Accepted Manuscript

Effects of embedded cracks and residual stresses on the integrity

of a reactor pressure vessel

Guian Qian, V.F. González-Albuixech, Markus Niffenegger

PII: S1350-6307(18)30073-6
DOI: doi:10.1016/j.engfailanal.2018.04.009
Reference: EFA 3437
To appear in: Engineering Failure Analysis
Received date: 15 January 2018
Revised date: 28 February 2018
Accepted date: 5 April 2018

Please cite this article as: Guian Qian, V.F. González-Albuixech, Markus Niffenegger ,
Effects of embedded cracks and residual stresses on the integrity of a reactor pressure
vessel. The address for the corresponding author was captured as affiliation for all authors.
Please check if appropriate. Efa(2018), doi:10.1016/j.engfailanal.2018.04.009

This is a PDF file of an unedited manuscript that has been accepted for publication. As
a service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.
 ACCEPTED MANUSCRIPT

Effects of embedded cracks and residual stresses on the integrity of a reactor

pressure vessel

T
Guian Qian1,*, V.F. González-Albuixech2, Markus Niffenegger1

IP
CR
1
Paul Scherrer Institute, Nuclear Energy and Safety Department, Laboratory for Nuclear Materials, Villigen

PSI, Switzerland

US
2
CIIM Universitat Politècnica de València. Camino de Vera, s/n 46022 Valencia, Spain
AN
M

*Corresponding author
ED

Guian Qian

Tel. +41 56 3102865

PT

Fax. +41 56 3102199

CE

E-mail address: [email protected]

Paul Scherrer Institute

AC

Nuclear Energy and Safety Research Department, Laboratory for Nuclear Materials, OHSA/06

5232 Villigen PSI, Switzerland

 ACCEPTED MANUSCRIPT

T
IP
CR
US
AN
M
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

Abstract

One potential challenge to the integrity of the reactor pressure vessel (RPV) of a pressurized water reactor is

posed by a pressurized thermal shock (PTS), which is associated with rapid cooling of sections of the hot and

still pressurized RPV by injection of relatively cold emergency coolant. PTS transients lead to high tensile

circumferential and axial stresses in the RPV wall. If the stress intensity factor (SIF) is large enough, a critical

T
crack may grow. Thus, the RPV has to be assessed against cleavage fracture.

IP
In this paper, two kinds of embedded cracks, i.e. semi-elliptical and elliptical crack with depth of 17 mm and

CR
length of 102 mm are considered. The extended finite element method (XFEM) is used to model such

postulated cracks. The embedded crack with tip in the cladding/base interface causes a high KI. This is due to

US
the stress discontinuities at the interface between the materials. In the FAVOR (probabilistic fracture mechanics
AN
code) calculation, for such cracks the point closest to the inner surface is considered in order to be conservative.

However, due to the highly ductile cladding material, it is unlikely that the embedded crack will propagate
M

through the cladding. Thus, it is more appropriate to consider the outer surface point of the crack front.
ED

The effect of welding residual stress and cladding/base interface residual stress on the crack driving force is

studied. Surface cracks are assumed in the study of residual stresses. Results show that considering realistic
PT

welding residual stresses may increase KI by about 5 MPa·m0.5, while the cladding/base interface residual stress
CE

has a negligible effect on KI. The reason is that the cladding residual stress is only localized to the interface and

it decreases significantly through the vessel wall. Considering welding residual stress increases the Weibull
AC

stress and fracture probability of the RPV.

Keywords: reactor pressure vessel, pressurized thermal shock, welding residual stress, embedded crack
 ACCEPTED MANUSCRIPT

Nomenclature

a crack depth, mm

KI Mode I linear elastic stress intensity factor, MPa·m0.5

KIc material fracture toughness, MPa·m0.5

m, 0 model parameters known as Weibull modulus

T
Mm free-surface correction for membrane stress

IP
Mb free-surface correction for bending stress

CR
P fracture probability

US
Q crack shape correction factor for stress intensity factor calculation

t vessel wall thickness, mm

AN
T temperature, °C

V0 elementary volume representing the mean volume occupied by each

M

micro-crack in a solid, mm3

ED

Vpl volume of the plastic deformation zone as the cleavage fracture process

zone, mm3
PT

x distance from the inner surface of the vessel, mm

CE

σR residual stress, MPa

σm membrane stress, MPa

AC

σb bending stress, MPa

σw Weibull stress, MPa

σth threshold stress, MPa

1 maximum tensile principal stress, MPa

CRS cladding residual stress

 ACCEPTED MANUSCRIPT

FEM finite element method

MBLOCA medium break loss-of-coolant accident

PTS pressurized thermal shock

RPV reactor pressure vessel

SIF stress intensity factor

T
SBLOCA small break loss-of-coolant accident

IP
WRS welding residual stress

CR
XFEM extended finite element method

US
AN
M
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

1. Introduction

The integrity of the reactor pressure vessel (RPV) of a nuclear power plant has to be assured throughout the

