View source for B17695, Forwards Response to NRC 990204 RAI Re LBB Evaluation of Pressurizer Surge Piping.Response Discusses Loads,Fatigue & Fatigue Crack Growth Analysis That Were Part of Util Original 981109 Submittal

{{Adams
| number = ML20207H553
| issue date = 03/05/1999
| title = Forwards Response to NRC 990204 RAI Re LBB Evaluation of Pressurizer Surge Piping.Response Discusses Loads,Fatigue & Fatigue Crack Growth Analysis That Were Part of Util Original 981109 Submittal
| author name = Necci R
| author affiliation = NORTHEAST NUCLEAR ENERGY CO.
| addressee name = 
| addressee affiliation = NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
| docket = 05000336
| license number = 
| contact person = 
| document report number = B17695, NUDOCS 9903160028
| document type = CORRESPONDENCE-LETTERS, INCOMING CORRESPONDENCE
| page count = 54
}}

=Text=
{{#Wiki_filter:e          Northeast                                                     Rope Ferry Rd. (Route 156), Waterford, Cr 96385 Nuclear Energy                                                eston, Noae., ro.., station Northeast Nudcar Energy Company P.O. Box 128 Waterford CT06385-0128 (860) 447 1791 Faz (860) 444 4277 The Northeast Utilities System MAR       5 1999 Docket No. 50-336 B17695 U.S. Nuclear Regulatory Commission Attention: Document Control Desk Washington, DC 20555.
Millstone Nuclear Power Station, Unit No. 2 Response to Second Request For Additional Information Concerning Leak Before Break Evaluation of the Pressurizer Surae Pioina                                             i i
The purpose of this letter is to provide the NRC Staff with additional information pertaining to the NRC review and approval of Leak Before Break (LBB) methodology for                                    i the Millstone Unit No. 2 Pressurizer Surge piping. In a {{letter dated|date=November 9, 1998|text=letter dated November 9,1998}},*
Northeast Nuclear Energy Company (NNECO) requested that the NRC Staff review and approve the plant-specific LBB evaluation of the Millstone Unit No. 2 Pressurizer Surge piping. In a {{letter dated|date=February 4, 1999|text=letter dated February 4,1999}},* the NRC requested that NNECO provide a response to eleven questions raised by the NRC Staff during their review of the November 9,1998 LBB submittal.
The eleven questions relate to the loads, fatigue and fatigue crack growth analyses that were pa.t of our original submittal of November 9,1998. NNECO's detailed response to each of these questions is provided in Attachment 1. As stated in this attachment, a                                f/
l complete ASME Section lit fatigue analysis was performed as part of the Combustion Engineering Owner's Group report CEN-387-P, Rev 1 in response to NRC Bulletin 88-11 which identified thermal stratification as a major contributor to fatigue. This report was accepted by the NRC Staff as fulfilling the requirements for fatigue usage for a 40                             @f/
year design life considering the phenomenon of thermal stratification and thermal striping. As part of our leak before break submittal, we performed an extensive fatigue crack growth analysis which considered the impact of reactor water environment. The
* Letter from Martin L. Bowling to U. S. Nuclear Regulatory Commission, ' Millstone Nuclear Power Station, Unit No. 2, Request For Permission to Apply Leak Before Break Methodology To The Pressurizer Surge Piping,' dated November 9,1998.
* Letter from Stephen Dembek to Martin L. Bowling, Jr., "Second Request For Additional Information Regarding Submittal Requesting Leak-Before-Break Approval- Millstone Nuclear Power Station, Unit No. 2 (TAC No. MA4126)," dated February 4,1999.
                                       '^
9903160028 990305 PDR    ADOCK 05000336 P                     PDR            g

l l
4 U.S. Nucirr Regulatory Commission B17695\Page 2 fatigue crack growth analyses demonstrated that an undetected flaw located at the most limiting location in the surge line with a hypothetical 6:1 aspect ratio (that is length: depth) and with an initial flaw depth of 11.1% (maximum depth allowed by ASME Section XI IWB-3500) would not grow through the pipe wall even after considering the        1 anticipated number of heatup and cooldown cycles through the end of the Unit's current license. To account for the reactor water environment, this evaluation applied a factor of 2 on the number of allowable cycles to the ASME Section XI crack growth curve for austenitic stainless steel (in air). For the carbon steel, the ASME Section XI fatigue crack growth curve in a reactor water environment was utilized. Thus, the effect of a reactor water environment has been adequately addressed in the flaw tolerance evaluations. Furthermore, considering that (a) Millstone Nuclear Station Unit Number 2 has only experienced approximately 56 heatup and cooldown cycles to date, (b) the results of previous in-Service-Inspections (ISI) did not identify . any unacceptable indications, and (c) the future ASME Section XI required inspections planned for the surge line would be expected to identify any potential flaws that may exist prior to these flaws _ reaching significant depth, NNECO concludes that fatigue concerns are adequately addressed such that it will not be a significant contributor to the probability of pipe rupture.
There are no regulatory commitments contained within this letter.                           l if you have any additional questions concerning this submittal, please contact Mr. Ravi G. Joshi at (860) 440-2080.                                                        ,
Very truly yours, NORTHEAST NUCLEAR ENERGY COMPANY Raymond P. Necci Vice President - Oversight and Regulatory Affairs Attachments cc:     H. J. Miller, Region 1 Administrator J. C. Linville, Chief, inspections Branch E. V. Imbro, Director, Millstone ICAVP inspections S. Dembek, NRC Project Manager, Millstone Unit No. 2                               ;
D. P. Beaulieu, Senior Resident inspector, Millstone Unit No. 2 j

o Docket No. 50-3M B17695 l
l l
l l
l Attachment 1 Millstone Nuclear Power Station, Unit No. 2 Response to Second Request For Additional information Concerning Leak Before Break Evaluation of the Pressurizer Surge Piping Response to Questions 1 Through 11 I
i i
i 1
March 1999 i
i

, 3 Table of Contents Section                                                                                                                                                         Page

==1.0 INTRODUCTION==
... .. .. .. .. .. .. .... .. .......... ...... .... ...... ............. ... . .... .... ..... ..... . . . ... .... .... . . . .. .. .. .. . . . . 1 - 1 2.0 RES PONSE TO QUESTIONS ............................ ...................................... ....................... 2- 1 2.1   Question # 1 - Environmental Effects .......................................................................... 2- 1 2.2   Question #2 - Turbulent Penetration ............................................ ............................. 2-3
       -- 2.3  Question #3 - Tme Stress Strain Curve ....................................................... .............. 2-8 2.4   Question #4 - S MiRT Paper .... .............. ................ ... ................................ ........... 2-9 2.5   Question #5 - Determination of Appendix B Moments ............................................ 2-10 2.6    Question #6 -Detailed Analysis for Throughwall Stratified Imading Stresses ........ 2-11 2.7   Question #7 - Maximum Stratified Flow Condition.................................................. 2-17 2.8   Question #8 - Model for Bimetallic Effects .............................................................. 2- 18 2.9   Question #9 - Polynomial Fit......... ........ ................................................................ 2-21 2.10 Question #10- Austenitic Versus Ferritic Bimetallic Weld...................................... 2-22 2.11 Question #11 - Bimetallic Stress Curve Fit........................................................... ... 2-23

==3.0 REFERENCES==
 .. ....... ... . .. .. ..... .. ................ .... ...... ............ ...... ...... ...... ... .. . .. . .. . . ......... ... . .... .. . . . . 3 - 1 APPENDIX A REPRINT FROM VOLUME L OF 1989 SMIRT PROCEEDINGS... ............ A-0 APPENDIX B REPRINT FROM JOURNAL OF PRESSURE VESSELTECHNOLOGY s                VOL. 108, AUGUST 198 6 ............................................................ ................ .... B -0 9
                                                                                .1 e

List of Tables Table                                                                                                                          Page Table 2.61 surge Line Pipe Wall Temperature Distribution for Pressurizer Outsurge
                      -(AT=200*F)........................................................................................................2-13
         ' Table 2.6-2 Surge Line Pipe Wall Local Stresses for Pressurizer Outsurge (AT=200*F).........................................................................................................2-15 i
i I
11                                                                          l 1
I L,                                                                                                                                              1

i i~   ,
List of Figures Page Figure 2.2-1. Schematic of Surge Line Showing Turbulence Penetration................................. 2-6 Figure 2.2-2. Hypothetical Surge Line Where Turbulence Penetration Could Have an Effect. 2-7
          - Figure 2.8-1. Bimetallic Joint Model and Stress Distribution .......................... ..................... 2-20 l
4 l
l i
iik

F l
l o
l               ,
i l

==1.0     INTRODUCTION==

The NRC has completed a review of the information provided in the Northeast Nuclear Energy Company (NNECO) November 9,1998 submittal requesting approval ofleak-before-break status for the Millstone Unit 2 surge line. Questions were submitted based on Structural Integrity Associates (SI) report SIR-98-096, an enclosure to the November 9,1998, submittal letter.
l This attachment provides a response to the second set of questions in the letter from Stephen Dernbeck (NRC) to Mr. Martin Bowling (NNECO), dated 2/4/99. The responses are based in large part from repon SIR-99-027, dated March 1,1999 by Structural Integrity Associates.
1-1

   .- e
        . 2.0      RESPONSE TO QUESTIONS 2.1-     Question #1-Environmental Effects General Design Criterion (GDC) 4 permits the exclusion of dynamic effects associated with postulated pipe ruptures in nuclear power units, from the design basis when analyses reviewed and approved by the Commission demonstrate that the probability of fluid system piping rupture is extremely low under conditions consistent with the design basis for the piping. In 52 FR 41291, the Commission indicated that the concept of Leak-Before-Break (LBB) is applicable
        ' only to high quality piping. The Commission further indicated that LBB is applicable to systems
                                                               ~
where failure mechanisms such as fatigue are not significant contributors'to the potential for pipe rupture. The staff has interpreted the Commission direction to mean that the piping has a low fatigue usage. Given that current fatigue data has shown that the pressurized water reactor (PWR) environment can have a significant impact on the fatigue life of stainless steel components
        - (Referencei NRC letter to NEl dated September 2,1998, providing Argonne National Laboratory assessment of fatigue data), provide an assessment of the influence of the PWR water environmental effects on the ASME Section III fatigue analysis of the surge line.
Response to Ouestion #1:
The fatigue analysis that forms the licensing basis for the Millstone Unit 2 surge line is based upon the work reported in the CEOG report on surge line stratification [1]. The NRC Safety Evaluation of that report concluded that "... analysis adequately demonstrates that the bounding surge line and nozzles meet ASME Code stress and fatigue requirements for the 40-year design life of CEOG facilities considering the phenomenon of thermal stratification and thermal striping."
The effects of the pressurized water reactor environment on surge line fatigue usage have been        !
recently evaluated by the Electric Power Research Institute, using data from Calven Cliffs, a plant very similar to Millstone Unit 2 [2]. These studies showed that there would be minimal effects due to the environment, with a resulting environmental multiplier for the controlling         l Calvert Cliffs unit'of about 1.4. This factor of 1.4 on cycles is well within the factor of safety of l 20 on cycles included in the ASME Section III fatigue (air) curves.
l 1
2-1 l
bv

i f
I It may be further noted that the CEOG surge line fatigue analysis (Report CEN-387-P, Rev.1)          !
considered 500 heatup and cooldown cycles. Considering that Millstone Unit 2 has only                 ,
t experienced approximately 56 heatup/cooldown cycles in its 20 years of operation, significant
{
fatigue margin is expected to remain through the end of design life of Millstone Unit Number 2.
I NNECO selected a very conservative approach with respect to the fatigue for the LBB evaluation of the pressurizer surge piping. In performing the fatigue crack growth evaluation, design basis loadings for the plant included both the system cycling as well as associated thermal i
stratification cycling. The crack growth analysis conservatively considered that maximum AT's         I occur at normal operating pressures. In actual practice, the maximum AT's occur at relatively lower pressures. Furthermore this analysis was conducted using fatigue crack growth curves recommended by ASME [3], with a factor of two applied to account for the reactor water                i environmental effects for austenitic stainless steel; for carbon steel, ASME XI water environment crack growth rates for ferritic steel were utilized. Using these crack growth predictions as a guideline, the planned ISI interval at Millstone Unit 2 is sufficient to ensure that ASME Section XI allowable flaw size will not be exceeded, even for very high hypothetical (20:1) aspect ratio flaws.
Thus, although the effects of reactor water environment on fatigue usage were not explicitly evaluated, the conservatisms enumerated above are adequately offsetting. Coupled with the results of the flaw tolerance evaluation, the results of previous In - Service- Inspections (ISI) which did not identify any unacceptable indications, and the ISI interval currently in place for the surge line adequately manages fatigue, such the.t it will not be a significant contributor to the probability of pipe rupture.
l l
f 1
l 2-2

a lt L
!       -2.2     Question #2 -Turbulent Penetration i
L                                 .
Providejustification why thermal cycling due to turbulent penetration will not affect the fatigue crack propagation calculations of the welds closest to the hot leg (welds 2 through 7 in Figure B-1.)
Response to Ouestion #2:
(In the work performed to evaluate the effects of surge line thermal stratification and reported in the CEOG report [1], data from a number of plants was evaluated. There was no identification of turbulent penetration as a loading mechanism, although it may have been a contributor to some of the observed stratification cycling. The observations from the testing were conservatively incorporated into the CEOG report loading definition for fatigue usage evaluation.
These test observations and the associated stratification loading spectra have been conservatively incorporated into the loading definition used for the crack growth analysis for the LBB evaluation.
Turbulence penetration was the subject of considerable research during the Electric Power Research Institute (EPRI) Thermal Stratification, Cycling and Striping (TASCS) program [4]. In this research, it was found that turbulent penetration is important for predicting cyclic stress
       ' behavior for nominally stagnant piping with inleakage of cold water toward reactor coolant system piping. Turbulence penetration was shown to be a contributor to the cracking observed in the Safety Injection piping at Farley and Tihange as described in USNRC Bulletin 88-08 [5]. In these cases, there was a shallow layer of cold water leaking toward the reactor coolant system (RCS) main loop piping. As the very low flow leakage approached the relatively warmer reactor coolant piping, turbulence from the RCS interacted with the shallow layer of water, such that the local stratification dissipated. Then, the leaking flow re-established stratification unti' it was wiped away by another turbulence penetration burst. This resulted in cyclic stresses that caused th'e observed cracking. In this case, the turbulent penetration could completely disrupt the flowing layer of cold water, causing it to mix with the hot water in the piping.
2-3

This phenomenon does not occur in the case of turbulent penetration into the surge line. Figure 2.2-1 shows the conditions expected in a surge line. The following describes the thermal-hydraulic phenomena involved:
: 1)          The more-or-less continuous flow of water from the pressurizer (normal spray flow) tends
               - to fill the top of the surge line with water slightly warmer than that in the hot leg. The density difference between the water in the pressurizer and the water in the hot leg is            i sufficient to cause the entire line to stratify. The geometry is such that the stratificatic. , is stable. There will be some axial gradients due to heat transfer between the two fluids and between the fluids and the pipe wall.
There is significant turbulence in the hot leg. Turbulence penetration into the region just        l 2).
above the hot leg nozzle tends to keep the bottom of the surge line filled with relatively colder water from the hot leg, since the distance to the horizontal piping is only about two diameters.
: 3)          The stratification layer of flow coming from the pressurizer that is near the hot leg nozzle is relatively deep, on the order of half the surge line diameter. Turbulence penetration into the' surge line does not have the strength to " wipe out" the relatively deep hot water layer in the top of the piping, 4);         Since this flow situation is relatively stable, no significant cycling is expected to' occur.      I i
On the other hand, if the surge line came into the bottom of the hot leg piping, then turbulence
    .. penetration might be'a contributor, In this case, the depth of the water approaching the hot leg
          ~
i nozzle would be relatively shallow, in this case, turbulence penetration could potentially disrupt             )
i I
2-4                                                     j

f!
1 I
i  ..
c the flow and cause stratification cycling. (See Figure 2.2-2). In this case, the geometry and stratification cycling would be much more similar to that reported for the cracking in NRC
                                                                                                        )
Bulletin 88-08. Fortunately, this is not the case for the Millstone Unit 2 surge line.
These observations are supported by review of actual plant data in the CEOG surge line report.
The stratification is observed to be relatively stable, and does not change much except for conditions with significant insurge or outsurge. Thus, any additional stresses due to turbulent penetration are judged to be insignificant as compared to the stresses caused by the overall stratification cycling that has been evaluated.
2-5

Pressurizer Relatively Warmer Water from Pressurizer
                                                  ~
b_______'____t_-~--~~~~-+------------4--_
                                                                            -------y Turbulence      g             i from Hot Leg    g Relatively Cold Water in Bottom of Pipe Hot Leg                                                                l 1
l 99019r0 Figure 2.2-1. Schematic of Surge Line Showing Turbu!ence Penetration i
2-6 l
f-

o Pressurizer Hot Leg Relatively WarmerWater
                  \                                    from Pressurizer Turbulence Penetration               #['"'''''
g
                                              '----------.____,,,,,,,__ ----~~ ,,       &
Relatively Cold Water peo2oro -
in Bottom of Pipe Figure 2.2-2.11ypothetical Surge Line Where Turbulence Penetration Could llave an Effect 2-7 L:

