Eurobitume TF Data Collection

Position Paper on Test Methods

used during the Data Collection

May 2009
 ISBN 2-930160-11-X
D/2009/7512/12
© May 2009
Published by the European Bitumen Association
Boulevard du Souverain 165
B – 1160 Brussels, Belgium
Tel.: +32/(0)2 566 91 40 * Fax: +32/0)2 566 91 49
[email protected]
www.eurobitume.eu
Position Paper on Test Methods
used during the Data Collection

Considerable effort has been made to assure the accuracy and reliability of the information
contained in this publication. However, Eurobitume cannot accept liability for any loss, damage
or injury whatsoever resulting from the use of this information.

ACKNOWLEDGEMENT

Eurobitume is grateful to the members of the Task Force Data Collection for producing this
document
Anja Soerensen – ARBIT (Project Leader)
Jeroen Besamusca – Kuwait
Alice Bomers – Total
Theodor Hagemann – BP
Laurence Lapalu – Total
Claire Laurens – Shell
Stéphane Le Clerc – Shell
Bernard Lombardi – GPB
Alberto Madella – API
Sophie Nigen-Chaidron – Shell
Antonio Paez Dueñas – Repsol
Ignacio Perez – CEPSA
Chris Rayner – ExxonMobil
Mike Southern – Eurobitume
Chris Southwell – RBA
Markus Spiegl – OMV
André Stawiarski – Total
Wim Teugels – Nynas
Maria Rita Valentinetti - Eni
Table of Contents
Executive Summary ......................................................................................... 1

Background ...................................................................................................... 2

What was done, issues encountered? ........................................................... 2

Use of the database ......................................................................................... 3

Caveat ............................................................................................................... 3

Statistical analysis ........................................................................................... 3

Glossary ............................................................................................................ 4

Position towards individual Test / Calculation Methods or Procedures ..... 5

Part 1 - Conventional Characteristics............................................................. 5

1.1. Nominal Penetration Range ............................................................................................... 5
1.2. Needle penetration [EN 1426] ............................................................................................ 5
1.3. Softening Point Ring and Ball [EN 1427]............................................................................ 6
1.4. Penetration Index [EN 1427] .............................................................................................. 7
1.5. Dynamic Viscosity at 60°C [EN 12596] .............................................................................. 8
1.6. Kinematic Viscosity at 135°C [EN 12595]........................................................................... 9
1.7. Fraass Breaking Point [EN 12593] ................................................................................... 10
1.8. Elastic Recovery (for PMB only) [EN 13398].................................................................... 10
1.9. Storage Stability (for PMB only) [EN 13399] .................................................................... 11
1.10. Flash Point [EN ISO 2592] ............................................................................................. 11
1.11. Solubility [EN 12592] ...................................................................................................... 11
1.12. Density [EN ISO 3838].................................................................................................... 12

Part 2: Ageing Procedures ............................................................................ 13

2.1. Resistance to Hardening – RTFOT-Method [EN 12607-1]............................................... 13
2.2. Pressure Ageing Vessel – PAV [EN 14769] ..................................................................... 13
2.3. Rotating Cylinder Ageing Test – RCAT [EN 15323]......................................................... 14

Part 3 - Resistance to Hardening .................................................................. 15

3.1. Change in mass [EN 12607-1] ......................................................................................... 15
3.2. Retained Penetration [EN 12607-1 in combination with EN 1426]................................... 15
3.3. Change in Softening Point Ring and Ball [EN 12607-1 in combination with EN 1427] .... 15
Part 4: High Service Temperature Characteristics ...................................... 16
4.1. Complex Modulus in DSR [EN 14770]: temperature sweep between 40 and 80°C (at
given frequencies) and frequency sweep between 0.1 and 10 Hz (at given temperatures) ... 16
4.2. Low Shear Viscosity in DSR – LSV [WI 336067, later prTS 15324]: EVT1 and EVT2 .... 18
4.3. Zero Shear Viscosity [WI 336068, later prTS 15325] ....................................................... 20

Part 5: Intermediate Service Temperature Characteristics......................... 22

5.1. Complex Modulus G* and phase angle in DSR [EN 14770]: temperature sweep between
10 and 40°C (at given frequencies) and frequency sweep between 0.1 and 10 Hz (at given
temperatures)........................................................................................................................... 22

Part 6: Low Service Temperature Characteristics....................................... 23

6.1. Bending Beam Rheometer BBR [EN 14771]: Determination of Stiffness and m-value at -
16°C as well as determination of temperature at which stiffness S = 300 MPa and m-value =
0.3 respectively........................................................................................................................ 23

Part 7: Cohesion............................................................................................. 26
7.1. Deformation Energy with Force Ductility Method [EN 13589 and EN 13703] .................. 26
7.2. Deformation Energy with Tensile Test Method [EN 13587 and EN 13703] ..................... 27
7.3. Vialit Pendulum [EN 13588].............................................................................................. 28

Annex 1 – Data Collection template.............................................................. 29

Annex 2 – List of test methods used during the project............................. 31

Executive Summary
Eurobitume’s Task Force Data Collection has created a comprehensive database of bitumen
properties for many commercially available bitumens in Europe. The database includes paving
grades, hard paving grades, PMB and special binders that were commercially available at the
time of the project.
TF Data Collection was established in response to a request from CEN to collect test data on
bituminous binders for the purpose of development of new performance related specifications.
This Data Collection is not giving any statements on how performance related specification
should look like but is addressing binder tests only.
The purpose of the project was to allow assessment of the various test methods proposed as
candidates to measure properties which may be suitable for future bitumen specifications. Many
of the traditional test methods were developed for paving grade and hard paving grade bitumen,
but have demonstrated deficiencies when measuring the characteristics of PMB.
The outcome of the data collection supports positions taken earlier by bitumen industry
representatives in the CEN specification discussions with hard evidence, e.g. on the applicability
and limits of test methods.
Simple tests can and should be used for simple binders i.e. Normal paving grades and Hard
grade bitumens. The introduction of more complex tests for simple binders should not be
considered unless they provide substantial improvement versus today. Needle penetration is a
suitable test for all types of binders, but the interpretation of ring and ball softening point is
different for paving grade and hard paving grades than for PMB.
More complex binders require more tests. Dynamic Shear Rheometer (DSR) equipment offers
the possibility to replace several traditional test methods for PMB, e.g. ring and ball softening
point, or dynamic viscosity at 60°C, but parameters and limits still need to be agreed on. DSR
also offers opportunities for testing simple binders, but this is not the recommended route as
simple tests are satisfactory.
For PMB there is a need to measure high temperature properties in a better way than traditional
tests allow. The determination of EVT1 from LSV-testing in DSR might offer a solution, but
because the method and the standard are relatively new, further evaluation is necessary.
The Dynamic Shear Rheometer is considered to be an important tool for testing in future,
especially for modified binders. Some test methods do not appear to add value: e.g. Zero Shear
Viscosity.
None of the test methods measuring low temperature properties (Fraass Breaking Point and
BBR) are completely satisfactory the way they are currently used.
Some of the correlations between properties reported in the BiTVal Phase 1 Report have not
been supported by the data collection project. The database will help discussions on these
issues.

TF Data Collection Position Paper – May 2009 1

Introduction
Eurobitume’s Task Force Data Collection has created a comprehensive database of bitumen
properties for many of the commercially available bitumens in Europe. The database includes
paving grades, hard paving grades, PMB and special binders that were commercially available
at the time of the project.
The purpose of the project is to allow assessment of the various test methods proposed to
measure properties which may be suitable for future bitumen specifications. It is recognised that
the binder properties alone do not determine pavement, or even mixture performance. Other
parameters, such as aggregate characteristics, mix design, manufacture and laying are also
considered important. Future specifications must ensure that bitumens which are currently
available and performing satisfactorily are not excluded and also that innovation is not
discouraged.
Eurobitume has compiled a database comprising 146 unique datasets representing
commercially available binders including 69 paving grades, 58 Polymer Modified Binders (PMB),
15 hard paving grades and 4 special binders, i.e. those not supplied to an existing European
Specification. The datasets were collected between 2006 and 2007.

Background
TF Data Collection was established in response to a request from CEN to collect data on
bituminous binders for the purpose of development of new performance related specifications.
This report does not provide any statements on how performance related specifications should
be structured, but is addressing binder tests only.
The European Standards EN 12591, EN 13924 and EN 14023 are based on various national
standards which were in existence when the process of harmonisation started, i.e. in 1990.
These standards are based on so-called “empirical” specifications and are, by common consent,
reasonably successful and adequate for use in Europe for most binders. The relationship of
many of the test methods to performance is fairly well understood, but they cannot be said to be
directly performance related.
The next step in the process of European Standardisation is to progress to the development of
new specifications which are more-directly performance-related. It will require agreement to be
reached on the properties to be specified and the test methods to measure them. It will also
require agreement to be reached on how best to differentiate the specifications to allow for
differing climate and traffic.

What was done, issues encountered?