whole lifetime of the plant, and the proof of the exclusion of brittle failure of the RPV for all possible load cases

is an important task in the safety assessment. During operation, certain abnormal conditions could result in the

so called pressurized thermal shock (PTS), i.e. rapid cooling of sections of the hot and still pressurized RPV by

T
injection of relatively cold emergency coolant. Brittle failure under PTS conditions and increasing neutron

IP
embrittlement during operation is generally considered to be the major threat to RPV integrity. Thus, the PTS

CR
analysis has to be performed according to the material property obtained from the surveillance program.

During the last two decades, substantial progress has been achieved in the safety assessment of RPVs under

US
PTS loadings [1-13]. Integrity analysis of a model RPV subjected to a small break loss-of-coolant accident
AN
(SBLOCA) and a medium break loss-of-coolant accident (MBLOCA) is performed by assuming crack depths of

two times the nondestructive examination limit [1-2]. The constraint effect of crack tip on the fracture toughness
M

is also analyzed in [1, 2]. Elastic-plastic analysis and cleavage fracture analysis have been performed in RPV
ED

materials considering different loading transients [3-7]. Chou et al. [8] performed a probabilistic analysis for a

RPV considering the uncertainty of crack distributions and fracture toughness according to the US Nuclear
PT

Regulatory Commission [9]. Testing of fracture toughness in the ductile-to-brittle transition temperature is
CE

documented in [10]. The solutions of crack driving force and modeling of fracture toughness are presented [11,

12]. The general procedure for structural integrity assessment of a RPV subjected to a PTS is briefly introduced.
AC

In the structural integrity analysis, firstly the temperature distribution is calculated through the RPV wall for a

certain transient. The input parameters, which consist of water temperature, pressure and heat transfer

coefficient between water and the inner wall of the RPV, stem from thermal hydraulic calculation (e.g. RELAP).

Based on the temperature field, the circumferential and axial stress histories for the whole transient are

calculated. Finally, cracks with different depths and orientations are assumed in the beltline region and the
 ACCEPTED MANUSCRIPT

calculation of SIFs is performed. Different models, such as the ASME model, the Master Curve and the FAVOR

model [13], are used to characterize fracture toughness of the material at different temperatures by using the

material chemical composition and the neutron fluence. In the integrity analysis of the RPV, both deterministic

and PFM analyses are performed. In fact, the probabilistic analysis is always based on a deterministic analysis.

In the deterministic method the maximum allowed RTNDT to exclude crack initiation is determined.

T
It is widely studied that postulated embedded cracks close to the inner-surface of an RPV contribute

IP
significantly to the conditional probability of cleavage fracture of the RPV during the PTS events. Therefore, it

CR
is necessary that an accurate KI computation for embedded cracks is available. In the FAVOR code [13]

developed by ORNL, the crack tip close to the inner surface of the vessel wall is considered in order to be

US
conservative. However, the propagation of the crack tip close to the inner surface is questionable due to the
AN
ductile material of the cladding. Thus, KI calculation of the embedded crack is presented and discussed using

FAVOR and extended finite element methods (XFEM).

M

In this paper, the effect of welding residual stress and cladding/base interface residual stress on the crack driving
ED

force is studied. The welding process to connect different RPV rings together creates the welding residual stress.

Finite element analysis was applied to compare the crack driving force obtained using residual stress
PT

distribution formulas and the stress free temperature model.

CE

The paper is organized as follows: The first part is introduction. The second part presents the fracture mechanics

analysis for embedded cracks. Then follow the effects of residual stress and cladding on K I. The last part is
AC

conclusions. In the modeling of embedded cracks in Figs.1, 2,4 (partly),5,7,8, XFEM is used to calculate KI. In

Figs. 6 ,4 (partly), 12, FAVOR is used to calculate KI. In the analysis of residual stress in Figs. 10 and 11, 3D

FEM is used to analyze KI.

