicitées en fa-
tigue. Dans le cas d'applications industrielles, chaque évaluation est potentiellement coûteuse en temps
de calcul et en espace de stockage. De ce fait, seul un nombre fini de calcul peut être réalisé ce qui sup-
pose de travailler sur l'algorithme de fiabilité mais aussi d'accélérer les calculs mécaniques. A partir de
ce constat, deux méthodes sont présentées, la méthode MRCP (Méthode de Ré-échantillonnage du
Chaos Polynomial) et l'approche SLDLT (décomposition LDLT Stochastique). La méthode MRCP se
concentre sur l’algorithme de fiabilité alors que l'approche SLDLT vise à accélérer les calculs méca-
niques.
ABSTRACT
This work takes place in the context of the estimation of the reliability of structures under fatigue loa-
ding. In the case of industrial applications, each model evaluation may be time and storage consuming.
This way, only a few number of evaluations can be performe which implies to work on the reliability
algorithm as well as the speeding up of mechanical computations. To achievve this goal, two method
are developed, the RPCM (Resampling Polynomial Chaos Method) approach and the SLDLT (Stochas-
tic LDLT) decomposition. The RPCM method targets the reliability algorithm as the SLDLT approach
is dedicated to the speeding up of mechanical computations.
142
INDEX çç Sommaire
��������� ������ ����������������������������������������������������� ��������
5�%�-��<�5�%�-�
� #�� $���%
����(������%=��%8(��&��%����,��������2���(���&���=�%��-������&�����*��1������&�%�%�(�,��(�%�%����,����%����*���2����>��%����,�%�&=�77��,�����%���&�%�(�����%"�,'�/���
������������%��7����������-����,�?���%�������-7%�&��,��,����������%7�,��&��%��,@�2���>��,��*���"�%���������-1(��*����&��,��,���7����A�(��(����%��,��/���%�77�%��
INDEX çç Sommaire
&�� �(�������(� %�(� �=��2�(��'-�� &�� *��1������ -��%� ��%%�� &=�,,���(�(� ��%� ,��,��%� -�,���/��%��0� 7�(��(� &�� ,�� ,��%���"� &���� -��'�&�%� %���� 7(�%�����%"� ��� -��'�&��
�5� �B���'�&��&��5��,'����������2��&���'��%� ��4��-���C�����=�77(�,'����>���B&�,�-7�%�������>������,'�%��/��C�����-��'�&���5� �%��,��,���(��%�(�
�=��2�(��'-��&��*��1���������(%�/����=�77(�,'����>�����%��D��,,���(�(���%�,��,��%�-�,���/��%�
� � �����%
�'�%�E�(@���@�%�7��,������'��,��������*��'���%��-�������*��'��(����1����4��*�%�(�,��(�%���&�(�*���2������&��2������'��,�%���*���&�%�(�����77��,�����%"���,'�-�&���
�����������-�4�1����-����&�%��(�2��,��%�-��2���'�%�E�4"����4���*�E���-1�(��*�����������%�,���1��7�(*�(-��E'�,'��-7���%����E�(@�����'��(����1����4���2�(��'-�
�%�E�����%��'��%7��&��2��7��*�-�,'���,���,�-7�������%������,'�������'�%�2���"��E��-��'�&��(��&�����7�&"��'��5 ���B5�%�-7���2� ��4��-�����'��%����'�&C�
�77(��,'� ��&� �'�� ��>��� B���,'�%��,� �>��C� &�,�-7�%��������'�� 5 ��� -��'�&� ��(2��%� �'�� (����1����4� ��2�(��'-� �%� �'�� ��>��� �77(��,'� �%� &�&�,���&� ��� �'��
%7��&��2��7��*�-�,'���,���,�-7�������%�
���'�&���5� ���'�&����>��
� �� ��&�% � �� ��&�%
�'�(�������%
� ����-�%�(������-1(��&��,��,��%F
� 6����%�(����-���-�-��G��*�(-������&�%7���1���
�� � Q � � Q
���-��(�,�����%����&�7��&�����&�%�(����%�����%�&���=��������&�7��&
���/��-����&�����
� ��?����-�(�/���B,�-7������C����*��,�����&����-1(����&����(�2�%
���&�����&�-����(2��(�&��1��&��������'�&��(����1���%�I�*��+,��-���
���,��%���
���-��'�&���5� ����82(�����2�%�����&G�((��(�B7�(�(�77�(��D�����1)�,��*�&��*��1�����C��������7'�%��&G�77(����%%�2��7��(�,��%�(��(�����-���-�&8���&��-���8(��
�&�7�����������,���(����&���G�((��(�%G�**�,����7�(���%�����(�����%�&��,��*���,�����%�(�%�����%�-���(����/������,'����&��������2���(�&���G����(������&��,��*���,��
,�--����&�,����(�&�����/�������&��-���-�&8����%��7�(�������
���-��'�&����>���(�7�%��%�(�����'47��'8%��*�(���/���,��%�%���D�%�77�%�(�/�����%�*��,�������%�&�����-��(�,���(���2����(����*�(���(�� ��%������2��2��1��%�,��/���
(�������D�(�7�(��(��G�����%�(���%���(-�%�&�����-��(�,��&��2�����������%�(�%�����%�-���(�������2����7�����������(�)�%/�=D����*�,���(�3���
� �
RÉSUMÉ / ABSTRACT
117P
SOUDAGE D’ACIERS AU CRMO POUR LA PRODUCTION D’ÉNERGIE,
LE RAFFINAGE DE PÉTRÔLE
ET LA PÉTROCHIMIE – PASSÉ, PRÉSENT, FUTUR.
J. Hilkes, V. Gross
Bôhler Schweisstechnik Deutschland GmbH
Unionstrasse 1, D-59067 Hann, Germany
P.-A. Legait
Groupe Böhler Soudage France SAS
14 Place Georges Pompidou 78180 Montigny-le-Bretonneux, France
RÉSUMÉ
Cet article donne un aperçu de l’évolution et des applications des aciers CrMo classiques jusqu’aux
aciers 12cr1mo, en passant par les nouveaux aciers CrMoV. En outre, les consommables de soudages
associés à la production d’énergie, au raffinage de pétrole et à l’industrie pétrochimique sont pris en
compte. Il sera fait référence aux normes internationales applicables et aux propriétés spécifiques et
exigences actuelles établies dans l’industrie.
ABSTRACT
This paper provides an overview of the development and applications of the classic CrMo, the new
CrMoV steels and CrMo steels all the way up to 12 Cr1Mo. Also, the corresponding welding consu-
mables for the power generation and the petrochemical industry will be discussed. Reference wil be
made to the various international standards that are applicable and the specific properties and requi-
rements as set by today’s industries.
INDEX çç Sommaire
Soudage d’aciers au CrMo pour la production d’énergie, le raffinage de pétrole
et la pétrochimie
- Passé, présent et futur -
2
Groupe Böhler Soudage France S.A.S
14, place Georges Pompidou, 78180 Montigny-le-Bretonneux, France
+33 1 30 05 49 04, [email protected]
Mots clés: GMAW, SAW, GTAW, SMAW, aciers résistant au fluage, haute température, production
d’énergie, raffineries de pétrole, pétrochimie, CrMo, CrMoV, applications, modes opératoires
de soudage, TTAS, énergie de soudage, Tp, Ti, ferrite, perlite, bainite, martensite, propriétés
mécaniques, tôles, tubes, refroidissement par paliers, ZAT.
RESUME
Cet article donne un aperçu de l'évolution et des applications des aciers CrMo classiques jusqu’aux aciers
12Cr1Mo, en passant par les nouveaux aciers CrMoV. En outre, les consommables de soudage associés à
la production d'énergie, au raffinage de pétrole et à l'industrie pétrochimique sont pris en compte. Il sera fait
référence aux normes internationales applicables et aux propriétés spécifiques et exigences actuelles
établies dans l’industrie.
INTRODUCTION
Les aciers alliés au CrMo résistant au fluage à haute température existent depuis très longtemps et ont
trouvé une utilisation avec succès pour des applications dans les secteurs de la pétrochimie, du raffinage et
dans l'industrie de la production d'énergie. Les fabrications types de ces industries sont les chaudières,
fours, échangeurs de chaleur, réacteurs, unités d’hydrocraquage, usuellement construits à sous la forme
d’appareils à pression à parois épaisses.
Dans un souci permanent d'optimisation des rendements de différentes installations dans ces industries, les
pressions et/ou les températures de service ont été augmentées. Pour répondre à ces exigences
croissantes, il était nécessaire d’envisager l’une des possibilités suivantes : la première consiste à envisager
la mise à disposition de matériaux d’épaisseurs plus importantes ; la seconde consiste à améliorer la
résistance mécanique et la résistance au choc des matériaux. Une augmentation des propriétés mécaniques
permet de réduire ou au moins de limiter l'épaisseur de paroi nécessaire, ce qui génère un avantage
économique supplémentaire dans la production, la manutention et l'installation d’équipements lourds. Une
partie d’unité d’hydrocraquage est présentée en figure 1 à titre d’exemple d'appareil à pression à parois
épaisses.
