Transactions, SMiRT-23

Manchester, United Kingdom - August 10-14, 2015

Division II: Fracture Mechanics and Structural Integrity

THE R6 DEFECT ASSESSMENT PROCEDURE:

STATUS AND KEY RECENT DEVELOPMENTS
Peter J Budden1 and John K Sharples2
1
R6 Lead Engineer, EDF Energy, UK
2
Structural Integrity Consultant, Amec Foster Wheeler, Clean Energy, UK

ABSTRACT

The R6 procedure for the assessment of the integrity of structures containing defects has been
continuously developed since 1976 to meet the needs of the UK power industry for defect tolerance
arguments supporting the continued operation of nuclear power plant. It is routinely used by EDF Energy
for the fracture assessment of metallic components in its Advanced Gas-Cooled Reactor (AGR) and
Pressurised Water Reactor (PWR) fleet. As the plant ages, requirements to minimise unnecessary
conservatism increase, and the challenges to maintaining margins increase. Hence the potential for the use
of novel, more advanced methods also grows.

This paper first summarises the current status of the R6 documentation and its maintenance and
development programme. Particular developments in three technical areas are then briefly highlighted: (1)
the treatment of welding residual stress in fracture assessments, including bounding residual stress
profiles and numerical simulation of residual stress; (2) the treatment of secondary stresses in fracture
assessment, including the interaction of primary and secondary stress; and (3) limit load solutions. Other
ongoing key developments are also briefly outlined.

INTRODUCTION

The R6 procedure (EDF Energy, 2015) was first published in 1976. Fracture mechanics was being used in
the design assessment of proposed new civil nuclear plant in the UK; in particular, R6 was used for the
pre-service safety case for the Sizewell PWR. At the time of its first issue, a failure assessment diagram
(FAD) was used to assess elastic-plastic fracture and plastic collapse by evaluating the ratio, Sr, of
primary load to limit load based on flow stress, σf, defined as the average of the 0.2% proof stress, σy, and
the ultimate tensile stress, σu. The shape of the failure assessment curve, f(Sr), and hence the value of the
J-integral crack driving force, was defined by a strip-yielding approximation, with fracture conceded
when K r = f (S r ) where, for primary loads alone, Kr is the ratio of stress intensity factor to fracture
toughness. The scope of R6 was expanded by the issue of Revisions 1 and 2 in 1977 and 1980,
respectively. However, the first major update was Revision 3 in 1986. This introduced the current
notation Lr for the abscissa on the FAD, where Lr is defined using σy for the limit load, so that
Lr = (σf / σy )Sr . Collapse is conceded at L r = Lmax
r = σ f / σ y > 1 . Options were also introduced for the
shape of the failure assessment curve, K r = f ( L r ) , with Option 2 a material-specific curve based on the
reference stress J-estimation scheme (Ainsworth, 1984), and Option 1 an explicit function of Lr derived as
an approximate lower bound to Option 2 curves for a range of austenitic and ferritic steels.

Development of R6 between 1986 and 2000 focussed on issuing new appendices which significantly
expanded the scope of the document. However, following the production of the European SINTAP
procedure (Ainsworth et al., 2001), it was decided that a further major update was appropriate and this
appeared as Revision 4 in 2001. This adopted the current structure of R6 based on 5 chapters each
containing a number of sections, with Chapters I-II providing the basic procedure, Chapter III a range of
 23rd Conference on Structural Mechanics in Reactor Technology
Manchester, United Kingdom - August 10-14, 2015
Division II: Fracture Mechanics and Structural Integrity

alternative, more advanced procedures, Chapter IV stress analysis and residual stress profile inputs, and
Chapter V comprising validation and worked examples. Since 2001, a series of Amendments have been
issued, the most recent being in March 2015 as Amendment 11.

Maintenance and development of R6 is managed by EDF Energy under a Structural Integrity Assessment
Procedures collaboration which involves Amec Foster Wheeler, Rolls-Royce, TWI, Frazer-Nash
Consultancy and NRG, Netherlands. R6 continues to be developed by EDF Energy and its partners to
meet the requirements for the defect tolerance leg of safety cases, primarily driven by the requirements of
the UK AGR and PWR nuclear reactors, the challenges of ageing plant, and the research requirements of
the UK civil nuclear industry regulator, the Office for Nuclear Regulation (ONR).