2. Fracture Mechanics analysis for embedded cracks

 ACCEPTED MANUSCRIPT

In FAVOR code, the stress intensity factor (SIF) of an embedded crack is calculated by the EPRI model in the

ASME procedure [13]. It is a weight function method based on the resolution of nonlinear applied stresses

through the RPV wall thickness into the linear superposition of approximate membrane and bending stress

components. It is expressed as

KI   a  M m m  M b b 
Q 

T
. (1)

IP
In order to simplify the modelling, SIF can also be calculated by the interaction integral using XFEM, which

CR
enriches the finite element approach space with special functions that are able to describe the discontinuity and

introduce the singular behaviour associated with the crack front, and makes its analysis, up to a certain point,

US
independent of the mesh [14]. The essential idea in XFEM is to use a displacement field approximation that can

model any crack face discontinuity and the near crack-tip asymptotic stress field. As a consequence it is not
AN
necessary to modify the mesh to consider a specific crack; at most, moderate refinement must be introduced
M

around the crack to achieve good accuracy. XFEM is implemented in Abaqus [15]. In this study, FAVOR and
ED

XFEM are used to calculate the SIFs based on the reference transient.

Two kinds of embedded cracks, i.e. semi-elliptical and elliptical crack with depth of 17 mm and length of 102
PT

mm are considered, as shown in Fig. 1. The depth of the crack is two times the nondestructive detection limit,
CE

according to the German standard KTA 3201.2. The two kinds of embedded cracks in Fig. 1 are totally different

crack types. The crack shown in Figure 1a is an underclad crack and the crack shown in Figure 1b is a
AC

subsurface (embedded) crack in base metal. The embedded crack in Fig.1b is considered in FAVOR but the

underclad crack isn’t considered in FAVOR. However, according to the report [16], the underclad crack is found

in European reactors. In order to study the integrity of the underclad on the RPV, an underclad crack is

postulated in this study. The vessel has the inner radius of 1668.5 mm and thickness of 171 mm including a 5

mm cladding. XFEM is used to model such cracks. The assembling of the RPV and the embedded crack and the

mesh are shown in Fig. 2. The submodeling technique is used, as shown in Fig. 2. SIFs for crack front point A
 ACCEPTED MANUSCRIPT

and B are calculated. The postulated transients considered in this paper, as shown in Fig.3, are MBLOCA and

SBLOCA described in [1, 2]. RPV material properties are described in [1, 2] and presented in Table 1. The

results from XFEM are also compared with those from FAVOR for point A, as shown in Fig. 4. A good

agreement is obtained, which validated the results calculated by XFEM. In Fig. 5, it is shown that the embedded

crack with tip in the cladding/base interface (point A) causes a high KI. KI for point A is between 50%-100%

T
higher than that for point B, which is due to the stress discontinuities at the interface between the materials. This

IP
confirms the conservative assumption in the FAVOR code. In the FAVOR calculation, for such cracks the point

CR
closest to the inner surface is considered in order to be conservative. The SIFs are compared with fracture

toughness of the base material and cladding of the RPV. SIFs for both surface and embedded cracks are

US
calculated. The surface crack and the depth of the crack are according to the German standard KTA 3201.2. K Ic
AN
database of cladding material (both irradiated and unirradiated) are collected from [17, 18]. It is shown in Fig. 6

that SIFs at the tip closest to the inner surface of the RPV is higher than the K Ic of the base material while lower
M

than that of cladding. This indicates that the embedded crack may initiate in the cladding direction because of a
ED

high SIF. However, due to the highly ductile cladding material, it is unlikely that the embedded crack will

propagate through the cladding. Thus, it is more appropriate to consider the outer surface point B of the crack
PT

front in the integrity assessment. In FAVOR, If the SIF of an embedded crack (completely contained in the base
CE

material) exceeds the fracture toughness of the base material, then the following is assumed: The crack initiates

in cleavage fracture in the base material; The embedded flaw instantaneously becomes an infinite surface crack,
AC

i. e., the cladding above the crack is assumed to have failed completely and the crack to have extended length-

wise, either in the axial or circumferential directions. From this study, a detailed crack propagation is needed.

It is noted that in the integrity assessment of core region of RPV, the important embedded cracks are cracks with

distances to the inner surface smaller than 3/8 of the wall thickness. In this study, we considered the most
 ACCEPTED MANUSCRIPT

critical case of the embedded cracks. The embedded crack located close to the inner surface of the RPV wall is

more critical due to the higher KI and lower KIc.