Détails techniques d'une unité
d’hydrocraquage:
INDEX çç
Soudage d’aciers au Cr Mo pour la production d’énergie, le raffinage du pétrole et la pétrochimie
- Passé, présent et futur -
Sommaire
Les aciers de base classiques CrMo sont alliés avec 0.5% Mo / 1% Cr - 0,5% Mo / 2,25% Cr - 0,5% Mo /
5% Cr - 1% Mo / 9% Cr - 1% Mo/ et 12% Cr / 1% de Mo. Une amélioration significative de ces aciers a été
permise par l’ajout d’éléments d’addition tels que V, W, Ni, Ti, Nb, B et/ou N qui ont conduit au
développement de nouvelles nuances connues à ce jour sous les désignations telles que T/P22V, T/P23,
T/P24, T/P91, T/P92 et VM 12-SHC. Un grand nombre de ces nouvelles nuances ont été utilisées avec
succès dans l'industrie, mais le développement continue.
Évidemment, le développement des consommables de soudage a du et doit toujours suivre l’évolution des
métaux de base avec l'assurance de satisfaire des exigences aussi strictes, d’autant plus que la ZAT est
aussi généralement considérée comme partie intégrante de la soudure. Des recherches approfondies ont
été conduites par la société Böhler Schweisstechnik en Allemagne pour permettre l’obtention d’une gamme
de consommables complète pour la nouvelle génération d’aciers CrMo (V), consommables pour lesquels
des résultats d’essais de fluage jusqu'à 60.000 heures ont pu être obtenus.
Les teneurs en éléments d’addition augmentant, les modes opératoires de soudage doivent être ajustés ;
En outre, les paramètres de soudage et de traitement thermique doivent être contrôlés de façon plus précise
et rigoureuse.
La résistance au fluage d'un acier CrMo est basée sur la formation de précipités stables tels que des
carbures dans un alliage de microstructure ferritique, bainitique et/ou martensitique dans l'état normalisé. En
raison d'un traitement de revenu ultérieur, une microstructure contenant des précipités est générée et reste
stable à la température de service pour laquelle l'acier a été développé. Les précipités formés vont
notamment bloquer les joints de grains et empêcher les déplacements relatifs des plans de glissement,
permettant d’obtenir les propriétés souhaitées de résistance au fluage. Ils devraient donc être de forme
appropriée, être présents en quantité adéquate et être uniformément répartis pour obtenir une structure
homogène et des propriétés homogènes. Selon la quantité d'éléments d’addition et selon le traitement
thermique, certains types de précipités seront formés dans une certaine quantité. Les paramètres essentiels
pour le traitement thermique sont la température et le temps. Les différents types de précipités qui peuvent
être observés et qui sont principalement utilisés dans la conception des aciers CrMo classiques et modernes
sont listés dans le tableau 2.
INDEX çç Sommaire
Précipités et phases possibles dans les aciers CrMo
Graphite
Epsilon = Fe2.4C
Cémentite = Fe3C
Chi = Fe2C
M2 X M6 C M23C6
M7C3 Laves
M5C2 Z-phase
Mo2C Cr3C
NbC NbN VN
Tableau 2: Précipités qui peuvent être trouvés dans les aciers CrMo résistant au fluage / 1, 2 /.
1000 1000
+ 0 heating rate
800 760°C- 10 80 - 120°C/h
800
740°C
Temperature [°C]
Temperature [°C]
2-4h
Heating rate Cooling rate 540 - 560°C
600 80 - 120°C/h 100 - 150°C/h 600 min. 1h
cooling rate
100 - 150°C/h
wall thickness
intermediate
depends on
PWHT-time
·
heating
pre-
20 - 100°C
· ·
0 0
00,51 1,52 2,53 3,54 4,555,566,577,588,599,51010,5
11 1,5 1313,5
1212,5 time
Time
Si lors d’un TTAS ultérieur, d’un traitement thermique de détensionnement intermédiaire ou en service, la
température du traitement thermique du matériau de base était trop largement et trop longtemps dépassée,
les précipités pourraient se dissoudre à nouveau, ce qui entraînerait une réduction des propriétés
mécaniques du matériau de base. C’est la raison pour laquelle la température maximale de 760°C ne doit
pas être dépassée pour le P91 (voir figure 2). Pour le T/P23 de la figure 2, un traitement thermique de
détensionnement intermédiaire ou Intermediate Stress Relieving (ISR) est indiqué pour les constructions
d’épaisseurs de matériaux différentes. Pour chaque type d’application, un TTAS optimal doit être déterminé.
Cette question est approfondie dans le chapitre soudage de ce document. (Tableau 4).
INDEX çç Sommaire
Fragilisation au revenu
Lorsque le métal de base CrMo et le métal fondu sont exposés à une température comprise entre 400°C et
500°C pendant une longue durée, il y a un risque de fragilisation au revenu. Ce type de fragilisation est
causé par les traces d’éléments comme P, Sb, Sn et As qui migrent vers les joints de grains et peuvent
réduire la ductilité du métal de base et du métal fondu. La dimension de ce phénomène dépend simplement
de la température et du temps. Afin de déterminer la sensibilité d'un matériau à la fragilisation au revenu,
une simulation est effectuée par traitement thermique, lors d’un essai dit de refroidissement par paliers ou
Step Cooling (STC) qui est effectué à des températures comprises entre 593°C et 316°C pendant 240
heures. La différence de température de transition ductile-fragile, ou le décalage de la courbe de transition
avant et après l'essai STC, est mesuré pour évaluer la sensibilité à la fragilisation au revenu du matériau. Un
déplacement maximal acceptable de la température de transition après le traitement thermique de STC peut
être alors spécifié comme exigence requise pour le matériau de base et le métal fondu. Afin de réduire le
risque de fragilisation au revenu, les teneurs en traces d’éléments préalablement cités doivent être
restreintes. Bruscato et Watanabe ont développé des formules pour estimer cette tendance à la fragilisation
au revenu /3, 4/.
La formule de Watanabe est uniquement valable pour le matériau de base et est généralement limitée à une
valeur maximale de J <160, mais les limites J <120 ou 80 sont également spécifiées aujourd'hui par
l'industrie.
La formule de Bruscato, aussi appelée le facteur X, est valable tant pour le métal fondu que pour le métal de
base. En ce qui concerne le métal fondu, les spécifications deviennent de plus en plus strictes avec
l’augmentation des épaisseurs et du désir d’obtention en outre de propriétés mécaniques satisfaisantes.
Initialement, la valeur maximale requise pour le facteur X était de X<15, mais les spécifications actuelles
demandent déjà X<10. Une exigence supplémentaire s’applique aux teneurs en Mn et Si, devant satisfaire
Mn + Si <1,1%.
Concernant spécifiquement le procédé de soudage à l’arc submergé (SAW), les traces d’éléments peuvent
provenir à la fois du fil et du flux, l'association des deux devant être testée pour s’assurer de la conformité
aux exigences. Cela signifie qu’une seule source d’approvisionnement pour le fil et le flux serait
recommandée / 5 /.
INDEX çç Sommaire
SOUDAGE ET CONSOMMABLES POUR LES ACIERS CRMO
En général, les aciers CrMo résistant au fluage sont soudés avec des métaux d’apport appropriés en vue de
l’obtention d’un joint soudé homogène ayant des propriétés mécaniques à peu près équivalentes. La
similarité des compositions permet aussi l’obtention d’un même coefficient de dilatation thermique, ce qui
empêche, ou du moins réduit les risques de fatigue thermique en service. À cet égard, la zone affectée
thermiquement (ZAT) est une zone vulnérable. En principe, tous les procédés de soudage à l'arc peuvent
être envisagés, comme par exemple les procédés SMAW, GTAW, GMAW, FCAW et SAW. Concernant les
procédés de soudage manuels, il est important de prendre des mesures suffisantes pour protéger les
soudeurs de la chaleur, sachant qu’il est d'une importance primordiale que les températures de préchauffage
et d’entre passes minimales soient respectés. Ces dernières ne devraient pas être réduites en vue
d’apporter un meilleur confort de travail au soudeur, même en ce qui concerne le pointage par soudage.
Avec les procédés de soudage sous gaz de protection, il est essentiel d'assurer une bonne protection du
bain de fusion. Lorsque la température de préchauffage est élevée, la protection gazeuse peut être
perturbée. Des buses spécifiques sont disponibles pour limiter ce problème.
Au cours des dernières décennies, Böhler Schweisstechnik Allemagne a développé une large gamme de
consommables pour le soudage des aciers CrMo avec les procédés: SMAW, GTAW, SAW, GMAW et
FCAW. Une sélection de consommables associés aux divers procédés de soudage des aciers résistant au
fluage CrMo est disponible dans le tableau 3.
Selon le type d’alliage, d’un alliage à 0,5%Cr à un alliage à 12%Cr-1%Mo, les conditions de soudage, de
préchauffage (Tp) et de température entre passes (Ti), ainsi que les cycles de traitements thermiques après
soudage (TTAS) et notamment ISR et STC, peuvent avoir une influence notable. Un aperçu des exemples
types à ce sujet est donné au tableau 4. La figure 2 ci-dessus donne des exemples de traitements
thermiques complexes. Le traitement thermique nécessaire dépend aussi de l'épaisseur de la construction et
doit être déterminé par le fabricant dans le cadre du développement des procédures de soudage.
