ML20205G724

Toughness Criteria for Thermally Aged Cast Stainless Steel

ML20205G724
Person / Time
Site:	Prairie Island
Issue date:	09/30/1985
From:	Chirigos J, Kim C, Witt F WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
To:
Shared Package
ML20205G675	List: ML20205G668 ML20205G710 ML20205G724
References
WCAP-10930, NUDOCS 8511130324
Download: ML20205G724 (73)

Text

.

Director of NRR~

Oct ber 21, 1985 WCAP 10930 TOUGHNESS CRITERIA FOR THERMALLY AGED CAST STAINLESS STEEL F. J. Witt C. C. Kim September 1985 7 O IN N APPROVED:

APPROVED:

J),s. Chirigos, Minager T. A. Meyer, Manager Structural Materials Structural Materials And Engineering Reliability Technology WESTINGHOUSE ELECTRIC CORPORATION NUCLEAR ENERGY SYSTEMS P.O. Box 355 Pittsburgh, Pennsylvania 15230 8511130324 851021g2 DR ADOCK O

FOREWORD This document contains Westinghouse Electric Corporation proprietary information and data which has been identified by brackets. Coding associated with the brackets set forth the basis on which the informaton is considered proprietary. These codes are listed with their meanings in WCAP-7211.

The proprietary information and data contained in this report were obtained at considerable Westinghouse expense and its release could seriously affect our competitive position. This information is to be withheld from public disclosure in accordance with the Rules of Practice 10 CFR 2.903.

,, Withholding of this information does not adversely affect the public interest.

This information has been provided for your internal use only and, should n'ot be released to persons or organizations outs'ide the Directorate of Regulation and the ACRS without the express written approval of Westinghouse Electric Corporation. Should it become necessary to release this information to such persons as part of the review procedure, please contact Westinghouse Electric Corporation, which will make the necessary arrangements required to protect the Corporations' proprietary interests.

The proprietary information is deleted in this unclassified version.

l i

5 l

ggg

~

TA8LE OF CONTENTS Section Title Page

~

1.0 INTR 000CTIOM'AND 8ACKGROUNO 1-1 2.0 A GENERAL TOUGHNESS CRITERIA FOR THE THERMALLY AGED CAST STAINLESS STEEL - [

]a,c.e.

2-1 2.1 Introduction 2-1 2.2 Metallurgy of Thermally Aged Cast Stainless Steel 2-1 2.3 General. Toughness Behavior of Thermally Aged Cast Stainless Steel 2-2 2.4 Energy Oisposition in Impact Specimen Testing 2-5

}b'# Criteria - CIRCA 1984 2-11 2.5 Summary of the [

2.6 Additional ('

]"' ' # Data 2-15 2.7 The KCU for [

]a,c.e 2-35 2.8 The Alternate [

] E # Toughness Criteria 2-36 2.9 Experimental Verification of the Alternate Toughness Criteria 2-46 2.10 Sample Application of the Alternate 2-49 Toughness Criteria 3.0 DISCUSSION AND CONCLUSIONS 3-1

4.0 REFERENCES

4-1 l

l l

V l

LIST OF FIGURES fl1EC1 Title Pgge 2-1 Full impact curves for 6.4 mm (0.25 in.) Allegheny Ludlum-melted E-8 RITE alloy laboratory-annealed and air-cooled or water-quenched 2-3 2-2 Charpy V-Notch Energy Curve for Various Aging Times -

Four Inch Pipe Material (The aging temperature was 800*F (427'C))

2-4 2-3 Effect of Aging Time on Charpy Impact Results 2-6 2-4 DVM impact transition curves showing the effect of longtime aging at 280*C 2-7 2-Sa Fracturing Process of the Duplex Material 2-9 2-Sb Ligaments of Duplex Structure Exhibiting Ductile Behavior 2-9 2-Sc Sketch of Fracture Surface of Austenitic - Ferritic 4

Impact Specimen with completely embrittled ferrite showing a tortuous fracture path for a beam 2-9 i

2-6a Dimensions on Swiss VSM and German DVM keyhole

' specimens 2-12 2-6b Charpy'(Simple-Beam) Impact Test Specimens, Types A, 8 and C (ASTM E-23) 2-12 2-7 Charpy U-Notch Energy at Room Temperature for [

Ja.c.e 2-14 2-8 Charpy V-Notch Energy at Room Temperature and 550*F for (

2-16 l

Ja c.e 2-9 J-integral.Results at 600*F for [

]a,c.e, j

8 = 0.34 in. (Based on Deformation J Evaluation) 2-17 2-10 J-integral Results at 200*F for [

]a.c.e, B = 0.34 in. (Based on Defonnation J Evaluation) 2-18 2-11 J-&a Curves at Different Temperatures for Aged Material

[

]a.c.e 2-19 2-12 Full Size Plan View of Material Used in Investigation, Thickness was 0.772 in. (20 mm) 2-20 2-13 Load Time Trace for Charpy V-Notch Impact Specimen 78A of (

Ja.c.e Material Tested at. Room Temperature 2-21 vii

- ~

LIST OF FIGURES f.1SE1 111.11.

E191 2-14 Load-01splacement Curves for the Five 0.394 in.

Thick Compact Specimens Tested at Temperatures 8etween -200*F and 550*F (Note differences in scales and crack ratios) 2-23 2-15 True-Stress True-Strain Curves at Room Temperature and 600*F for [ a.c.e 3

2-26 2-16 Load-01splacement Curves for 0.34 in. Thick Specimens of

[

]a.c.e Material Tested Statically 8etween

-320*F and 600*F using Unioading Compliance 2-27 2-17 J-Integral Results at -320*F for [

Ja.c.e (8ased,on Deformation J Evaluation) 2-28 2-18 J-Integral Results at Room Temperature (75'F) for

[

Ja.c.e (Based on Deformation J Evaluation) 2-29 2-19 Comparison of Static Load-Displacement Curve with Dyna-nic Load-Displacement Curves at 600*F ([

])a,c.e 2-30 2-20 Jge as a function of Test Temperature for [

]a,c.e 2-33 2-21 Variation of Tmat with Temperature for (

3a,c.e 2-34 2-22 Variation of J c with Charpy V-Notch Impact Energy 2-38 I

2-23 Micrographic features of two-phase specimens as a-function of chemical composition Ni'eq/Cr'eq and thickness. The ferrite is colored by electrolytic etching in a 10-N sodium hydroxide solution. dp =

average segment intercepted by ferrite (um); 4% =

ferrite content measured by micrography 2-39 2-24 Correspondence Between the KCV Energy and KCU Energy of Austenitic Stainless Steels tested at 20*C 2-44 Plots of J c. Tmat and J nr at Operating Temperature 2-25 i

as a Function of (

Ja.c.e 2-45 2-26 Typical Layout of the Primary Loops for a Westinghouse Two-Loop Plant 2-52 2-27 Identification of Hests With Location For Cold Leg 2-53 viii

LIST OF FIGURES (Cont)

Fiaure Title fiS1 2-28 Identification of Heats With Location for Hot leg 2-54 2-29 Identification of Heats With Location for Crossover leg 2-55 2-30 Fracture Toughness Criteria for Heat X

([

])a.c.e of Sample Application 2-58 4

4 ix m - ----,

w-,---

-,.,_.,_,,_.__n.,

yn,,

- - - - - -~., -- -

--,-n,.,,

LIST OF TA8LES IAlLl1 111.1.1 f.121

~

2-1 Materials Specifications, Chemistry and Material Properties of [

]a.c.e 2-13 2-2 Static Fracture Toughness Properties for [

Ja.c.e 2-24 2-3 Charpy Impact Energies from Standard and Non-Standard Specimens at Different Temperatures for

[

Ja.c.e 2-31 2-4

.I e, Tmat and.lmax at Operating Temperatures for I

which the [

Ja.c.e is known 2-42 2-5 Example Chemistry and calculated [

la.c.e Values for Pipe and Fittings 2-50 2-6 Fracture Toughness Criteria for the Primary Piping Components of the Example Two-Loop Plant 2-57 i

l i

s xi

--_.,-_.-m-

,_.,___.-,.--____,--,.-.,__,._-__-y,

l I

1.0 INTRODUCTION

AND BACKGROUNO It has been demonstrated by various analyses conducted over the past several years that piping used in the primary systems of nuclear power plants will in general leak stably should large through-wall penetrations occur for the loads imposed (1.1,1.2,1.3)*

This demonstration has prompted both the nuclear power utilities and the Nuclear Regulatory Commission (NRC) to take advantage of this situation in the as-built' design and construction of primary piping systems. Specifically, many of the utilities have prepared integrity evaluations on a plant specific basis which demonstrate that leak-before-break conditions exist for reasonably sized cracks when subjected to the applicable loadings. For Westinghouse type pressurized water reactors, the utilities have, in the main, subcontracted to Westinghouse to develop the integrity evaluations. The evaluations have been used by the utilities as the technical basis for requesting exemptions from the NRC.for installing plant specific pipe whip restraints and jet shields. The NRC has responded very favorably to the requests having granted to date six exemptions. Also they have initiated an interim rule making session to revise General Design Criteria-4 to permit the use of advanced fracture mechanics analyses in lieu of postulated reactor coolant system main loop breaksI*I.