2.3      Question #3 -True Stress Strain Curve                                                        i Providejustification for basing Equation 20 in Appendix A on tme stress-true strain values of the material. State, or provide a reference, that the stresses and strains in the equations on which the elastic-plastic fracture mechanics analysis in Section 6 and Appendix D is based are true stress-true strain quantities, and the analysis uses a true stress-tme strain curve.                  3 Response to Ouestion #3:
i l
In the fracture mechanics models presented in Appendix D of SIR-98-096, tme stress and true strain values are used as stated in the ASME section XI technical basis document for flaw evaluation of austenitic stainless steel pipes (Reference 3, page 360, Eq.13, attached as Appendix B). In addition, the difference between the engineering stress / strain and true stress / strain is not very significant at strain values normally encountered in engineering analysis (less than 5%) considering the fact that they are related by the following relationship:
a = S (1+e) c = In (1+c) where:      o=       true stress                                                             j c = true strain S =      engineering stress e = engineering strain l
4 I
2-8 i
* i

f
    +
2.4     Question #4-SMIRT Paper In Appendix A, Equations 19 and 20 are taken from the paper (shown as Reference 19) by Cofie, N. G., et, al.;'" Stress-Strain Parameters in Elastic-Plastic Fracture Mechanics", presented at the ,
10th SMIRT International Conference, August 14-18,1989. This paper was not published in the
      ~ Proceedings of the Conference. Please provide a copy of this paper. ~                                {
I lAtf use to Ouestion #4:
The referenced paper was published in the 1989 SMiRT Proceedings in Volume L, page 91. A copy is provided in Appendix A.
                                                                                                             \
                                                                                                             )
I             2-9                          >

i
                                                                                                          )
l 2.5      Question #5 - Determination of Appendix B Moments The thermalltratification moments :ced in the fatigue calculations described in CEN-387-P, Rev.1-P-A, were based on maximum stratified flow conditions. This was accepted by the staff in its safety evaluation report dated July 14,1993, and included in the report. In Appendix B, the thermal stratification moments are based on a linear top-to-bottom temperature distribution.
Providejustification why these moments were determined on this basis, instead of the maximum         )
I stratified flow condition, as stated in CEN-387-P, Rev.1-P-A.
    ' Response to Question #5:
Appendix B of Report SIR-98-096 (Enclosed with NNECO's November 9,1988 Submittal),                   I provides the results of stress analysis to define unit loads due to dead weight, surge line thermal expansion, RCS piping thermal expansion and global stratification. The analysis was conducted with the ANSYS computer program, using pipe (beam) elements. This model pmvides the                  I overall distribution of moments in the piping system due to the specified loadings.
When considering stratification loading, the resulting piping stresses may be broken into those due to global effects (resulting from the overall piping system deformation), and those due to       ,
local effects (local stresses at the location where stratification exists). The results presented in Appendix B are for the global effect, and are based on a linear AT of 100*F, arbitrarily chosen as a " unit" load in the piping structural model. As explained in Appendix D of the report, these results were scaled to maximum stratified flow conditions using an appropriate scaling multiplier so that the approach accepted by the NRC could be duplicated. The local effects due to stratification flow were added separately. (See further discussion in response to Question 7 that follows.)
l l
l l
i 2-10

7 2.6      Question #6 - Detailed Analyris for Throughwall Stratified Loading Stresses Provide the [etailed analysis of the maximum throughwall thermal stresses, showing all relevant input parameters for both the temperature and the stress distributions, for the thermal stratification loading condition.
Response to Ouestion #6:
This analysis was conducted using Stmetural Integrity Associates program called TOPBOT.
This program has been verified under the requirements of SI's Quality Assurance Program. The program is set up to analyze an infinitely long stratified section of pipe, and requires the following input:
      -     Pipe thermal conductivity, specific heat, density, modulus of elasticity, Poisson's ratio and coefficient of thermal expansion
      -     Heat transfer coefficients for the outside of the pipe, inside of the pipe (for the hot fluid) and I
inside of the pipe (for the cold fluid).
      -     Height of the hot / cold level interface                                                           ;
i The temperature solution is determined based upon classical finite difference techniques; the            )
stress solution is based on axial force compatibility for a pipe end condition which is free to expand axially but not allowed to rotate (due to the stratification) and with the additional constraint stresses due to through-wall thermal gradients.
l 2-11 l

r.
     ,                                                                                                         i The program output is temperature distribution in the piping system, stresses for the fixed-end       I condition, moments generated by the stratified condition, and stress distribution for a free-free I
constraint condition at the ends of the pipe.
For the evaluations mported in SIR-98-096, the model was modified to allow the bottom node to be held at the temperature of the hot leg (for pressurizer insurge) or the top node to be held at the temperature of the pressurizer (for outsurge) as described in the CEOG report. These conditions were reponed to provide the best fit of the observed data for the fully: stratified conditions. The .
analysis used a modulus of elasticity of 25,800 ksi and a coefficient of thermal expansion of 4
        ' 10.75 x 10 in/in *F. Heat transfer coefficients from the CEOG repon were used. The pipe model used in this evaluation was broken up into 8 thrcugh-wall elements and 73 circumferential elements This analysis is described in Section D.3 of Appendix D.
The following tables provides detailed output for the case identified in Figures D-1 and D-2 of SIR-98-096. The temperature distribution is shown in Table 2.6-1 for the top of the pipe held at 653'F and the bottom of the pipe held at 453'F for a unit load case with AT = 200*F. Table 2.6-2 shows the associated local stresses.
       . Thus, the analysis has implemented the same loading con'ditions that were utilized in the CEOG report, such that no further evaluation is required.
e i
                                           -1
* I
                                                         '2-12 I

1
                 -                                                                             I l
Table 2.6-1                                            1 Surge Line Pipe Wall Ternperature Distribution for Pressurizer Outsurge (AT = 200*F) i
                                                    *    * *E**'"'
Angle from Top of Pipe,   Fluid                Radial Location in Pipe Wall, inches degrees    Temp., 'F 5.145   5.309   5.473   5.637   5.801  5.965     6.129 6.293 0.0        653     653     653     653     653     653     653      653   653 2.5        653     651     651     651     651     651     651      651   651 5.0        653     649     649     649     649     649     649      649   649 7.5         653     647     647     647     647     647     647      647   647 10.0        653     644     644     644     644     644     644      644   644 12.5        653     642     642     642     642     642     642      642   642      )
     -     15.0        653     640     640     640     640      640    640      640   640      l 173         653     638     638     638     638      638    638      638   638 20.0        653     636     636     636     635      635    635      635   635 22.5        653     633     633     633     633      633    6h.s     633   633 25.0        653     631     631     631     631      631    631      631   631 273         653     629     629     629     628      628    628      628   628 30.0        653     626     626     626     626      626    626      626   626 323         653     624     624     624     624      624    623      623   623 35.0        653     622     621     621     621      621    621      621   621 373         653     619     619     619     619      618    618      618   618 40.0        653     616     616     616     616      616    616      616   616 423         653     614     614     613     613      613    613      613   613 45.0        653     611     611     611     611      610    610      610   610 47.5        653     608     608     608     608      608    607      607   607 50.0        653     605     605     605     605      605    605      604   604      I 52.5        653     603     602     602     602      602    602      601   601      l 55.0        653     599     599     599     599      599    598      598   598      l 57.5        653     596     596     596     596      595    595      $95   595      l 60.0        653     593     593     593     592      592    592      592   592 62.5        653     590     589     589     589      589    589      589   589 65.0        653     586     586     586     585      585    585      585   5 85_ __
67 3        653     583     582     582     582      582    582      581   581 70.0        653     579     579     578     578      578    578      578   578 723         653     575     575     575     574      574    574      574   574 75.0        653     571     571     571     570      570     570     570   570 77.5        653     567     567     567     566      566     566     566   566 80.0        653     563     563     562     562      562     562     562   561 82.5        653     559     558     558     558      557     557     557   557      j 85.0        653     554     554     553     553      553     553     553   553 87.5        653     549     549     548     548      548     548     548   548 90.0        553     543     543     543     543      543     544     544   544 923         453     537     538     538     539      539     539     539   539 95.0        453     532     533     534     534      534     534     535   535 97.5        453     528     529     529     530      530     530     530   530 100.0       453     524     524     525     525      526     526     526   526 102.5       453     520     521     521     522      522     522     522   522      1 105.0       453     516     517     517     518      518     518     519   519 2-13                             3

Table 2.6-1 (concluded)
Surge Line Pipe Wall Temperature Distribution for Pressurizer Outsurge (AT = 200*F) g                                     WallTemperature *F Top of Pipe,    Fluid                Radial Location in Pipe Wall, inches degrees     Temp., *F 5.145  5.309    5.473  5.637    5.801  5.965     6.129 6.293 110.0         453     510    $10      511     511     512     512      512   512 112.5         453     507    507      508     508     508     509      509   509 115.0         453     504    504      505     505     5%      506      506   506 117.5         453     501    502      502     503     503     503      503   503    l 120.0         453     499    499      500     500     500     501      501   501
             -122.5         453     497    497-     497     498     498     498      498   498    ,
125.0         453     494    495      495     4%      4%      496      496   496
                                                                                                  ]
127.5         453     492     '93     493     494     494     494      494   494 130.0         453     491    4n       491     492     492     492      492   492 132.5         453     489    489      490     490     490     490      490   490 135.0         453     487    488      488     488     488     489      489   489 137.5         453     486    486      486     487     487     487      487   487 140.0         453     484    485      485     485     485     486      486   486 1423          453     483    483      484     484     484     484      484   484 145.0         453     482    482      482     483     483     483      483   483 1473          453     481    481      481     482     482     482      482   482 150.0         453     480    480      480     481     481     481      481   481 1523          453     479    479      479     480     480     480      480   480 155.0         453     478    478      479     479     479     479      479   479 157.5         453     477    478      478     478     478     478      478   478 160.0         453     477    477      477     477     478     478      478   478 162.5         453     476    476      477     477     477     477      477   477 165.0         453     476    476      476     476     476     477      477   477 1673          453     475    476      476     476     476     476      476   476 170.0         453     475    475      475     476     476     476      476   476    l l
172.5         453     475    475      475     475     475     476      476   476 175.0         453     475    475      475     475     475     475      475   475 177 5         453     474    475      475     475     475     475      475   475    l 180.0         453     474    475      475     475     475     475      475   475    )
l l
i l
2-14                             ,

E                                                                                                                   1 I
l Table 2.6-2 Surge Line Pipe Wall Local Stresses for Pressunr.er Outsurge (AT = 200*F)
Angle from                                      Axial Stress in Pipe Wall. ksi Top of Pipe,                               Radial Location in Pipe Will, inches degrees      5.063     5.145   5309       5.473        5.637       5.301     5.965    6.129  6.293   6375 0.0   -7.02     -6.70   -6.05      -5.40        -4.75       -4 10      3.45    -2.80  -2.16   -1.83 25    -6.48     -6.16   -5.50      -4.85        -4.20       -3.55     -2.90    -2.25  -1.60   -1.28 5.0    5.98      5.65   -5.00      -435         -3.69        3.04     -239     -1.74  -1.10   -0.77 7.5   -5.52     -5.19   -4.53      -3.88        -3.22       -2.57     -1.93    -1.28  -0.63   -031 10.0   -5.09     -4.76   -4.10      -3.45        -2.80        2.15     -1.50    -0.86  -0.22    0.10
  ^
12.5   -4.69      436    -3.71      -3.05        -2.40       -1.76     -1.12    -0.48   0.16    0.48 15.0   -433      -4.00     334      -2.69        -2.05       -1.41     -0.77    -0.14   0.49    0.81  I 17.5   -4.00     -3.67   -3.02      -237         -1.73       -1.09     -0.46     0.16   0.79    1.10  l 20.0   -3.69     -337      2.72     -2.08        -1.44        0.81     -0.19     0.43    1.04   135 22.5    3.42     -3.09   -2.45        1.82       -1.19       -0.57      0.05     0.66    1.26   1.56 25.0    3.17     -2.85   -2.21      -1.59        -0.97       -0.35      0.25     0.85    1.44   1.74 27.5   -2.95     -2.63   -2.00      -138         -0.77       -0.17      0.43      1.01   1.60   1.88  l 30.0   -2.75     -2.43   -1.81      -1.20        -0.60       -0.01      0.57     1.15    1.72   2.00  I 32.5   -2.57     -2.26   -1.65      -1.05        -0.46        0.12      0.69     1.25    1.81   2.08 l
35.0   -2.41      2.10   -1.50      -0.91        -034         0.23      0.78     133     1.87   2.14  )
37.5   -2.26     -1.96     137      -0.80        -0.24        032       0.86      1.39   1.91   2.17  !
40.0   -2,14     -1.84   -1.26      -0.70        -0.15        038       0.91     1.42    1.93   2.18 42.5    2.02     -1.73   -1.17      -0.62        -0.09        0.43      0.94      1.44   1.93   2.17 45.0   -1.92       1.63    1.08     -0.55        -0.03        0.47      0.96      1.44   1.91   2.14 47.5   -1.82       1.54  -1.01      -0.49         0.01        0.50      0.97      1.43   1.88   2.10 50.0   -1.73     -1.46   -0.94      -0.44         0.04        0.51      0.97      1.41   1.84   2.05 52.5   -1.64       138    -0.88     -039          0.07        0.52      0.96      138    1.79   1.98 l
55.0   -1.56     -130    -0.81      -035          0.10        0.53      0.94      134    1.73   1.92 57.5   -1.47     -1.22    -0.75     -030          0.13        0.54      0.93      131    1.67   1.84 60.0   -137      -1.14    -0.68     -0.25         0.16        0.55      0.92      1.27   1.61   1.77 62.5   -1.27     -1.04   -0.61      -0.19         0.19        0.56      0.91      1.24   1.56   1.71 65.0   -1.16     -0.94    -0.52     -0.13         0.24        0.58      0.91      1.22   1.51   1.65 67.5   -1.03     -0.82    -0.42     -0.05         030         0.62      0.92      1.21   1.47    1.60 70.0   -0.89     -0.68    -030       0.05         037         0.67      0.95      1.21   1.45    1.56 72.5   -0.72     -0.53    -0.17      0.16         0.47        0.74      0.99      1.23   1.44    1.54 75.0   -0.53     -034      0.00      031          0.58        0.83       1.06     1.27   1.45    1.54 77.5   -031      -0.13     0.19      0.48         0.73        0.95       1.15     133    1.48    1.56 80.0   -0.05      0.13     0.42      0.68         0.90         1.09      1.26     1.41   1.54'   l.60 82.5    0.27      0.43     0.70      0.92          1.11        1.27      1.40     1.52   1.61    1.M 85.0    0.65      0.81   .1.04        ' 22         136         1.47      1.56     1.64   1.70    1.73 87.5 1.15      130     1.46       1b           1.64        1.69      1.73     1.77   1.80    1.81 90.0     1.98     2.00     1.99       1.96         1.93        1.91      1.89     1.88   1.88    1.88 l             92.5    3.03      2.80     2.53      235           2.22        2.12      2.04     1.99   1.95    1.93 l             95.0     3.52     3.29     2.94      2.68          2.47        231       2.18    2.08    2.01    1.98 2-15                            >

y i
Table 2.6-2 (concluded)
Surge Line Pipe Wall Local Stresses for Pressurizer Outsurge (AT = 200*F)
                                                                                                                 )
Angle from                                     Axial Stress in Pipe Wall, ksi Top of Pipe,                              Radial Location in Pipe Wall, inches degrees    5.063     5.145   5309       5.473        5.637      5.801      5.965    6.129  6.293  6375 97.5  3.85       3.62    3.23       2.92         2.67       2.46       2.29     2.15   2.04   2.00 100.0  4.01       3.84    3.43       3.09         2.80       2.56       235      2.18   2.04   1.99 102.5  4.22       3.98    3.56       33           2.88       2.61       237      2.17   2.00   1.93  I 105.0  431        4.06    3.63       3.24         2.90       2.61       235      2.12   1.92   1.84  ]
2.56       2.18     2.02   1.80 107.5  4.34       4.09    3.64       3.24         2.88                                         1.70
                                                                                                                 )
110.0  433        4.08    3.62       3.20         2.82       2.47       2.16     1.89   1.64   1.53   I 112.5  4.29       4.03    3.55       3.11         2.71       235        2.02     1.72   1.44   132 115.0  4.20       3.94    3.45       2.99         2.58       2.19       1.84     1.51   1.22   1.08 117.5  4.09       3.82    331        2.85         2.41       2.01       1.63     1.28   0.96   0.81 120.0  3.96       3.68    3.16       2.67         2.22       1.80       1.40     1.03   0.68   0.52 122.5  3.80       3.52    2.98       2.48         2.01       1.57       1.15     0.76   039    0.21 125.0  3.63       3.34    2.79       2.27         1.78       132        0.88     0.47   0.07 -0.11    l l
127.5  3.44       3.14    2.58       2 04         1.54       1.06       0.60     0.17 -0.25  -0.45 130.0  3.24       2.94    2.36       1.81         1.28       0.79       031    -0.15  -0.58  -0.79 132.5  3.03       2.72    2.13       1.56         1.02       0.51       0.01   -0.46  -0.92  -1.14 135.0  2.82       2.50    1.89       132          0.76       0.23     -0.29    -0.78  -1.25    1.48 137.5  2.60       2.28    1.66       1.06         0.49      -0.06     -0.59    -1.10  -1.59  -1.83 140.0  238        2.05    1.42       0.81         0.23      -034      -0.88    -1.41  -1.92  -2.17 142.5  2.17       1.83    1.19       0.57       -0.03       -0.61     -1.17    -1.72  -2.25  -2.51 145.0   1.95      1.62    0.96       0.33        -0.29      -0.88      -1.46   -2.02  -2.56  -2.83 147.5   1.75      1.40    0.74       0.09       -0.53        1.14      -1.73   -231   -2.87  -3.14 150.0   1.55      1.20    0.52     -0.13         -0.77      -139       -1.99    -2.58  -3.16 -3.44 152.5   1.36      1.01    032       -0.35        -0.99      -1.63      -2.24   -2.84  -3.43  -3.72 155.0   1.18      0.83    0.13      -0.55        -1.21      -1.85      -2.48    -3.09 -3.69  -3.98 157.5   1.02      0.66   -0.05      -0.73        -1.40      -2.05      -2.69   -331    -3.93 -4.23 160.0  0.87       0.50   -0.21      -0.90        -1.58      -2.24      -2.89    -3.52  -4.14  -4.45   !
162.5  0.73       036    -035       -1.06        -1.74      -2.41      -3.06    -3.71 -434   -4.65 165.0  0.61       0.24   -0.48      -1.19         1.88      -2.56      -3.22     3.87  -4.51 -4.82 167.5  0.51       0.14   -0.59      -131         -2.00      -2.68      -335     -4.01  -4.65  -4.97 170.0   0.42      0.05   -0.68      -1.40        -2.10      -2.79      -3.46    -4.12  -4.77  -5.09 172.5  036       -0.02   -0.76      -1.48        -2.18      -2.87      -3.55    -4.21  -4.87  -5.19 175.0   031      -0.07   -0.81      -1.53        -2.24      -2.93      -3.61    -4.28  -4.94  -5.26 177.5   0.28     -0.10   -0.84      -1.56
                                                             -2.27      -2.97      -3.65    -432   -4.98  -530 180.0   0.27     -0.11   -0.85      -1.57        -2.28      -2.98      -3.66    -433   -4.99  -531 2-16                            ,