Eurobitume members used the template developed by CEN TC336 WG1 (Annex 1) and carried
out tests on binders from their portfolio. The process was complicated by the need to ensure
that members could submit the data confidentially, to ensure that the identity of the binders
submitted could not be determined once data were in the database. This was achieved by using
an independent contractor to receive the data and enter it into the database.
Data integrity was identified as an issue from early in the process. Due to the necessity to
maintain confidentiality, correction of erroneous data once entered into the database was
problematic. Participants were able to identify their own data within the database, but the
requirement to maintain confidentiality meant that they needed to request the contractor to
delete erroneous datasets and replace them with corrected versions. This led to strict version
control being necessary to ensure that the numerous corrections necessary were incorporated
properly. Examples of commonly encountered problems were;
¾ Wrong units being used, e.g. g/ml instead of kg/m³.
¾ Incorrect binder type being selected, e.g. paving grade bitumen rather than PMB.
¾ Typographical errors, e.g. omission of decimal point.

TF Data Collection Position Paper – May 2009 2

For future versions Eurobitume concluded that the template should incorporate limit values for
data entries in order to limit a number of the most commonly encountered problems.
At the time of writing it is considered that the data contained within the database are correct in
respect of reported values and that no transcription errors are included. However, the issue of
data integrity is likely to be encountered for future versions of the database and in future data
collection projects and has the potential to significantly diminish the usefulness/applicability of
the database.

Use of the database

As previously stated, the data collection project was devised to assess candidate test methods
for the development of performance related specifications. The present position paper has been
developed to present the views of Eurobitume on the suitability of test methods for future
specifications, taking into account the interpretation of the results and the experience obtained
with carrying out the tests.

Caveat
During the analysis of the database a number of good correlations were identified between
properties that are not fundamentally related, e.g. for paving grade bitumen a strong correlation
was identified between Needle Penetration and Softening Point. The binders tested during this
project constitute a reasonable subset of commercially available products, but certain classes,
e.g. paving grade bitumen can be assumed to comply with existing European Specifications.
Consequently, for the population in the database, the fact that certain properties are required to
comply with specification limits means that the samples were not random. Therefore
correlations can be anticipated that are artifacts of the sample population, rather than genuine
correlations between test methods. In this paper we have commented under each test method
whether we believe a correlation to be a consequence of the specification, or a genuine
relationship between a given property and another property.
European Test Methods are subject to revision on a regular basis. During the period of this
project a number of the test methods are known to have undergone revision, some of which will
make comparison of test results that use different versions of a test invalid. The list of test
methods used and the version of the standard are listed in Annex 2. Developments on test
methods or on new tests in Europe, or in other areas, were not assessed during this project.
The Eurobitume project only considered binder test methods, therefore the database should
only be used to evaluate properties of binders and cannot, at this time, provide information on
mixture or pavement properties.

Statistical analysis
In order to provide the statements on test methods used during the data collection, practical
experience with test methods was collected in a systemic approach. Additionally, correlation
checks were undertaken for all available test methods. The results are listed in tables in the
respective chapters. In order to avoid duplication, correlations are only stated for one of the two
test methods reviewed.
The statistics were carried out based on 90% reliability and only where at least ten data points
were available to be checked. Correlation between test results was checked for each of the four
different types of binder and for groups of them. However, only those results are given where
most of the binders are included: e.g. if a correlation was determined for all binders but also for
paving grade bitumen only, only the correlation result for all binders is given.
TF Data Collection considered that a correlation existed if R² ≥ 0.9 for the regression graph
determined.

TF Data Collection Position Paper – May 2009 3

Glossary
BiTVal Bitumen Test Validation
BBR Bending Beam Rheometer
CEN Comité Européen de Normalisation (European Committee for Standardization)
CPD Construction Products Directive
DSR Dynamic Shear Rheometer
EN European Standard
EVT Equi-Viscous Temperature
FD Force Ductility
G* Complex Shear Modulus
LSV Low Shear Viscosity
LTA Long Term Ageing
NPG Normal Paving Grades
PAV Pressure Ageing Vessel
Pen Penetration
PMB Polymer Modified Bitumen
R&B Ring and Ball
RCAT Rotating Cylinder Ageing Test
RTFOT Rolling Thin Film Oven Test
SP R&B Softening Point Ring and Ball
STA Short Term Ageing
TS Technical Specification
WI Work Item

TF Data Collection Position Paper – May 2009 4

 Position towards individual Test / Calculation Methods or
Procedures

Part 1 - Conventional Characteristics

1.1. Nominal Penetration Range

General Comments: It is not a test but an indicator for conventional bitumen (paving grades)
and currently used to differentiate between and name the grades.
Suitability for future specification: Not relevant, because nominal penetration range is a
designation and not a test. However, this is historically used for designation of conventional
binders.

1.2. Needle penetration [EN 1426]

General Comments: Needle penetration is a well known and established test which is
generally applicable.
Precision: Precision data are stated in the standards. However, a number of Round Robin tests
suggest that, in practice, the precision is not as good as stated [Ref. ESSAIS CIRCULAIRES
2007 PROGRAMME N° 86, « BITUMES »: BNPé – Département Produits et Méthodes,
Septembre 2007].
Timing: Including preparation approximately 3 h.
Comments from BiTVal: For unmodified bitumens, the penetration test correlates well with the
stiffness of the bitumen measured, using the DSR, at the same temperature (25°C) and at a
frequency of 0.4 Hz, with the equivalent loading time. In rheological terms, a good correlation
has been identified between log (G*), the complex shear modulus and log(pen).
It is generally considered that, for paving grade bitumens, the R&B softening point is equivalent
to a penetration of 800 x 0.1 mm.
Correlation with other test methods from this Data Collection:

Penetration on fresh binders

Paving
Correlation with All binders grades, hard Paving grades Paving PMB only
paving and hard grades only
grades and paving grades
special
binders
LSV – EVT1 (fresh) R² = 0.91, log
Penetration (short R² = 0.94, lin
term aged)
R&B Softening Point R² = 0.91, log
(fresh)

Penetration on short term aged binders

Paving
Correlation with All binders grades, hard Paving grades Paving PMB only
paving and hard grades only
grades and paving grades
special
binders
R&B Softening Point R² = 0.90, log
(fresh)

The correlation recorded between Penetration and R&B softening point is most probably forced
by the fact that the binders studied comply with existing specifications.

TF Data Collection Position Paper – May 2009 5

 Observation / remarks: BiTVal stated that correlation has been found in the papers with DSR
measurements at 25°C. During the Data Collection tests were performed at 20° and 30°C and
no correlation was found at any frequency between 0.1 and 10 Hz. TF Data Collection believes
that R2 values of 0.8 do not show a correlation while literature sometimes regards this value as
a proof for correlation.
Suitability for production / quality control: Yes, but the relatively long time needed until
results are obtained should be taken into account.
Suitability for future specification: Yes, but the limited precision for some binders should be
taken into account. Though the method is well established, it is not performed accurately by all
users.

1.3. Softening Point Ring and Ball [EN 1427]

General Comments: It is a well known and established test applicable for conventional binders.
The significance of the test is uncertain for PMB, in particular highly modified PMB.
Precision: The precision data are provided in the test method standards having been
established in several international round robin tests. However, some round robin tests have
shown borderline (and higher) results [CEN TC 336 WG 1 TG 9].
Timing: 45 minutes.
Comments from BiTVal: It is generally considered that, for paving grade bitumen, the R&B
softening point is equivalent to a penetration of 800 x 0.1 mm. Although many test methods
measure related properties and therefore there will be some relationship, no formal correlation
has been found in the papers reviewed between the softening point test and other bitumen
tests.
Correlation with other test methods from this Data Collection:

Softening Point Ring and Ball on fresh binders

Paving grades,
Correlation with All binders hard paving Paving Paving PMB only
grades and grades and grades only
special binders hard paving
grades
Dynamic Viscosity @ R² = 0.91, exp
60°C (fresh)
LSV – EVT1 (fresh) R² = 0.95, exp
LSV – EVT2 (fresh) R² = 0.94, lin
ZSV (fresh) R² = 0.96, exp
G* @ 60°C and 1,6 Hz R² = 0.94, exp
(fresh)
Temperature @ G*/sin R² = 0.96, exp
δ = 1 kPa (fresh)
R&B softening point R² = 0.94, exp
(short term aged)
Dynamic Viscosity @ R² = 0.93, exp
60°C (short term aged)
LSV – EVT1 (short term R² = 0.94, lin
aged)
ZSV (short term aged) R² = 0.94, exp
G* @ 60°C and 1,6 Hz R² = 0.91, exp
(short term aged)
Temperature @ G*/sin R² = 0.90, exp
δ = 2.2 kPa (short term
aged)

TF Data Collection Position Paper – May 2009 6

The correlation recorded between Penetration and R&B softening point is most probably forced
by the fact that studied binders comply with existing specifications.