Le facteur principal est d'avoir un refroidissement et un réchauffage contrôlé, lent et régulier, afin d’éviter la
formation de contraintes supplémentaires dans le joint soudé. Pour les fortes épaisseurs, ceci implique une
distribution de chauffage contrôlée de plusieurs côtés, afin d’optimiser la répartition de la chaleur dans le
matériau. Ces précautions doivent être prises pour préserver le métal de base, le métal fondu et la zone
affectée thermiquement (ZAT). Les évolutions récentes sur les P22V, P23, P24, P92 et VM12-SHC ont
conduit à l’élaboration de procédures de production et de soudage plus précises pour garder le contrôle
nécessaire sur la qualité de la construction soudée. Bien que ces matériaux ne soient pas aussi indulgents
que les aciers plus conventionnels au CrMo, la soudabilité reste excellente dès lors que les procédures
adaptées sont correctement respectées. En fonction de l'application, il peut y avoir des exigences
concernant le refroidissement par paliers et le facteur X de Bruscato. Dans le cas de fortes épaisseurs en
P22V il pourrait être nécessaire d'appliquer un traitement thermique ISR en vue de réduire le niveau de
contraintes résiduelles de l’ensemble de la pièce avant d’envisager les traitements thermiques finaux. Avec
l'expérience que Böhler Schweisstechnik Allemagne a accumulée au fil des dernières décennies,
l’assistance qui peut être fournie aux clients est devenue un maillon vital dans la chaîne des fournitures de
contrats actuels.
Comme mentionné précédemment, les épaisseurs des constructions soudées pour la production d’énergie,
le raffinage de pétrole et la pétrochimique ne cessent d'augmenter ; des matériaux de résistances
mécaniques plus élevées, offrant une composition chimique plus stricte et plus élaborée, sont utilisés pour
permettre la réalisation des fabrications. Cela signifie que les consommables de soudage doivent être
adaptés pour suivre cette évolution. La figure 3 présente un exemple d’assemblage de forte épaisseur sur
une jonction de tube de vapeur en P91 dans une centrale électrique.
INDEX çç Sommaire
Figure 3: préparation et soudure finale dans une jonction de tube de
vapeur en P91, soudé en procédé SMAW grâce à Thermanit Chromo 9 V
Comme indiqué, les cycles thermiques et notamment le préchauffage et la température entre passes doivent
être strictement contrôlés pour réussir ces types de joints soudés. Les fenêtres de paramètres telles que les
températures de préchauffage et d’entre passes données dans le tableau 4 doivent être respectées tout au
long de l’exécution du joint soudé. Le procédé SMAW est très adapté pour cette application en raison de sa
flexibilité et de son faible niveau d'investissement pour l’équipement nécessaire. Afin d'accroître la
productivité, il s’agit d’accroître le taux de dépôt par unité de temps ; certains fils fourrés destinés au
soudage des aciers CrMo par le procédé FCAW sont justement en cours de développement. Un certain
nombre de fils fourrés sont d’ores et déjà disponibles, comme présenté au tableau 3, toutefois la gamme
sera étendue pour répondre à la demande industrielle.
La gamme de produits d’apport destinés au soudage par le procédé SAW couvre l'ensemble des aciers
CrMo disponibles aujourd'hui. Le procédé GTAW est principalement utilisé pour le soudage des passes de
racine ou parfois pour le soudage automatisé. La gamme de métaux d’apport destinés au soudage par le
procédé GMAW est disponible, mais impopulaire.
La figure 4 illustre un autre exemple de réacteur construit en acier 2,25%Cr-1%Mo. L'appareil à pression de
forte épaisseur a été fabriqué par Godrej (Inde). Lorsque l’on travaille avec des pièces de forte épaisseur,
avec des aciers modernes alliés au CrMo répondant à des spécifications très strictes, il est absolument
nécessaire d’acquérir une expérience suffisante pour être en mesure de satisfaire les exigences des
entreprises d'ingénierie ainsi que des donneurs d’ordre tels que les sociétés de l’industrie pétrolière ou de la
production d’énergie.
- Produits d’apport :
SAW: Union S1CrMo2/UV 420TTR
SMAW: Phoenix SH Chromo 2 KS
Figure 4: Réacteur de forte épaisseur en 2.25%Cr-
1%Mo fabriqué par GODREJ, INDE
INDEX çç Sommaire
Contrôle des paramètres et conseils pratiques
Le métal fondu CrMo(V) présente habituellement une microstructure de type bainito-martensitique, qui se
montre très sensible à tout type de cycle thermique imposé notamment par le soudage ou les traitements
thermiques. En outre, la résistance mécanique élevée obtenue à l’état brut de soudage nécessite des
précautions d’exécution en termes de sensibilité à l'hydrogène et au traitement thermique ISR afin d'éviter la
formation de fissures dues à hydrogène et/ou aux contraintes résiduelles des soudures, par exemple dans le
cas de piquages de fortes épaisseurs.
Pour illustrer certains phénomènes, des observations typiques de soudures réalisées sur des aciers
résistant au fluage CrMo (V) sont présentées dans le paragraphe suivant.
La figure 5 montre des précipités de ferrite dans un alliage P11 formés suite à un maintien à une
température excessive lors du TTAS. La micrographie de la figure 6 représente le dommage causé par
l'hydrogène dû à un traitement thermique incorrectement appliqué, laissant trop d'hydrogène résiduel dans le
métal fondu. La figure 7 montre l’influence de l'épaisseur des passes en procédé SMAW, conduisant à un
déplacement de la courbe de transition fragile-ductile, en raison d'une structure à grains plus grossiers.
250
---PWHT
200 ---PWHT + step cooling
C V N [J ]
150
flat beads
100
thick beads
55 J
50
0
-80 -70 -60 -50 -40 -30 -20 -10 0
Figure 5: Précipités de ferrite sur Figure 6: Aspect de surface d’une fissure Figure 7: Influence de l’épaisseur des
soudures SAW de P11 due à l’hydrogène sur soudures SAW de passes sur les valeurs de résilience
P22V
Les paramètres de fabrication, dont les paramètres de soudage, la fiabilité de l'équipement de soudage
utilisé et le niveau de compétence des soudeurs, ont vu leur importance croître avec les nouvelles exigences
d’augmentation de la résistance des matériaux mis en œuvre. La fenêtre des paramètres opératoires se
restreint. C'est pourquoi des mécanismes de contrôle et des procédures appropriées doivent être mis en
place pour assurer la bonne application des paramètres requis. En particulier, le contrôle des actions listées
ci-après ne doit pas être négligé afin de permettre l’obtention d’une qualité optimale des soudures:
- Respect des épaisseurs et de la répartition des passes spécifiées Low alloy steel
- Vérification de la température lors des traitements thermiques. C-
steel
La majorité des problèmes rencontrés lors du soudage des aciers
- optimum
CrMo peuvent être liés au non-respect des points mentionnés ci- parameters +
dessus. Par conséquent, des mécanismes de contrôle appropriés Figure 8: Assurance qualité -
doivent être développés pour s’assurer de la production de contrôle des paramètres requis
soudures saines. L'assurance qualité (AQ) devient un facteur
majeur pour le soudage des aciers CrMo. L’AQ doit être
considérée comme une variable essentielle, comme l'illustre la
figure 8.
En conclusion, nous pouvons affirmer que les aciers alliés au CrMo résistant au fluage sont globalement
utilisés avec succès pour la production d’énergie, le raffinage de pétrole et la pétrochimie. La tendance à
l’augmentation des températures de service requiert de nouveaux matériaux, aussi bien pour les matériaux
de base que pour les consommables de soudage.
INDEX çç Sommaire
Des exemples types d’application des aciers CrMo dans les unités de production d’énergie et les unités
pétrochimiques sont donnés dans le tableau 5.
Cet article a été élaboré afin de fournir une vue d'ensemble sur les matériaux employés et les normes
correspondantes, mais aussi pour sensibiliser les lecteurs à propos des implications et des conséquences
en matière de soudage, de traitements thermiques et de fabrication en général. Les projets peuvent être
correctement menés à leurs termes, dès lors que des procédures correctes sont élaborées et suivies tout au
long de la production.
