EUROPEAN STANDARD DRAFT

NORME EUROPÉENNE prEN 1993-1-9

EUROPÄISCHE NORM
March 2023

ICS 91.010.30; 91.080.13 Will supersede EN 1993-1-9:2005

English Version

Eurocode 3: Design of steel structures - Part 1-9: Fatigue

Eurocode 3: Calcul des structures en acier - Partie 1-9: Eurocode 3: Bemessung und Konstruktion von
Fatigue Stahlbauten - Teil 1-9: Ermüdung

This draft European Standard is submitted to CEN members for enquiry. It has been drawn up by the Technical Committee
CEN/TC 250.

If this draft becomes a European Standard, CEN members are bound to comply with the CEN/CENELEC Internal Regulations
which stipulate the conditions for giving this European Standard the status of a national standard without any alteration.

This draft European Standard was established by CEN in three official versions (English, French, German). A version in any other
language made by translation under the responsibility of a CEN member into its own language and notified to the CEN-CENELEC
Management Centre has the same status as the official versions.

CEN members are the national standards bodies of Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia,
Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Norway,
Poland, Portugal, Republic of North Macedonia, Romania, Serbia, Slovakia, Slovenia, Spain, Sweden, Switzerland, Türkiye and
United Kingdom.

Recipients of this draft are invited to submit, with their comments, notification of any relevant patent rights of which they are
aware and to provide supporting documentation.

Warning : This document is not a European Standard. It is distributed for review and comments. It is subject to change without
notice and shall not be referred to as a European Standard.

EUROPEAN COMMITTEE FOR STANDARDIZATION

COMITÉ EUROPÉEN DE NORMALISATION
EUROPÄISCHES KOMITEE FÜR NORMUNG

CEN-CENELEC Management Centre: Rue de la Science 23, B-1040 Brussels

© 2023 CEN All rights of exploitation in any form and by any means reserved Ref. No. prEN 1993-1-9:2023 E
worldwide for CEN national Members.
prEN 1993-1-9:2023 (E)

Contents Page

1 Scope.......................................................................................................................................................... 9
1.1 Scope of EN 1993-1-9 ........................................................................................................................... 9
1.2 Assumptions ........................................................................................................................................... 9
2 Normative references........................................................................................................................ 10
3 Terms, definitions and symbols .................................................................................................... 10
3.1 Terms and definitions ....................................................................................................................... 10
3.1.1 General.................................................................................................................................................... 10
3.1.2 Fatigue actions ..................................................................................................................................... 15
3.1.3 Fatigue action effect ........................................................................................................................... 17
3.1.4 Fatigue resistance ............................................................................................................................... 19
3.1.5 Fatigue verification ............................................................................................................................ 24
3.2 Symbols .................................................................................................................................................. 25
4 Basis of fatigue design ....................................................................................................................... 27
5 Fatigue design concepts.................................................................................................................... 28
6 Fatigue design methods .................................................................................................................... 29
6.1 Design stress methods ...................................................................................................................... 29
6.2 Verification methods ......................................................................................................................... 29
7 Fatigue action effect ........................................................................................................................... 30
7.1 Calculation of nominal stresses ..................................................................................................... 30
7.2 Relevant nominal stresses ............................................................................................................... 30
7.3 Calculation of nominal stress ranges ........................................................................................... 33
7.3.1 General.................................................................................................................................................... 33
7.3.2 Design value of nominal stress range .......................................................................................... 33
7.4 Effective design value of stress range.......................................................................................... 34
8 Fatigue resistance ............................................................................................................................... 35
8.1 Fatigue resistance curves................................................................................................................. 35
8.2 Classification of constructional details ....................................................................................... 41
8.3 Fatigue resistance modifications .................................................................................................. 42
8.3.1 Size effect ............................................................................................................................................... 42
8.3.2 Post-fabrication treatment.............................................................................................................. 42
9 Fatigue verification ............................................................................................................................ 43
9.1 Verification with respect to elastic behaviour ......................................................................... 43
9.2 Verification with respect to reference value ............................................................................ 43
9.3 Verification with respect to fatigue limit ................................................................................... 44
9.4 Verification for multiaxial fatigue ................................................................................................ 44
10 Classified constructional details for the nominal stress method ...................................... 45
Annex A (normative) Verification using cumulative linear damage model ............................... 84
A.1 Use of this annex ................................................................................................................................. 84
A.2 Scope and field of application ........................................................................................................ 84
A.3 Fatigue action effect ........................................................................................................................... 84
A.3.1 Stresses from fatigue actions .......................................................................................................... 84
A.3.2 Calculation of stress ranges ............................................................................................................ 84
A.4 Fatigue resistance ............................................................................................................................... 85
A.4.1 Endurance for the nominal stress method ................................................................................ 85
A.4.2 Endurance for the hot spot stress method................................................................................. 87
A.4.3 Endurance for the effective notch stress method.................................................................... 87

2
 prEN 1993-1-9:2023 (E)

A.4.4 Endurance for welded joints subjected to High Frequency Mechanical Impact
Treatment.............................................................................................................................................. 87
A.5 Fatigue verification ............................................................................................................................ 88
Annex B (normative) Hot spot stress method ....................................................................................... 91
B.1 Use of this annex ................................................................................................................................. 91
B.2 Scope and field of application ........................................................................................................ 91
B.3 Fatigue action effect........................................................................................................................... 91
B.3.1 Stresses from fatigue actions.......................................................................................................... 91
B.3.2 Calculation of stress ranges ............................................................................................................ 93
B.4 Fatigue resistance............................................................................................................................... 94
B.4.1 Fatigue resistance curves ................................................................................................................ 94
B.4.2 Classification of constructional details ....................................................................................... 96
B.4.3 Fatigue resistance modification .................................................................................................... 99
B.5 Fatigue verification ......................................................................................................................... 100
Annex C (normative) Effective notch stress method ........................................................................ 101
C.1 Use of this annex .............................................................................................................................. 101
C.2 Scope and field of application ..................................................................................................... 101
C.3 Fatigue action effect........................................................................................................................ 101
C.3.1 Stresses from fatigue action......................................................................................................... 101
C.3.2 Calculation of stress ranges ......................................................................................................... 102
C.4 Fatigue resistance............................................................................................................................ 103
C.4.1 Fatigue resistance curves ............................................................................................................. 103
C.4.2 Classification of constructional details .................................................................................... 103
C.5 Fatigue verification ......................................................................................................................... 104
Annex D (informative) Recommendations for magnification factors k1 and stress
concentration factors kf ................................................................................................................. 105
D.1 Use of this annex .............................................................................................................................. 105
D.2 Scope and field of application ..................................................................................................... 105
D.3 Secondary moments in lattice girders ..................................................................................... 105
D.4 Flanges of ⌶-section girders with transitions in thickness or width ........................... 106
D.5 Thickness transitions in plates................................................................................................... 108
D.6 Shell structures ................................................................................................................................ 108
Annex E (informative) Recommendations for preloaded bolts and rods subject to tension
................................................................................................................................................................ 109
E.1 Use of this annex .............................................................................................................................. 109
E.2 Scope and field of application ..................................................................................................... 109
E.3 Simplified calculation method .................................................................................................... 110
Annex F (informative) Fatigue design of welded joints subjected to High Frequency
Mechanical Impact Treatment .................................................................................................... 112
F.1 Use of this annex .............................................................................................................................. 112
F.2 Scope and field of application ..................................................................................................... 112
F.3 Fatigue action effect........................................................................................................................ 113
F.3.1 Stresses from fatigue actions....................................................................................................... 113
F.3.2 Calculation of the stress ranges .................................................................................................. 113
F.4 Fatigue resistance............................................................................................................................ 114
F.4.1 Fatigue resistance curves ............................................................................................................. 114
F.4.2 Classification of constructional details .................................................................................... 115
F.4.3 Alternative formulae for determination of detail category ............................................. 119
F.4.4 Fatigue resistance modification ................................................................................................. 119
F.5 Fatigue verification ......................................................................................................................... 120

3
prEN 1993-1-9:2023 (E)

F.6 Requirements for application ..................................................................................................... 120

F.6.1 Requirements for welds before HFMI treatment ................................................................. 120
F.6.2 Requirements for welds after HFMI treatment..................................................................... 121
F.6.3 Quality control .................................................................................................................................. 121
F.7 Treatment of variable amplitude loading ............................................................................... 121
Annex G (informative) Hot spot stress reference detail method ................................................. 123
G.1 Use of this annex .............................................................................................................................. 123
G.2 Scope and field of application ..................................................................................................... 123
G.3 Fatigue action effect ........................................................................................................................ 123
G.4 Fatigue resistance ............................................................................................................................ 123
G.5 Fatigue verification ......................................................................................................................... 124
Bibliography .................................................................................................................................................... 125

4
 prEN 1993-1-9:2023 (E)

European foreword

This document (prEN 1993-1-9:2023) has been prepared by Technical Committee CEN/TC 250
“Structural Codes”, the secretariat of which is held by BSI. CEN/TC 250 is responsible for all Structural
Eurocodes and has been assigned responsibility for structural and geotechnical design matters by CEN.
This document is currently submitted to the CEN Enquiry.
This document will supersede EN 1993-1-9:2005 and EN 1993-1-9:2005/AC:2009.
The first generation of EN Eurocodes was published between 2002 and 2007. This document forms part
of the second generation of the Eurocodes, which have been prepared under Mandate M/515 issued to
CEN by the European Commission and the European Free Trade Association.
The Eurocodes have been drafted to be used in conjunction with relevant execution, material, product
and test standards, and to identify requirements for execution, materials, products and testing that are
relied upon by the Eurocodes.
The Eurocodes recognise the responsibility of each Member State and have safeguarded their right to
determine values related to regulatory safety matters at national level through the use of national
annexes.

5
prEN 1993-1-9:2023 (E)

Introduction

0.1 Introduction to the Eurocodes

The Structural Eurocodes comprise the following standards generally consisting of a number of parts:
• EN 1990 Eurocode: Basis of structural and geotechnical design;

• EN 1991 Eurocode 1: Actions on structures;

• EN 1992 Eurocode 2: Design of concrete structures;

• EN 1993 Eurocode 3: Design of steel structures;

• EN 1994 Eurocode 4: Design of composite steel and concrete structures;

• EN 1995 Eurocode 5: Design of timber structures;

• EN 1996 Eurocode 6: Design of masonry structures;

• EN 1997 Eurocode 7: Geotechnical design;

• EN 1998 Eurocode 8: Design of structures for earthquake resistance;

• EN 1999 Eurocode 9: Design of aluminium structures;

• New parts are under development, e.g. Eurocode for design of structural glass.

The Eurocodes are intended for use by designers, clients, manufacturers, constructors, relevant
authorities (in exercising their duties in accordance with national or international regulations),
educators, soft-ware developers, and committees drafting standards for related product, testing and
execution standards.
NOTE Some aspects of design are most appropriately specified by relevant authorities or, where not specified,
can be agreed on a project-specific basis between relevant parties such as designers and clients. The Eurocodes
identify such aspects making explicit reference to relevant authorities and relevant parties.

0.2 Introduction to EN 1993 (all parts)

EN 1993 (all parts) applies to the design of buildings and civil engineering works in steel. It complies with
the principles and requirements for the safety and serviceability of structures, the basis of their design
and verification that are given in EN 1990 – Basis of structural design.
EN 1993 (all parts) is concerned only with requirements for resistance, serviceability, durability and fire
resistance of steel structures. Other requirements, e.g. concerning thermal or sound insulation, are not
covered.
EN 1993 is subdivided in various parts:
• EN 1993-1, Design of Steel Structures — Part 1: General rules and rules for buildings;

• EN 1993-2, Design of Steel Structures — Part 2: Steel bridges;

• EN 1993-3, Design of Steel Structures — Part 3: Towers, masts and chimneys;

• EN 1993-4, Design of Steel Structures — Part 4: Silos and tanks;

6
 prEN 1993-1-9:2023 (E)

• EN 1993-5, Design of Steel Structures — Part 5: Piling;

• EN 1993-6, Design of Steel Structures — Part 6: Crane supporting structures;

• EN 1993-7, Design of steel structures — Part 7: Design of sandwich panels.

EN 1993-1 in itself does not exist as a physical document, but as a document series that comprises the
following 14 separate parts, the basic part being EN 1993-1-1:
• EN 1993-1-1, Design of Steel Structures — Part 1-1: General rules and rules for buildings;

• EN 1993-1-2, Design of Steel Structures — Part 1-2: Structural fire design;

• EN 1993-1-3, Design of Steel Structures — Part 1-3: Cold-formed members and sheeting;

NOTE Cold formed hollow sections supplied according to EN 10219 are covered in EN 1993-1-1.

• EN 1993-1-4, Design of Steel Structures — Part 1-4: Stainless steels;

• EN 1993-1-5, Design of Steel Structures — Part 1-5: Plated structural elements;

• EN 1993-1-6, Design of Steel Structures — Part 1-6: Strength and stability of shell structures;

• EN 1993-1-7, Design of Steel Structures — Part 1-7: Strength and stability of planar plated structures
transversely loaded;

• EN 1993-1-8, Design of Steel Structures — Part 1-8: Design of joints;

• EN 1993-1-9, Design of Steel Structures — Part 1-9: Fatigue strength of steel structures;

• EN 1993-1-10, Design of Steel Structures — Part 1-10: Material toughness and through-thickness
properties;

• EN 1993-1-11, Design of Steel Structures — Part 1-11: Design of structures with tension components
made of steel;

• EN 1993-1-12, Design of Steel Structures — Part 1-12: Additional rules for steel grades up to S960;

• EN 1993-1-13, Design of Steel Structures — Part 1-13: Beams with large web openings;

• EN 1993-1-14, Design of Steel Structures — Part 1-14: Design assisted by finite element analysis.

All subsequent parts numbered EN 1993-1-2 to EN 1993-1-14 treat general topics that are independent
from the structural type like structural fire design, cold-formed members and sheeting, stainless steels,
plated structural elements, etc.
All subsequent parts numbered EN 1993-2 to EN 1993-7 treat topics relevant for a specific structural
type like steel bridges, towers, masts and chimneys, silos and tanks, piling, crane supporting structures,
etc. EN 1993-2 to EN 1993-7 refer to the generic rules in EN 1993-1 and supplement, modify or supersede
them.

7
prEN 1993-1-9:2023 (E)

0.3 Introduction to EN 1993-1-9

EN 1993-1-9 gives specific design rules for verification of fatigue resistance of steel structures. It is
intended to be used with EN 1990, EN 1991 and EN 1993-1. Matters that are already covered in those
documents are not repeated. The focus in EN 1993-1-9 is on design rules that supplement, modify or
supersede the equivalent provisions given in EN 1993-1.
0.4 Verbal forms used in the Eurocodes

The verb “shall" expresses a requirement strictly to be followed and from which no deviation is permitted
in order to comply with the Eurocodes.
The verb “should” expresses a highly recommended choice or course of action. Subject to national
regulation and/or any relevant contractual provisions, alternative approaches could be used/adopted
where technically justified.
The verb “may" expresses a course of action permissible within the limits of the Eurocodes.
The verb “can" expresses possibility and capability; it is used for statements of fact and clarification of
concepts.
0.5 National annex for EN 1993-1-9

National choice is allowed in this standard where explicitly stated within notes. National choice includes
the selection of values for Nationally Determined Parameters (NDPs).

The national standard implementing EN 1993-1-9 can have a National Annex containing all national
choices to be used for the design of steel structures to be constructed in the relevant country.

When no national choice is given, the default choice given in this standard is to be used.

When no national choice is made and no default is given in this standard, the choice can be specified by a
relevant authority or, where not specified, agreed for a specific project by appropriate parties.
National choice is allowed in EN 1993-1-9 through notes to the following:
1.1(8) 4(6) 5(4) 5(6)
6.1(3) – 3 choices 7.1(4) 8.2(1) – 2 choices 9.1(1)
9.4(3) B.2(1) B.2(1) C.2(4)
C.2(5) F.2(2) F.2(5) F.2(6)
F.3.2(1) F.4.2.1(3)
National choice is allowed in EN 1993-1-9 on the application of the following informative annexes:
Annex D Annex E Annex F Annex G
The National annex may contain, directly or by reference, non-contradictory complementary information
for ease of implementation, provided it does not alter any provisions of the Eurocodes.

8
 prEN 1993-1-9:2023 (E)

1 Scope

1.1 Scope of EN 1993-1-9

(1) EN 1993-1-9 gives design methods for the verification of the fatigue design situation of steel
structures.
NOTE Steel structures consist of members and their joints. Each member and joint can be represented as a
constructional detail or as several of the latter.

(2) Design methods other than the stress-based methods, such as the notch strain method or fracture
mechanics methods, are not covered by EN 1993-1-9.
(3) EN 1993-1-9 only applies to structures made of all grades of structural steels which conform to
EN 1993-1 (all parts), in accordance with the provisions noted in the detail category tables or annexes.
(4) EN 1993-1-9 only applies to structures where execution conforms to EN 1090-2.
NOTE Supplementary execution requirements are indicated in the detail category tables.

(5) EN 1993-1-9 applies to structures operating under normal atmospheric conditions and with
sufficient corrosion protection and regular maintenance. The effect of seawater corrosion is not covered.
(6) EN 1993-1-9 applies to structures with hot dip galvanizing in accordance with the provisions
noted in the detail category tables or annexes.
(7) Microstructural damage from high temperature (> 150°C) that occurs during the design service
life is not covered.
(8) EN 1993-1-9 gives guidance of how to consider post-fabrication treatments that are intended to
improve the fatigue resistance of constructional details.

1.2 Assumptions

(1) Unless specifically stated, EN 1990, EN 1991 (all parts) and the other relevant parts of EN 1993-1
(all parts) apply.
(2) The design methods given in EN 1993-1-9 are applicable if:
• the execution quality is as specified in EN 1090-2, and

• the construction materials and products used are as specified in the relevant parts on EN 1993 (all
parts), or in the relevant material and product specifications.

(3) The design methods of EN 1993-1-9 are generally derived from fatigue tests on constructional
details with large scale specimens that include effects of geometrical and structural imperfections from
material production and execution (e.g. the effects of tolerances and residual stresses from welding).

9
prEN 1993-1-9:2023 (E)

2 Normative references
The following documents are referred to in the text in such a way that some or all of their content
constitutes requirements of this document. For dated references, only the edition cited applies. For
undated references, the latest edition of the referenced document (including any amendments) applies.
NOTE See the Bibliography for a list of other documents cited that are not normative references, including
those referenced as recommendations (i.e. through ‘should’ clauses) and permissions (i.e. through ‘may’ clauses).

EN 1090-2, Execution of steel structures and aluminium structures — Part 2: Technical requirements for
steel structures

EN 1990, Eurocode - Basis of structural design

EN 1991 (all parts), Eurocode 1 - Actions on structures

EN 1993-1 (all parts), Eurocode 3 - Design of steel structures

3 Terms, definitions and symbols

3.1 Terms and definitions

For the purpose of this document terms and definitions given in EN 1990, EN 1991 (all parts),
EN 1993-1-1, EN 1993-1-5, EN 1993-1-8 and the following apply.

3.1.1 General

3.1.1.1
fatigue
gradually progressive, localised damaging process of a constructional detail within a structure subject to
fatigue action (see 3.1.2.1) that may culminate in failure caused by crack initiation and propagation

Note 1 to entry: The type of failure depends on the definition of fatigue resistance, see 3.1.4.1.

3.1.1.2
design service life
reference period of time that depends on the type of structure for which its constructional details are
required to perform safely with an appropriate level of reliability that failure by fatigue cracking will not
occur

Note 1 to entry: EN 1990 gives provisions on design service life.

3.1.1.3
safe life concept
design concept in which an appropriate level of reliability for the fatigue design situation is obtained
without the need for regular in-service inspection or monitoring for fatigue during the design service life

3.1.1.4
damage tolerant concept
design concept in which an appropriate level of reliability for the fatigue design situation is obtained by
implementing prescribed inspection and maintenance for detecting and mitigating fatigue during the
design service life

10
 prEN 1993-1-9:2023 (E)

3.1.1.5
constructional detail
part of a member or joint containing a stress raising effect

3.1.1.6
hollow section joint
joint consisting of structural circular hollow sections (CHS) or structural rectangular hollow sections
(RHS), or their combinations as used in uniplanar or multi-planar trusses or girders, such as T-, Y-, X-, K-
, XX-, and KK-joints

3.1.1.7
rod
circular solid threaded member made of structural steel including stainless steel

3.1.1.8
stress raising effect
local increase in stress caused by discontinuity in loading and/or geometry and/or material

3.1.1.9
stress concentration
computable part of stress raising effect, expressed by the stress concentration factor kf, see Figure 3.1

Note 1 to entry: Stress concentration factors are usually only available for concentrated load effects and geometric
effects.

a) reference: b) concentrated c) macro-geometric d) macro-geometric effects

without stress load effect effects accounting accounting for eccentricity
concentration for large opening (maximum tensile stress shown)
(kf = 1)

Key

σ nominal stress
kf stress concentration factor
A cross-sectional area
F concentrated load

Figure 3.1 — Examples of stress concentration factor kf

11
prEN 1993-1-9:2023 (E)

3.1.1.10
concentrated load effect
stress raising effect arising from discontinuity in loading caused by single loads, usually not taken into
account in the detail category tables, e.g. Figure 3.1 b)

3.1.1.11
macro-geometric effect
stress raising effect arising from discontinuity in gross shape of a member, see e.g. Figure 3.1 c) and d),
usually not taken into account in the detail category tables

Note 1 to entry: Examples are apertures, re-entrant corners, large openings, shear lag, curved members, secondary
bending caused by eccentricities and misalignments beyond the limits accounted for by the detail category tables.

3.1.1.12
misalignment
unintended offset or out-of-straightness (angular mismatch) due to the arrangement or position of
jointed elements arising during the manufacturing process

3.1.1.13
eccentricity
intended offset of jointed elements

3.1.1.14
joint-geometric effect
stress raising effect arising from discontinuity in local shape of a member caused by attachments or
connected members, see Figure 3.2 c)

Note 1 to entry: Examples are shell bending stresses in addition to membrane stresses in plates caused by one-sided
attachment.

3.1.1.15
notch-geometric effect
stress raising effect arising from discontinuity in local geometry of a member at a microscopic scale
caused by notch geometry (notch radius), see Figure 3.2 d)

Note 1 to entry: Examples for non-welded member are scratches, corrosion pits and rolling defects. Examples for
welded members are weld profile shape, weld toes, weld roots, lack of fusion, slag inclusion, lack of penetration,
cold laps and porosity.

3.1.1.16
material effect
stress raising effect arising from discontinuity in material properties, such as regions with different yield
strengths in the heat affected zone of welds, that are accounted for within the detail category tables

12
 prEN 1993-1-9:2023 (E)

a) transverse weld toe b) nominal stress c) hot spot stress d) effective notch
distribution distribution stress distribution

Key
1 potential crack
2 stress distribution on surface accounting for weld with sharply edged weld toes
3 linear stress extrapolation
4 stress distribution on surface accounting for weld with rounded off weld toes
5 round off radius for weld toe
σ nominal stress at potential crack location (here: weld toe)
σHS hot spot stress at potential crack location (see 3.1.1.20)
σENS effective notch stress at potential crack location (see 3.1.1.22)

Figure 3.2 — Examples of different types of normal stress distribution in the vicinity of
transverse weld toe

3.1.1.17
nominal stress
σ or τ
elastic stress in a constructional detail adjacent to a potential crack location, disregarding any stress
raising effect, Figure 3.2 b)

Note 1 to entry: The nominal stress as specified in EN 1993-1-9 can be a normal stress, a shear stress, a principal
stress or an equivalent stress.

Note 2 to entry: The joint-geometric (see 3.1.1.14), the notch-geometric (see 3.1.1.15) and the material effects (see
3.1.1.16) are accounted for by the nominal stress-based detail categories. See 3.1.1.18 if macro-geometric and/or
concentrated load effects exist.

Note 3 to entry: For beam-like components with uniform loading, the nominal stress can be calculated by beam
theory.

3.1.1.18
modified nominal stress
nominal stress multiplied by an appropriate stress concentration factor kf to allow for geometric and/or
concentrated load effects, see Figures 3.1 b) to d)

Note 1 to the entry: Instead of stress concentration factors, fatigue notch factors kt can be used. Examples are given
in EN 1999-1-3.

