Reactor Pressure Vessel Structural Integrity Research In the U.S. NuclearRegulatory Commission HSST and HSSI Programs*

W. E. Pennell and W. R. CorwinOak Ridge National Laboratory

"The submittedmanuscripthas beenauthoredby a contractorof the U.S. Government under contract DE-AC05-84OR21400. Accordingly, the U.S. Government retains apaid-up, nonexclusive, irrevocable, worldwide license topublish or reproducethe published form of this contribution,preparederivative works, distributecopies to the public, andperform publicly and display publicly, or allow othersto doso, for U.S. Governmentpurposes."

DISCLAIMER

This report was prepared as an account of work sponsored by an agency of the United StatesGovernment. Neither the United States Governmentnor any agency thereof, nor any of their

employees, makes any warranty, express or implied, or assumes any legal liability or responsi-bility for the accuracy, completeness, or usefulness of any information, apparatus, product, orprocess disclosed, or represents that its use would not infringe privately owned rights, Refer-ence herein to any specific commercial product, process, or service by trade name, trademark,manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recom-mendation, or favoring by the United States Government or any agency thereof. The viewsand opinions of authors expressed herein do not necessarily state or reflect those of theUnited States Governmentor any agency thereof.

DISTRIBUTION OF THIS DOCUMENT IS UNLIMITED

(

i,,l!,_ i

W.E. PenneltandW.R.Corwln 1

Reactor Pressure Vessel Structural Integrity Research In the U.S. NuclearRegulatory Commission HSST and HSSI Programs

W. E. Pennell and W. R. CorwinOak Ridge National Laboratory

Development continues on the technology used to assess the safety of irradiation-embfittled nuclearreactor pressure vessels containing flaws. Fracture mechanics tests on reactor pressure vessel steel haveshown that (a) local brittle zones do not significantly degrade the material fracture toughness,(b) constraint relaxation at the crack tip of shallow surface flaws results in increased fracture toughness,and (c) biaxial loading reduces but does not eliminate the shallow-flaw fracture toughness elevation.Experimental irradiation investigations have shown that (a) the irradiation-induced shift in CharpyV-notch versus temperature behavior may not be adequate to conservatively assess fracture toughnessshifts due to embrittlement and Co)the wide global variations of initial chemistry and fracture propertiesof a nominally uniform material within a pressure vessel may confound accurate integrity assessmentsthat require baseline properties.

INTRODUCTION Technology (HSST) and Heavy Section Steel1. Regulatory requirements limit the permis- Irradiation (HSSI) Programsat Oak Ridge National

sible accumulation of irradiationdamage in the Laboratory(ORNL) areperforming the researchmaterial of a reactor pressure vessel. Irradiation required to resolve these issues and furtherdevelopdamage limits areset such that required fracture and refine the fracturemargin assessmentprevention margins aremaintained throughout the technology.nuclear plant licensed operating period. 3. This paperpresents abrief overview of theRegulatory requirements are based on fracture current statusof fractureprevention regulatorymechanics technology and utilize materials aging requirements and the associated fracturemargindata drawn from mandatory reactor vessel irradia- assessment technology. Issues identified with thetion damage surveillance programs, technology are reviewed, and some of the research

2. In recent years it has become evident that a programs implemented to resolve those issues arenumber of nuclear plants will exceed the regula- described.tory limits on irradiation damage to the reactorvessel material before the end of their current REGULATORY REQUIREMENTS ANDoperating license period (ref. l ). One result from RESEARCH NEEDSthis development is that a number of nuclear 4. Irradiationembrittlement and low vesselindustry organizations have gained experience in temperatures both act to reduce the cleavagethe application of fracture margin assessment fracture toughness of reactor vessel material.technology. This experience has resulted in the Regulatory requirements for fracture prevention,identification of a number of issues with the therefore, focus on operating and potential accidenttechnology in its present form. Data from conditions that generate both low temoeratures andirradiation testing programs, operating plant high stresses in the reactor vessel material. Forsurveillance programs, and large-scale fracture some vessels, however, an additional concerntechnology validation tests have identified exists at higher operating temperatures. Theseadditional issues. The Nuclear Regulatory vessels contain materials with a reduced resistanceCommission (NRC)-funded Heavy Section Steel to failure by ductile tearing. Additional

requirements are placed on vessels in this lattercategory to assure that adequate ductile tearing

"ResearchsponsoredbyOfficeof NuclearRegulatory fracture prevention margins are maintained.Research,U.S.NuclearRegulatoryCommissionunderInteragencyAgreement1886-8011-9BwiththeU.S. 5. A summary of the regulatory process forDepartmentof EnergyundercontractDE-AC05- assuring retention of adequate fracture prevention84OR21400withMartinMariettaEnergySystems,Inc. margins in reactor pressure vessels is presented in

' W.E. Penncll and W. R. Corwin 2

Fig. I. A mandatory reactor vessel surveillance OPERATINGPRESSURE.

program(refs2,3) provides data on the progressive _rrscu_v_ / _Rcugvg L__effect of irradiation embrittlement on the vessel

material fracture toughness. Changes in fracture _ OFLIFE / VESSEL / _N,,x__"_'_\_)k_" _\\UJ

toughness are measured using small Charpy _: -..,,,=/ _ .__-_-,_.._,specimens rather than large fracture toughness test _ 7_ ARTNDT _._.=_ ;;ERA'I3N'G'_x__yspecimens ,and,are expressed in terms of (a) lateral ,"I, ,,I Jr" 7 _xx ENVELOPE Q_,q

translation (ARTNDT)of the Charpy energy curve _. _,at the 41-J (30 ft-lb) energy level and (b) decrease zin the upper shelf fracture energy (AUSE) of the _material. Regulatory requirements for preservation o

of required fracture prevention margins are i MINIMUM PRESS UPtt".... _expressed in terms of an adjusted reference nil- ' __ductility transition temperature (RTNDT)(ret's 4-6) COdlin"TEMPERATURE_ I

and a minimum acceptable level of upper shelf Fig. 2. Adjustment of reactor P-T curve inenergy (USE) (ref. 4). response to irradiation embrittlement of vessel