The integrity evaluations rest seinly 'on the application of elastic-plastic fracture mechanics and leak calculation methodologiesI*I. One of the primary inputs to an evaluation is the elastic-plastic fracture criteria with which the calculated anplied stress intensity factors may be compared.

In general, the J-integral approach has been applied with the criteria II*II being l

Superscripts refer to similarly numbered items in Section 4.0, References.

1 -1

1._ J,,, < Jgg, or (1-1)

(1-2) j 2.

If J,,, 1 Jgg, then T,,, < Test where T is the tearing modulus and subscript app designates est applied.

The NRC has addended these criteria by(1.5) 3.

Sufficient margin must be demonstrated (1-3) 4.

J,,, < J,,,

where J,, does not exceed the maximum value of J determined from fracture toughness specimens.

Fracture toughness J properties are in general not available for all cases for nuclear primary piping; however, Westinghouse has developed a bounding approach which has been favorably received by the NRCI*I. The approach generally divides the piping system into two categories: (1) forged and (2) cast piping and welds. The analyses focus on the cast piping and fittings and welds since toughness degradation due to thennal aging is known to occur for these two product fones, the cast material being limitingI1' ' I* I.

A correlation has been developed based on the chemistry of the cast material

(

].

Fortunately, a significant amount of fracture toughness data has been generated on a highly sensitive heat of cast stainless steel pipe identified as [

]***.

The KCU impact value was estimated to be [

].

Thus, any heat of pipe estimated to exhibit more than [

3.c.e a

JIc " E I

II-4)

T,,g = [

]' d '

(1-5)

J,,, = [

]W (1-6) 1-2 l

i

~

Of all the nearly 25 primary systems examined through the end of 1984, the applied loads were such that the above criteria were easily met with substantial margin being demonstrated. However, the correlation based on

' chemistry does not produce the required [

]" energy for a limited number of heats of all the primary piping for some of the plants.

In such a case where the [

j.c.e energy is not a

demonstrated, there is at present no explicit data base from which component I-specific criteria may be obtained. Hence, the adequacy of calculated applied fracture toughness values cannot currently be quantitatively assessed using the above approach.

It is the objective of this report to develop alternate fracture toughness criteria for primary piping system components for any calculated [

]"

energy such that leak-before-break integrity evaluations may be perforined.

Specifically, general toughness criteria for thermally aged cast stainless steel are developed for calculated [

]' d impact values less than [

y.c.e a

I l

j e

1-3 n

..s-,-.

--e,n--,,-.-----,-,,n..,-w--,,

.,--.-----.,,--.,-,-.,,,,_.,-.,..,--,,,,._,,-,-,,,.._,y-

2.0 A GENERAL TOUGHNESS CRITERIA FOR THERMALLY AGED CAST STAINLESS STEEL -

g j.c.e a

l 2.1' INTRODUCTION The general background for this report has been discussed in the previous 4

section. In this section alternate fracturt toughness criteria are set forth. First the metallurgy of themally aged cast stainless steel is reviewed. A general summary of the toughness behav'ior of thermally aged stainless steel is provided. Impact test results are also provided and the nature of such results are explained. A suemary of the [

]"'# # data is given supplemented by new results. Implications of impact energy to fractura toughness (.1, K. T) are set forth. Finally alternate [

]"'C

l-fracture criteria are developed and supportive experimental data are set forth. A sample application is given.

The discussion of this chapter is strongly referenced to References 2.1, 2.2 4

and 2.3.

2.2 METALLURGY OF THERMALLY AGEO CAST STAINLESS STEEL t

CF8A and CF8M type cast stainless steels in the primary loops of many Westinghouse type pressurized water reactors are two-phase alloys consisting of austenite and ferrite. The difference between CF8A and CFBM is that CF8M has molybdenum and CF8A does not. The ferrite content very* seldom exceeds 25%

and the ferritic phase of these alloys has a c'homical composition very similar to pure ferritic low carbon stainless steel (2.1)

It has been found that the chrome enriched ferrite of the two-phase alloys becomes hardened and embrittled when thermally aged at temperatures up to approximately $20*C (968'F). This embrittlement is caused mainly by the successive precipitation of chromium in ferrite due to the large miscibility l

gap in the Fe-Cr binary system. CF8A and CF8M type cast stainless steel in the primary loop system can experience the embrittlement after long time i

service because they are operated at temperatures generally from 550*F (288'C) 1 to 600*F (316*C). The nature of this thermal aging has been extensively explained in a series of articles (2.1 through 2.7)

These investigations were carried out on both pure ferritic alloys and ferritic-austenitic cast stainless steels.

2-1

--~w ev--,..,

c w r---,

,-,,,y,.- - -, - _ -,, _, _ _ -.. _.. _ _

)

References 2.4 and 2.7 investigated pure ferritic alloys and observed embrittlement effects on the alloys when thermally aged. The embrittlement was also observed in the ferrite phase of ferritic-austenitic duplex structure I

cast stainless steels by References 2.1 and 2.2.

Reference 2.2 concluded that I

the austenite matrix of the duplex structure cast stainless steels remains ductile while the ferrite phase shows cleavage rupture when the fractured surface of thermally aged Charpy impact specimens were examined. Based on these investigations it can be concluded that the ferrite phase of ferritic-austenitic cast stainless steels becomes hardened and embrittled when thermally aged and behaves more or less as a ferritic stainless steel.

2.3 GENERAL TOUGHNESS BEHAVIOR OF THERMALLY AGED CAST STAINLESS STEEL The fracture morphology of thermally aged cast stainless steel indicates that the hardened and embrittled ferrite behaves as a more or less ferritic stainless steel while the austenitic matrix is nonembrittling and continues to exhibit ductile behavior (2.1, 2.2)

On this basis alone, a quantitative explanation should be forthcoming for the observed fracture toughness behavior of thermally aged cast stainless steel.

i Ferritic stainless steels have been extensively investigated and an ASTM special technical publication has been devoted to the toughness of ferritic stainless steels (2.8)

Examples of Charpy impact curves taken from Reference 2.Ba are given in Figure 2-1.

It is seen that ferritic stainless steels exhibit a ductile-brittle transition very similar,to low carbon steels. It is expected then that the ferritic phase of a thernally aged cast stainless steel is the primary contributor to the loss of ductility as this or T results phase becomes embrittled. Interestingly, no K7g, Jgg mat are presented Reference 2.8.

An example of loss of ductility due to thermal aging embrittlement and test temperature is given in Figure 2-2 taken from Reference 2.9.

The material is SA351 CFSM (Type 316). The aging temperature was 800*F (427'C). For the embrittlement time of 2000 hours0.0231 days <br />0.556 hours <br />0.00331 weeks <br />7.61e-4 months <br /> the upper Charpy V-notch energy is seen to drop by 7.round 80 f t-lb down to approximately 100 ft-lb. The transition temperature is seen to increase by about 150*F. The major effect is seen to l

occur within the first 500 hours0.00579 days <br />0.139 hours <br />8.267196e-4 weeks <br />1.9025e-4 months <br /> with perhaps a gradual lowering of the shelf i

s 2-2 i

4

.,e-

,.-,_-,,,vr

,_n,,.,e=_

_._.m,ve-.,,,,,--,,,_,.,_------__,,----,-----m-,-w--,-,,------

TEMPERATURE PF)

-10 0

-50 0

50 10 0 15 0 200

• WATER QUENCHED

• AIR COOLED 2 71 200 1

203 15 0

+

(Y 8

9

(

U 1

b 100 $

13 6 -

68 50

~2 ft-1bs.

j 3

1 i

-73

-46

-18 10 38 65 93 TEMPERATURE PC )

Figure 2-1 Full impact curves for 6.4 m (0.25 in.) Allegheny Ludlum-melted E-BRITE alloy laboratory-annealed and air-cooled or water-quenched 2-3

TEMPERATURE (*C1 144 129 73.3 17.7 37.7 S3.3 149 2:4 200 316 371 3:1 i

i i

i l

l l

280 o AS RECEIVED e 1200 HRS.