13
prEN 1993-1-9:2023 (E)

3.1.1.19
geometric stress
structural stress
elastic stress within a welded constructional detail accounting for joint-geometric effects (and macro-
geometric and concentrated load effects if applicable) and neglecting the notch-geometric and material
effects, see Figure 3.2 c)

Note 1 to entry: The notch-geometric (see 3.1.1.13) and the material effects (see 3.1.1.14) are accounted for in Table
B.1.

3.1.1.20
hot spot stress
σHS
stress at the weld toe of the considered constructional detail derived from the geometric stress (see
3.1.1.18) through stress extrapolation

Note 1 to entry: See EN 1993-1-14 for determination of hot spot stress.

3.1.1.21
notch stress
elastic stress in a constructional detail taking into account all stress concentrations, Figure 3.2 d)

3.1.1.22
effective notch stress
σENS
peak value of notch stress at potential crack location modelled with a specified effective notch radius,
Figure 3.2 d)

Note 1 to entry: See EN 1993-1-14 for determination of effective notch stress.

3.1.1.23
residual stresses
permanent stresses in a member or structure in the absence of any external action

Note 1 to entry: Residual stresses can arise from rolling, cutting and forming processes, thermal treatment, weld
shrinkage or lack of fit between members. As external action is absent, the residual stresses locked in a member are
self-balancing.

3.1.1.24
inspection
examination for conformity by measuring, observing, or testing relevant characteristics

14
 prEN 1993-1-9:2023 (E)

3.1.2 Fatigue actions

3.1.2.1
fatigue action
action on a structure that is composed of loading events for which the number of reoccurrences cannot
be neglected for structural design as the action effect on the constructional details may cause fatigue

Note 1 to entry: Examples of fatigue actions are:

− axle loads of lorries on road bridges,

− transverse forces due to alternate vortex-shedding on masts, towers and chimneys,

− wheel loads of cranes on crane runway beams.

Note 2 to entry: EN 1990 gives representative values of the actions on structures for the fatigue design situation.

3.1.2.2
loading event (load cycle)
period of time with a defined variation in magnitude and/or point of application of the fatigue action that
can be considered to reoccur a number of times

Note 1 to entry: Examples of loading events are:

− sequence summarizing approach, passage and departure of a lorry or a railway train in case of bridges,

− shedding of a single vortex in case of masts, towers, chimneys,

− sequence of crane operations commencing when a payload is hoisted and ending when the crane is ready to
hoist the next payload in case of crane runway beams.

3.1.2.3
loading history
presentation of the expected fatigue action on a structure (considering prediction inaccuracy) during its
design service life by arranging the loading events in chronological sequence

15
prEN 1993-1-9:2023 (E)

3.1.2.4

action spectrum
evaluation of the loading history of a structure presenting the different levels of fatigue action with the
associated relative frequency (number of stress cycles) in descending order by neglecting sequence
effects

Note 1 to entry: The fatigue action can be described by ordinary spectra (relationship of different action levels Qi
and associated numbers of load cycles Ni) or by cumulative spectra (also called sum-spectra; relationship of
different action levels Qi and associated number of load cycles ΣNi for which this action level is reached or exceeded).
Figure 3.3 shows selected types of action spectra that are commonly used:

− discrete action spectrum of different action levels Qi, Figure 3.3 a)

− continuous loading spectrum characterized by Qmax, Nmax and a standardized spectrum shape, Figure 3.3 b)

− equivalent constant loading spectrum characterized by Qe and Nmax, Figure 3.3 c)

− equivalent constant loading spectrum characterized by Qe,2 and 2×106 load cycles representing a simplified
fatigue load model in EN 1991, Figure 3.3 d).

Note 2 to entry: EN 1990 contains general provisions for structures for which EN 1991 does not provide loading
spectra.

Note 3 to entry: The equivalent constant loading spectra characterized by Qe,2 and 2×106 load cycles replace real
discrete or continuous load spectra.

a) ordinary and corresponding cumulative discrete b) ordinary and corresponding cumulative

multiple action level spectra continuous multiple action level spectra

c) equivalent constant ‘single action level spectrum d) equivalent constant ‘single action level
spectrum with 2×106 load cycles
Key
Ni number of cycles at action level Qi
ΣNi number of cycles for which action level Qi is reached or exceeded
Nmax total number of cycles
Qi action level
Qe load level of damage equivalent constant action spectrum
Qe,2 load level of damage equivalent constant action spectrum with 2×106 load cycles
Qmax representative value of fatigue action spectrum
Figure 3.3 — Commonly used representations of action spectra

16
 prEN 1993-1-9:2023 (E)

3.1.3 Fatigue action effect

3.1.3.1
fatigue action effect
resulting stress effect from the application of the fatigue action on a constructional detail that is
composed of stress cycles

3.1.3.2
stress cycle
period of time (denoted ‘1’ in Figure 3.4) with a defined stress variation between a maximum and
minimum stress starting and ending at the same stress level

Key

σmin minimum stress

σmax maximum stress
∆σ stress range
t time
1 stress cycle
2 stress amplitude
3 mean stress

Figure 3.4 — Stress cycle parameters (also applicable for shear stress cycles)

3.1.3.3
stress history
presentation of the expected fatigue action effect by arranging the stress cycles in chronological sequence

3.1.3.4
stress range
difference between maximum and minimum stress of a stress cycle, see Figure 3.4

17
prEN 1993-1-9:2023 (E)

3.1.3.5
stress ratio
ratio of minimum and maximum stress of a particular stress cycle, with stresses calculated including the
static load effects from the relevant combination of actions

Note 1 to entry: The influence of stress ratio only exists for non-welded constructional details and for welded
constructional details with thermal stress relief or post-weld treatment as specified in Annex F.

3.1.3.6
thermal stress relief
reduction in residual stress as a result of thermal treatment (e.g. post weld heat treatment)

3.1.3.7
stress amplitude
half of stress range of a particular stress cycle, denoted ‘2’ in Figure 3.4

3.1.3.8
constant amplitude fatigue action effect
fatigue action effect where all stress cycles have the same stress range

3.1.3.9
variable amplitude fatigue action effect
fatigue action effect where the stress ranges vary between stress cycles

3.1.3.10
stress-range spectrum
evaluation of expected stress history presenting the different stress ranges and the associated relative
frequency (number of stress cycles) commonly presented in descending order neglecting sequence
effects through cycle counting methods, such as the rainflow and reservoir methods

a) discrete cumulative b) continuous cumulative c) equivalent constant d) equivalent constant

variable amplitude variable amplitude amplitude spectrum amplitude spectrum
spectrum spectrum with 2×106 stress cycles

Key
∆σi,E stress range level
∆σmax,E maximum stress range
Ni,E number of stress cycles at level ∆σi,E
ΣNi,E number of stress cycles for which ∆σi,E is reached or exceeded
Nmax,E total number of stress cycles
∆σe,E damage equivalent stress range
∆σe,2,E damage equivalent stress range with 2×106 stress cycles
Figure 3.5 — Commonly used representations of stress-range spectra

18
 prEN 1993-1-9:2023 (E)

Note 1 to entry: As for action spectra in Figure 3.3, the fatigue action effect can also be described by ordinary spectra
(relationship of applied stress ranges ∆σi,E or ∆τi,E and associated numbers of stress cycles Ni,E) or by cumulative
spectra (also called sum-spectra; relationship of applied stress ranges ∆σi,E or ∆τi,E and associated number of stress
cycles ΣNi,E for which this stress range is reached or exceeded). The following types of stress-range spectra are
commonly used:

− discrete spectrum of different stress range levels ∆σi,E or ∆τi,E and associated numbers of stress cycles Ni,E, see
Figure 3.5 a),

− continuous spectrum characterized by maximum stress range ∆σmax,E or ∆τmax,E, total number of stress cycles
Nmax,E and a standardized spectrum shape, see Figure 3.5b),

− equivalent constant stress range spectrum with ∆σe,E or ∆τe,E and total number of stress cycles Nmax,E, see Figure
3.5 c),

− equivalent constant stress range spectrum with ∆σe,2,E or ∆τe,2,E and 2×106 stress cycles, see Figure 3.5 d).

Note 2 to entry: The equivalent constant stress range spectrum with ∆σe,2,E and 2×106 stress cycles is the effect of a
simplified fatigue load model in EN 1991 representing an equivalent loading spectrum with Qe,2 and 2×106 load
cycles (see 3.1.2.4). The fatigue load model causes a stress range that is adjusted by λ-values to obtain ∆σe,2,E. The
λ-values account for the influence of the real loading spectrum and are dependent on the type of structure (bridge,
mast, crane runway etc.) and the kind of stress.

Note 3 to entry: For some applications, a load cycle can cause multiple stress cycles for the considered
constructional detail. The number of stress cycles (NE) rather than the number of load cycles (N) is always decisive
for the fatigue design situation.

3.1.3.11
cycle counting
process of transforming a variable amplitude stress history into a stress range spectrum, each with a
particular stress range, such as the rainflow method and reservoir method

3.1.3.12
rainflow method
particular cycle counting method of producing a stress-range spectrum from a given stress history

Note 1 to entry: For the mathematical determination see prEN 1990:2021 Annex F.

3.1.3.13
reservoir method
particular cycle counting method of producing a stress-range spectrum from a given stress history

Note 1 to entry: For the mathematical determination see prEN 1990:2021 Annex F.

3.1.4 Fatigue resistance

3.1.4.1
fatigue resistance
capacity of a constructional detail to withstand fatigue actions without fatigue failure, with an
appropriate level of reliability

19
prEN 1993-1-9:2023 (E)

3.1.4.2
fatigue life
number of cycles that a constructional detail can withstand without fatigue failure

Note 1 to entry: The fatigue life depends on the stress range spectrum. For a particular structure, the fatigue life can
alternatively be expressed as a period of time.

3.1.4.3
fatigue resistance curve
calculation model for the fatigue resistance of a constructional detail that establishes a quantitative
relationship between stress range Δσ and the corresponding endurance NR, see Figure 3.6 a)

a) Endurance of any stress range b) Reference value at 2×106 stress cycles

Key
a fatigue resistance curve
m slope parameter of fatigue resistance curve
∆σ stress range
N number of cycles
NR endurance associated with ∆σ
∆σC reference value at 2×106 stress cycles

Figure 3.6 — Definition of endurance and reference value of fatigue resistance curve
(also applicable for shear stressing)

Note 1 to entry: The characteristic fatigue resistance curves given in EN 1993-1-9 are lower bound values based on
the evaluation of fatigue tests with large scale test specimens.

Note 2 to entry: The characteristic fatigue resistance curves are also known as S-N curves or Wöhler curves.

Note 3 to entry: For simplification, the characteristic fatigue resistance curves of the constructional details in this
document are attributed to detail categories.

20
 prEN 1993-1-9:2023 (E)

3.1.4.4
endurance
NR
fatigue life of a constructional detail under a stress range ∆σ or ∆τ that is obtained from the
corresponding fatigue resistance curve, see Figure 3.6 a)

3.1.4.5
characteristic reference value of fatigue resistance
∆σC, ∆τC
mechanical property of a constructional detail expressing its characteristic fatigue resistance in terms of
a stress range ∆σC or ∆τC for an endurance of NR = NC = 2×106 stress cycles, see Figure 3.6 b)

Note 1 to entry: ‘Reference value of fatigue resistance’ is also commonly referred to as ‘fatigue strength’.

3.1.4.6
detail category
characteristic reference value ∆σC or ∆τC of fatigue resistance in N/mm2 identifying a particular fatigue
resistance curve

Note 1 to entry: The detail category attributed to a particular constructional detail can be obtained from the first
row of the detail category tables depending on the chosen design stress method (see 6.1).

3.1.4.7
slope parameter
m
parameter of the fatigue resistance curve expressing the intensity of fatigue damage, see Figure 3.6 a)

Note 1 to entry: In mathematics, the slope parameter is identical with the negative inverse slope of the fatigue
resistance curve.

3.1.4.8
constant amplitude fatigue limit
CAFL
∆σD, ∆τD
characteristic maximum stress range, for constant or variable stress range spectra, for which a
constructional detail can withstand an infinite number of stress cycles without fatigue damage, see Figure
3.7

Note 1 to entry: Figure 3.7 is a simplification. The verification with respect to the CAFL, that takes partial factors
into account, is described in 9.3.

Note 2 to entry: The verification with respect to the CAFL for design situations with variable amplitude fatigue
action effect is only possible where the real maximum stress range is known and stays below ∆σD or ∆τD, Figure 3.7
b). The fatigue load models in EN 1991 do not provide the real maximum stress range. See details in 9.3.

21
prEN 1993-1-9:2023 (E)

a) constant amplitude fatigue action effect b) variable amplitude fatigue action effect with
known maximum stress range

Key
1 stress range spectrum not contributing to fatigue damage
a fatigue resistance curve
∆σD CAFL
ND number of stress cycles linked with the CAFL

Figure 3.7 — Spectra with all stress ranges below CAFL (also applicable for shear stressing)

3.1.4.9
variable amplitude fatigue limit
VAFL
∆σL
characteristic maximum stress range, for variable amplitude stress range spectra, below which the stress
ranges do not need to be considered in calculating the fatigue damage
Note 1 to entry: For constructional details with a fatigue resistance curve having only a slope parameter m1, the
value of the CAFL is the same as that of the VAFL. These are:

− lattice girder joints made of hollow sections subject to normal stress ranges,

− all constructional details subject to shear stress ranges.

Thus, all stress ranges below ∆σD or ∆τD are not considered in the fatigue damage calculation for constructional
details with such fatigue resistance curves. Figure 3.8 is a simplification neglecting partial factors.

Note 2 to entry: For constructional details with a fatigue resistance curve having slope parameters m1 and m2, the
stress ranges between ∆σD and ∆σL contribute to fatigue damage, if the stress range spectrum also comprises stress
ranges above ∆σD.

22
 prEN 1993-1-9:2023 (E)

a) Fatigue resistance curve with same CAFL and b) Fatigue resistance curve with separate CAFL
VAFL levels and VAFL levels

Key
1 stress ranges not contributing to fatigue damage
2 stress ranges contributing to fatigue damage with slope parameter m2
3 stress ranges contributing to fatigue damage with slope parameter m1
a fatigue resistance curve
∆σD CAFL for details subject to normal stresses
∆τD CAFL for details subject to shear stresses
∆σL VAFL for details subject to normal stresses and two-slope fatigue resistance curve
ND number of stress cycles linked with the CAFL
NL number of stress cycles linked with the VAFL

Figure 3.8 — Spectra with stress ranges above and below CAFL

3.1.4.10
manual welding
welding in which all operations are carried out manually (the electrode holder, gun, torch or blowpipe is
manipulated by hand)

3.1.4.11
fully mechanized welding
welding in which all main operations (excluding the handling of the work piece) are performed
automatically

Note 1 to entry: Manual adjustment of welding variables during welding is possible.

3.1.4.12
automatic welding
welding in which all operations are performed automatically

Note 1 to entry: Manual adjustment of welding variables during welding is not possible.

23
prEN 1993-1-9:2023 (E)

3.1.4.13
High Frequency Mechanical Impact (HFMI) treatment
continuous peening process carried out by traversing a hardened rounded tipped tool along a surface
whilst it is subjected to high frequency hammer blows produced by machine

Note 1 to entry: Qualification involves application of a pre-approved procedure to a specific welded joint geometry,
which is then subjected to fatigue testing.

Note 2 to entry: Investigation in [1] showed that the following HFMI technologies are comparable:

− HiFIT (High Frequency Impact Treatment)

− PIT (Pneumatic Impact Treatment)

− UIT (Ultrasonic Impact Treatment).

3.1.4.14
operator
appropriately trained person who is able to operate the HFMI-device properly

3.1.4.15
device manufacturer
company that manufactures the HFMI-device

3.1.5 Fatigue verification

3.1.5.1
fatigue damage
Di
fatigue damage due to a stress range Δσi,E with Ni,E stress cycles (Di = Ni,E / Ni,R), the ratio of the number of
cycles, Ni,E, of a particular stress range applied to a constructional detail and the endurance, Ni,R, of this
constructional detail under the same stress range

3.1.5.2
accumulated fatigue damage
D
sum of fatigue damages from all stress ranges in a spectrum

3.1.5.3
Miner's summation
linear cumulative summation of the fatigue damage from all stress cycles accounted for in a stress-range
spectrum, based on the Palmgren-Miner rule

24
 prEN 1993-1-9:2023 (E)

3.2 Symbols
Latin upper-case letters

A cross-sectional area
As stress area of the bolt or rod
CAFL Constant Amplitude Fatigue Limit
CHS Circular Hollow Section
D accumulated fatigue damage due to different stress ranges (D = Σ Di)
Di fatigue damage due to a stress range Δσi,E with Ni,E stress cycles (Di = Ni,E / Ni,R)
EVM Equivalent von Mises
F concentrated load
FB force acting on the bolt or rod
FC reduction of the compressive force of the clamped components due to FT
FT external tensile force applied to the preloaded bolted joint or joint with rod
FEM Finite Element Method
FLM Fatigue Load Model
HFMI High Frequency Mechanical Impact treated state
I second moment of area
N number of cycles
NC number of stress cycles associated with the characteristic reference value of fatigue
resistance
ND number of stress cycles associated with the characteristic constant amplitude fatigue limit
NDT Non Destructive Testing
NDP Nationally Determined Parameter
Ni number of load cycles of an applied fatigue action level Qi
Ni,E number of stress cycles of an applied stress range Δσi,E
Ni,R number of stress cycles which can be resisted by a constructional detail under an applied
stress range Δσi,E until failure (endurance)
NL number of stress cycles associated with the variable amplitude fatigue limit
Nmax total number of load cycles
Nmax,E total number of stress cycles
NR endurance
NRd design value of endurance
PS principal stress
Qfat fatigue action
Qe equivalent action applied for Nmax load cycles
Qe,2 equivalent action applied for 2×106 load cycles

25
prEN 1993-1-9:2023 (E)

Qi action value applied for Ni load cycles

Qmax maximum action value
R stress ratio (R = σmin/σmax)
RHS Rectangular Hollow Section
S(t) first moment of area
V shear force in a cross section
VAFL Variable Amplitude Fatigue Limit
VB shear force per shear plane acting on the bolt or rod
VM Von Mises

Latin lower-case letters

aw as-welded
k1 magnification factor for nominal stress ranges to account for secondary bending moments
in trusses
kf stress concentration factor
ks reduction factor for fatigue resistance (in terms of stress) to account for size effects
m slope parameter of a fatigue resistance curve
maw slope parameter of fatigue resistance curve for as-welded state
mHFMI slope parameter of fatigue resistance curve for HFMI state
t plate thickness

Greek upper-case letters

ΔσC, ΔτC characteristic reference value of fatigue resistance at NC = 2×106 stress cycles
ΔσC,HFMI characteristic reference value of the fatigue resistance at NC = 2×106 stress cycles
for HFMI-treated state
ΔσC,red reduced characteristic reference value of the fatigue resistance
ΔσD, ΔτD characteristic constant amplitude fatigue limit at ND stress cycles
ΔσE, ΔτE fatigue action effect (stress ranges)
Δσe,2,E, Δτe,2,E equivalent stress range applied 2×106 stress cycles
Δσe,E, Δτe,E equivalent stress range applied Nmax,E stress cycles
Δσe,2,HFMI,E equivalent stress range for HFMI treated joints, i.e. accounting for λHFMI
Δσeq,E equivalent stress range for connections in webs of orthotropic decks
Δσi,E, Δτi,E stress range level applied Ni,E stress cycles
ΔσL characteristic variable amplitude fatigue limit at NL stress cycles
∆σmax,Ed, correspond to max∆σi,Ed and max∆τi,Ed
∆τmax,Ed

26
 prEN 1993-1-9:2023 (E)

Δσs characteristic stress range at intersection of fatigue resistance curves for as-
welded and HFMI state
ΣNi accumulated number of load cycles for which a fatigue action level Qi is reached
or exceeded
ΣNi,E accumulated number of stress cycles for which an applied stress range Δσi,E or Δτi,E
is reached or exceeded

Greek lower-case letters

γFf partial factor for applied stress ranges ΔσE, ΔτE

γMf partial factor for fatigue resistance
λHFMI supplementary damage equivalent factor to consider the mean stress effect in spectrum in
case of HFMI treatment
λi damage equivalent factors depending on, for example, the load situation and the structural
characteristics
σ nominal normal stress
σENS effective notch stress
σHS hot spot stress
τ nominal shear stress

4 Basis of fatigue design

(1) The fatigue design of steel structures shall be in accordance with the general rules given in
EN 1990 and EN 1991 (all parts) and the specific design provisions for steel structures given in the other
relevant parts of EN 1993-1 (all parts).
NOTE EN 1991 (all parts) gives provisions on fatigue actions for particular types of structures such as bridges,
masts, towers, chimneys and crane supporting structures that meet these requirements. See definitions in 3.1.2.

(2) Steel structures designed according to this document shall be executed according to EN 1090-2.
(3) The methods for the verification of the fatigue design situation given in Clause 6 should follow the
principle of verification by the partial factor method by comparing fatigue action effects and compatible
fatigue resistances.
(4) Fatigue cracks should be either monitored in a specific inspection regime or repaired with
particular care to avoid introducing more severe stress raising effects.
NOTE The action to undertake depends on the consequence of failure, see Clause 5.

(5) Fatigue tests on specimens may be carried out to determine:

• the fatigue resistance for constructional details not included in this document,

• the fatigue life of prototypes.

NOTE Prototype testing considering the design loading spectrum, in connection with a statistical evaluation of
the test results according to prEN 1990:2021, Annex D, can be used to determine the characteristic fatigue life of
the tested component. The actual characteristic fatigue resistance (ΔσC) of the component’s constructional details
remains unknown.

27
prEN 1993-1-9:2023 (E)

(6) The statistical evaluation of fatigue tests should be in accordance with EN 1990.
NOTE 1 Note 2 to 8.2(1) contains detailed information on evaluation of fatigue tests.

NOTE 2 The National annex can give requirements for determining fatigue resistance from tests.

5 Fatigue design concepts

(1) The fatigue design situation should be verified using either:
a) safe life concept or
b) damage tolerant concept.
(2) The safe life concept should provide an appropriate level of reliability that the performance of a
structure will not be impaired throughout its design service life without the need for regular in-service
inspection or monitoring for fatigue cracks.
(3) The safe life concept should be applied in cases where local formation of cracks in one component
could rapidly lead to failure of a structural element or structure.
(4) The damage tolerant concept should provide an appropriate level of reliability that the
performance of a structure will not be impaired throughout its design service life, provided that readily
inspectable details are used and prescribed inspection and maintenance, for detecting and preventing
fatigue cracks, is implemented.
NOTE The National annex can give provisions for inspection and maintenance. The inspection intervals and
methods can be based on crack growth calculations, on experience, monitoring, or previous inspection results and
can be a function of the consequence of failure. Maintenance can include component replacements at given intervals.
It can further give provisions for changes in inspection intervals and methods in case of monitoring or after
repairing detected cracks.

(5) Constructional details, materials and stress levels should be selected for the appropriate design
concept to ensure an appropriate level of reliability in accordance with EN 1990, at least equal to the
reliability required for ultimate limit state verification, is achieved for:
a) safe-life concept
• at the end of the design service life;

b) damage tolerant concept

• at the end of each in-service inspection interval, so that in the event of the formation of cracks one
or all of the following occurs:

• propagation rates are low and cracks are easily detectable prior to failure,
• multiple load paths exist,
• crack-arresting constructional details prevent progressive damage.

NOTE The verification of the fatigue design situation includes long-term uncertainties, related to the fatigue
resistance and actions, such that the possibility for fatigue cracking cannot be completely excluded but only reduced
to an acceptable likelihood of occurrence.

(6) For the purpose of verification of fatigue design situation using this document, an appropriate
level of reliability should be achieved by adjustment of the partial factor for fatigue resistance, γMf, taking
into account the consequences of failure and the design concept used.
NOTE 1 The values of the partial factors γMf are given in Table 5.1 (NDP) unless the National annex gives different
values.