6. Surveillance program data are used to material severely restricts permitted reactorperiodically adjust the upper bound to the reactor operating envelope.pressure-temperature (P-T) operating envelope(refs 5, 7), as shown in the left-hand branch in operations. This is particularly true for reactorsFig. 1. Adjustment is required to maintain margins with a fixed setpoint tow-temperature overpressureagainst fracture of 2 on pressure loading and 1 on protection (LTOP) system. This difficulty hassimultaneously applied thermal loading. Fracture sparked interest in refinement of the technologyprevention margins are calculated assuming an used to define the P-T curve and the associatedinner surface flaw, having a depth corresponding to LTOP setpoint. Specific research topics relating to25% of the wall thickness. The fracture margin this issue include (a) the effect of local brittleassessment analysis must be performed using zones and pop-ins on the material fracturelower-bound (KIIO dynamic fracture toughness toughness and (b) improved correlations forproperties (ref. 5). Adjustment of the P-T curve in predicting the irradiation-induced shift in theresponse to irradiation embrittlement of the reactor material RTNDT.

vessel material has the effect of restricting the 7. The center branch of Fig. 1 summarizes thepermissible reactor P-T operating envelope, as regulatory requirements for demonstration ofillustrated in Fig. 2. It is e, !dent from Fig. 2 that a acceptably low failure probability when the reactorhigh irradiation-induced RTNDT can restrict the vessel is exposed to pressurized thermal shockP-T operating envelope to the point where reactor (PTS) loading. A probabilistic PTS analysis isheatup and cooldown become very difficult required when irradiation damage has caused the

adjusted RTNDTfor the vessel material to reach

_ _1_=,*o=,_,, I [ screening limits. These limits have been set at[ _w.,_ _, ,-.-1 =,,._,_,,=,. 132.2°C (270°F) for plates, forgings, and axial

I nm.oT Ix, ,.o..,_._Ls weld material and 148.9°C (300°F) for circum-o.,, ,E.,_,,,.o._ ....._ ferential weld material (ref. 6). Acceptable

_ methodology for performing a plant-specific[ ,rr_._ ] [ usE._ ] probabilistic PTS analysis is defined in NRC

[ [ [ [ Regulatory Guide 1.154 (ref. 8).[ _._L=,s ] [ ,.T_.L,s j [ usE,=_ I 8. Principal features of the fracture mechanics

[ I elements of the PTS analysis procedure are. ,,cF,_ ,,,.G 1.,oct,,, . ,,e,, _O.A,,.o illustrated in Fig. 3. Thermal stresses are highest

• ..rm_,..ero._ [•,..ou,o_,.,. "=_'_"'"" adjacent to the inner surface of the vessel where

• DETERMINISTICANAL. • MIXED-MODEFRACTURE • DUCTILETEARING• ,,,,F_.. ,,o.,_.._-_,,.,L • co_Ec,_.,,= the effects of irradiation embrittlement and• 8.S. : 2.0 FLAWPOPULATION ,, DRAFTREGULATORY

_(F')£$ = 104 GUIDEDG.1023I transienttemperaturescombineto producethe

maximumreductionin thematerial fractureFig. 1. Regulatoryrequirementslimit permissible toughness.Thenet result from this combinationofirradiationembrittlementof reactorvessel conditionsis thatthemajority of crackinitiationsmaterialssuchthatadequatefractureprevention arepredictedto originatefrom shallow flawsmargins are maintained under both operating and located in the inner surface of the vessel. Thepotential accident loading conditions, dominant influence of shallow surface flaws

W. E. Pcnncll and W'. R. Corwm 3

' ' ' ' FRACTURE MECHANICS RESEARCH10. Local brittle zones and pop-ins. McCabe

I_ K,c ._ ._ et al. have evaluated the Kit scatter band and|\ \ ,NmA'r,ON/ / / ] lower-bound toughness for pressure vessel material

I \ \ (DEEP FLAWJ/ _ ARREST ] both with and without pop-ins (ret. l 1). Results

I__/__ from their evaluation are reproduced in Fig. 4.

They show that in general the pop-in results laywithin the Kit scatter band for specimens thatfailed without pop-ins. At most, the inclusion of

_'" TEMPERATURE the pop-in data marginally lowers the lower bound/J X "/_

]o/ _...Jx// ofthe Kic data. The interim conclusion from thisresearch on irradiated material is that it appears

[-- (SHALLOW /N _,

toughness curve higher than the KIR curve fordefining reactor LTOP setpoints. The selected

__'-/" ___ - curve, however, may not be as high as the Klc/ FLUENC E

ININER SURFACE I _ curve. Additional data for irradiated material mustbe included in the data base before this interimo o.a 0.4 o.s 0.8 1.o conclusion can be validated.

a/W, FRACTIONALWALLANDFLAWDEPTH 11. Crack-tip constraint effects for shallowflaws. Fracture occurs when the opening-mode