354 a 1000 HRS.

A 1500 HRS.

240

--- 327 g j 220 C 500 HRS.

E 2000 HRS.

m A

l h 200 272

_m

$ 180 Y

245 b a

a E 100 i

217.7 E w

w g 140 190.5 3 E 120 1s3.35 n-U a0 (

[

--- y

^j-

. =

> 100 s-13s.1 >

s 10s.s$

g so 31.s 5 40 54.4 m

2n 27.2 I

I I

I I

I i

300 200 100 0

100 200 300 400 500 400 700 TEMPERATURE (*F)

Figure i.-2: Charpy V-Notch Energy Curve for Various Aging Times-Four Inch Pipe Material (The aging temperature was 800*F(427'C)).

r l

l P

l 2-4 i

with aging time thereafter. The transition behavior with aging time is not well defined in Figure 2-2.

An example of the expected transition behavior is shown in Figure 2-3 taken from References 2.9.

An example of the generic benavior was calculated in Reference 2.1 and is shown in Figure 2-4 It is thus concluded that the change in the Charpy V-notch (also U-notch) impact curves with thermal aging is a direct function of the ferrite hardening and embrittling. Indeed the unbrittled ef fect on the ferrite is judged to be

^

very similar to the ef fect of lower temperature embrittlement of ferritic stainless steel. Noting as in Figures 2-2 and 2-3 that the major embrittling ef fects occur in a relatively short time and appear to systematically saturate over such longer periods, thermal aging emerittlement of the ferrite-pnase is entire.ly consistent with' the rapid drop of Charpy energy in the transition temcerature region of ferritic stainless steel followed by a much more gradual decrease in the lower shelf region (see Figure 2-1).

In Reference 2.1, 2.2, and 2.3, the austenitic matrix of the duplex structured cast stainless steel is seen to remain ductile and nonemarittled. This austenitic matrix makes up from 75% to 85% by volume of the nuclear grade cast piping of interest. Austenitic stainless steel with little or no ferrite is very tough in terms of Charpy energy. For example, in Ref erence 2.1 Heat 291 contained 6% ferrite and exhibited over 150 f c-lbs (200J) impact energy. The point here is that an embrittlement of roughly 20% of the thermally aged cast stainless steel produces very significant changes in the impact energy while around 80% of the material (the austenitic matrix) remains unemerittled and very ductile. Indeed the austenitic matrix must be a major contributor to Charpy values of around 150 f t-lbs in the unaged state; yet, it is unable to exhibit such toughness in the thermally aged state. This apparent anomaly is discussed in the next section.

2.4 ENERGY DISPOSITION IN IMPACT SPECIMEN TESTING In this section the above anomaly is discussed.

For the unaged cast stainless steels at operating temperatures of interest the Charpy V-notch toughness in quite high, usually greater than 150 ft-lbs, of ten challenging the Charpy machine capacity of -250 f t-lbs. Ferritic stainless l

2-5 i

o g 180

f LEGEND: TESTS DONE AT ROOM E

160 TEMPERATURE l

g O ENERGY-AXIAL ORIENTATION y 140 7 LATERAL EXPANSION AXIAL

(

ORIENTATION Xw 120 C ENERGY-CIRCUMFERENTIAL ORIENTATION

$ 100 <

a LATERAL EXPANSION CIRCUMFERENTIAL ORIENTATION 4

I so 3

e0 a

%,A g

do_

i l

l

=

m 0 100 500 1000 2000 3000 4000 TIME (HOURS) AT 300'@(427*C)

Figure 2-3: Effect of Aging Time on Charpy Impact Results 2-6

E".

MERT 250ft k

4 9

g 5

5

&r, -,

$ /,f g

w.

~

/

q#

L-4 8 sp-

@ 2 ;-

p 4

g 'I a.

/

S 3

P g

AfAWg.

-33

-100 O

100 84 300 400 TEST TEMPERATURE DEG.C Figure 2-4 DVM impact transition curves showing the effect of longtime aging at 280*C 2-7

steels exhibit about the same Charpy toughness. Thus it is reasonable to assefoe that the contributions to the total energy by the austenitic matrix and the ferrite are approximately in proportion to the percent volumes of each components, perhaps 80% for the austenitic matrix and 20% for the ferrite.

However, with aging the ferrite is embrittled and the impact energy may dramatically decrease. The embrittled fracture behavior is depicted in Figure i

2-5a taken from Reference 2.2.

In Figure 2-5b, sketched to the right, the width (i.e., line of fracture) of the cross section consisting of the austenitic matrix is seen to be about 60% of the total cross sectional width.

In general one might expect the austenitic matrix length to be around perhaps 80% of the total cross sectional width. Naturally, one would expect some j

reduction in Charpy energy if one assumes the ferrite makes little or no contribution to the fracture resistance. However, one might be surprised to observe around an order of magnitude reduction of Charpy energy when only 20%

4 of the material has been embrittled. This behavior is explained below.

Some years ago a rather intensive study was made of the energy disposition for various size impact specimens (2.W A model was developed and verified which accounts for both the specimen width and depth when an impact test is made.

For the problem at hand, the ferrite is assumed to make no contribution to the measured energy and the austenitic matrix is assumed to be fully ductile.

Under this assumption the basic equation for a single beam from Reference 2.10 becomes:

Et = A j 8 dW

( 2-1 )

where E is the total fracture, energy, A is a material constant, W is the t

length of the fracture surf ace and 8 is the width of the beam and is a function of W.

i For computational purposes. Equation (2-1) may be expressed in the sununation form I

k 2

Et=AI 8 W (2-2) g 1.,

2-8

w f

efmp xN

..:........-....-.e.

arh A

P, l(1) y 3(2)

.(3) s j

l (4) s

/

Figure 2-Sa: Fracturing Process of the Figure 2-Sb: Ligaments of Duplex Structure Duplex Material Exhibiting Ductile Behavior Direction of Loading y

• Austenite f

l 9-rerri te p* M 9 1 C

D :for tuous i

Beam fracture I

[

p Surface iPm_

.e_.

)

e l

Figure 2-Sc Sketch of Fracture Surface of Austenitic - Ferritic Impact Specimen with completely embrittled ferrite showing a tortuous fracture path for a beam 2-9 i

_._ _. _ _ _ _ _ _ ~. _ _ _. _.. _ _

where k is the number of increments into which W is divided. For a set of individual beams of the same material, the total energy, E, is given by T

a k

2 E7=A I I S W

(2-3) gj gj j-i 1-1 where a is the number of individual strands (beams) of the austenitic matrix (e.g., in Figure 2-gh, e is 4 at the location observed).

As an example, in Figure 2-5b, the expected impact energy for the four ligaments of the austenitic matrix, assuming they are distinct beams, is only about 25% of.that had the ligament been as one. Of course the austenite is certainly not expected to be aligned into separate ~ straight beams. However, a fracture surface in which there are originally voids (in this case the embrittled ferrite) would tear as beams with varying thicknesses with the cross constraints defined by the voids detemining the actual distinct lines of material separation; hence the fracture surface is indeed representable by a series of beams. An example of the more or less random dispersion model for the ferrite is shown in Figure 2-5c.

This cross section has separated as a series of beams of varying depths and widths; thus Equation (2-3) could be used to estimate the impact energy to fracture the specimen if the exact paths were known. An example of a fracture path for a beac is also sketchui in Figure 2-Sc.

For later discussion, such fracture paths will be called-tortuous fracture paths and the model representing the total fracture energy as given by' Equation (2-3) is designated the tortuous beam model. For general applications'the tortuous fracture paths can only be determined by microscopic j.

examination of the fracture surface and could be highly subjective. This is not a problem for the applications made later of the tortuous beam model.