In a general 3D stress state, the critical crack orientation is not obvious as in axisymmetric cases where it is

sufficient to analyse circumferential and axial cracks lying perpendicular to the principal stress directions. We

therefore performed an additional analysis of a crack which is inclined by 45 degree to the axial direction of the

T
RPV. In addition a planar crack was postulated. With a planar crack we mean that the crack is parallel to the

IP
inner surface of the vessel, similar to the hydrogen flakes found in Belgian RPVs. Both surface and embedded

CR
cracks have been found in US research reactors, according to [16]. It is shown in Fig. 7 that the SIF for a 45

degree inclined crack is between that for axial and circumferential cracks. However, the SIF for the planar crack

US
is much lower and this crack is not likely to propagate, as shown in Fig. 8. It is noted that the negative K in Fig.
AN
8 may be due to crack closure in the starting stage of the transient.

3. Effects of residual stress on KI

M

The effect of welding residual stress and cladding/interface residual stress is studied. According to IAEA [19],
ED

the welding residual stress in the welds along the vessel wall is expressed as

 2 x 
PT

 R  56  cos   (MPa)
 t  . (2)
CE

where t is the RPV thickness without cladding, x is the distance starting from the inner surface of the vessel

wall.
AC

A similar formula is used in FAVOR [13] to account for the welding residual stress as

 2 x 
 R  44  cos   (MPa)
 t  . (3)

In this study, both equations are used to study the effect of welding residual stress. The prescribed residual

stresses are applied to the ABAQUS models with a user subroutine which assigned stress at material points

based on their radial and circumferential location. The stress was initialized with the model at room
 ACCEPTED MANUSCRIPT

temperature, and a subsequent analysis step allowed the stresses to self-equilibrate. A three-dimensional,

axisymmetric model is built with linear elastic, temperature dependent, material properties applied for the base

and cladding, as shown in Fig. 9. A semi-elliptical crack with depth of 17 mm is assumed and KI of the deepest

point in the crack front is calculated. As shown in Fig. 10, the welding residual stress results in a KI of about 5

MPa·m0.5 along the crack front for both axial and circumferential cracks.

T
Moreover, the modified Beremin model is used to scale fracture toughness in different specimens. As a local

IP
approach to fracture, Beremin model [20] is essentially a two-parameter Weibull distribution as below:

CR
 
P  1  exp  (  σ1m dV / V0 ) / σ 0m   1  exp   σW / σ 0   ,
m

 Vpl   

US
(4)

1/ m
 
σW    σ1m dV / V0 
AN
,
V 
with  pl  (5)
M

where P is the cumulative probability of fracture, Vpl denotes the volume of the plastic deformation zone as the

cleavage fracture process zone, m and 0 are the two model parameters known as Weibull modulus and the scale
ED

parameter, respectively, 1 is the maximum tensile principal stress, V0 is an elementary volume representing the
PT

mean volume occupied by each micro-crack in a solid, dV is the differential volume. In order to consider the
CE

plastic deformation effect, the cumulative failure probability formulation is modified to adopt a fixed-value

threshold stress (th) by Gao et al. [21],

AC

  σ  σ m 
P  1  exp    W th
 
  σ 0  σth  
. (6)

According to the European program [22], the calibrated parameters for this RPV material are m=6.36, σth=1546

MPa, σ0=2076 MPa, V0=0.001 mm3. The integration zone is volume where the von Mises stress is over two

times the yielding stress. It is shown in Fig. 11 (a) that the Weibull stresses σ w for the case with WRS is greater

than that for the case without WRS. σw captures the stress and strain distributions ahead of the crack tip at the
 ACCEPTED MANUSCRIPT

same J. The probability of cleavage initiation for the RPV with WRS is calculated. It is clear in Fig. 11 (b) that

the RPV with WRS has a higher failure probability due to the increasing of KI by WRS.