REFERENCES
/1/ Cole, D. and Bhadeshia, H.K.D.H., Design of Creep-Resistant Steel Welds, Research work,
University of Cambridge, Department of Materials Science and Metallurgy, 1998
/2/ Bhadeshia, H.K.D.H, Design of Creep-Resistant Steels, Proceedings of Ultra-Steel 2000,
National Research Institute for Metals, Tsukuba, Japan 2000, pp.89-108
/3/ Bruscato, R., Temper Embrittlement and Creep Embrittlement of 2.25%Cr - 1%Mo shielded
metal arc weld deposits, Welding Journal 49 (4), 1973, pp. 148-156
/4/ Watanabe, J., et. al., Temper Embrittlement of 2.25%Cr - 1%Mo Pressure Vessel Steel,
ASME 29th Petroleum Mechanical Engineering Conference, Dallas, USA, 1974
/5/ Gross, V., Heuser, H., and Jochum, C., Schweisstechnische Herausforderung bei der Verarbeitung
von CrMo(V)-Stählen für Hydrocracker, Publication of Böhler Thyssen Schweisstechnik,
Germany, 2007
/6/ Handel, Geert van den, Chroom-Molybdeen staalsoorten, Lastechniek, Nederlands Instituut
voor Lastechniek (NIL), Nr. 5, May 2008, pp.10-14
INDEX çç Sommaire
Classifications normalisées internationales composition chimique typique (% massique)
Température
Nuances CrMo
DIN/VdTÜV EN C Si Mn Cr Mo Ni V W Nb autre de service
ASTM
& ASME % % % % % % % % % % °C
0.5Mo T/P 1 16 Mo 3 8MoB 5-4 0,16 0,30 0,82 < 0,30 0,32 < 030 < 460
1.25Cr-0.5Mo T/P 11 10 CrMo 5-5 10 CrMo 5-5 0,10 0,32 0,68 1,25 0,50 < 545
1,00Cr-0.5Mo T/P 12 13 CrMo 4-5 13 CrMo 4-5 0,13 0,70 0,60 1,00 0,50 < 545
1.25Cr-1MoV 15 CrMoV 5-10 0,15 0,30 0,75 1,25 1,05 0,26 < 545
-/- 15 NiCuMoNb 5 15 NiCuNb 5 0,15 0,35 0,95 0,45 1,12 0,22 Cu: 0,62 < 545
(WB 36)
2.25Cr-1Mo T/P 22 10 CrMo 9-10 10 CrMo 9-10 0,10 0,36 0,69 2,20 1,02 < 545
2.25Cr-1MoV T/P 22V 0,12 0,08 0,50 2,25 1,00 0,30 < 545
2.25Cr-MoVW T/P 23 HCM 2S 7CrWVMoNb 9-6 0,08 0,34 0,42 2,32 < 0,30 0,02 1,55 0,06 N < 0,010 < 550
2.25Cr-1MoVTiB T/P 24 7CrMoVTiB 10-10 7CrMoVTiB 10-10 0,07 0,28 0,60 2,25 1,04 0,24 N < 0,010 < 550
B: 15-70 ppm
Ti: 0,05-0,10
5Cr-0.5Mo T/P 502 12 CrMo 19-5 0,12 0,35 0,65 5,10 0,54 < 550
9Cr-1Mo T/P 9 X12 CrMo 9-1 X12 CrMo 9-1 0,12 0,60 0,40 9,00 1,00 < 585
9Cr-1Mo mod. T/P 91 X10 CrMoVNb 9-1 X10 CrMoVNb 9-1 0,10 0,36 0,52 8,82 1,02 < 0,40 0,22 0,08 N: 30-70 ppm < 585
9Cr-0.5MoWV T/P 911 X11 CrMoWVNb 9-1-1 X11 CrMoWVNb 9-1-1 0,11 0,28 0,54 8,80 1,02 0,25 0,22 1,05 0,08 N: 0,05-0,09 < 625
9Cr-0.5MoWV T/P 92 X10 CrWMoNb 9-2 0,10 < 0,50 0,55 8,80 0,52 < 0,40 0,23 1,55 0,08 N: 0,03-0,07 < 625
B: 0,001-0,006
12Cr-0.25Mo X12 CrCoWVNb 11-2-2 0,11 0,45 0,20 11,50 0,23 0,28 0,24 1,40 0,07 Co: 1,30 < 650
+1.4W1.3Co0.2
V (VM 12-SHC) t<10mm N: 0,055
B: 0,003
12Cr-1MoNiV X20 CrMoV 12-1 X20 CrMoV 11-1 0,20 < 0,50 < 1,00 12,10 1,05 0,65 0,28 < 585
Tableau 1: Vue d’ensemble des classifications normalisées internationales, des compositions chimiques et des températures de service des aciers
CrMo résistant au fluage les plus courants.
Nuances CrMo Métaux de base Consommables de soudage pour les aciers CrMo
ASTM & EN SMAW GTAW GMAW SAW FCAW
ASME fil Flux
0.5Mo T/P 1 8MoB 5-4 Phoenix SH Schwarz 3 K Union I Mo Union I Mo Union S 2 Mo UV 420 TT Union TG Mo R
1.25Cr-0.5Mo T/P 11 10 CrMo 5-5 Phoenix Chromo 1 Union I CrMo Union I CrMo Union S 2 CrMo UV 420 TT Union TG CrMo R
1.00Cr-0.5Mo T/P 12 13 CoMo 4-5 Phoenix Chromo 1 Union I CrMo Union I CrMo Union S 2 CrMo UV 420 TT Union TG CrMo R
1.25Cr-1MoV 15 CrMoV 5-10 Phoenix SH Kupfer 3 K
-/- 15 NiCuNb 5 Phoenix SH Schwarz 3 K Ni Union I Mo Union I Mo Union S 3 NiMo 1 UV 420 TT(R)
(WB 36)
20 MnMoNi 5-5 Phoenix SH Schwarz 3 K Ni Union I MoMn Union I MoMn Union S 3 NiMo 1 UV 420 TT(R) Union TG Mo R
2.25Cr-1Mo T/P 22 10 CrMo 9-10 Phoenix SH Chromo 2 KS Union I CrMo 910 Union I CrMo 910 Union S 1 CrMo 2 UV 420 TTR Union TG CrMo 9 10 R
2.25Cr-1MoV T/P 22V Phoenix SH Chromo 2 V Union S 1 CrMo 2V UV 430 TTR-W
2.25Cr-MoVW T/P 23 7CrMoWVMoNb 9-6 Thermanit P23 Union I P23 Union I P23 Union S P23 UV 430 TTR-W UV P23
2.25Cr-1MoV T/P 24 7CrMoVTiB 10-10 Thermanit P24 Union I P24 Union I P24 Union S P24 UV 430 TTR-W UV P24
5Cr-0.5Mo T/P 502 12 CrMo 19-5 Phoenix Chromo 5 Union I CrMo 5 Union I CrMo 5 Union S1 CrMo 5 Marathon 543
9Cr-1Mo T/P 9 X12 CrMo 9-1 Thermanit Chromo 9 V Thermanit MTS 3 Thermanit MTS 3 Thermanit MTS 3 Marathon 543 Thermanit MTS 3 PW
9Cr-1Mo mod. T/P 91 X10 CrMoVNb 9-1 Thermanit Chromo 9 V Thermanit MTS 3 Thermanit MTS 3 Thermanit MTS 3 Marathon 543 Thermanit MTS 3 PW
Thermanit Chromo T91
9Cr-0.5MoWV T/P 911 X11 CrMoWVNb 9-1-1 Thermanit MTS 911 Thermanit MTS 911 Thermanit MTS 911 Thermanit MTS 911 Marathon 543
9Cr-0.5MoWV T/P 92 X10 CrWMoNb 9-2 Thermanit MTS 616 Thermanit MTS 616 Thermanit MTS 616 Thermanit MTS 616 Marathon 543
12Cr-0.25Mo X12 CrCoWVNb 11-2-2 Thermanit MTS 5 CoT Thermanit MTS 5 CoT
+1.4W1.3Co0.2V (VM 12-SHC) t<10mm
12Cr-1MoNiV X20 CrMoV 11-1 Thermanit MTS 4 Thermanit MTS 4 Si Thermanit MTS 4 Si Thermanit MTS 4 Marathon 543
Tableau 3: Sélection de métaux d’apport destinés à différents procédés utilisés pour le soudage les aciers CrMo résistant au fluage
Suggestions de températures de préchauffage, d’entre-passes, TTAS comme SR, ISR et traitement
Nuance CrMo Classifications normalisées thermique de STC pour aciers CrMo
ASTM & EN Tp Ti SR ISR PWHT/STC
ASME °C °C h, °C h, °C h, °C
0.5Mo T/P 1 8MoB 5-4 RT RT 2-4h @ 580-630°C
t
1.25Cr-0.5Mo T/P 11 10 CrMo 5-5 200-250°C > 200°C 2-4h @ 660-700°C Refroiss par paliers selon application
1,00Cr-0.5Mo T/P 12 13 CrMo 4-5 200-250°C > 200°C 2-4h @ 660-700°C
1.25Cr-1MoV 15 CrMoV 5-10 200-250°C > 200°C 2-4h @ 660-700°C
-/- 15 NiCuNb 5 200-250°C > 200°C 2-4h @ 580-620°C 60h @ 550°C + 40h @ 620°C
(WB 36)
21 MnMoNi 5-5 200-250°C > 200°C 2-4h @ 580-620°C
2.25Cr-1Mo T/P 22 10 CrMo 9-10 200-300°C 200-300°C 2-4h @ 670-720°C
2.25Cr-1MoV T/P 22V 200-300°C 200-250°C 1h @ 680°C 8h @ 705°C + STC + 32h @ 705°C
2.25Cr-MoVW T/P 23 7CrWVMoNb 9-6 200-300°C 200-300°C 1h @ 540-560°C* 0.5-4h @ 740°C**
2.25Cr-1MoV T/P 24 7CrMoVTiB 10-10 200-280°C 200-280°C 0.5-4h @ 740°C**
5Cr-0.5Mo T/P 502 12 CrMo 19-5 225-300°C > 225°C 2-4h @ 730-760°C
9Cr-1Mo T/P 9 X12 CrMo 9-1 200-300°C 200-300°C Refroid. lent après soudage xh @ 750°C
9Cr-1Mo mod. T/P 91 X10 CrMoVNb 9-1 200-300°C 200-300°C Refroid. lent après soudage
9Cr-0.5MoWV T/P 911 X11 CrMoWVNb 9-1-1 200-300°C 200-300°C Refroid. lent après soudage xh @ 730-780°C
9Cr-0.5MoWV T/P 92 X10 CrWMoNb 9-2 200-300°C 200-300°C Refroid. lent après soudage xh @ 730-780°C
12Cr-0.25Mo X12 CrCoWVNb 11-2-2 200-280°C 200-280°C Refroid. lent après soudage xh @ 770°C
+1.4W1.3Co0.2V (VM 12-SHC) t<10mm
12Cr-1MoNiV X20 CrMoV 12-1 200-280°C 200-280°C Refroid. lent après soudage xh @ 760°C
x fonction épaisseur
* avec de grandes différences ** pas de TTAS requis pour GTAW
dans les épaisseurs pour épaisseur jusqu’à 10mm
Tableau 4: Suggestions de paramètres thermiques tels que températures de préchauffage, d’entre-passes, TTAS comme SR, ISR et traitement
thermique de STC pour aciers CrMo. Voir aussi Figure 2.