I Since unaged Charpy V-notch energies of at least 100 ft-lbs are expected and I

energies of around 20 ft-Ibs have been measured on thermally aged stainless steel (2.2), one might inquire somewhat of the disposition of ligaments in the austenitic matrix to produce an energy perhaps representative of fully aged stainless steel. An energy of 10 f t-lbs will be used. The austenitic entrix is taken as 80% of the material. Equal width straight beams are j

assumed since nothing is known of the tortuous fracture paths. solving Equation 2-3 by trial and error gives 7 ligaments each of width 0.045 in. to produce an energy of 10 ft-Ib, assuming the unaged impact energy is 100 f t-lb.

2-10

}

1

P.,

The objective of this section was to demonstrate that the low Charpy energies measuredfromhighly(thermallyagedcaststainlesssteelwithaduplex structure of austenitic'And ferrite is consistent with the metallurgically observed ductiliaust$nitikmetrix with embrittled ferrite accounting for

^

y about 1/5 of the material by weight. A complementary objective is to define an energy model for impact testing. It is judged that this objective has been successfully accomplished.

d

, At this time one may inquire if static results would be the same as those ob' served for impact results. This question is answered in the affirmative in a liter section.

2.5 $UMMARY OF THE ['

]' ' ' # CRITERIA - CIRCA 1984 Reference 2.3 provides much of the available information upon which the

[

]"' criteria previously discussed are based. In this section the data is_ again presented for completeness and to provide background for the new data presented in the next section.

First however, since the correlation of Reference 2.2 relates to Charpy U-notch (KCU) data, a more specific definition of the specimens tested is needed. Figure 2-6a, taken from Reference 2.1, shows the two types of specimens tested in Reference 2.1.

The standard specimens of ASTM I*

E-23 are shown in Figure 2-6b.

In this figure the Type A specimen is the standard Charpy V-notch (KCV) specimen. Note that the specimens of Figure 2-6a have a different flaw depth than the Type C specimen of Figure 2-6b'.

The KCU data presented in Reference 2.2 are from the Type C specimen of Figure 2-6b.

In Reference 2.1, results from the V5M and DVM specimens are used interchangeably even though the notches have different depths.

[

. ]' ' " # The material specification, chemistry and material properties are given in Table 2-1.

The material corresponds to ASME SA351-CFSM. Room temperature Charpy U-notch data (Type C specimen of ASTM E-23) are provided in Figure 2-7 as a function of aging time. Charpy V-notch 2-11

KEYH01.ESPECIMEN[m) 2

~4 s

s.

GERMAN DVM 8, kR t

I te l is

y SWISS VSM I'

8, :M i

l isl ss

\\

Figure'2-6a Dimensions on Swiss VSM and German DVM keyhole specimens i

\\

.5 I

iii PL/2 H

-2== (o. ort")

o.2s==

5 a

($2;s*) Y4s*Y TtI.!a's")

(O"".)

L TYPE A s==(o.197')

l H LA-=

l --lhg,;,, 1t

,=q

=

e C!.'iIs") ~

g3

.)

1.e== 0/16*)

==

sawcut L

TYPE B "L"

l-- LA H s =m p.197*)

0

• M L

ss --

i s _p,.o_i,)

(2'*5') _-

+(o,o..

.s*41 7y,,e Figure 2-6b Charpy (Simple-Beam) Impact Test Specimens, Types A, B and C (ASTM E-23) 2-12

)

' 's e

1 h~

r i

r, j.

TA8LE 2-1 MATERIALS. SPECIFICATIONS, CHEMISTR,Y AND MATERIAL PROPERTIES a c,.

g OF [

3 y

t I

l 6

P

?

l l

l.

l 4

i e

T e

'M f.

4 I

i 2-13

, \\-

a,c.e 4

gI i

r fp.

~

Figure 2-7: Charpy U-Notch Energy at Room Temperature for [

Ja.cie 2-14

___,.w

i data obtained by Westinghouse (2.3) are given in Figure 2-8.

Fracture toughness J-R data obtained by Westinghouse are given in Figures 2-9 and 2-10 at 600*F and 200*F, respectively. A sununary of fracture toughness J-R data obtained from Reference 2.2 and Westinghouse (2-3) is given in Figure 2-11.

The above results were used to establish the [

]

for operating temperatures as given by Equations (1-4), (1-5) and (1-6).

2.6 A001TIONAL [

l***** OATA For the fracture toughness test on [

]

performed by Westinghouse only one half of a non-standard 2T compact specimen (2.12) (8 = 0.772 in., 20 sun) used in a fatigue test was available and was obtained from

~

(

J.c.e A full s<ize pl.an view (Xerox copy of the a

specimen plan) is shown in Figure 2-12.

The material was to 'be an example of a low toughness material per the techniques outlined in References 2.14 and 2.15 as part of a Westinghouse general research program. Three smaller bars of (

]

were obtained at a later time from which tensile specimens were made.

The specimen half of Figure 2-12 was first split to obtain two blanks, one

-0.4 in. (10 nut) thick and the other -0.35 (9 nui) thick. The test program reported here consisted of two parts. Initially, from the thicker section six standard compact specimens nominally 0.394 in. (10 nun) thick (2J2) and one standard Charpy V-notch specimen were obtained. The thinner section was saved for potential later investigations. The six 0.394 in, thick compact specimens were numbered 78A9-1 through 6.

The Charpy specimen was numbered 78A.

l

.The Charpy specimen was tested at 75'F and [

J c.e The a

load-time curve is given in Figure 2-13.

The compact specimens were tested at temperatures between -200*F (-129'C) and 550*F (288'C). There was an instrumentation failure during one test and no data were obtained.

i Essentially these five 0.394 in. (10 num) thick compact specimens were tested per ASTM E 992-84.(

• I Both load / displacement and load / time curves were obtained. If an instability was noted an~ unloading was attempted in an effort to determine the crack length as discussed in Reference 2.15.

2-15

1

?

a,c.e 9

I e

t Figure 2-8: Charpy V-Notch Energy at Room Temperature and 500'F for [

ja.c.e 2-16 f

O 1

1 i

a,c.e 4

4 s

Figure 2-9: J-integral Results at 600*F for [

3*'C,B=0.34in.

(Based on Deformation J Evaluation) r L

L 2-17 r

4 e

.----m.,--..

.,.-.--.m-

- - -,~.-

.r.

..i L

a,c.e s-l 4

t e

4 Figure 2-10: J-integral Results at 200*F for [

]a.c.e B = 0.34 in. (Based on Deformation J Evaluation) 2-18 m

- _. =

4 T

i I

a c.e 3

d t

I J

t Figure 2-11: J-da Curves at Different Temperatures for Aged Material-

[

]a,c.e l

2-19 l

9 1

l m-

7&"'~ i,~AihQA ~~~

M ' e y ::

9$ i._ b.: 43,y t.y M LM^iDM.%M.56d.

~

%.. WW.3 W-2.:~. -

$$.'.514. #

.3g Portion of Specimen er[

}

i

';W" lj j

.;,3 p

b..'.%g,

3 a,c.e./

. ~;. :g*:.

'i p.' W ".y.'..s.

~

l t..:.:

5 U. >.. ". '7 a.

.I pq

._g._.,

y J

crack orientation Figure 212: Full Size Plan View of Material Used in Investigation.

Thickness was 0.772 in. (20 nn).

2-20

a.c.e

~"

T 4

t

->s Figure 2-13: Load Time Trace for Charpy V-Notch Impact Specimen 78A of [--

]**C'8 Material Tested at Room Temperature 2-21 L____

The load displacement curves for the five compact specimens are shown in Figure 2-14.

Instability type behaviors past maximum load are noted for the specimens tested at -200*F (-129*C),150*F (66*C), and 200*F (93*C). The results at 550*F (288'C) did not display instability indications and oddly enough neither did the room temperature result. For these specimens the maximum load fracture toughness, K-EE(.16) and tearing modulus, Tu t' 35 obtained from Reference 2.14 were measured.

The best way to handle the tearing evaluation is not obvious. Generally T

is measured only for shelf behavior with no instability, such as noted at in the -200*F (-129'C),150*F (66*C) and 200*F (93*C) test results. Tmat was calculated per Reference 2.14 both by ignoring the instability and by taking it into account by limiting the final displacement to the displacement where the instability occurred. These results are given in Table 2-2.