This paper first briefly summarises the principal content of the changes to R6 Revision 4 at Amendment
11 and then provides some more detail of recent work in three particular technical areas: welding residual
stress; secondary stress; and limit loads. However, the R6 programme covers work in a number of other
technical areas, the details depending on priorities and varying over time. These currently include leak
rate calculation methods (e.g. Gill, 2014) probabilistic leak-before-break arguments (e.g. Bhimanadam
and Blom, 2014), FAD methods for notch-like defects (e.g. Horn and Budden, 2014), load history effects
such as warm pre-stress (e.g. Teng and Sharples, 2012), and constraint effects due to thin sections (e.g.
Kulka and Sherry, 2012) or residual stresses.

CHANGES TO R6 AT AMENDMENT 11

Amendment 11 has introduced revisions to 19 sections within Chapters I-IV of R6 and added one new
section in Chapter V. The major changes are to: Section II.6 on the treatment of secondary stresses;
Sections II.4 and IV.1 on limit loads; and to Sections II.7, III.15 and IV.4, including a new Section V.6,
on welding residual stresses. These developments are summarised in the following three sections of this
paper. Additionally, the leak-before-break methodology of Section III.11 has been updated to refer to
recent work under the R6 programme on evaluating the crack opening area under both elastic-plastic and
creep conditions (e.g. Madew et al., 2011), and to offer additional advice on determination of an
appropriate local or global surface roughness, dependent on the crack opening, for use in determining
friction factors and hence leak rates (e.g. Taggart and Budden, 2008). These are not described further in
this paper. Minor changes have been made to Section II.1 on tensile properties and Section II.2 on
fracture toughness values to maintain these sections up-to-date. The other revisions to R6 at Amendment
11, including to the status notes of Section I.12, are principally to reflect the significant changes to the
methodology of Section II.6, in particular the use of the multiplicative V factor alone to quantify the
effect of the interaction between primary and secondary stresses on the R6 Kr parameter, and hence the
corresponding estimate of J, with the earlier additive ρ term being removed.

WELDING RESIDUAL STRESSES

R6 Section II.7 gives advice on the treatment of weld residual stress and describes methods for estimating
the corresponding stress distribution. It refers in particular to the residual stress compendium of R6
Section IV.4, and to R6 Section III.15 on numerical simulation of weld residual stress for austenitic
materials that was issued in 2009. In addition, validation and worked example sheets for the calculation
of residual stress in weldments have been incorporated in R6 Sections V.5 and V.6, respectively. At
Amendment 11, additional advice has been added to Section II.7 on evaluating the stress intensity factor
due to residual stress, on validation of the post-weld heat treatment rules from BS7910:2013 and on a
range of residual stress measurement methods. On stress intensity factor, KI evaluation, it has been noted
that KI can be calculated using solutions for arbitrary applied stress distributions when these are available
for the geometry being addressed. Alternatively, the stress distribution can be decomposed into
membrane, bending and self balancing components. Solutions for the membrane and bending KI
 23rd Conference on Structural Mechanics in Reactor Technology
Manchester, United Kingdom - August 10-14, 2015
Division II: Fracture Mechanics and Structural Integrity

components may be more readily available. Methods for determining the self balancing component are
suggested in the text and also in the status notes.

The residual stress compendium of Section IV.4 contains upper bound residual stress profiles (so-called
Level 2 profiles) for a range of common welded joints including: Plate Butt and Pipe Seam Welds, Pipe
Butt Welds, T-Butt Welds, Nozzles, Repair Welds, Transition Welds and Weld T-Intersections. Annual
reviews of new information are undertaken to maintain R6 Section IV.4 up-to-date, but have not led to
major revisions to Section IV.4 in recent years. However, for Amendment 11, slight modifications have
been made such that the recommended profiles are now fully consistent with the recent advice in
BS7910:2013 (British Standards, 2013), and additional advice has been added to cover both ferritic and
austenitic steels for the range of weldment geometries addressed.