           ;2.7       - Qwnes #7 - Maximum Stratifled Flow Condition -                                             4 The calculation of the global moments and local stresses due to thermal stratification were based on a linear top-to-bottom temperature gradient. In Combustion Engineering Owners Group
           ~ (CEOG) repon CEN-387-P, Rev.1-P-A, dated May 1994, the thermal stratification moments used in the ASME Section III fatigue calculations were based on maximum stratified flow conditions. This was accepted by the staff in its safety evaluation repon dated July 14,1993, and
           . included in the CEOG repon. The staff considers maximum stratified flow conditions also -
applicable to the LBB calculations. Provide a reevaluation of all relevant quantities to the LBB application based on maximum stratified flow conditions.
Response to Ouestion #7:
ThE entire spectrum of thermal stratification, including the maximum stratified flow conditions have been considered in the LBB calculations for the pressurizer surge piping. As described in
            ' Appendix D of SIR-98-096, the analysis using the CEOG maximum stratified flow model (not the linear model) was used to derive a multiplier on the linear moment such that the global moments could be derived for each of the stratified loading conditions in the thermal stratification loading spectmm. The thermal stratification analysis showed that the maximum thermal stratification conditions (for a given top-to-bottom temperature differential) were 1.038 times those for a linear temperature assumption (100/96.34). Thus, this multiplier was used on the global stratification moments determined in Appendix B of SIR-98-096. There was also a temperature scaling factor applied based upon the. actual stratification conditions being evaluated.
In addition, the stratification throughwall gradient stress was also derived from the stratification model (also scaled based upon the actual temperature conditions) and was added to arrive at the total stress due to stratification.-                                                                ,
Thus, the analysis has implemented the same loading conditions that were utilized in the CEOG       l repon, such that no funher evaluation is required.
2-17                           .

2.8     Question #8 - Model for Bimetallic Effects Provide the model for calculating the additional local stresses at the nozzle-to-safe-end welds due to bimetallic stress effects.
Response to Ouestion #8:
For evaluation of the bimetallic stresses, a simplified axisymetric finite element model was used. j The model represented the intersection of two equal diameter and thickness cylinders placed end on end. The 2D axisymmetric model was constructed using the ANSYS finite element software package. The model consists of a pipe-to-pipe connection, using pipes of dissimilar materials.
The basic dimensions were 12.75 inches outside diameter by 10.125 inches inside diameter, representive of the safe end -to-nozzle weld dimensions.
The details of the weld between the two materials was not modeled. Instead, there was a material discontinuity at the center of the weld region. The length of each material section was   1 approximately 10 inches (> 38). The shell was modeled with approximately 2:1 aspect ratio elements (length to thickness) with 10 elements through the wall.
The two pipe sections were modeled using the following materials properties at a temperature of 550'"F:
Mean Coefficient of Modulus of Elasticity          Thermal Expansion Material                      E (ksi)                     a (in/in *F) 4 ASTM A-182 F1                          26,700                      7.18 x 10 4
ASTM A-351 CF8M                        25,500                      9.50 x 10
;                                         I           2-18 3
l l

A uniform 550*F temperature was applied to the entire model with the stress-free temperature set at 70*F. To account for the difference between the surge line material properties and the        1 properties used in the analysis, the stresses were ratioed to account for the actual surge line material properties as described in SIR-98-096 Appendix D.                                       I 1                                                                                                         l
\
l l        The axial stresses are shown in Figure 2.8-1 and show the effects of bending across the l
bimetallic location. For the purposes of this evaluation, the stress distributions were derived from the maximum stress locations for each side of the weld.
l l
l l
I l
l l
l 2-19                             >
l

ANSYS 5.3 2
                                         -j                          JAN 27 1999 R                          18:21:49
                                           "                          FLOT No. 1
                                         -E                          NODAL SOLUTION
                                      =                               STEP =1 SUB -1
                                           .                          TIME =1 E                          SY         (AVG)
RSTS=C DMX =.083032 M                                  SMN --14 57 9 SMNB=-27960 SMX =13806 SMXB=76808
_                             g      -14579
_                           g      -11425
                                           . ,                              -8271
                                         '                                  -5117
                                         - E ,,
                                                                     "      -1963 1191 4345 7499
                                                                     "      10653 13806
                                            -a 1
i Figure 2.8-1. Bimetallic Joint Model and Stress Distribution 2-20                    ,

l' l                   .           .
         ; 2.9     Question #9-Polynomial Fit State the purpose for curve fitting the axial stress distribution with a cubic polynomial.
Response to Ouestion #9:
The fracture mechanics model used for the crack growth calculation described in D.7 of SIR            096 Appendix D requires that all stresses be input as coefficients of a cubic polynomial. Thus, all detailed stress distributions were curve fit to determine the polynomial coefficients. Only some of the loading conditions required that the higher order terms be included. This is a commonly accepted practice in performing fracture mechanics analysis.
2-21                          i

I 2.10    Question #10 - Austenitic Versus Ferritic Bimetallic Weld Provide an explanation for why the axial stress distribution in Figure D-5 for ferritic steel is apparently the mirror image of the axial stress distribution in Figure D-6 for austenitic steel, both at 550'F.
Response to Ouestion #10:
The model used for determining the bimetallic stresses was a finite element model of a bimetallic weld as explained in Question 8. Due to the difference in modulus of elasticity and coefficient of thermal expansion across the weldjoint, there is a stress reversal across thejoint. The stress pass chosen was slightly removed from the theoretical bimetallic interface, and was at the location of maximum magnitude of stress on the inside smface of the model. Thus, the bimetallic stresses will theoretically be mirror images.
It should be noted that the bimetallic stresses were much smaller than the weld residual stresses that were also applied at each weld location. Here, the inside, surface residual stress was 30 ksi.
In addition, the cyclic stress ranges due to the bimettalic weld stresses were not a significant contributor to the crack growth evaluation.
i I
l 2-22

     ~2.11   Question #11 - Bimetallic Stress Curve Fit
                   ~
In Figure D-65 the axial stress determined from the curve fit underestimates the axial stmss calculated from the stmss analysis at the outer surface by about 15%. Provide a discussion of the significance of this underestimation in the fatigue crack growth analysis.
ReSDonFa to Ouestion #4:
The underestimation of the bimetallic stress at the outside surface has an insignificant effect on the results. In performing fracture mechanics analysis, it is important to have a good estimate of the stress distribution (and thus the polynomial coefficients) over the region between the surface ID and the crack tip. Thus, the curve fits were purposely selected to match the inside surface stresses and to be representative over the rest of the wall thickness. In addition, the cyclic nature of the bimetallic stresses was not significant except for the startup/ shutdown transient and had essentially no contribution to crack growth.
I 2
i i            2-23                            $

==3.0 REFERENCES==
: 1. CEOG Report CEN-387-P, Rev I-P-A, " Pressurizer Surge Line Flow Stratification Evaluation," May 1994.
: 2. EPRI Report TR-107515, " Evaluation of Thermal Fatigue Effects on Systems Requiring Aging Management Review for License Renewal for the Calvert Cliffs Nuclear Power Plant," December,1997 (transmitted to Mr. P. T. Kuo (NRC) by Mr. John J. Carey (EPRI) on Febmary 9,1998)
: 3.  " Evaluation of Flaws in Austenitic Stainless Steel Piping," Journal of Pressure Vessel !
Technology,352Nol.108, August 1986 (Attached as Appendix B).
: 4. EPRI Report TR-103581, " Thermal Stratification, Cycling and Striping (TASCS),"
March 1994.
i
: 5. United States Nuclear Regulatory Commission, Bulletin No. 88-08 " Thermal Stresses in Piping Connected to Reactor Coolant Systems," June 22,1988.
* l 3-1

m                                                                               ,
i APPENDIX A Reprint from Volume L of 1989 SMiRT Proceedings
            " Stress - Strain Parameters in Elastic-Plastic Fracture Mechanics" by Nathaniel G. Cofie Georges A.Miessi Arthur F. Deardorff Pages 91 -96 4
i I
1 A-0                           ,

    .                Str:ss - Strain Pcramst rs in E1:stic-Plcstic Fracture Mcchtnics Nathaniel G. Cofle, Georges A. Miessi NUTECH Engineers, San Jose, CA USA Arthur F. Deardorff StructuralIntegrity Associates, San Jose, CA USA ABSTRACT The Ramberg-Osgood stress-strain relationship plays a key role in the fracture mechanics analysis of flawed pipes. A methodology is proposed that can be used to determine parameters for this relationship in the absence of experi-mental' data based on ASME Code mechanical properties.
INTRODUCTION Recent advances in elastic-plastic fracture mechanics have reduced problems that previously necessitated complicated finite element analyses to simple handbook solutions which only require a good description of the crack geometry and the material stress-strain characterization. Solutions for several crack geometries and loading conditions have been provided through studies performed under the auspices of the Electric Power Research Institute (EPRI) [1-3]. In these studies, the material stress-strain characterization is of significant importance and is described by the Ramberg-Osgood (R-0) model in the form:
T                *     "
(1)
                       =     +a["]"
o     o           o where o and c are true stress and true strain, respectively, and c , o , a and n are model parameters that must be determined for the SateSiat of interest. The parameters e and c are reference stress and strain, respac-tively. Usually, o is tak8n as tiie yield stress, o Y, and e is taken to be the pseudo-elastic field strain (o /E).               If the stress-strain curve of the material is available, these paradters can be determined by curve fitting.
Although there have been attempts to provide complete material data for nuclear power plant materials in the open literature, the stress-strain curves for most materials under various conditions are generally not available. In the absence of experimental stress-strain curves, it would be desirable to determine the R-0 parameters based on available mechanical properties. This paper suggests a simple method for the d? termination of the R-0 parameters using ASME Code-specified mechanical properties and compares the parameters so obtained with parameters determined elsewhere in the literature.
DETERMINATION OF RAPEERG-OSGOOD PARAMETERS The value of the parameter a is determined from geometrical considerations of the engineering stress-strain (S-e) diagram shown in Figure 1.                 In this figure, point A on the curve represents the ASME Code-specified 0.2% offset yield stress (5 ). The corresponding engineering strain at this p(ointS /E)           is given by eos =e     + 0.002 where e is the pseudo-elastic yield stress y y
91

shown as' point C in Figure 1.            These values of stress and strain at point A on the curve   are known equations:
converted'to      true    stress and true strain values using the well o = S (1 + e)'                                                          (2) s = An (1 + e)                                                          (3)
        !When the true stress and true strain values at the 0.2% offset yield stress are substit_uted into Eq.1, an expression for a is obtained as follows:
(  An(1  in(l'++e,,)   0.002 + e(1+e)       )) y~ ((1 +0.0020.002 I4)
                                                                                      + e ))
11+0.002+e')n                         e#
C (1 + ey )     )
Given the parameter a, the.value of the-parameter n can be determined directly from Eq.1 tif the values of stress and strain are known at one other )oint on              j the stress-strain curve.
ultimate strain (eu ). The An            obvious choice is -the ultimate stress (S ) and value of Su15 provided in the ASME Code for au wide range of temperatures encountered during normal cperation of nuclear power                 {
plants. The value of- e can       u be estimated based on ASME Code minimum percent-age elongation which is provided in the Code at room temperature. Converting the values of S and e u parameter n is 8etermined           as:to true stress and strain values and using Eq.1, the in(1 + e )       S (1 + e )
I"      I                                  )3 an(1 + e ) ~ S (1 + e I                                (}             j
                                   'ADE  S (1uS
                                              + eIl) + 'u 3 y       y                                                 l t
      -Other expressions have been suggested in the literature for the determination                i of .the R-0 parameters in the absence of experimental data. In one method                    i suggested by Bloom and Malik [4], the parameters a and n are obtained from an                !
iterative solution of the following expressions:
S u            e1 /n                 1/n         1 i
5      (1.002 + e ) , [In(1.002 + e )) /n                       (6)            i y                                ;
g       [tn(1.002 + e   y )]II"
                    " *Ty (          (1.002 + e1 )        )n                         (}
I Another method was suggested by Gerber et al. [5] where these parameters .are                '
determined as follows:
n.= 1/in(1+eu )                                                  (0) in(1+e )        Su I1**u)
                   **E in(1+eu )            ~ yS (1+e         Sy)(1+e 3 yE) u(1*u) -n(}
i y                y The values of the parameters a and n determined for Type 304 stainless steel at 20*C- (68'F) and ~ 288'C (550'F) using the various formulations in Eqs. 4
     ' through 9 are shown in Table 1. 'It can be observed from Eqs. 6 and 7 that the 92

method of Malik and Bloom is independent of the ultimate strain while the other two methods show some dep:ndence on the ultimate strain.        Because the ultimate strain (taken as the percentage elongation) is not specified at all temperatures in the ASME Code, the parameters are computed at two values of ultimate strain, the minimum specified value in the Code at room temperature (30%) and a value of 60% which is considered to be more representative [6]. It can be seen that the parameters ralculated using Eqs. 4 and 5 are not sensi-tive to the choice of ultimate straL' while there is great dependency of the parameters on the ultimate strain in the method suggested by Gerber et al. in Eq. 8 and 9. It can also be observed from Table 1 that if the minimum elong-ation (e u) specified in the ASME Code (30%) is used for this material, the parameters obtained from the methodology proposed in this paper compare reasonably well with those obtained from the other two methods. However, there are significant differences between those of Gerber et al. (Eqs. 8 and
: 9) from the other two methods when eu = 60%. Parameters determined from experimental data for Type 304 stainless steel in [7] are also shown in Table
: 1. Figure 2 compares the stress-strain curves obtained with the parameters from the various methods at 288'C (550'F).
FRACTURE MECHANICS ANALYSIS The different sets of R-0 parameters determined in Table 1 have been used in-elastic-plastic fracture mechanics analyses to determine the sensitivity of the different sets of parameters in fracture prediction.
Simple handbook solutions provided in [1-3] were used in this study. Figure 3 shows typical results of the J-intergral 'vs. the crack length obtained for a 24-inch Sch. 120 pipe with a through-wall fl aw. As can be expected, the differences in the results are not very significant when the R-0 parameters with the ASME Code minimum elongation are used in the analyses for this material, otherwise significant differences are observed.
DISCUSSIONS AND CONCLUSI3NS In the formulation presented in this paper, the R-0 parameters a and n are determined by considering only two points on the stress-strain curve, hence, they can be considered as approximations in the absence of experimental data. The parameters obtained with the proposed formulation for Type 304 stainless steel base metal were compared with those obtained from other formulations in the literature [4, 5] and were found to compare well when ASME Code minimum percentage elongation is used.        The advantage of the proposed formulation is that it is not sensitive to the choice of the ultimate strain and it ensures that both the ASME Code yield and ultimate values are satisfied on the curve.       Comparison with experimental data shows some differences; however, it should be noted that parameters determined from experimental data are heavily dependent upon the stress-strain range used for the curve fitting. In general, it is difficult to obtain one set of parameters to describe the whole stress-strain range of the experimental data as shown in Figure 4 from [7]. Figure 5 from [7] shows how the experimentally determined      !
parameters can change depending on the stress-strain range used for the curve     l fit. Furthermore, Table 1 shows that the average parameters determined experimentally can change for the same material types obtained from different sources. Hence, although the R-0 parameters suggested in this paper are       ;
approximate, they provide a consistent and unambigious set of parameters based    !
on ASME Code mechanical properties.         In order to use these parameters in fracture mechanics analyses to determine unstable flaw sizes, it will be necessary to use an anologous approach to derive an appropriate material J-resistance curve.
93