5Cr-0.5Mo T/P502 12CrMo19-5 --- Appareils à pression sollicités à la corrosion par le soufre à chaud < 550
fours et réacteurs
Réacteurs, Résistance à la corrosion par le soufre à chaud, fours
9Cr-1Mo T/P9 X12CrMo9-1 --- tuyauterie et réacteurs < 585
12Cr-1MoNiV X20CrMoV11-1 Collecteurs de vapeur, surchauffeurs; Tuyauteries sous H2S < 585
Rp0.2 > 500 MPa, Rm 700-850 MPa
Table 5: Vue d’ensemble d’applications types en aciers CrMo pour la production d’énergie, le raffinage de pétrole et la pétrochimie.
Soudage d’aciers au CrMo pour la production d’énergie,
le raffinage de pétrole et la pétrochimie
- Passé, présent et futur -
JAN HILKES1, VOLKER GROSS1, PIERRE-ALEXANDRE LEGAIT2
1
Böhler Schweisstechnik Deutschland GmbH, 2Groupe Böhler Soudage France S.A.S
250
---PWHT
200 ---PWHT+stepcooling
[J]
150
flatbeads
CVN
100
thickbeads
55J
50
0
-80 -70 -60 -50 -40 -30 -20 -10 0
Testtemperature[°C]
Précipités de ferrite sur soudures Aspect de surface d’une fissure due à Influence de l’épaisseur des passes sur
SAW P11 l’hydrogène sur sou dures SAW P22V les valeurs de résilience
RECOMMANDATIONS
Nuance ASTM EN TTAS (°C)/ x hrs*
temp. type préchauffage/ entre-passes (°C) - Métal de base: 2,25%Cr 1%Mo
CrMo Classification T/P Classification **,***
1,25Cr.-0,5Mo 11/12 13CrMo-4-5/Si5-5 200-250 660-700 - Dimensions:
-épaisseur: 358 mm
2,25Cr-1Mo 22 10CrMo910 200-250 650/690 -Longueur: 21 m
2,25Cr-1MoV 22V 13CrMoV10 180-230 705 -Diamètre: 5,3 m
INDEX çç
Virole d'une unité d’hydrocraquage,ATB, Italie
Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page144
144
INDEX çç Sommaire
00-MEP ESOPE programme 2010_Mise en page 1 07/09/10 08:28 Page145
INDEX çç Sommaire
INDEX ESOPE 2010_Mise en page 1 08/09/10 15:45 Page1
INDEX
ABITTAN E. Contrôle en fabrication des soudures de tuyauteries : Présentation du guide des alternatives à la
gammagraphie à l’Ir 192 / Inspection during fabrication on piping welds : Presentation of the guide of
alternatives to the gammagraphy using Ir 195 68D
AKKUS N. Investigation par la méthode des éléments finis sur des appareils à pression composites renforcées par des
fibres sous chargement de choc / FEM investigation of the fiber reinforced composite pressure vessel under
impact loading 70G
ALLART E. Référentiel d’audit UFIP/UIC pour la sélection des réparateurs de soupapes de sûreté /
Reference system of audit UFIP/UIC for the selection of safety valves repairing 55F
ANCELET O. Comportements mécaniques des matériaux pour les réacteurs HTR : Développement en supports
de l’intégrité des structures. / Mechanicals behaviour of HTR materials: development in support of
defect assessment, structural integrity and lifetime evaluation. 34B
ANTALFFY L. Vanadium modified 2 1/4Cr-1Mo steel reactor fabrication and the resolution of recently experienced
fabrication problems 90C
BAOUDOUR J.-L. Guide des Bonnes Pratiques pour le contrôle par émission acoustique des équipements sous pression /
Guide to good practice for acoustic emission testing of pressure equipment 9Ea
BEIGE J Evaluation of non destructive TESTING for the detection of hot temperature hydrogen attack 25E 26E
BIRCH D. “Robinets de bouteilles de gaz : l’équipement “”Cinderella”“ / Gas Cylinder Valves, the Cinderella device “ 27H
BIREMBAUT Y. Emissions fugitives des robinets industriels - Le point sur la normalisation et les méthodes d’essai /
FUGITIVE EMISSIONS OF INDUSTRIAL VALVES:STANDARDS AND TESTING PROCEDURES 53F
BLANCHARD S Evaluation of non destructive TESTING for the detection of hot temperature hydrogen attack 25E 26E
BLETTNER A. Contrôle en fabrication des soudures de tuyauteries : Présentation du guide des alternatives à la
gammagraphie à l’Ir 192 / Inspection during fabrication on piping welds : Presentation of the guide of
alternatives to the gammagraphy using Ir 194 68D
BLOSSIER J.-L. Suivi en Service des Équipements Sous Pression à couvercle amovible utilisés dans l’industrie
aéronautique et spatiale / In-service inspection and requalification of autoclaves in the aircraft and
aerospace industry 45Ea
çç Sommaire
INDEX ESOPE 2010_Mise en page 1 08/09/10 15:45 Page2
BOUCHAUD Y Suivi en Service des Équipements Sous Pression à couvercle amovible utilisés dans l’industrie
aéronautique et spatiale / In-service inspection and requalification of autoclaves in the aircraft and
aerospace industry 45Ea
BOURGEOIS G. Incidence de l’application de la norme EN15085 pour les fabricants de produits chaudronnés /
Impact of the application of the EN15085 standard for fabricators 78H
BOURGES Ph. Expérience acquise en comprtement des aciers résistants à la corrosion dans des milieux acqueux
contenant du sulfure d’hydrogène - Conséquence des impositions de PWHT (Post Welded Heat Treatment) /
Experience achieved on sour service resistant steels behaviour - consequences of PWHT requirements 19B
BOURGES Ph. EFFET DU VIEILLISSEMENT THERMIQUE ET DE LA FRAGILISATION PAR L’HYDROGENE SUR LES
PROPRIETES DE MECANIQUE DE LA RUPTURE ET DE RESILIENCE DES NUANCES TYPE 2,25CR1MO –
APPLICATION A LA DETERMINATION DE LA TEMPERATURE MINIMALE DE PRESSURISATION (MPT) /
EFFECT OF TEMPER AND HYDROGEN EMBRITTLEMENT ON FRACTURE MECHANICS AND CVN PROPERTIES
OF 2,25CR1MO STEEL GRADES – APPLICATION TO MINIMUM PRESSURIZING TEMPERATURE (MPT) ISSUES 23B
BRAGATTO PA. La représentation numérique paramétrique d’équipement sous pression pour la certification de sûreté /
The parametric digital representation of pressure equipment for safety certification 95D
BREART N. Exemple d’évaluation de conformité menée par l’OIU d’EDF pour l’EPR Flamanville 3 /
An exemple of the evaluation of conformity carried out by OIU of EDF for EPR Flamanville 3 8D 7D
C. Faidy Comparaison des règles codifiées d’analyse à la fatigue - Une proposition pour améliorer les EN /
Comparaison of codified fatigue analysis rules - A proposal to improve EN standards 75J
C. Petesch “Comparaison des règles codifiées d’analyse à la fatigue - Une proposition pour améliorer les EN /
Comparaison of codified fatigue analysis rules - A proposal to improve EN standards 75J
CARPREAU J.-M. Procédés innovants pour la réparation des centrales nucléaires / Innovated processes for the repair of nuclear units 13D
CATTY J. “Diagnostic global des ESP : contrôle, calcul et matériaux (session “”Vie des appareils”“ ?) /
Global diagnosis of pressure equipment : inspection , calculation and materials “ 4E
CATTY J. Contrôle par Emission Acoustique d’équipement sous pression - Exemple d’application concernant une
unité de stockage de gaz de sphères sous talus. 87Ea
CAWELIUS R. Enveloppes cylindriques constituées de plaques épaisses, une alternative aux pièces forgées /
Shell courses made of heavy plates, an alternative to forged shell rings - Enveloppes cylindriques constituées
de plaques épaisses, une alternative aux pièces forgées / Shell courses made of heavy plates, an alternative to
forged shell rings 30B 33B
CAZES F. Modélisation de la déchirure ductile de l’apparition de la plasticité aux grandes propagations de fissure /
Modelling ductile tearing from plasticity onset to large crack propagation 37A
CHAUVEAU D. Contrôle en fabrication des soudures de tuyauteries : Présentation du guide des alternatives à la
gammagraphie à l’Ir 192 / Inspection during fabrication on piping welds : Presentation of the guide of
alternatives to the gammagraphy using Ir 192 68D
CHAUVY C. Expérience acquise en comprtement des aciers résistants à la corrosion dans des milieux acqueux contenant
du sulfure d’hydrogène - Conséquence des impositions de PWHT (Post Welded Heat Treatment) /
Experience achieved on sour service resistant steels behaviour - consequences of PWHT requirements 19B
CHAUVY C. EFFET DU VIEILLISSEMENT THERMIQUE ET DE LA FRAGILISATION PAR L’HYDROGENE SUR LES PROPRIETES
DE MECANIQUE DE LA RUPTURE ET DE RESILIENCE DES NUANCES TYPE 2,25CR1MO – APPLICATION A LA
DETERMINATION DE LA TEMPERATURE MINIMALE DE PRESSURISATION (MPT) / EFFECT OF TEMPER AND
HYDROGEN EMBRITTLEMENT ON FRACTURE MECHANICS AND CVN PROPERTIES OF 2,25CR1MO STEEL
GRADES – APPLICATION TO MINIMUM PRESSURIZING TEMPERATURE (MPT) ISSUES 23B
CHEFAOUI M. “Diagnostic global des ESP : contrôle, calcul et matériaux (session ”Vie des appareils”) /
Global diagnosis of pressure equipment : inspection , calculation and materials “ 4E