The welding process to build the cladding on the base material causes residual stresses, partly according to the

different thermal expansion of the two materials. In FAVOR, it is assumed that the residual stress at room

temperature is 146.9 MPa [13, 23]. The effect of welding residual stress and cladding residual stress is studied

T
with FAVOR and the result for KI is shown in Fig. 12 (a). It is shown that considering the cladding residual

IP
stress has a negligible effect on KI for a surface crack. The reason is that the cladding residual stress only is

CR
localized to the interface and it decreases significantly through the vessel wall. It is notice that this analysis is

based on the single crack calculation according to the German KTA rule. However, it has been shown [24] that a

US
very shallow surface which just penetrates the cladding into the base material is highly influenced by the
AN
cladding. This very shallow flaw can demonstrate the highest probability of cleavage initiation. Thus, in this

case, cladding residual stress plays a very important role on KI and the fracture probability.
M

4. Effects of cladding on KI
ED

In 3D FE calculations the question raised is whether the relatively thin cladding (austenitic stainless steel) has to

be modelled, which causes problems due to the large difference in size of the different parts in the model, or
PT

whether it could be neglected.

CE

The effect of cladding on KI is therefore studied in Fig. 12 (b). The comparison of KI and KIc shows an overlap

between KI and KIc, meaning that crack initiation may occur in the PTS transient. Considering the residual
AC

stress will increase KI and thus increase the crack initiation probability of the RPV. It is shown that neglecting

the cladding leads to an increase of peak KI of about 2 MPa·m0.5 and a temperature shift of about -15 ºC. We

thus conclude that neglecting the cladding is a conservative approximation based on analyses of the single

surface crack. It should be noted that a very shallow surface which just penetrates the cladding into the base

material is highly influenced by the cladding [23]. For this very shallow flaw neglecting the cladding is a non-
 ACCEPTED MANUSCRIPT

conservative approximation. However, our assumption of the crack is according to the German KTA rule, which

assumes the crack has a depth of 1/10 of the vessel wall thickness.

5. Conclusions

Based on this study, the following conclusions are drawn:

T
1. For the embedded crack, KI for point A is between 50%-100% higher than that for point B. However,

IP
crack propagation through the cladding material is not likely, which confirms the conservative

CR
assumption in the FAVOR code.

2. The SIF for a 45 degree inclined crack is between that for axial and circumferential cracks. However, the

US
SIF for the planar crack is much lower and this crack is not likely to propagate.
AN
3. Considering welding residual stresses increase KI by about 5 MPa·m0.5 (This is about 5% the peak KI for

the surface crack.) and the cladding residual stress has a negligible effect on KI. Neglecting the cladding
M

is a conservative approximation for the crack has a depth of 1/10 of the vessel wall thickness. It should be
ED

noted that a very shallow surface which just penetrates the cladding into the base material is highly

influenced by the cladding. For this very shallow flaw neglecting the cladding is a non-conservative
PT

approximation.
CE

4. Considering welding residual stress increases the Weibull stress and fracture probability of the RPV by

about 20% for KI=100 MPa m0.5.

AC

Acknowledgements

The authors are grateful for the financial support of the PROBAB Project provided by the Swiss Federal

Nuclear Safety Inspectorate (ENSI) (DIS-Vertrag Nr. H-100668). V.F. González-Albuixech is thankful for the

research program Juan de la Cierva Incorporación 2015, IJCI-2015-23245, financed by the Spanish Ministerio

de Economía, Industría y Competitividad.

 ACCEPTED MANUSCRIPT

References

1. Qian G, Gonzalez-Albuixech VF, Niffenegger M. In-plane and out-of-plane constraint effects under

pressurized thermal shocks. Int J Solids Struc 2014; 6:1311-21.

2. Qian G, Niffenegger, M. Integrity analysis of a reactor pressure vessel subjected to pressurized thermal

shocks by considering constraint effect. Eng Fract Mech 2013; (112-113): 14-25.

T
3. Qian XD, Jr. RH, Yin SJ, Bass R. Cleavage fracture modeling of pressure vessels under transient thermo-

IP
mechanical loading. Eng Fract Mech 2008; 75: 4167-89.

CR
4. Sun X, Chai G, Bao Y. Elastic and elastoplastic fracture analysis of a reactor pressure vessel under

US
pressurized thermal shock loading. Eur J Mech A‐Solid. 2017;66:69‐78.

5. Yu M, Luo Z, Chao YJ. Correlations between Charpy V-notch impact energy and fracture toughness of
AN
nuclear reactor pressure vessel (RPV) steels. Eng Fract Mech 2015; 147: 187-202

6. Hohe J, Brand M, Siegele D. Behaviour of sub‐clad and through‐clad cracks under consideration of the
M

residual stress field. Eng Fract Mech 2010; 77:217‐228.