Softening Point Ring and Ball on short term aged binders

Paving grades,
Correlation with All binders hard paving Paving grades Paving PMB only
grades and and hard grades only
special binders paving grades
LSV – EVT1 (fresh) R² = 0.91, lin
LSV – EVT2 (fresh) R² = 0.91, lin
ZSV (fresh) R² = 0.95, exp
G* @ 60°C and 1,6 R² = 0.90, exp
Hz (fresh)
Temperature @ G*/sin R² = 0.91, exp
δ = 1 kPa (fresh)
Def. Energy FD 10°C R² = 0.95, lin
(fresh)
Dynamic Viscosity @ R² = 0.96, exp
60°C (short term
aged)
LSV – EVT1 (short R² = 0.95, lin
term aged)
LSV – EVT2 (short R² = 0.93, lin
term aged)
ZSV (short term aged) R² = 0.97, exp
G* @ 60°C and 1,6 R² = 0.92, exp
Hz (short term aged)
Temperature @ G*/sin R² = 0.95, exp
δ = 2.2 kPa (short
term aged)
Def. Energy FD 10°C R² = 0.99, lin
(long term aged)

Observation / remarks: BiTVal stated that no formal correlation has been found in the papers
reviewed between the softening point test and other bitumen tests. However, the Data
Collection results proved clearly that the softening point ring and ball is at least for some groups
of binders correlating very well with other tests.
Suitability for production / quality control: Yes, useful for production and quality control,
because the method is well known, relatively quick, and correlated to other properties.
Suitability for future specification: Yes, suitable for specification purposes for paving grade
and hard paving grade binders, but not necessarily for PMB.

1.4. Penetration Index [EN 1427]

General Comments: Penetration index is obtained by calculation.
The calculation formula is based on the assumption that the penetration of the binder is 800 0.1
mm at its softening point ring and ball. This is correct for most conventional binders but not for
PMB and special binders. The result of the calculation could be improved if it is based on
penetration values at different temperatures.
Precision: It is related to the precision of penetration and softening point ring and ball tests.
The variability can be +/- 0.5 units.
Timing: To determine the penetration index, a calculation is carried out. The determination of
the necessary characteristics takes approximately 3 hours.

TF Data Collection Position Paper – May 2009 7

Comments from BiTVal: Although many test methods measure related properties and
therefore there will be some relationship, no formal correlation has been found in the papers
reviewed between the penetration index and other bitumen tests.
Suitability for production / quality control: Yes, for paving grade bitumen. Penetration index
gives a fair approximation of the temperature susceptibility between 25°C and the R&B SP
temperature of these binders.
Suitability for future specification: Because penetration index gives an approximation of the
temperature susceptibility of normal paving grade binders it could be used for future
specifications for these binders only.

1.5. Dynamic Viscosity at 60°C [EN 12596]

General Comments: The applicability is limited to binder with softening point R&B below 60°C.
The standard requires a vacuum of "40 000 Pa +/- 67 Pa": It is not stated if this is a relative
vacuum (in comparison with atmospheric conditions or not) and the recommended precision of
(+/- 67 Pa) cannot be met with any equipment.
It is essential to use a suitable capillary and cleaning of the capillaries must be carried out
properly.
Precision: Good for Newtonian liquids.
Timing: Less than two hours.
Comments from BiTVal: The dynamic viscosity at 60°C measured with the capillary
viscometer test to the Australian standard AS 2341.02 was correlated with other binder
properties for multigrade binders. A good correlation is reported with G*/sin δ at a frequency of
10 rad/sec and 60°C before and after RTFO-ageing. The correlation with the Ring and Ball
softening point was good after RTFO-ageing, but not before.
Correlation with other test methods from this Data Collection:

Dynamic Viscosity on fresh binders

Paving
Correlation with All binders grades, hard Paving Paving PMB only
paving grades grades and grades only
and special hard paving
binders grades
LSV – EVT1 (fresh) R² = 0.91, log
LSV – EVT2 (fresh) R² = 0.91, log
G* @ 60°C and 1,6 Hz R² = 0.92, lin
(fresh)
Temperature @ G*/sin δ R² = 0.94, log
= 1 kPa (fresh)
Def. Energy FD 25°C R² = 0.92, lin
(fresh)
Dynamic Viscosity @ R² = 0.94, lin
60°C (short term aged)

TF Data Collection Position Paper – May 2009 8

Dynamic Viscosity on short term aged binders
Paving
Correlation with All binders grades, hard Paving grades Paving PMB
paving grades and hard grades only only
and special paving grades
binders
ZSV (fresh ) R² = 0.96, lin
Temperature @ G*/sin δ = R² = 0.91, log
1 kPa (fresh)
LSV – EVT1 (short term R² = 0.92, log
aged)
ZSV (short term aged) R² = 0.99, lin
Temperature @ G*/sin δ = R² = 0.93, log
2.2 kPa (short term aged)

Observation / remarks: The statements from BiTVal can be confirmed only in some respects.
It cannot be seen from the Data Collection that very good correlations exist with softening point
ring and ball whatever the binder type: this is true only for the group of paving grades, hard
paving grades and special binders but not for PMBs. For this group, a good correlation exists
before and after short term ageing. The good correlation between viscosity at 60°C and G*
(60°C, 10 rad/s) is confirmed only for paving grade and hard paving grades, while the
correlation with Temp @ G*/sin (δ) =1.0 kPa (fresh) is good whatever the binder type. It can be
seen that correlations also exist with LSV - EVT1 (fresh), LSV - EVT2 (fresh), ZSV (STA) and
with the Temp @ G*/sin (δ) = 2.2 kPa (aged) for certain groups of binders.
Suitability for production / quality control: Yes, useful for production and quality control,
because the method is well known, relatively quick, and correlated to other properties.
Suitability for future specification: No. Due to the limitations on applicability (only Newtonian
liquids), this test cannot be recommended for inclusion in a general specification. However,
dynamic viscosity is a useful indicator for soft binders and could be part of specifications for
binders that show Newtonian behaviour.

1.6. Kinematic Viscosity at 135°C [EN 12595]

General Comments: The test is applicable for most grades and on fresh as well as on aged
binders. However, only Newtonian liquids can be tested properly.
Precision: Good for fresh binders and Newtonian liquids, becomes worse for non-Newtonian
liquids.
Timing: Less than two hours.
Comments from BiTVal: No comments to the kinematic viscosity at 135°C in the BiTVal report.
Correlation with other test methods from this Data Collection:

Kinematic Viscosity on fresh binders

Paving
Correlation with All binders grades, hard Paving Paving PMB only
paving grades and grades only
grades and hard paving
special grades
binders
Def. Energy FD 20°C R² = 0,93, lin
(long term aged)

TF Data Collection Position Paper – May 2009 9

Kinematic Viscosity on short term aged binders
Paving
Correlation with All binders grades, hard Paving Paving PMB only
paving grades and grades only
grades and hard paving
special grades
binders
G* @ 60°C and 1,6 Hz R² = 0.92, lin
(fresh)
Temperature @ G*/sin δ = R² = 0.94, log
1 kPa (fresh)
LSV – EVT1 (fresh) R² = 0.92, log
LSV – EVT2 (fresh) R² = 0.91, log
Def. Energy FD 20°C R² = 0,99, lin
(long term aged)

Observation / remarks: There are no comments in the BiTVal report. The Data Collection
shows that there are some very good correlations between LSV - EVT1 and the Temp @ G*/sin
(δ) = 1kPa on short term aged paving grade and hard paving grade binders.
Suitability for production / quality control: Yes, useful for production and quality control.
Suitability for future specification: No. Kinematic viscosity is needed for information on
handling only and is not related to the performance of the binder.

1.7. Fraass Breaking Point [EN 12593]

General Comments: Relatively easy on routine testing for conventional grades, but more
difficult for PMB. In the case of PMB the temperature specified for sample preparation (80°C) is
not always sufficient, the viscosity being too high.
The test shows a high variability on all binders. With automatic apparatus, an improvement of
the precision was encountered.
Precision: The precision of results varies depending on the equipment used, the binder tested
and the experience of the person performing the test. Precision data are stated in the
standards. However, a number of Round Robin tests suggest that, in practice, the precision is
not as good as stated, with Reproducibility value around twice the value stated in the method,
which is already quite large (6°C) [Ref. ESSAIS CIRCULAIRES 2007 PROGRAMME N° 86,
« BITUMES »: BNPé – Département Produits et Méthodes, Septembre 2007].
Timing: Up to three hours.
Comments from BiTVal: There is a broad correlation of the Fraass breaking point test with the
BBR for unmodified bitumen and also some suggestion with PMB.
Correlation with other test methods from this Data Collection: No correlation between
Fraass Breaking Point and other test methods was determined on the base of R² ≥ 0.9.
Observation / remarks: The statements from BiTVal cannot be confirmed.
Suitability for future specification: No. The difficulties with sample preparation and poor
precision make this test unsuitable for future specifications.

