ML17059A031

Elastic Plastic Fracture Mechanics Assessment...Nine Mile Point,Unit 1:Response to NRC RAI Re GL 92-01

ML17059A031
Person / Time
Site:	Nine Mile Point
Issue date:	08/31/1993
From:	NIAGARA MOHAWK POWER CORP.
To:
Shared Package
ML17059A032	List: ML17059A031 ML18040A095
References
GL-92-01, GL-92-1, TAC-M83486, NUDOCS 9309140275
Download: ML17059A031 (74)

Text

NineMilePointUnit1DocketNo.50-220DPR-63TACNo.M83486GenericLetter92-01ElasticPlasticFractureMechanics Assessment forNineMilePointVnitI:ResponsetoNRCRequestforAdditional Information

'August,1993,9309i40275 930908PDRADOCK05000220PDR

TABLEOFCONTENTS

1.0INTRODUCTION

......

2.0 RESPONSES

TOENCLOSURE 1REQUESTSFORADDITIONAL INFORMATION

-SERVICELEVELSAANDB2.1Information Request1.-J-RModel2.2Information Request2.-Mechanics Model2.3Information Request3.-EffectofCladding~..............

~.....5515163.0RESPONSES TOENCLOSURE 2REQUESTSFORADDITIONAL INFORMATION

-SERVICELEVELSCANDD3.1Information Request1.-Temperature Dependencies 3.2Information Request2.-95%Confidence Properties

...............

3.3Information Request3.-J-Material Values3.4Information Request4.-Transient Duration.....~....~~.........

3.5Information Request5.-ThermalTransient Parameters 3.6Information Request6.-CladEquivalent Stress3.7Information Request7.-StressIntensity FactorEquation...........'

.3.8.Information Request8.-SampleCalculation 18182122242631323

44.0REFERENCES

.........

37Appendix-ExampleLevelCFlawStability Calculation

........,...

~.....~...38

1.0INTRODUCTION

NiagaraMohawkPowerCorporation (NMPC)submitted theReference

[MA92]reporttotheNRCbyletterdatedOctober16,1992.CommentsprovidedbytheNRCwereincorporated intotheanalysisandarevisedreportwassubmitted onDecember17,1992[MA92b].TheNRClaterconcurred withNMPCthattheA302Bmaterialmodelisappropriate foranalysisoftheNineMilePointUnit1(NMP-1)beltlineplates,andareport[MA93]waspreparedwhichcontainsonlytheA302Bmaterialmodel(theA533Bmodelwasdeleted).

The[MA93]reportwasnotsenttotheNRCbecausetheA302Bmodelandresultsareidentical tothosereportedinReference'[MA92b].

Thesesubmittals containaplant-specific elastic-plastic fracturemechanics assessment forNMP-1underServiceLevelAandBloadings.

AreportwhichcontainstheresultsforServiceLevelCandDloadings[MA93b]wassubmitted totheNRConFebruary26,1993.Theanalysesdescribed inthesereportswereperformed inaccordance withthedraftASMEAppendixX[ASME92],

anddemonstrate thatsufficient marginsofsafetyagainstfractureexistthroughend-of-license (EOL).InaletterdatedJuly22,1993,theNRCindicated thatapreliminary reviewofthesereportshasbeencompleted andthatadditional information isrequiredtocompletethereview.ThisreportwaspreparedinresponsetotheNRC'srequestforadditional information andisfullyresponsive toallinformation requestsprovidedinEnclosures 1and2oftheJuly22,1993letter.

0

2.0 RESPONSES

TOENCLOSURE 1REQUESTSFORADDITIONAL INFORMATION

-SERVICELEVELSAANDB2.1Information Request1.-J-RModel"Thereportindicates thattheJ-Rcurvefora6Tspecimentestedat180'1'isdrawntomeettheJaxisatJc=525in-lblin',

thenthiscurveisshif)eddowntomaketheJpointcoincidewiththeestimated Jicpoint,leavingthedifference betweentheplateaulevelofJandJicconstantat175in-Iblin',

independent ofbothtemperature andUSE.Providejustification fortheassertedindependence oftheJdifference (175in-Iblin) withrespecttotemperature andUSEvalues.AlsojustifythattheproposedJ-Rmodelshouldbreakdown whenUSEvaluesreachzero.(Although thisissuewasaddressed inatelephone conference heldinJanuary1993,awrittenresponseisrequired)"

RESPONSE

~BackcaadIncontrastwiththeJ-Rcurvedatatrendsforotherpressurevesselmaterials, Reference

[H189]reportedanunprecedented sizeeffectforA302Bsteel.AsshowninFigure2.1-1,thethickerthespecimen, thelowertheJ-RresponselevelaAerinitiation.