Fig. 3. Intersection of applied KI curve (stress and tensile stresses at the tip of a crack in a brittleflaw geometry dependent) with material fracture material exceed a critical value over a finite lengthtoughness curves (temperature, fluence, and

(ref. 12). Local yielding of the material at thematerial dependent) defines crack initiation and crack tip limits the buildup of opening-modearrest behavior at point in time during a PTS stresses and thereby directly influences fracturetransient, toughness. Tensile hydrostatic stresses contribute

directly to the crack-tip opening-mode tensilegenerates a need for quantitative assessments of stresses but do not influence yielding. Crack-tip(a) the effect of reduced crack-tip constraint on the constraint is the term used to describe conditionsmaterial fracture toughness associated with that influence the hydrostatic component of theshallow flaws, (b) the effect of prototypical biaxial crack-tip stress field. Low constraint reduces thestress states on the material shallow-flaw fracture hydrostatic stress contribution to the opening-modetoughness, and (c) the effect of stainless steel stress and thereby increases fracture toughness.cladding on the initiation of cracks from shallow Shallow flaws in a reactor pressure vessel havesurface flaws, reduced crack-tip constraint because of the

9. The requirement for the reactor vessel proximity of the inner surface of the vessel.material to maintain a minimum USE not less than Shallow-flaw fracture toughness would thereforethe 68-J (50-ft-lb) regulatory limit is reflected inthe right-hand branch of Fig. 1. The regulatory 3oo ....

requirement (ref. 4) states that this condition must zx,r.K,,°;r.K_. _ K_[_ _

be maintained "unless it is demonstrated in a u__rn,r.K.,, i i F.il

manner approved by the Director, Office of • ,r-Pop.,,I• ,'-Po,.. ,o

Nuclear Power Regulation, that lower values of 2oo Im,T.,_,.,,upper shelf energy will provide margins of safety _ - PO_-IN¶

against fracture equivalent to those required by , i i= I_ _

Appendix G of the ASME Code." Acceptable _ 1_

procedures for demonstrating the retention of IpOmNI .... "'" """adequate margins against fracture for low-upper- loo _ N__Kjc LOWERBOUNDshelf (LUS) material have recently been published ." ........in American Society of Mechanical Engineers "_-POP-INLOWERBOUND(ASME) Code Case N-5 12 (ref. 9) and Draft 5° 5o 85' _5 lo5'Regulatory Guide DG- 1023 (ref. 10). These r_srTEMPERATURE(oc)

procedures evaluate the stability of a reference Fig. 4. Test data for irradiated high-copper weldflaw in LUS material using an irradiation-adjusted (73W) indicate that lower-bound fracturematerial J-R curve. A need exists therefore for a toughness curve reflecting pop-in test data differsJ-R curve data base for LUS materials, very little from one derived from Kjc data.

• _V. E, Pcnncll and \V. R. Corwin 4

be expected to be higher than the toughness RTNDT between the two data sets. It is appropriatederived from tests on deep-flaw specimens, to note that most of"the shallow-flaw fracture

12. Fracture toughness tests have been toughness data plotted in Fig. 6 are for a tlawperformed on single-edge-notch-bending (SENB) depth of 10 mm (0.4 in.). Shallow-flaw fracturetest specimens using both deep (a/W = 0.5) and toughness would be expected to be flaw-depth-shallow (a/W = 0.1 ) flaws (ref. 13). Beam dependent, with larger shifts in the RTNDT

specimens used in these tests are shown in Fig. 5. anticipated as the flaw depth decreases. Note thatThe beams were fabricated from A533B material the shallow-flaw effect on fracture toughness isand were nominally 100 mm (4 in.) deep. Use of significant in the lower transition range of thebeams with this depth permitted testing of shallow fracture toughness curve, which is the area of theflaws with depths in the range identified as the curve important to PTS analysis, but is notcritical depth range for PTS analysis. Use of significant on the lower-shelf portion of the curve.prototypical flaw depths reduced the uncertainties 14. Dual parameter fracture toughnessassociated with extrapolation of shallow-flaw corrections and correlations have been proposed tofracture toughness data for application to full-scale provide a quantitative assessment of the effect ofstructures, reduction of crack-tip constraint on fracture

13. Shallow-flaw and deep-flaw fracture toughness. Principal features of the J-ACr fracturetoughness data generated in these tests are plotted toughness correction proposed by Dodds,in Fig. 6. The data are seen to be grouped into two Anderson, and Kirk (ref. 14) and the J-Q dualseparate families corresponding to the shallow and parameter fracture toughness correlation proposeddeep flaws, respectively. Lower-bound curves by O'Dowd and Shih (refs 15,16) are illustrated in

fitted to the data show a 35°C (63°F) shift in Fig. 7. In the J-Q fracture toughness correlation, Qdefines the departure of the stress-state-dependent

J -AerI l l 'I' I l I I

_ (_RIOo--kSTRESS _ /

-'o - CONTOUR/ (

.._0 _' .... _,_

_ CRACK

Fig. 5. 100-mm-deep beams were used in shallow- , , I I I J _ Iflaw test program to permit full-scale testing of 0surfaceflawshavingdepthsintherangethatPTS x/(J/oo)analysis has shown to be the controlling range for _r is the area overwhichstressescrack initiation, exceedthereferencestressratio

(URIC0) forcrackinitiation

'°°....o,o.,Wo•'o8,o.....s " /°"1 :'-° ........g ""_'=,,mj

400 [- • 14 100 50 • RACK , .......2.4 ..................o 50 100 100

n 10 100 100' " 50 100 150 TSHIF-r=35oC

_;_ 300 " 10 100 150 • 7 / "_ 1.6_" _ V ..... 0.22 I

_ 200 0.81 ----- 0 n=10 I

o/ -F-°, 7° , ,, I100 0 I 2 3 4 5

d(Jloo)

.............. ' ......... Q is a parameter characterizing the0150 -100 -50 0 50 100 hydrostatic stresses and therefore

T-RTNDT(C) the maximum principal stresses

Fig. 6. Data from shallow-flaw fracture toughness aheadof thecracktesting program show that shallow-flaw effect Fig. 7. Dual-parameter corrections andproduces substantial increase in toughness at correlations have been proposed as means fortemperatures in transition region of fracture adjusting fracture toughness data for effects oftoughness curve, crack-tip constraint.