  ,  REFERENCES
[1] V. Kumar, M. D. German and C. F. Shih, "An Engineering Approach for Elastic-Plastic Fracture Analysis," Electric Power Research Institute, Report No. EPRI NP-1931, Research Project 1237-1, July 1981.
[2] V. Kumar, et al., " Advances in Elastic-Plastic Fracture Analysis,"
Electric Power Research Institute, Report No. EPRI NP-3607 Research Project ~1237-1, August 1984
[3] V. Kumar and M. D. German, " Elastic-Plastic Fracture Analysis of Through-wall and Surface Flaws in Cylinders," Topical Report No. EPRI NP-5596, Research Project 1237-5, January 1988.
[4] J. M. Bloom and S. N. M'alik, " Procedure for the Assessment of the Integrity of Nuclear Pressure Vessels and Piping Containing Defects,"
Electric Power Research Institute, Report No. EPRI NP-2431, Project 1237-2, June 1982
[5] T. L. Gerber et al., " Evaluation of High-Energy Pipe Rupture Experiments,"
Electric Power Research Institute, Report No. EPRI-5531, Project 2176-2, January 1988.
[6]" Mechanical      and     Physical Properties of Austenitic Chromium-Nickel Stainless Steel at Ambient Temperatures," The Internal Nickel Company, Inc., Bulletin A, New York, New York,1963.
[7] K. Kishida    and    A. Zahoor,     " Crack-0pening     Area  Calculations for Circumferential Through-wall Pipe Cracks," Electric                    Power Research Institute, Report No. EPRI NP-5959-SR, August 1988.
Table 1 Ramberg-Osgood parameters for Type 304 stainless steel base metal.
                                               'o     E       'o     a     n     a     n 3
x10     x10*3 Material     Relationships       (ksi) (ksi) (in./in.)  eu = 0.30   eu = 0.60 Eqs. 4 and 5        30.0  28.3      1.060  1.89 4.13   1.89 3.94 Eqs. 6 and 7        30.0  28.3      1.060  2.85 3.79   2.85 3.79 Eqs. 8 and 9        30.0  28.3      1.060  2.75 3.81  23.07 2.13 34.7  28.3      1.226  3.82 5.04   3.82 5.04 st an ss at 1N1 and TT-02 (7)
(680F)    Experimental        39.3  28.3      1.389  3.46 5.68   3.46 5.68 JAERI TT-102 (7)
Experimental        45.3  30.0      1.510  9.16 3.20   9.16 3.20 EPRI/GE (7)
Eqs. 4 and 5        18.8  25.75     0.730  2.74 3.29   2.74 3.23 Type 304    Eqs. 6 and 7        18.8  25.75     0.730  3.70 3.11   3.70 3.11 a     ss
                ,     g Eqs. 8 and 9        18.8  25.75    'O.730  1.27 3.81  17.65 2.13 2880C    Experimental        20.2  25.50     0.788  6.86 3.16   6.86 3.16 (5500F)     NRC/8CL 4131 1 AND 4141-1 (7)
Erperimental        26.1  25.50     1.024  4.08 3.73   4.08 3.73 NRC/BCL 4141-3E (7) 94

Sh e
Y _
m C        A/,-
                                 'Y               I
                                                 /               /lI
                                                /               /I
                                               /               / I
                                              /              /     I
                                            /               /
                                           /               /       l i
                                                          /        I
                                                        )          l E            /  E I
1 j /1                                                          I f                                  =
e                           l 0.002         (ey+ 0.002)
Fig. 1 Detennination of parameter a from engineering stress-strain curve.
100 90 -
50 -
70 -
5        60 -
an 50 -
id        40 -
30 -
20 -                                                       a PROPOSED (a - 2.74, n - 3.29) l       -                                           O BLOOM AND MAltK (a - 3.70, n - 3.11)
                                                                         + GERBER ET AL (a - 1.27, n - 3.81) 10 J                                                     O GERBER ET AL (= - 17.65, n - 2.13) 0        .     .    .       .        .       .       .   .    .    .     .      .      .     .
O        O.04          0.08             0.12        0.18      0.2         0.24          0.25 TRUE STRAIN (IN/IN)
Fig. 2 Comparison of true stress - true strain curves of Type 304 stainless steel at 288'C.                                                                               .
95

l            ioo                   _ . -
l                                                                     A PROPOSED (= = 2.74, n - 3.29)
I              90 -             40 kal                               O BLOOM AND MALIK(a -3.70, n -3.11) l                                                                     + GERBER ET AL (a - 1.27, n - 3.81) l                                                                                                                                                               I O GERBER ET AL (a - 17.65, n - 2.13)
      ?        70 -
                         ~
5 1
s0 -
50 -
c l
1 40 -
4        30 -
20 -
10 -
10 ksi 0    .
y     y       y      7     7      7       y    .      ,    ,       ,         ,     ,      ,     ,       ,       ,      ,      ,
O             2              4            6          8          10                  12          14             15             18        20 HALF CRACK LENGTH (IN)
Fig. 3 Parametric curves of J-integral versus through-wall crack length for a 24-inch sch.120 stainless steel pipe at 288'C (remote tension stress                                                                          I
                  = 10 ksi and 40 ksi).
Low strain fit
                                                                                                      - o = 3.33 n = 6.20                                             l l
100     --
2 High strain fit 7 50 g
                                                                                                                         ,,$%                  . 21.50
                                                                                           ,                                                   n = 1.91 m-                                             D                              I           -
0o                                 3           #                     Mid strain fit
                                                                                                      ~
B                                    o = 5.98 l                                               0 O                                                                                 n = 4.29 60  -
Ramberg-Osgood                        {           --
_                                               curve                            {a 10       ~
A                                                                                {                 e=1%
l#
65
                ~"
c U                 oo= 20.1 (ksi) 80
* 7.882E-04
                                                                                          $ 5
                                                                                                    ;                                  ,, . 20.1 (ksi)
I e = 8.800                                                                                E = 25.500 (ksi) 20                                n = 3.100                                               -                                g = co/E D = Experimental data                                                                    O = Test data 0          !        !         !         I       I       !                            1                            i                i        !
0      0.05     0.10      f 15     0.20     0.25    0.30                              1                        2                 3       4 Strain                                                                      Strees Ratio (o/co)
Fig. 4 True stress - true strain                                             Fig. 5 Ramberg-Osgood law fit of
}                   curve for SA 376 TP304 base                                                           true stress - tiue strain material in 15tC/BCL 4131-1                                                            for NRC/BCL 4131-1 and and 4141-1 pipes [7].                                                                 4141-1 base metal [7].
l l
96

r.
l I
r i:
l APPENDIX B Reprint from Journal of Pressure Vessel Technology Vol.108, August 1986
            " Evaluation of Flaws in Austenitic Steel Piping" by Section XI Task Group for Piping Flaw Evaluation ASME Code Pages 357.- 366 i
B-0 i-i i

r
    ; ps4 .. . . e                                                                                                                                                       .
c;.4 ..                                           -
&,;;,; . :.a $; V s   o 3
Pressure Vessel and Piping Codes l
l
                                                                                                                                                                           \
l l
1 Evaluation of Dawsin Austenl4c Steel                                  shielded metal arc welds (SMAW). Allowable naw sizes were                     l Pipmg                                            developed for both categories.                        .                       ]
During operation in the LWR temperature range, cast austenitic piping material experiences an aging phenomenon that changes the material ;oughness and strength properties.
Section XI Task Group for Piping Flaw Evaluation, ASME Because work is still in progress to define the magnitude and Code                                                                               effect of these changes, cast austenitic piping is not iden" ed explicitly in either material category. Consequently, riaw evaluation for cast pipe requires appropriate material This report summarizes the methods and bases used by the classification on a case by case basis.
Task Groupfor Piping Flaw Evaluation to detelop allowable                              ne base metal and nonflux weld evalation is based ca a               -
flaw sizesfor the American Society of Mechanical Engineers plastic collapse iailure mechanism and the allowable tiaw sizes                                   e tASME) Boiler and Pressure Vessel Code, Section XI,                               were generated from limit-load analyses. The flux weld CB-3640, " Evaluation Procedures and Acceptance Criteria evnluation is based on an unstable crack tearing failure p Austenitic Piping." Because the Task Group identified mechanism and the allowable flaw sizes were determined using                                              !
tu s flaw-related failure mechanisms for light water reactor clastic plastic fracture mechanics analysis methods.                                                j (L VR) austenitie stee/ piping, two sets of allowableflaw size                        Allowable circumferential flaw sizes were developed using            i tables were deteloped. One set is presentedfor base metal and margins of 2.77 (for normal / upset conditions) and 1.39 (for nortflux welds, which have highflaw tolerance; a second set is emergency / failed conditions) on the sum of the primaty mem-providedforflux welds that can hatt lower relative toughness branc and bending stresses. Circumferential flaw evaluation cndflaw tolerance. Application of the flaw evaluation pro-                         for the lius welds also requires including the expansion stress ceduresfor cast austenitic piping requires material classifica- with a margin of 1.0. Allowable longitudinal 11aw sizes werc                                i    l tion on a case by case basis. Allowableflawsizesareprese. 't i developed using marsins oi 3.0 and 1.5 on primary membrane                                   ;
in tabularform, with the allowable part-through flaw depth stress for normal / upset and emergency / faulted conditions,                                    t    1 defined as a func; ion of load and flaw length. Tabler are respectively.                                                                                    i presentedfor both longitudinal and circuntferentialflaws, and                           ne Task Group also recommends tatigue and corrosion                  i normal / upset and emergency / faulted conditions. This report crack growth rates, and residual stress distributions that may                               I also recommends crack growth rate and weld residualstress be employed in tae subcritical crack growth nnalysis outlined                                     {
data needed to perform subcritical crack growth analyses in in Secties XI, Appendix C, " Evaluation of Flaws in                                       _
accordance wita Section XI, Appendit C, " Evaluation of Austenitsc Piping.
Flaws in Austenitic Piping." These data and the allowable                                                                                                    ,
flaw size tables provide the means to determine the accep- 1 Intrt& tion                                                                                     1
:bialityfor continued service of austenitic piping containing Sectio 1 Xl of the ASME Code l1) and 10CFR$0 of the Code flaws whose sizes exceed the standards spectfled in Sect on XI'                     of Federal Regulations (2) require periodic inspection of reac-           .
IWB-35,1,4J, " Allowable Indication Standardsfor Austenitic tor coolant pressure boundary and related safety systems com.                                   {
          #IPI "I'                                                                           ponents in operating commercial nuclear power generation                 t his report describes the bases and methods used by the Sec-facilities. When the inspection results indicate flaws, the flaws may & evaluated uskg me aegtance standa@ in Sem
[
j tion XI Task Group for pip ng       i Flaw Evaluation to develop XI, IWB-3500 to determine if the flawed component can bc                                   <
flaw evalunu.on procedures and allowable flaw sizes for safely returned to service. Flaw sizes larger them those custenitic steel and high n;ckel alloy pipmg.
i
                                                                                         . specified in IWB-3500 may be evaluated further using analysis The evaluation procedures, flaw sizes and data contained in procedures and acceptance criteria specified in IWB-3600.                                  I this report are applicable to materials normally employed in                           Prior to publication of the Winter 1983 Addendum to the              i the design and construction of light water reactor (LWR) Code, evaluation procedures in IWB-3600 were at allable only                                       j rustenin,e ntng systems. The austenitic materials have been for ferritic steel components 4 in. or greater in thickness. Flaw                               s placed into two categor':s based on their flaw toleraace. The evaluation procedures and allowable flaw sizes for LWR                                        f.
first category includes high toughness materials such as austenitic piping first appeared in IWB-3640 in the Winter wrought material and nonflux welds. ne second category in-                          1983 Addendum. The evaluation is based on a plastic collapse           i(
cludes flux welds, such as submerged are welds (SAW) and failure mechanism and the allowable flaw sizes were developed                                     f,j using limit-load analysis. Because plastic collapse is the an-          tg ticipated failure mechanism, secondary stress is assumed to be          $
relaxed at failure and only primary membrane end bending               g Conuibuted by the Pressure Vesseis and Piping Division for publication in the stresses are used when performing flaw evaluations in accor-            g Jaumm or Panssums vsssu Tsc                  . Manuscripi reemed by the            dance with IWB-3640. ne evaluauon procedure in the 1983                p Pressure Veuels and Piping Division. March 26.1986.                                Winter Addendum is applicabic to all wrought and cast                   p-4 3S2 / Vol.108, AUGUST 1986                                                                                          Transacllonsof the ASME              f
  .--.-...<u-....:-...,.-___.                                  .- ...... - .               .      .    ._                           . . .         . .      . . .-. 4

i                                                                                  !
l' sra eP                        ,
I Fa=sama L
                                                                  '-'                                          n,,   % sU,n.
                                                  ,,,                                                        -       _ s     ,,n.,                               '
er.<-                                                                   ::y:ll;'
4     t,,,,,     l                          tll:                                               l n,                      fem a.,                        Ase.mtm M, slas w .en.
a.n                   saae                       a.
* Normal Censlaans
                                                'a**as                     a,                        ,.
                                                        'N             ^ :'::":'
a, .
as sa, H v., I consieris l           l Fig.1 Section XI pipe flaw evaluation sequence custenitic product forms, and weldments normall'y associated corrosion crack growth rates, and residual stress distributions with LWR austenitic steel piping.                                            are included in this report.                                               ,
Paragraph IWB-3640 includes allowable flaw sizes for both                  The following sections of this report present the evaluation circumferential and longitudinal flaw orientations. For cir- procedures and flaw sizes developed for the two classes of cumferential flaws, the allowable flaw size tables include austenitic materials, and the information needed to perform margins on load of 2.77 and 1.39 for normal / upset and suberitical crack growth analyes for flawed austenitic piping.
emergency / faulted conditions, respectively. For longitudinal Section 2 presents the methodology and bases used to develop                             <
flaws, the margin associated with normal / upset and emergen- the allowable flaw sizes. Section 3 contains the Task Group's                             '
cy/ faulted conditions ate 3.0 and 1.5, respectively.                        recommendations for the crack growth rates and residual in March 1985, the Pipe Flaw Evaluation Task Group and stress data needed for suberitical crack growth analysis. Con-Working Group for Flaw Evaluation proposed modifications clusions and recommendatiocs are listed in Section 4.
[3] to the austenitic piping flaw evaluation procedures presented in IWB-3640. De modifications were proposed to 2 Haw Evaluation Methodology and Bases distinguish between high toughness materials and certain lower toughness flux welds, which include shielded metal are                   2.1 Materials. The criteria described in this document welds (SMAW) and submerged are welds (SAW). This distmc- may be used for the evaluation of piping products of the tion became necessary because some small specimen cx- following material types included in Table 1-1.2 cf Section 111, perimental data suggest that the applicable failure mecham,sm ASME Boiler and Pressure Vessel Code:
for the flux welds is unstable crack extension that would occur ct loads lower than the plasue collapse load.                                 I w       ht 'austenide steeln The net effect of the modification is to reduce the allowable 2 cast austenitic steels with a ferrite content less than either flaw size at any specified load for flux welds relative ' the                   20 pem or 20 M4 higher toughness materials. Because limit load may be ' :hed                 3 wrought nickel-chro um-iron Alloy 600; prior to failure, it is recommended that the expansion. ... esses            4 weldments associated with the base materials in items I,2, with a margin of 1.0 be included along with primary mem-                         and 3*
brane and bending stress when evaluating flaws in flux welds.
Recognition also is given to thermal aging of cast products and Applicabilhy is limited to materials with a minimum specified reqmrements are presented f9r case by case evaluations. The yield strength less than or equal to 45 ksi at room temperature.
evaluation procedure and associated allowable flaw size tables                  Different criteria have been developed for these materials.
were adopted by the Code in the 1985 Winter Addendum.                        The   different criteria arise from the possibility that the IWB-390 presents the allowable flaw size at the end of any materials may have varying degrees of fracture toughness.
specified operating interval. To assess the integrity of a flawed Restrictions and limhations for these materials are discussed component an analysts must be performed when the flaw is later in this document.
detected to determine how much larger the detected flaw will                    2.2 Failure Csiteria Definition. Developlag accurate be at the end of any subsequent specified operating interval. flaw evaluation procedures and allowables requires defining The end of interval flaw size is compared to the allowable size the appropriate flaw-related failure mechanism and using the in IWB-3640 to determine if the component can be safely analysis method associated with that mechanism to determine returned to service for the desired time period. The overall the relationshp between allowable flaw size and load. Four flaw evaluation sequence is illustrated in Fig.1.                            flaw related failure mechanisms can be considered in struc-The procedures and criteria for performing the suberitical tural applications. These are: 1) rapid, nonductile crack crack growth analysis are specified in Section XI, Appendix C.               extension, 2; unstable ductile tearing prior to limit load, Ts support the evaluation procedures for subcritical crack 3) plastic collapse (limit load), when the remaining ligament growth, the Task Group recommendations for fatigue and of the cracked pipe section becomes fully plastic prior to crack Joumal of Pressure VesselTechn31ogy                                                                                 AUGUST 1986, Vol.108/ 353
              ..    - . - -       - - - - - - - - - - -     -     -              -                                                                .   . _. . d .