28
 prEN 1993-1-9:2023 (E)

NOTE 2 The partial factors in Table 5.1 (NDP) for low consequences satisfy the reliability requirement for class
of consequences CC1 according to EN 1990, those for medium consequence satisfy CC2, and those for high
consequence satisfy CC3.

NOTE 3 The partial factors in Table 5.1 (NDP) are based on the possibility to perform visual inspections.

Table 5.1 (NDP) — Recommended values of the partial factors for fatigue resistance γMf

Consequence of failure
Design concept
Low consequence Medium consequence High consequence
Safe life 1,15 1,25 1,35
Damage tolerant 1,00 1,15 1,25

6 Fatigue design methods

6.1 Design stress methods

(1) All stress raising effects that are relevant to the fatigue design situation shall be considered in the
verification of the fatigue design situation independent of the chosen design stress method.
NOTE Design stress methods are the nominal stress method, hot spot-stress method and effective notch stress
method.

(2) Where stress raising effects are not included within the fatigue action effect according to a given
design stress method, the stress raising effects should be accounted for within the fatigue resistance.
NOTE The different kinds of stress raising effects are explained in 3.1.1.7.

(3) The requirements expressed in (1) and (2) are satisfied by the fatigue design situation verification
according to Clause 9, where the fatigue action effects are based on nominal or modified nominal stresses
according to Clause 7, and the fatigue resistances according to Clause 8 are used.
NOTE 1 Annex B gives provisions on the determination of fatigue action effects and corresponding fatigue
resistances according to the hot spot stress method unless the National annex specifies differently.

NOTE 2 Annex C gives provisions on the determination of fatigue action effects and corresponding fatigue
resistances according to the effective notch stress method unless the National annex specifies differently.

NOTE 3 Annex F gives guidelines for the determination of fatigue action effects and corresponding fatigue
resistance of constructional details treated by HFMI unless the National annex specifies differently.

6.2 Verification methods

(1) The method of verification shall be based upon calculation models that are appropriate for the
fatigue design situation.
(2) For fatigue design situations where the fatigue action effect is defined by an equivalent constant
stress range spectrum, with ∆σe,2,Ed and 2×106 stress cycles, the verification method according to Clause 9
should be used.
NOTE See 3.1.3.9 for equivalent constant stress range spectrum. An equivalent constant stress range spectrum
is usually based on a simplified fatigue load model of EN 1991 with appropriate λi-values.

(3) For fatigue design situations with any other kind of stress range spectrum, the verification
method of Annex A should be used.

29
prEN 1993-1-9:2023 (E)

NOTE Examples of such spectra are spectra with constant stress range but more, or less, than 2×106 stress
cycles, see Figure 3.5 a) and b).

7 Fatigue action effect

7.1 Calculation of nominal stresses

(1) The nominal stresses from fatigue actions should be calculated using linear elastic analysis
according to EN 1993-1-1.
NOTE EN 1993-1-1:2022 includes rules for:

− section properties of net area in 8.2.2.2,

− shear lag effects in 8.2.2.3,

− elastic stresses due to shear in 8.2.6(4),

− elastic stresses due to torsion in 8.2.7(3).

(2) For constructional details of steel members in composite structures, the nominal stresses should
be calculated considering all effects caused by other materials in the fatigue design situation.
NOTE For steel concrete composite structures, concrete cracking under normal service conditions can affect
the nominal stresses of the steel members. In these cases, actions such as the permanent actions, variable actions,
shrinkage, or creep, can have an influence on the nominal stresses.

(3) Modified nominal stresses should be used for the verification where stress concentrations due to
macro-geometric effects and/or concentrated load effects occur close to the potential crack location
shown in the illustrations of Table 10.1 to Table 10.12 and where such effects are not included in the
detail category of the considered detail according to the detail description of the tables.
(4) Class 4 cross sections may be treated as Class 3 cross sections for fatigue if the width to thickness
ratios are less than the limiting proportions for Class 3 obtained from EN 1993-1-1 when ε is multiplied
by �𝑓𝑓y,k ∙ 𝛾𝛾m0 / 𝜎𝜎com,f , where σcom,f is the maximum design compressive stress in the part using the load
combination for fatigue of EN 1990.
(5) The nominal stresses in cross sections classified as Class 4 for fatigue should be divided by the
reduction factors for buckling as for the reduced stress method of EN 1993-1-5, where the plate
slenderness may be determined with the maximum design compressive stress in the element determined
using the effective area of the section caused by all simultaneous actions σcom,f.
NOTE The National annex can give limitations for class 4 sections subject to fatigue actions.

7.2 Relevant nominal stresses

(1) All nominal stresses, relevant for the constructional detail, shall be accounted for in the
verification of the fatigue design situation.
(2) For each relevant nominal stress, the corresponding detail category should be determined from
Table 10.1 to Table 10.12 accounting for the constructional detail, the type of stress (σ or τ) and the stress
component. The supplementary requirements of the detail category on the stress calculation should be
considered in addition to the provisions of 7.1.
NOTE For example, the detail category can require the stress calculation at a particular location of the
constructional detail or can forbid simplifications that are allowed in EN 1993-1-1.

30
 prEN 1993-1-9:2023 (E)

(3) For unwelded constructional details and for welded constructional details with weld toe failure,
the relevant nominal stresses should be:
• nominal normal stresses σ in the parent metal, and

• nominal shear stresses τ in the parent metal.

(4) For weld root failure of load-carrying partial penetration or fillet welds, the forces transmitted by
a unit length of weld should be resolved into components transverse and parallel to the longitudinal axis
of the weld.
(5) The relevant nominal stresses in the weld throat should be taken as, see Figure 7.1:
• normal stresses σwf transverse to the axis of the weld, and

• shear stresses τwf longitudinal to the axis of the weld.

Figure 7.1 — Relevant stresses in (a) fillet welds and (b) partial penetration welds

(6) The nominal stresses σwf and τwf should be obtained by dividing the relevant component of the
force transmitted per unit length of weld by the throat size a, see Figure 7.1.
(7) The stress components σ⊥ and τ⊥ on the throat section (see Figure 7.2), determined by the
directional approach according to EN 1993-1-8 due to a force F transverse to the axis of the weld, may be
transferred into the nominal stress component σwf of the weld by:
𝜎𝜎wf = �𝜎𝜎⊥2 + 𝜏𝜏⊥2 (7.1)

31
prEN 1993-1-9:2023 (E)

Figure 7.2 — Stresses on the throat section of fillet welds according to EN 1993-1-8

(8) The stress component τ|| on the throat section, determined by the directional approach according
to EN 1993-1-8 due to a shear force parallel to the axis of the weld, should be the nominal shear stress
component τwf.
(9) The stress component σ|| parallel to the axis of the weld may be neglected for weld root failure.
(10) The relevant weld stress in double fillet welds due to transverse bending of the plate may be
calculated using:
𝑀𝑀
𝜎𝜎wf = (7.2)
𝑎𝑎 𝐿𝐿𝑤𝑤 (𝑎𝑎+𝑡𝑡)

where:
t is the plate thickness under bending
a is the throat size of the weld
M is the bending moment
Lw is the weld length according to Figure 7.3

Figure 7.3 — Relevant stresses in double fillet welds due to transverse bending

32
 prEN 1993-1-9:2023 (E)

(11) For bolted joints or joints with rods with rolled or cut threads, the relevant nominal stresses
should be:
𝐹𝐹𝐵𝐵
• nominal normal stresses 𝜎𝜎 = 𝐴𝐴𝑠𝑠
(7.3)

𝑉𝑉𝐵𝐵
• nominal shear stress on shank 𝜏𝜏 = 𝐴𝐴
(7.4)

where:
A is the area of the shank of the bolt or rod
As is the stress area of the bolt or rod
FB is the force acting on the bolt or rod
VB is the shear force per shear plane acting on the bolt or rod
NOTE For joints with preloaded bolts or rods:

− FB can be determined according to Annex E,

− VB can be disregarded.

(12) Prying and/or bending effects in bolted joints or joints with rods should be accounted for when
determining the force FB. Effects due to imperfections in the bolt or rod may be neglected.
(13) Prying and/or bending effects on the stress in the bolt may be neglected in simple and symmetric
joints with preloaded bolts such as the T-stub in prEN 1993-1-8:2021, Table 6.2, where plates are either
in full contact and have sufficient stiffness, or are without tip contact after assembly.

7.3 Calculation of nominal stress ranges

7.3.1 General

(1) Nominal stress ranges in relevant parts of constructional details should be determined at the
locations indicated in the details shown in Table 10.1 to Table 10.12.
(2) Effective nominal stress ranges may be considered in the verification of non-welded or thermally
stress-relieved welded constructional details in partial or total compression or in cases with partial or
total compression confirmed by modelling or measurements.
NOTE See 7.4 for effective nominal stress ranges.

7.3.2 Design value of nominal stress range

(1) For fatigue design situations where the fatigue action effect is defined by an equivalent constant
stress range, the design value of the nominal stress ranges Δσe,2,Ed and Δτe,2,Ed should be determined as
follows:

Δσe,2,Ed = λ1 λ2 ... λn ΔσE(γFf Qfat)

(7.5)
Δτ,e,2,Ed = λ1 λ2 ... λn ΔτE(γFf Qfat)

where
λi are damage equivalent factors depending on, for example, the load situation and the structural
characteristics

33
prEN 1993-1-9:2023 (E)

(2) For fatigue design situations where the fatigue action effect is defined by an equivalent constant
stress range, the design value of the modified nominal stress ranges Δσe,2,Ed and Δτe,2,Ed should be
determined as follows:

Δσe,2,Ed = kf λ1 λ2 ... λn ΔσE(γFf Qfat)

(7.6)
Δτe,2,Ed = kf λ1 λ2 ... λn ΔτE(γFf Qfat)

where
kf is the stress concentration factor to take account of concentrated load and/or macro-geometric
effects not included in the detail category
NOTE 1 See Figure 3.1 for examples of stress concentrations.

NOTE 2 Stress concentration factors kf can be obtained from handbooks or from finite element calculations.
Annex D gives stress concentration factors kf for specific cases.

(3) For welded joints of hollow sections in lattice girders the design value of the modified nominal
stress range Δσe,2,Ed should be determined as follows using the simplified model in Annex D.3 unless more
accurate calculations are carried out:
Δσe,2,Ed = k1 Δσ**e,2,Ed (7.7)

where:
k1 is the stress concentration factor according to Table D.1 and Table D.2
Δσ**e,2,Ed is the design value of the stress range caused by the simplified fatigue load model specified
in EN 1991 calculated for a simplified truss model with continuous chords and pinned
braces
(4) For load models other than those resulting in an equivalent constant stress range, the design
value of the stress ranges Δσi,E and Δτi,E should be determined using Annex A.

7.4 Effective design value of stress range

(1) In non-welded details, or welded details which have been subjected to full thermal stress relief
after all welding operations have been completed, the mean stress level influence on the fatigue life may
be taken into account by considering a reduced design value of each stress cycle in the design spectrum
where part, or all, of the stress cycle is compressive.
(2) The mean stress level according to (1) should be confirmed by modelling or measurements
including the serviceability stresses due to the permanent action and the most tensile stress resulting
from the fatigue action.
(3) The effective design value of the stress range ∆σ may be calculated by adding the tensile portion
of the stress range (if any) to 60% of the magnitude of the compressive portion of the stress range (if any)
as follows, see Figure 7.4:
• if the entire stress range is in tension as in Figure 7.4 a): Δσ = σmax – σmin
• if a part of the stress range is in compression as in Figure 7.4 b): Δσ = σmax – 0,6 σmin (7.9)
• if the entire stress range is in compression as in Figure 7.4 c): Δσ = 0,6 σmax – 0,6 σmin

34
 prEN 1993-1-9:2023 (E)

a) fully tensile b) partly compressive c) fully compressive

NOTE Tension stresses are positive.

Figure 7.4 — Effective design value of stress range for non-welded or stress relieved details

8 Fatigue resistance
8.1 Fatigue resistance curves

(1) The fatigue resistance of constructional details should be calculated based on applicable fatigue
resistance curves.
NOTE 1 The characteristic fatigue resistance is represented by a series of parallel and equally spaced fatigue
resistance curves defining the number of cycles, NR (endurance), that a stress range Δσ or Δτ can withstand until
failure occurs, as shown in Figures 8.1 to 8.4.

NOTE 2 The fatigue resistance curves are attributed to detail categories. Each detail category is designated by a
number which represents the characteristic reference value ΔσC or ΔτC in N/mm2 for the fatigue resistance at
NC = 2×106 cycles.

NOTE 3 For constant amplitude loading, the fatigue resistance curves consist of a finite life region expressed by
a linear relationship with slope parameter m1 and an infinite life region expressed by a constant relationship on
logarithmic scales. The transition between both regions is defined by the constant amplitude fatigue limit ΔσD or
ΔτD.

NOTE 4 For variable amplitude loading under normal stresses, the finite life region of the fatigue resistance
curves extends beyond ΔσD to the variable amplitude fatigue limit ΔσL by a linear relationship with slope parameter
m2 on logarithmic scales.

(2) The fatigue resistance curves of Figure 8.1a should be applied for non-welded constructional
details with light notch effect subject to Δσ including:
• plain members free from welding in Table 10.1;

• bolted details in Table 10.2 with holes drilled or reamed.

NOTE See also Table 8.1.

35
prEN 1993-1-9:2023 (E)

(3) The fatigue resistance curves of Figure 8.1b should be applied for non-welded constructional
details with sharp notch effect subject to Δσ including:
• bolted details in Table 10.2 with holes thermally cut and punched without subsequent reaming;

• bolt detail ⑥ in Table 10.2.

NOTE See also Table 8.1.

(4) The fatigue resistance curves of Figure 8.2 should be applied for welded constructional details
with sharp notch effect subject to Δσ including Table 10.3 to Table 10.11 except for Table 10.8.
NOTE See also Table 8.1.

(5) The fatigue resistance curves of Figure 8.3 should be applied for hollow section joint details in
Table 10.8 subject to Δσ.
NOTE See also Table 8.1.

(6) The fatigue resistance curves of Figure 8.4 should be applied for constructional details subject to
Δτ including:
• non-welded details ⑥, ⑦, ⑧ in Table 10.1;

• bolt detail ⑦ in Table 10.2;

• welded constructional detail ⑩ in Table 10.6.

NOTE See also Table 8.2.

36
 prEN 1993-1-9:2023 (E)

Nominal stress range ∆σ [N/mm²]

NL=108
∆σL=0,647∆σC

Number of cycles N
(a) light notch effect
Nominal stress range ∆σ [N/mm²]

NL=108
∆σL=0,457∆σC

Number of cycles N
(b) sharp notch effect
Key
a – Fatigue resistance curve for constant amplitude loading
b – Extended fatigue resistance curve for variable amplitude loading
c – Detail category
Figure 8.1 — Characteristic fatigue resistance curves of non-welded constructional details
subject to nominal normal stress ranges

37
prEN 1993-1-9:2023 (E)

Nominal stress range ∆σ [N/mm²]

ND=5×106
∆σD=0,737∆σC
NL=108
∆σL=0,405∆σC

Number of cycles N
(a) Detail category 71 and above
Nominal stress range ∆σ [N/mm²]

ND=107
∆σD=0,585∆σC
NL=108
∆σL=0,369∆σC

Number of cycles N
(b) Detail category below 71
Key
a – Fatigue resistance curve for constant amplitude loading
b – Extended fatigue resistance curve for variable amplitude loading
c – Detail category
NOTE See explanations and symbols in Figure 8.1.
Figure 8.2 — Characteristic fatigue resistance curves of welded constructional details with sharp
notch effect subject to nominal normal stress ranges

38
 prEN 1993-1-9:2023 (E)

Nominal stress range ∆σ [N/mm²]

ND=107
∆σD=0,725∆σC
NL=108
∆σL=0,561∆σC

Number of cycles N
NOTE See explanations and symbols in Figure 8.1.

Figure 8.3 — Characteristic fatigue resistance curves of lattice girder joints made of hollow
sections according to Table 10.8 subject to nominal normal stress ranges
Nominal stress range ∆τ [N/mm²]

ND=107
∆𝜏𝜏D=0,457∆𝜏𝜏C

Number of cycles N
NOTE See explanations and symbols in Figure 8.1.

Figure 8.4 — Characteristic fatigue resistance curves of constructional details subject to nominal
shear stress ranges

39
prEN 1993-1-9:2023 (E)

Table 8.1 — Characteristic fatigue resistance values in N/mm2 from curves in Figures 8.1 to 8.3

Figure Detail Slope parameter ∆σD ∆σL

category [N/mm2] [N/mm2]
m1 m2

8.1a 180 5 9 180 117

160 160 104
140 140 90,6
125 125 80,9
112 112 72,5
100 100 64,7
90 90,0 58,2
80 80,0 51,8
8.1b 112 3 5 112 51,2
100 100 45,7
90 90,0 41,1
80 80,0 36,6
71 71,0 32,4
63 63,0 28,8
56 56,0 25,6
50 50,0 22,9
8.2a 160 3 5 118 64,8
140 103 56,7
125 92,1 50,6
112 82,5 45,4
100 73,7 40,5
90 66,3 36,5
80 59,0 32,4
71 52,3 28,8
8.2b 63 3 5 36,9 23,2
56 32,8 20,7
50 29,3 18,5
45 26,3 16,6
40 23,4 14,8
36 21,1 13,3
8.3 90 5 9 65,3 50,5
71 51,5 39,8
56 40,6 31,4
50 36,3 28,1
45 32,6 25,2
36 26,1 20,2

40
 prEN 1993-1-9:2023 (E)

Table 8.2 — Characteristic fatigue resistance values in N/mm2 from curves in Figure 8.4

Figure Detail Slope parameter ∆τD ∆τL

category [N/mm2] [N/mm2]
8.4 100 5 ꟷ 45,7 45,7
80 36,6 36,6

8.2 Classification of constructional details

(1) For the nominal stress method, the detail categories ΔσC and ΔτC should be used in accordance
with Clause 10:
• Table 10.1 for plain members free from welding,

• Table 10.2 for mechanically fastened joints,

• Table 10.3 for welded built-up sections and longitudinal welds,

• Table 10.4 for transverse butt welds (full penetration),

• Table 10.5 for weld attachments and stiffeners,

• Table 10.6 for load carrying welded joints,

• Table 10.7 for hollow section joints,

• Table 10.8 for lattice girder joints made of hollow sections,

• Table 10.9 for orthotropic decks – closed stiffeners,

• Table 10.10 for orthotropic decks – open stiffeners,

• Table 10.11 for crane runway beams,

• Table 10.12 for chimneys, masts and towers.

NOTE 1 Tables 10.1 to 10.12 describe the constructional details together with the stress distribution in the
vicinity of the site of the potential crack location that corresponds with the detail category.

NOTE 2 Where test data has been used to determine the detail category for a particular constructional detail, the
stress range ΔσC or ΔτC corresponding to NC = 2×106 cycles has been calculated using a lower prediction bound of
survival for log N with 95% probability, taking into account the standard deviation and the sample size. The number
of data points (not lower than 12) has been considered in the statistical analysis. Accounting for pre-existing
information on the slope parameter of the fatigue resistance curves, the aforementioned approach to determine the
stress range ΔσC or ΔτC is identical with prEN 1990:2021, Annex D in respect of the statistical determination of a
single property.

NOTE 3 The test data used to determine detail categories includes detail specific properties such as residual
stress effects.

NOTE 4 The National annex can permit the determination of a more appropriate detail category for a particular
application based on tests according to 4(5) and (6) provided that the test data is statistically evaluated in
accordance with NOTE 2.

41
prEN 1993-1-9:2023 (E)

NOTE 5 The National annex can give detail categories ΔσC or ΔτC for details not classified by Table 10.1 to Table
10.12.

(2) For the application of Tables 10.1 to 10.12, Annexes B, C and F, weld quality level B according to
EN ISO 5817 (assuming execution class EXC3 or higher), assignment of qualified personnel and an extent
of non-destructive testing (NDT) as specified by EN 1090-2 should be implemented. Additionally, the
supplementary requirements described by the detail category tables in Clause 10 should be met.
NOTE For structures and components in execution class EXC2, weld quality level C can be sufficient.

8.3 Fatigue resistance modifications

8.3.1 Size effect

(1) The size effect due to thickness or other dimensional effects should be taken into account as
required in Tables 10.1 to 10.12. The fatigue resistance, considering size effect, should be determined as
follows:
∆𝜎𝜎C = 𝑘𝑘s Δ𝜎𝜎C,Table (8.1)

where:

∆σC,Table is the value of the considered detail category provided in the first column of table in
Clause 10
ks is the parameter size effect specified in the tables of Clause 10
NOTE For some constructional details, the size effect is already taken into account by specifying different detail
categories.

(2) In case ∆σC according to (1) is smaller than 71 N/mm², but ∆σC,Table is greater than 71 N/mm², ∆σD
shall remain as 0,737 ∆σC.

8.3.2 Post-fabrication treatment

(1) The effect of a post-fabrication treatment on fatigue resistance may be taken into account.
NOTE 1 The National annex can give provisions for post-fabrication treatments including post weld treatments
such as toe grinding, TIG re-melting of weld toe region, hammer peening, HFMI treatment and shot peening.

NOTE 2 Guidance is given in Annex F regarding fatigue design of welded joints subject to High Frequency
Mechanical Impact Treatment (HFMI treatment).

(2) The effect of a post fabrication treatment other than grinding should not be combined with the
beneficial effect of full thermal stress relief according to 7.4(3).

42
 prEN 1993-1-9:2023 (E)

9 Fatigue verification
9.1 Verification with respect to elastic behaviour

(1) The nominal stresses σEd and τEd resulting from the characteristic combination of actions
according to EN 1990, making due allowance where relevant for the macro-geometric and/or
concentrated load effects, should be limited as follows:
𝜎𝜎Ed ≤ 𝑓𝑓y /𝛾𝛾M,ser (9.1)

and
𝑓𝑓y
𝜏𝜏Ed ≤ (9.2)
𝛾𝛾 M,ser √3

NOTE 1 Elastic behaviour is a necessary condition for all stress-based verification methods of this document
including Annexes A, B, C and F.

NOTE 2 The numerical value for γM,ser is set to 1,0 unless a different value is specified by the National annex.

(2) In order to avoid low cycle fatigue, the maximum nominal stresses ranges Δσmax,Ed and Δτmax,Ed
resulting from the frequent combination of actions according to EN 1990 and calculated according to
Clause 7, making due allowance where relevant for the macro-geometric and/or concentrated load
effects, should be limited as follows:
∆𝜎𝜎max,Ed ≤ 1,5𝑓𝑓y /𝛾𝛾Mf (9.3)

and
𝑓𝑓y
∆𝜏𝜏max,Ed ≤ 1,5 𝛾𝛾 (9.4)
Mf √3

NOTE High nominal stress ranges can lead to plastic deformation at the spots of stress concentration. As a
consequence, alternating plastic deformations can be induced in cases of reversed loading that need to belimited to
avoid low cycle fatigue.

(3) Where the actual stress range spectrum is replaced by an equivalent constant stress range Δσe,2,Ed
or Δτe,2,Ed according to EN 1991, the requirement of (2) may be assumed to be fulfilled.

9.2 Verification with respect to reference value

(1) This verification should be performed for design situations with unknown maximum stress range
Δσmax,Ed or Δτmax,Ed.
NOTE Design situations with unknown maximum stress range can be:

− The variable amplitude stress range spectrum is replaced by the equivalent constant stress range Δσe,2,Ed or
Δτe,2,Ed.

− The variable amplitude stress range spectrum is calculated using a simplified fatigue load model according to
EN 1991 that neglects the largest load cycles with very low frequencies.

(2) The normal and shear stress ranges for constructional details should satisfy the following
relationships:
Δ𝜎𝜎e,2,Ed
Δ𝜎𝜎C /𝛾𝛾Mf
≤ 1,0 (9.5)

and

43
prEN 1993-1-9:2023 (E)

Δ𝜏𝜏e,2,Ed
Δ𝜏𝜏C /𝛾𝛾Mf
≤ 1,0 (9.6)

NOTE The stress components to be considered for a particular constructional detail, see 7.2, are indicated by
Tables 10.1 to 10.12. See also 7.3.