Development of statistical methods, Teng and Bate (2014), Mathew et. al. (2014) has been ongoing to
improve the current Level 2 profiles. A heuristic method has been developed for determining residual
stress profiles, Teng et al. (2008), based on a combination of the weighted least-squares method and the
application of expert judgement. This has been applied to the residual stress data used to determine the
current profiles for transverse stresses in pipe girth welds and to more recent measurement data. The
method was used to generate a mean residual stress profile based on the combination of the membrane,
bending and self-balancing components. An upper bound profile was then determined on the basis of
95% confidence limits. An example is shown in Figure 1, where stress and distance through the pipe wall
are normalised by the 0.2% proof stress and wall thickness, respectively. This compares the mean
residual stress profile determined using the heuristic method (AMEC, mean) with measured Block
Removal Splitting and Layering (BRSL) data for a low carbon micro-alloyed steel pipe, X52, submerged
arc (SAW) girth weld from He et al. (2010). Also shown is the R6 Level 2 profile for low heat input
welds. The figure shows that the upper bound profile determined using the heuristic approach (AMEC,
95% UB) is not as conservative as the current Level 2 profile but still provides a reasonable upper bound
to the measured data. A neural network approach has also been developed to characterise through-
thickness residual stress profiles in austenitic steel girth welds, Mathew et al. (2014). The network was
trained using measured data collated by Bouchard (2007). An upper bound profile determined using the
neural network approach has been compared to the neutron diffraction measurements on experiments
from the European STYLE project, Keim et al. (2014). Based on the work described above and other
complementary studies, including the provision of additional experimental data, the recommended
transverse and longitudinal residual stress profiles for pipe girth welds are planned to be revised based on
an analysis of the enhanced database using the developed statistical methods. Furthermore, it is proposed
to introduce new upper bound residual stress profiles for electron beam welds. These are likely to be
based on studies undertaken by Hurrell et al. (2014) on four electron beam welded specimens: 316L plate,
304L pipe, SA508-3 forging and an A533B plate.

Finite element analysis (FEA) is being applied more frequently nowadays to predict residual stresses in
welded components for assessment purposes. This involves complex non-linear analysis with many
assumptions and, as a consequence, the accuracy and reliability of solutions is variable. In order to
improve the consistency of weld modelling, and hence the accuracy and confidence in their use, Section
III.15 of R6 was introduced to provide a set of guidelines covering the calculation of residual stresses.
This guidance is currently focused on austenitic materials and provides detailed information under nine
key steps. These steps are (i) problem definition, (ii) identify and compile basic information, (iii)
resource requirements in terms of personnel, software and computer hardware, (iv) decide modelling, (v)
finite element model, (vi) thermal analysis, (vii) mechanical analysis, (viii) validation of weld residual
stress calculations, and (ix) sensitivity studies. Mixed isotropic-kinematic hardening models have
recently been introduced in order to improve the numerical simulation guidance.
 23rd Conference on Structural Mechanics in Reactor Technology
Manchester, United Kingdom - August 10-14, 2015
Division II: Fracture Mechanics and Structural Integrity

AMEC, 95% UB AMEC, mean

R6 BRSL Measurement
1.5

1.0

Normalised stress
0.5

0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2

-0.5

-1.0
Normalised distance

Figure 1. Comparison of residual stress profiles with axial residual stress measurements on a low
carbon micro-alloyed SAW pipe girth weld, He et. al. (2010)

Work is ongoing to extend the guidance for austenitics and for ferritic materials and dissimilar metal
welds where phase changes need to be considered in the modelling process. The validation part of the
guidelines provide “standards” against which the predicted stresses can be validated. A series of “Weld
Residual Stress Benchmarks” are being developed for inclusion in R6 Sections V.5 and V.6. These
describe welded mock-ups in which residual stresses have been measured using various techniques.
Users can refer to these when validating their finite element modelling techniques and thus provide
greater confidence in the predicted results. A number of welded mock-ups have been generated to
support a European collaborative round-robin and industrial research programmes, Bate et al. (2012),
Bate (2013). These include an austenitic single bead-on-plate weldment, the information of which was
introduced in Section V.5 of R6 in 2011, and an austenitic groove welded plate, the information of which
has been introduced at Amendment 11 in Section V.6 (Figure 2). Planned validation cases include an
austenitic pipe girth weld, an austenitic repair weld, a ferritic groove weld and a ferritic bead-on-plate
weldment.