ED

7. Keim E, Schmidt C, Schöpper A, Hertlein R. Life management of reactor pressure vessels under pressurized
PT

thermal shock loading: deterministic procedure and application to Western and Eastern type of reactors. Int J

Pres Ves Piping 2001; 78: 85-98.

CE

8. Chou HW, Huang CC. Effects of fracture toughness curves of ASME Section XI - Appendix G on a reactor
AC

pressure vessel under pressure-temperature limit operation. Nucl Eng Des 2014; 280: 404-412.

9. U.S. Nuclear Regulatory Commission, Technical Basis for Revision of the Pressurized Thermal Shock

(PTS) Screening Limit in the PTS Rule (10 CFR 50.61), NUREG-1806, Vol. 1, US NRC, Washington

(2007).

10. ASTM-E1921-02, 1997. Test method for determination of reference temperature, T0, for ferritic steels in the

transition range.
 ACCEPTED MANUSCRIPT

11. Wang X. Elastic T-stress solutions for semi-elliptical surface cracks in finite thickness plates. Eng Fract

Mech 2003; 70: 731-56.

12. Zhu XK, Joyce JA. Review of fracture toughness (G, K, J, CTOD, CTOA) testing and standardization. Eng

Fract Mech 2012; 85: 1-46.

13. Williams PT, Dickson TL, Yin S. Fracture Analysis of Vessels – Oak Ridge FAVOR, v12.1,Computer Code:

T
Theory and Implementation of Algorithms, Methods, and Correlations, ORNL/TM-2012/567, Oak Ridge

IP
National Laboratory, Oak Ridge, TN, November 2012.

CR
14. Moës, N., Dolbow, J., Belytschko, T. A finite element method for crack growth without remeshing. Int. J

Numer. Methods Engng 1999; 46: 131–150.

15. Hibbitt, Karlsson, Sorensen, 2013. Abaqus 6.13.3 Manual.

US
AN
16. EricksonKirk MT, Dickson TL. Recommended Screening Limits for Pressurized Thermal Shock (PTS),

2010,NRC, NUREG-1874.
M

17. Brumovsky M, Kytka M., Kopriva R. Cladding in RPV integrity and Lifetime Evaluation. Procedia
ED

Engineering, 2015;130: 1544-53.

18. Viehrig H.W., Altstadt E., Houska M. Radiation response of the overlay cladding from the
PT

decommissionedWWER-440 Greifswald Unit 4 reactor pressure vessel. Nucl Eng Des 2015; 286: 227-36.
CE

19. International atomic energy agency, 2010. Pressurised thermal shock in nuclear power plants: good practices

for assessment, IAEA, Austria, IAEA-TECDOC-1627.

AC

20. Beremin F, 1983. A Local Criterion for Cleavage Fracture of a Nuclear Pressure Vessel Steel. Metall Trans A

14A, 2277-87.

21. Gao, X., Zhang, G., and Srivatsan, T. S. Prediction of cleavage fracture in ferritic steels: A modified Weibull

stress model. Mater. Sci. Eng., A 2005;394:210-219.

22. Hümmer M., Keim E., Hofmann H. TIMES: An international project on transferability of fracture toughness
 ACCEPTED MANUSCRIPT

values for irradiated RPV steels. Proceedings of the PVP, July 22-26, 2007 ASME Pressure Vessels and

Piping Division Conference, San Antonio, Texas, USA

23. Kusnick J., Kirk M., Bass B., Williams P., Dickson T. Effect of cladding residual stress modeling technique

on shallow flaw stress intensity factor in a reactor pressure vessel. Proceedings of ASME PVP 2015.

24. Bass B., Dickson T, Williams P, Klasky H, Jr RH. The Effect of Shallow Inside-Surface-Breaking Flaws on

T
the Probability of Brittle Fracture of Reactors Subjected to Postulated and Actual Operational Cool- Down

IP
Transients: A Status Report, ORNL/TM-2015/59531/REV-01, Oak Ridge National Laboratory, February

CR
2016.

US
AN
M
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

Tab. 1 Thermo-mechanical properties of the base material and cladding of the RPV.