9.3 Verification with respect to fatigue limit

(1) This verification may be performed for design situations with known maximum stress range
Δσmax,Ed or Δτmax,Ed.
NOTE See note to 9.2(1).

(2) An infinite life for constructional details may be assumed if the maximum stress range of the
applied normal or shear stress spectrum satisfy the following relationships:
Δ𝜎𝜎max,Ed
≤ 1,0 (9.7)
Δ𝜎𝜎D /𝛾𝛾Mf

or
Δ𝜏𝜏max,Ed
≤ 1,0 (9.8)
Δ𝜏𝜏D /𝛾𝛾Mf

9.4 Verification for multiaxial fatigue

(1) If the considered location of a constructional detail (see cracks indicated in Table 10.1 to Table
10.12) is subject to a combination of normal and shear stresses, the fatigue verification should consider
their combined effects as follows:
a) If the normal and shear stresses simultaneously occur during each loading event, the principal
stresses should be considered and the corresponding stress range should be verified according to
Formula (9.5);
b) If the normal and shear stresses do not simultaneously occur during each loading event, the
components of damage should be added according to Miner’s summation:
• for failure in parent metal (including crack initiation at weld toes) subjected to nominal normal
stress ranges ∆σx, ∆σy, ∆σz and nominal shear stress ranges ∆τxy, ∆τxz, ∆τyz:

𝑚𝑚σ 𝑚𝑚τ
Δ𝜎𝜎j,e,2,Ed Δ𝜏𝜏
� �
Δ𝜎𝜎j,C /𝛾𝛾Mf
� +� �Δ𝜏𝜏 k,e,2,Ed
/𝛾𝛾
� ≤ 1,0 (9.9)
k,C Mf
j=x,y,z k=xy,xz,yz

• for weld failure (crack initiation at weld root) due to weld stresses according to 7.2(5):
𝑚𝑚σ 𝑚𝑚τ
Δ𝜎𝜎 Δ𝜏𝜏
�Δ𝜎𝜎 wf,e,2,Ed
/𝛾𝛾
� + �Δ𝜏𝜏 wf,e,2,Ed
/𝛾𝛾
� ≤ 1,0 (9.10)
wf,C Mf wf,C Mf

where:

mσ is the first slope parameter m1 of the fatigue resistance curve for the considered constructional
detail under normal stress loading
mτ is the first slope parameter m1 of the fatigue resistance curve for the considered constructional
detail under shear stress loading
(2) If normal and shear stresses cause the formation of fatigue cracks at different locations, a separate
fatigue verification for both locations should be performed.
(3) Where no data for Δσe,2,Ed or Δτe,2,Ed is available the verification format in Annex A may be used.

44
 prEN 1993-1-9:2023 (E)

NOTE The National annex can give information on the use of Annex A.

10 Classified constructional details for the nominal stress method

(1) The classifications contained in Tables 10.1 to 10.12 should be used for the nominal, or modified
nominal, stress method only.
NOTE 1 The weld symbols used in Tables 10.3 to 10.12 are explained by Table 10.13.

NOTE 2 The last column of Tables 10.1 to 10.11 can contain supplementary requirements beyond the provisions
of EN 1090-2, EN 1993-1-1 and EN 1993-1-8.

(2) In situations where non uniform stress distributions may be caused by global and/or local
bending, the stresses at the potential crack location should be considered.
NOTE The illustrations of the constructional details subject to normal stress generally show the distribution of
the stress σ at the mid-plane of the member as a result of normal forces.

(3) The classifications of fillet welded details may be applied to convex, concave or flat fillet welds,
although the illustrations only show one indicative fillet weld type.
(4) For N < 2×106, fatigue verification should consider whether constructional details of plain
members adjacent to joints have a lower fatigue resistance.
NOTE Due to the different slopes of details in plain members, mechanically fastened joints and welded joints,
the plain members can have a relatively lower fatigue resistance for N < 2×106 (intersection of fatigue resistance
curves).

(5) For structures with hot dip galvanizing, the particular provisions in Tables 10.1, 10.2, 10.4, 10.6,
10.9 and 10.11 should be applied.
NOTE 1 Regarding the particular provisions, for the use of the next lower detail categories as required by the
footnotes in Tables 10.1, 10,2, 10.4, 10.6, 10.9 and 10.11, see 8.1(2), 8.1(3) and 8.1(6).

NOTE 2 These provisions do not apply to thermal sprayed zinc/aluminium details.

(6) For structures made of weathering steels, the particular provisions in Tables 10.1 and 10.2 should
be applied.
NOTE Regarding the particular provisions, for the use of the next lower detail categories as required by the
footnotes in Tables 10.1 and 10.2, see 8.1(2) and 8.1(3).

45
prEN 1993-1-9:2023 (E)

Table 10.1 — Plain members

Detail
Constructional detail Description Supplementary Requirements
category
Rolled or extruded products ①, ③: Defects, sharp edges and rolling flaws should
subject to normal stress: be removed by grinding until a smooth transition is
achieved with mean surface roughness depth in the
① plates and flats order of mill scale (Rz ≤ 200 μm).
180 with as rolled edges; Repair by welding not allowed, requiring following
m1 = 5 ② rolled sections technical delivery conditions:
with as rolled edges; ①: Class A3 acc. to EN 10163-2.
③ seamless hollow sections, ③: Option 1.5 acc. to EN 10210
either rectangular or
circular
①, ②, ③: Defects, sharp edges and rolling flaws
should be removed by grinding until a smooth
transition is achieved.
160 Repair by welding not allowed, requiring following
m1 = 5 technical delivery conditions:
①: Class A3 acc. to EN 10163-2.
②: Class C3 acc. to EN 10163-3.
③: Option 1.5 acc. to EN 10210
①, ②, ③: Defects, sharp edges and rolling flaws
should be removed by grinding until a smooth
125 transition is achieved.
m1 = 5 Repair by welding should be followed by grinding
until a smooth transition is achieved.

Sheared or thermally cut Stress concentrations ④ Mean surface

material subject to normal due to macro-geometric roughness depth after
160 stress: effects should be grinding in the order of
m1 = 5 accounted for. Δσ should mill scale (Rz ≤ 200 μm).
④ with subsequent grinding; be calculated using net Repair by welding not
⑤ with subsequent deburring section including allowed.
appropriate stress
concentration factors. ⑤ Thermal cut quality
acc. to EN 9013 with
mean surface profile
125 range 2.
m1 = 5 Repair by welding
followed by grinding
until a smooth transition
is achieved.
Rolled or extruded products For shear loads in webs ⑥, ⑦ and ⑧:
subject to shear stress: of any section class, Requirements as for ①
Formula (8.25) of to ③ in detail category
⑥ to ⑧ EN 1993-1-1 should be 125.
same description as for used to calculate Δτ.
① to ③

100
m1 = 5

For ① to ⑧ made of weathering steel the next lower detail category of Figure 8.1 should be used, but not higher than Detail Category 140.
For ① to ⑧ with hot dip galvanizing the next lower detail category of Figure 8.1 should be used, but not higher than Detail Category 140.

46
 prEN 1993-1-9:2023 (E)

Table 10.2 — Mechanically fastened joints

Detail
Constructional detail Description Supplementary Requirements
category

① Double covered symmetrical joint Δσ should be calculated for all

subject to normal stress members containing a potential
112 with preloaded high strength bolts crack site using the gross cross-
m1 = 5 or preloaded injection bolts section.

② Double covered symmetrical joint For all members containing a

subject to normal stress potential crack site, Δσ should be
calculated with
with fitted bolts 𝑑𝑑0 3
∆𝜎𝜎 = ∆𝜎𝜎net �𝑎𝑎 + �𝑏𝑏 − 𝑐𝑐 � �
or non-preloaded 𝑤𝑤
90 injection bolts where σnet is the net cross-section
m1 = 5 Holes stress.
drilled
with non-preloaded or For fitted bolts and non-preloaded
normal bolts in holes reamed injection bolts:
with normal clearance 1 bolt row: a = 1; b = 1,6; c = 2,7
without load reversal 2 bolt rows: a = 1; b = 1,3; c = 2,2
≥ 3 bolt rows: a = 1; b = 1,1; c = 1,8
with fitted bolts For non-preloaded normal bolts in
or non-preloaded Holes holes with normal clearance:
injection bolts ther- a = 1; b = 1,6; c = 2,7
71 mally See note at the end of table for
m1 = 3 with non-preloaded cut dimensions d0 and w.
normal bolts in holes or
The bolts should be checked
with normal clearance punched
using ⑦.
without load reversal
③ One-sided fully supported Δσ should be calculated for all
connection subject to normal stress members containing a potential
with preloaded high strength bolts crack site using the gross cross-
or preloaded injection bolts section.
One-sided unsupported
100
connections should be avoided or
m1 = 5
the effect of eccentricity should be
taken into account in the stress
calculation.

④ One-sided fully supported For all members containing a

connection subject to normal stress potential crack site, Δσ should be
calculated as for ②.
with fitted bolts
The bolts should be checked
or non-preloaded
80 Holes using ⑦.
injection bolts
m1 = 5 drilled One-sided unsupported
with non-preloaded or connections should be avoided or
normal bolts in holes reamed the effect of eccentricity should be
with normal clearance taken into account in the stress
without load reversal calculation.
with fitted bolts
or non-preloaded Holes
injection bolts ther-
56 mally
m1 = 3 with non-preloaded cut
normal bolts in holes or
with normal clearance punched
without load reversal

47
prEN 1993-1-9:2023 (E)

Detail Supplementary
Constructional detail Description
category Requirements

⑤ Structural element Δσ should be calculated using

90 subject to normal stress net cross-section.
m1 = 5 with drilled or reamed
round holes

as aforementioned, but
50 with thermally cut or
m1 = 3 punched round holes

⑥ Black bolts and rods Δσ should be calculated using

subject to normal stress the tensile stress area of the
71
with normal metric screw bolt or rod according to
m1 = 3
Size effect for threads EN 1993-1-8, see 5.7.
d > 30 mm. rolled after heat treatment See Annex E for preloaded
bolts and rods.
56 ks = (30/d)0,25 as aforementioned, but
m1 = 3 rolled before heat treatment
where d is the
50 nominal bolt as aforementioned, but
diameter in hot-dip or electric galvanized
m1 = 3
mm. after rolling

50 as aforementioned, but
m1 = 3 with cut threads

⑦ Fitted bolts Carbon steel bolts should be

subject to single or of Grade 5.6, 5.8, 6.8, 8.8 or
double shear 10.9 acc. to EN 1993-1-8.
Stainless steel bolts should
have nominal values of yield
as aforementioned, but strength and ultimate tensile
with normal bolts in holes strength equal to or higher
100
with normal clearance than those of Grade 70
m1 = 5
without load reversal according to EN 1993-1-4.
Thread should not be in
shear plane(s).
Δτ should be calculated using
the shank area of the bolt.

NOTE It applies w = max (p2; 2e2). See EN 1993-1-8 for definition of d0, p2 and e2.

For ⑤ made of weathering steel the next lower detail category of Figure 8.1 should be used.
For details ②, ④ and ⑤ with hot dip galvanizing and with drilled or reamed holes, the next lower detail category of Figure 8.1 should be
used.

48
 prEN 1993-1-9:2023 (E)

Table 10.3 — Welded built-up sections and longitudinal welds

Detail
Constructional detail Symbol Description Supplementary Requirements
category

① Automatic or fully mecha- None.

nised butt welds,
125
welded from both sides,
without stop-starts
as aforementioned, but
112
with stop-starts
② Automatic or fully mecha- For cover plates two parallel
nised fillet welds, single fillet welds are necessary.
125 without stop-starts Cover plate ends should be
checked using ⑥, ⑦ or ⑧ of
Tab. 10.6.

as aforementioned, but
with stop-starts
112

③ Details ① and ② as manual None.

100 ③ Manual welding See above
welds
④ Automatic or fully In case of permanent root
mechanised butt welds, backing, tack welds should
112 welded from one side, attach the root backing inside
on continuous root the groove.
backing, without stop-starts
as aforementioned, but
100 with stop-starts or manual
butt welds
⑤ Butt welds, welded from Extent of NDT according to
both sides, ground flush with EN 1090-2: 100%.
125
plate surface in direction of
stress, without stop-starts
as aforementioned, but None.
112 without grinding

as aforementioned, but None.

90 without grinding and
with stop-starts
⑥ Repaired automatic or fully None.
100 mechanised or manual fillet
See above or butt welds for ① to ④
as aforementioned, but
90
⑥ Repair welding for ⑤
as aforementioned, but
Original
with grinding or blending
detail
smoothly the weld toes
category
for ① to ⑤
⑦ Chain or staggered Δσ should be calculated using
intermittent fillet welds normal stress in the parent
with g/h ≤ 2,5 metal assuming the weld were
80 continuous.

49
prEN 1993-1-9:2023 (E)

Detail Supplementary
Constructional detail Symbol Description
category Requirements

⑧ Butt or fillet or intermittent Δσ should be calculated

welds with a cope hole of using normal stress in the
a height ≤ 60mm parent metal assuming the
For cope holes with a height weld were continuous.
71 Longitudinal weld welded
> 60mm see ① in Table 10.5.
all around inside cope hole.
NOTE EN 1993-1-5 can have Transverse weld should be
stricter rules on cope holes. checked using Table 10.4.

⑨ Automatic or fully mechanized None.

or manual butt or fillet welds
in box girders,
welded from one side
100

⑩ Automatic or fully mechanised Wall thickness t ≤ 12,5mm

140 longitudinal welds
in hollow sections
according to EN 10210 or Wall thickness t > 12,5mm
EN 10219 (without stop-starts
125 as manufactured in a conti-
nuous process)
⑪ Automatic or fully mechanized None.
longitudinal welds
in hollow sections
90 with stop-starts

NOTE 1 Welding other than automatic and fully mechanised welding should be treated as manual welding.
NOTE 2 Slope parameter m1 = 3 of fatigue resistance curve unless otherwise stated in detail category.

50
 prEN 1993-1-9:2023 (E)

Table 10.4 — Transverse welded butt joints (full penetration)

Detail
Supplementary
cate- Constructional detail Symbol Description
Requirements
gory
① Splices in plates and Weld all-around ground
flats of same thickness, flush with plate surface in
112 welded from both sides, direction of stress.
ground flush Misalignment ≤ 5% of plate
thickness, see NOTE 2.
② as aforementioned, but ②: Welded in welding
as-welded position PA acc. to
with flank angle ≥ 150° EN ISO 6947.
90 ②③:
Weld ground flush at plate
edges in direction of stress,
See NOTE 1 for size effect where relevant, after
removing weld run-off
③ as aforementioned, but pieces.
as-welded
with flank angle ≥ 110° Misalignment ≤ 5% of plate
80 thickness, see NOTE 2.

④ Splices in plates and See ② and ③.

flats of same thickness, Tack welds should attach
welded from one side on the root backing inside the
71 permanent root backing, groove.
tack welds terminate
not closer than 10 mm to
the plate edges
⑤ as aforementioned, but
tack welds terminate
50 closer than 10 mm to the
plate edge and/or if a good
fit cannot be guaranteed

⑥ Splices in plates and See ② and ③.

71 flats of same thickness, Full penetration checked by
welded from one side appropriate NDT
without root backing
None.
36
See NOTE 1 for size effect of Detail
Category 71
⑦ Flange and web splices ⑦⑧⑨:
in plate girders, No full cross-section joint.
112
welded from both sides, Splices welded before
ground flush assembly of girder.
⑧ as aforementioned, but Longitudinal weld should be
as-welded checked using Table 10.3
with flank angle ≥ 150° ⑦: see ①.
90 ⑧: see ②
⑨: see ③.

⑨ as aforementioned, but
as-welded,
See NOTE 1 for size effect with flank angle ≥ 110°
80

51
prEN 1993-1-9:2023 (E)

⑩ Flange splices of full cross- Rolled sections cut and

section butt welds of rewelded.
rolled sections of same See ①.
90 dimensions without tole-
rance difference, ground
flush, with semi-circular
cope holes
⑪ as aforementioned, but See ②.
as-welded,
with flank angle ≥ 150° or
80 with tolerance differences

See NOTE 1 for size effect

52
 prEN 1993-1-9:2023 (E)

Detail
Supplementary
cate- Constructional detail Symbol Description
Requirements
gory
⑫ Web splices in full cross- Δσ should be calculated
section joints with cope using the modified
112
holes, welded from both nominal stress in the web
sides, ground flush at location A taking into
account stress raising
⑬ as aforementioned, but effect of the cope hole.
as-welded
with flank angle ≥ 150°
90

⑭ as aforementioned, but
as-welded,
with flank angle ≥ 110°
80

See NOTE 1 for size effect

⑮ Full cross-section butt Rolled sec-
welds of rolled sections tions cut
of same dimensions and rewel-
112
without tolerance ded.
difference, ground flush, See ①.
without cope holes
Full pene-
⑯ as aforementioned, but See ②. tration at
as-welded the region
with flank angle ≥ 150°, of web-to-
90 and with tolerance flange
differences junction
should be
checked by
⑰ as aforementioned, but See ③. NDT.
as-welded
with flank angle ≥ 110°
80 See NOTE 1 for size effect
and with tolerance
except for Detail Category 63
differences

See ① to Penetration
63 As for ⑮ to ⑰
③. not checked.
⑱ Splices in plates and See ①.
flats tapered in width Δσ should be calculated
or in thickness using normal stress in the
112
with slopes ≤ 1:4, thinner plate or flat.
Taper in thickness welded from both sides, See NOTE 3.
ground flush
⑲ as aforementioned, but See ②.
as-welded Weld toes blended
with flank angle ≥ 150° smoothly.
90 See NOTE 3.

See NOTE 1 for size effect Taper in thickness

For ①, ⑦, ⑫, ⑮ and ⑱ with hot dip galvanizing the next lower detail category of Figure 8.2 should be used.

53
prEN 1993-1-9:2023 (E)

Detail Supplementary
Constructional detail Symbol Description
category Requirements
⑳ Splices in plates and See ④.
flats tapered in width Δσ should be calculated
or in thickness using normal stress in the
with slopes ≤ 1:4, thinner plate or flat.
71
welded from one side See NOTE 3.
on permanent root
backing and tack welds
as for ④
See ⑤.
as aforementioned, but
Taper in thickness with tack welds as Δσ should be calculated
for ⑤ using normal stress in the
50 thinner plate or flat.
See NOTE 3.
See NOTE 1 for size effect

Splices in plates and Misalignment ≤ 5% of

flats of different thick- thinner plate thickness,
ness see NOTE 2.
with transition in weld, Weld toes blended
71 centrelines aligned smoothly.
Δσ should be calculated
using normal stress in the
See NOTE 1 for size effect thinner plate or flat.
See NOTE 3.
Continuous flange should
Transverse butt weld at
be checked using ④ or ⑤
intersecting flanges as
or ⑥ of Table 10.5.
④ or ⑤ in Table 10.5

as aforementioned, but
as ② in with transition radius r
Tab. 10.6 acc. to ⑥ in Table 10.5

1 − ground
See ②.
Splices in stacked
plates, The end groove welds
as-welded should not melt while
with flank angle ≥ 150°, laying the butt weld.
The root of the butt weld
should be positioned in
the center of one of the
See NOTE 1 for size effect 1 − end groove weld plates.
90
as aforementioned, but See .
tapered in thickness Weld toes blended
with slopes ≤ 1:4 smoothly.
See NOTE 3.

See NOTE 1 for size effect 1 − end groove weld

NOTE 1 Size effect for t > 25mm is considered by stress modification with ks = (25/t)0,2 for as-welded details and ks = (25/t)0,1 for details that
are ground flush where t is the thinner plate thickness in mm for which the stress range is calculated.
NOTE 2 Misalignment due to fabrication ≤ 5% of plate thickness. An eccentricity is considered by appropriate nominal stress modification, see
D.4(2).
NOTE 3 For⑱ to and , as parts of girder flanges in bending, the modified nominal stress range approach is used. Recommended stress
concentration factors kf are provided by Annex D.
NOTE 4 Slope parameter m1 = 3 of fatigue resistance curve unless otherwise stated in detail category.
For ①, ⑦, ⑫, ⑮ and ⑱ with hot dip galvanizing the next lower detail category of Figure 8.2 should be used.

54
 prEN 1993-1-9:2023 (E)

Table 10.5 — Weld attachments and stiffeners

Detail Supplementary
Constructional detail Symbol Description
category Requirements

① Plates and beam flanges Attachment welded all

subject to normal stress around. Attachment
80 with longitudinal attachment thickness is less than its
at their surfaces, height. Otherwise see ⑤
as-welded,  ≤ 50mm or ⑥ of Table 10.6.
Δσ should be calculated
as aforementioned, but using the normal stress
71 in the parent metal
 = distance of weld toes 50 <  ≤ 200mm
neglecting the attach-
63 as aforementioned, but  > 200mm ment.

② as aforementioned, but See ①.

with chamfered attachment ends,
as-welded

as ①

Before and after welding

③ Plates and beam flanges See ①.

subject to normal stress Transition of attached
with longitudinal attachment plate by machining or
weld at their surfaces, gas cutting before
termi- with transition radius r ≥ 150 mm welding, reinforced weld
nation; and reinforced (full penetration)
80 terminations ground
weld terminations smooth until flush with
rest of
plate surface in direction
weld as
of stress after welding.
① and ②
Before and after welding (1 − ground)

④ Plates or beam flanges Attachment welded all

subject to normal stress around.
with longitudinal attachment Δσ should be calculated
56 at their edges, using normal stress in
as-welded the parent metal
neglecting the attach-
ment.
⑤ as aforementioned, but See ④.
with chamfered attachment ends
as ④ as-welded

⑥ Plates or beam flanges See ④.

subject to normal stress Transition of attached
with longitudinal attachment plate by machining or
weld at their edges, gas cutting before
termi- with transition radius r ≥ 150mm welding, weld termina-
nation;
tions ground smooth
until flush with plate
71 rest of
edge in direction of
weld
stress after welding.

55
prEN 1993-1-9:2023 (E)

Detail Symb Supplementary

Constructional detail Description
category ol Requirements

⑦ Plates and beam Δσ should be calculated

flanges using normal stress in
subject to normal parent metal neglecting
stress the attachment.
with transverse Ends of welds should be
attachment ground to remove
at their surfaces, undercut if exists.
as-welded:

80 ℓ ≤ 50mm
.
71 50 < ℓ ≤ 80mm
⑧ Flanges and webs of See ⑦.
rolled sections Longitudinal welds, if
subject to normal stress exist, should be checked
with fitted
as ⑦ using Table 10.3.
transverse attachment
at their surfaces,
welded all round, as-
welded

as aforementioned,
but
of built-up sections
(weld intersection)
as ⑦

as aforementioned,
but
with cut holes,
welded all around
as ⑦

as aforementioned,
but
not welded all
around
as ⑦

⑨ Webs subject to ∆σ should be calculated

combined using principal stress in
normal and shear the web neglecting the
stresses attachment.
with transverse Ends of welds should be
as ⑦ attachment ground to remove
terminating in the undercut if exists.
web,
welded all round,
as-welded

56
 prEN 1993-1-9:2023 (E)

Detail Supplementary
Constructional detail Symbol Description
category Requirements

⑩ Box girder sections with See ⑦.

diaphragms welded to flanges
and/or webs subject to normal
stress

as ⑦

⑪ Plates, flanges or webs Δσ should be calculated

subject to normal stress using normal stress in
with small attachment (such as parent metal neglecting
welded shear stud, bushings etc.) the small attachment.
of any shape at their surfaces

Note: It applies ℓ ≤ 50mm for small

attachments. See ⑦ for definition
of ℓ.

80

NOTE Slope parameter m1 = 3 of fatigue resistance curve unless otherwise stated in detail category

57
prEN 1993-1-9:2023 (E)

Table 10.6 — Load carrying welded joints

Detail Supplementary
Constructional detail Symbol Description
category Requirements

① Load carrying Δσ should be calculated

attachment subject using normal stress in the
to normal stress load carrying attachment.
whose stress pattern is The eccentricity of the load
not affected by local carrying attachments (or
deformation of the continuity plating, if exists)
member to which it is should not exceed 15% of
welded, see NOTE 1 to the thickness of the
①, with weld toe intermediate plate.
failure For load carrying
attachments with partial
penetration butt welds or
fillet welds, weld root
failure should additionally
be checked using ③, see
Note 2 to ①.