W. E, Penncll and W. R. Corwm 5

opening-mode stress distribution on the crack constraint is reduced. Both of these trends canplane from the opening-mode stress distribution have a significant influence on the outcome of aderived by Hutchinson (ref. 17)and Rice and probabilistic analysis of reactor vessel failure ratesRosengren (ref. 18) (HRR) for a highly constrained under PTS loading.crack tip. Q is therefore expressed as follows: 16. Biaxial fracture toughness testing of

shallow flaws. Biaxial stress fields are produced in

Q = (yyy - (Gyy)HRR (1) a reactor pressure vessel wall by both pressurea o ' loading and through-wall temperature gradients. A

typical biaxial stress field is shown in Fig. 9,

where (Yyy is the crack-tip opening-mode stress together with a constant-depth, shallow-surfacedistribution for a specific constraint condition, flaw. One of the principal stresses is seen to be(('Tyy)HRRis the corresponding stress distribution aligned parallel to the crack front. There is nofor the HRR constraint condition, and _o is the counterpart of this far-field, out-of-plane stress inmaterial yield stress. In physical terms, Q is the the shallow-flaw fracture toughness tests described

previously. The far-field, out-of-plane stress hasdifference between the hydrostatic stressthe potential to increase stress triaxialityassociated with the reference HRR crack-tip stress(constraint) at the crack tip and thereby reducedistribution and the crack-tip hydrostatic stresssome of the fracture toughness elevation associatedassociated with a specific constraint condition. with shallow flaws. The HSST biaxial test15. The data set of reference 13 covers two

constraint configurations. These are deep flaws program was instituted to investigate this effectwith a crack depth (a) to beam depth (W) ratio (ref. 20).

17. A cruciform test specimen was developed(a/W) of 0.5 and shallow flaws with an a/W ratioof 0.1 Professor R. Dodds of the University of to investigate the effects of biaxial loading on the

• shallow-flaw fracture toughness of pressure vesselIllinois has performed a J-Q analysis of these testspecimens under an HSST Program subcontract, steels. Conceptual features of the specimen areJ-Q trajectories for both the shallow- and deep- illustrated in Fig. 10. The specimen design isflaw beams are shown in Fig. 8 together with a J-Q capable of reproducing a linear approximation ofthe nonlinear biaxial stress distribution shown inloading trajectory from one of the HSST Program Fig. 9. The cruciform design, coupled with awide.-plate tests (ref. 19). Fracture toughness data statically determinate load reaction system, permitsfrom each of the test series are shown in Fig. 8 the specimen to be loaded in either uniaxialsuperimposed on the appropriate J-Q loading (four-point bending) or biaxial (eight-pointtrajectory. Upper and lower bounds to the bending) configurations. Tests of nominallyresulting J-Q scatter band are represented in Fig. 8 identical specimens can thus be performed with theby straight lines. The dual parameter (J-Q) fracture level of stress biaxiality as the only test variable.toughness scatter band shows that both fracture 18. Test values for KJc are shown in Fig. 11toughness and data scatter increase as crack-tip

superimposed on data from the shallow-flaw anddeep-flaw SENB tests. The cruciform beam data

2.0I _ ; J-QTRAJECTORIES• _ , -] 250

',\ w,,,l\ /" " " \ SENB

LOCUS b _ / _ '"_" SCATTER _ _ I %

>. , ,T-RTNDT= -25°Csnd-2B°C SEI_BTM ,_. '_ m

0 I xa/W=0.1 _'_l/-1.0 -0.5 0

NORMALIZEDO IREDUCINGCONSTRAINTI

Fig. 8. J-Q failure locus scatter band shows Fig. 9. PTS loading produces biaxial stresses inincreased toughness and increased scatter as crack- reactor vessel wall with one of the principaltip constraint decreases (Q becomes more stresses aligned parallel with tip of constant-depthnegative), shallow surface flaw.

W. E. Pennell and W. R. Corwin _,

Ip;2 set. The two data points from biaxial loading tests

_ on the cruciform specimen are seen to be closelymatched.