Base material Cladding

Temperature [°C] 0 20 100 200 300 400 0 20 100 200 300 400
Elastic modulus [103 MPa] 206 206 199 190 181 172 200 200 194 186 179 172
Mean linear thermal expansion 10.3 10.3 11.1 12.1 12.9 13.5 16 16 16 17 17 18
coefficient [10-6 °C-1]
Thermal conductivity [W/(m∙K)]

T
44.4 44.4 44.4 43.2 41.8 39.4 15 15 16 17 19 21
Specific heat capacity [J/(kg∙K)] 450 450 490 520 560 610 500 500 500 540 540 590

IP
Density [103 kg/m3] 7.6 7.6 7.6 7.6 7.6 7.6 7.6 7.6 7.6 7.6 7.6 7.6
Poisson’s ratio 0.3 0.3

CR
Yield stress of the unirradiated 449.3
material [MPa]
Stress free temperature [°C] 280.3

US
AN
M
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

T
IP
CR
Fig. 1a Half-elliptical (Semi-elliptical) crack postulated in the study, both point A and B are considered in the calculation.

US
AN
M
ED
PT
CE
AC

Fig. 1b Elliptical crack postulated in the study, both point A and B are considered in the calculation.
 ACCEPTED MANUSCRIPT

T
IP
CR
(a) (b)

US
AN
M
ED
PT

(c) (d)
CE
AC

(e)

Fig. 2a the whole RPV model, Fig. 2b the meshed RPV, Fig. 2c stress distribution of the RPV, Fig. 2d XFEM assembling of the
submodel RPV with an embedded crack, Fig. 2e XFEM modeling of an embedded crack in the submodeling.
 ACCEPTED MANUSCRIPT

20
300
Water temperature [C]

MBLOCA
MBLOCA
SBLOCA
15 SBLOCA
250

Pressure [MPa]

T
200
10

IP
150
5

CR
100

0
50

US
0 6000 12000 18000 24000 30000
0 6000 12000 18000 24000 30000
Time [second] Time [second]

(b)
(a)
AN
25
M
Heat transfer coefficient

20
ED

15 heat transfer coefficien, MBLOCA

[kW/m K]

heat transfer coefficien, SBLOCA

2

10
PT

5
CE

0 6000 12000 18000 24000 30000

Time [second]
AC

(c)

Fig. 3. MBLOCA and SBLOCA transients (a) water temperature history, (b) pressure history, (c) heat transfer coefficient
history.
 ACCEPTED MANUSCRIPT

150 Circumferential embedded elliptical cracks,

a=17 mm 2c/a=6, closest point A
125 MBLOCA, XFEM
MBLOCA, FAVOR
KI [MPam0.5]

100 SBLOCA, XFEM

T
SBLOCA, FAVOR

IP
75

CR
50

25

US
0 ti ti
AN
50 100 150 200 250 300
Crack tip temperature [C]
M

Fig. 4 KI for the embedded elliptical crack, MBLOCA and SBLOCA, point A is calculated by FAVOR and XFEM.
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

150
MBLOCA
Half-elliptical embebed crack, a=17 mm, 2c/a=6
125

T
100
KI [MPam0.5]

IP
75

CR
50

25 A (closest to inner surface)

B (closest to outer surface)

US
0 ti

0 50 100 150 200 250 300

Crack tip temperature [C]
AN
Fig. 5a KI for the half-elliptical crack, MBLOCA, both point A and B are calculated
M
ED

150 MBLOCA
Elliptical embebed crack, a=17 mm, 2c/a=6
125 A (closest to inner surface)
PT

B (closest to outer surface)

KI [MPam0.5]

100
CE

75

50
AC

25

0
0 50 100 150 200 250 300
Crack tip temperature [C]

Fig. 5b KI for the elliptical crack, MBLOCA, both point A and B are calculated.
 ACCEPTED MANUSCRIPT

KIc, base, ASME KI, 2c/a=6, axial cracks, elliptical, point A

RTNDT=93 [C]Embedded elliptical cracks, point A, a=17 mm
400 Cladding toughness 5 mm from inner surface
KI and KIc [MPam0.5]

Irradia., Viehrig, welding 7 mm from inner surface

Irradia., Viehrig, forged ring 9 mm from inner surface

T
Cladding toughness, Brumovsky, a=12 mm
5 mm from inner surface

IP
300 Irradia., WWER 440
Irradia., WWER 1000 a=10 mm
Unirradia.
5 mm from inner surface

CR
Surface crack
200
a=5 mm

100
US
AN
0 ti
M

-100 0 100 200 300 400 500 600

Crack tip temperature [C]
ED
PT

Fig. 6 Comparison of KI and KIc for both surface and embedded cracks subjected to MBLOCA, KIc for both base and cladding
are compared.
CE
AC
 ACCEPTED MANUSCRIPT