71 ℓ ≤ 80

63 80 < ℓ ≤ 100 For all t

[mm]
Key 100 < ℓ ≤
56 1 load carrying attachment 120
2 member with load-carrying attachment
ℓ > 120 t ≤ 20
3 intermediate plate
4 continuity plating 120 < ℓ ≤ t > 20
200
50
ℓ > 200 20 < t ≤
30
200 < ℓ ≤ t > 30
300
45
ℓ > 300 30 < t ≤
50
40 ℓ > 300 t > 50
NOTE 1 to ① Typical examples: symmetrical load carrying attachments at opposite surfaces of a plate (cruciform joint); load carrying
attachments welded to I sections or hollow sections stiffened by continuity plating.
NOTE 2 to ① Effective full penetration butt welds according to EN 1993-1-8 are considered as partial penetration butt welds in respect of
fatigue.

58
 prEN 1993-1-9:2023 (E)

② as ①, but See ①.
where the stress pattern Δσ should be calculated
of the load carrying using modified nominal
attachment is affected stress (compare Sec. 7.3.3)
by local deformation in the load carrying
of the member to which attachment.
it is welded, see NOTE to ②.

as ① as ①

NOTE to ② Typical examples: load carrying attachments on unstiffened ⌶ sections or unstiffened hollow sections; intersection of plates in
the same plane like intersecting flanges of main and cross beam. These details do not show best practice. The calculation of the modified
stresses for these details is generally not simple.

59
prEN 1993-1-9:2023 (E)

Detail
Constructional detail Symbol Description Supplementary Requirements
category

③ Load carrying attachment Δσ should be calculated using

with partial penetration nominal stress in the weld.
butt welds or fillet The eccentricity of the load-
weldssubject to normal stress carrying attachments (or
whose stress pattern is not continuity plating, if exists)
affected by local deformation should not exceed 15% of the
of the member to which it is thickness of the intermediate
welded, see also Notes to plate.
①,with weld root failure
For great weld sizes, weld toe
failure should additionally be
checked using ①, see Note 2 to
①.
as aforementioned, but as aforementioned, but
where the stress pattern of Δσ should be calculated using
40 the load carrying attachment modified nominal stress
is affected by local (compare Sec. 7.3.3) in the weld.
deformation of the
member to which it is
welded, see also Note to
②.
Key
1 load carrying attachment
2 member with load-carrying attachment
3 intermediate plate
4 continuity plating

④ Fillet welded overlapped Δσ should be calculated using

joint the area as shown in the figure
subject to normal stress and normal stress in main plate.
with failure of main plate Weld should terminate at least
10mm from plate edge.
Failure in the overlapping plates
should be verified using ⑤.
Shear stress in the weld should
be verified using ⑩.
Parameter ℓ in ① should be
taken as equal to weld length
As ① measured in stress direction.

Key
1 overlapping plate
2 main plate

60
 prEN 1993-1-9:2023 (E)

⑤ Fillet welded overlapped See ④, but

joint subject to normal stress Δσ should be calculated using
with failure of overlapping normal stress in the overlapping
plates plates.
50

61
prEN 1993-1-9:2023 (E)

Detail
Symb Supplementary
cate- Constructional detail Description
ol Requirements
gory
⑥ Single or stacked cover If the cover plate is wider
plates on flanges in hot- than the flange, a frontal
rolled beams and transverse weld should be
plate girders, with cover laid. The weld should be
plate length ≥ 300mm, ground to remove undercut.
with or without frontal
transverse weld

NOTE See ① for size

effect of shorter cover
plates.

Key tc < t [mm] tc ≥ t

[mm]
1 longitudinal weld
56 2 frontal transverse weld t ≤ 20 −
50 20 < t ≤ 30 t ≤ 20

30 < t ≤ 50 20 < t ≤
45
30
t > 50 30 < t ≤
40
50
36 − t > 50

⑦ Single cover plates on Δσ should be calculated

flanges in hot-rolled using normal stress in the
beams and plate girders, flange.
with cover plate Surface of frontal transverse
length ≥ 300mm, with weld ground, weld toes
reinforced frontal blended smoothly.
transverse weld,
chamfered with slope
80
1:2

Before and after welding weld

termi-
natio ⑧ as ⑦, but
n chamfered with slope
1:3

80

Before and after welding

⑨ Joints transmitting shear For shear ∆τ should

stress, with continuous loads in be
80 fillet welds and partial flange-to-web calculated
m1 = 5 penetra- tion butt welds connections in using shear
sections of stress in the
any class, weld.

62
 prEN 1993-1-9:2023 (E)

as aforementioned, but Eq. (8.25) of ∆τ should

with full penetration butt EN 1993-1-1 be
welds should be calculated
used to using shear
NOTE Effective full pene- calculate Δτ. stress in the
tration butt welds according Stress parent
100 to EN 1993-1-8 are consi- concentra- metal.
m1 = 5 dered as partial tions due to
penetration butt welds for macro-
fatigue. geometric
effects should
be accounted
for.

63
prEN 1993-1-9:2023 (E)

Detail Supplementary
Constructional detail Symbol Description
category Requirements

⑩ Fillet welded overlapped As for ④, but

joint subject to shear ∆τ should be calculated
stress using shear stress in the
with weld root failure weld considering the total
80 weld length.
m1 = 5

⑪ Welded stud shear ∆τ should be calculated

connectors for using cross section of the
90 composite stud shear connector.
m1 = 8, applications See EN 1994-2.
see subject to shear stress
EN 1994
Part 2

⑫ Ring flange connection Δσ should be calculated

with full penetration using modified nominal
welds where the stress stress (compare Sec. 7.3.3)
pattern of the load in the load carrying member.
71 carrying member is Bolts should be checked
affected by local using
deformation of the ring ⑥ of Table 10.2.
flange to which it is
welded
⑬ as aforementioned, but As for ⑫, but
with fillet welds Δσ should be calculated
36 using modified nominal
stress (compare Sec. 7.3.3)
in the weld.
NOTE Slope parameter m1 = 3 of fatigue resistance curve unless otherwise stated in detail category
For ⑪ with hot dip galvanizing the next lower detail category of Figure 8.2 should be used.

64
 prEN 1993-1-9:2023 (E)

Table 10.7 — Structural hollow sections

Detail Symb Supplementary

Constructional detail Description
category ol Requirements

① Hollow-section-to-plate Δσ should be calculated

joint with flatted hollow using normal stress in
section with ∅ ≤ 200mm the hollow section.
and
71 wall thickness t ≤ 12,5mm

② Hollow-section-to-plate Δσ should be calculated

joint with slitted hollow using normal stress in
section,hole at the end of slit, the net hollow section
hollow section chamferedwith with
slopes α ≤ 45° and gusset plate Anet = A – 2 dH
with semi-circular cut-out where dH is the hole dia-
45 meter at the end of slit.
The welds subject to
shear should be verified
using ⑨ of Table 10.6.

1 – interior sealing plate, if necessary

③ Hollow-section-to-plate Δσ should be calculated
joint with slitted hollow using normal stress in
section, the hollow section.
with full penetration Sealing plate should be
welds, hollow section welded after hollow-
chamfered with section-to-plate joint to
56 slopes α ≤ 45°, enable NDT of the full
gusset plate with semi- penetration welds.
circular cut-out and rounded
plate- tube tip, exterior
sealing plate for corrosion
protection (single
fillet welded)
as aforementioned, but Δσ should be calculated
with partial penetration using normal stress in
welds, the hollow section.
50
1 – exterior sealing plate gusset plate with semi-
Z – plate-tube tip, chamfered or rounded circular cut-out and rounded
or chamfered plate-tube tip
as aforementioned, but
36 with fillet welds

④ Hollow-section-to-plate Δσ should be calculated

joint with slitted hollow using normal stress in
section, hollow section the hollow section.
unchamfered, semi-
circular shaped gusset
50
plate with rounded or
chamfered plate-tube tip,
exterior sealing plate for
corrosion protection
(single fillet welded)
as aforementioned, but
36 with fillet welds
1 – exterior sealing plate
Z – plate-tube tip, chamfered or rounded

65
prEN 1993-1-9:2023 (E)

⑤ Weld joint between Width parallel to stress

circular or rectangular direction ℓ ≤ 100mm,
structural hollow section and else like longitudinal
another section with fillet attachment, see
welds subject to normal stress Table 10.5.
in the continuous hollow
71
section with wall thickness
t ≤ 12,5mm

66
 prEN 1993-1-9:2023 (E)

Detail Supplementary
Constructional detail Symbol Description
category Requirements

⑥ Butt-welded splices in Δσ should be calculated

circular structural hollow using normal stress in the
sections, welded from one hollow section.
71 side, with flank angle ≥ Welded in welding
150°, position PA or PB
wall thickness 12,5 < t ≤ according to EN ISO 6947.
20mm
as aforementioned, but
90 wall thickness 8 < t ≤ 12,5
mm
as aforementioned, but
71
wall thickness t ≤ 8mm
as aforementioned, but Δσ should be calculated
on root backing, using normal stress in the
wall thickness 20 < t ≤ hollow section.
60mm Welded in welding
71 position PA, PB, PC or PF
m1 = 5 according to EN ISO 6947.
In case of permanent root
backing, tack welds should
attach the root backing
inside the groove.
⑦ Butt-welded splices in Δσ should be calculated
circular structural hollow using normal stress in the
sections of different wall hollow section with
thickness, welded from smaller wall thickness.
one side Welded in welding
on root backing, position PA, PB, PC or PF
with flank angle ≥ 150°, according to EN ISO 6947.
wall thicknesses 20 < ti ≤ In case of permanent root
60mm, t2 / t1 > 0,6 backing, tack welds should
71
attach the root backing
m1 = 5
inside the groove.

as aforementioned, but
without root backing

⑧ Weld splices in circular Δσ should be calculated

structural hollow sections using normal stress in the
with intermediate plate, hollow section.
56 wall thickness 8 < t ≤ 12,5
mm,
with butt welds,
with weld toe failure
as aforementioned, but
50 with wall thickness t ≤
8mm

⑨ As ⑧ with 8 < t ≤ 12,5mm, Δσ should be calculated

but with fillet welds or partial using weld stress.
45
penetration butt welds, with For greater weld sizes,
weld root failure weld toe failure should
additionally be checked
as aforementioned, but
using ⑧, see Note 1.
40 with wall thickness t ≤
8mm

67
prEN 1993-1-9:2023 (E)

Detail Supplementary
Constructional detail Symbol Description
category Requirements

⑩ Butt-welded splices in Δσ should be calculated

rectangular structural hollow using normal stress in the
71 sections, welded from one side, hollow section.
with flank angle ≥ 150°, Welded in welding position
wall thickness 8 < t ≤ 12,5mm PA or PB according to
EN ISO 6947.
as aforementioned, but
56 wall thickness t ≤ 8mm

⑪ Weld splices in rectangular Δσ should be calculated

structural hollow sections using normal stress in the
with intermediate plate, hollow section.
50
wall thickness 8 < t ≤ 12,5mm,
with butt welds,
with weld toe failure
as aforementioned, but
45
with wall thickness t ≤ 8mm

⑫ As ⑪ with 8 < t ≤ 12,5mm, Δσ should be calculated

but with fillet welds or partial using weld stress.
40
penetration butt welds, For great weld sizes, weld
with weld root failure toe failure should
additionally be checked
as aforementioned, but
36 using ⑪, see Note 1.
with wall thickness t ≤ 8mm
NOTE 1 For partial penetration butt welds or fillet welds, weld root failure as assumed for ⑨ and ⑫ is expected for normal weld sizes
a ≤ t. For greater weld sizes, weld toe failure can also occur. Effective full penetration butt welds according to EN 1993-1-8 are considered as partial
penetration butt welds in respect of fatigue.
NOTE 2 Slope parameter m1 = 3 of fatigue resistance curve unless otherwise stated in detail category.

68
 prEN 1993-1-9:2023 (E)

Table 10.8 — Lattice girder joints made of structural hollow sections

Detail
catego Constructional detail Symbol Description Supplementary Requirements
ry
① Gapped K- and N-joints Chords and braces should sepa-
made of circular rately be verified accounting for
90 structural hollow secondary bending according to
m1 = 5 sections Tab. D.2.
with ti/t0 ≤ 0,5 For fillet welds, see Note 1.
for all ti and t0 – Dimensions: 35° ≤ θ ≤ 60° ;
as aforementioned, but 4 < t0 ≤ 20mm ; 4 < ti ≤ 20mm ;
71
with 0,5 < ti/t0 ≤ 0,7 d0/t0⋅t0/ti ≤ 40 ; d0 ≤ 325mm ;
m1 = 5
for t0 ≤ 10mm 0,3 ≤ di/d0 ≤ 0,67
– In-plane eccentricity:
as aforementioned, but
63 -0,5d0 ≤ ei/p ≤ 0,25d0
with 0,5 < ti/t0 ≤ 0,7
m1 = 5 – Out-of-plane eccentricity:
for t0 > 10mm
eo/p ≤ 0,02d0
as aforementioned, but
36 – Steel: fy,0 and fy,i ≤ 700 N/mm²
with 0,7 < ti/t0 ≤ 1,0
m1 = 5 – Gap: g ≥ 2ti
for all ti and t0
② Gapped K- and N-joints Chords and braces should sepa-
made of rectangular rately be verified accounting for
71 structural hollow secondary bending according to
m1 = 5 sections with ti/t0 ≤ 0,5 Tab. D.2. Detail categories
should be linearly interpolated
for 0,5 ≤ ti/t0 ≤ 1,0.
Fillet welds with a = ti
as aforementioned, but
– Dimensions: 35° ≤ θ ≤ 50°;
with ti/t0 = 1,0
t0 ≤ 20mm ; ti ≤ 20mm;
b0/t0⋅t0/ti ≤ 25 ; b0 ≤ 200mm;
0,4 ≤ bi/b0 ≤ 1,0;
– In-plane eccentricity:
36 -0,5h0 ≤ ei/p ≤ 0,25h0
m1 = 5
– Out-of-plane eccentricity:
eo/p ≤ 0,02b0
– Steel: fy,0 and fy,i ≤ 700 N/mm²
– Gap:
0,5(b0 - bi) ≤ g ≤ 1,1(b0 - bi)
and g ≥ 2t0
③ Overlapped K-joints Chords and braces should sepa-
made of circular or rately be verified accounting for
rectangular structural secondary bending according to
hollow sections Tab. D.2. Detail categories
with ti/t0 ≤ 0,7 should be linearly interpolated
for
71
0,7 ≤ ti/t0 ≤ 1,0.
m1 = 5
Fillet welds with a = ti
– Dimensions: 35° ≤ θ ≤ 50° ;
t0 and ti ≤ 8mm;
b0 and d0 ≤ 200mm;
0,4 ≤ bi/b0 ≤ 1 ; 0,25 ≤ di/d0 ≤ 1;

69
prEN 1993-1-9:2023 (E)

as aforementioned, but b0/t0⋅t0/ti ≤ 25 ; d0/t0⋅t0/ti ≤ 25;

with ti/t0 = 1,0 – In-plane eccentricity:
-0,5h0 ≤ ei/p ≤ 0,25h0
-0,5d0 ≤ ei/p ≤ 0,25d0
– Out-of-plane eccentricity:
eo/p ≤ 0,02b0 or ≤ 0,02d0
– Overlap:
30% ≤ overlap ≤ 100%
overlap = (q/p) × 100%
See Note 2 for definition of
p and q.

56
m1 = 5

Welds are not shown in these figures.

70
 prEN 1993-1-9:2023 (E)

Detail Supplementary
Constructional detail Symbol Description
category Requirements

④ Overlapped N-joints See ③.

made of circular or
rectangular
structural hollow
sections
71 with ti/t0 ≤ 0,7
m1 = 5

as aforementioned,
but
with ti/t0 =1,0

50
m1 = 5

Welds are not shown in these figures.

NOTE 1 Fillet welds are permitted. For fillet welded K-joints with wall thickness ti > 8mm, weld failure can occur.

NOTE 2 Definition of p and q for overlapped joints

NOTE 3 The usage of the hot-spot stress method, see Annex B, can lead to a more efficient design.

71
prEN 1993-1-9:2023 (E)

Table 10.9 — Orthotropic decks – closed stiffeners

Detail Supplementary
Constructional detail Symbol Description
category Requirements

① Continuous stiffener to cross- Δσ should be calculated

beam joint with extended cut- using normal stress in the
out in crossbeam stiffener web, near bottom
subject to normal stress of the weld between the
in stiffener, crossbeam and stiffener
stiffener failure, (spot ‘A’).
80 crossbeam web thickness Dimensions b, c, ce , rl , ru
t ≤ 12mm and wt for extended cut-
out shape as specified in
EN 1993-2, Annex C.

as aforementioned, but
with t > 12mm

as aforementioned, but Δσ should be calculated

with extended cut-out shape using normal stress in the
71 according to Fig. C.13 or C.14 stiffener web, near bottom
of EN 1993-2 of the weld between the
Key crossbeam and stiffener.
1 deck plate
2 continuous stiffener
3 cross beam
4 extended cut-out
I before welding
II after welding
② Continuous stiffener to cross- Δσ should be calculated
beam joint with extended cut- using normal stress in the
out in crossbeam critical section (spot ‘A’) of
subject to normal stress the crossbeam modeled as
in crossbeam, Vierendeel beam as
crossbeam web failure at free specified in 9.4.2 of
71 − edge of extended cut-out EN 1993-2.

③ As ①, but Δσ should be calculated

with close fit cut-out in using normal stress in the
crossbeam stiffener bottom flange
(spot ‘A’).

As
①

Key
1 deck plate
2 continuous stiffener
3 cross beam
I before welding
II after welding

72
 prEN 1993-1-9:2023 (E)

Detail Supplementary
Constructional detail Symbol Description
category Requirements

④ Discontinuous stiffener to Δσ should be calculated

crossbeam joint (crossbeams using normal stress in the
separates stiffeners) stiffener bottom flange
subject to normal stress
in stiffener,
with butt welds,
stiffener failure

71

Key
1 deck plate
2 discontinuous stiffener
3 cross beam
⑤ As ④, but Δσ should be calculated
with fillet welds or partial using weld stress
penetration butt welds,
weld root failure

36

⑥ Splice in stiffener Δσ should be calculated

subject to normal stress, using normal stress in the
welded from one side bottom flange of stiffener.
100
on permanent root backing, Tack welds should attach
gap between weld the root backing inside the
preparation edges g > 6 mm groove.
as aforementioned, but
4mm ≤ g ≤ 6mm
71

as aforementioned, but
g < 4mm
36

⑦ Splice in stiffener Δσ should be calculated

As ① in subject to normal stress, using normal stress in the
112 Table
welded from both sides bottom flange of stiffener.
10.4
See ①, ② or ③ in
Tab. 10.4.
As ② in
90 Table
10.4

73
prEN 1993-1-9:2023 (E)

As ③ in
80 Table
10.4

74
 prEN 1993-1-9:2023 (E)

Detail Supplementary
Constructional detail Symbol Description
category Requirements

⑧ Stiffener to deck plate joint Δσ should be calculated

subject to normal stress in using normal stress in
stiffener web accounting for the stiffener web at the
local bending, location of joint.
fillet weld with throat size a ≥ t, Calculation model for
stiffener pre-chamfered, local bending as specified
weld gap g ≤ 0,5mm in EN 1993-2.

63

⑨ Stiffener to deck plate joint Δσ should be calculated

subject to normal stress in using normal stress in
stiffener web, the stiffener web at the
automatic partial penetration location of joint.
100
welds with throat size a ≥ t, Calculation model for
stiffener with weld local bending as speci-
preparation, fied in EN 1993-2.
weld gap g ≤ 0,5mm Melt-through should not
Partial Joint as aforementioned, but occur.
penetration as manual partial penetration weld
90 percentage: with included angle α ≥ 50°

PJP = 100(t–
p/t)
as aforementioned, but
75 ≤ PJP ≤ 95% weld gap 0,5mm ≤ g ≤ 2mm

50

75
prEN 1993-1-9:2023 (E)

Detail Supplementary
Constructional detail Symbol Description
category Requirements

As ① in ⑩ Deck plate splice joint, Δσ should be calculated

112 Table longitudinal or transverse, using normal stress in the
10.4 subject to normal stress, deck plate.
welded from both sides See ①, ② or ③ of
As ② in Table 10.4.
90 Table
10.4

As ③ in
80 Table
10.4

As ⑪ Deck plate to stiffener web Δσ should be calculated

①/112 in joint, using normal stress in the
112
Table subject to normal stress deck plate at the location
10.3 of the joint.
See ①, ③ or ⑥ of
As ③ in Table 10.3.
Table
10.3
100
As ⑥ in
Table
10.3

⑫ Deck plate to crossbeam web Δσ should be calculated

joint, subject to normal stress using normal stress in the
in deck plate, deck plate at the cross-
80 crossbeam web thickness beam web location.
t ≤ 12mm

as aforementioned, but
with t > 12mm

71

NOTE 1 Slope parameter m1 = 3 of fatigue resistance curve unless otherwise stated in detail category.
For ⑦ and ⑩ with symbol note “As ① in Table 10.4” with hot dip galvanizing the next lower detail category of Figure 8.2 should be used.

76
 prEN 1993-1-9:2023 (E)

Table 10.10 — Orthotropic decks – open stiffeners

Detail Supplementary
Constructional detail Symbol Description
category Requirements

① Continuous stiffener to Δσ should be calculated using

crossbeam joint normal stress in the stiffener,
subject to normal stress near the bottom of the
in stiffener, vertical weld between the
80
with extended cut-out in crossbeam and the stiffener
crossbeam, (spot ‘A’).
crossbeam web thickness
t ≤ 12mm
As aforementioned, but
crossbeam web thickness
t > 12mm

71
Key
1 deck plate
2 continuous stiffener
3 cross beam
4 extended cut-out
② Continuous stiffener to See ①.
crossbeam joint as ①, but
close fit cut-out in crossbeam

As ① As ①

③ Continuous stiffener to Δσeq should be calculated

crossbeam joint combining normal stress σ
subject to normal and shear and shear stress τ in the web
stresses in crossbeam, of the crossbeam at the
with weld toe failure critical spot of cross section
‘S’ by

∆σeq =
0,5 [∆σ + (∆σ2 + 4∆τ2)0,5 ]

with
56 σ = MS / Wnet,S
τ = VS / Aw,net,S
with
Wnet,S is net section modulus;
Aw,net,S is net cross-sectional
area of web;
MS internal moment;
VS internal shear force.

77
prEN 1993-1-9:2023 (E)

Detail Supplementary
Constructional detail Symbol Description
category Requirements

④ Continuous stiffener to Δσ should be calculated

crossbeam joint, using weld stress σw
subject to normal due to internal moment at
stress in crossbeam, the critical spot of cross
weld root failure section ‘S’ by

σw = σ t /(2a)

where
σ is normal stress in
crossbeam web, see ③;
t is the thickness of
crossbeam web;
a is the throat size of weld.
36

Key
1 deck plate
2 continuous stiffener
3 cross beam
4 extended cut-out
⑤ Continuous stiffener to Δτ should be calculated
crossbeam joint, using shear stress τw in the
subject to shear stress weld due to internal shear
in crossbeam, force at the critical spot of
weld root failure cross section S by

τw = τ t /(2a)

where
τ is shear stress in
crossbeam web, see ③;
t is the thickness of
crossbeam web;
a is the throat size of weld.
80
m1 = 5

Key
1 deck plate
2 continuous stiffener
3 cross beam
4 extended cut-out
NOTE 1 Slope parameter m1 = 3 of fatigue resistance curve unless otherwise stated in detail category.

78
 prEN 1993-1-9:2023 (E)

Table 10.11 — Crane runway beam related details

Detail Supplementary
Constructional detail Symbol Description
category Requirements

① Flange-to-web connection Δσ should be calculated

subject to wheel loads using vertical
in rolled sections compressive stress in the
web due to wheel loads
according to 7.6.2 of
EN 1993-6.
160 none

② Welded flange-to-web connection

subject to wheel loads,
with continuous full penetration
butt welds

NOTE Effective full penetration butt

100
joints according to EN 1993-1-8 are
considered as partial penetration butt
joints for fatigue.