Figure 2. Validation example – austenitic groove weld

SECONDARY STRESSES

The R6 approach to quantifying the contribution of secondary stresses to fracture is to modify the
definition of the Kr parameter, leaving Lr unchanged so that secondary stresses by definition in R6 do not
influence collapse. Prior to Revision 4, this was achieved via an additive ρ parameter, based on Ainsworth
(1986). At Revision 4, an additional approach, based on a multiplicative V-factor, see Ainsworth et al.
(2000), was introduced in parallel to the route based on ρ. This additional approach is represented by the
 23rd Conference on Structural Mechanics in Reactor Technology
Manchester, United Kingdom - August 10-14, 2015
Division II: Fracture Mechanics and Structural Integrity

following equation, with the superscripts p and s on the stress intensity factor KI or J-integral denoting
primary and secondary, respectively:

K r = (K pI + VK sI ) / K mat , equivalent to J = (K pI + VK sI ) 2 /(E′(f (L r ) 2 ) (1)

with E ′ = E /(1 − ν 2 ) , where E and ν are Young’s modulus and Poisson’s ratio, respectively. The V and ρ
approaches are based on the same underlying J-estimation methodology so the results of an assessment
are essentially independent of the choice of method. All recent developments in the R6 approach to
treating secondary stresses have been based on V so at Amendment 11 it was decided to remove ρ and to
focus on the use of V. It should be noted that the use of a multiplicative factor on the secondary terms is
consistent with that employed in both the API 579-1/ASME FFS-1 (2007) and RSE-M (AFCEN, 2005)
procedures.

At Amendment 11, R6 Section II.6 has been re-written. The “simplified V procedure” has been retained
but with a lower limiting value of V of 0.4 at high values of Lr rather than the previous lower limit on V
of unity. This corresponds to an allowance for mechanical stress relief of the secondary stress, noting that
V < 1 is equivalent to ρ < 0. The previous “detailed V procedure” has been retained without change with
V = (K sJ / K sI )ξ where K sJ = E′J s and values of ξ are tabulated as a function of Lr and βsJ = K sJ L r / K pI .
The main changes to Section II.6 are in the provision of a range of additional options to the user.

Firstly, a second “detailed procedure” is set down where (Ainsworth, 2012) ξ is expressed as an explicit
function of Lr alone:

(
V = (K sJ / K sI ) f (L r ) + 0.42L r (0.72 + L r ){f (L r )}2 ) (2)

with f(Lr) being the shape of the failure assessment curve that is used in the overall failure assessment. A
comparison of the tabular and Equation (2) approaches to ξ is shown in Figure 3.
1.4
Table - β sJ = 1

1.2 Table - βs = 4
J
Equation (2)
1
ξ
0.8

0.6

0.4

0.2

0
0 0.5 Lr 1 1.5 2

Figure 3. Comparison of the detailed approaches for ξ based on tabulated values and Equation (2)

Secondly, new guidance on elastic follow-up is introduced into Section II.6. The approaches to estimating
ξ allow VK sI / K sJ = ξ > 1 at small values of Lr, up to a maximum of about 1.2 using Equation (2). At larger
values, VKsI / K sJ falls below unity corresponding to plastic relaxation of secondary stress. The estimates
are intended to be conservative relative to typical behaviour. However, it is known that in some cases
VK sI / K sJ can be less than unity for all values of Lr and, in other cases, the maximum value of
VK sI / K sJ can be markedly greater than 1.2. Examples are given in Section II.6 of both scenarios to aid the
 23rd Conference on Structural Mechanics in Reactor Technology
Manchester, United Kingdom - August 10-14, 2015
Division II: Fracture Mechanics and Structural Integrity

assessor. The cases where VK sI / K sJ is significantly greater than unity correspond to significant elastic
follow-up where the basic procedures, which are appropriate to low or moderate follow-up, are not
appropriate. In cases of significant follow-up, two alternatives are now given. The first is to treat the
secondary stress as primary in the calculation of V; this leads to (Ainsworth, 2012):

K sJ (β s + L r )f ( L r ) − L r f ( L r + β s )
V= s
f (β s ) (3)
KI βs f (L r + βs )

where β s = K sI L r / K pI . The second option involves the use of an elastic follow-up factor, Z. Such a factor
has been used for some time for creep crack growth under combined loading in the R5 procedure (EDF
Energy, 2014) but its explicit use is new to R6. Section II.6 refers to R5 for the quantification of Z. For
Z<3, the simplified and detailed R6 approaches are adequate. However, for Z>3, an estimate of V is

K sJ  Z −1 2 KI 
s
V=  f (L r ) + 0.75 L (β s
+ L ){f ( L )}  (4)
K sI  r r r
K J 
s
 Z

Secondary stress acting as primary corresponds to the limit Z → ∞ . However, Equation (4) does not fully
capture the approach of Equation (3) in this limit, but does reduce to it for small values of primary and
secondary loads with L r + β ≤ 0.7 and so sensitivity studies should be performed when large values of Z
are obtained.