Whilesimilardatatrendshavebeenobservedforsomepressurevesselmaterials, decreases intheJ-Rcurvesofthemagnitude reportedbyHiserhavenotbeenreportedearlier.Basedonchemicalandmicrostructural considerations, itwasdetermined thatthemodifiedA302B(A302M)NMP-1plateswouldexhibitductilefracturebehaviorsimilartothatpresented inReference

[HI89].Reference

[HI89]reportedJ-Rdatafor0.5T,1T,2T,and4Tspecimens, butonlyone6Ttestwasperformed (180'F,T-Lorientation).

Themicromechanical explanation fortheJ-RcurvebehaviorshowninFigure2.1-1hasnotbeendefinitively established.

Hiser[HI89]hasreportedbrittle-like splits,orlaminatetearing,forallofthespecimens tested.Thesesplitsareorientedinthedirection ofcrackgrowthwithsmallamountsofmicrovoid coalescence intheregionbetweenthesplits.Thesize,relativenumber,anddistribution ofthesplitsareapproximately constantforvariousspecimensizes.Hiserconcluded thatthesplitsresultedfromseparation of,theinterface betweenthematerialmatrixandtheinclusions (sulfides, aluminides) and/orthesplitting ofthemorebrittlealloyrichbondedstructure (possibly bainite).

Theonlyapparentdifference inthefractureofsmallandlargespecimens isthetotalnumberofsplitsandnottherelativeproportion, Acomplete~micromechanical explanation isnotyetavailable.

Reference MA92AnalsisSincetherearenotsufficient thick-specimen data(6Tto8T)available atpresenttodefinitively establish therelationship betweenJ<<andtheJplateau(hJ),asafunctionoftoughness level(inparticular, USElevel),theReference

[MA92]analysiswasperformed assumingthatthedifference betweentheplateaulevelofJandJ<<isaconstantequalto175in-

Ib/in'over therangeofUSElevelsfrom10it-lbsto100A-lbs).Atthetimetheanalysiswasperformed, itwasrecognized thatthe175in-lb/indifference maychangesomewhatasthetoughness ofthesteelvaries.HowevertheUSElevelforthissteelis52ft-lbs(T-L),whichisroughlyinthemiddleoftherangeoverwhichtheJ-Rcurvescalingwasdone.Therefore, itwasjudgedthatthedifference betweentheactualmaterialbehavior, andthematerialmodelbasedontheassumption ofaconstantB,J=175in-lb/in',

wouldbesmallandadequately represented byotherconservatism inthemodel.SincethereisnophysicalbasisuponwhichtovaryhJastheUSElevelischanged,thechoiceofaconstanthJobtainedfrom6Tdataisareasonable modelling assumption.

6JCharacterization TheNRChasrequested thatjustification fortheconstantb,Jusedinthe[MA92]calculations beprovided.

Unfortunately, asdiscussed above,withoutextensive additional testingandanalysis, completejustification cannotbeprovided.

Inparticular, sincetheplateauforthe6TA302Btestissolowat52A-lbs,itispossiblethattheh,Jvariation atlowerUSElevelsmaynotscale,inthesamemannerasotherRPVmaterials.

Intheabsenceofadditional data,calculations havebeenperformed using0.5Tand1TdatatoassessthehJvariation atlowtoughness.

Sinceitislikelythatthesedataareconservative incomparison with6TA302Bdata,thecalculations providedbelowshouldbeviewedasworstcaseimpactassessments.

Inandefforttocharacterize theh,Jvariation withtoughness, 0.5Tand1TdatafromReferences

[MEA90]and[MEA83]wereanalyzed.

Thephysicalcrackextension (ha,)fortheanalysesreportedinReference

[MA92]isontheorderof0.1in.Therefore, 6Jforthe0.5Tand1Tdatawascalculated bysubtracting J<<fromJatha;-0.1in.(J).Itisimportant tonotethatthethinspecimens atintermediate tohightoughness levelsdonotexhibitaplateauatsmallh,aaswiththe6TA302Bdata.However,thesmallspecimendatacanbeusedtoobtainanestimateofthelLJvariation withtoughness.

Infact,atthepresenttime,thisistheonlymethodavailable forcharacterizing thehJvariation.

Thesedataarepresented inFigure2.1-2.TheReference

[MEA83]J-Rpowerlawformulation wasusedtomodelthedatashowninFigure2.1-2.Themodel,determined fromleastsquaresregression, isgivenby:where,J=C(ha)'=J-Integral (in-lb/in')

C=1000[-0.4876 (USE/100)+

7.5611(USE/100)']

(in-lb/in')

ha=crackextension (in)n=0.267(C/1000)'""

Figures2.1-3and2.1-4illustrate thefunctional formofCandn.TheresultsobtainedusingthepowerlawmodelareshowninTable2.1-1andinFigure2.1-2.Themodelrepresents the0.5Tand1Tdatawell,andapproaches aphysically meaningful limitatlowtoughness.