1.8. Elastic Recovery (for PMB only) [EN 13398]

General Comments: Test method provides indication of the elastic behaviour of the product
and is only applicable to elastomer modified PMB.
The results allow a screening of the status of the binder for the presence of an elastomeric
polymer and therefore some indication of the polymer modification level in case of elastomer
modified PMB. It also gives some indication of the binder type and can help to differentiate
between an elastomer modified binder and others.

TF Data Collection Position Paper – May 2009 10

The requested elongation of 200 mm cannot be met with all binders at all temperatures (e.g.
hard binders or aged binders). This shortcoming could be overcome if different temperatures
than 10 and 25°C were allowed.
Precision: Acceptable but impact of thermal history of the binder on precision.
Timing: Approximately 3 hours.
Correlation with other test methods from this Data Collection: No correlation between
Elastic Recovery and other test methods was determined on the base of R² ≥ 0.9.
Suitability for production / quality control: Yes, to indicate the presence of an elastomer in
PMB.
Suitability for future specification: Because elastic recovery gives an indication of the
presence of an elastomer in PMB it could be used as such for future specifications for elastomer
modified PMB.

1.9. Storage Stability (for PMB only) [EN 13399]

General Comments: The test evaluates the storage stability of polymer modified binders, which
relates to the handling/storage characteristics of a binder.
Depending of the type of modification, softening point ring and ball is not always a good
parameter to judge storage stability; sometimes penetration is giving differences not observed
by softening point ring and ball.
It is uncertain if sample size and duration of the conditioning period are representative of real
storage conditions. The LAST test [Ref. Evaluation of the Laboratory Asphalt Stability Test,
Publication No. FHWA-HRT-04-111] may be a better tool to compare with actual storage
conditions.
Precision: The test uses a difference in properties between the top and bottom sections of a
sample and therefore the precision of the softening point and the needle penetration apply here.
Timing: Three days of conditioning and 3 hours of testing.
Suitability for future specification: This property is critically important for handling and
storage and therefore must be addressed by manufacturers of PMB; however, the test results
obtained may not necessarily indicate actual behaviour in practice.

1.10. Flash Point [EN ISO 2592]

General Comments: A well known test needed for HSE and handling purposes.
Precision: Good.
Timing: Less than 2 hours.
Suitability for future specification: Yes, as it is linked to essential requirement 4 of CPD:
Safety in use.
Flash point is not related to the performance of the binder in the pavement but is needed for
handling information and HSE issues. It can be an indicator of the overall quality of a binder.

1.11. Solubility [EN 12592]

General Comments: Filtration can be very time consuming (up to 2 days), but the method does
not specify any time limit and it is not clear if this affects the measurements. In some countries
filtering membranes are used which may lead to better results. Problems can occur when too
fine glass powder is used. Problems can also occur with some modified binders.
Some bitumens need to be heated to be dissolved properly in xylene, but the method does not
specify anything on this (e.g. effects on the measurements).
The solubility of a binder is determined as an indicator of the purity of the binder only and not for
performance purposes.

TF Data Collection Position Paper – May 2009 11

Precision: Good.
Timing: Normally less than 3 hours, but can be up to two days due to filtration.
Suitability for future specification: Solubility is deemed to be an indicator of the overall quality
of a binder. It is not related to the performance of the binder in the pavement but requested for
information only.

1.12. Density [EN ISO 3838]

General Comments: The density of a binder is determined for information only and not for
performance purposes.
Precision: Good.
Timing: Less than 2 hours.
Suitability for future specification: Density is not related to the performance of the binder in
the pavement but needed for information only.

TF Data Collection Position Paper – May 2009 12

Part 2: Ageing Procedures

2.1. Resistance to Hardening – RTFOT- Method [EN 12607-1]

General Comments: The procedure is typically carried out at a single temperature, 163°C. This
might be too high for soft paving grades or for binders used in warm mixes and too low for hard
grades and some PMB. For future specifications it may be better to perform the procedure at a
defined EVT.
For paving grade binders a good correlation with the bitumen ageing during asphalt production
in a discontinuous (batch pugmill) asphalt plant is stated and the procedure also gives
information on the ageing susceptibility of the binder.
It may be difficult to obtain sufficient sample for further tests, especially when the product is
viscous (PMB and hard paving grade bitumen).
To ensure accurate test results it is essential that representative samples of binder are obtained
from the ageing flasks.
Precision: As it is a conditioning procedure and not a test in itself, no precision data can be
given on it. Nevertheless, the procedure is considered to be good.
Timing: Approximately 4 hours: The procedure can be rather long when several repetitions are
needed to obtain enough sample for further testing.
Comments from BiTVal: Four bitumens were used to make a hundred samples of bitumen and
corresponding asphalt mixtures, manufactured with the three main types of asphalt plants
(discontinuous, continuous, rotary dryer mixer) [3.07]. The main properties of the bitumens
(penetration at 25°C, R&B softening point, ductility at 17°C) were determined before and after
RTFOT and on the corresponding bitumens taken from the coated materials. A comparative
statistical analysis of the results after RTFOT and coating was carried out.
The study showed that:
¾ The bitumen source and the grade of the bitumen have a major role on the thermal
susceptibility to hardening with coating.
¾ The manufacturing process and the composition of the asphalt mixture did not have a
significant effect, on average, on the hardening of the bitumen in the experiment.
¾ The predictive capacity of the RTFOT method is satisfactory. In particular, it makes it
possible to assess the change of R&B softening point with an acceptable precision.
The RTFOT was a little more severe overall than mixing asphalt for the experiment conducted.
Therefore, the RTFOT would appear to be a good method to indicate the degree of hardening of
asphalt mixtures during asphalt production for the binders tested in the experiment.
Suitability for future specification: This is a conditioning procedure only and not a test. Good
correlation is found between individual properties on short term aged and fresh binder in most
cases (for details, see results on individual properties). This good correlation is probably a
consequence of the use of this test in the present bitumen specifications. To have a short term
ageing procedure and the related parameters in the specifications is recommended.

2.2. Pressure Ageing Vessel – PAV [EN 14769]

General Comments: The reliability of the equipment (maintenance, spare parts) is currently
under question.
Potentially this is a good procedure, but test temperature depends on the binder type. It is
uncertain whether the current test method for PAV ageing is the best procedure, as RTFOT-
ageing is required prior to testing, which leads to a relatively small sample volume being
available despite the long test time.

TF Data Collection Position Paper – May 2009 13

PAV is intended to give information of long term ageing susceptibility, but at present there are
limited data to correlate the whole spectrum of European binders against their ageing in asphalt
pavements.
It is questioned whether PAV is suitable in all cases of PMB.
There are still questions about the applicability of this ageing procedure as simulation of actual
long term ageing due to the prolonged high temperature experienced by the binder.
Precision: As it is a conditioning procedure and not a test in itself, no precision data can be
given on it.
Timing: All in all two working days are needed to be able to test the long-term aged binders.
For some participants it is a relatively long procedure to be performed in quality control, but it
might be used in a type testing scheme.
Comments from BiTVal: 20 h of PAV ageing at 100°C was found to correspond to 178 h of
RCAT ageing. However, subsequent comparisons found the value was (176 ± 16) h with RCAT
at 85°C. 5 h of PAV ageing at 100°C and 2.07 MPa was found to be equivalent to standard
RTFOT ageing, and 25 h of PAV ageing at 100°C and 2.07 MPa was found to be equivalent to
standard RTFOT plus PAV ageing.
Suitability for future specification: This is a conditioning procedure only and not a test.
PAV is currently the only widely used long term ageing procedure for bituminous binders.

2.3. Rotating Cylinder Ageing Test – RCAT [EN 15323]

General Comments: RCAT equipment is only in very limited use in Europe.
RCAT could be considered as a more promising procedure than PAV as short and long term
ageing are performed in the procedure. Ageing kinetics of PMB can be very different in PAV and
RCAT and there is uncertainty which method is the better predictor of ageing in the field. It
seems to be easier to perform short and long term ageing in RCAT than in RTFOT plus PAV,
but it takes much longer.
RCAT was developed as an easier conditioning procedure than PAV and was indirectly checked
against PAV [Ref.: A. Verhasselt and A. Vanelstraete - Long-term ageing – Comparison
between PAV and RCAT ageing tests. Second Eurasphalt & Eurobitume Congress, Sessions 1
and 4, pp. 897-905, Barcelona 2000].
There are still questions about the applicability of this ageing procedure as simulation of actual
long term ageing due to the prolonged high temperature experienced by the binder.
Precision: As it is a conditioning procedure and not a test in itself, no precision data can be
given on it.
Timing: 5 hours are needed for short term ageing and 6 working days to be able to condition
the long-term aged binders. The actual manual handling time needed is less than that needed
for carrying out RTFOT plus PAV.
Suitability for future specification: This is a conditioning procedure only and not a test.
Insufficient data are available to conclude on the usefulness of the procedure for future
specifications. Some practical issues exist which may make other ageing regimes preferable.