               . 3    .
       *
* extension.and 4) unstible ductile tearing subsequent t211mit to ratcheting. In addition, limits on pe k stress levels are
* load. The failure mechanism fir any specified cpplicati:n is specified 13 protect against f:tigue failure.
        . dictited by the material flaw t:lerance (t:ughness) and              To cpply the evaluati:n procedures cnd acceptance criteria
       ', estrength pr:perties, Haw size and shape, and stress distribu- in IWB-3640, the loading conditions specified in the design tion in the component.                                             specification and/or stress reports for normal operating con-Rapid, nonductile failure is applicable for ferritic materials ditions (including upset and test conditions) as well as ct low temperatures and generally is not applicable to emergency and faulted conditions need to be established. He custenitic materials. The third and fourth mechanisms appropriate primary membrane (P,,,) and bending (P ) stresses described in the previous paragraph apply for very tough for both normal, and emergency and faulted conditions are in materials where the remaining ligament of the cracked pipe the piping stress reports. As Section 2.5 discusses, when section becomes fully plastic prior to failure. Normally, plastic evaluating flaws in flux weldments the thermal expansion collapse is associated with load control application, while stress (P,) should be includes in the stress ratio as a primary unstable ductile tearing subsequent to limit load can occur for stress in addition to the primary membrane and bending displacement control conditions. The second mechanism ap-          stresses. Furthermore, the sum of the primary membrane and plies when some stable ductile tearing occurs and the load car-    bending stresses, and the thermal expansion stresses should rying capacity of the structure is less than that predicted for not exceed 25, (where S, is the mimmum ASME Section III limit load.                                                        yield stress) for the structural integrity assessment of flawed Results from experimental and analytical studies (see Sec-    nux weldments.
tion 2.4.4) indicate that austenitic mateirals, such as wrought       The primary membrane and primary bending stresses used product and nonflux welds, will reach limit load prior to in ASME Section XI, IWB-3640 correspond to the unconcen-                           j failure. Consequently, the Task Group selected plastic col- trated primary stress intensity values defined in equation (9) of                l lapse as the appropriate failure criterion for wrought material ASME III Section NB-3650. The expansion stress (P,) is the                   !
and nonflux welds, and used limit-load analysis to develop the unconcentrated stress intensity value for moment loads de-                    I allowable flaw sizes.                                              fined in equation (10) of ASME NB-3650. The use of un-Results from a limited number of small specimen laboratory concentrated primary membrane and bending, as well as tests (see Seedon 2.5) indicate a relatively lo v toughness for general expansion stresses, is based on the fact that all girth custenitic flux welds compared to the wrought material and butt welds are assemed to be in relatively straight pipe and ninflux welds. These data and subsequent analyses described that local surface irregularities at th'e welds have little effect on in Section 2.5 indicated failure by unstable tearing prior to the weld joint strength with respect to gross plastic deforma-limit load. Based on this informadon the allowable flaw sizes tion. It is noted that the evaluation procedures limit the region f r flux welds were developed from clastic plastic fracture of applicability to a distance of V(Rt) from the centerline of a mechanics analyses using the /-integral crack driving potential girth butt weld. This limitation ensures the validity of the and associated ductile tearing modulus instability criterion.       straight pipe assumption and the use of unconcentrated During operation in the LWR temperature range, cast stresses in the flaw evaluation procedure. Application of the custenitic piping material experiences an aging phenomenon flaw evaluation procedure presented in ASME Section XI thn changes the material toughness and strength. Because IWB-3640 to other regions or other piping product forms re-work is stillin progress to define the magnitude and effect of quires the use of appropriate piping stresses, these changes, cast austenitic piping is not identified explicitly in either of the two material categories. Consequently, flaw           2.3.2 Margirts. The margins of safety against postulated             I evaluation for cast pipe requires appropriate material              failure modes in piping systems are based on the relative im-classification on a case by case basts.                             portance of the piping system or component and the The flaw evaluation procedures and a!!owable flaw sizes for likelihood of the defined loading condition. Therefore, piping the plastic co!! apse and ductile tearing mechanisms are systems are designed as ASME Class 1,2 or 3 consistent with described in Sections 2.4 and 2.5, respectively.                    their safety classification. The allowable flaw sizes presented m@
m mMe @ m N 1 piph h a 2.3 Loads and Margins                                               cept of service level categories is used to adjust the margin of 2.3.1 Loads. To perform limit-load and fracture                safety based upon the likelihood of the event. The categories are: Service Level A-Normal Operating Conditions; Ser-mechanics analyses of austenitic piping systems,it is necessary vice Level B-Upset sad Test Conditions; Service Level to esatblish the appropriate loads acting on the pipe section C-Emergency Conditions; and Service Level D-Faulted containing the flaw. In the case of circumferential flaws, the Conditions.
axial forces and bending moments need to be determined,                The allowable primary stress limits for ASME Section Ill while the primary membrane stress due to pressure is required Class I, piping as a function of the service levels are:               i for axial flaws.
j Nuclear piping desirn criteria are governed by a combina.                Service level ti:n of ASME See6.un 111 and Nuclear Regulatory Commis-Class I limits                  i t
si n requirements. The ASME pis.ing criteria have been                              A                         1.5S.
established to provide margin against failure under the static                                                                              ;
E                    1.85 s 1.5S,                     t loads encountered in normal service, and dynamic loads                              C                   2.25S. s 1.8S,                    i associated with low probability, abnormal events, such as                           D                    3.05, s 2.0S, earthquakes and postulated loss of coolant accidents. The                                                                                  f stresses due to applied loads are categorized consistent with For convenience in developing and applying flaw evaluation                 -
their contribution to potential failure modes. The primary procedures, Service Levels A and B have been placed into one stress satisfies equilibrium between internal and external category and Service Levels C and D have been placed into a f:rces and moments. Secondary stress is the stress developed second. The Class I limit for the normal / upset category is 1.5 by the self-constraint of a structure and is self-limiting as local S., while the limit for the emergency / faulted category is           .-
yielding and/or minor distortions will reduce or eliminate the 3.0S..
stress. The primary stress limits in the ASME Code Section !!!         One of the primary objectives of the Pipe Flaw Evaluation          i, are specified to protect against plastic collapse failure modes. Task Group was to develop the ASME Section XI IWB.3640 The primary plus secondary stress limits assure shakedown to allowable flaw sizes consistent with the safety margins con-j elastic behavior and thus protect against potential failure due tained in the applicable piping design criteria of the ASME               j 1
3S4 / Vol.108, AUGUST 1986                                                                             Transactions of the ASME i
                                                                                                                                                 -mn

         .' I..        f.
Section 111. For normal conditions the minimum safety margin on primary membrane stress (P.) is 3 because the                              1
       ;         Code requires P. < S., where S. Is the lower of S,/3 or 0.9                               )~-
S, given in Appendix 1 of Section 111, and 5, and S, are the 8
p . sm ,              [
Code aBowable ultimate and yield strengths. De margm for                             ~{       a
                                                                                                                                   ~'
the emegency/fauhed co agory is taken as half the margin for                                                     "'""*
normal / upset condition, or margin on load equals 1.5. This                                      .          .           _
i                criterion is consistent with the Code design basis.
Fig. 3 Lensumenes swoose new geomewy l                     The margin against failure la pure bending can be deter-mined from the product of strength and computed stress margins ==aded with Code practice. Using equation (9) of NB-3600 for straight pipe at normal / upset conditions (P. +
l P, s 1.5S.), and setting P. = 0 gives P. s 1.5S.. Taking which plastic collapse is predicted is determined from the the ratio of the failure stress at limit load 3S., (see Section foHowing equations:
2.4.2) to the 1.55, aBowable gives a margin of 3S./1.5S. =                                  P,' = 2af[2 Sin A-(a/t).(Sin O)]/r              (1) 2.0. In addition, the ratio of the moment needed to produce full plasticity in the (uncracked) pipe section to the moment re-             whm quired to produce flow stress, af at the pipe surface, or                                      # = [(1- Oa/t)- (P. '/aj)s-]/2 def8t /rafR8t , is 4/r. nus, the marsm at normal / upset con-                 or if 0 + $ > r, then
* diuons for pipes in pure bending is (2).(4/r) = 2.55.                                                                                                 i The safety factor used for developing allowable flaw sizes                                 P,' = 2af[(2-a/t) Sin $]/r                 (2) for normal / upset conditions is the average of the Code where minimum safety margins, or 1/2(3.0 + 2.55) = 2.77. For fauhed conditions, the safety margin is one half of the value                                 g,,gj _,j,_#",/# YI2-###I I
for normal conditions, i.e., 1/2 (2.77) = 1.39. While these                  P.' and P,' in the two foregoing equations are the primary margins are developed from plastic collapse considerations, membrane and bending stresses, respectively, corresponding they also were used to generate the aHowable flaw size tables to plastic collapse for any specified crack depth and angle.
presented in Section 2.4.5 for the unstable ductile tearing I"U"''              ""*""                                                         2.4.1.2 I ongitudinal Flaws. An empirical formulation              ,
for the hoop stress at failure was developed by Eiber et al. [7]      l for pipes with axial throughwan flaws and is                          i 2.4 Flaw Evaulation for Plastic Conspee. The critical
                                                                                                                          ,, . ,,fy                        (3) crack size for high toughness materials was detennined using a net-sectaon plastic collapse criterion. Plastic collapse failure              *h*''
assumes that at failure the remaininF ligament of the cracked                   M= [l + (1.61/4Rt)/8]i/2                                    (4) section of the pipe is fully plastk prior to any extension of the                l= total axial flaw length crack. His criterion implies that the flawed pipe is at the point               R = mean pipe radius ofincipient failure when the net section in the crack plane first                t= pipe wall thickness forms a plastic hinge. Failure is assumed to occur at a critical flow stress, af, which is considered a material property. The For part-throughwall axial cracks (see Fig. 3) the hoop stress at failure is value of the flow stress typically is defined as (a, + a.)/2, where a, is the 0.2 percent offset yield and a, is the ultimate.                                   aa = a/[(1 -x)/(1 -x/M)]                 (5) tensile strength.
where x = a/ti ne remaining portions of this subsection describe the ap-plication of the plastic collapse criterion for developing flaw                   2.4.2 Plow Strest Definition. ne flaw acceptance evaluation procedures for ductDe piping containing i$cr criteria in this document assumes that flow strength at net sec.
            - 1:nsitudinal or circumferential flaws.                                         tion plastic collapse is 35.. The basis for this assumption is a 2.4.1 Conqputation Procedurer                                            cou.h # 3S. with flow stress determined from ex-perimei,M issults for circumferentiaDy cracked piping (see 2.4.1.1 Circumferential Flaws. De basis for the flaw Section 2.4.4) and tensile strength properties, evaluatson procedure for high toughness austenitic matenals                       During completion of the work described in [4], tensile test has been described previously by a==gaanth and co-workers data from plate and pipe matenals were used to determine the 14,5] and is based on earlier work by Fanniaan et al. [6]. The acceptabilty of using the relationship relationship between the couapse load and flaw size is ob-tLined by requiring force and moment equilibrium of the pipe                                           af = 1.15(S, + S,)/2                  (6)    I section (see Fig. 2). The crack depth, a, and half angle,0, at to define the flow stress (af) as a function of the ASME Code yield (S,) and ultimate (S.) strengths for austeniuc piping steel.
                                                     ,,,,,,,,,,,,,,,,,,,,,                      Figure 4 shows flow strength values calculated from equa-tion (6) using yield and ultimate strength d1tta for the sampling womenes semen         -=
                                                                                  . + P. er of seepipe "Tl    l--P materials; applicable ASME Code and weMment
                                           *   ::                                           values of 35, are also shown. The results show that 3S. is a
:E  - - - - - - -
lower bound to the flow strength calculated for base metal at f' /                g                                             room temperature, and is a reasonable estimate of the median
                   'l l            8      -
::                                            base metal flow stress at 550F.
                        \
                                ._          4                                l                  De plot also shows that welds have a higher flow strength
                                          \         -----                    ;              than base metal at $50F (and at room temperature, although
                                  "*''*' "*1                                                not shown in the figure). This is due almost entirely to the
                                                  't"'''""'==            ~~I   '"
higher yield strengths of Type 308 welds compared to Type pig,3 m,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,i,,,,,,g.                               304/316 base material. The ultimse s'.rengths are comparable ispse ses esennemen                                                            for the two groups of materials.
Journalof PressureVesselTechnology                                                                               AUGUST 1986,Vol.1081355 J

                      ~
: , ' . r.         '
: j.          *'
materials. De allowchic flaw size tables in IWB-3640 specify aw                    the allowable flaw depth relative to pipe wall thickness, a/t, as ss=                                  ao.mo as %e
                             ,  _                            1 a function of flaw length reladve to pipe circumference, O/r, l                                            and the applied stress ratio 4  .                            I                                                                         SR = (P , + P,)/S.
I                                                                                                                   (7) g                                            where P, and P, are the applied primary membrane and 3  -
I bending stress, respectively, at the location of interest.
g'                The allowable loads and flaw sizes were defined by main-e.,    '
i
[           l l   l
                                                                                          ,               taining adequate safety margins on the stresses at plastic col-lapse (P,', and P,,' in equations (1) and (2)). Using equation
   ;                     ,s     -                                                                         (7), this criterion can be written as j                    j       -
ga s.
8 ng (P,, + P,)/5., = (P,' + F, ')/(S,, *SF)                (8)
{                    .1                                                                                where SF is the margin on load and is equal to either 2.77 for
   '                    1                    I
   ,'                                                                                                     normal / upset or 1.39 for emergency / faulted.
                        }'      -
I
                                             !                                                               Tables IWB-3641-1 and 2 were constructed from equations
                        **      -                                                                         (1) and (2) by assuming a constant P.,', af = 3S,,, and soMng l                                                            for P,' as a function of relative flaw length and depth. ne
                             ,      ,        i             ll       ,        ,      ,      ,        ,     allowable stress ratio at any specified flaw length and depth
                               .o         so                no               ro           ao combination was then obtained by substituting P.,', P,', and the appropriate margin into equation (8). As a convenience in 8  -
y constructing Tables I and 2, the axial stress at plastic collapse, 11 Samosus s  -                    3
                                             %see                                   s50'F to I                                                                                                    \
g l                                                                                       1 ss
                                ~
I                                                                        u   -
                                                                                                                                     \                 '%
l                                                                    E                        Pm + Pe = Sm               '"""**"
tg           i         i                                 e                       fcollapse has from o     '       I!      I     IIl      I II       I I '         '        '
* Eas tu and t21l 40         So                to              ro            40
{      _
Calculated F)ow Stromrth, kal                                    y                                     p, + pb = 2.773 Sm Fig. 4 Cornparison of flow strength and a 3,, values for stainless steel pipes and weld. Flow stress = 1.15(37 + 3, A f=
                                                                                                                                                       * * " #~II*
{os 1
The selection of 3S,, as the flow stress simplifies the flaw                               ja2 _      swa ssirs evaluation procedure and provides readily available data for                                                 "****''
using equations (1), (2), (3) and (5).
2.L3 Allowable Flaw Size Tables
* I          1           l          1 a         u          o.,        u           o.e     , .o 2.4.3.1     Circumferential Flaws. Equations (1) and (2)                                                       Fact $aa a' cma*aace em were used to generate tables of allowable flaw sizes for cir.
Fig. 5 Cornparison of typical computed s!!owable ftsw stre using equa-cumferential flaws in high toughness austenitic piping tions (1) and (2) with staw evaluation standard, tws.ssia.s Table 1 (IWB 36411) Allowable end-of-evaluation period flaw depthi'8 to thickness ratio for circumferential flaws-normal operating (locluding upset and test) conditions (P., + P,)/S (88                                    Ratio of flaw length. lfto pipe circumferencek' O.0             0.1            0.2            0.3             0.4              0.5 or greater g,3                       to              to            to             to              to                to l                                                    I.4                     0.75            0.40           0.21           0.15             84                (4 l                                                   1.3                     0.75            0.75           0.39           0.27            0.22              0.19 1.2                     0.75            0.75           0.56           0.40           0.32               0.27 l                                                   1.1                     0.75            0.75           0.73           0.51           0.42               0.34 i                                                   1.0                     0.75            0.75           0.75           0.63           0.51               0.41 0.9                     0.75            0.75           0.75           0.73           0.59               0.47 0.8                     0.75            0.75           0.75           0.75           0.68               0.53 0.7                     0.75            0.75           0.75           0.75           0.75               0.58 50.6                      0.75             0.75           0.75           0.75           0.75               0.63
(''Flawdepth=a, for a surface flaw i                                                             2a, for a subsurface flaw I                                                        t= nominal thickness Linear interpolation is permissible
(*' P. = primary longitudinal membrane stress (P,, s 0.5 5,,,)
P = primary bending stress S.,, = allowable design stress intensity (in accordance with Section !!!)
I" Circumference based on nominal pipe diameter (4 1WB-3514.3 sha!! be used l
1 3561Vol.108, AUGUST 1986                                                                                                         Transactions of the ASME t
l                                                                                                 .