Asexpected,

themodelshowsthataconstanthJ=175in-lb/in's conservative forUSElevelsaboveabout40ft-lb,butissomewhatnon-conservative forUSElevelsbelow40ft-lb.Inordertoassesstheimpactofadecreasing hJwithtoughness, thefollowing materialmodelwasanalyzed:

USEft-ib10203040-100~EJinib/in/02082175Theabovedescribed J-Rmaterialmodelisthesameasthatdescribed inReference

[MA92];exceptthatbelow40ft-ibtheb,Jvariedinaccordance withtheabovelisteddata.TheresultsofthisanalysisareshowninTable2.1-2.Reviewofthesedatashowsthatevenifh,Jweretodecreasedramatically atUSElevelsbelow40ft-lb,theminimumallowable USEisbelowtheprojected materialUSEatEOL.MaterialModelTemeratureDeendenceWithregardtothequestionoftemperature dependence oftheJ-Rcurves,the6TJ-Rtestat180'F[HI89]isexpectedtoconservatively represent thematerialbehavioruptoreactoroperating temperature.

AsshowninFigure2.1-5,the6Ttestwasperformed atatemperature slightlyhigherthantheon-setoftheuppershelf.TheCharpydataindicatetemperature independence fromabout165'Fuptoreactoroperating temperature.

NMPCPositionItisNMPC'spositionthattheresultsoftheAppendixXanalysisreportedinReference

[MA92]areaccurateandconservative.

Atpresent,therearenotsufficient dataavailable tocharacterize thevariation ofh,Jwithtoughness forthicksectioncomponents.

Therefore, theuseofaconstant4J=175inlb/in'sreasonable andisexpectedtoyieldamaterialmodelwhichaccurately represents thicksectionbehavior.

~J-USEModelBehavioratLowTouhnessTheJ-RmodelfortheA302Bmaterialreliesonthecorrelation ofJ<<withUSEasshowninFigure4-12oftheDecember17,1992submittal.

IfitwerepossibletoproduceamaterialwithUSE=0(i.e.,noenergyrequiredtodriveacrack),thenJ<<mustalsobezero(i.e.,'ocrackdrivingforcerequired).

Therefore, thetheoretical limitforaJ<<vs.USEcorrelation astoughness decreases istheorigin.Thisdatatrendisclearlydemonstrated inFigure4-12.However,asapractical consideration, theUSEforferriticRPVsteelswouldnotbeexpectedtodropbelowthelowershelfenergylevel.Reference

[MEA90]showsthatthelowershelfforA302Bsteelisintherangeof4-18ft-lbs.Therefore, asthematerialtoughness decreases, theJ,c-USEcorrelation isexpectedtodescribethematerialfracturebehaviorastheUSElevelapproaches theCharpylowershelfenergylevel.

A302BJ-RDATAFORVARIOUSSPECIMENTHICKNESSES

'15001000XlICO05500CDO0k~~;~kaJ~~~~~~~+MID~~cI0.5TDATA<0.5TDATA00.5TDATA40.5TDATA*0.5TDATA*0.5TDATA01TDATA+1TDATAo2TDATAa2TDATA<4TDATA44TDATA~6TDATADeltaa(ln.)Figure2.1-1Comparison ofJ;RCurvesforA302BPlate(DataTakenfrom[HI89])

J-RCurveDeltaversusJicA302BandA533BMaterial30002500Ol4tOIQI20005500C)u1000cd500PowerLawModel100020003000Jtc(in-Ibs/in**2)

T-L~L-T+A533A302A302Figure2.1-2h,JasaFunctionofJ,cfor0.5Tand1TSpecimens

HUCLERRVESSELSTEELS288C,1T~,28-25%SGFILLEDSYNOLSRREIRRRDIRTED

~-lKUNENTS~i-HROOGHTN4gkgJ~wCS~~C/18887.5611%(Cv/188)

+2".4878%(cv/188)

I.BCv/IBB(ft-lb)2.BFigure2.1-3Correlation ofNormalized Coefficients withNormalized CharpyUpperShelfEnergyValuesPvKA83]10

1.88NUCLERRVESSELSTEELS288oC~iTCT~2825~SGFILLEDSYMBOLSRREIRRRDIRTED

.68ggSgWggSOhgh0glhn-8.2SC(CiBBB)82'62aa-gQMENTS>>-NROUGHT8.888812Cti888(4roaEq.3-i)28Figure2.1-4Correlation ofPowerLawExponent"n"withCoefficient "C"[MEA83]

188TEHPERRTURE

('F)I88288388A302-8PLATE(V50)NewData6858previousData48382818188TEHPERRTURE

('C)Figure2.1-5Comparison oftheAverageCurvefits totheNewandthePreviousCDatafortheA302-BPlate.TheNewDataIndicateHigherOverallToughness, withaHigherUpperShelfEnergyLevelandLowerTransition Tempeiatures.