TF Data Collection Position Paper – May 2009 14

Part 3 - Resistance to Hardening

3.1. Change in mass [EN 12607-1]

General Comments: Change in mass is obtained by weighing and calculation.
Precision: Good.
Timing: After finishing the RTFOT procedure, another 2.5 hours are necessary to have the
results of change in mass as the hot binder needs to cool down.
Comments from BiTVal: see item 2.1
Suitability for future specification: The test may provide some information on loss of volatile
material, which may be HSE related.

3.2. Retained Penetration [EN 12607-1 in combination with EN 1426]

General Comments: The retained penetration is obtained by calculation. After RTFOT, the
penetration of the aged binder is tested and compared with the value of the fresh binder. In
addition, see the general comments given in 2.1.
The change of the binder property due to the impact of heat and air is observed which may be
relevant to evaluate the binder quality.
Precision: The precision depends on the precision of the penetration test.
Timing: Between 1.5 and 4 hours for RTFOT and 3 hours for penetration testing.
Suitability for future specification: For paving grade binder the test appears to be useful.
However, for hard paving grade bitumen penetration values close to the precision limits are
measured, which may limit the usefulness of this parameter for future specifications.

3.3. Change in Softening Point Ring and Ball [EN 12607-1 in combination with EN
1427]
General Comments: This parameter is also based on calculation. As the softening point ring
and ball of some PMB may decrease after RTFOT, the wording should be “change” in softening
point ring and ball instead of “increase”, which is currently used in EN 14023. In addition, see
the general comments given in 2.1.
The change of the binder property due to the impact of heat and air is observed which may be
relevant to evaluate the binder quality.
Precision: The precision depends on the precision of the softening point test.
Timing: Between 1.5 and 4 hours for RTFOT and 45 minutes for softening point testing.
Suitability for future specification: For conventional binders the test appears to be useful.
However, the issues mentioned above concerning PMB may limit the usefulness of the
parameter for future specifications.

TF Data Collection Position Paper – May 2009 15

Part 4: High Service Temperature Characteristics

4.1. Complex Modulus in DSR [EN 14770]: temperature sweep between 40 and
80°C (at given frequencies) and frequency sweep between 0.1 and 10 Hz (at given
temperatures)
General Comments: The test can be applied on most binders, including highly modified PMB.
Frequency sweep interval may need to be extended to lower frequencies for further use of data
for EVT1 calculations.
The test is performed within the linear visco-elastic range, while permanent deformation in
pavements can be linked to non-linearity at high stresses.
The test enables discrimination between elastomer modified binders and paving grade binders
by analysis of the phase angle curve.
The results can be dependent on the apparatus used. The standard is imprecise, allowing
different interpretations of the data. Consequently, analysis is difficult and experience is needed
for meaningful interpretation.
The test evaluates visco-elastic properties of binder, but it is necessary to solve practical
aspects related to the definition of a final numerical result (curve, reference values).
Maybe more repetitions are necessary for good accuracy. Care should be taken with sample
preparation and thermal history.
Precision: The repeatability can be quite high, up to 20-30% on G* [Ref. “Complex Modulus of
Bituminous Binders: Results of the Round Robin Test of the GE1 working group (France)” - B.
Eckmann et al, Eurasphalt & Eurobitume 2008 Congress, Copenhagen 21-23 May 2008].
In general, the standard sets a rather wide precision range; this needs to be assessed.
Timing: Full testing may take some time (temperatures and frequencies), but can be fully
automatic with modern DSR equipment. Testing over a full sweep is not considered as a quality
control test. Temperature sweep at a specific frequency could be a faster option and in general
faster than e.g. R&B or penetration test.
The test takes 2 hours to one day per sample, depending on the sample preparation and testing
conditions. To obtain results this test is faster than LSV and ZSV.
Comments from BiTVal: Penetration has been correlated with DSR measurements. It is
generally considered that, for paving grade bitumen, the R&B softening point is equivalent to a
penetration of 800 x 0.1 mm. From the relationship between log(G*) and log(pen), it is possible
to calculate the value of G* which equates to 800 pen (G*800 pen). Measurement of G* at more
than one temperature enables a relationship between G* and temperature to be established and
it is then possible to determine the temperature which corresponds to G*800 pen, nominally the
R&B softening point. It has been found that the stiffness of the binder can be predicted from the
penetration index and R&B softening point for paving grade bitumen. At very low testing
frequency, the ratio G*/sin δ is related to the oscillation ZSV. Hence, there is also a relation with
creep ZSV.

TF Data Collection Position Paper – May 2009 16

Correlation with other test methods from this Data Collection:

Complex Modulus G* at 60°C and 1.59 Hz (10 rad/s) on fresh binders

Paving
Correlation with All binders grades, hard Paving Paving PMB only
paving grades and grades only
grades and hard paving
special grades
binders
LSV – EVT1 (fresh) R² = 0.96, log
LSV – EVT2 (fresh) R² = 0.92, log
Temperature @ G*/sin δ = R² = 0.96, log
1 kPa (fresh)
LSV – EVT1 (short term R² = 0.91, log
aged)
Def. Energy FD 20°C R² = 0.95, lin
(long term aged)

Complex Modulus G* at 60°C and 1.59 Hz on short term aged binders

Paving
Correlation with All binders grades, hard Paving Paving PMB only
paving grades and grades only
grades and hard paving
special grades
binders
LSV – EVT1 (fresh) R² = 0.92, log
LSV – EVT2 (fresh) R² = 0.91, exp
ZSV (fresh) R² = 0.96, lin
LSV – EVT1 (short term R² = 0.93, log
aged)
ZSV (short term aged) R² = 0.94, lin
Temperature @ G*/sin δ = R² = 0.92, log
2.2 kPa (short term aged)
Def. Energy FD 20°C R² = 0.94, lin
(long term aged)

Temperature at which G*/sin δ = 1 kPa (fresh binders)

Paving
Correlation with All binders grades, hard Paving Paving PMB only
paving grades and grades only
grades and hard paving
special grades
binders
LSV – EVT1 (fresh) R² = 0.98, lin
LSV – EVT2 (fresh) R² = 0.96, lin
ZSV (fresh) R² = 0.97, exp
LSV – EVT1 (short term R² = 0.92, lin
aged)
LSV – EVT2 (short term R² = 0.90, lin
aged)
ZSV (short term aged) R² = 0.92, exp

TF Data Collection Position Paper – May 2009 17

Temperature at which G*/sin δ = 2.2 kPa (short term aged binders)
Paving
Correlation with All binders grades, hard Paving Paving PMB only
paving grades and grades only
grades and hard paving
special grades
binders
LSV – EVT1 (fresh) R² = 0.91, log
LSV – EVT2 (fresh) R² = 0.94, lin
ZSV (fresh) R² = 0.94, exp
LSV – EVT1 (short term R² = 0.95, lin
aged)
LSV – EVT2 (short term R² = 0.93, lin
aged)
ZSV (short term aged) R² = 0.98, exp

Observation/remarks: No correlation was found between DSR results at high temperature and
penetration. A good correlation is confirmed between G* and R&B temperature but only if PMB
are not taken into account.
A mathematical correlation is found between shear modulus and FD results on long term aged
binder, but no physical reason is seen for that correlation.
Suitability for future specification: Temperature and frequency susceptibilities are
fundamental properties of binders and, as such, may be useful for future specifications.
However the standard, as such, provides too much data to be used in specifications. Suitable
parameters and limits would need to be determined from mixture properties and field
performance.
A combination with ageing procedures could be considered.

4.2. Low Shear Viscosity in DSR – LSV [WI 336067, later prTS 15324]: EVT1 and
EVT2
General Comments: The procedure can be used for all binders, but the protocol could be
improved:
¾ No need is seen to carry out the temperature sweep at each temperature degree in order
to generate acceptable EVT1 data.
¾ Indication of a starting temperature (lowest temperature) or range of temperatures would
be useful for the temperature sweep (only a viscosity range is recommended).
¾ The calculation of EVT2 is not straight-forward, which can easily lead to mistakes. The
extrapolation supposes that there is a plateau at low frequencies, which may not be the
case for all binders, especially PMB.
Note that data for EVT1 calculations might also be obtained by frequency sweep in DSR, if the
frequency interval is extended to include lower frequency values (see 4.1). Doing this would
avoid performing several tests on the same binder and thus save time.
Precision: According to the precision data given in the project standard (Round Robin test
performed in 2003), it is estimated that R (paving grade bitumen) = 2.0 to 2.8°C and R (PMB) =
4.3 to 5.3°C. This is comparable or worse than Reproducibility indicated for R&B softening point
(resp. 2.0°C for paving grade bitumen and 3.5°C for PMB).
Note that this Round Robin test was performed on an earlier version of the protocol.
Timing: Between half a day and a full working day. The testing is shorter than that of ZSV, but
can be longer than a frequency sweep at high temperature (complex modulus see 4.1). EVT2
results need long testing time because of low frequency.
Comments from BiTVal: Probably there is an acceptable link with rutting for EVT1, but in case
of EVT2, the frequency may be too low to be linked with rutting.