For cases of multiaxial remote stresses with significant stress parallel to the defect plane, an estimate of V
for cases of moderate elastic follow-up is given, following James (2010), as

−1 / 2
K sJ  Eε mod (σ mod
ref / σ y )
2

V = s f ( L r )  mod − A
ref
mod 
(5)
KI  σ ref ( Eε mod
ref / σ ref ) 


where A = 0.8σ1p / σMises

p
and σ mod mod
ref = L r σ y / 1.25 with ε ref denoting the strain on the material uniaxial

stress-strain curve at the modified reference stress, σ mod

ref . The factor of 0.8 in the expression for A was
chosen for conservatism based on a range of finite element calculations and applies for uniform remote
tensile stress. For remote bending, σ pI and σMises
p
are the peak bending stress divided by 3 and 3/2,
p
respectively. More generally, σMises can be defined from the limit load for the uncracked section with
p
σMises = σref and σ pI as the peak tensile stress. This approach may be used for sensitivity studies,
particularly when there are remote applied stresses parallel to the defect plane, but is not applicable to
cases of large elastic follow-up.

Further work since Amendment 11 was finalised is being performed to improve the advice in Section II.6
on the treatment of follow-up (Song et al., 2013) and on the calculation of K sJ (e.g. James, 2015a).
Treating the secondary stresses as primary is clearly conservative but may be overly so. Finite element
analysis may be used to provide a closer definition of margins in particular cases. Sensitivity analysis is
also useful as assessments may not be particularly sensitive to the value of V when β s V / L r is small
(Ainsworth, 2012). Load-order effects are also being considered (James, 2015b; Oh et al., 2014). FE
elastic-plastic J values were compared by Oh et al. (2014) with those derived using the new and the
 23rd Conference on Structural Mechanics in Reactor Technology
Manchester, United Kingdom - August 10-14, 2015
Division II: Fracture Mechanics and Structural Integrity

previous methods in Section II.6. It was shown that there are significant load-order effects for large
secondary stresses but these are successfully treated by the new methods in R6 Section II.6.

LIMIT LOADS

The limit load is one of the principal inputs to R6; it defines the parameter Lr and hence quantifies
proximity to plastic collapse, and to fracture via the definition of the failure assessment curve Kr = f(Lr).
Equivalently, for primary loads alone, it leads to an estimate of the J-integral as J p = K 2 / E ′f 2 (L r ) . A
compendium of limit loads for defective homogeneous test specimens, plates, cylinders and spheres,
under various loads, was included in R6 Revision 4 as Section IV.1 and has been updated several times
since 2001. At Amendment 11, the previous solutions in R6 Section IV.1 for circumferential defects in
thin-walled cylinders under combined internal pressure, p, axial tension, N, and global bending, M, have
been extended to general thick-walled solutions by Lei et al (2014a). Some comparisons of the solutions
against detailed finite element limit load calculations have also been reported by Li et al. (2014). An
example is shown in Figure 4, where nL and mL are normalised limit values of N and M. Generally, the
analytical limit loads are conservatively lower than the finite element data. The solutions have also been
used with R6 Option 2 to compare the estimated J-values with published data for a range of cylinder
radius to thickness ratios, crack sizes, loading types and material strain hardening exponents, n; see Figure
5 for an example. The J-data have also been used to give better advice in R6 Section II.4 for both axial
and circumferential defects in cylinders. In particular, cases where the global limit loads are conservative
for J-estimation are more clearly identified and, in other cases, the previous advice that factors of 0.8 to
0.9 on the global limit load are sometimes necessary to ensure conservatism is confirmed; e.g. Figure 5
(Lei, 2013).