③ Welded flange-to-web connection Δσ should be calculated

subject to wheel loads, using vertical compres-
with continuous partial penetration sive stress in welds due
butt welds and weld sizes a ≤ 0,7 t to wheel loads according
to 7.6.2 of EN 1993-6.
For great welds with
50 a > 0,7 t, Δσ should be
calculated using stress in
the web.
NOTE For greater
welds, weld toe failure
can additionally occur.

④ Welded flange-to-web connection

subject to wheel loads,
with continuous double fillet weld,
and weld sizes a ≤ 0,7 t

50

⑤ Welded T-section flange to web See ②.

connection subject to wheel loads,
100 with continuous full penetration butt
welds, see Note on ②.

as aforementioned, but See ③ and ④.

with continuous partial penetration
50
butt welds and weld sizes a ≤ 0,7 t

79
prEN 1993-1-9:2023 (E)

as aforementioned, but
with continuous double fillet weld
and weld sizes a ≤ 0,7 t

80
 prEN 1993-1-9:2023 (E)

Detail Supplementary
Constructional detail Symbol Description
category Requirements

⑥ Crane rail as hot-rolled flat or Δσ should be calculated

square steel bar using vertical compres-
subject to wheel loads, sive stress in rail welds
fastened to the top flange due to wheel loads
above supporting webs according to 7.8.2 of
by continuous longitudinal rail welds EN 1993-6
56

⑦ Rail clamps welded to top flange, Δσ should be calculated

not directly subject to fatigue crane using longitudinal
action normal stress in top
flange

L = distance of weld toes [mm]

80 L ≤ 50, all t

71 50 < L ≤ 80, all t

63 80 < L ≤ 100, all t

100 < L ≤ 120, all t

56
L > 120, t ≤ 20

50 120 < L ≤ 200, t > 20

as aforementioned, but
112 all
in single span runway beams
NOTE Slope parameter m1 = 3 of fatigue resistance curve unless otherwise stated in detail category.
For ① with hot dip galvanizing the next lower detail category of Figure 8.1 should be used.

81
prEN 1993-1-9:2023 (E)

Table 10.12 — Constructional details related to towers, masts and chimneys

No. 1 2 3 4 5

Description

Classification Table 10.2: ⑥ Table 10.3: ① Table 10.3: ⑤ Table 10.3: ⑦ Table 10.4: ⑱⑲

No. 6 7 8 9 10

Description

Classification Table 10.4: ⑳ Table 10.4: ⑳ Table 10.5: ① Table 10.5: ① Table 10.5: ②

No. 11 12 13 14 15

Description

See note to table

Classification Table 10.5: ⑦ Table 10.5: ⑪ Table 10.5: ⑦ Table 10.6: ① Table 10.6: ③

No. 16 17 18

Description

Table 10.6: ⑫ Table 10.6: ⑬

Classification Table 10.6: ⑨
See note to table See note to table
NOTE Bolts to be checked with detail No. 1 of this table.

82
 prEN 1993-1-9:2023 (E)

Table 10.13 — Explanation of weld symbols in Tables 10.3 to 10.12

No. Weld type Description Weld Symbol

− welded from both sides,

1
− root of the first weld back gouged or ground

− welded from both sides,

Butt weld with full penetration − root of the first weld back gouged or ground,
2
− weld faces flat and flush with the plate surface,
surface finish parallel to load direction

3 − welded from one side on root backing

− welded from both sides,

4
− root of first weld safely melt through
Butt weld with full penetration
with additional fillet weld passes − welded from both sides,
5
− root of first weld gouged or ground, back weld

6 − welded from one side

Butt weld with partial penetration and

7
additional fillet weld passes

8 − weld toes blended smoothly

Fillet weld, double fillet welds
9

10 Stud welding

NOTE Weld symbols according to EN ISO 2553.

83
prEN 1993-1-9:2023 (E)

Annex A
(normative)

Verification using cumulative linear damage model

A.1 Use of this annex

(1) This Normative annex contains additional provisions to 6.2, 7.3.2 and 9.4 where verification is
undertaken using a cumulative linear damage model.
NOTE This Normative annex can be used independently of the type of verification (nominal, hot spot or
effective notch stress method).

A.2 Scope and field of application

(1) This Normative Annex applies to fatigue design situations where the fatigue action effect is not
defined by an equivalent constant stress range spectrum with ∆σe,2,Ed or Δτe,2,Ed and 2×106 stress cycles.
(2) Multi-axial verification using a cumulative linear damage model (according to A.5) is not
applicable to hot spot and effective notch stress methods.

A.3 Fatigue action effect

A.3.1 Stresses from fatigue actions

(1) The stresses used for fatigue verification according to this Annex may be derived using one of the
fatigue load models in EN 1991, or of any other fatigue load model defined in agreement with
prEN 1990:2021, 6.1.3.3.
(2) The stresses should be calculated according to 7.1, 7.2, B.3.1 and C.3.1.
(3) A stress history should be determined from the loading events for the constructional detail under
consideration taking account of effects of dynamic magnification of the structural response.

A.3.2 Calculation of stress ranges

(1) Stress ranges should be derived from stress histories using the rainflow or reservoir counting
method.
(2) The design value of the stress ranges Δσi,Ed and Δτi,Ed should be determined considering the load
effect from the fatigue load model multiplied by the partial factor γFf, considering the effective value of
the stress range as determined in 7.4, as follows:

Δσi,Ed = Δσ(γFf Qfat)

(A.1)
Δτi,Ed = Δτ(γFf Qfat)

(3) Nominal stress ranges should be determined using 7.3.1 and 7.3.2.
(4) Hot spot stress ranges should be determined using B.3.2.
(5) Effective notch stress ranges should be determined using C.3.2.

84
 prEN 1993-1-9:2023 (E)

A.4 Fatigue resistance

A.4.1 Endurance for the nominal stress method

(1) For constant amplitude loading with normal stress ranges ΔσEd, the design value of endurance,
NRd, should be determined as follows:
ΔσC /𝛾𝛾Mf 𝑚𝑚1
𝑁𝑁Rd = 2 × 106 � Δ𝜎𝜎Ed
� for Δ𝜎𝜎Ed ≥ Δ𝜎𝜎𝐷𝐷 /𝛾𝛾Mf (A.2)

where:

∆σD is the constant amplitude fatigue limit at ND, Figures 8.1 to 8.3
m1 is the first slope parameter of the fatigue resistance curve, Figures 8.1 to 8.3

∆σEd is the design value of the applied stress range, see 7.3
(2) Constant amplitude stress ranges ΔσEd<ΔσD/γMf may be neglected.
(3) For variable amplitude loading with normal stress ranges above and below the constant
amplitude fatigue limit ΔσD, the design value of endurance, Ni,Rd, corresponding to the stress range Δσi,Ed
should be determined for each stress range in the spectrum based on the extended fatigue resistance
curves as follows:
𝑚𝑚1
ΔσC /𝛾𝛾Mf
𝑁𝑁𝑖𝑖,Rd = 2 × 106 � Δ𝜎𝜎𝑖𝑖,Ed
� for Δ𝜎𝜎𝑖𝑖,Ed ≥ Δ𝜎𝜎𝐷𝐷 /𝛾𝛾Mf (A.3)

𝑚𝑚2
Δ𝜎𝜎D /𝛾𝛾Mf
𝑁𝑁𝑖𝑖,Rd = 𝑁𝑁D � � for Δ𝜎𝜎L /𝛾𝛾Mf ≤ Δ𝜎𝜎𝑖𝑖,Ed ≤ Δ𝜎𝜎D /𝛾𝛾Mf (A.4)
Δ𝜎𝜎𝑖𝑖,Ed

where:

∆σD is the constant amplitude fatigue limit at ND, see Figures 8.1 to 8.3

∆σL is the variable amplitude fatigue limit at NL, see Figures 8.1 to 8.3
m1 is the first slope parameter of the fatigue resistance curve, see Figures 8.1 to 8.3
m2 is the slope parameter of the extended fatigue resistance curve, see Figures 8.1 to 8.3

∆σi,Ed is the design value of the applied stress range, see 7.3
(4) Stress ranges in variable amplitude loading with Δσi,Ed<ΔσL/γMf may be neglected.
(5) For either constant or variable amplitude loading with shear stress ranges, the design value of
endurance, Ni,Rd, corresponding with the stress range Δτi,Ed should be determined for each stress range in
the spectrum as follows:
𝑚𝑚1
Δ𝜏𝜏C /𝛾𝛾Mf
𝑁𝑁𝑖𝑖,Rd = 2 × 106 � Δ𝜏𝜏𝑖𝑖,Ed
� for Δ𝜏𝜏𝑖𝑖,Ed ≥ Δ𝜏𝜏D /𝛾𝛾Mf (A.5)

where:

∆τD is the constant amplitude fatigue limit at ND, see Figure 8.4
m1 is the slope parameter of the fatigue resistance curve, see Figure 8.4

∆τi,Ed is the design value of the applied stress range, see 7.3
(6) Stress ranges Δτi,Ed<ΔτD/γMf may be neglected.

85
prEN 1993-1-9:2023 (E)

(7) The size effect due to thickness or other dimensional effects should be considered according to
8.3.1 of this standard.
NOTE Characteristic fatigue resistance values for the nominal stress method without consideration of size
effects are given in Table A.1.

Table A.1 — Characteristic fatigue resistance values in N/mm2 for curves in Figures 8.1 to 8.4

Figure Detail category Slope parameter Characteristic CAFL at Characteristic values [N/mm2]
m1 m2 ND [cycles] At 104 cycles ΔσD ΔσL
8.1 180 5 9 2×106 519 180 116
160 462 160 103
140 404 140 90,6
125 361 125 80,9
112 323 112 72,5
100 288 100 64,7
90 260 90,0 58,3
80 231 80,0 51,8
8.2a 160 3 5 5×106 936 118 64,8
140 819 103 56,7
125 731 92,1 50,6
112 655 82,5 45,3
100 585 73,7 40,5
90 526 66,3 36,4
80 468 58,9 32,4
71 415 52,3 28,7
8.2b 63 3 5 1×107 368 36,8 23,2
56 327 32,7 20,7
50 292 29,2 18,4
45 263 26,3 16,6
40 234 23,4 14,8
36 210 21,1 13,3
8.3 90 5 9 1×107 260 65,2 50,5
71 205 51,5 39,8
56 162 40,6 31,4
50 144 36,2 28,1
45 130 32,6 25,3
35 104 26,1 20,2
8.4a 100 5 ꟷ 1×108 288 45,7 45,7
80 231 36,6 36,6
8.4b 90 8 ꟷ 1×108 162 58,3 58,3

86
 prEN 1993-1-9:2023 (E)

A.4.2 Endurance for the hot spot stress method

(1) When using the hot spot stress method, the design value of endurance Ni,Rd for each stress range
in the spectrum should be calculated using B.4.
NOTE When using B.4, characteristic fatigue resistance values are given in Table A.2.

Table A.2 — Characteristic fatigue resistance values in N/mm2 for curves in Figures B.3 and B.4

Figure Detail category Slope parameter Characteristic CAFL at Characteristic values [N/mm2]
m1 m2 ND [cycles] At 104 cycles ΔσHS,D ΔσHS,L
B.3 112 3 5 5×106 655 82,5 45,3
100 585 73,7 40,5
90 526 66,3 36,4
B.4 193 3,36 5 5×106 933 147 80,8
178 3,30 884 134 73,9
149 3,17 790 111 61,2
128 3,07 717 94,6 52,0
114 3,00 669 84,3 46,3
97 2,90 601 70,4 38,7
88 2,85 566 63,8 35,0
74 2,75 509 53,3 29,3
69 2,72 485 49,2 27,0

A.4.3 Endurance for the effective notch stress method

(1) When using the effective notch stress method the design value of endurance Ni,Rd for each stress
range in the spectrum should be calculated using C.4.
NOTE When using C.4, characteristic fatigue resistance values are given in Table A.3.

Table A.3 — Characteristic fatigue resistance values in N/mm2 for curves in Figure C.2

Figure Detail category Slope parameter Characteristic CAFL at Characteristic values [N/mm2]
m1 m2 ND [cycles] At 104 cycles ΔσENS,D ΔσENS,L
C.2 225 3 5 1×107 1316 132 83,3
200 1170 117 73,8

A.4.4 Endurance for welded joints subjected to High Frequency Mechanical Impact
Treatment

(1) When dealing with welded joints subjected to High Frequency Mechanical Impact Treatment, the
design value of endurance Ni,Rd for each stress range in the spectrum should be calculated using F.4.
NOTE When using F.4, characteristic fatigue resistance values are given in Table F.2 to Table F.7.

87
prEN 1993-1-9:2023 (E)

A.5 Fatigue verification

(1) The verifications with respect to elastic behaviour, considering nominal stresses, given in 9.1
should be performed.
(2) The verifications with respect to fatigue limit given in 9.3 of this standard may be used for infinite
life design.
NOTE Stress ranges ∆σmax,Ed and ∆τmax,Ed in 9.3 are max∆σi,Ed and max∆τi,Ed.

(3) For finite life design, the design value of the damage D accumulated during the design service life
should be calculated from:
𝑛𝑛
𝑁𝑁𝑖𝑖,Ed
𝐷𝐷 = � (A.6)
𝑖𝑖 𝑁𝑁𝑖𝑖,Rd

where
Ni,Ed is the number of cycles in the spectrum corresponding with the stress range Δσi,Ed or Δτi,Ed
Ni,Rd is the design value of endurance for the stress range Δσi,Ed or Δτi,Ed.
NOTE For the example of a four-band ordinary stress range spectrum according to Figure A.1, Figure A.2 gives
the corresponding endurances. The design values of the stress range Δσi,Ed are plotted in these figures.

Figure A.1 — Example of a four-band ordinary stress range spectrum

88
 prEN 1993-1-9:2023 (E)

NOTE

− The blue curve is the extended S-N curve.

− N1,R and N2,R are not provided because the corresponding stress ranges are below VAFL ∆σL.

Figure A.2 — Example of determination of endurances for four bands of the spectrum

(4) The design value of the damage D should satisfy following relationship:
D ≤ 1,0 (A.7)

(5) If the considered location of a constructional detail (see cracks indicated in Table 10.1 to Table
10.12) is subject to a combination of nominal normal and shear stresses, the fatigue verification should
consider their combined effects as follows:
a) If the normal and shear stresses simultaneously occur during each loading event, the principal
stresses should be considered and the corresponding stress ranges should be verified according to
Formula (A.7);
b) If the normal and shear stresses do not simultaneously occur during each loading event, the
components of damage should be added according to Miner’s summation:
• for failure in parent metal (including crack initiation at weld toes) subjected to nominal normal
stress ranges ∆σx, ∆σy, ∆σz and nominal shear stress ranges ∆τxy, ∆τxz, ∆τyz:

∑j=x,y,z 𝐷𝐷𝑗𝑗,σ + ∑k=xy,xz,yz 𝐷𝐷𝑘𝑘,𝜏𝜏 ≤ 1,0 (A.8)

where:
Dj,σ is the damage determined using Formula (A.6) for the ranges of the normal stress
component σj with j = x, y or z
Dk,τ is the damage determined using Formula (A.6) for the ranges of the shear stress
component τk with k = xy, yz or xz

89
prEN 1993-1-9:2023 (E)

• for weld failure (crack initiation at weld root) due to weld stresses according to 7.2(5):

Dσ + Dτ ≤ 1,0 (A.9)

where
Dσ is the damage determined using Formula (A.6) for normal stress ranges, and

Dτ is the damage determined using Formula (A.6) for shear stress ranges.

(6) If normal and shear nominal stresses cause the formation of fatigue cracks at different locations,
a separate fatigue verification for both locations should be performed according to Formula (A.7).

90
 prEN 1993-1-9:2023 (E)

Annex B
(normative)

Hot spot stress method

B.1 Use of this annex

(1) This normative annex contains additional provisions to 6.1(3) for the determination of fatigue
effects and corresponding fatigue resistances for the verification of welded constructional details using
the hot spot stress method.

B.2 Scope and field of application

(1) This normative annex applies only to welded constructional details with a potential crack location
at the weld toe.
NOTE 1 It does not apply for welded constructional details with potential cracks starting from the weld root.
Either the nominal stress method or the effective notch stress method can be used for evaluating constructional
details with a potential crack location at the weld root.

NOTE 2 Guidance on the modelling using FE analysis is given in EN 1993-1-14.

(2) This normative annex supplements and/or modifies the provisions according to Clause 9 and
Annex A for the hot spot stress method.
NOTE The national annex can give conditions for the application of the methods given in this annex. It can give
guidance on which method takes priority when the fatigue verification of a classified constructional detail by
another method gives a different result.

(3) Verification of orthotropic bridge decks using geometrical stresses may be done according to
TS 1993-1-901.
NOTE The National annex can give information on the use of TS 1993-1-901.

B.3 Fatigue action effect

B.3.1 Stresses from fatigue actions

(1) The hot spot stress should be calculated assuming linear-elastic material behaviour and an
idealized geometry.
NOTE Guidance on the constructional detail geometry is given in EN 1993-1-14.

(2) The distinction between hot spot type “a”, type “b” and type “c” in Figure B.1 should be made for
the determination of the hot spot stress to account for the stress distribution through thickness of the
plate or wall thickness of the structural hollow section joint with the potential fatigue crack.
(3) Hot spot type “a”, indicating cracking at the weld toe on a plate surface or on a section, should be
considered when the stress considerably varies through thickness of the plate with the potential fatigue
crack.
(4) Hot spot type “b”, indicating cracking at a plate edge or a section, should be considered if the stress
at the potential fatigue crack location (hot spot) is independent of the plate or cross-sectional wall
thickness.

91
prEN 1993-1-9:2023 (E)

(5) Hot spot type “c”, indicating cracking in a structural hollow section, should be considered if the
stress at the potential fatigue crack location (hot spot) is within a lattice girder joint made of structural
hollow sections.

Key
1 Chord
2 Brace
3 Crown heel
4 Crown toe
5 Saddle

Figure B.1 — Definition of hot spot type “a”, “b” and “c”

(6) When calculating the hot spot stress type “c”, only the stress component perpendicular to the
weld toe should be considered.
NOTE In case of hot spot type “a” and “b”, further guidance is given in B.3.2(6) to (8).

(7) For constructional details covered by Table B.1 and Table B.2, the stresses should be calculated
as given in EN 1993-1-14.

92
 prEN 1993-1-9:2023 (E)

B.3.2 Calculation of stress ranges

(1) The design value of the hot spot stress ranges ΔσHS,Ed should be determined from the load effect
of the fatigue load model multiplied by the partial factor γFf.
(2) For fatigue design situations where the fatigue action effect is defined by an equivalent stress
range, the design value of the hot spot stress range Δσe,2,HS, Ed should be determined as follows:
Δσe,2,HS, Ed = λ1 λ2 … λn ΔσHS (γFf Qfat) (B.1)

where

ΔσHS (γFf Qfat) is the design value of the hot spot stress range caused by the fatigue actions specified
in EN 1991
λi are damage equivalent factors depending on the spectra as specified in the relevant
parts of EN 1993 and EN 1994, i = 1 to n
(3) For any other kind of stress range spectrum, the design value of each stress range Δσi,HS,Ed in the
spectrum should be determined in accordance with the method for Δσe,2,Ed using A.3.2(2).
(4) In welded details which have been subjected to full thermal stress relief after all welding
operations have been completed, the mean stress influence on the fatigue life may be taken into account
by considering a reduced design value of each stress range in the design spectrum where part, or all, of
the stress cycle is compressive as specified in 7.4.
(5) For calculation of hot spot stress range in hot spot types “a” and “b”, the principal stress range
with the largest absolute stress difference should be used if its orientation is closer than α = ±45˚ from
that of the normal to the weld line as shown in Figure B.2.
NOTE When the principal stresses have the same sign, the maximum principal stress range can be used.

(6) Stress σ1 should be replaced by σ2 in Figure B.2 case of a compressive nominal stresses in the
plate.

93
prEN 1993-1-9:2023 (E)

(a) For α ≤ 45° (b) For α > 45°

σHS = σ1 σHS = σ⊥

Key
σ1 : first principal stress
σ2 : second principal stress
σ⊥ : stress perpendicular to the weld toe
1 Main plate
2 Attached plate
3 Weld

Figure B.2 — Stress components for hot spot stress ranges, planar view of a plate surface (in case
of biaxial stress, σ1 and σ2 can be time-dependent)

(7) Where the orientation of the principal stress with the largest absolute stress difference does not
satisfy the requirement in (5) and (6), the normal stress component acting perpendicular to the weld line
should be used. Other stress components (e.g. parallel to the weld line) should be checked using the
nominal stress method.
(8) Paragraphs (5) to (7) should also be applied in situations where the directions of principal
stresses change during the load cycle. In this case, the principal stress range should be determined as the
greatest algebraic difference between principal stresses planes no more than 45° apart.
(9) The calculated hot spot stress range should consider misalignments that exceed the more onerous
of the tolerances indicated in the detail category tables and EN 1090-2. In cases of eccentricities in joints,
the nominal value of eccentricity should be used.
NOTE To account for macro-geometric effects not included in the hot spot detail category, the stress range can
be calculated directly by a FE analysis or indirectly accounted for by means of magnification factor kf that is available
for different geometries in Annex D or in the literature. The hot spot stress range can be calculated considering only
the extent of misalignment that exceeds the more onerous of the tolerances indicated in the detail category tables
and EN 1090-2.

B.4 Fatigue resistance

B.4.1 Fatigue resistance curves

(1) The characteristic fatigue resistance curves for hot spot stress ranges in Figure B.3 should be used
for details in plated structures and for details in sections, and combinations of sections. The characteristic
fatigue resistance curves for hot spot stress ranges in Figure B.4 should be used for lattice girder joints
made of structural hollow sections .
NOTE Further information on resistance curves for lattice girder joints made of structural hollow sections can
be found in ISO 14347.

94
 prEN 1993-1-9:2023 (E)

(2) The fatigue resistance curves for hot spot stress ranges under constant and variable amplitude
loading should be determined in accordance with the recommendation in 8.2(1).
NOTE A summary of the characteristic fatigue resistance values can be found in Annex A.
Hot spot stress range ∆σHS [N/mm²]

ND=5×106
∆σD=0,737∆σC
NL=108
∆σL=0,405∆σC

Number of cycles N
∆𝜎𝜎𝐻𝐻𝐻𝐻,𝐷𝐷 = 0,737∆𝜎𝜎𝐻𝐻𝐻𝐻,𝐶𝐶 ∆𝜎𝜎𝐻𝐻𝐻𝐻,𝐿𝐿 = 0,405∆𝜎𝜎𝐻𝐻𝐻𝐻,𝐶𝐶 (values are given in Table A.2)

NC = 2×106 cycles, ND = 5×106 cycles, NL = 108 cycles

Figure B.3 — Characteristic fatigue resistance curves for hot spot stress ranges
in conjunction with Table B.1, types “a” and “b”

95
prEN 1993-1-9:2023 (E)

Hot spot stress range ∆σHS [N/mm²]

Number of cycles N
∆𝜎𝜎𝐻𝐻𝐻𝐻,𝐷𝐷 , ∆𝜎𝜎𝐻𝐻𝐻𝐻,𝐿𝐿 : values are given in Table A.2

NC = 2×106 cycles, ND = 5×106 cycles, NL = 108 cycles

Figure B.4 — Characteristic fatigue resistance curves for hot spot stress ranges
in conjunction with Table B.2 (type “c”, lattice girder joints made of structural hollow sections)

B.4.2 Classification of constructional details

(1) For the application of the hot spot stress method, constructional details should be classified in
accordance with the following tables:
• Table B.1 for details with potential fatigue cracks at the weld toes of:

• butt welds,
• fillet welded attachments,
• fillet welds in cruciform joints.

• Table B.2 for details with cracks at weld toes of lattice girder joints made of structural hollow sections.

NOTE The hot spot stress reference detail method can be used as an alternative to classify a constructional
detail. Further guidance is given in Annex G.

(2) The classifications of fillet welded details may be applied to convex, concave or flat weld caps,
although the illustrations only show one indicative weld profile type.