As interest in and concern for theriaal aging of cast stainless steel increased in the early 1980's, it was realized that the results obtained above were more than just another set of fracture toughness results for a low upper Charpy V-notch shelf material. Attempts to obtain additional [

]' #

i material yielded only three small bars from which tensile specimens could be machined as previously noted. These tests were performed at roc:t temperature and 600*F (316*C). The remaining - 0.35 in. thick blank mentioned above was then machined to obtain 6 compact specimens nominally 0.34 in. thick and four subsize (ieduced cross section) Charpy y-notch specimens. The compact specimens were numbered CT1 through 6.

Four of the specimens were tested per ASTM E-813(2.U) for determining J using unioading compliance.

g Estimates for J an T were obtained. Tests were perfomed at -320*F Ic mat

(-196*C), room temperature, 200*F (93*C) and 600*F (316*C).

J-R curves for the 200*F (93*C) and 600*F (316*C) s. ave been previously given in Figures 2-9 and 2-10, respectively, and were used as part of the data base in setting the 1

[

- ]* criteria discussed in the previous section. Two compact specimens were tested dynamically to resolve speculation about strain-rate i

sensitivity (2.1)

The subsize Charpy specimens were tested at room temperature, 200*F (93*C) and 550*F (288'C).

f 2-22

t o

a,c.e c

^

I e

l l

d Figure 2-14: Load-Displacement Curves for the Five 0.394 in, Thick Compact Specimens Tested at Temperatures Between -200*F and 550*F (Note differences -in scales and crack ratios) 2-23 T=

y 9-, -

p-.

m 4

e,,-p

a i

e 4

U.

O l

l 4

EU C*

6e m

- M W

e-.

' N EW LQ EA

. M MW Z

O

>=

W M3>=

(J u.

b)

-W 4>=

sn N

a N-W J

S4>=

2-24

l The true stress-true strain curves obtained at room temperature and 600*F(316*C) are given in Figure 2-15. Tabular data are also given in the figure and supplement the results given in Table 2-1.

Load displacement curves for the four statically tested 0.34 in. thick compact specimens are given in Figure 2-16.

The room temperature and 600*F result exhibited no instability. Again the 200*F result would appear to give some indication of tearing instability although somewhat masked by the unloading and reloading. At -320*F (-196*C) an instability actually occurred right after an unloading compliances was performed; that is, the specimen just continued to tear as the load dropped. The J-Aa data for the specimens tested at -320*F and, room temperatur,e.are given in Figure 2-17 and 2-18, respectively. The results from this second set of static tests are given in Table 2-2 also. The J-R curve of the 600*F (316*C) static test was previously

~

plotted in Figure 2-11 for comparison with the results of Reference 2.2.

The comparison is very complimentary.

Two compact specimens were tested dynamically at 600*F. Comparisons of the load-displacement curves with the static test curve are given in Figure 2-19.

For the dynamic tests only the maximum load toughness values were calculated.

These values are also given in Table 2-2.

The four subsize Charpy Y-notch speciser.s had the standard length of 2.165 in.

(55 mm). The cross section was square and the V-notch was 20% of the total depth (i.e., except for length the subsize specimens were geometrically f

similar to the standard specimen). Three specimens'had a width' equal to ene-half of the standard Charpy specimen. The remaining one had a width equal to 87% of the standard Charpy specimen. The smaller specimens were tested at room temperature and 200*F (93*C). An adequate energy measurement was not obtained from one test. The larger specimen was tested at 550*F (288'C).

Results from these tests and the prior full size specimen test are given in Table 2-3.

2-25

e a,c.e L

Y i T T

f x

[.a 0

m Figure 2-15:

rue-Stress True-Strain Curves at Room T pgrgture and 600*F for 2-26 m.

e a,c.e 4

s I

t a,c.e Figure 2-16: _ toad Displacement Curves for[

3 Thick Specimens of[

3a,c.e Material Tested Statically Between -320*F and 600*F Using Unioading-Compliance 2-27

e e' -'

=

4 s

1 a,c.e-i i

e g 1

r r

Figure 2-17: J-Integral Results at -320*F for [

'ja.ce(Based on Deformation J Evaluation) 4 2-28

g a,c.e i

Figure 2-18: J-IntegralResu{tgatRoomTemperature(75'F)for[

3

(Based on Defomation J Evaluation) 2-29

-. ~

a,c.e 1

P

~

Figure 2-19: Comparison of Static Load-Displacement Curve with Ovr.2mic Load-Displacement Curves at 600*F ([

]a,c.e) 2-30

TABLE 2-3: CHARPY IMPACT ENERGIES FROM STANDARD AND NON-STANDARD SPECIMENS AT DIFFERENT IEMPERATURES FOR [

Ja.c.e a.c.e I

r,a l'd aWestinghouse Data bDimension of square cross section. Notch was 1/5 of B.

The length was as for a Standard Charpy Test (2.165 in., 55 mm)

Estimated assuming specimens C3 and C4 were tested at same temperature c

1-i i

1 9

It has been demonstrated that additional aging of (

]a,c.e could result in further embrittlement of the ferrite as depicted by the lower temperature fracture toughness tests. The behavior of the J estimates are plotted in R

Figure 2-20.

The most significant observation is that at -320'F (-196*C) where the ferrite is surely completely embrittled, the J estimate is still gg around [

]a,c.e This value then is the toughness of the austenitic matrix for [

]

in the duplex geometry which exists.

Since thermally aged cast stainless steel nuclear grade pipes of Type CF8M are expected to have quite similar micro-structure and chemistry, a value of

(

]a.c.e for J;c is a reasonable her bound for such CF8M pioes, even when the ferrite is completely emerittled if, as seen later, the ferrite content is that of ('

]

The load displacement curves given in Figures 2-14 and 2-16 show that plastic instability type behavior can occur up to 200'F (93*C). Such behaviors suggest that the tearing resistance is degrading with decreasing temperature.

Supplementing this, the tearing moduli of Table 2-2 also do not appear to

m sent as an-optimistic view as J The tearing moduli are plotted in 7g.

' igure 2-21 as a function of temperature. [

~

]#' *'

The dynamic results given in Table 2-2 suggest that the material is not

[

]a,c.e at operating temperatures.

In this sense, Charpy impact results are judged to be indicative of ( -

aj.c.e In Reference 2.2 the room temperature Charpy U-notch results are [

] # On a per area basis the 20 ft-lbs obtained from specimen 78A of Table 2-3 is [

] #, being in very good agreement with

.the U-notch results.

2-32

~

a,c.e

~

l l

a i

l

.I l

1 i

l i

l i

t Figure 2-20: J as a function of Test Temperature for[

]*

Ic 2-33

.....,, ~s...._ _.. _ _.. _

4.-_.,_

5

[

}

j

\\

r,-

1'3._

b

- =

r 4

o

. r I.

p

..-4

.t' l'^

s s

?

" ' I x. h 4

M db-t r

f a

4.t T

,,.L

~

a c.e -

i

.m,'

I a,

r.

1;.t s

s.

5-

.vx o

J-'

e

> i 3

p e.m t

7

r. -

,n..

E...

+

+

I

+

i

{-

V.

s' i

.-t t

L

,2 L

t l.

e t- -','.i- -

. j,

.i 4

4

d. ce i

f.

{ :"

. t

- 1

.,,d

=.,f',

ie T

f v

?

j 4

-- 7 7

[,

.t

.s..

3 i

e

..m.:

?

P J.

- k'

1. Figure 2-21: J. Variation of Tut with Temperature for[

3,c.e a

a,1 9

w g

F I

s k

4 2-34. -

i s

e

~

5

+

k q_-

t

,. J.

r i

J In Table 2-3, the subsize and full Charpy results are consistent. In the lower transition region one would expect the energy to vary somewhat as the cross sectional area just as in brittle fracture testing for K room temperature results, the areas vary by a factor of [ ]a,cN. For the which is also the ratio of the energies [

].

For very high upper shelf Charpy behavior, the energies for subsize and full size Charpy specimens would vary somewhat as the cube of the scale factor.

For lower upper shelf values, variation with the cube of the scale factor is not expected to be realized. Thus the energy [

]C for the subsize Charpy specimen tested at 550*F should be scaled by some factor between [

]a,c.e to obtain the full size Charpy energy. This gives between [

]"

If the result at 200*F (93*C) and 550*F _(288'C) were assumed to have been perfonned at the same temperature, a direct scaling can be p'erfonned. Such a scaling is reasonable since the 200'F (93*C) result does exhibit a slight increase in toughness over the corresponding room temperature test. The scaling produced an energy estimate for the full size Charpy specimen at 550*F (288'C) of [

]*** and this is the value plotted in Figure 2-8.