19. Test data generated to date are consistentwith expectations for the biaxial tests. Biaxialloading appears to reduce Kjc below the valuesassociated with shallow flaws, but not down to thelower-bound values associated with highlyconstrained deep flaws. There are also indicationsthat biaxial loading may reduce the scatter ofshallow-flaw fracture toughness data. The biaxialtesting program is, however, in its infancy, and thedata currently available from the program are notsufficient to support an assessment of biaxialloading effects on shallow-flaw fracture toughness

i _ if" SHALLOW at this time.REACTOR.-..._J" -.-'_,._ /"_,.."[ SURFACEVESSEL _ ""-"C_.4,,f__ CRACKINNER I _ ":),r_ I IRRADIATION EFFECTS RESEARCH

SURFACE ,,_./ y "._,,, 20. Klc curve shifts in high-copper welds. To" '" account for the effects of neutron irradiation on

-._ ] A _ouT toughness, the initiation and arrest fracture=N.p_ _ -OF-PLANE toughness curves as described in Sect. XI of theSTRESSES-- / _ ""_ - STRESSES ASME Boiler and Pressure Vessel Code are shifted

Fig. 10. Conceptual features of cruciform shallow- upward in temperature without change in shape byflaw biaxial fracture toughness test specimen, an amount equal to the shift (plus a margin term)

of the Charpy V-notch (CVN) impact energy curveat the 41-J (30-ft-lb) level. Such a procedure

500; • a" W'B'(n_rn) .... HAL' W/' ' !" " " implies that the shifts in the fracture toughness050 100 50 S LO• 10 10050 CRACK t/ /DEEP curves are the same as those of the CVN 41-J

400 • 14 100 50 •/ /CRACK" 50 loo loo / / energy level and that irradiation does not change• 1o loo loo / /

'-" * 50 100 150 B_--T SHIFT - :3R°C the shapes of the fracture toughness curves.1_ 300 , 10 100150 II' " -; ........ 21 The objectivesof the HSSI Fifth andSixth

BCRUCIFORMUNDER • •1 ]a. BIAXIAL LOADING e/ / Irradiation Seriesare to determinetheKlc andKia"-" ®CRUCIFORMUNDER • i/%.UNIAXIAL

_q200 _1_=7 nu_'ru_M_,J curve shifts and shapes for two irradiated high-

UNAXIALLOADING • _ 7-,.............

" copper, 0.23 and 0.31 wt %, submerged-arc welds_,_j,_BIAXIAL (72W and 73W, respectively). Irradiations were

1O0 _L_...._......_ CRUCIFORM (2)_ performed at 288°C (550°F) to average fluences of0 ........................ about 1.5 x 1019neutrons/cm 2 (E > 1 MeV). Tests-_5o -loo -so o 50 loo included tensile,CVN impact, drop-weight,and

T-RTNDT(C) fracture toughness. Compact specimens up to 203Fig. 11. Biaxial loading resulted in shallow-flaw and 101mm (8 and 4 in.) in thickness were testedfracture toughness values below lower-bound in the unirradiated and irradiated conditions,values obtained under uniaxial loading, respectively. The detailed results of testing have

been presented previously (ref. 21). For the CVNplotted in Fig. 11 correspond with the point on the results, the 41-J (30-ft-lb) transition temperaturecrack front at which posttest examination showed shifts were 72 and 82°C (130 and 148°F), whilethe crack to have initiated. The cruciform specimentested under uniaxial loading produced a fracture the 68-J (50-ft-lb) shifts were 82 and 105°C (148toughness close to the lower bound of the fracture and 189°F) for welds 72W and 73W, respectively.toughness scatter band from the shallow-flaw 22. For those fracture specimens that met theSENB tests. Biaxial loading of the cruciform American Society for Testing and Materialsspecimen produced two results below the lower (ASTM_ E 399 criteria for a valid KIo the Kicbound of results from the shallow-flaw SENB data value is used in the analysis. For those specimens

\_,.E.Pcnnclland \V.R.(',u-_vm "

that exhibited curvature in the load-displacement +,........................................................................H581 73W WgI..D l0 31'_ t.l.u

record, indicative of plastic deformation and.. ,.,. ,o,,,....perhaps, stable ductile tearing, the Kjc value was _"° . .:: _ ........... •

used. To include both linear-elastic and elastic- : .....:_ . . {._,..._,,. _,,................. !5'

plastic fracture mechanics calculations, data have :__ . .' _::_"'............... _ _i,__,been designated Kcl for cleavage fracture ::....., . _:,.'"toughness. As ,,,.'asillustrated in Fig. 4, an _ : "_ _'"'"_"'"_'_ : " :unexpectedh, large number of cleavage pop-ins _,_0- J,',_:,.-._",...... . _, 2 ... 'occurred in the irradiated data set. Of 156 z " i .... r',. S, ar, .... ' r '

unirradiated compact specimens, only two :-" ; I,_,_-_,,_,,,...... :-. '. ,, .'• _' '_ , l .--_ _ .."_-..."

exhibitedpop-insascomparedto36pop-insfor _o-- .' .....--.-_..x;.----.- -.the 110 irradiated specimens. To be conservative.

0 _

only the initial pop-in is used herein to determine _00 ,_o ,o_ _,_ ., _o ,oo .._o _ ..._TEST TEMPERATURE I°C_

cleavage fracture toughness for those specimensexhibiting pop-ins. Fig. 12. Fracture toughness, Kcl, versus test