96
 prEN 1993-1-9:2023 (E)

Table B.1 — Detail categories for use with hot spot stress method, types “a” and “b”

Detail Weld Supplementary

β1) Constructional detail Description
category symbol Requirements

① Full penetration Weld all-around

butt joint, ground flush to plate
welded from both surface in direction of
sides, stress, where
ground flush. relevant, after
removing weld run-
112 0,1 off pieces
Only for Misalignment due to
evaluation of fabrication to be
misalignment neglected if ≤ 5% of
effects plate thickness, see
also B.3.2(9).

② Full penetration Fabrication: Welding

butt joint, position PA.
welded from both Smooth transition of
sides, weld to plate surface
Plate edges ground
In case of butt- parallel to direction of
welded splices in stress, where
structural hollow relevant, after
sections, welded removing weld run-
100 0,2 from one side. off pieces
Misalignment due to
fabrication to be
neglected if ≤ 5% of
plate thickness, see
also B.3.2(9).
Modelling: FE model
should include weld
convexity of weld cap.

③ Cruciform joint Weld flank angle ≥

with full 120°,
penetration see Note 2.
butt welds, The eccentricity of the
welded from both load carrying plates
100 0,3 sides due to fabrication to
be neglected if ≤15%
of the thickness of the
intermediate plate,
see also B.3.2(9)

97
prEN 1993-1-9:2023 (E)

Detail Weld Supplementary

β1) Constructional detail Description
category symbol Requirements

④ Cruciform joint Fabrication: Weld

with load- flank angle ≥ 120°,
carrying partial see Note 2.
penetration butt The eccentricity of the
welds and fillet load carrying plates
welds due to fabrication to
be neglected if ≤15%
of the thickness of the
intermediate plate,
see also B.3.2(9)
90 0,3 Modelling: FE model
with full connectivity
between the plates
independent of the
weld type.
For partial
penetration butt
welds or fillet welds,
weld root failure
should be checked
using ③ of Tab. 10.6,
see Note 3.

⑤ Non load-carrying Weld flank angle ≥

fillet welds 120°.
100 0,3

⑥ Cover plate ends Weld flank angle ≥

and similar joints. 120°.
100 0,3

⑦ Type “b” joint Fillet or full

with longitudinal penetration weld.
attachment.
100 -

1)The thickness correction only applies to hot spot type a, see B.4.3.
NOTE 1 Fatigue resistance curve with slope parameter m1 = 3 unless otherwise stated in detail category
NOTE 2 For load carrying attachments with partial penetration butt welds or fillet welds, for weld sizes a > 0,7 t weld
root failure should also be checked.
NOTE 3 Effective full penetration butt welds according to EN 1993-1-8 are considered as partial penetration butt
welds in respect of fatigue.

98
 prEN 1993-1-9:2023 (E)

Table B.2 — Detail categories for use with hot spot stress method for lattice girder joints made
of structural hollow sections, type “c”, including a thickness correction

Detail Thickness
Constructional detail Requirements
category t1) [mm]
193 ①, ② and ③
4
m1=3,36 Stop-starts for non-continuous welds should
not be located at points of high stress
178
5 ① CHS and RHS joints concentration.
m1=3,30
② and ③
149 Partial or full penetration welds are
8
m1=3,17 recommended.
128 Fillet welds only if a ≥ t and if local dihedral
12 angle at the intersection between the brace
m1=3,07 and the chord surfaces does not exceed 120°.
② CHS and RHS joints
114 For further guidance, see ISO 14347.
16
m1=3,00
97
25
m1=2,90
88
32
m1=2,85
③ CHS joints
74
50
m1=2,75
69
61
m1=2,72
1) The
thickness of the member, brace or chord, in which the potential crack develops.
NOTE Fatigue resistance curves have variable slopes, Figure B.4 and Formulae (B.2) and (B.3).

B.4.3 Fatigue resistance modification

(1) For hot spot type “a”, where the plate or wall thickness is greater than the reference thickness,
the stress range should be modified by the application of a thickness factor that should be determined as
follows:
𝑡𝑡ref 𝛽𝛽
𝑘𝑘𝑠𝑠 = � � for t > tref (B.2)
𝑡𝑡eff

where:
tref reference thickness equal to 25mm for plated structures and for details in sections and
combinations of sections. For structural hollow sections (details with geometries as in Table 10.7
only), the reference thickness is 16mm
t thickness of element through which a potential crack can develop
β thickness exponent on fatigue resistance.
teff effective thickness
(2) The effective thickness for details of Table B.1 should be calculated as follows:

99
prEN 1993-1-9:2023 (E)

• For details 1 and 6: 𝑡𝑡eff = t

• For details 2, 3, 4, 5: 𝑡𝑡eff = min (14 + 0,66ℓ; 𝑡𝑡) for teff ≥ tref (B.3)where the parameters where
the parameters teff,  and t are measured in mm and  and t are defined in Table B.1 and Figure B.5.

Figure B.5 — Definition of attachment length in details 3, 4, 5 and weld width in details 1, 2 of
Table B.2.

B.5 Fatigue verification

(1) Within a constructional detail with different potential crack locations, fatigue verification may
require the use of different design stress methods. In those cases, each fatigue verification shall be made
with reference to the value of the fatigue resistance relevant for that method.
NOTE For example, potential weld root cracking under shear stresses is verified according to the nominal
stress method, while potential weld toe cracking under normal stresses is verified according to the hot spot stress
method of this annex.

(2) For verification with the equivalent constant stress range Δσe,2,HS,Ed, the hot spot stress range of a
constructional detail should satisfy following relationship:
Δ𝜎𝜎e,2,HS,Ed
≤ 1,0 (B.4)
Δ𝜎𝜎HS,C /𝛾𝛾Mf

(3) Where no data for Δσe,2,HS,Ed is available the verification format in Annex A may be used.
(4) An infinite life for constructional details may be assumed if the maximum stress range of the
applied hot spot stress spectrum Δσmax,HS,Ed satisfies following relationship:
Δ𝜎𝜎max, HS,Ed
Δ𝜎𝜎HS,D /𝛾𝛾Mf
≤ 1,0 (B.5)

100
 prEN 1993-1-9:2023 (E)

Annex C
(normative)

Effective notch stress method

C.1 Use of this annex

(1) This normative annex contains additional provisions to 6.1(3) for the determination of fatigue
load effects and corresponding fatigue resistances for the verification of welded constructional details
using the effective notch stress method.

C.2 Scope and field of application

(1) This normative annex applies to welded constructional details with potential crack location at the
weld toe or the weld root.
NOTE Guidance on the calculation of effective notch stress using FE analysis is given in EN 1993-1-14.

(2) This normative annex only covers welded constructional details with plate thicknesses t ≥ 5 mm.
(3) This normative annex does not cover fatigue verification of parent metal away from weld toes or
of welded details with mild notches.
NOTE A mild notch is one for which the effective notch stress is less than 2 times the nominal stress. This can
e.g. be the case in transverse butt welds with small or no weld convexity and small eccentricity.

(4) This normative annex does not cover fatigue verification of welded details with potential crack
locations at the root or inner weld imperfections respecting the weld quality levels given in 8.2(2).
(5) This normative annex supplements and/or modifies the provisions according to Clause 9 and
Annex A for the effective notch stress method.
NOTE The national annex can give conditions for the application of the methods given in this annex. It can give
guidance on which method takes priority when the fatigue verification of a classified constructional detail by
another method gives a different result.

C.3 Fatigue action effect

C.3.1 Stresses from fatigue action

(1) The effective notch stress should be calculated assuming linear-elastic material behaviour and an
idealized weld geometry.
(2) Butt welds should be modelled with a weld toe angle of θ = 30˚, see Figure C.1. Fillet welds should
be modelled with a weld toe angle of θ = 45˚. The cases designed for other angles should be modelled with
the nominal values of these angles.
NOTE The weld toe angle θ can be calculated as the complementary angle of the weld toe flank angle.

(3) The effective notch stress should be obtained by rounding the weld toe or root with a notch of
radius, r = 1 mm, see Figure C.1.
NOTE Guidance on the rounding of the weld toe and root is given in EN 1993-1-14.

101
prEN 1993-1-9:2023 (E)

a) Rounding of a butt weld b) Rounding of a fillet weld

Figure C.1 — Weld toe angle and rounding of weld toe for different types of welded details

C.3.2 Calculation of stress ranges

(1) The design value of the effective notch stress ranges ΔσENS,Ed should be determined considering
the load effect from the fatigue load model multiplied by the partial factor γFf.
(2) For fatigue design situations where the fatigue action effect is defined by an equivalent stress
range, the design value of the effective notch stress range, Δσe,2,ENS,Ed, should be determined as follows:
Δσe,2,ENS,Ed = λ1 λ2 λi ... λn ΔσENS (γFf Qfat) (C.1)

where:

ΔσENS (γFf Qfat) is the design value of the effective notch stress range caused by the fatigue actions
specified in EN 1991
λi are damage equivalent factors depending on the spectra as specified in the relevant
parts of EN 1993 and EN 1994, i = 1 to n
(3) For any other kind of stress range spectrum, the design value of each stress range Δσi,ENS,Ed in the
spectrum should be determined in accordance with the method for Δσe,2,Ed using 7.3.2(1) and the
verification method of Annex A should be used.
(4) In welded details which have been subjected to full thermal stress relief after all welding
operations have been completed, the mean stress influence on the fatigue life may be taken into account
by considering a reduced design value of each stress range in the design spectrum where part, or all, of
the stress cycle is compressive as specified in 7.4.
(5) The effective notch stress range may be calculated using either:
• Principal stress (PS), as the maximum difference of the principal stress range,

• Equivalent von Mises (EVM): as the maximum von Mises equivalent stress range calculated from the
range of notch stress components.

(6) The calculated effective notch stress range should consider misalignments that exceed the more
onerous of the tolerances indicated in the detail category tables and EN 1090-2. In cases of eccentricities
in joints, the nominal value of eccentricity should be used.
NOTE To account for macro-geometric effects, the effective notch stress range can be calculated directly by a
FE analysis or indirectly accounted for by means of stress concentration factor kf that is available for different
geometries in Annex D or in the literature. The effective notch stress range can be calculated considering only the
extent of misalignment that can exceed the more onerous of the tolerances indicated in the detail category tables
and EN 1090-2.

102
 prEN 1993-1-9:2023 (E)

C.4 Fatigue resistance

C.4.1 Fatigue resistance curves

(1) The characteristic fatigue resistance curves for effective notch stress ranges in Figure C.2 should
be used for details calculated using the PS or the EVM.
(2) The fatigue resistance curve under constant and variable amplitude loading should be
determined in accordance with the principle for Δσ using 8.1(1).
Effective notch stress range ∆σENS [N/mm²]

ND=107
∆σD=0,585∆σC
NL=108
∆σL=0,369∆σC

Number of cycles N
∆𝜎𝜎𝐸𝐸𝐸𝐸𝐸𝐸,𝐷𝐷 = 0,585∆𝜎𝜎𝐸𝐸𝐸𝐸𝐸𝐸,𝐶𝐶 ∆𝜎𝜎𝐸𝐸𝐸𝐸𝐸𝐸,𝐿𝐿 = 0,405∆𝜎𝜎𝐻𝐻𝐻𝐻,𝐶𝐶 (values are given in Table A.3)

NC = 2×106 cycles, ND = 107 cycles, NL = 108 cycles

Figure C.2 — Characteristic fatigue resistance curves for effective notch stress ranges (PS and
VM methods)

C.4.2 Classification of constructional details

(1) The constructional detail categories in Table C.1 should be used with the effective notch stress
ranges and the relevant calculation of stresses (with PS or EVM).

103
prEN 1993-1-9:2023 (E)

Table C.1 — Detail categories for use with effective notch stress method

Detail Stress components Supplementary

Constructional detail
category / Description requirements
PS. For misalignment
Notch as-welded. see C.3.2(6).

225 Effective notch

radius equal to 1mm
replacing weld toe
and weld root notch
EVM.
Notch as-welded.
200 Effective notch
radius equal to 1mm
replacing weld toe
and weld root notch
NOTE Fatigue resistance curve with slope m1 = 3

C.5 Fatigue verification

(1) The equivalent constant effective notch stress range should be compared with the fatigue relevant
resistance curve (PS or EVM).
(2) It should be checked that the fatigue resistance of the parent metal is not exceeded in the direct
vicinity of the weld (e.g. using the nominal stress method and the fatigue strength relevant for that
method).
(3) For verification with the equivalent constant stress range Δσe,2,ENS,Ed, the effective notch stress
range of a constructional detail should satisfy following relationship:
Δ𝜎𝜎e,2,ENS,Ed
Δ𝜎𝜎ENS,C /𝛾𝛾Mf
≤ 1,0 (C.2)

(4) Where no data for Δσe,2,ENS,Ed is available the verification format in Annex A may be used.
(5) An infinite life for constructional details may be assumed if the maximum stress range of the
applied effective notch stress spectrum Δσmax,ENS,Ed satisfies following relationship:
Δ𝜎𝜎max, ENS,Ed
Δ𝜎𝜎ENS,D /𝛾𝛾Mf
≤ 1,0 (C.3)

104
 prEN 1993-1-9:2023 (E)

Annex D
(informative)

Recommendations for magnification factors k1 and stress

concentration factors kf

D.1 Use of this annex

(1) This informative annex provides supplementary guidance to 7.3.2 and 8.3 for magnification
factors k1 and stress concentration factors kf.
NOTE National choice on the application of this informative annex is given in the national annex. If the national
annex contains no information on the application of this informative annex, it can be used.

D.2 Scope and field of application

(1) This informative annex covers magnification factors k1 and stress concentration factors kf for the
calculation of modified nominal stresses.

D.3 Secondary moments in lattice girders

(1) For lattice girders made of hollow sections the modelling may be based on a simplified truss
model with continuous chords and pinned braces. Provided that the stresses due to external loading
applied to members between joints are taken into account, the effects from secondary moments due to
the stiffness of the connection may be considered by the use of k1 factors according to:
• Table D.1 for circular hollow sections;

• Table D.2 for rectangular hollow sections;

where these sections are subject to the geometrical restrictions according to Table 10.8.

Table D.1 — k1 factors for circular hollow sections under in-plane loading

Type of joint Chords Verticals Diagonals

K type 1,5 – 1,3
Gap joints
N type / KT type 1,5 1,8 1,4
K type 1,5 – 1,2
Overlap joints
N type / KT type 1,5 1,65 1,25

Table D.2 — k1 factors for rectangular hollow sections under in-plane loading

Type of joint Chords Verticals Diagonals

K type 1,5 – 1,5
Gap joints
N type / KT type 1,5 2,2 1,6
K type 1,5 – 1,3
Overlap joints
N type / KT type 1,5 2,0 1,4

105
prEN 1993-1-9:2023 (E)

NOTE For the definition of joint types see EN 1993-1-8.

D.4 Flanges of ⌶-section girders with transitions in thickness or width

(1) For the details with transverse butt welds (details ⑱, ⑲, ⑳, and of Table 10.4, also Figure
D.1 and Figure D.2) the recommended stress concentration factor kf is defined as follows:
𝑘𝑘f = 𝑘𝑘f,N 𝑘𝑘f,m (D.1)

where:
kf,N is the stress concentration factor for the plate membrane stress effect
kf,m is the stress concentration factor for the plate bending stress effect

a) tapering to the outside b) tapering to the inside c) centric tapering

(both sides tapered)

Figure D.1 — Flange thickness tapering in girders

Figure D.2 — Flange width tapering in girders

(2) The stress concentration factor kf,N for the plate membrane stress effect is determined as follows:
• Details ⑱, ⑲, ⑳, with thickness tapering to the outside or to the inside or with centric tapering
(Figure D.1 a), b) and c)) and detail :
𝜎𝜎 𝑡𝑡
𝑘𝑘f,N = 1 + 𝛼𝛼 �𝜎𝜎2 𝑡𝑡2 − 1� (D.2)
1 1

106
 prEN 1993-1-9:2023 (E)

where

α = 0,5 for details ⑱, ⑲, ⑳, with thickness tapering to the outside or with centric
tapering and detail ;
α = 0,33 for details ⑱, ⑲, ⑳ and with thickness tapering to the inside.
• Details ⑱, ⑲, ⑳ and with width tapering (Figure D.2):
𝜎𝜎2 𝑏𝑏2
𝑘𝑘f,N = 1 + 0,5 � − 1� (D.3)
𝜎𝜎1 𝑏𝑏1

NOTE The stresses σ1 and σ2 are calculated along the centreline of the flanges. The resultant normal forces in
the flanges are not identical due to the displacement of cross-section centroids.

(3) The stress concentration factor kf,m for the plate bending stress effect, due to the eccentricity of
the plates of the girders, should be determined as follows:
• Details ⑱, ⑲, ⑳, with thickness tapering to the outside or to the inside (Figure D.1 a) and b)):
𝑡𝑡
1 𝜆𝜆 � 2−1�
𝑡𝑡1
𝑘𝑘f,m = �1 + 𝑡𝑡 𝑛𝑛
� ≥ 1,0 (D.4)
𝛽𝛽 2 �1+� 2� �
𝑡𝑡1

where:
β = 1,25 (recommended value)
𝜆𝜆 = 4 + (𝑐𝑐⁄𝑡𝑡1 )/10 ≤ 6
• Details ⑱, ⑲, ⑳, with centric tapering (Figure D.1 c)), details ⑱, ⑲, ⑳, with width tapering
(Figure D.2) and detail :

kf,m = 1,0 (D.5)

(4) The recommended stress concentration factors should only be applied if the following
requirements for the throat weld between web and flange are fulfilled:
• Girders have no cope holes in the web in proximity of the transverse butt joint of the flange.

NOTE For girders with cope holes the above recommended stress concentration factors do not apply.

• Where the change of normal forces in the flanges is not considered in the design of the longitudinal
weld, longitudinal full penetration butt welds or double fillet welds with a ≥ tw/2 within a distance of
0,5 h on each side of the transverse butt weld are provided for.

NOTE The change of normal force in the flange is b (σ2 t2 – σ1 t1) for the cases covered by Figure D.1 and (b2 σ2
t2 – b1 σ1 t1) for those covered by Figure D.2.

• Where the additional vertical stresses σz in the longitudinal weld are not considered in the design,
longitudinal full penetration butt welds within a distance of 0,5 h on each side of the transverse butt
weld are provided for.

NOTE Additional vertical stresses σz can occur in cases with flange thickness tapering in girders to the inside
(Figure D.1 a)) or to the outside (Figure D.1 b)).

107
prEN 1993-1-9:2023 (E)

D.5 Thickness transitions in plates

(1) For the details with transverse butt welds of plates with tapered thickness transitions and no
vertical support of the plates (details ⑱, ⑲, ⑳, of Table 10.4, also Figure D.3) the stress
concentration factor kf should be determined as follows:
6𝑒𝑒 𝑡𝑡11,5
𝑘𝑘f = �1 + � (D.6)
𝑡𝑡1 𝑡𝑡11,5 +𝑡𝑡21,5

where
e is the eccentricity of centrelines.

Figure D.3 — Tapered thickness transitions without vertical support of the plates

D.6 Shell structures

(1) Stress concentration factors for shell structures may be derived using prEN 1993-1-6.

108
 prEN 1993-1-9:2023 (E)

Annex E
(informative)

Recommendations for preloaded bolts and rods subject to tension

E.1 Use of this annex

(1) This informative annex provides supplementary guidance to 7.2 on the principles of calculation
of the forces (neglecting partial factors) in preloaded bolts and rods subject to tension.
NOTE National choice on the application of this informative annex is given in the national annex. If the national
annex contains no information on the application of this informative annex, it can be used.

E.2 Scope and field of application

(1) This informative annex covers a simplified calculation method for preloaded bolts or rods under
a centrically applied external tensile force in order to account for the beneficial influence of preloading
under fatigue action effects.

a) dimensions b) spring model c) spring model d) spring model with

before after preloading forces due to
preloading external force

Key
1 volume of clamped components under compression due to preloading
2 spring representing stiffness of clamped components
3 spring representing stiffness of bolt or rod

Figure E.1 — Preloaded bolted joint

109
prEN 1993-1-9:2023 (E)

NOTE 1 The simplified calculation method using a spring model calculates the distribution of forces between
bolt or rod and clamped components in a preloaded bolted joint or joint with rod under a centrically applied
external tensile force, see Figure E.1 a). In the absence of an external force, due to the preloading of the joint, the
clamped components are exposed to compression and the bolt to tension, see Figure E.1 c). If an external tensile
force FT is applied to the joint, it is distributed between the bolt or rod and the clamped components based on the
ratio of elastic stiffness (inverse of flexibility) illustrated as springs in Figure E.1 b) to d). Due to the external force,
the compression of the clamped components due to preloading reduces by FC while the tensile force of the bolt or
rod due to preloading increases by FB, see Figure E.1 d). As the spring representing the clamped components has a
greater stiffness it carries a relatively greater portion of the applied external force than the bolt or rod. This is
favourable in cases where FT is a fatigue action effect.

NOTE 2 EN 1993-1-14 can be used for a more accurate analysis based on finite elements.

E.3 Simplified calculation method

(1) For bolts or rods with preloading force FP,c, the simplified calculation method may be taken into
account provided that the following condition applies to the reduction of the compressive force FC on the
clamped components due to a external tensile force FT centrally applied to the joint (Figure E.1 d)),
𝐹𝐹𝐶𝐶 ≤ 𝐹𝐹P,c, (E.1)

where:
FP.c is the preloading force of the bolt or rod
FC is the reduction of the compressive force of the clamped components due to FT
FT is the external tensile force applied to the preloaded bolted joint or joint with rod
(2) The additional tensile force FB on the bolt or rod and the reduction of the compressive force FC on
the clamped components due to the external tensile force FT may be determined as follows:
𝛿𝛿c
𝐹𝐹B = 𝐹𝐹
𝛿𝛿b +𝛿𝛿c T
(E.2)

𝛿𝛿b
𝐹𝐹C = 𝐹𝐹 (E.3)
𝛿𝛿b +𝛿𝛿c T

where:
δb is the flexibility of the bolt or rod
δc is the flexibility of clamped components
(3) The flexibility of the bolt or rod δb and the flexibility of the clamped components δc may be
determined as follows (Figure E.1 a)):
1 𝐿𝐿 𝐿𝐿
𝛿𝛿b = 𝐸𝐸 (1,1𝐴𝐴 )
+ 𝐸𝐸 2𝐴𝐴 (E.4)
S

𝐿𝐿
𝛿𝛿c = (E.5)
𝐴𝐴𝑝𝑝 𝐸𝐸

110
 prEN 1993-1-9:2023 (E)

where:
L1 is the bolt length without thread according to Figure E.1 a)
L2 is the bolt length with thread according to Figure E.1 a)
L is the total thickness of clamped components including washers (Figure E.1 a))
AS is the stress area of the bolt or rod according to EN 1993-1-8
A is the area of the shank of the bolt or rod
Ap is the base area of compressed volume of the clamped components due to preload (truncated
cone)
(4) The base area of compressed volume of the clamped components (truncated cone in Figure E.1 a))
due to preload may be determined as follows:
𝜋𝜋 𝐿𝐿 2
𝐴𝐴p = ��𝑠𝑠 + � − 𝑑𝑑02 � (E.6)
4 10

where:
s is the width across flats (wrench size)
d0 is the hole diameter of the bolt or rod
L is the total thickness of clamped components including washers (Figure E.1 a))
NOTE Instead of Formulae (E.4) and (E.5), more accurate bolt and plate assembly flexibilities δb and δc can be
obtained according to VDI 2230 or using a model of the structure.

(5) Where bearing surfaces between clamped plates are not flat the effects of plate assembly
flexibility should be taken into account.
NOTE Plate assembly flexibility can be determined according to VDI 2230 or using a model of the structure.

111
prEN 1993-1-9:2023 (E)

Annex F
(informative)

Fatigue design of welded joints subjected to High Frequency

Mechanical Impact Treatment

F.1 Use of this annex

(1) This informative annex provides supplementary guidance to 6.1(3) and 8.4.2(1) for taking into
account the High Frequency Mechanical Impact Treatment carried out on welded joints in the verification
of the fatigue design situation.
NOTE The national choice on the application of this informative annex is given in the national annex. If the
national annex contains no information on the application of this informative annex, it can be used.