fi               .
l.,             ,
     ~. ;     .
                   .?
i      ,              Table 2 (IW?-36412) Allowable end-of-eva!=tio] period ft:w depth'** ts thickness ratia fzr circimfere ti;l
        *-             flaws-esmergemey and faulted conditions Ratio of flaw length. P f, to pipe circumference (''
(pg p,)fs, ten O.0            0.1         0.2              0.3              0.4               0.5             0.75           1.0 3.0                   80             (O          (O               M                 (0               (*              (O             (*
2.8                  0.75           0.46        0.24             0.17              0.13              M               M              to 2.6                  0.75           0.75        0.39             0.27              0.22             0.19            0.17           0.17 2.4          -
0.75           0.75        0.54             0.38              0.30             0.26            0.24           0.24 2.2                  0.75           0.75        0.68             0.48              0.38             0.33            0.30           0.29 2.0                  0.75           0.75        0.75             0.58              0.46             0.40            0.35           0.35 1.8                  0.75           0.75        0.75             0.67              0.54             0.47            0.41           0.40 1.6                  0.75           0.75        0.75             0.75              0.62             0.53            0.46           0.46 1.4                  0.75           0.75        0.75             0.75              0.69             0.60            0.51           0.51 s 1.2                   0.75           0.75        0.75             0.75              0.75             0.66            0.56           0.55
("Flawdepth=a, for a surface flaw 2a, for a subsurface flaw t = nominal thickness Linear interpolation is permissible
(*l P. = primary longitudinal membrane stress (P , < 1.05.)
P, = primary bending stress S. = allowable design stress intensity (in accordance with Section !!D                                                                              l The sum (P , + P )3sha!! not exceed 2 5,, where S yis the Section !!! specified minimum yic!d stress M Circumference based       on nominal pipe diameter M1WB-3514.3 shall be used P.', was assumed equal to 0.55, for normal / upset condi-                                          3a, = 3S.,[(1 -x)/(1 -x/M)]                     (9a) tions and 1.05, for emergency /fauhed conditions, respective-              ,,
ly, nis assumption is based on the fact that piping typically is designed with hoop stress for normal / upset conditions equal                                      aa /S., = [(1 -x)/(1 -x/M)]                   (9b) to about S., so that the axial stress is 0.5S.,.                           Using equation (9b), the allowable stress ratio as/S,,, can be Tables IWB.3641-1 and 2 for normal / upset and emergen- determined as a function any specified crack depth and length, cy/ faulted conditions, respectively, are repeated in this report             S milarly, evaluation sizes were generated for emergency for completeness as Tables 1 and 2. Tables IWB-3641-1 and 2 and faulted conditions by using a margin on load of 1.5, which                                I contain three modifications to the allowable flaw sizes ob. results in the right side of equation (9b) being multiplied by tained from the indicated development procedure. First, max- 2.0 imum nondimensional crack depth, a/t, was limited to 0.75 to                  This procedure was used to develop Section XI Tables provide additional conservatism. Also, for crack lengths IWB-3641-3 and IWB-3641-4 for normal / upset and emergen-greater than or equal to 50 percent of the circumference, the cy/ faulted conditions, respectively, ne tables present the allowable flaw depth was assumed to be constant and, equal to allowable stress ratio a /S,,, as a function of normalized crack the depth for a 360-deg flaw. Because the variation in the length, IN(Rt), and depth, a/t, and are repeated for com-allowable flaw size for naw lengths exceeding 50 percent of the pleteness in this report as Tables 3 and 4 circumference was small, the simplified procedure is                         The allowable flaw size tables contain two modifications to reasonable and conservative. Finally, the lower bound allowable values were fixed at allowable flaw standards of IWB-3514-3, Section XI, ASME Code. His is consistent with the Code because Daws below the allowable flaw standards of IWB-3514.3 need not be evaluated. As an example, the                                                                              a allowable relative flaw depth is illustrated in Fig. 5 as a func-                        ro                                       t tion of relative flaw length for a single stress ratio.                                                                         2a As an alternative to using Tables IWB-36411 and 2,                                    "  -
                                                                                                                                                      -l F allowable flaw sizes and loads can be obtained directly by up ing equations (1), (2), (7) and (8) on a case by case basis for                    ]                                              g flaw specific evaluations. In this application, P ,' wou'd be set                   i so    -
equal to the actual primary membrane stress at the flaw loca-                      f, tion of interest rather than 0.55., and P,' can be obtained                        ; ,,     .,,,
from equations (1) or (2) for the measured flaw depth and length. The allowable stress at measured flaw size is obtained
                                                                                                           }                                       r.        g3 .
30 -
by substituting P,' and P.,' into equation (8). His procedure appears in an approved Code Case, " Alternative Equations for Evaluation of Flaws in Austenitic Piping," and will appear                          so _
in Section XI, Appendix C of the 1986 Winter Addendum to                                        c,     w     ist w the Code.                                                                               ,,
                                                                                                                     ,,,, w cmia v e on w 2.4.3.2 _ '.ongitudinal Flaws. Allowable flaw sizes for longitudinal flaws in high toughness austenitic pipe materials                           ,           I     i       1      I         I      o o         s     *       *      *        'o    58 were generated using equation (5) and a margin on load of 3.0 for normal / upset conditions and 1.5 for emergency / faulted                                             co.es = ts.i.ia.=
conditions (see Section 2.3.2).                                           Fig. s noeutte troen steineen steel center cracked penas tests showing For example, for normal / upset operation, equation (5) can renouve p euen of creen inausuon and conspee u a funcuan of strees be written as [8]                                                         and creet sine Journalof PressureVesselTechnology                                                                                AUGUST 1986, Vol.1081357

              '   ,,, Table 3 (IWB-3641-3) A!!owable end-of. evaluation period flaw dep3M12 Cickness ratto for axial flaws-normal e pe ating
       ,'                 (inclnding upset and test)condhions
       *n .
Sticss ratiote                                                   Nondimensional flaw length (d [lf /Jer]
O.0      0.5       1.0     2.0     3.0       4.0     5.0     6.0     7.0         8.0   9.0    10.0     11.0   12.0 or greater s 0.4            0.75     0.75      0.75    0.75    0.74      0.70    0.68    0.67    0.66        0.65  0.64    0.64     0.64    (0 0.5            0.75     0.75      0.75    0.72    0.65      0.61    0.59    0.58    0.57        0.56 0.55     to       (o      u) 0.6            0.75     0.75      0.75    0.64    0.55      0.51   0.49     0.48    0.47         (O    (*     (*       44      (0 0.7            0.75     0.75      0.73    0.53    0.44      0.40    0.38    0.37     (*          (*    (4     to       (0      (0 0.8            0.75     0.75      0.62    0.40    0.32      0.28    0.26     (*      (4          (4    (*     (0       (4      (4 0.9            0.75
(*      0.70      0.42    0.23    0.17      0.15   0.14      to      (o          (4    (*     (0       to      (o 1.0                     (o        (*      (*      (O        (*      (*      (4      (4          (o    (*     (O       (0      (4
(*)flawdepth=a,, for a surface flaw 2a, for a subsurface flaw Linear interpolation is permissible
(*) Stress ratio =(PD/2t)/S.
I                      where i                          P = maximura pressure for normal operating conditions l                          D = nominal outside diameter of the pipe t = nominal thickness S,,, = allowable design stress intensity (in accordance with Section !!D tol, =end of evaluauon period for flaw length
                              / = nominal radius of the pipe t = nomina 1 thickness (A!WB.35413 shall be used Table 4 (IWB-3641-4) Allowable end-evaluation period flaw depth'd to thickness ratio for axial flaws-emergency and faulted condidons Stress ratioco                                     Nondimensional flaw length (d [(If /drt))
0.0         0.5         1.0        2.0         3.0           4.0        5.0         6.0 or greater s 0.8             0.75        0.75        0.75       0.75        0.74          0.70       0.68          (o 1.0             0.75        0.75        0.75       0.72        0.65          0.61        40           (4 1.2             0.75        0.75        0.75       0.64        0.55           (O         (O           to 1.4             0.75        0.75        0.73       0.53        0.44           (4         (4           to 1.6             0.75        0.75        0.62       0.40         (*            (4         (0           (4 1.8             0.75        0.70        0.42       0.23         (4            (0         (0           (o 2.0              to          (*          (4         to          (O            (o         (0           (o
(" Flaw depth =a. for a surface flaw 2ir for a subsurface flaw Linear interpolan,on is permissible toStress Ratio =(PD/2t)/S,,
where P = maximum pressure for normal operating conditions D = nominal outside diameter of the pipe t = nominal thickness S,,, = allowable design stress imensity (in accordance with Section 110
(" If =end-of-evaluation period for flaw Icagth r = nominal radius of the pipe t = nominal thickness
(*IWB.35145.3 shall be used
    ,                    the flaw sizes determined by equation (9b). Allowable crack pipe. These panels were tested at room temperature under a
    ,                    depths are limited to 75 percent of wall thickness. Also the variety of loading conditions, including monotonic, inter-nondimensional flaw lengths, IN(Rt), are limited to the rupted, cyclic, and seismic, and with cracks that included both allowable lengths for a throughwall flaw calculated using saw cut and fatigue initiators in sensitized aad unsensitized equation (3). This flaw length limit guarantees that surface material. The results of all these tests (see Fig. 6) indicate that flaws will remain below critical size (based on the plastic col- plastic collapse is appropriate for circumfercutially flawed lapse conditions) should they grow through the wall. Flaws ex.                high toughness austenitic piping.
ceeding this length are limited to the a/t spec.fications in                      Eighteen 4-point bending tests were performed by IWB-3514-3.                                                                   Wilkowski as described is [9] on circumferntially flawed 2,4 I
2.4.4 Erperimental and Analytical Veryication. The                        and 16-in-dia wrought type 304 stainless steel pipe. Nine 4.in.
l                                                                     .                          pipes and one 16-m-dia pipe were tested with internal surface j                   following two subsections present additional detail from pipe flaws. Two 2-in., si.t 4-in. and one 16-in-dia pipes were tested
    ;                   expenments and pipe system analysts to verify that limit-load              ,  with throuf,hwall flaws. One 4-in. pipe had a combined analysts is an appropriate basts for high toughness austenitic piping material.                                                              throughwall/part-throughwall crack geometry. All tests were
    .                                                                                                 conducted at room temperature.
2.4.4.1 Pipes With Circumferential Flaws. Early ex-                          The test results showed that the load at the onset of stable penmenta.1 work by Brock and Marschall [6, 9) used 12-in, crack extension was at least 93 percent of the maximum load i                   wide, center-cracked Type 304 stainless steel panels tested in for pipes with throughwall cracks, and the 4-in. pipes with sur-
     !                  uniaxial tension to simulate circumfer-ntially flawed 4-in-dia face and compound crack geometries. For the 16-in.dia 3581Vol.108, AUGUST 1986                                                                                         Transactions of the ASME

            .# 1-
           * * *
* surface-flawed pipe stable radial flaw growth started at about (10) was extended to acccunt f:r the c:mplex, three-
   . ..          73 percent cf maximum load. In all tests, maximum load digiensi nal pipe system geometry. This involved determining ranged from a few percent below to 110 percent of the load an effective pipe length, L,,r, that could be used with equation predicted by the not-section collapse criterion. 'Ihe flow stress (10) to represent any pipe system specific geometry. The from the pipe experiments was determined to be about 70 ksi. details of this work was published later in a separate report by This value indice n that 33. (S. = 20 kai at RT)is a conser. Paris, Tada and Macek [15].
vative estimate for flow stress for circumferentially flawed          Crack stability analysis using the Last parameter also were austenitic pipes at room temperature.                              performed for a wide range of LWR piping by Cotter, Chang, Wilkowski and co-workers [10] recently performed tests at and Zahoor [16]. The results showed that large pipe breaks
                 $50F on three 6-in-dia pipes having circumferential, part- were unlikely for high toughness austenitic materials at loads throughwall cracks. The test results show that the maximum significantly in excess of faulted loads typically postulated for load is between 86 and 116 percent of the predicted collapse LWR piping system design. Subsequently, simite analyses load when flow stress equal 35, is assumed. These data are performed by Zahoor and Gamble (17] showed that the poten-                          I consistent with the ben metal tensile strength data shown for tial for unstable crack extension for a throughwall crack in 550F in Fig. 4.                                                    piping systems having part-throughwall cracks at other weld The data in [9 and 10] demonstrate that limit load analysis locations does not differ from that predicted for pipes con-with flow stress equal 3S. is appropriate for defining taining a single throughwall flaw.
allowable flaw size for high toughness austenitic piping.                                           .
In addition to the experimental verification, the results from    2.4.4.2 Pipes With Longitudinal Flaws. Experimental previo% performed fracture mechanics analyses can be used data for austenitic piping with axial flaws are limited. In 1971 to demonstrate that plastic collapse is a conservative criterion Eiber et al. [7] published the results from 34 experiments on for developing allowable flaw sizes for piping materials with Mpnsentative Type 316 stainless steel and A106B carbon steel high toughness. German and Kumar [11] used elastic plastic nuclear piping having diameters up to 24 in. Four experiments fracture mechanics analyses to show that flaw instability were performed for 24-m-dia stainless steel piping having part-would occur very near predicted limit load for high wughn ss through flaws; the test temperatures ranged from 600 to 675F.
austenitic piping. The analyses included 4 and 26-in. pipe The experimentai results from the austenitic pipe tes'.s indicate diameters and conservative esti==res for Ju and stable crack that plastic collapse is an appropriate flaw evaluation pro-growth. Load controlled conditions were assumed.                   cedure for axially flawed austenitic piping.
Implicit in the plastic collapse criterion is an assumed          Using the reported failure pressures, the hoop stress % =
deadweight (load control) load application. However, an- #,/2t, and equation (5), the indicated experimental flow ticipated and postulated load applications for LWR piping stress is about 40 ksi, or about 2.4S, at the test temperature.
systems often may be displacement limited rather tnan load This value is about 20 percent less than 3S., which was shown controlled. When displacement limited loading prevails and to be a reasonable estimate of average flow stress at $50F the remaining ligament of the cracked pipe section is fully based on material tensile properties (see Section 2.4.2) and cir-plastic, crack extension will reduce the load (and stored elastic curnfenndally flawed pipe (see Section 2.4.4.1). Although the energy of the pipe system) and the pipe can tolerate deforma. experimentally determined flow stress for axially flawed pipe tions larger than those associated with the plastic collapse is 2.45.,3S. is used as the bas,si for the allowable flaw sizes load.                                                              for the following nasom The conditions under which tearing instability will occur subsequent to predicted limit load was first described as part        1 Because the Code S. values for austenitic on an NRC Pipe Crack Study Group effort [12] to evaluate the elevated temperature are generally governed                     by ield, the the y, ste integrity of BWR piping containing IGSCC. The details of this actual safen margins on ultimate strength are greater than analysis was described in a subsequent report by Tada Paris,       th"8-       .
and Gamble (13] These studies showed that the potential for           2 Lu.nsting the maximum allowable crack depth to 75 per-unstable crack extension for fully plastic, displacement- cent of wall thickness and limiting allowable crack length (see controiled conditions could be estimated for straight pipe sec. Section 2.4.3.2) provides additional margin above that used to tions by comparing the material and applied tearing modulii, develop the allowable flaw size tables.
or instability occurs when                                            2.5 Duetile Teoring Evalua*los for Circumferential Fisws. As Section 2.4 describes, the allowable flaw sizes 4 m T.,                            W   presented in IWB-3640 for high toughness materials were ob-where                                                              tained using the assumption that pipe sections containing T"" = material tearing modulus                                     flaws will reach limit load prior to any significant increase in T                                                                  initial flaw size. However, some experimental data from small F i .(L/R)
LM == ratio     of pipe+length Es*j to radius                             laboratory specimens indicate that austenitic steel flux welds              ,
F      = 2F/r                                                      have reduced toughness, which may produce sigmficant or                     l
               'F                                                                   unstable crack extension before the flawed pipe section                     i
                        = [Cos(0/2) - 2 Sine] E/2FRa/                               reaches the limit load associated with the flux weld flow str*ss.
F      = Sin (0/2) + Cose O      = half-angle of throughwall flaw                               The reduced load-carrying capacity for flux. welded
                 /      = / integral                                                austenitic steel piping was determined by Zahoor and Gamble E      = modulus of elasticity.                                    [18] by calculating the ratio of the load to produce unstable tearing to the limit load. The reciprocal of this ratio was used For typical BWR pipe systems and loadings F, and F are to define a modification factor that multiplies the applied on the order of 1.0, and L/R is the dominant term in equation stresses from the stress report; this increased stress is used to (10); L/R was estimated to range from about 20 to 30. For enter Tables IWB-3641 1 and 2 and define the allowable errsk very tough materials T., was estimated to be at least 200; depth for any crack length. This procedure in effect reduces consequently, it was concluded [13] that unstable ductile tear- the allowable flaw sizes for flux welds relative to austenitic ing would not occur subsequent to the remaming ligament of wrought product and non flux weld materials.                             ,
the cracked pipe section im " ; fully plastic. Later, as part         Definition of the modification factor required parametne of another NRC Pipe Crack Study Group effort (14], the analyses to define the effect of pipe and flaw size, load crack stability estimate in stra ght pipe represented in equation distribution, and weld toughness and strength properties. The Journal of Pressure VesselTechnology                                                                  AUGUST 1986 Vol.108f 359