23. Linearized two- and three-parameter temperature for irradiated HSSI weld 73W. Thenonlinear regression analyses similar in form to the ASME Kic curve for the unirradiated data isKlc curve in Sect. XI of the ASME Code gave shown, as is the same curve after shifting it upwardfracture toughness temperature shifts, measured at in temperature equal to the Charpy 41-J shift. Thethe 100-MPa-,/-m(91-ksi i-V]_.)level, of abcat 83 curves labeled 1,2, and 3 represent the ASMEand 99°C (149 and 178°F) for 72W and 73W, curve shifted by the indicated criterion, whererespectively. The analyses show some decreases in margin is 15.6°C. The Kia curve represents theslopes for the irradiated data for both welds. These ASME KIa curve shifted by the Charpy 41-J shift•decreases, however, are only about 4.1 and 6.9% The lq).05 curve is the five-percentile curve for allfor 72W and 73W, respectively, with large enough the HSSI 72W and 73W combined data using thestandard errors to imply a low statistical Wallin procedure.significance of the slope changes. For thecombined data sets, with temperature normalized that the irradiated Kcl curves for these two weldsto RTNDT, the differences are about 10, 15, and appear to have exhibited some shape change after17°C (18, 27, and 31°F) between the unirradiated irradiation. Figure 13 shows a plot of all theand irradiated mean fracture toughness curves at irradiated fracture toughness data for 72W andKcl values of 50, 100, and 200 MPa-fm (46, 91, 73W plotted versus temperature normalized to theand 182 ksii.qq-n7),reflecting the average change in RTNDT. As shown in the figure, a total of eightcurve shape, data points fall below the ASME Klc curve. To

24. Figure 12 shows a plot of the irradiatedfracture toughness data and various curves for _o

73W. The ASME KIc curve is shown for both the "'_"'"°'_°'_"'"_"°"*'"'' ' ° 'o ' ' _iunirradiated and the irradiated conditions after _oo.. ,2,,,o.,._,_ ........................o...... ,

shifting the curve upward. The dashed curves _ -- _"_-_','_¢'_, °', ,-,_-,_,-,-,,,].--.-I ° " l --o _. ........ Ic"vNslulla ._

labeled 1 through 3 represent different methods for _e_5o--I . ,T.,_" ,I _o ..---/" ,'

shifting the Kic curve. The curve labeled 4 __oo ° , . ./ • ,?,_oo.,ol _lJ.:represents the ASME KIa curve shifted upward in _ _'-" _._,__'....................o.__i_..... j

temperature equal to the Charpy 41-J (30-ft-lb) _'_° ° _ °_ ishift (ATT4t). The curve labeled 5 is the five- -_ _--,4 ,_.100

percentile curve produced using the method of _ __ _ ,/_...Wallin (ref. 22). For 72W, the data are bounded _°_......................_ " " ....... -_1by the Aq-T41+ margin curve and the AKcl curve, i

2 _ i ,, i, i _ 1 , ,bu_neither of these curves quite bound all the data to .,_o .,oo ._o o _o ,oo ,_o ,,oo _oTEMPERATURE ('C)

for 73W. The margin is 15.6°C (28°F) asdefined in Regulator3' Guide 1.99 (Rev. 2) Fig. 13. Fracture toughness, Kcl, versus norma-assuming credible surveillance data. The Kla lized temperature, T - RTNDT, for irradiated weldscurve is shown to allow for comparison of that 72W and 73W. The dashed curve is the ASMEcurve wi:h the shifted Kic curves, especially curve shifted upward in temperature to just boundregarding curve shape, in view of the observation the irradiated data.

\k [ tk'm_clland \V t,: ('_,r-,vn_

bound all data, the dashed Kit curve must be combined irradiated data for 72W and 73W

shifted upward in temperature 18°C (32°F). normalized to the RTND T. The mean irradiated25. Observation_ from the HSSI Fifth and Sixth Kcl curve has been shitted much closer to the

Irradiation Series included the irradiation-induced irradiated Ka curve than is the case for thetemperature shift. Statistical analyses and curve unirradiated conditions. The fact that the average

fitting showed that the temperature shifts at a fracture separation in initiation and arrest toughness at anytoughness of 100 MPa_,fn_ (91 ksi-,]_-.. ) were greater given temperature is reduced in the irradiatedthan those at a Charpy energy ot"41 J (30-ft-lb) but condition may help explain the enhancedwere in good agreement with the Charpy 68-J propensity ibr pop-in events following irradiation.(50-ft-lb) transition shifts. The 68-J temperature 27. IrradiatiQn embrittlernent in a commercial

shifts were greater than the 41-J shifts, reflecting the LUS weld. The HSS1 Program includeschange in the slope of the CVN curves following examination of the fracture resistance of LUS

irradiation, welds. This class of submerged-arc welds was26. Results from the HSSI Sixth Irradiation produced using Linde 80 welding flux because it

Series on crack-arrest toughness indicate no produced a very fine dispersion of inclusionsirradiation-induced curve shape changes in the Kla within the weld and a resultant low number ofcurve. Similar shifts were measured at the 41-J reportable defects observed by radiography.(30-ft-lb) level for CVN specimens and the Unfortunately, this fine dispersion of inclusions100-MPax/-m (91-ksi i_qm.) level for KIa (ref. 23). provided such a large number of microvoidFigure 14 shows a comparison of the fracture initiation sites that the macroscopic resistance oftoughness and crack-arrest toughness for the these welds to ductile crack extension by the

microvoid growth and coalescence process wassignificantly reduced. This fine inclusion

III , II I , I I ' I I , l I , 1111 i IIII ,

400 ,HS$1 WELDS 72W AND 73W /_ _ dispersion, in combination with the common early350 -UNIRRADIATED o / I_ MEAN "J

'COMBINED72W & 73W o / practice of using copper-coated welding wire, has..... 300 -MEAN FITS , / resulted in a significant number of pressure vessels

O O

E [ o K_ DATA] . ,'/ : in which major fabrication welds have botha-_ 250 I " K, DATA I Oj_,t /-%. relatively low resistance to ductile fracture in the

200 , ,' ' ' . _,_. ,/ x, uE,m unirradiated condition and a high sensitivity to_" . _%_.._ ,,/ further degradation from neutron exposure.•, 1so 8_._'_,"