At 200*F (93*C) the increasing toughness of the subsize specimen suggests the full size Charpy specimen would exhibit increased toughness over the room temperature results. The near shelf result at 200*F (93*C) is based on experience and judgment and is somewhat speculative.

2.7 THE KCU FOR [

]

2 In the prior section a J of [

]W in-M n was demonstrated for gg

[

]a,c.e with fully embrittled ferrite. However, no impact tests were perfonned to obtain a corresponding KCU energy for the fully embrittled ferrite. In this section such a KCU is set forth.

In Figure 2-16, the test at -320*F produced the J of (

gC g

. ]' ' ' d The energy for the J;g obtained from the R-curve is very close to the energy at maximum load of the load displacement curve.

~

2-35

A maxisma load ctrrelction cf K with Charpy impact energy (KCV) is giv;n in Reference 2.18.

Although not verified for mateHals with low Charpy shelf behaving in a ductile manner such as the unedHttled austenite, a Charpy impact energy may be calculated from 3, using K = (EJ )

The 7

Charpy impact energy so obtained was [

J.c' Application of the a

Barsom-Rolfe correlation (2.19) for E using the K obtained from J was gg gc

[

3 c.e a

From the J results plotted in Figure 2-20, it is obvious that Charpy gg energies at temperatures below the room temperature result of Figure 2-8 will be less than the [

]

obtained at room temperature. While there are no definitive instabilities noted in Figure 2-13, pop-in type behaviors (blips) are noted up to point A.

It would be frivolous to say instabilities are occurring since machine instrumentation could be responsible for such blips. However, if one assumes the ferrite is breaking in cleavage up to point A, an estimate of the energy to fracture the austenite is approximately

[

,)a c.,

It is concluded from this discussion that for fully embrittled ferrite in

[

J.c.e the Charpy V-notch energy is around ['

a

,)a,c.e 2.8 THE ALTERNATE [

]" # TOUGHNESS CRITERIA In this section, alternate [

lC# cHteria for cast stainless steel nuclear primary piping are set forth. The new criteria consists of two parts. First if the KCU correlation of Reference 2.2 for a piping component provides an end of life KCU greater than that for [

]a,c.e.(4,,,,

~

greater than [

']')',

then the current [.

] C #

cHteria reviewed in Section 1.0 are applicable. This could be a restrictive criteria for cast piping less prone to thermal aging embrittlement. However, i

for leak-before-break evaluations involving some twenty-five Westinghouse nuclear plants, the current [

l

criteria have not proven restHctive. A less restrictive criteria than that based on [

]#

is not addressed in this report.

\\.

The second part of the alternate criteria addresses the situation when the calculated KCU falls below that of (

.]"

As shown in the previous section, a minimum J of [-

gg 4

2-36

h1

-]"'C'" daJ/cm ) with Tm[

]"'C may 2

l always be taken for [

]"'C

For this case the Charpy V-notch energy has been conservatively estimated in Section 2.7 to be [

]" ' C #. However, using this fully embrittled case as the criteria when the current [

l.c.e criteria are not met is unduly restrictive a

since as has been demonstrated in the,'rior sections, the embrittling effect in terms of ispact energy after long periods of time is very gradual. Jgc in turns reflects a gradual change with impact energy at low values as shown in Figure 2-22 (data sources are identified in the figure).

For the embrittling situation considered, the ferrite content in relation to that of [

l.c.e must also be addressed. The calculated ferrite a

content for [

l.c.e varies between [,

]C percent a

weight, having an average valve of [

]" # percent.

For the fully embrittled case as discussed for [

1 'C# in the previous 8

section, the ferrite is assumed to make no contribution to the fracture resistance. Referring to the discussion in Section 2.4, the amount of ferrite for the fully embrittled case is seen to have the potential for being a significant contributor to the fully embrittled fracture resistance. This is discussed below.

Since data from [

l

"# are essentially the only data for which full b

embrittlement of the ferrite may be addressed, [

(2-4) aj,c.e 2-37

s_

c, a

ygren E

tcapm I

hc to N

V ypra hC h

t iw c

i J

fo no i

t a

a ira V

I2 2

erug i

F r,M a

4 l

~

l M 1.04 OA1 0.63 Thickness. [: ;,. :,

- - ~ ' - - -

(mm) :-

.. : -.J 5. :;. ;..: c.,,d s.

dr :, 4.2 4.4 s.1

$% 1A 13.8 19.5 dp: e.s 7.4 e.s 5% 3.1 14.8 23.4

/

de:

7.7 so.e s.9 S% 2.9 17.2 2a.4 10 M Q

b1 C

Figure 2-23: Micrographic features of two-phase specimens as a function of chemical composition Ni'eq/Cr'eq and thickness. The i

ferrite is colored by electrolytic etching in a 10-N sodium i

hydroxide solution. dF = average segment intercepted by, ferrite (um); 6% = ferrite content measured by micrography i

2-39 r

[

I~

First the sumation term of Equation (2-3),

i.e.,

m t

I I

S W.

(2-5) il IJ j.1 11 is :ensiderec.

f.

e 2-s)

(2-7)

(2-8)

(2-9)

)a,c.e 2-40

4

.T

~f >;

[

1 1,c,e a

1-The question remains as to how the [

] # #

First if a Charpy impact specimen is tested affects Jgg, Tmat ""d Jm.

in a slow bend mode in the fully ductile region, the impact energy and energy from the slow bend mode are about the same. The model of Equation (2-3) was validated by dynamic tear (sharp crack) tests. Again in the ductile region,

'he impact and slow bend test energies are abcut the same. A slow bend test of a specimen having a sharp crack is a fracture toughness test from whien Jgg, T and J could be detennined. Thus the total energy from a fracture toughness test would vary with [- ]a,c.e ju'st as for impact tests.

Since J could be obtained for the full energy to instability, J, would vary as [ ]C

Observing the -320*F result (Spec. CT4) of Figure 2-16, the energy to J as seen in Table 2-4 (to be discussed later) is R

about [

]C of that of J,,,.

A ferrite-austenite ratio of only

(

']a,c.e for J yields the J value. It.is conserv'att've then to

] # gg assume J varies with [

gg Thus, for a generalized heat of material (call it Heat X) with a variable a.c.e g

.j Jgg (Heat X) = [

]a c.e

( 2-10)

J,,, (Heat X) = [

l 'C#

( 2-11)

Since [

l

(2-12)

' )*

• C # for thermally aged At the present there are (

s cast stainless steel available for which the operating temperature J and g

the room temperature impact energy are estimated. These are tabulated in Table 2-4.

The room temperature impact results must be in U-notch energy not V-notch energy. A plot of KCU energies versus KCV energies are given in 2-41

.L

A 4

-O e

4 1

I TABLE 2-4 JIc' Tmat AND J AT OPERATING TEMPERATURES FOR WHICH THE- [

]

is kn0wn a,c.e 4

f4

~

d 4

I*

I.

The [

l c.e fit of the data was used to a

Figure 2-24

- calculate KCU for the three available data points; the values are also given in Table 2-4 With the above information in hand, the alternate criteria when the calculated KCU value falls below [

} ' # can be addressed. The first step is to determine the fully embrittled fracture toughness values for the material under consideration. Having determined the [

3,c.e a

Jgg = [

] C #

(2-13)

T

=[]

(2-14) mat J,,, = [

] #

(2-15)

• KCU = [

]#

(2-16)

The embrittlement is a slow process and the f racture tougnness values are assumed to vary [

]a,c.e values between the fully embrittlea toughness and the current [

} full service lif e criteria. The assumed [

1** is justified based on the near [

)*'C of the results of Table 2-4 which are plotted in Figure 2-25.

If the calculated KCU value is less than the KCU value [

3,c.e a

In sunenary the alterr. ate toughness criteria for thermally aged cast stainless steel may be divided into three categories. Those are suninarized below.

f

[

a.c.e j

J;c = [

1

'C

(2-17) 8 T

=[

]

(2-18) at J,,, = [

]

(2-19) a [

3a,c.e 2-43 n-,

e

.m,.