çç Sommaire
INDEX ESOPE 2010_Mise en page 1 08/09/10 15:45 Page3
CHERFAOUI M. De la nécessité du contrôle non destructif pour la qualité des équipements en composite /
The need of non-destructive to assure the quality of composite pressure equipement 88G
CHERFAOUI M. Guide des Bonnes Pratiques pour le contrôle par émission acoustique des équipements sous pression /
Guide to good practice for acoustic emission testing of pressure equipment 9Ea
CLOISEAU D. Suivi en Service des Équipements Sous Pression à couvercle amovible utilisés dans l’industrie
aéronautique et spatiale / In-service inspection and requalification of autoclaves in the aircraft and
aerospace industry 45Ea
COLIN J. Conception selon l’EN 13445 d’un réacteur-échangeur à tubes baïonnettes de grandes dimensions et
à haute température / Design according to EN 13445 of a large heat exchanger reactor with bayonet tubes
under high temerature 2A
COLONNA F. Anomalies de pose de bouchons mécaniques dans les tubes des générateurs de vapeur des réacteurs
nucléaires d’Electricité de France (EDF) / Plug installation faults on EDF Nnuclear reactor steam generator tubes 100E
COMBESCURE A. Modélisation de la déchirure ductile de l’apparition de la plasticité aux grandes propagations de fissure /
Modelling ductile tearing from plasticity onset to large crack propagation 37A
CORRE S. EFFET DU VIEILLISSEMENT THERMIQUE ET DE LA FRAGILISATION PAR L’HYDROGENE SUR LES PROPRIETES
DE MECANIQUE DE LA RUPTURE ET DE RESILIENCE DES NUANCES TYPE 2,25CR1MO – APPLICATION A LA
DETERMINATION DE LA TEMPERATURE MINIMALE DE PRESSURISATION (MPT) / EFFECT OF TEMPER AND
HYDROGEN EMBRITTLEMENT ON FRACTURE MECHANICS AND CVN PROPERTIES OF 2,25CR1MO STEEL GRADES
– APPLICATION TO MINIMUM PRESSURIZING TEMPERATURE (MPT) ISSUES 23B
COUDREUSE L. Expérience acquise en comprtement des aciers résistants à la corrosion dans des milieux acqueux contenant
du sulfure d’hydrogène - Conséquence des impositions de PWHT (Post Welded Heat Treatment) /
Experience achieved on sour service resistant steels behaviour - consequences of PWHT requirements 19B
COUDREUSE L. EFFET DU VIEILLISSEMENT THERMIQUE ET DE LA FRAGILISATION PAR L’HYDROGENE SUR LES PROPRIETES
DE MECANIQUE DE LA RUPTURE ET DE RESILIENCE DES NUANCES TYPE 2,25CR1MO – APPLICATION A LA
DETERMINATION DE LA TEMPERATURE MINIMALE DE PRESSURISATION (MPT) / EFFECT OF TEMPER AND
HYDROGEN EMBRITTLEMENT ON FRACTURE MECHANICS AND CVN PROPERTIES OF 2,25CR1MO STEEL GRADES
– APPLICATION TO MINIMUM PRESSURIZING TEMPERATURE (MPT) ISSUES 23B
COUGNON L. Emissions fugitives des robinets industriels - Le point sur la normalisation et les méthodes d’essai /
FUGITIVE EMISSIONS OF INDUSTRIAL VALVES:STANDARDS AND TESTING PROCEDURES 53F
CRIPPS P.R. Facteurs de sélection d’aciers alliés pour des appareils à pression contenant de l’hydrogen et travaillant
à haute température pour hydrocarbures / Selection Factors of alloy steels for pressure vessels in high
temperature hydrocarbon services containing hydrogen 99B
DELAGE J Contrôle de fonctionnement de soupapes de sûreté sur site / Inspection and adjustment in service of safety valves 51F
DELILLE B. Enceintes sous vide pour les accélérateurs de particules : des récipients de taille imposante soumis
à la pression extérieure / Vacuum vessels for particle accelerators : large size negative pressure vessels 29A
DEMONTE S Evaluation of non destructive TESTING for the detection of hot temperature hydrogen attack 25E 26E
DESOMBRE D. Générateur de Vapeur en situation de colmatage – Statut des internes supportant le faisceau tubulaire –
Vérification du respect des critères de tenue mécanique de ces internes, pour un GV en situation de colmatage
des plaques entretoises, avec la réglementation applicable. / Steam generator with clogging – Tubes bundle
support status – Verification of mechanical strength criteria with applicable regulations for steam generator with
tubes support plates clogging. 66A
DETEMPLE I. Enveloppes cylindriques constituées de plaques épaisses, une alternative aux pièces forgées /
Shell courses made of heavy plates, an alternative to forged shell rings 30B 33B
çç Sommaire
INDEX ESOPE 2010_Mise en page 1 08/09/10 15:45 Page4
DHENNIN D. Exemple d’évaluation de conformité menée par l’OIU d’EDF pour l’EPR Flamanville 3 /
An exemple of the evaluation of conformity carried out by OIU of EDF for EPR Flamanville 4 8D 7D
DUNOYER R. Référentiel d’audit UFIP/UIC pour la sélection des réparateurs de soupapes de sûreté /
Reference system of audit UFIP/UIC for the selection of safety valves repairing 55F
DUPLESSIS P Evaluation des dommages et réparation des réservoirs à charbons / Assessment and repair of bulges in coke drums 15E
EADY C. De l’importance des compétences des personnels chargés du soudage dans la fabrication /
The Special importance of personnel competence in welding fabrication 20C
EAS L. Suivi en Service des Équipements Sous Pression à couvercle amovible utilisés dans l’industrie
aéronautique et spatiale / In-service inspection and requalification of autoclaves in the aircraft and
aerospace industry 45Ea
FAIDY C Utilisation de la norme EN 13480 pour l’évaluation des équipements sous pression nucléaires
en situations accidentelles / Use of EN 13480 standards for the evaluation of nuclear pressure
equipment subjected to faulted conditions 76J 82J
GALVEZ J. Etude soudabilité de joints en matériaux différents utilisés dans la fabrication de réchauffeurs de production
d’énergie / Weldability study of dissimilar metal joints used in power generation reheaters manufactory 98C
GAMBHAVA D. Effet de l’énergie de chauffage et du revenu sur la ténacité des joints soudés en Cr-Mo-V /
Effect of Heat Input & Tempering parameter on Notch-Toughness property of Cr-Mo-V weldments 14C
GENC G. Investigation par la méthode des éléments finis sur des appareils à pression composites renforcées
par des fibres sous chargement de choc / FEM investigation of the fiber reinforced composite pressure
vessel under impact loading 70G
GENETTE P. Générateur de Vapeur en situation de colmatage – Statut des internes supportant le faisceau tubulaire –
Vérification du respect des critères de tenue mécanique de ces internes, pour un GV en situation de colmatage
des plaques entretoises, avec la réglementation applicable. / Steam generator with clogging – Tubes bundle
support status – Verification of mechanical strength criteria with applicable regulations for steam generator
with tubes support plates clogging. 66A
GHOSH M. Effet de l’énergie de chauffage et du revenu sur la ténacité des joints soudés en Cr-Mo-V /
Effect of Heat Input & Tempering parameter on Notch-Toughness property of Cr-Mo-V weldments 14C
GIROURDIERE F. Conception selon l’EN 13445 d’un réacteur-échangeur à tubes baïonnettes de grandes dimensions
et à haute température / Design according to EN 13445 of a large heat exchanger reactor with bayonet
tubes under high temerature 2A
çç Sommaire
INDEX ESOPE 2010_Mise en page 1 08/09/10 15:45 Page5
GOETCHELUNCK JF Générateur de Vapeur en situation de colmatage – Statut des internes supportant le faisceau tubulaire –
Vérification du respect des critères de tenue mécanique de ces internes, pour un GV en situation de colmatage
des plaques entretoises, avec la réglementation applicable. / Steam generator with clogging – Tubes bundle
support status – Verification of mechanical strength criteria with applicable regulations for steam generator
with tubes support plates clogging. 66A
GRANDEMANGE JM Comparaison des règles codifiées d’analyse à la fatigue - Une proposition pour améliorer les EN /
Comparaison of codified fatigue analysis rules - A proposal to improve EN standards 75J
GRANDEMANGE JM. Utilisation de la norme EN 13480 pour l’évaluation des équipements sous pression nucléaires
en situations accidentelles / Use of EN 13480 standards for the evaluation of nuclear pressure
equipment subjected to faulted conditions 76J 82J
HATSCH J. Contrôle en fabrication des soudures de tuyauteries : Présentation du guide des alternatives à la
gammagraphie à l’Ir 192 / Inspection during fabrication on piping welds : Presentation of the guide of
alternatives to the gammagraphy using Ir 193 68D