The treatment of defects in pipe bends has been extended in R6 Section II.4 to include references to
sources of information on stress distributions in the uncracked bend under various pressure and bending
situations, which can be used to construct approximate limit loads for the defective bends.

Recent work via the R6 programme has concentrated on extending the scope of the recommended limit
loads to more complex geometries, such as pipe bends and pipe branches, and to more complex loading
conditions, for example plates under biaxial loading conditions. Although, published limit load solutions
for both cracked and uncracked pipe bends have been reviewed within EDF Energy, explicit advice for
inclusion in R6 Section IV.1 awaits further finite element J-solutions, and also the development of
corresponding stress intensity factor solutions for R6 Section IV.3.

Figure 4. Internal surface crack, θ π = 0.2 , a t = 0.8 , N/(πRi2)=1, k = R o R i (Lei et al., 2014a)
 23rd Conference on Structural Mechanics in Reactor Technology
Manchester, United Kingdom - August 10-14, 2015
Division II: Fracture Mechanics and Structural Integrity

15
FE a/t
0.1
0.2
10 0.4
0.5

J/Je
Predicted (Opt-2 FAC based)
Predicted (Opt-1 FAC based)
5 Predicted (Opt-2, 80% limit load)
θ/π = 0.12

0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

M /M L (σ 0 )

Figure 5. External semi-elliptical surface crack, cylinder under bending, θ π = 0.12 , k = 1.25 , n=10 (Lei,
2013). J is normalised by the corresponding elastic value, Je and M by its limit value, ML

The case of the limit load for an extended surface crack in a plate under combined biaxial force and
through-thickness bending is separately addressed in this conference by Aird and Lei (2015) who
compare finite element values of the limit load with theoretical estimates of Lei and Budden (2014). The
latter are shown to be conservative in general, that is, lower than the corresponding finite element values.
Some numerical validation of the R6 solution, by means of comparison of R6 Option 2 J-estimates with
published J values, was given by Lei et al. (2014b). The theoretical solutions have recently been
generalised to surface defects of finite length by Lei and Budden (2015). These solutions for biaxial
loading will be included in future updates to R6 following suitable validation against estimates of J. The
stress intensity factor is not affected by the force parallel to the crack front so changes to R6 Section IV.3
will not be required.

CONCLUSIONS

This paper has briefly described the evolution and current status of the R6 procedure. Recent
developments in the areas of welding residual stresses, secondary stresses and limit loads have been
described. The paper has also discussed ongoing work in the R6 programme, focussing on these 3
technical areas, which is expected to lead to further enhancements to R6 advice in the future. The changes
to R6 Revision 4 at Amendment 11 have been highlighted and include the following.

● The advice on numerical simulation of welding residual stress in austenitic steels in R6 Section III.15
has been improved by the use of mixed isotropic-kinematic hardening models.
● The upper bound, Level 2 welding residual stress profiles in R6 Section IV.4 are now fully consistent
with the recent advice in BS7910:2013, and additional advice has been added to cover both ferritic and
austenitic steels for the range of weldment geometries addressed.
● Additional advice has been added to R6 Section II.7 on evaluating the stress intensity factor due to
residual stress, on validation of the post-weld heat treatment rules from BS7910 and on the range of
residual stress measurement methods.
● A new R6 Section V.6 has been added as a worked example of the Section III.15 approach.
● The treatment of secondary stress in R6 Section II.6 now uses the multiplicative V factor alone. The
conservatism of the simplified V route has been lessened for values of Lr greater than unity by allowing
for mechanical stress relief of the secondary loading.
● A range of new detailed approaches for treating secondary stresses has been added which include
methods based on quantifying elastic follow-up and multiaxiality of the remote loads. Improved advice is
given on situations where follow-up may be significant.
● The previous global limit load solution for thin-walled circumferentially-cracked cylinders under
combined tension, global bending and internal pressure in R6 Section IV.1 has been extended to thick-
 23rd Conference on Structural Mechanics in Reactor Technology
Manchester, United Kingdom - August 10-14, 2015
Division II: Fracture Mechanics and Structural Integrity

walled cylinders. The advice on the use of global limit loads for cylinders with axial or circumferential
cracks in R6 assessments has been expanded and clarified.