F.2 Scope and field of application

(1) This informative annex applies to post-weld treatment of the constructional details in Table F.1
through qualified High Frequency Mechanical Impact (HFMI) technologies whose beneficial effect on the
fatigue resistance may be taken into account for the fatigue verification.
NOTE 1 The covered constructional details are those with potential crack location at the weld toes in as-welded
state. Therefore, a post-weld treatment of the weld toes affects the fatigue resistance beneficially.

NOTE 2 The applied stress ratio affects the behaviour of HFMI treated details.

(2) This informative annex covers qualified post-weld treatment technologies and the resulting
improved fatigue resistances from such treatment.
NOTE 1 Qualified post-weld treatment technologies whose effectiveness has been proved by experimental
research are (as listed in 3.1.4.13): HiFIT (High Frequency Impact Treatment), PIT (Pneumatic Impact Treatment),
UIT (Ultrasonic Impact Treatment).

NOTE 2 Other treatment technologies do not fall in the scope of this annex unless the national annex gives
different provisions.

(3) This informative annex covers improved fatigue resistance following the requirements for
application given in F.6.
(4) This informative annex only covers welded constructional details with plate thicknesses t ≥ 5 mm.
(5) This informative annex applies to steels covered by EN 1993-1-1; it does not apply to weathering
steels according to EN 10025-5 and stainless steels according to EN 10088.
NOTE Conditions for the application on other steels and welded detail can be specified by the national annex.

(6) This informative annex applies for fatigue verification that is carried out according to the
provisions given in Clause 9 and in Annex A.

112
 prEN 1993-1-9:2023 (E)

Table F.1 — Constructional details

Design of the constructional detail Weld symbol Description

Transverse stiffener
Tab. 10.5,
Details ⑦ to ⑩
(not all cases presented)

Transverse butt weld

Tab. 10.4,
Details ②, ③, ⑲ and

(not all cases presented)

Longitudinal stiffener
Tab. 10.5,
Details ① and ②
L = attachment length
(not all cases presented)

F.3 Fatigue action effect

F.3.1 Stresses from fatigue actions

(1) The applied stresses should be calculated in accordance with the requirements in 7.1.
(2) The applied stresses should be calculated including the static load effects from the frequent
combination of actions.
NOTE The fatigue resistance tables in F.4.2 contain limitations of applied stresses

F.3.2 Calculation of the stress ranges

(1) For fatigue design situations where the fatigue action effect is defined by an equivalent stress
range, the design value of the nominal stress range Δσe,2,HFMI,Ed should be determined as follows:
Δσe,2,HFMI,Ed = λ1 λ2 … λn λHFMI Δσ (γFf Qfat) (F.1)

and λ1 λ2 … λn≤ λmax

113
prEN 1993-1-9:2023 (E)

where:

Δσ(γFf Qfat) is the design value of the normal stress range caused by the simplified fatigue load model
specified by EN 1991 (= σFLM,max – σFLM,min)
λi are damage equivalent factors depending on the spectra as specified in the relevant parts
of EN 1993 and EN 1994, i = 1, 2, … , n
λmax is the maximum λ-value taking into account of the fatigue limit, as specified in the
relevant parts of EN 1993 and EN 1994
λHFMI is the factor to consider the mean stress effect in spectrum, see F.7
NOTE The national annex can give alternative methods to calculate Δσe,2,HFMI,Ed or λHFMI. The following
conservative value may be recommended λHFMI = 2,0.

(2) In case of fatigue action effect expressed as a load spectrum, the treatment of variable amplitude
loading is given in F.7.

F.4 Fatigue resistance

F.4.1 Fatigue resistance curves

(1) The characteristic fatigue resistance curves in Figure F.1 should be used for HFMI treated details.
NOTE For numbers of cycles N < 5×106, the improved fatigue resistance curve of HFMI treated constructional
details has a slope parameter mHFMI = 5. For numbers of cycles 5×106 < N ≤ 108, the improved fatigue resistance
curve of HFMI treated constructional details has a slope parameter mHFMI,D = 9 (Figure F.1).

(2) For stress range Δσs and corresponding Nmin,HFMI, the fatigue resistance curve for as-welded state
and for HFMI treated state intersect in Figure F.1. The HFMI fatigue resistance curves should only be used
for stress range below Δσs/γMf or number of cycles N > Nmin,HFMI.
(3) For stress range above or equal to Δσs/γMf or for N ≤ Nmin,HFMI, the as-welded fatigue resistance
should be used (no improvement of the fatigue resistance).
(4) The minimum cycles Nmin,HFMI and the related stress range Δσs provided by Tables F.3, F.5 and F.7
should be used for the considered constructional details.

114
 prEN 1993-1-9:2023 (E)

Key
1 Intersection (threshold)
2 No effect
3 Beneficial effect of HFMI treatment

Figure F.1 — Fatigue resistance curves for HFMI treated state

F.4.2 Classification of constructional details

F.4.2.1 General

(1) The detail category of HFMI treated details ∆σHFMI,C in Tables F.2, F.4 and F.6 that are dependent
on steel grade fy and stress ratio R may be used.
(2) As an alternative to the tabulated values given in Tables F.2, F.4 and F.6 the detail category of
HFMI treated details ΔσHFMI,C may be calculated for any yield strength fy and stress ratio R according to
F.4.3.
(3) The steel grade is decisive for the detail category. It may be assumed that a reduction of the
nominal yield strength depending on the material thickness does not affect the detail category.
NOTE The national annex can specify a reduction of the nominal yield strength in function of the material
thickness.

(4) Linear interpolation may be used for stress ratios R different from those given in the tables unless
the less favourable limit value is chosen.

115
prEN 1993-1-9:2023 (E)

F.4.2.2 Transverse stiffener

Table F.2 — Reference value of detail category of HFMI treated details ΔσHFMI,C of transverse
stiffeners due to qualified HFMI treatments for the nominal stress method

Detail categorya) c)

Stress ratio R [-]b)

Steel grade according to EN 10025 -1,0 0,1 0,5
S235 ≤ S < S355 125 125 80
S355 ≤ S < S650 160 140 90
S650 ≤ S ≤ S700 160 160 125
a) Table applies for ℓ ≤ 50mm; if 50 < ℓ ≤ 80mm, ΔσC,HFMI should be reduced by one fatigue detail category.
b) For other stress ratios R, linear interpolation is allowed.
c) Limitations of applied stresses calculated according to F.3.1 : -0,7 fy < σ ≤ fy

116
 prEN 1993-1-9:2023 (E)

Table F.3 — Threshold for beneficial effect of HFMI treatment for transverse stiffeners
(Figure F.1)

Detail category Threshold for beneficial effect of HFMI

t t tt t t
Transverse As- HFMI- Maximum Minimum number
stiffener welded treated stress range of cycles
ΔσC ΔσHFMI,C Δσs Nmin,HFMI
[N/mm²] [N/mm²] [N/mm²] [-]
80 96 817100
90 128 337800
100 167 153250
50 < ℓ ≤ 80mm 71 112 222 65500
125 292 28750
140 388 12300
160 542 4500
90 107 826600
100 140 375000
112 186 160300
ℓ ≤ 50mm 80
125 244 70350
140 324 30100
160 453 11050
F.4.2.3 Transverse butt weld

Table F.4 — Reference value of detail category of HFMI treated details ΔσHFMI of transverse butt
welds without or with tapering in width or in thickness (slope ≤ 1:4) due to qualified HFMI
treatments for the nominal stress method

Detail categoryb) c)

Stress ratio R [-]a)

Steel grade according to EN 10025 -1,0 0,1 0,5
S235 ≤ S < S355 140 140 90
S355 ≤ S < S650 160 160 100
S650 ≤ S ≤ S700 160 160 140
a) For other stress ratios R, linear interpolation is allowed.
b) For tmin > 25mm, ΔσC,HFMI should be reduced by ks = (25/tmin)0,2
c) Limitations of applied stresses calculated according to F.3.1: -0,9 fy < σ ≤ fy

117
prEN 1993-1-9:2023 (E)

Table F.5 — Threshold for beneficial effect of HFMI-treatment for transverse butt welds
(Figure F.1)

Threshold for beneficial effect of HFMI-

Detail category
treatment
Transverse butt As- HFMI- Maximum Minimum number
weld welded treated stress range of cycles
ΔσC ΔσHFMI,C Δσs Nmin,HFMI
[N/mm²] [N/mm²] [N/mm²] [-]
100 117 907200
112 155 387900
t ≤ 25mma) 90 125 205 170200
140 272 72750
160 379 26750
a) For t > 25mm, ΔσC,HFMI should be reduced by ks = (25/t)0,2

F.4.2.4 Longitudinal stiffener

Table F.6 — Reference value of detail category of HFMI treated details ΔσHFMI,C of longitudinal
stiffener due to qualified HFMI treatments for the nominal stress method

Detail categorya),d) e)

Stress ratio R [-]c)

Steel grade according to EN 10025 -1,0 0,1 0,5

S235 ≤ S < S355 b) b) b)

S355 ≤ S < S650 100 100 71

S650 ≤ S ≤ S700 125 125 80
a) Table is valid for any stiffener length.
b) No improved detail category due to missing data base.
c) For other stress ratios R, linear interpolation is allowed.

d) For longitudinal stiffener in pure bending, the detail category of HFMI-treated details ΔσHFMI,C should be

reduced by one detail category.

e) Limitations of applied stresses calculated according to F.3.1: -0,5 fy < σ ≤ fy

118
 prEN 1993-1-9:2023 (E)

Table F.7 — Threshold for beneficial effect of HFMI treatment for longitudinal stiffener
(Figure F.1)

Threshold for beneficial effect of HFMI-

Detail category
treatment
Longitudinal As- HFMI- Maximum Minimum number
stiffener welded treated stress range of cycles
ΔσC ΔσHFMI,C Δσs Nmin,HFMI
[N/mm²] [N/mm²] [N/mm²] [-]
80 114 333400
90 154 137800
ℓ > 100mma) 63 100 200 62500
112 265 26700
125 349 11750
a) See footnote 1 of Table F.6

F.4.3 Alternative formulae for determination of detail category

(1) As an alternative to the tabulated values given in Tables F.2, F.4 and F.6 the detail category of
HFMI treated details ΔσHFMI,C may be calculated for any yield strength fy and stress ratio R using the fatigue
resistance in these tables for S355 and R = 0,1 as a reference value modified as follows:
ΔσHFMI,C = f1.f2.ΔσHFMI,C,ref (F.2)

where:
ΔσHFMI,C is the detail category of HFMI treated detail of any yield strength fy and any stress ratio R;
ΔσHFMI,C,ref is the detail category for same detail with fy = 355 N/mm² and R = 0,1 as specified in Tables
F.2, F.4 and F.6;
f1 is the modification factor accounting for the effect of yield strength fy determined as
follows:
0,1�𝑓𝑓y −355�
𝑓𝑓1 = 1 + (F.3)
𝛥𝛥𝜎𝜎HFMI,C,ref

f2 is the modification factor accounting for the effect of stress ratio R determined as follows:
1
𝑓𝑓2 = if 0,1 < R < 1,0, otherwise f2 = 1,0 (F.4)
0,5𝑅𝑅2 +0,95𝑅𝑅+0,9

(2) When calculating the modification factor f1, the nominal value for yield strength fy may be used
without correction for plate thickness.

F.4.4 Fatigue resistance modification

(1) A thickness correction may be neglected since it is already included in Tables F.2, F.4 and F.6.
(2) The correction for steel grade fy and the stress ratio R should be made according to F.4.2 to F.4.4.

119
prEN 1993-1-9:2023 (E)

F.5 Fatigue verification

(1) For verification with the equivalent constant stress range Δσe,2,HFMI,Ed, , the nominal stress ranges
from HFMI treated constructional details should satisfy the following relationship:
Δ𝜎𝜎e,2,HFMI,Ed
≤ 1,0 (F.5)
Δ𝜎𝜎HFMI,C /𝛾𝛾Mf

(2) Stress components that are not perpendicular to the HFMI treated weld toe shall be verified in
accordance with the procedures for welded details according to Clause 9.
(3) Where no data for Δσe,2,HFMI,Ed is available the calculation should be made in accordance with F.7.
(4) An infinite life for constructional details may be assumed if the maximum stress range of the
applied nominal stress spectrum Δσmax,HFMI,Ed satisfies following relationship:
Δ𝜎𝜎max, HFMI,Ed
Δ𝜎𝜎HFMI,D /𝛾𝛾Mf
≤ 1,0 (F.6)

(5) In case of the combination of normal and shear stresses, the fatigue verification should consider
their combined effects by adding the components of damage according to Miner’s summation with
respect to cracking at weld toe subjected to nominal normal stress ranges ∆σx, ∆σy, ∆σz and nominal shear
stress ranges ∆τxy, ∆τxz, ∆τyz:
𝑚𝑚σ 𝑚𝑚τ
Δ𝜎𝜎j,e,2,Ed Δ𝜏𝜏k,e,2,Ed
� � � +� � � ≤ 1,0 (F.7)
Δ𝜎𝜎j,C /𝛾𝛾Mf Δ𝜏𝜏k,C /𝛾𝛾Mf
j=x,y,z k=xy,xz,yz

where:

mσ is the first slope parameter m1 of the fatigue resistance curve for the considered constructional
detail under normal stress loading
mτ is the first slope parameter m1 of the fatigue resistance curve for the considered constructional
detail under shear stress loading
Δ𝜎𝜎x,C corresponds to the detail category for stresses acting perpendicularly to the weld toe of HFMI
treated details ∆σHFMI,C
(6) Where normal and shear stresses may cause the formation of fatigue cracks at different locations,
a separate fatigue verification for both locations should be performed.

F.6 Requirements for application

F.6.1 Requirements for welds before HFMI treatment

(1) The requirement from the original detail category tables for a detail before HFMI treatment
should be met.
(2) The weld toes to be treated shall be accessible for treatment.
(3) The post-weld treatment that change the weld toe geometry should not be applied before HFMI
treatment.
NOTE Such treatments are grinding, TIG welding and vibration method [9].

(4) Cleaning of the welds (e.g. removal of weld spatter) to comply with quality requirements of
EN 1090-2 is permitted. The weld toes to be treated should be visible after cleaning.

120
 prEN 1993-1-9:2023 (E)

F.6.2 Requirements for welds after HFMI treatment

(1) Any further post-weld treatment should not be applied after HFMI treatment (e.g. grinding, TIG
welding, stress relief treatment, vibration method [9]).
NOTE The beneficial effect of HFMI treatment is mainly based on the introduction of compressive residual
stresses. These stresses can be relieved by a further post-weld treatment afterwards.

(2) Blast cleaning for surface preparations for corrosion protection is permitted.
(3) Thermal treatments to straighten welded components is not permitted.
(4) Hot dip galvanizing of HFMI treated constructional details is not permitted.
(5) The limitations in Formula (9.3) should also be satisfied during manufacturing and assembly.

F.6.3 Quality control

(1) HFMI treatments should only be performed by trained and qualified operators.
(2) The qualification of operators should be based on suitable training related to the device
manufacturer.
NOTE Guidance on training can be found in the literature [4],[9],[10].

(3) For HFMI treatment, a visual inspection of HFMI treatment (trace of indentation) should be
carried out by the operator and welding supervisor (extent of inspection: 100%). It should be ensured
that the relevant weld toes have been completely treated and that the original weld toe notches have been
entirely removed.
NOTE Guidance on visual inspection of HFMI treated details can be found in the literature [4],[9],[10].

(4) The HFMI treated sections should be marked in the constructional execution documents.

F.7 Treatment of variable amplitude loading

(1) When using the design spectrum, the applied stress range Δσi should be multiplied by γFf and the
fatigue resistance ΔσHFMI,C divided by γMf. Annex A provides fatigue verification formats for fatigue action
effect expressed as a stress range spectrum.
(2) For HFMI treated details subject to variable amplitude loading where the mean stress (or stress
ratio R) for each stress range is known, the design value of a modified equivalent stress range that
accounts for the mean stress effect may be determined as follows:

5 5
∑�𝑛𝑛𝑖𝑖 �∆𝜎𝜎𝑖𝑖,𝐸𝐸𝐸𝐸 ⁄𝑓𝑓2,𝑖𝑖 � �
∆𝜎𝜎eq,R,Ed = � ∑ 𝑛𝑛𝑖𝑖
(F.6)

where:
f2,i is the modification factor for stress ratio effect for the cycle i calculated acc. to Formula (F.4)
Δσi,Ed is the stress range for the ith stress in the spectrum (design values)
ni is the number of cycles associated with the stress range Δσi,Ed.
(3) The design value of the equivalent stress range at 2×106 cycles Δσe,2,HFMI,Ed may be determined as
follows:
∑ 𝑛𝑛
∆𝜎𝜎e,2,HFMI,Ed = 2×10𝑖𝑖 6 ∆𝜎𝜎eq,R,Ed (F.7)

121
prEN 1993-1-9:2023 (E)

where

� 𝑛𝑛𝑖𝑖 is the total number of cycles of the spectrum.

(4) The detail category to be used in fatigue verification based on Δσeq,R,Ed should be determined as
follows:
∆𝜎𝜎HFMI,C = 𝑓𝑓1 ∆𝜎𝜎HFMI,C,ref (F.8)

(5) When using an equivalent stress range based on a simplified fatigue load model of EN 1991 and
appropriate λ-values, the factor to consider the mean stress effect in spectrum, λHFMI, may be determined
as follows:
Δ𝜎𝜎eq,R,Ed
𝜆𝜆HFMI = but 1,0 ≤ 𝜆𝜆HFMI ≤ 2,0 (F.9)
Δ𝜎𝜎eq,Ed

where:
Δσeq,R,Ed is the modified equivalent stress range that accounts for the mean stress effect according to
Formula (F.6)
Δσeq,Ed is the design value of the equivalent stress range acc. to Formula (F.6) but setting f2,i = 1,0
NOTE More guidance can be found in ref. [3], including values for λHFMI, in case of road bridges.

122
 prEN 1993-1-9:2023 (E)

Annex G
(informative)

Hot spot stress reference detail method

G.1 Use of this annex

(1) This informative annex provides complementary guidance to B.4.2(1) to the application the hot spot
stress method for fatigue verification using a reference detail.
NOTE The national choice on the application of this informative annex is given in the national annex. If the
national annex contains no information on the application of this informative annex, it can be used.

G.2 Scope and field of application

(1) This informative annex covers non-classified details as similar as possible to reference details,
which satisfy the requirements of Tables 10.1 to 10.11 and Table B.2.
NOTE Further guidance on the application of this method can be found in ref. [1].

(2) This informative annex applies for fatigue verification that is carried out according to the
provisions given in Clause 9 and in Annex A.

G.3 Fatigue action effect

(1) Fatigue action effect should be calculated in accordance with B.3.

G.4 Fatigue resistance

(1) The calculation procedure for the determination of the fatigue resistance should be as follows:
a) Select a reference detail with known fatigue resistance from the detail category tables, which is as
similar as possible to the detail to be verified with respect to geometric, weld quality and loading
parameters;
b) Identify the type of stress in which the fatigue resistance is expressed. This is usually the nominal
stress;
c) Establish a FEM model of the reference detail and the detail to be verified with the same type of
meshing and elements following the recommendations given in EN 1993-1-14;
d) Use the reference detail and the detail to be verified with the stress identified in b);
e) Determine the hot spot stress ranges ∆σHS,ref of the reference detail and the hot spot stress ranges
∆σHS,nc of the non-classified detail to be verified;
f) The detail category of the non-classified detail to be verified ∆σHS,C,nc is determined from the detail
category of the reference detail ∆σHS,C,ref as follows:
Δ𝜎𝜎HS,ref
Δ𝜎𝜎HS,C,nc = Δ𝜎𝜎HS,nc
Δ𝜎𝜎HS,C,ref (G.1)

(2) Where control measurements are taken to verify the calculated stresses on the non-classified
detail or the reference detail, strain gauges should be positioned outside the heat affected zone.

123
prEN 1993-1-9:2023 (E)

G.5 Fatigue verification

(1) For the fatigue verification of the detail, the slope parameters m1, m2 and the number of cycles ND
of the reference detail may be assumed.
(2) For verification with respect to reference value, the normal stress ranges for a non-classified
constructional detail should satisfy the following relationship:
Δ𝜎𝜎HS,,e,2,Ed
≤ 1,0 (G.2)
Δ𝜎𝜎HS,C,nc /𝛾𝛾Mf

(3) When no data for ΔσHS,e,2,Ed are available the verification format in Annex A may be used.
(4) An infinite life for constructional details may be assumed if the maximum stress range of the
applied normal stress spectrum satisfies following relationship:
Δ𝜎𝜎max, HS,Ed
Δ𝜎𝜎HS,D,nc /𝛾𝛾Mf
≤ 1,0 (G.3)

124
 prEN 1993-1-9:2023 (E)

Bibliography
References contained in recommendations (i.e. “should” clauses)
The following documents are referred to in the text in such a way that some or all of their content
constitutes highly recommended choices or course of action of this document. Subject to national
regulation and/or any relevant contractual provisions, alternative documents could be used/adopted
where technically justified. For dated references, only the edition cited applies. For undated references,
the latest edition of the referenced document (including any amendments) applies.
[1] EN 10163-2, Delivery requirements for surface condition of hot-rolled steel plates, wide flats and
sections - Part 2: Plate and wide flats;

[2] EN 10163-3, Delivery requirements for surface condition of hot-rolled steel plates, wide flats and
sections - Part 3: Sections;

[3] EN 10210, Hot finished steel structural hollow sections;

[4] EN 10219, Cold formed welded structural hollow sections of non-alloy and fine grain steels;

[5] EN ISO 2553, Welding and allied processes — Symbolic representation on drawings — Welded joints;

[6] EN ISO 5817, Welding — Fusion-welded joints in steel, nickel, titanium and their alloys (beam welding
excluded) — Quality levels for imperfections;

[7] EN ISO 6947, Welding and allied processes — Welding positions;

[8] EN ISO 9013, Thermal cutting — Classification of thermal cuts — Geometrical product specification
and quality tolerances;

[9] ISO 14347, Fatigue – Design procedure for welded hollow section joints – recommendations.
References contained in possibilities (i.e. “can” clauses) and notes
The following documents are cited informatively in the document, for example in "can" clauses and in
notes.
[1] Hobbacher, A.F.: Recommendations for Fatigue Design of Welded Joints and Components
Second Edition, IIW document IIW-2259-15, 2016.
[2] Kuhlmann, U., Breunig, S., Ummenhofer, T., Weidner, P.: Entwicklung einer DASt-Richtlinie für
höherfrequente Hämmerverfahren, DASt-AiF-IGF-Nr. 17886, Final report, 2017.
[3] Shams-Hakimi, P., Carlsson, F., Al-Emrani, M., Al-Karawi, H.: Assessment of in-service stresses in steel
bridges for high-frequency mechanical impact applications, Engineering Structures, Elsevier, Vol.
241, 112498, May 2021.
[4] Kuhlmann, U., Breunig, S., Gölz, L.: Untersuchungen zur Auswirkung variierender
Ausführungsqualitäten einer HFH-Nachbehandlung und Beurteilung einfacher Methoden zur
Überprüfung der Nachbehandlungsspur, Schlussbericht, BAW;
http://www.baw.de/content/publikationen/www-dokumente-
oeffentlich/0/2019_01_30_Schlussbericht_Final.pdf.
[5] VDI 2230-1: Systematic calculation of highly stressed bolted joints – Joints with one cylindrical bolt;
Verlag des Vereins Deutscher Ingenieure, 2015.
[6] EN 1994, Design of Composite Steel and Concrete Structures.

125
prEN 1993-1-9:2023 (E)

[7] EN ISO 14732, Welding personnel – Qualification testing of welding operators and weld setters for
mechanized and automatic welding of metallic materials.
[8] CEN/TR 14599:2005: Terms and definitions for welding purposes in relation with EN 1792, CEN,
June 2005.
[9] DASt Richtlinie 026, Ermüdungsbemessung bei Anwendung höherfrequenter Hämmerverfahren,
Stahlbau Verlag, 2019, ISBN 978-3-941687-37-0.
[10] Marquis, G. B., and Barsoum, Z: IIW Recommendations for the HFMI Treatment: For Improving
the Fatigue Strength of Welded Joints. IIW Collection, Springer Publishers, 2016, ISBN 978-981-
10-2503-7.

126