• e°# •

°** _ _i;r" , 28. The principal current activity within the" loo i L.hi.l-" :t-,- a - 4_'c HSSI Program to examine LUS welds is the Tenth

_i _ Irradiation Series in which the effects of irradiationso "_'_"(a) ; on the fracture toughness of commercially

o ....... _ ' ' ' --J fabricated LUS submerged-arc welds from the_00 .... ' , , , . , , , . reactor pressure vessel of the canceled Midland

_HSSI WELDS72W AND 73W ' Unit 1 nuclear plant are being investigated. The350 -IRRADIATED,"-28S'C1.5(Kj¢)& 1.9 {K,) o /2 welds from the Midland plant carry the Babcock

_ 300 • x 10 II n/era 2 (>1 MeV) oo /,,r'x,,.. and Wilcox Company designation WF-70, a

_fi COMBINED 72W & 73W o • O/,' K_, MEAN specific combination of weld wire and welding250 -MEAN FITS o _/,'_,....._

_- {o K,, DATAI o_}, ' K, MEAN flux that exists in several commercial pressurized-200:_" I • K, DATA @,_r/'o water reactors. The initial part of this study

l _ o _I_ o involved the determination of variations in

o chemical composition, RTNDT, tensile properties,

" looa _;_." - "'o_" _,- _'c and fracture toughness throughout the welds (ref.so ._.___:_7;_.._,,,,.o • 24). Four 1.17-m-long (46-in.) sections ofbeltline

• (b) weld and two similar sections of nozzle course

o ' _ ' ' ' _ ' ' " ' _ ' weld have been examined. Nil-ductility transition-'5o -_oo -so o so _oo _so

T - RTNoT ('C) temperatures ranged from --40 to ---60"C (-40 to-76°F). Because the Charpy impact energy did not

Fig 14. Comparison of mean fracture toughness achieve 68 J (50 ft-lb) at NDT + 33°C (59°F), the

and crack-arrest toughness versus normalized RTND T values are all controlled by the Charpytemperature for welds 72 and 72W in behavior. The RTND T values vary from -20 to(a) unirradiated and (b) irradiated conditions. 37°C (-4 to +99°F) with position in the vessel

,. W.E. Pcnncll and W. R. Corwin _J

(Fig. 15) while the upper-shelf energies varied aoo , , .-, ....., , C----'-_--T---, ,from 77 to 108 J (57 to 80 ft-lb). Analysis of the • ,_Tcombined data revealed a mean 41-J (30 ft-lb) _. _5o r_ T I_temperature of-8°C (18°F) with a mean USE of _. [] 2T[] 4T !'1

v 200

88 J (65 fl-lb). Even though both welds carry the EWF-70 designation, their bulk copper contentsrange widely, from 0.21 to 0.34 and 0.37 to _ ,so

0.46 wt % in the beltline and nozzle course weld, _ lan

respectively. _ loo _ !i 1 129. Tensile and fracture toughness properties 2_ 17were determined on nozzle and beltline weld _ so

metals at six temperatures ranging from -100 to , , _ , , , ,288°C (-148 to 550°F). The yield strength of the o -,oo -Ts -so -2s o -;oo -Ts -so -2s o

(oc)nozzle weld metal was significantly higher than NozzLE BELTLINEthat of thebelt]ine weld, on theorder of 100MPa(14.5 ksi). All the fracture toughness tests to Fig 16. Mean values of unirradiated transition-characterize the unirradiated material, using temperature-range fracture toughness of MidlandLUS beltline and nozzle course welds as functioncompact specimens ranging up to 101 mm (4 in.)

of test temperature and specimen size.in thickness, have been completed. Data tocharacterize ductile-to-brittle transition

temperature were evaluated using a test standard 30. The irradiation of the Midland weld is incurrently under development by the American progress. The exposure of the first of the two largeSociety for Testing and Materials (ASTM). This irradiation capsules, containing tensile, CVN, andinvolves the determination of the position of a fracture toughness specimens, to the primary targetmedian fracture toughness transition curve (master fluence of 1x 1019neutrons/cm2 (>1 MeV) hascurve), using only the data from six 12.5-mm-thick bcen completed, and the second one has begun.(0.5-in.) (1/2T) compact specimens, to compare Small fracture toughness and CVN specimens arewith data from large specimens. The "reference also being exposed in low- and high-fluencetemperature" for the master curve was found to be scoping capsules to 5 x 1018and 5 x 1019-60"C (-76°F) for the beltline weld and --43°C neutrons/cm 2 to examine fluence effects over this

(--45°F) for the nozzle weld. This appears to agree range.well with the fact that the mean fracture toughnessversus temperature behavior for the beltline weld is INTERIM CONCLUSIONShigher than for the nozzle weld (Fig. 16). 31. Fracture toughness tests on irradiated

material containing local brittle zones gave+o , , , , toughness values that suggest that the current

_ procedure for defining a reactor vessel P-T curve isP_ 1/2 T I

_0 _ ,, _ r I conservative. Shallow-flaw fracture toughness tests,-- k_ 3,/4 T I have shown that both fracture toughness and data_o_ _ _"_+ I

g _ _ n scatter increase as crack-tip constraint is reduced.