HAUVILLER C. Enceintes sous vide pour les accélérateurs de particules : des récipients de taille imposante
soumis à la pression extérieure / Vacuum vessels for particle accelerators : large size negative pressure vessels 29A
HERVE C. De la nécessité du contrôle non destructif pour la qualité des équipements en composite /
The need of non-destructive to assure the quality of composite pressure equipement 88G
HOULLE P. Perspectives offertes par un nouvel alliage de nickel à forte teneur en molybdène et chrome /
Shell courses made of heavy plates, an alternative to forged shell rings 42B
HUGHES P. Comment justifier l’absence de contrôles internes dans des usines à hauts risques /
Establishing a justification for-invasive examination of high-Hazard Process Plant 61F
JARDIN N. Procédés innovants pour la réparation des centrales nucléaires / Innovated processes for the repair of nuclear units 13D
JESSOP T. De l’importance des compétences des personnels chargés du soudage dans la fabrication /
The Special importance of personnel competence in welding fabrication 20C
KARCHER G. Développement les plus récents dans les codes et normes Asme dédiés aux équipements sous pression /
Latest developments in ASME pressure equipment codes and standards 71J
KASPARI P. (DR) Point de vue des producteurs d’aciers sur les demandes de tôle en acier CMn pour des équipements
sous pression en milieu acide / Steel producer’s view on demands for CMn steel plates processed to
sour service pressure vessels 31B 32B
KOCH K-M Matériaux en titane pour applications chimiques / Titanium materials for chemicals applications 91B
KRAUT R. Enveloppes cylindriques constituées de plaques épaisses, une alternative aux pièces forgées /
Shell courses made of heavy plates, an alternative to forged shell rings 30B 33B
LAENEN C. Retour d’expérience de la méthode des END des ondes guidées ultrasonores /
Feed back on the method en essai non destructif of ultrasonic guided waves 72D
çç Sommaire
INDEX ESOPE 2010_Mise en page 1 08/09/10 15:45 Page6
LE GALL A. Conception selon l’EN 13445 d’un réacteur-échangeur à tubes baïonnettes de grandes dimensions
et à haute température / Design according to EN 13445 of a large heat exchanger reactor with bayonet
tubes under high temerature 2A
LE JEUNE H. Emissions fugitives des robinets industriels - Le point sur la normalisation et les méthodes d’essai /
FUGITIVE EMISSIONS OF INDUSTRIAL VALVES:STANDARDS AND TESTING PROCEDURES 53F
LELIEVRE C Evaluation of non destructive TESTING for the detection of hot temperature hydrogen attack 25E 26E
LENAIN J.C. Suivi en Service des Équipements Sous Pression à couvercle amovible utilisés dans l’industrie
aéronautique et spatiale / In-service inspection and requalification of autoclaves in the aircraft
and aerospace industry 45Ea
LENAIN J.C. Surveillance en service et à distance des ESP par émission acoustique /
On line and remote monitoring of presure equpments by acousitc emission 47Ea
LENEVE C Evaluation of non destructive TESTING for the detection of hot temperature hydrogen attack 25E 26E
LEWIS A. Effet de l’énergie de chauffage et du revenu sur la ténacité des joints soudés en Cr-Mo-V /
Effect of Heat Input & Tempering parameter on Notch-Toughness property of Cr-Mo-V weldments 14C
LIMOUSIN S. Anomalies de pose de bouchons mécaniques dans les tubes des générateurs de vapeur des réacteurs
nucléaires d’Electricité de France (EDF) / Plug installation faults on EDF Nnuclear reactor steam generator tubes 100E
LOPEZ A Etude soudabilité de joints en matériaux différents utilisés dans la fabrication de réchauffeurs de
production d’énergie / Weldability study of dissimilar metal joints used in power generation reheaters manufactory 98C
LOUKACHENKO N. Expérience acquise en comprtement des aciers résistants à la corrosion dans des milieux acqueux contenant
du sulfure d’hydrogène - Conséquence des impositions de PWHT (Post Welded Heat Treatment) /
Experience achieved on sour service resistant steels behaviour - consequences of PWHT requirements 19B
LUCIEN E. Conception selon l’EN 13445 d’un réacteur-échangeur à tubes baïonnettes de grandes dimensions
et à haute température / Design according to EN 13445 of a large heat exchanger reactor with bayonet
tubes under high temerature 2A
MAFFERT J. Point de vue des producteurs d’aciers sur les demandes de tôle en acier CMn pour des équipements
sous pression en milieu acide / Steel producer’s view on demands for CMn steel plates processed to
sour service pressure vessels 31B 32B
MAFFERT J. Propriétés spéciales des tôles ex pour une meilleure mise en service pour les facteurs d’équipements
sous pression / Specific properties of heavy steel plates leading to better efficiency for pressure vessel fabricators 31B 32B
çç Sommaire
INDEX ESOPE 2010_Mise en page 1 08/09/10 15:45 Page7
MALOUINES P. Générateur de Vapeur en situation de colmatage – Statut des internes supportant le faisceau tubulaire –
Vérification du respect des critères de tenue mécanique de ces internes, pour un GV en situation de colmatage
des plaques entretoises, avec la réglementation applicable. / Steam generator with clogging –
Tubes bundle support status – Verification of mechanical strength criteria with applicable regulations for
steam generator with tubes support plates clogging. 66A
MARIE O. Comportements mécaniques des matériaux pour les réacteurs HTR : Développement en supports de
l’intégrité des structures. / Mechanicals behaviour of HTR materials: development in support of defect
assessment, structural integrity and lifetime evaluation. 34B
MARIE S. Modélisation de la déchirure ductile de l’apparition de la plasticité aux grandes propagations de fissure /
Modelling ductile tearing from plasticity onset to large crack propagation 37A
MARMONIER M. Exemple d’évaluation de conformité menée par l’OIU d’EDF pour l’EPR Flamanville 3 /
An exemple of the evaluation of conformity carried out by OIU of EDF for EPR Flamanville 5 8D 7D
MECK S. Perspectives offertes par un nouvel alliage de nickel à forte teneur en molybdène et chrome /
Shell courses made of heavy plates, an alternative to forged shell rings 42B
MOORE K. Vanadium modified 2 1/4Cr-1Mo steel reactor fabrication and the resolution of recently
experienced fabrication problems 90C
MOORE P. Comment justifier l’absence de contrôles internes dans des usines à hauts risques /
Establishing a justification for-invasive examination of high-Hazard Process Plant 61F
MORAN D. Etude soudabilité de joints en matériaux différents utilisés dans la fabrication de réchauffeurs de production
d’énergie / Weldability study of dissimilar metal joints used in power generation reheaters manufactory 98C
MORIN A. Anomalies de pose de bouchons mécaniques dans les tubes des générateurs de vapeur des réacteurs
nucléaires d’Electricité de France (EDF) / Plug installation faults on EDF Nnuclear reactor steam generator tubes 100E
MÜNSTERMANN S. Strain-based design approaches for unfired pressure vessels - an extension of DBA methology based based
on ductile failurelocus criteria 89A
NARJOZ C. Rechargement sur site d’une colonne de distillation avec revêtement anti-corrosion /
On site overlay of distillation column with corrosion protection layer 86E
OSAGE D. A. Fondements techniques des nouvelles règles de ténacité dans la section VIII Division 2 du B&PV code l’Asme /
Technical Basis of material toughness requirements in the Asme B&PV code Section VIII, Division 2 10A
OSAGE D. A. Fondements théoriques du Code Case 2605 de l’ASME pour les matériaux en 2,25Cr-1Mo-V -
Extension de la limite d’utilisation dans le domaine du fluage. / Background to the Development
of ASME B&PV Code Case 2605 - Extension of 2.25Cr-1Mo-V materials to 851 11B
OSAGE D. A. Gestion et suivi en service des équipements sous pression / Life-Cycle Management of pressurized fixed
Equipment using the latest API and ASME Codes and Standards 12E
PASTOR T. Développement les plus récents dans les codes et normes Asme dédiés aux équipements sous pression /
Latest developments in ASME pressure equipment codes and standards 71J
PATEL R. Effet de l’énergie de chauffage et du revenu sur la ténacité des joints soudés en Cr-Mo-V /
Effect of Heat Input & Tempering parameter on Notch-Toughness property of Cr-Mo-V weldments 14C