.c

4 a.c.e r

Figure 2-24: Correspondence Between the KCV Energy and KCU Energy of Austenitic Stainless Steels tested at 20*C I

2-44 i

r t

a,c.e a

7

- F.

s 5

a Function of [ mat and d ax Figure 2-25:

Plots of JIc, T m

at Operating Temperature as Ja,c.e 2-45

C 3,c.e,a a

Jgg = [

]" '" #

(2-20)

Tat = [

]

(2-21) 3,,x = (-

]

(2-22)

L

.[

i j.c.e.

a JIc = C

]

(2-23)

T

(]

a at (2-24)

J,,, = [

]

(2-25) 2.9 EXPERIMENTAL VERIFICATION OF THE ALTERNATE TOUGHNESS CRITERIA Research is continuing on the long tenu effects of thennal aging on cast stainless steels at several facilities. The work of Trautwein and Gysel (Reference 2.1) continue to be extended at George Fischer, Ltd., Schaf fhausen, Switzerland.

The U.S. Nuclear Regulatory Commission. has sponsored aging investigations at Argonne National Laboratory ( *

}

Unfortunately only preliminary results are available which pertain to verifying the methodology set forth in this report. The extensive program at t

l Argonne National Laboratory anticipates some test results during 1985 but none l

are currently available.

The work at George Fischer, Ltd. apparently is not oriented toward fracture toughness (Jgg, T t) testing but general trends and additional correlative type results are expected.

)a,c.e 2-46 3

r-.

m

,y_

.-,s..e

,m.

,___~,r...,

i i

1 Some preliminary results, however, are available from [

]*'C.

A static heat of cast stainless steel was specially made to l

obtain a high aging sensitivity. The check analysis chemistry of the heat is as follows (weight %):

[

i '

j.c.e a

Aging was at 750*F (400*C) for 1000 hours0.0116 days <br />0.278 hours <br />0.00165 weeks <br />3.805e-4 months <br /> and 5000 hours0.0579 days <br />1.389 hours <br />0.00827 weeks <br />0.0019 months <br />. The 4-ferrite is quoted as being between [

]' ' ' #. Specifically the ASTM method based on the check analysis gave [

]***, the magnetic method gave

[

]

and the ASTM micrographic method gave [

].

It should be noted that the chromium content exceeds that of the specification for CF8M (18 to 21%). Also the 4-ferrite is well above the maximum anticipated'4-ferrite content [

]"

for cast stainless steel pipe

~

and fittings used in the Westinghouse type piping systems. However, metallurgical examinations confirmed that the microstructure of the test heat was typical of CFBM and that the same mechanism of aging existed for this material as exists for CF8M.

The available experimental results are as follows:

['

j.c.e a

Based on a comparison of the KCU values, the material is judged to be fully aged after [

] C d.

The.l was obtained by averaging values at 68*F (20*C), 212*F (100*C) and gg 572*F (300*C) with variation within 120%. These results were obtained from 1T I2"I )

compact specimens with 20% side grooves Unloading compliance was used. The room temperature yield strength was [

]

and the ultimate strength was [

J,c.e T

is thus seen a

mat to be between [

l.c.e a

i 2-47 l

i.

I

b The Charpy KCU values were obtained from a standard Charpy machine. The values quoted are thought to compensate for machine energy Tosses as related to such small values. Tests are planned, however, using a small capacity

. impact machine to validate these numbers.

These results as related to the methodology of the alternate toughness criteria are discussed below.

First, as noted the Charpy KCU energy values are [

]' ' " #. Precise infonantion on exactly how the energy from the KCU values quoted were compensated for was not given. The values of (

]C # are thus subject to confirmatica.

Side-grooving fracture toughness specimens tend to lower the J, measured.

7 For example, at room temperature J determined from a non-side grooved k

specimen gave [

j c.e (Table 2-2 and Figure 2-18) while with a

side grooves a valua of [

]a.c.e was found (Figure 2-11), a difference of 48%.

Using the above chemistry, the ASTN procedure as applied by Westinghouse gives a'a-ferrite of [

].

The discrepancy with the [

]" # quoted is not yet resolved. Based on the 3-ferrite values given, it would seem that prisently the. interval [

]

is appropriate.

Applying the technique of Section 2.8, the following are determined:

[

3,c.e a

2-48 g

v It is seen that the calculated KCU values are somewhat less than the [

] # measured. However, more exact compensation for energy dispensation could easily bring the predicted and experimental values into agreement. As shown above the real J ob W ned from non-side grooved gc 4

specimen could be as much as (

i l' ' ' #. Thus based on the available data, the experiental and i

calculated J values are in reasonable agreement..

Ic It is concluded that the accuracy with which the tortuous beam model predicted this rather extreme situation is quite good and suggests that the model is a

- viable tool for evaluating the embrittling effects of long teria thermal aging on cast stainless steel. Dathanticipatedfromcontinuingstudiesare exp6cted to quantify the model further.

2.10 SAMPLE APPLICATION OF-THE ALTERNATIVE TOUGHNESS CRITERIA The steps for establishing the alternate toughness criteria for a specific plant piping system are outlined below. A two-loop p1' ant is considered in this example; however, the procedure would be siellar for three-loop and four loop plants.

Chemistry and KCU Touchness.

i C

l j c.e a

l Thus the fully embrittled toughness values are by Equations (2-13 through 2-16),

2-49

~

- l

n:%.

14g.m l

' l

-TABLE 2-5' EXAMPLE CHEMISTRY AND CALCULATED [

3

'C VALUES FOR 8

PIPE AND FITTINGS-a,c.e 7

t 4

L 1

j.

9 '

'4 A

4

+

q f

e E

I' A

4 i

I d

t i

4

(

i 2-30 s

i 4

I

~

-4 m

em, m.-.--

.,-o--,-wsm,-

>am,---.-e,.--~,-,wve-,+v,,.,

-e;+-.v~w-me,w-ew,-w.-w,~-e,+,ww.v--u,r-vre-g-~--,

-rw w--x-

[

aj.c.e Since the end-of-service life KCU is [

]"', Heat X falls under the [

]'"

criteria. For Heat Y, the

[

} is:

(

j c.e a

Thus, the fully embrittled toughness is:

C j c.e a

~ Since the calculated end-of-service life KCU value of [

]"'#

is less than the fully aged KCU value of.(

]"', Heat Y falls i

under the [

]a,c.e,

The As-built Loon Central-File drawings are available for the as-built loops of each Westinghouse nuclear power plant. Thus, the location of each heat may be determined.

A layout for a two-loop plant is given in Figure 2-26.

The heats of Table 2-5 l

are identified with their locations in Figures 2-27, 2-28 and 2-29.

2-51

. =. -, -.

o., -,,

)

I Steam Generator

,.g Crossover Leg Hot leg i..

.,) Loop 8 Pump Reactor Vessel

,. ' ~..,

Cold Leg

[

'.._./

'~

Loop A s' '.

l'~~f

-4 wp

.I Figure 2-26: Typical Layout of the Primary Loops for a Westinghouse Two Loop Plant 2-52

9 Loop B i

Heat A-N REACTOR VESSEL

\\

\\

t

,7 p PUMP h

\\

\\

tb J

/

Loop A Figure 2-27: Identification of Heats with location for Cold Leg 2-53

,.,' o

./

Steam Loop B Generator (Reactor i

Heat X*

\\ Vessel i

s!

,f

)

Loco A

(

• 0oes not comply with current [

]

criteria Figure 2-28 Identification of Heats with Location for Hot Leg e

2-54

o.

~

4 5 team Loop B Generator

'es Pump HEAT Y* -

\\/

J Loop A Does not comply with current [

3a.c.e criteria Figure 2-29:

Identification of Heats with Location for Crossover Leg 2-55

4...o The Criteria for the Example Acclication The alternate toughness criteria of Section 2.8 is now applied to all piping components of this example calculation. The results are given in Table 2-6.

The toughness values for Heat X were obtained by the [

]" # (Equations 2-20, 2-21 and 2-22) and are-plotted along with the criteria lines for Heat X in Figure 2-30.