l , ,, ,      , . ,
Tm kT                             (15) d                                                                                        The instability cor.fiti:n m y be f:und by plotting / as a to                                   function of T for both the applied loads and flaw size, and material crack extension resistance. The intersection of the 1
R It --
material and applied J/T curves defines the instability condi-g                                               tion [22].
The instability computation is illustrated in Figs. 8(a) and (b). First, the applied 1and Tvalues are determined as a func-
                                         ~                                                   tion of applied load for a specified crack length using equa.
tions (11) through (14), and are plotted with the material Fig. F Pipe cross section containing a througtethe-wall crack resistance to ductile tearing on a 1/Tplot as shown in Fig.
8(a). The / at the instability point in Fig. 8(a) defines the load j                                                                                            at instability as illustrated in Fig. 8(b). The ratio of the in-                     l results were used to develop conservative estimates for stability load from Fig. 8(b) to the limit moment at the i                    modification facts t for both SAW and SMAW. The re- specified crack size indicates the reduction in load-carrying mainder of this section describes the analysis procedures and capacity relative to limit load for the fiux-welded pipe at the results,                                                                specified crack and pipe size and load distribution.
The moment at limit load for a pipe containing a
'.                       2.5.1 Computational Procedure. The instability load throughwall flaw and subjected to bending loads can be i                    was determined using the J-integral and the associated teanng calculated from equations (1) and (2) and can be placed into modulus crack stability criterion. Because / solutions are not the form [6]
available in the literature for finite length, part-throughwall flaws, the reduction in load carryms capacity was estimated                               M"48 Rs/    t
[Cos(0/2)-(1/2)  Sin 0]             (16) based on previously developed clastic plastic / solutions [11               Representative materials data needed to compute J T and at d 19] for the throughwall crack geometry illustrated in Fig.          limit moment were obtained from experiments conducted at
: 7.                                                                       550F for both SAW and SMAW by Landes and co-workers Using the formulation presented in [11 and 19], J for [23]. The Ramberg-Osgood stress strain parameters and other throughwall cracks can be expressed as the sum of clastic and matenal properties are:
plastic components, or J, = J, + J'                         (l1)                                  Submerged       Shielded metal Parameter               arc weld           are weld In this formulation the clastic portion of /is a                      11.0                9.0 J, = K,s/E                          (12)                 n                       6.9                9.8 where K,is found from LEFM solutions adjusted for plastici.                        ',, ksi                    33.7              49.4 a ,ksi 42.1              55.4 e eneral form for the plastic portion of1in bending is                         ,ksi               25,000.0           25,000.0 1,c, in-Ib/, ,m-              650.0             990.0 J, = aa,e.e-(a/b).hi .(M/M.)"+ 8                   (13) where the values of a, n, a,, and e, are defined by the Ramberg-Osgood true stress true strain relationship for the The material J resistance curves were obtained from 1-in.
,                    material                                                                compact tension specimens cut from weldments; the resultant 1/T plots are shown in Figs. 9 and 10 for SAW and SMAW, e/e, = a/a, + a (c/a,)"                          respectively.
and 2.5.2 Reruits. The reduced load-carrying capacity is il-2e = length of remaining circumferentid ligament of the lustrated in Fig.11 where the instability to limit-load ratio !s cracked portion of the pipe,                                  presented as a function of circumferential throughwall crack 1                   a/b = crack length normalized with respect to circumference length for SAW, bend loading, and several pipe diameters.
I                             (2a/rD)                                                       The results in the figure show the pipe size effect and indicate l                   h i = function which accounts for relative crack and compo-l                              nent size, and material work hardening,
 ;                  M, = reference moment associated with an average stress of magnitude a, in the cracked section, and                        J                                    Ja M = applied moment.
The applied tearing modulus, due to Paris and co-workers
[20] and Hutchinson and Paris [21], used in this. work cor-g ">
eu2.n, responds to assumed load control conditions and is T.=(d1./da)(E/a/)                            (14) where d1./da indicates the increment of / needed to produce a specified increment of crack extension at any specified load                                                                          I and crack stae. Eis Young's modulus, and af is the material                                     =='*                                    I flow stress.                                                                                                                            I u
The material resistance to unstable crack extension is deter.                                                                    ey I l
mined experimentally from test data that relateland crack ex-tension. Using the experimental data the material tearing                                    7                                Las
                                                                                                                                                      '" --dv l                  modulus, T,,, can be determined from the slope of the 1ver.
I sus crack extension curve using equation (14). At any specified                            ,                                    .
                    / level, where J. m Ja. unstable crack extension will occur rio. s o. termination of instabmey ,t. r, and s.sociated load for load j                   when [20]                                                               control etastic plastic fracture meenanics ane#y,is 360lVol.108, AUGUST 1980                                                                                        Transactions of the ASME

f . , ' .'l ' -
  ' . ' ,, U                                                                                    'the multiplier approach represented by equatlins (17)
[.'                 ""~                                                              thr: ugh (20) was later modified in discussi:ns with the NRC.
These modifications involved limiting allowable flaw depth to yesmes                                                   60 percent of wall, thickness, and specifying that the multipliers in equations (17) and (18) be computed for o.d. =
                     ,,                                                                     24 in. for pipe sizes less than 24 in. This later requirement was j'"'-                                                                 Intended to account for uncertamties in the thermal expansion I        .
stress for smaDer pipe sizes.
t        -
s                                                             Based on these restrictions, n9w tables were generated for
                           ,,,,,     '%[                is flux welds [3] and appear in the code as Tables IWB-3641-5 and 6 for normal / upset and emergency / faulted conditions.
respectively. These new tables are based on equations (17) through (20) and Tables IWB-3641-1 and 2, but inicude the o    ....'....i....t....t..                       ..          restrictions indicated in the previous paragraphs, and are
                                "          'oo    son          mo      em          me shown here as Tables 5 and 6.
The multipliers defined in equations (17) and (18) also can Fig.9 OWenninetton of J and 7 at erask InstaWuty for Saw at 850er be applied if the alternative evaluation method described in eaun -
[                                                                                               Pise Dense,. supe u-                                  .
   .                 's acoo ;                                                                                                              ''
3                                                                                                                    as
[*********                                                         se-                            f             p J..eoo 1
1
_         ///
U                      A u
                              'o              E'      '      ' '     '  '     ''
400 Fig.10 Detennination of J and T st crack instaWNty for SMAW at s50F u   -
that the load-carrying capacity ranges from 60 to 80 percent of                                      ,           ,            ,         ,
limit load for pipe diameten # sips froin 4 rp 34 in., respec-                             .        ai          u           u          a. u tively. Similar computatioti - $rformeo ? 9 SMAW welds;                                                 ,,,,,,,,,,,a,,,,,,,,,,,,,,p3 typical results are presented in Fig.12 for a 28-in-dia p!pe and tension and bend loading.                                                   Fis.11 Ra#o of instaWuty enoment to tinm enoment as a function of car.
                                                                                               "                - z new len9Ui for wwad pipe diameters. SAW at Using the results in Figs.11 and 12, estimates were obtained            ',,,","''*"*'l for the stress multipliers by taking conservative bounds of the results. The recommended multipuers are (18]
Z= 1.15[1 + 0.013(o.d. -4)]                   (17) for SMAW, and Z= 1.30[1 + 0.010(o.d.-4))                    (18) for SAW where o.d. is the pipe outer diarneter in inches.                               u_                                        _
2.5.3 Allowable Flaw Site Tables. As discussed                                  }                    W previously, the multipliers lo equations (17) and (18) are used                                                                    s to increase the applied loads when using Tables !WB-3641 1                              u    -
                                                                                                                                       /
and 2 to define allowable flaw sizes for flux welds. Because                                             s==.=__/
flux welds are predicted not to reach limit load prior to failure.
the secondary stresses may not be relaxed near the failure                              s.-
load. Consequently, the Task Group i+_ ---- +i that the expansion stress, p,, be included in the stress estio with a margin of 1.0 when the evaluating flux welds. Using either                              u    -
equations (17) or (18), the r--dda= used to define the stress ratio for flux welds in Tables IWB-3641 1 and 2 are Stress Ratio = Z(p, + p. +p/2.77)/S,                         (19)                ',         d          E-           E         4    a fh normal and upset conditions, and                                                                   r       .e c=         - . tem Fig.12 Reno of instaWuty leau to umit load as a function of cir.
Stress Ratio = Z(p, + p, + p/1.39)/S,                        (20)   sumsesenties amonghwou new sength ser beneing and asias leadings.
for emergency and faulted conditions.                                       as.in.dia pipe, salaW st $80F Journalof PressureVesselTechnology                                                                                  AUGUST 1986, Vol.108 / 361

       . 6
  <.-                         Table 5 (IWB-3MI 5) Allowable end-of-evaluation period flaw depthl'8 13 thickness ratis for cir.
  '* " * *
* eumferential flaws la flux welds normal operating (loci: ding t pset and test) conditions
     *'                       Stress                                            Ratio of flaw length. l f, to pipe circumference ('l ratiot*l                                                                                                                             l 0.0              0.1               0.2                  0.3              0.4             0.5 or greater 1.05                    (*                (*               (4                   to                (4              to 1.00                   0.60             0.25              0.16                 0.13               (4              to 0.95                   0.60             0.47              0.25                 0.18             0.13             0.12 0.90                   0.60             0.60              0.39                 0.27             0.22             0.19           1 0.85                   0.60             0.60              0.51                 0.36             0.29             0.25 0.80                   0.60             0.60              0.60                                                                  l 0.45             0.36             0.30 0.75                  0.60              0.60                                                                                    l 0.60                 0.52             0.43             0.35 0.70                   0.60             0.60              0.60                 0.60             0.51             0.41 0.65                   0.60             0.60              0.60                 0.60             0.55             0.45 5 0.60                    0.60             0.60              0.60                 0.60             0.60             0.49
(*8 Flaw depth =a, for a surface flaw,2a, for a subsurface naw t= nominal thickness Linear interpolation is                                                                                                           i I
(6) Stress ratio =M.[P.+permissibleP, + (P,/2.77)]/S.                                                                                 I where                                                                                                                                 l S. = a!!owable design stress intensity (in accordance with Section llI)
P,,, = primary longitudinal membrane stress (P. s 0.5 S.)                                                                         l i
P. = primary bending stress                                                                                                       1 P
M ==unconcentrated 1.0 for SMAW expansion when o.d. stresses s 24 in.
M = 1.0 + 0.01 (o.d. - 24) for SMAW when o.d. > 24 in.
M = 1.08 for SAW when o.d. $ 24 in., and M = 1.08 + 0.009 (o.d. - 24) for SAW when o.d. > 24 in.
8'ICircumference based on nominal pipe diameter (41WB-3514.3 shad be used Table 6 (IWB-341-6) Allowable end-of-evaluation period flaw depth (88 to thickness ratio for cir-eumferential flaws In flux welds - emergency and faulted conditions Stress                                              Ratio of flaw length. I f, to pipe circumference ('8                               i ratio (*'                                                                                                                           '
O.0,             0.1                0.2                  0.3              0.4            0.5 or greater          i 2,3                    to               (o                 to                   to               (o               to            l 2.0                   0.60             0.28               0.17                 0.14              (4               (4 1.9                   0.60             0.52               0.27                 0.19             0.15            0.12 1.8                   0.60             0.60               0.39                 0.27             0.22            0.17 1.7                   0.60             0.60               0.49                 0.34             0.27            0.22 1.6                   0.60             0.60              0.59                  0.42             0.33            0.26
  '                                1.5                   0.60             0.60              0.60                  0.44             0.39            0.30
  !                                1.4                   0.60             0.60              0.60                  0.56             0.44            0.34 1.3                   0.60             0.60              0.60                  0.60             0.51            0.38 s 1.2                    0.60             0.60              0.60                  0.60             0.56            0.47
('' Flaw depth =a, for a surface flaw i                                                 24 for a subsurface flaw i                                             t = nominal thickness
  ;                              Linear interpolation is permissible l
                            '''Suess Ratio = M.[P. + P, + (P,/1.39)]/S.
where S. = a!!owable design stress utensity (in accordance with Sec      J on !!!)
:                              P. = primary longitudinal membrane stress (P. s 1.0 S.)
  !                              P. = primary bending stress P
M == 1.0 unconcentrated for SMAW when  expansion  stresses o.d. 5 24  in.
M = 1.0 + 0.01 (o.d. - 24) for SMAW when o.d. > 24 in.
M = 1.08 for 3AW when o.d. s 24 in.. and M = 1.08 + 0.009 (o.d. - 24) for SAW when o.d. > 24 in.
The sum (P. + P )3sha!! not exce:d 2S ,f where S, is the Section III specified minimum yield stress.
I'8 Circumference based on nominal pipe diameter toIWB-3514.3 shall be used 3
Sect cn 2.4.3.1 is used (i.e., P.' = P. from the stress report Tables IWB-3641 1 and 2 with the maxtmum flaw depth and at the flux weld location ofinterest is used to solve equations mmtmum outside diameter restrictions provide the same (1) and (2), rather than 0.55.).                                                results shown in Tables IWB-36415 and 6, the Task Group Tables IWB-3641-5 and 6 include the stress ratio believed construction of separate allowable flaw size tables for multipliers, M, for SMAW and SAW as a function of outside flux welds will allow more efficient flaw evaluation.
diameter relative to 24 in. The equation for M shown in the tables were derived from equations (17) through (20) and Tables IWB-36411 and 2 using o.d. - 24 in. and SMAW 3 Subcritical Haw Growth material as a reference. While the Z multipliers applied to                       The allowable flaw sizes presented in Section 2 are the 3621 Vol.108, AUGUST 1986                                                                                               Transactions of the ASME

nr''          ~
g . .'    ,
     **     " **b*                                                                                                "-
   .                      auowable flaw sises at the end af any specified operating inter.
I .?                val. As Fig. I shows, to determine the length of time a flawed                          =-
pipe can remain la servise a subcritical flaw growth analysis                           se-       [8=a .,, %
must be performed to ensure the initial flaw sine does not grow                     Iw . #(            7"{ s g N,' -
by fatigue and/or corrosion to exceed the size allowed by                           ( -->  .                                                  . .
                        . !WB 3514.3 or IWB-3640 dunas the specified operating                                s.                              \ :"          f latervali Stress corrosion and fatigue crack powth analyses usually
                                                                                                              ] ,,,   ,
                                                                                                                                                  %Q            -
e are based on liniar elastic fracture ==ch==tes (I.EFM), which                      J..                                                     % ,,,,,, ''3 assumes crack growth is contround by the stress latensity fac-tor, r,. Elastic stras intensky factors ==aaef==d with cracks la piping can be calcalated ming previously developed                                  *   ~
i,,           ,          ,        ,         *         '
                          =had= if the crack size and shape, and stress distribution are                                  a.            e          a        se known. The a= lent =*ad i=*=== lam are then combined with the                                                   mee momener. inee experimentaDy determined crack growth rates to predict crack               Fle.14 Olsoumissensiel seendesel wold otsees in HAZ et the inside sur.
growth in piping over the specified operating laterval.                    '''''' 'h* P'P* *e e henction es deemeier et seheduse so pipe Due to the nature of the corrosion and fatigue crack growth
:==eh==3            , the weld r==utual stress must be included in the subentical flaw growth analysis, in addition to the primary and mandary strues typicaDy preunted in piping stnes reports. Bar== the applied stresses are known from the stress in magnitude to the axial residual stresses [24,31]. The data in 1
report, the ability to --, y predict crack growth depends Fig.14 and finite element calculations [26,1;,i suggest that the edge M the crack smwth rate and nsidual stress hoop stresses on the inner surface of large-diameter (k 16 in.)
pipe-to-pipe weldments become compressive, although this may                    on t e est input duing faMcadon.
Section 3CI, Appendix C' '' Evaluation of Flaws la Austenitic Piping" provides criteria and general guidehnes for                    Finite element calculations [26,28), which are supported by performing subcritical crack growth analyses but gives no e          ed, expental menantnents [24, 25], assm ht guidance for selecting an appropriate crack growth rate or there are sigmficant differences in throughwall residual tross weld residual stress distribution. The following paragraphs of distributions between small (4-in.) and large (=20-in.) dia this section summarize the Task Group's recommendations ents. W calcuMas aim magat that the fonn d for appropriate generic weld residual stress distributions, and the stress distributions in intermediate diameter (10 to 12-in.)
corrosion and fatigue crack growth rates.                                   weldments depends strongly on the weld heat input [28]. The raidual stras diffemaces are due primarily to the increased 3.1 Residual Stresses. There is significant residual stre.s heat capacity of thicker wall pipe and are only coincidentally
                -        variation from weldment to weldment even for large-diameter associated with diameter. Consequently, the recommended pipes, and substantial effort has been made to calculate residual stresses are given in terms of wn!! thickness with a                                   ,
and/or sie;                    lly measure the magnitude and distribu- transition chosen somewhat arbitrarily at 1 in.
tion of residual stresses in austenitic pipe weld =*ats [24-30].                  Based on the available data and finite element computa-Figure 13 presents the avaDable experimental data for the tions, the axial and circumferential throughwall residual stress axial, residual stress in large diameter pipes as a function of distributions recommended by the Task Group for subcritical                              !
radial distance into the pipe wall from the inside surface [4, crack growth analysis are presented in Fig. IS. Because there is                          i 25]. Few expenmental measurements of circumferential a lack of data, uniform distributions are conservatively recom-                                     I throushwall residual stress distributions have been reported. : ended for the circumferential throughwall residual stresses.                             i Figure 14 shows the available data for circumferential, inside                    Pipe-to-pipe welds represent only a portion of the welds in            i surface residual stress as a function of pipe diameter. Inside surface measurements on 4 to 12-in. pipe weldments show that .
the circumferatial residual stresses are tensile and are timilar TWwen neeeuel Stroes '
Anel               Circumnerenhet3 lastes wee                                         ousmies was S                             S eo
                                                                                                                     <14nch                       s o                       o
                                         =
4                                                                                                     N                           l m-        e                               ,                                                     =           =           =          =
3                                                                                                                                             ;
10
                                                .e                              ea gp         ; j eo e              ,
0                                           g o-hp    '
7 d a wl.-                                                                                             oss 1 s.
10 30
                                            .==.==
                                            ==
                                                     **e*
aa . e 8
4 t
d*
W**:  e   as m
11 inch              See Note 3            o%~~+ + I I
ano        t   t      i      !   1    I   1    I   t                  ' S = 30 hae 0 at . Sa e,a a4 as OS af as os t.0                              2 Conesearable veneteen with woes heat MeuL a
no iiseene c,mse asesi qwe -                         e - ei .olt - s.es awo + s.se two* - c.4e rwe . 2.os (usal a
Fig.13 sleasassed estal wow seeideal sisees es e hsneglen of                               ''*                         I*
seelal detenee Wesegh Wie pipe well ter sentenille seeinises steel pipe     Fig.1s fleessnmonded estal and shoessnfosential seeishsel etsees wolds                                                                       alleestipuolene Ier sustentie steiniese steel pipe welde Journal of Pressure VesselTechnology                                                                                    AUGUST 1966, Vol.108I363 4