TF Data Collection Position Paper – May 2009 18

Correlation with:

LSV – EVT1 (fresh binders)

Paving grades, Paving
Correlation with All binders hard paving grades and Paving PMB
grades and hard paving grades only
special binders grades only
LSV – EVT2 (fresh) R² = 0.96, exp
ZSV (fresh) R² = 0.97, exp
LSV – EVT1 (short term R² = 0.90, lin
aged)
LSV – EVT2 (short term R² = 0.91, lin
aged)
ZSV (short term aged) R² = 0.93, exp
Def. Energy FD 20°C (long R² = 0.95, log
term aged)

LSV – EVT1 (short term aged binders)

Paving grades, Paving grades
Correlation with All binders hard paving and hard Paving PMB
grades and paving grades grades only
special binders only
LSV – EVT2 (fresh) R² = 0.93, exp
ZSV (fresh) R² = 0.94, exp
LSV – EVT2 (short term R² = 0.96, exp
aged)
ZSV (short term aged R² = 0.96, exp
binder)
Vialit pendulum maximum R² =
(short term aged) 0.923, exp

LSV – EVT2 (fresh binders):

Paving grades, Paving
Correlation with All binders hard paving grades and Paving PMB only
grades and hard grades only
special binders paving
grades
ZSV (fresh) R² = 0.98, exp
SP R&B (short term aged) R² = 0.91, lin
LSV – EVT2 (short term R² = 0.93, lin
aged)
ZSV (short term aged) R² = 0.95, exp
BBR Temperature at R² = 0.93, log
which m = 0.3 (long term
aged)
Def. Energy FD 20°C R² = 0.99, log
(long term aged)

TF Data Collection Position Paper – May 2009 19

LSV – EVT2 (short term aged binders)
Paving grades, Paving
Correlation with All binders hard paving grades and Paving PMB only
grades and hard grades only
special binders paving
grades
ZSV (fresh) R² = 0.93, exp
Def. Energy FD 25 (fresh) R² = 0.97, log
G* @ 60°C and 1.6 Hz R² = 0.90,
(short term aged) exp
ZSV (short term aged) R² = 0.96, exp

Observation / remarks: EVT1 is well correlated with dynamic viscosity at 60°C and is
measurable on more datasets than dynamic viscosity at 60°C: EVT1 was measured on 131
binders (and no ‘non measurable’ values were recorded) while dynamic viscosity at 60°C was
measured only on 97 binders and was non measurable for 26 binders (23 PMB and 3 hard
grades). EVT1 could possibly be proposed instead of dynamic viscosity at 60°C.
A mathematical correlation is seen between EVT1 (fresh binder) and FD at 20°C (long term
aged binder), but no physical reason is seen for that correlation. The same remark applies for
the correlation seen between EVT2 and BBR results.
Suitability for future specification: This test may be useful for future specifications, especially
for PMB. EVT1 is more preferred than EVT2 as there is a very good correlation between EVT1
and EVT2. In addition, it is less time consuming to measure only EVT1 and this parameter could
also be deduced from a frequency sweep which would avoid performing several tests on the
same binder and thus save time.

4.3. Zero Shear Viscosity [WI 336068, later prTS 15325]

General Comments: The test method is based on coherent theoretical concepts and
fundamental rheological properties that can be addressed with the latest generations of DSRs.
However, the relevance of the test can be questioned for PMB due to the following reasons:
¾ A long time is needed to reach the testing plateau. The standard sets 4 h to reach the
plateau, which is not always sufficient. For most PMB the final result does not provide the
ZSV, but an approximation of the ZSV due to the fact that the viscosity depends on the
shear. A very poor repeatability is observed for PMB (see “precision” below).
¾ Very often acceptable results cannot be obtained, ZSV being higher than the
recommended measurement range (100-50 000 Pa·s); independent of polymer content or
ageing conditioning.
For soft binders characterised in this Data Collection (Penetration>150 0.1mm), the ZSV
measured is most often below the recommended measurement range.
Precision: The repeatability is reasonably good for unmodified bitumen, but poor for PMB with
variations up to more than 50% for the same binder tested in the same lab [Ref. Rheometrical
quantification of Bituminous Binders for specification purposes by S. Nigen-Chaidron, E&E 2008
congress].
In the provisional standard the results of a Round Robin test are displayed. Standard deviation
for reproducibility is as high as 17% meaning that reproducibility could reach a level of ~50%.
Timing: For unmodified bitumen, the test can be performed within a half working day while for
PMB it takes between 12 hours and two days to obtain results. For PMB it is often necessary to
perform 3 test repeats. If two values are close to each other and the 3rd is different it is not clear
which of them is valid.
Comments from BiTVal: None.

TF Data Collection Position Paper – May 2009 20

Correlation with:

ZSV (fresh binders)

Paving
Correlation with All binders grades, hard Paving Paving PMB
paving grades grades and grades only only
and special hard paving
binders grades
ZSV (short term aged) R² = 0.97, lin
Def. Energy FD 20°C (long R² = 0.96, lin
term aged)

ZSV (short term aged binders)

Paving
Correlation with All binders grades, hard Paving Paving PMB
paving grades grades and grades only only
and special hard paving
binders grades
Def. Energy FD 25°C (fresh) R² = 0.91, lin
Def. Energy FD 20°C (long R² = 0.94, lin
term aged)

Suitability for future specification: The shortcomings of this test make it unsuitable for
inclusion in a future specification. Ring and ball softening point is simpler to test and gives
similar interpretation for conventional binder. Complex modulus and phase angle or EVT1 are
preferred, especially for PMB (better precision). The determination of ZSV takes too much time
for specification purposes.

TF Data Collection Position Paper – May 2009 21

Part 5: Intermediate Service Temperature Characteristics

5.1. Complex Modulus G* and phase angle in DSR [EN 14770]: temperature
sweep between 10 and 40°C (at given frequencies) and frequency sweep between
0.1 and 10 Hz (at given temperatures)
General Comments: Problems occurred in some cases when measuring hard paving grade
bitumen (penetration ≤ 20 0.1 mm) at T=10ºC, due to the limitations of the DSR equipment.
Some older rheometers may have torque limitations in case of testing hard paving grades at low
temperatures. However, the procedure seems to be satisfactory for most binders when suitable
DSRs are used.
Measurement system compliance may be an issue when testing hard paving grade binders at
low temperatures.
Maybe more repetitions are necessary for good accuracy. Care should be taken with sample
preparation and thermal history.
Precision: The repeatability can be quite high, up to 20-30% on G* [Ref. “Complex Modulus of
Bituminous Binders: Results of the Round Robin Test of the GE1 working group (France)” B.
Eckmann et al, Eurasphalt & Eurobitume 2008 Congress, Copenhagen 21-23 May 2008].
In general, the standard sets a rather wide precision range, this needs to be assessed.
Timing: Full testing may take some time (temperatures and frequencies), but normally fully
automatic with recent DSRs. Testing over a full sweep is not considered as a quality control
test. A temperature sweep at a specific frequency could be a faster option and in general faster
than e.g. R&B or penetration test.
The test takes 2 hours to one day per sample, depending on the sample preparation and testing
conditions.
Comments from BiTVal: see item 4.1.
Correlation with:

Complex Modulus G* at 20°C on fresh binders

Paving
Correlation with All binders grades, hard Paving Paving PMB only
paving grades and grades only
grades and hard paving
special grades
binders
BBR Temperature at which R²= 0.91, log
m = 0.3 (fresh)

Observation / remarks: BiTVal stated that correlation has been found in the papers between
penetration and DSR measurements at 25°C. TF Data Collection performed tests at 20°C and
30°C and no such correlation was found at any frequency between 0.1 and 10 Hz. TF DC
believes that R2 values of 0.8 do not show a correlation while literature sometimes regards this
value as a proof for correlation.
Suitability for future specification: Temperature and frequency susceptibilities are
fundamental properties of binders and, as such, may be useful for future specifications.
However the standard, as such, provides too much data to be used in specifications. Suitable
parameters and limits would need to be determined from mixture properties and field
performance.
A combination with ageing procedures could be considered.