9 l

1 l

[

i j c.e a

l 2-56

O 4 6 *

.e I

I w

C=

1 r

E 1-El si A

se w

I *3 w

5*

$52" m

W t

l 1

I l

l t

2-57 I

o..,.

a.c.e Figure 2-30:

Fracture Toughness Criteria for Heat)X ([ Sample Application Ja.c.e of 2-58

I 3.0 DISCUSSION ANO CONCLUSION The primary objective of this report is to develop a general methodology for 4

estimating fracture toughness criteria applicable for leak-before-break evaluations of cast stainless steel piping which is quite sensitive to long term theriaal aging, falling below ths currently accepted criteria. This has been accomplished in several steps.

F.'rst, available information is reviewed and insight into the fracture behavior of steels is developed. Specifically, the energy disposition in fracturing specimens is demonstrated and applicability of an energy model developed for carbon steels is suggested for application to fully aged stainless steels. Additional data are presented for the criterion material ([

]#) and a generic model, keyed to

[

]* # fracture toughness criteria is eset forth. A comparison of model predictions and experimental results for a most extreme embrittlement 15 positive. Investigations in progress can be reasonably expected to further substantiate the model. Finally, application of the model for determining alternate fracture toughness criteria for thermally aged cast stainless steel l

[

]"'C # is demonstrated.

Based on the results presented, it is judged that a viable engineering methodology has been developed for establishing alternate fracture criteria for thermally aged cast piping [

]a,c.e,

applicable for leak-before-break analysis.

1 i

3-1

• , e, a

4.0 REFERENCES

1.1

(

j.c.e a

(Proprietary Class 2).

1.2 E.1. Landeman and W. H. Bamford, Fracture Toughness and Fatigue Characteristics of Centrifugally Cast Type 316 5tainless Steel After Simulated Thermal Service Conditions, presented at the Winter annual Meeting of the ASNE, San Francisco, CA.

December 10-15,1978 (MPC-8, ASME).

1.3 W. H. Bamford, E. I. Landerman and E. Diaz, Thermal Aging of Cast Stainless Steel and Its Impact on Piping Integrity. ASME PVP-Vol.

95. Circumferential Cracks in Pressure Vessels and Pining, Vol. II,

)

pp 137-172, (1984).

1 1.4 Wuclear Regulatory Comunission 10 CFR Part 50, Modification of General Design Criteria 4 Requirements for Protection Against Dynamic Effects of Postulated Pipe Rupture, Federal Register /Vol. 50, No. 126/ Monday, July 1, 1985/ Proposed Rules, pp. 27006-27009.

1.5 Letter

8. J. Youngblood, Chief Licensing Branch # 1, Division of Licensing, USNRC, Washington, D. C. to R. J. Gary, Executive Vice President Texas Uti.lities Generating Company, Dallas, Texas, dated March 2, 1984,

Subject:

Request For Additional Information Concerning Leak Before Break Analysis For Comanche Peak Steam Electric Station (Units 1 and 2); Docket Numbers 50-445/446.

4 I

N 4-1

<a e e 1.6 0 1

j.c.e a

1.7 51ama, G.,

' requin, P., Masson, S.

H., and Mager T.

R., 'Ef fect of Aging on Mechanical Properties of Austenitic Stainless Steel Casting Welds", presented at SMiRT 7 Post Conference Seminar 6 - Assuring Structural Integrity of Steel Reactor Pressure Soundary Components, August 29/30, 1983, Monterey, CA.

2.1 A. Trautwein, and W. Gysel, " Influence of Long-time Aging of CFB and CF8M Cast Steel at Temperatures 8etween 300 and 500 Degree C on the Impact Toughness and the Structural Properties". Stainless Steel Castines, ASTM STP 756, V. G. Behal and A. 5. Melilli, Eds, American Society for Testing and Materials,1981, pp 165-189.

2.2 51ama el al; same as Reference 1.7.

2.3 [

]a.c.e; same as Reference 1.6.

2.4 T. J. Nichol. A. Batta, and G. Aggen, Transactions, American Institute of Mining, Metallurgical and Petroleum Engineers, Vol. II A,1980, pp. 573.

2.5 R. O. Williams and H. W. Paxton, J. Iron Steel Inst., Vol. 185, 1957, pp.

358-314.

2.6 R. O. Williams, Trans. AIME, Vol. 212,1958, pp. 497-502.

2.7 H. Hendry, Z. F. Mazur and K. H. Jack, Metal Science, August American Society for Testing and Materials,1980, pp. 482-486.

2.8 R. R. Lula, Ed., Touchness of Ferritic Stainless Steels, ASTM STP 706 American Society For Testing and Materials,1980.

l 4-2

1 3%, b.

j 2.8a H. E. Deverall, " Toughness Properties of Vacuum Induction Melted High-Chromium Ferritic Stainless Steels," Touchness of Ferritic Ltainless Steels. ASTM STP 706.

R. A. Lula, Ed., American Society for Testing and Materials,1980, pp.184-201. Also Letter from G. Aggen, Allegheny Ludlum Steel Corporation Brackenridge, PA. to C. C. Kim, Westinghouse Electric Corp.,

dated 2/26/85,

Subject:

Impact Properties of E-8 rite (T.M.).

2.9 W. H. Samford, et al. same as Reference 1.3.

2.10 F. J. Witt and R. G.~ 8erggren, " Size Effects and Energy Oisposition in Impact ' specimen Testing of ASTM Ar 533 Grade 8 Steel" Experimental Mechanics, vol.11, No. 5, pp.193-201 (May 1971).

2.11 E 23 - Standard Methods for Notched Bar Impact Testing of Metallic Materials, Annual Book of ASTM Standards, Section 3 - Metals Test Methods and Analytical Procedures, Vol. 03.01, Metals - Mechanical Testing; Elevated.and Low-Temperature Tests.

2.12 E 399 - Standard Test Method for Plane-Stain Fracture Toughness of Metallic Materials, Annual ASTM Book of Standards, Section 3 - Metals Test Methods and Analytical Procedures, Vol. 03.01, Metals - Mechanic Testing, Elevated and low-Temperature Tests.

2.13 [

aj.c.e 2.14 F. J. Witt, Fracture Toughness Parameters obtained from Single Small Specimen Tests, WCAP-9397, Westinghouse Electric Corp;, October 1978.

2.15 F. J. Witt, "An Engineering Interpretation of Pop-in Arrest and Tearing Arrest in Terms of Static Crack Arrest, E,,' Engineering Fracture g

Mechanics, Vol.18, No. 5, pp. 997-1010 (1983).

4-3

r.

< e.s 4 2.16 E 992 Standard Practice for Determination of a Fracture Toughness of Steels using Equivalent Energy Methodology, Annual ASTM 8cok of Standards, Section 3 - Metals Test Methods and Analytical Procedures, Vol. 03.01, Metals - Mechanic Testing Elevated and Low-Temperature Tests.

2.17 E 813 - Standard Test Method for JIc, A Measure f Fracture Toughness, Annual Book of Standards, Section 3 - Metals Test Methods and Analytical Procedures Vol. 03.01, Metals - Mechanical Testing; Elevated and Low-Temperature Tests.

2.18 F. J. Witt, " Relationships Bet. teen Charpy Impact Energies and Upper Shelf E

Values for Reactor Pressure Vessel Steels", Int. Journal of gg Pres.sure Vessels and Piping, Vol. 11, pp. 47-63 (1983).

2.19 5. T. Rolfe and J. M. Barson, Fracture and Fatique Control in Structures

- ADolications to Fracture Mechanics, Prentice-Hull, Chapter 6 (1977).

2.20 H. A. Ernst, " Material Response and ' Instability Seyond J Controlled Crack Growth", Elastic - Plastic Fracture: 2nd Sym. Vol. 1, In-elastic Crack Growth, ASTM STP 803, C. H. Shih and J. P. Gudas, Eds. American Society for Testing and Materials,1983, pp.1-191 to I-213.

2.21 M. T. Leger, " Predicting and Evalt ating Ferrite Content in Austenitic Stainless Steel Casting,' Stainle ss Steel Castinos, ASTM STP 75G, V. G.

i-Behal and A. 5. Melilli, Eds., American Soci.e.ty for Testing and i

. Mater.ials 1982, pp.105-125.

2.22 [

i

)a c.e 2.23 0. K. Chopra and 6. Ayrault, Long Term Embrittlement of Cast Duplex Stainless Steels in LWR, Systems: Annual Report, October 1982.

September 1983, NUPEG/CR.3857 or ANL-84-44, Argonne National Laboratory, j

August 1984.

s 4 -4 s