ACKNOWLEDGEMENTS

This paper was produced as part of the R6 development programme and is published by permission of
EDF Energy and Amec Foster Wheeler. The authors acknowledge the contributions of colleagues in the
R6 programme to the work summarised herein and R Ainsworth for his comments on this paper.

REFERENCES

AFCEN (2005). “RSE-M: In-service inspection rules for mechanical components of PWR nuclear
islands”, 1997 Edition plus 1998, 2000 and 2005 Addenda, Paris.
Ainsworth, R. A. (1984). “The assessment of defects in strain hardening material”, Engng Frac. Mech.,
19, 633-642.
Ainsworth, R. A. (1986). “The treatment of thermal and residual stresses in fracture assessments”, Engng
Frac. Mech., 24, 65-76.
Ainsworth, R. A., Sharples, J. K. and Smith, S. D. (2000). “Effects of residual stresses on fracture
behaviour – experimental results and assessment methods”, J Strain Analysis, 35, 307-316.
Ainsworth, R. A., Bannister, A. C. and Zerbst, U. (2001). “An overview of the European flaw assessment
procedure SINTAP and its validation”, Int. J. Pres. Ves. Piping, 77, 869-876.
Ainsworth, R. A. (2012). “Consideration of elastic follow-up in the treatment of combined primary and
secondary stresses in fracture assessments”, Engng Frac. Mech., 96, 558-569.
Aird, C. J. and Lei, Y. (2015). “Comparison of analytical and finite element based limit loads for wide
plates containing an extended surface crack under combined biaxial forces and cross-thickness
bending”, Trans. SMiRT-23, August, Manchester.
American Petroleum Institute/ASME (2007). “Fitness-for-Service”, API 579-1/ASME FFS-1,
Washington DC.
Bate, S., Shallcross, N., Stone, K., Francis, J., Mark, A. and Gill, C. (2012), “Development of material
model parameters suitable for the finite element simulation of ferritic welds”, Paper PVP2012-
78299, Proc. ASME PVP Conference, July, Toronto.
Bate, S. (2013). “The development of ferritic weld benchmarks for validating residual stress weld
simulations”, Trans. SMiRT-22, August, San Francisco.
Bhimanadam, V. R. and Blom, F. J. (2014). “Probabilistic fracture analysis of circumferentially cracked
pipes”, Paper PVP2014-28150, Proc. ASME PVP Conference, 20-24 July, Anaheim.
Bouchard, J. (2007). “Validated residual stress profiles for fracture assessments of stainless steel pipe
girth welds”, Int. J. Pres. Ves. Piping, 84,195-222.
British Standards Institution (2013). “Guide to methods for assessing the acceptability of flaws in metallic
structures”, BS7910:2013, London.
EDF Energy (2015). “R6: Assessment of the integrity of structures containing defects”, Revision 4, with
amendments to Amendment 11, Gloucester.
EDF Energy (2014). “R5: Assessment procedure for the high temperature response of structures”,
including revisions to Revision 002, Gloucester.
Gill, P. J. (2014). “A coupled fluid-solid model to investigate leak rates for leak-before-break
sssessments”, Paper PVP2014-28170, Proc. ASME PVP Conference, 20-24 July, Anaheim.
He, W., Wei, L. and Smith, S. (2010), “Intensive validation of computer prediction of welding residual
stresses in a multi-pass butt weld”, 29th Int. Conf. on Offshore Mech. and Arctic Engng (OMAE),
June, China.
Horn, A. J. and Budden, P.J. (2014). “A comparison of notch failure assessment diagram methods for
assessing the fracture resistance of structures containing non-sharp defects”, Paper PVP2014-
28656, Proc. ASME PVP Conference, 20-24 July, Anaheim.
 23rd Conference on Structural Mechanics in Reactor Technology
Manchester, United Kingdom - August 10-14, 2015
Division II: Fracture Mechanics and Structural Integrity