TF Data Collection Position Paper – May 2009 22

Part 6: Low Service Temperature Characteristics

6.1. Bending Beam Rheometer BBR [EN 14771]: Determination of Stiffness and
m-value at -16°C as well as determination of temperature at which stiffness S =
300 MPa and m-value = 0.3 respectively
General Comments: Stiffness is a fundamental property of a binder, whereas m-value is not.
Low service temperature properties of binder are generally important in a pavement and low
temperature failure typically occurs late in service life. Therefore it is sensible to address the low
temperature behaviour of long term aged binders. However, because very good correlation was
found between some BBR properties on fresh and on long term aged binder it is under
discussion if there is a need to carry out the long term ageing step.
To determine the temperature at which S = 300 MPa or m = 0.3 respectively, the testing is
carried out at two different temperatures. However, for the most commonly used paving grade
bitumens in Europe (50/70 and 70/100) stiffness is approximately 300 MPa at -16°C, so for
these binders no additional information is gained when carrying out this test protocol at different
temperatures.
For modified binders, some trends can be observed but there is an issue that the low
temperature performance of PMB may be underestimated due to the m-value being the limiting
factor. Evidence from field performance suggests that PMBs resist fracture propagation better
than unmodified bitumen but this phenomenon could not be determined by this study. Some
field experience suggests better low temperature behaviour of PMB compared to unmodified
binders, but this is not always evident from BBR results. Nevertheless, BBR is probably less
prone to problems than Fraass breaking point, but the test method as currently written still has
deficiencies.
Sample preparation is important and de-moulding of the specimen is critical, as such problems
occur when working with very hard and soft binders. Alternative moulds should be allowed in the
standard, provided that the beam conforms to the size and tolerances specified in the standard.
Precision: Results confirm a good repeatability, but it depends on the specimen preparation
and the final geometry (of the prepared specimen). The precision is considered to be better than
that of Fraass breaking point.
Timing: The test takes 2 to 4 hours per sample and can be carried out faster if it is only
performed at one temperature, but that will also lead to some loss of information (temperature
susceptibility ...).
Correlation with:

BBR S @ -16°C on fresh binders

Paving grades, Paving
Correlation with All binders hard paving grades and Paving PMB only
grades and hard paving grades only
special binders grades
BBR Temp. @ S = 300 R² = 0.93, log
(fresh)
BBR Temp. @ m = 0.3 R² = 0.91, log
(fresh)
BBR S @ -16°C (long R² = 0.94, lin
term aged)
BBR Temp. @ S = 300 R² = 0.92, log
(long term aged)

TF Data Collection Position Paper – May 2009 23

BBR S @ -16°C on long term aged binders
Paving grades, Paving
Correlation with All binders hard paving grades and Paving PMB only
grades and hard paving grades only
special binders grades
BBR Temp. @ S = 300 R² = 0.92, log
(fresh)
BBR Temp. @ m = 0.3 R² = 0.91, log
(fresh)
BBR Temp. @ S = 300 R² = 0.94, log
(long term aged)

BBR m @ -16°C on fresh binders

Paving grades, Paving
Correlation with All binders hard paving grades and Paving PMB only
grades and hard paving grades only
special binders grades
BBR Temp. @ m = 0.3 R² = 0.91, log
(fresh)

BBR m @ -16°C on long term aged binders

Paving grades, Paving
Correlation with All binders hard paving grades and Paving PMB only
grades and hard paving grades only
special binders grades
BBR Temp. @ m = 0.3 R² = 0.91, log
(long term aged)

BBR Temperature at which S = 300 on fresh binders

Paving grades, Paving
Correlation with All binders hard paving grades and Paving PMB only
grades and hard paving grades only
special binders grades
BBR Temp. @ m = 0.3 R² = 0.93, lin
(fresh)
BBR Temp. @ S = 300 R² = 0.96, lin
(long term aged)

BBR Temperature at which S = 300 on long term aged binders

Paving grades, Paving
Correlation with All binders hard paving grades and Paving PMB only
grades and hard paving grades only
special binders grades
BBR Temp. @ m = 0.3 R² = 0.93, lin
(fresh)

BBR Temperature at which m = 0.3 on fresh binders

Paving grades, Paving
Correlation with All binders hard paving grades and Paving PMB only
grades and hard paving grades only
special binders grades
DSR G* at 20°C, 1.6 Hz R² = 0.91, exp
(fresh)

TF Data Collection Position Paper – May 2009 24

Suitability for future specification: BBR is potentially suitable for specification purposes, but
the issue of a possible under-estimation of the performance of PMB needs to be resolved.
Stiffness related data are preferred over m-value data. If used for specifications, a temperature
result, e.g. the temperature at which a particular stiffness is reached, is more preferred than a
stiffness value as it is easier to understand.
In general, no other test but BBR and Fraass breaking point used during this data collection
measured the low temperature properties of a binder. Unless other low temperature tests are
proposed and evaluated one of these two tests may be necessary part of a specification.

TF Data Collection Position Paper – May 2009 25

Part 7: Cohesion

7.1. Deformation Energy with Force Ductility Method [EN 13589 and EN 13703]
General Comments: When carried out according to the present specification for PMB, the
application is limited as not all products reach elongations between 20-40 cm (as required) at
the currently specified temperatures.
The force ductility curve gives some information on the polymer modification, which may be
interesting for the producer and the user of the binder. Focusing on pure numeric figures and
not interpreting the full curve of testing means that useful information may be lost.
Sample preparation is crucial, problems may occur when working with very hard binders. The
de-moulding of the specimen is critical, because they may break or be damaged easily.
As the equipment and the sample preparation is similar or identical to that used for testing
elastic recovery and, in some countries ductility, both the equipment and the experience gained
with it are widely spread over Europe.
The procedure described in the standard allows the use of too many temperatures.
The standard requires the measurement to be started at 5°C and the temperature increased by
5°C steps (up to 25°C), but for example, measurements are not possible on hard paving grades
or PMB at 5 or 10°C. Furthermore, after PAV the test cannot always be carried out due to the
hardness of the aged binder. Therefore the temperature should be indicated for each type of
binder according to type (conventional/PMB) and penetration range. Other options are to start at
25°C and decrease temperature or do the testing only at a given temperature (e.g. 20°C) to be
able to compare all bitumen under similar conditions.
Precision: The precision data of the method were evaluated in an European round robin test
and are given in the standard. They are deemed to be satisfactory, especially if the test is
performed on 3 samples, taking 2 of the results into account.
Timing: The time needed to prepare and perform the test is acceptable, with 2 to 4 hours
needed per sample and temperature.
Comments from BiTVal: Correlation between the maximum energy of the force ductility curve
and penetration has been found. The force ductility curve gives qualitative information on the
cohesive and elastic properties of polymer modified bitumen. These properties are determined
by the polymers in terms of their type, distribution, concentration and network. The force ductility
device can also be used for the determination of the elastic recovery, EN 13398.
Correlation with:

Force ductility on fresh binders

Paving grades, Paving
Correlation with All binders hard paving grades and Paving PMB only
grades and hard paving grades only
special binders grades
FD 10: Def. Energy FD 10 R² = 0.92, lin
(short term aged)
FD 5: Def. Energy FD 15 R² = 0.91, exp
(long term aged)

TF Data Collection Position Paper – May 2009 26

Force ductility on short term aged binders
Paving grades, Paving
Correlation with All binders hard paving grades and Paving PMB only
grades and hard paving grades only
special binders grades
FD 15: Def. Energy FD 10 R² = 0.93, log
(fresh)
FD 5: Def. Energy FD 5 R² = 0.99, lin
(fresh)

Suitability for production / quality control: Yes, under clearly specified testing conditions, i.e.
temperature. The data curves give useful information on the binder properties and the testing
procedure is easy to perform. Additionally the testing equipment is already widely used.
Suitability for future specification: The test may be suitable for future specification, but it is
recommended that the test temperature should be clearly specified for each binder type, or a
temperature relating to the binder properties should be chosen. The test is not recommended
for specification purposes after long term ageing, because either not all binders can be tested
when long term aged or a good correlation exists between deformation energy on fresh and on
long term aged binders. Furthermore, as the procedure is limited to one specific speed it is
therefore also limited in performance prediction.
The results allow a differentiation between polymer modified and unmodified binders.

7.2. Deformation Energy with Tensile Test Method [EN 13587 and EN 13703]
General Comments: The equipment to perform the tensile test is not widely spread throughout
laboratories. Within this data collection, only 3 datasets contained results from testing the fresh
and the short term aged binder with tensile test and the test was carried out at 5°C only. The
limited use of this test must be taken into account whenever discussing future specification
schemes.
For the data collection project the test was limited to a temperature range between 5 and 25°C.
The test does not add new information compared to the Force Ductility Test. If performed at
lower temperatures and with low speed, additional information on the low temperature
properties of a binder might be obtained.
Precision: The repeatability of the test is deemed to be fair. When working at low temperatures,
the sample preparation is difficult.
Timing: Approximately half a working day is needed to carry out the test.
Comments from BiTVal: The tensile test has several common features with the direct tensile
test (DTT) despite the procedures being intended for different purposes. In the DTT, a ductile-
brittle transition temperature is sought whereas, in the tensile test to EN 13587, a sample is
rejected if a brittle break occurs. The tensile test is similar to the force-ductility test where the
elastic/rubbery properties are tested with the elongation in these tests normally being >100 %.
DTT is a low temperature test and the result is a temperature whereas the tensile test is an
elongation test and the result is an elongation and force. Nevertheless, the specification for the
main components of the equipment for both tests, except the attachment devices, are
compatible, making it possible to use the same basic equipment with most of the test machines
and temperature chambers available on the market for both tests with the necessary
modifications. A single stretching speed of 1 mm/min is required for both tests and the test
temperature range is common to both procedures.
Similarly defined, but not equivalent, tensile properties can be measured through the force
ductility test to EN 13589. Although readily available and less costly than the tensile test
equipment (providing the ductilometer has not also to be purchased), the force ductility
apparatus has the drawback of a narrower test temperature range of 5ºC to 25ºC.
EN 13703 for deformation energy states two calculation procedures for the conventional energy,
depending on the method followed. Different specification limits have to be fixed for each test
result.