     .   ". '' piping systems. Few experimental data are availab8* for other da/dN                     = change in crack depth, a, per fatigue cycle, configurations. However, finite element calculauons have                            ' in./ cycle been carried out for a variety of weld geometries [28), in. C,n                    = material constants, n = 3.3, C = 2 x 10-l' ciuding a 12-in. pipe-to-sweepolet weld, a 12-in. pipe-to-                            (in./cyc!c)/(psiVin)"
reducer weld, and a 22-in. pipe to-valve weld. In all three cases S
                                                                                                       =R ratio correction factor = [1.0 - 0.5R3]-4              ,
the weld-induced axial stresses are compressive, and the stress R                  = K./K,,,                              .                   l distributien shown in Fig.15 is covervative for these E                            -environmental factor (equal 1.0, 2.0, and 10.0 geometries.                                                                            for air, PWR, and BWR environments, 3.2 Fatigue Enck Growth Rate. A description of the                                   respectively) fatigue crack growth rates for austenitic piping materials has AK                  = K,.a - Keiin. psi V(in), and been developed recently by a working group under the spon- Keim K aiu = rninimum and maximum values, respectively, of sorship of the Pressure Vessel Research Committee and Metals                          applied stress intensity factor Properties Council [32]. The recommended fatigue crack grc,wth rate is a modification of [321 and has the form                 3.3 Corrosion Crack Growth Rate. Horn and co-workers [4] and Bush and co-workers (331 have compiled sub-da/dN= C.E.S. ( AK)*                     (21)    critical IGSCC growth rate data. Some crack growth rate data where                                                                also are available from constant rate extension tests (CERT).
0 10     .
i CURVE A
                                         ~
P                                                                        l D/
                                                                                 /             o 10-4   -
a               CURVEB
                                         -                           0             g/                         e I
w E             -
                                                                /@A                 'a
                           .5
                                                               /e           >       c
* i            -
f/                   S E                               /
                                                          /
[ 10'S       -
y             , .., ., _ ,u A .f 1*.'"'.'!,;J..u
                                                                                                                                 .. n ,., n .
                            }
O
                                         .'         d/ Y               +
y                                          .
0 ..E1*." NIJ8.. uw n. .
                                                                                                                                        ,n.
y                     y                           .
O o ... . .              i,       u y       y                        o . - ., ,.        .u      .. u.,n. .
y o           -                                                          + . gt.       " '' u8 :,';*'   . .
E O                                                                        m .'2( % %:.'.Th e ,.=.c 10 4   -
a.34'O'l&".M'"'!.-i
                                                                                                            ; !'.,. :.'.'u ." "'=',s
                                                                                                     * . f;'O'                     ne ..
o . 18.P.f,tlF..)
                                                                                                              ..,        . t.(.u ,In .. ne l=.i ,
                                                                                                            ;!. r;<                             t e . ;.,. gir,.g<;; .u .. ie,     . =. u='scm p .,.,;. u==> , . ... .
                                          -                                                           x . .,.           .u      .. innen. .
jp., ,s. u='t # . .. .
i            i               i              i                    I                  I 10 70               10           20              30              40                 50                  60          70 Stress Intensity Factor, ksi Vin Fig.16 Recommended corrosion crack growth rates, for austenitic stainless steel; curve A-fumace sensitiaod material. O.2 ppm osygen; Curve 8-weed sensitized material. 0.2 ppm oxygen 3641Vol.108, AUGUST 1986                                                                                           Transa'ctions of the ASME

           *     [.      ,V                           .                                                                                                                                      1 r,..-
While the CERT data are helpful for ranking material growth fatigue and corresbn crack grewth rates, and weld residual
             ,         rate, they were not used by the Task Group because the CERT stress distributions are defined, strain rates are not representative of field applications.
The data presented in [4] are shown in Fig.16. These data A+                    _
ts were obtained from tests conducted in high purity water at constant load: some cyclic tests were run at a high load ratio               lhe work of the ASME Section XI Task Group on Piping (R = 0.9) to simulate minor temperature and pressure fluctus- Flaw Evaluation and other interested individuals is tions during operating. Variations in dissolved oxygen content acknowledged. Members of the Task Group are: W. H. Bam.
and material       **an'i=daa       also were included as test ford, J. M. Bloom, R. C. Cipolla, T. 'J. Griesbach, J. P.
parametaa                                                                Houstrup, R. E. Johnson, M. Kupmski, D. M. Norns (Chair.                                     j For any hen material heat, +U ruults M man), J. S. Panesar, P. C. Paris, S. Ranganath, W. L. Server,                                                               j indicate that crack growth rate lacreases with increasing                F. A. Simonen (Secretary), G. M. Wilkowski, K. K. Yoon, l
degree of <*aaid,.rian and dissolved oxygen content from 0.2             and S. Yukawa. Other important contributions to completing l
to 8 ppm, although this is not always the case [29, 34]. In addi- this report either in performing computations, or providans tion, variations in stress can have significant effect on stress         text material or critique of the manuscript are also corrosion crack growth rate. However, the overall effect of acknowledged. These contributors include R. M. Gamble, R.
these variables is often not discernable from heat to beat Klecker, T. U. Marston, H. S. Mehta, E. C. Rodabaugh, W.
J. Shack, and A. Zahoor. The individual research contribu-matenal variations, as can be seen from the data in Fig.16.                                                                                                           ;
0 ' "" '
Curves A and B together with the suggested throughwall                             ",',' gg",            , ist e residual stress distributions give conservative estiramen of                                                                             d the Pm. .                  i throughwall crack growth in austenitic piping when compard Vessel Research Committee, Subcommittee on Pipmg,
                         .h t      Id                                                             Pumps,        and    Valves. Part  of  the  research     reported  here was sponsored by the Electric Power Research Institute and a Boil-C       A        .      reco manded as the crack growth rate for 0.2 ppm oxygen environments and furnace sensitized ing Water Reactor Owner's Group formed to address in.
material. Curve B is recommended as the crack growth rate tergranular stmss c rr s a crac% @ SCC).
for 0.2 ppm oxygen environment and weld sensitized material; lower growth rates may be recomunended for solution an.                   Refmaces nealed material. Lower growth rates also may be appropriate                   i American Society of Ma a-             Ensiseers sauer and Prusure Yessel B       water environments when hydrogen water chemistry                    ,         $7,,,/,'                  ou Par       Uniw Sima co wim                    I A crack growth rate plateau at high K values is often ob.                  3 Norria. D. M Chairman Task Group for Pipe Flow Evaluatioe. lener to served for subcritical stress corrosion growth rate data [35]            x. aeron ASME. " Evaluation of Stainless Smel Hus w=W===*= (s) Pro-ar.d may result from crack branching or increased distance               '""d Code Copy fw Mosncation to twa.3640. 93 Ta+-mi aasis Docu.
ment. March 29.19s5.
from the environment as the crack depth increases. However.                  4 "The crowth and Stabinty of Strees Corremon Cracks in Larse chaman the data in Fig.16 are insufficient to confirm the existence and nWR Pipins." ed., R. M. Horn. Electric Poew Research lasisme. Report magnitude of a plateau. If additional data become available NP.2472 (Vol.1. Sanmary; Vol. 2. Appendices). July 19s2.                                                  ,
that accurately show the existence of a plateau, they may be ment             s Ransanath. 5, and Mehta. H. 5, "Ensianrms Mahods for the Asum.                        j of Ductile Fracture Marsin in Nuclear Power Plant Pipins." asste used in corrosion crack growth rate analyses.                            pg ,,e 7,,,,,e secomf symposium, vol. 2. " Fracture R-==am Curva and Ensineerins Appncations." ASTM STP s03. American Society for Testing and                     i Materials. Philadelphia, Pa.1983.                                                            l 4 Conclusions and Recommendatloas                                             G Kanninen. M. F et al "Towari an Elastic Plastic Fracture Mechanics Predictive Capability for Reactor Piping." Nuciese Engbssering sad Desirs, 1 Allowable flaw sizes have been defined for plastic col-             Vol. 4s.197s. pp.117-t34.
lapse and ductile tearing mM=?=.: In austenitic steel piping,                 7 Eiber, R. J., et al.. " Investigation of the taalariaa and Extent of Ductile auteue Columbus Laborascries. Report aMI.1908 June 1971.
ang are present,g in an easy to use ttbular format.                       Pipe, ,,,,,,
R,apture.",h. S., Mehta. H. S., and Norris. D. M., " Structural Eval 2 Allowable flaw sizes have been determined for of Flaws in Power Plant Pipins." circuai/ersarief Cracks 6: Prs =nne vassair longitudinal and circumferential flaws, and normal / upset and sad Pipias. Vol.1. ASME.19s4. p. 91.
emetgency/ faulted conditions. The safety margins employed                    9 raaaia= M. F, a al, "!amabuity Pr=Ma== for Circumferensinuy in defining allowable flaw size are consistent with those nor- Cracked Type.304 Staisises Under Dymanue Loadins." Electric Power Reaperch Institute Report NP-2347. (Vol.1. Susunary, Vol. 2. App =adw=) Apr. Os2.
taally required by the ASME Code.                                           to Wukowski. G. M., as al. Desraded Piping Program Phase II,9a='aa-at 3 Tables for the plastic collapse enschanisin are to be used Report ost.19s4-Mer.19ss. NUREG/CR.40s2. aMI 2120. Vol. 2. July 1985.
for wrought base matenals and cast austenitic steels meeting                !! G**en. M. D and Emmar. V.," Elastic.Plamic Analysis of Crack Open.
the requirement of IWB-3641(C), and gas tungsten arc and gas             ins., 3,,,,,
i,,,    Stable ,7   Growth, p,,,,,and   I-hmy3,aehavior y,,,,,,,                la Flawed,,304,Sr.ainless p,,,,,, y,,,     ,, pg,,,,, pyp,y,i,    Staat Pip metal arc welds.                                                         33. Amwican Society of Whaa-a' Engineers. June 19s2.
4 Tables for the ductile tearing man haniarn are to be used              12 Ptpe Crece study Gem,p. im. "inv==ivaniaa and Evniuanos of Siress for shielded metal vc and submerged arc welds.                           Cwroman Cracking in Pisiing of Light Waa- Rasctor Plants." NUREG.0531.
5 The plastic collapse tables we s generated using limit- U.S.               Nuclear Resulatory e==i-iant. Washington. D.C Feb. IM9.
13 Tada. H Paris P. C and Gamble. R., " Stability Analyas of Cir.
load analysis and an assumed flow stress equal 3S.. Ex* confeensial Cracks la Raamor Piping Symens." NUREG/CR 083s. U.S.
perimental data and previously performed analyses show that Nuclear Resuasiwy e==- W=haaP- D.C June 1979.
this procadure is reasonable for circumferentially crac.ked Cracking        34 P" PW' C'"c* 3'*d7 0'""*. isso. "lamaission and Evaluanon of i
lacidents in Piping in Pressuriand Waner Reactors." NUREG 0691, austenitic piping. Although limited experimental data indicate U $. Nuclear Ragubsory Ceaunission. Washington. D.C Sept.19s0.
the assumed flow stress, 35., is 20 percent larger than that in-            is Paris, P. C Tada. H., and Macek. R., " Fracture Proof Design and dicated by test results for longitudinal flaws,3S. was used Annarsis of Nuclear Pipias." NUREG/CR MS, U.S. Nuclear Regulatory Sept 8 consistently because compensating conservatisms were employed in defining allowable longitudmal flaw sizes.
                                                                                                ]                g H',' Chang. H. Y., and Zahoor. A "Appiacanon of Tearing 6 The ductile tearing tables were generated using conser-             Moduim        Stabiniy Concepu to Nuclear Pipias.".EPRI NP-2261. Electric Poww Rosarra samkuie. Paio Aho. Carif.. Feb.19s2 vative assumptions for load distribution, pipe size, and                    17 Zahoor. A., and Gambia. R., "Laak nefore areak Aaaiyus for nWR material tensile propertnes; matenal toughness was represented Amarculauca Pisiing Hams Cracks a Mukisite Weld I ama== " EPRI NP.3322.LD, Electric Power Amearch Instituse. Palo Ake. CaEf., Apr.19sa.
by average values front available expenmental data.                         Is Zabsor. A., and Gesable. R ** Proposed Esaluasion Procedure for 7 To support the subcritical crack growth evaluation pro. Aumeenic $smains Steel Flus Weeds." mad to the Task Group on Pipias cedures in Appendix C of Section XI, ,w==M= tons for Flaw Evaluation. Palm Springs. Cauf., Feb.11.19s5.                                                     ,
Joufnalof PressureVesselTechnology                                                                                  AUGUST 1986,Vol.108f 365

m._      __m       ._            _  ,mmm                         ,
e,
           .l.  ,., ,-e
              ' 4tumar. Y Gennan M. D., Wilkening. W. W., Andrews. W. R.,                         Pressure Vessels." NUREG4376. U.S. Nuclear Regulatory Commiadon.
      * *eeLoreezi. H. G., and Mowbray. D. F., " Advances in Dasde. Plastic Fracture              Washington. D.C 1977.
                  ~
  . ,         Analysis." EPRI NP.3607. Dectric Power Research !nstitute. Pelo Ahe. Calif             24 Rybicki. E. P et al."W-+ ' Residua! Stress Analysts for induc.
      .   . Aug.1984.                                                                             tion Heating of Weided BWR Pipes." EPRI NP.2662.LD. Dectric Power 20 Paris. P. C Tada. H. Zahoor, and Ernst. H "A Tresunent of the Sub.             Research Insdtute. Palo Aho. Calif.,1982.
jocs of Tearing Instability." NUREG.0311. U.S. Nuclear Regulatory Comnus-               29 Shack. W. J et aL. " Environmentally Assisted Cracking in Light Water sion. Washington D.C Aug.1977.                                                      Reactors." Annual Report Oct.1983-Sept.1984. NUREG/CR.4237. Argonne 21 Hutchinson, J. W.. and Paris, P. C
* Stability Analyns of J. Controlled Nadonal Laboratory.1983.
Crack Growth." in EJestk. Plastic Fiscrure. eds.. J. D. LAndes, et al. ASTM            30 Hams. D. O "The influence of Crack Crowth Kinedes and inspecuan STP 668.1979, pp. 37-64.                                                            on the intregity of SeaN BWR Piping Welds." EPRI.NP.ll63. Dessric 22 Paris. P. C., and Johnson R. E., "A Method of Applicadon of Dasde Power Research lastitute. Palo Aho. Calif 1979.
Plasde Practure Mdaain to Nuclear Vessel Analysts " Elastk Plastr Fruc*                31 huh R et al. "Matigadon ofInside Surface Residual Stress of Type rure: Second Symposium ATTM STP 803, Vol 11. Philadelphia, Pa 1983,                 304 Stainless Steel Pipe Welds by Innde Water Cooling Method." Proceedurgs:
pp. 5-14.                                                                           Seminar on Countermaarwesfor Ptpe Crocking in BWRs. EPRI WS.79174, 23 Landes. J. D., and McCabe. D. E., " Toughness of Austeniuc Stainless           Vol i Dectric Power Research Insdrute. Palo Aho. Calif 1980.
Steel Pips Weldments." EPRI Project RP 1238 2. EPRI Report (in press), May             32 James. L. A and Jones. D. P " Predictive Capabilities in Environmen.
1986.                                                                               taDy Assisted Cracking." Special Publication, PVP.Vol 99. Amencan Society 24 Shack. W. J., E!!ington. W. A., and Pahis L., " Measurement of of Mechamcal Engineers. Nov.1983.
Residual $4ress in Type 304 hiale== Steel Piping Burt Weidments." EPRI                 33 Pipeng Reikw Commirrer. "lovestigation and Evaluadon of Stress Cor.
NP.1413. Dectric Power Research Insdtute. Palo Aho. Ca!!f.,1980.                    rosion Cracking in Piping of Boiling Water Rector Plants." NUREG 1060, Vol.
23 Shack. W. J., " Measurement of Throughwall Residual Stresses in Large          1. U.S. Nuclear Regulatory Condssion. Washington. D.C 1984.
Diameter Type 304 omiatea Steel Piping Butt Weidments." ANL.82-15.                     34 Shack. W. J et al. "EnvironmentaDy Assisted Cracking in Light Water Argonne National Laboratory.                                                         Reactors" Annual Report Oct.1982-Sept.1983. NUREG/CR4287. Argonne 26 Shack. W. J., et aL. "EnvironmentaDy Assisted Cracking in Light Water          National Laboratory.1984.
Reactors" Annual Report. NUREG/CR.3292. Argonne Nadonal Laboratory.                    35 Ford. F. P "Mechamsms of Environmental Cracking in Systems Oct.1981-Sept.1982.                                                                  Peculiar to the Power Generation Industry." EPRI.NP.2589. Electric Power 27 Rybicki. E. P.. et aL. " Residual Stresses at Girth.Buu Welds in Pipes and     Research insotute. Palo Aho. Calif.,1982.
h 9
i i
i a
t I
i l
l I
i 3661Vol.108, AUGUST 1986                                                                                                        Transactions of the ASME}}