_,o I __ _ _ii[ These effects could have a significant positive"_ impact on predictions of reactor vessel failure rates

_ [_ under PTS transient loading. Biaxial loading

_o m I _ u reduces but does not eliminate the shallow-flaw

_-_o effect. There are wide ranges in initial fractureproperties observed in some vessel materials, and

-20 the current methods using CVN-based indices maynot be adequately conservative to account for_3o i ..... £ . I l J .. I _

8L-9 RL-i! BL-,a BL-,S NZ-_ NZ-4 irradiation-induced shifts in fracture toughness.POSITION IN VESSEL,

Fig 15. Distribution of RTNDT values for different REFERENCESsections of Midland LUS beltline (BL) and nozzle 1. Enclosure 3, Regulatory Analysis, Revision 2(NZ) course welds as function of through-thickness to Regulatory Guide 1.99, Radiation Embrittlementposition, of Reactor Vessel Materials, November 1987,

W E. Pcnncll and W, R. Corwin Illw

Available from the U.S. Nuclear Regulatory 16. O'DOWD N. P. and SHIH C. F, TwoCommission public document room, Washington. Parameter Fracture Mechanics: Theory andD.C. Applications, USNRC Report NUREG/CR-59582. Code of Federal Regulations, Title 10, Part (CDNSWC/SME- CR-16-92), Brown Llniversity,50, Appendix H. Reactor Vessel Material February 1993.Surveillance Program Requirements. 17. HUTCHINSON J. W. Singular behavior at3. ASTM Standard E185, Standard Practice for the end of a tensile crack in a hardening material, J.Conducting Surveillance Tests for Light-Water Mech. Phys. Solids 16, 13-31 (1968).Cooled Nuclear Power Reactor Vessels. 18. RICE J, R. and ROSENGREN G. F. Plane

4. Code of Federal Regulations, Title 10, Part strain deformation near a crack-tip in a power-law50, Appendix G. Fracture Toughness hardening material," J. Mech. Phys. Solids 16,Requirements. 1- 12 (1968).5. American Society of Mechanical Engineers 19. NAUS D. J. et al. Crack-Arrest Behavior inBoiler and Pressure Vessel Code, Section X1, SEN Wide Plates of Quenched and TemperedAppendix G, Fracture Toughness Criteria for A533 Grade B Steel Tested Under NonisothermalProtection Against Failure. Conditions, USNRC Report NUREG/CR-49306. Code of Federal Regulations, Title !0, Part (ORNL-6388), Martin Marietta Energy Systems,50, Section 50.61. Fracture Toughness Inc., Oak Ridge National Laboratory, August 1987.Requirements for Protection Against Pressurized 20. THEISS T. J. et al. Initial Results of theThermal Shock Events. Influence of Biaxial Loading on Fracture7. Code of Federal Regulations, Title 10, Part Toughness, USNRC Report NUREG/CR-603650, Section 50.36. Technical Specifications. (ORNI.,/TM-12349), Martin Marietta Energy8. _J.S. Nuclear Regulatory Commission Systems, Inc., Oak Ridge National Laboratory,Regulatory Guide 1.154, Format and Content of June 1993.Plant-Specific Pressurized Thermal Shock Safety 21. NANSTAD R. K. et al. Irradiation Effects onAnalysis Reports for Pressurized Water Reactors. Fracture Toughness of Two High-Copper9. ASME Code Case N-512, Assessment of Submerged-Arc Welds, HSSI Series 5, USNRCReactor Vessels with Low Upper Shelf Charpy Report NUREG/CR-5913 Vol. 1 (ORNL/TM-Impact Energy Levels, Section XI, Division 1, 12156/V 1), Martin Marietta Energy Systems, Inc.,February 12, 1993. Oak Ridge National Laboratory, October 1992.10. Draft Regulatory Guide DG-1023, Evaluation 22. WALLIN K. The scatter in KIc results, Eng.of Reactor Pressure Vessels with Charpy Upper- Frac. Mech. 19(6), 1085-1093 (1984).Shelf Energy Less Than 50 ft-lb, U.S. Nuclear 23. ISKANDER S. K., CORWIN W. R., andRegulatory Commission, September 1993. NANSTAD R. K. Results of Crack-Arrest Tests11. McCABE D. E. et al. Investigation of the on Two Irradiated High-Copper Welds, USNRCBases for Use of the Klc Curve, ASME PVP-Vol. Report NUREG/CR-5584 (ORNI.2TM- 11575),213/MPC-Vol. 32, Pressure Vessel Integrity-1991, Martin Marietta Energy Systems, Inc., Oak Ridgepp. 141-148, June 1991. National Laboratory, December 1990.12. RICHIE R. O., KNOTT J. F., and RICE J.R. 24. NANSTAD R. K. et al. ChemicalOn the Relationship Between Critical Tensile Composition and RTNDT Determinations forStress and Fracture Toughness in Mild Steel, J. Midland Weld WF-70, USNRC ReportMech. Phys. Solids 21,394--410 (1973). NUREG/CR-5914 (ORNL-6740), Martin Marietta13. THEISS T. J., SHUM D. K. M. and ROLFE Energy Systems, Inc., Oak Ridge NationalS.T. Experimental and Analytical Investigation Laboratory, December 1992.of the Shallow-Flaw Effect in Reactor PressureVessels, USNRC Report NUREG/CR-5886(ORNL/TM-12115), Martin Marietta EnergySystems, Inc., Oak Ridge National Laboratory,July 1992.14. DODDS R. H. Jr., ANDERSON T. L., andKIRK M. T. A framework to correlate a/W ratioeffects on elastic-plastic fracture toughness (Jc),Int. J. Frac. 48, 1-22 (1991).15. O'DOWD N. P. and SHIH C. F. Family ofcrack-tip fields characterized by a triaxialityparameter: Part 1--Structure of fields, J. Mech.Phys. Solids 39, 989-1015 (1991).

I