TF Data Collection Position Paper – May 2009 27

Suitability for production / quality control: Not enough data are available to give a
statement.
Suitability for future specification: The test is not widely used and there is little experience at
European level.

7.3. Vialit Pendulum [EN 13588]

General Comments: The test is only applicable to surface dressing binders and should not be
used in other areas.
The choice of the test temperatures depends on the analyst. This leads to a limited
comparability between results from different sources.
Within this data collection, only 22 datasets contained results from testing the fresh binder with
Vialit Pendulum (16 on short term aged binders, 3 on long term aged binders). The limited use
and applicability of this test must be taken into account whenever discussing future specification
schemes.
Precision: A limited study on the repeatability was carried out by Energy Institute in UK; all in
all the repeatability of the test is rather poor.
Timing: The test is time consuming; it takes nearly a week to carry out the test under all
circumstances to be taken into account.
Comments from BiTVal: Although many test methods measure related properties and
therefore there will be some relationship, no formal correlation has been found in the papers
reviewed between the Vialit pendulum test and other bitumen tests.
Correlation with:

Vialit pendulum maximum on fresh binders

Paving grades,
Correlation with All binders hard paving Paving Paving PMB only
grades and grades and grades only
special binders hard paving
grades
Vialit maximum (short R² = 0.967, log
term aged)

Suitability for production / quality control: Limited to binders for surface dressing purposes.
Suitability for future specification: It must be kept in mind that this test is not appropriate for
all binders in general but for a limited application (surface dressings) only.

TF Data Collection Position Paper – May 2009 28

Annex 1 – Data Collection template

PART 1: PRODUCT DESCRIPTION ACCORDING TO CONVENTIONAL SPECIFICATIONS

EN 12591 Paving Grades - prEN 13924 Hard Grades - EN 14023 Polymer Modified Bitumens

Product type Normal Paving Grade

Hard Grade
PMB
Special Bitumen

Characteristic Test method Unit Reported value

Binder condition
Fresh STA LTA

Nominal penetration range EN 1426 0.1 mm

Penetration @ 25 °C EN 1426 0.1 mm

Softening point R&B EN 1427 °C
Penetration index (2) -
Dynamic viscosity @ 60 °C EN 12596 Pa.s
2
Kinematic viscosity @ 135 °C EN 12595 mm /s
Fraass breaking point EN 12593 °C

Elastic recovery @ 10 °C (1) EN 13398 %

Elastic recovery @ 25 °C (1) EN 13398 %
Storage stability (1) EN 13399 °C

Resistance to hardening EN 12607-1

Change in mass EN 12607-1 %
Retained penetration @ 25 °C EN 1426 %
Increase in softening point R&B EN 1427 °C

Flash point EN ISO 2592 °C

Solubility EN 12592 %
Density EN ISO 3838 -
(1) to be reported for PMB's only
(2) Normative Annex B of EN 12591:1999 Calculation of the penetration index (PI Pfeiffer)

TF Data Collection Position Paper – May 2009 29

 PART 2: PERFORMANCE-RELATED PROPERTIES

Indicate the Long Term Ageing Procedure tyemperature

PAV at ...°C [pr EN 14769] 11
RCAT [WI 336...] 5

Characteristics Test method Unit Reported value

Binder condition
Fresh STA LTA
At high service temperature
Complex modulus (DSR)
G* and phase angle prEN 14770 kPa / deg
- for temperature sweep 40 - 80 °C (at use separate input
certain frequencies) sheet
- for frequency sweep 0.1 - 10 Hz
(at certain temperatures)

Low Shear Viscosity (DSR)

Equiviscous temperature EVT1 WI 336067 °C
@ LSV = 2,0 kPa.s and 0,1 rad/s
Equiviscous temperature EVT2 WI 336067 °C
@ LSV = 2,0 kPa.s and 0,001 rad/s

Zero Shear Viscosity (DSR)

Zero shear viscosity (creep mode) WI 336068 kPa.s
@ 60°C

At intermediate service temperature

Complex Modulus (DSR):
G* and phase angle prEN 14770 kPa / deg
- for temperature sweep 10 - 40 °C (at use separate input
certain frequencies) sheet
- for frequency sweep 0.1 - 10 Hz (at
certain temperatures)

At low service temperature

Bending Beam Rheometer (BBR) prEN 14771
Stiffness @ -16°C MPa
m-value @ -16°C -
Temperature Stiffness @ 300 MPa °C
Temperature m-value @ 0.300 °C

Cohesion (choice of test)

2
Force ductility @ 5 °C EN 13589 -13703 J/cm
2
Force ductility @ 10 °C EN 13589 -13703 J/cm
2
Force ductility @ 15 °C EN 13589 -13703 J/cm
2
Force ductility @ 20 °C EN 13589 -13703 J/cm
2
Force ductility @ 25 °C EN 13589 -13703 J/cm

2
Tensile test @ 5 °C EN 13589 -13703 J/cm
2
Tensile test @ 10 °C EN 13589 -13703 J/cm
2
Tensile test @ 15 °C EN 13589 -13703 J/cm
2
Tensile test @ 20 °C EN 13589 -13703 J/cm
2
Tensile test @ 25 °C EN 13589 -13703 J/cm

2
Vialit pendulum: maximum EN 13588 J/cm

TF Data Collection Position Paper – May 2009 30

Annex 2 – List of test methods used during the project

In order to assure that the Eurobitume Data Collection project, which is a contribution to the
CEN Data Collection project, will be carried out according to similar editions of published test
methods, this document provides an overview of the standards which may be of relevance for
this project.

Standard Title of document Edition

EN 1426 Bitumen and bituminous binders – Determination of needle October 1999

penetration

EN 1427 Bitumen and bituminous binders – Determination of softening October 1999

point – Ring and Ball method

EN 12591 Bitumen and bituminous binders – Specifications for paving November 1999
grade bitumen

EN 12592 Bitumen and bituminous binders – Determination of solubility November 1999

EN 12593 Bitumen and bituminous binders – Determination of the November 1999

Fraass breaking point

EN 12594 Bitumen and bituminous binders – Preparation of test December 1999

samples

EN 12595 Bitumen and bituminous binders – Determination of kinematic November 1999

viscosity

EN 12596 Bitumen and bituminous binders – Determination of dynamic November 1999

viscosity by vacuum capillary

EN 12607-1 Bitumen and bituminous binders – Determination of the October 1999

resistance to hardening under influence of heat and air – Part
1: RTFOT method

EN 13398 Bitumen and bituminous binders – Determination of the December 2003

elastic recovery of modified binders

EN 13399 Bitumen and bituminous binders – Determination of storage December 2003

stability of modified binders

EN 13588 Bitumen and bituminous binders – Determination of cohesion February 2004

of bituminous binders with pendulum test

EN 13589 Bitumen and bituminous binders – Determination of the December 2003

tensile properties of modified bitumen by the force ductility
method

EN 13703 Bitumen and bituminous binders – Determination of December 2003

deformation energy

prEN 13924 Bitumen and bituminous binders – Specifications for hard December 2005
paving grade bitumens

TF Data Collection Position Paper – May 2009 31

 Standard Title of document Edition

EN 14023 Bitumen and bituminous binders – Framework specification October 2005

for polymer modified bitumens

EN 14769 Bitumen and bituminous binders – Accelerated long-term October 2005

ageing conditioning by a Pressure Ageing Vessel (PAV)

EN 14770 Bitumen and bituminous binders – Determination of complex October 2005

shear modulus and phase angle - Dynamic Shear Rheometer
(DSR)

EN 14771 Bitumen and bituminous binders – Determination of the May 2005

flexural creep stiffness – Bending Beam Rheometer (BBR)

prEN 15323 Bitumen and bituminous binders – Accelerated long-term August 2005
ageing conditioning by the rotating cylinder method (RCAT)

prEN 15324 Bitumen and bituminous binders – Determination of August 2005

equiviscous temperature based on Low Shear Viscosity using
a Dynamic Shear Rheometer in low frequency oscillation
mode

prEN 15325 Bitumen and bituminous binders – Determination of Zero August 2005
Shear Viscosity (ZSV) using a Dynamic Shear Rheometer in
creep mode

EN ISO 2592 Petroleum products – Determination of flash and fire points – August 2001
Cleveland open cup method

EN ISO 3838 Crude petroleum and liquid or solid products, Determination May 2004
of density or relative density – Capillary-stoppered
pyknometer and graduated bicapillary pyknometer methods

TF Data Collection Position Paper – May 2009 32