Hurrell, P. R., Pellereau, B. M. E., Gill, C. M., Kingston, E., Smith, D. and Bouchard, P. J. (2014),
“Development of residual stress profiles for defect tolerance assessments of thick section electron
beam welds”, Paper PVP2014-28809, Proc. ASME PVP Conference, 20-24 July, Anaheim.
James, P. (2010). “Treatment of secondary and residual stresses: guidance to R6 regarding the interaction
of primary and secondary stresses from work performed between 2005-09”, Serco Report
SERCO/TAS/3375.03/001 Issue 2, Warrington.
James, P. (2015a). “Calculation of KJ under secondary loads in isolation for R6”, Proc ASME PVP
Conference, Paper PVP215-45274, July 19-23, Boston.
James, P. (2015b). “Investigation into load order effects on the treatment of combined primary and
secondary stresses within R6”, Trans. SMiRT-23, Manchester.
Keim, E., Sharples, J. and Nicak, T. (2014), “FP7 project STYLE: Implications and recommendations,
discussion”, Proc. ASME PVP Conference, Paper PVP2014-28254, July, Anaheim.
Kulka, R. S. and Sherry, A. H. (2012), “Fracture toughness evaluation in C(T) specimens with reduced
out-of-plane constraint”, Proc. ASME PVP Conference, Paper PVP2012-78751, July, Toronto.
Lei, Y. (2013). “Numerical validation of limit load solutions for defective cylinders in R6”, EDF Energy
Report E/REP/BBGB/0114/GEN/13, Gloucester.
Lei, Y. and Budden, P. J. (2014). “Limit load solutions of plates with extended cracks under combined
biaxial forces and cross-thickness bending”, J. Strain Analysis Analysis, 49, 533-546.
Lei, Y., Li, Y. and Gao, Z. (2014a). “Global limit load solutions for thick-walled cylinders with
circumferential cracks under combined internal pressure, axial force and bending moment – part
I: theoretical solutions”, Int. J. Pres. Ves. Piping, 114-115, 23-40.
Lei, Y., Budden, P. J. and Aird, C. J. (2014b). “The effect of biaxial loading on the limit load and J
estimates for plates with extended surface cracks, 20th European Conference on Fracture,
Trondheim, Procedia Materials Science, 3, 667-672.
Lei, Y. and Budden, P. J. (2015). “Global limit load solutions for plates with surface cracks under
combined biaxial forces and cross-thickness bending”, under review.
Li, Y., Lei, Y. and Gao, Z. (2014). “Global limit load solutions for thick-walled cylinders with
circumferential cracks under combined internal pressure, axial force and bending moment – part
II: finite element validation”, Int. J. Pres. Ves. Piping, 41-60, 23-40.
Madew, C., Sharples, J., Charles, R. Gill, P. and Budden, P. (2011), “The evaluation of crack opening
areas for through-wall cracks in the vicinity of pipe branch connections”, Proc. ASME PVP
Conference, Paper PVP2011-57312, July 17-21, Baltimore.
Mathew. J, Moat, R. J. and Bouchard, P. J. (2014), “Optimised neural network prediction of residual
stress profiles for structural integrity assessment of pipe girth welds”, Proc. ASME PVP
Conference, Paper PVP2014-28845, July, Anaheim.
Oh, C.-Y., Nam, H.-S., Kim, Y.-J., Ainsworth, R. A. and Budden, P. J. (2014). “FE validation of R6
elastic-plastic J estimation for circumferentially cracked pipes under mechanical and thermal
loadings”, Engng Fract. Mech., 124-125, 64-79.
Song, T.-K., Oh, C.-Y., Kim, Y.-J., Ainsworth, R. A. and Nikbin, K. (2013). “Approximate J estimates
for combined primary and secondary stresses with large elastic follow-up”, Int. J. Pres. Ves.
Piping, 111-112, 217-231.
Taggart, J. P. and Budden, P. J. (2008). “Leak before break: studies in support of new R6 guidance on
leak rate evaluation”, J. Pres. Ves. Tech., 130(1), 011402.
Teng. H., Bate. S. K. and Beardsmore. D. W. (2008), “Statistical analysis of residual stress profiles using
a heuristic method”, Proc. ASME PVP Conference, Paper PVP2008-61378, July, Prague.
Teng, H. and Bate, S. (2014). “Residual stress profiles of pipe girth weld using a stress decomposition
method”, Proc. ASME PVP Conference, Paper PVP2014-28247, July, Anaheim.
Teng, H. and Sharples, J. K. (2012). “Further predictions of fracture experiments of a pre-cracked
pressurised thermal shock disk specimen under warm prestressing loading cycles”, Proc. ASME
PVP Conference, Paper PVP2012-78077, July, Toronto.