PETESCH C. Utilisation de la norme EN 13480 pour l’évaluation des équipements sous pression nucléaires
en situations accidentelles / Use of EN 13480 standards for the evaluation of nuclear pressure
equipment subjected to faulted conditions 76J 82J
çç Sommaire
INDEX ESOPE 2010_Mise en page 1 08/09/10 15:45 Page8
PILLOT S. EFFET DU VIEILLISSEMENT THERMIQUE ET DE LA FRAGILISATION PAR L’HYDROGENE SUR LES PROPRIETES
DE MECANIQUE DE LA RUPTURE ET DE RESILIENCE DES NUANCES TYPE 2,25CR1MO – APPLICATION A LA
DETERMINATION DE LA TEMPERATURE MINIMALE DE PRESSURISATION (MPT) / EFFECT OF TEMPER AND
HYDROGEN EMBRITTLEMENT ON FRACTURE MECHANICS AND CVN PROPERTIES OF 2,25CR1MO STEEL
GRADES – APPLICATION TO MINIMUM PRESSURIZING TEMPERATURE (MPT) ISSUES 23B
PINTOS V. Etude soudabilité de joints en matériaux différents utilisés dans la fabrication de réchauffeurs de
production d’énergie / Weldability study of dissimilar metal joints used in power generation reheaters manufactory 98C
PITROU B. La conception des tuyauteries en composite / Design of piping in composites 60G
PITTIGLIO P. La représentation numérique paramétrique d’équipement sous pression pour la certification de sûreté /
The parametric digital representation of pressure equipment for safety certification 95D
PRAGER M. Fondements techniques des nouvelles règles de ténacité dans la section VIII Division 2 du B&PV code l’Asme /
Technical Basis of material toughness requirements in the Asme B&PV code Section VIII, Division 3 10A
PRAGER M. Fondements théoriques du Code Case 2605 de l’ASME pour les matériaux en 2,25Cr-1Mo-V -
Extension de la limite d’utilisation dans le domaine du fluage. / Background to the Development of
ASME B&PV Code Case 2605 - Extension of 2.25Cr-1Mo-V materials to 850 F 11B
RENAUD L. Réservoirs de stockage en matériaux plastiques : cas vécus de rupture brutale - retour d’expérience
de contrôles par émission acoustique / Storages tanks in plastic : actual experience of a fracture -
Feedback about inspection by accoustic emission 73G
RICHEZ M Evaluation of non destructive TESTING for the detection of hot temperature hydrogen attack 25E 26E
SALAVY JF. Développement d’un module de test de couverture tritigène pour ITER /
Development of a breeding Blanket Test Module for ITER 35A
SAMMAN M Evaluation des dommages et réparation des réservoirs à charbons / Assessment and repair of bulges in coke drums 15E
SANZ GARCIA E. Conception selon l’EN 13445 d’un réacteur-échangeur à tubes baïonnettes de grandes dimensions
et à haute température / Design according to EN 13445 of a large heat exchanger reactor with bayonet
tubes under high temerature 2A
SAUGER E. Emissions fugitives des robinets industriels - Le point sur la normalisation et les méthodes d’essai /
FUGITIVE EMISSIONS OF INDUSTRIAL VALVES:STANDARDS AND TESTING PROCEDURES 53F
SCHÄFER D. Strain-based design approaches for unfired pressure vessels - an extension of DBA methology
based on ductile failurelocus criteria 89A
SCHAURITSCH G. Contrôle par émission acoustique de réservoirs pour GPL, enterrés en béton /
Acoustic Emission Testing (AT) of underground concrete convered LPG tanks 17Ea
SCHRUFF C. Strain-based design approaches for unfired pressure vessels - an extension of DBA methology
based on ductile failurelocus criteria 89A
SHARGAY C. Vanadium modified 2 1/4Cr-1Mo steel reactor fabrication and the resolution of recently experienced
fabrication problems 90C
SHEEHAN M. Développement les plus récents dans les codes et normes Asme dédiés aux équipements sous pression /
Latest developments in ASME pressure equipment codes and standards 71J
SIEIRO F. Etude soudabilité de joints en matériaux différents utilisés dans la fabrication de réchauffeurs de
production d’énergie / Weldability study of dissimilar metal joints used in power generation reheaters manufactory 98C
çç Sommaire
INDEX ESOPE 2010_Mise en page 1 08/09/10 15:45 Page9
SIMATOS A. Modélisation de la déchirure ductile de l’apparition de la plasticité aux grandes propagations de fissure /
Modelling ductile tearing from plasticity onset to large crack propagation 37A
SIMONET Y.
Méthodes de vérification de la résistance par analyse des modes de défaillance /
Design by analysis based on failure modes 5A
SMYTHE D. Vanadium modified 2 1/4Cr-1Mo steel reactor fabrication and the resolution of recently experienced
fabrication problems 90C
SOUTIF E. Conception selon l’EN 13445 d’un réacteur-échangeur à tubes baïonnettes de grandes dimensions
et à haute température / Design according to EN 13445 of a large heat exchanger reactor with bayonet
tubes under high temerature 2A
STEFFEN M. Propriétés spéciales des tôles ex pour une meilleure mise en service pour les facteurs d’équipements
sous pression / Specific properties of heavy steel plates leading to better efficiency for pressure vessel fabricators 31B 32B
TERRIEN N. De la nécessité du contrôle non destructif pour la qualité des équipements en composite /
The need of non-destructive to assure the quality of composite pressure equipement 88G
THIENE A. Point de vue des producteurs d’aciers sur les demandes de tôle en acier CMn pour des équipements
sous pression en milieu acide / Steel producer’s view on demands for CMn steel plates processed to sour
service pressure vessels 31B 32B
TONTI A. La représentation numérique paramétrique d’équipement sous pression pour la certification de sûreté /
The parametric digital representation of pressure equipment for safety certification 95D
TOPALIS P. Application de l’inspection basée sur criticité en Chine / Risk based inspection implementation in China 69E
TOPTAS E. Investigation par la méthode des éléments finis sur des appareils à pression composites renforcées par
des fibres sous chargement de choc / FEM investigation of the fiber reinforced composite pressure vessel
under impact loading 70G
TOUSSAINT P Expérience acquise en comprtement des aciers résistants à la corrosion dans des milieux acqueux
contenant du sulfure d’hydrogène - Conséquence des impositions de PWHT (Post Welded Heat Treatment) /
Experience achieved on sour service resistant steels behaviour - consequences of PWHT requirements 19B
TOUSSAINT P EFFET DU VIEILLISSEMENT THERMIQUE ET DE LA FRAGILISATION PAR L’HYDROGENE SUR LES PROPRIETES
DE MECANIQUE DE LA RUPTURE ET DE RESILIENCE DES NUANCES TYPE 2,25CR1MO – APPLICATION A LA
DETERMINATION DE LA TEMPERATURE MINIMALE DE PRESSURISATION (MPT) / EFFECT OF TEMPER AND
HYDROGEN EMBRITTLEMENT ON FRACTURE MECHANICS AND CVN PROPERTIES OF 2,25CR1MO STEEL
GRADES – APPLICATION TO MINIMUM PRESSURIZING TEMPERATURE (MPT) ISSUES 23B
TSCHELIESNIG P. Contrôle par émission acoustique de réservoirs pour GPL, enterrés en béton /
Acoustic Emission Testing (AT) of underground concrete convered LPG tanks 17Ea
VANDENBERGHE B. “Diagnostic global des ESP : contrôle, calcul et matériaux (session “”Vie des appareils”“ ?) /
Global diagnosis of pressure equipment : inspection , calculation and materials “ 4E
VIAUD A. Contrôle d’un ESP par émission acoustique dans le cadre d’une requalification aménagée :
Ne pas perdre de vue l’objectif ! / Inspection by accoustic emission as part of a planed requalification
for a pressure equipment . Do not lost the object ! 1Ea
WALASZEK H. De la nécessité du contrôle non destructif pour la qualité des équipements en composite /
The need of non-destructive to assure the quality of composite pressure equipement 88G
WANG H. Application de l’inspection basée sur criticité en Chine / Risk based inspection implementation in China 69E
WINTLE J. Comment justifier l’absence de contrôles internes dans des usines à hauts risques /
Establishing a justification for-invasive examination of high-Hazard Process Plant 61F
Yuan Y. “Comparaison des règles codifiées d’analyse à la fatigue - Une proposition pour améliorer les EN /
çç Sommaire
00-2e,3e,4e Couv. Esope 2010_Mise en page 1 13/09/10 10:42 Page2
Membres fondateurs /
Founder members
Partenaires officiels /
Official sponsors
APITI
Association pour la Promotion de
L'Inspection Technique chez
Les Industriels
A.F.I.A.P.
Association Française des Ingénieurs en Appareils à Pression
39-41, rue Louis Blanc – 92400 COURBEVOIE
Association régie par la loi du 1er juillet 1901