T
75
SBLOCA, CFD, mixing

IP
Half-elliptical surface crack, a=17 mm, 2c/a=6

CR
50
KI [MPam0.5]

US
25
Circumferential crack
AN
45 degrees crack
Axial crack ti
0
M

50 100 150 200 250

ED

Crack tip temperature [C]

PT

Fig. 7 KI for the circumferential, axial and 45 degree inclination cracks, SBLOCA
CE
AC
 ACCEPTED MANUSCRIPT

4 Embedded planar crack, a=17 mm, 2c/a=6,

5mm from the inner suface, at the interface
2 MBLOCA
SBLOCA
KI [MPam0.5]

T
-2

IP
-4

CR
-6
ti

US
-8
0 50 100 150 200 250 300
Crack tip temperature [C]
AN
Fig. 8 KI for the embedded planar crack, MBLOCA and SBLOCA.
M
ED
PT
CE
AC
 ACCEPTED MANUSCRIPT

T
IP
CR
(a) (b)

(b)

US
AN
M
ED
PT

(c) (d)

Fig. 9a Cylinder to introduce WRS in finite element simulation, Fig. 9b sub model of a circumferential crack in a cylinder, Fig.
CE

9c sub model of an axial crack in a cylinder, Fig. 9d crack tip mesh.

AC
 ACCEPTED MANUSCRIPT

10
MBLOCA
Semi-elliptical crack a=17 [mm], 2c/a=6, by FEM

Welding residual stress from FAVOR

Axial crack
KI [MPam0.5]
Circum. crack

T
5

IP
CR
US
0
0.0 0.5 1.0
Crack front angle [F/]
AN
Fig. 10a KI of surface cracks due to the welding residual stress documented in FAVOR, calculated by 3D FEM.
M

10
MBLOCA
Semi-elliptical crack a=17 [mm], 2c/a=6, by FEM
ED

Welding residual stress from IAEA recommendation

Axial crack
PT

Circum. crack
KI [MPam0.5]

5
CE
AC

0
0.0 0.5 1.0
Crack front angle [F/]

Fig. 10b KI of surface cracks due to the welding residual stress as documented in IAEA, calculated by 3D FEM.
 ACCEPTED MANUSCRIPT

2000
W [MPa]

T
IP
Considering WRS, axial crack, a=17 mm
Without considering WRS

CR
1000

0 50 100
KJ [MPam0.5] US 150 200
AN
Fig. 11a Comparison of Weibull stress of the RPV with considering WRS and without considering WRS.
M
ED

Considering WRS, axial crack, a=17 mm

1.0 Without considering WRS
PT

0.8

0.6
Pf
CE

WRS effect
0.4
AC

0.2

0.0
0 50 100 150
KJ [MPam0.5]

Fig. 11b Fracture probability of the RPV with considering WRS and without considering WRS .
 ACCEPTED MANUSCRIPT

KI, a=17 mm, 2c/a=6, axial surface crack, by FAVOR,

120
MBLOCA
CRS: cladding residual stress
100 WRS: welding residual stress

T
KI [MPam0.5]

80

IP
60

CR
40
WRS
CRS ti
20
With WRS and CRS

US
Without residual stresses
0
0 100 200 300

Crack tip temperature [C]

AN
Fig.12a Influence of cladding (CRS) and welding residual stress (WRS) on K I.
M

KIc, ASME
MBLOCA
RTNDT=93 [C]
ED
]

KI, 2c/a=6, a=17 mm

0.5

100
KI and KIc [MPam

PT

t=166 mm
50 Axial, without cladding
Circumf., without cladding
t=171 mm ti
CE

Circumferential, without cladding

0 Circumferential, with cladding
Axial, without cladding
Axial, with cladding
AC

100 200 300

Crack tip temperature [C]

Fig. 12b KI of surface cracks with and without considering cladding, calculated by FAVOR.
 ACCEPTED MANUSCRIPT

Highlights

1. It is unlikely for the embedded crack to propagate through the cladding

2. It is more appropriate to consider the outer surface point of the crack front
0.5
3. Considering welding residual stresses increases KI by about 5 MPa·m

4. Cladding/base interface residual stress has a negligible effect on KI

5. Considering welding residual stress increases Weibull stress and fracture probability

T
IP
CR
US
AN
M
ED
PT
CE
AC