IHI89]12

Table2.1-1PowerLawModelforb,JasaFunctionofToughness USE253040506080100J(0.1)(in-lb/in')

223321547807109117092360SmallSpecimenDataJic.(in-lb/in')

199239319399479639798hJ(in-lb/in')

248222840861210701562h,JUsedin[MA92](in-lb/in' 17517517517517517517513 0

Table2.1-2EAectof4JVariation ontheMinimumUpperShelfEnergyLevelforNMP-1PlateG-8-1PlateASMEServiceLevelMaterialModelHawGrowthof0.1in.Criterion Ji~Jo.iHawStability Criterion MinimumUSE(Ft-lbs)4J=175in-ib/in'law Growthof0.1in.Criterion Ji~o.iFlawStability Criterion MinimumUSE(Ft-Ibs)Variable4JG-8-1G-8-1G-8-1A8cBDA302BA302BA302B1310231020333136313014 0

2.2Information Request2.-Mechanics ModelltTheieportcontainsnodescription ofthefracturemechanics analysisprocedure, i.e.theequations usedforcalculating J,>,T,>,andP~,.Onlythenameofacomputer-programismentioned.

EitherconJirmthattheequations usedareidentical tothoseinAppendixXorlistalltheequations whichdier."RESPONSE:

Asmentioned inSection3.0ofReference IMA92],theprocedure andequations specified inAppendixX[ASME92]forServiceLevelsAandBareidentical tothoseusedtocalculate theappliedJ,theappliedtearingmodulus,andinternalpressureatflawinstability, undertheJ-Integral/Tearing ModulusProcedure.

15 C

2.3Information Request3.-EffectofCladding"Provideinforniation regarding theeffectofcladdingtothecalculated appliedJvalue."RESPONSE:

~BackcuadReference

[ASME92]doesnotexplicitly recommend norrequirethatcladstresseffectsbeincludedintheServiceLevelAandBanalysis.

Discussions withseveralmembersoftheASMEWorkingGrouponFlawEvaluation (WGFE)indicated thattheeffectsofcladdinghavebeendiscussed, butthegroupdoesnotplantorecommend incorporation ofcladstressanalysisprocedures intoAppendixX.ASMEarticleA-3000,"MethodforK,Determination",

doesrequireconsideration ofresidualandappliedstressofallforms,including clad-induced stress,tobeincludedinstressintensity factorformulation.

Therefore, NMPCincludedcladinducedstresseffectsforServiceLevelCandDloadings, becausetheServiceLevelCandDanalysesrequirecalculations tobeperformed forshallowsurfaceflawswherecladinducedstresscanbesignificant.

However,cladstresseffectswerenotincludedintheServiceLevelAandBanalysesbecause1/4Tflawsarepostulated intheseanalysesandthecladinducedstresswereassumedtobenegligible.

Estimated CladInducedStressEffectInresponsetotheNRCinformation request,theefFectofcladdingontheappliedJforServiceLevelAandBloadingshasbeenestimated.

Surfacetensilestressesresultfromdifferential thermalcontraction fromthestressreliefheattreatment at1150'F.Alinearelasticmodelwasformulated tocalculate thestressresulting fromcooldownfrom1150'F,andthemodelpredictsthatthehoopstressesexceedyieldbeforethevesselIDtemperature reaches100'F.Anelastic-plastic finiteelementanalysisofthecooldownfrom1150'Ftoroomtemperature, followedbyre-heating to528'F,withasubsequent 100'F/hrcooldown, wasperformed.

Theresultsofthefiniteelementanalysisconfirmed theanalytical modelprediction ofa36ksihoopstressinthecladduetodifFerential thermalcontraction whenthecooldownofthevesselwasterminated atavesselIDtemperature of100'F.Thestressintensity atthe1/4Tflawduetothecladstress~~)wascalculated andfoundtobe6.6ksiVin.Thestressintensity modelincludestheeffectsofthebasemetalcompressive reactionforce.Theminimumallowable USEwascalculated byaddingK~tothestressintensity factorsdefinedinAppendixX.TheAppendixXcalculative procedures werefollowedandtheevaluation criteriaapplied.Theresultsofthesecalculations areshowninTable2.3-1.ReviewofthesedatashowsthatifcladstresseffectswereincludedintheServiceLevelAandBanalysis, theminimumallowable USEisbelowtheprojected materialUSEatEOL.r16

Table2.3-1EffectofCladStressontheMinimumUpperShelfEnergyLevelforNMP-1PlateG-8-1PlateASMEServiceLevelMaterialModelMinimumUSE(Ft-Lbs)WithoutCladStressEffectMinimumUSE(Ft-Lbs)WithCladStressEffectG-8-1A&BA302BFlawGrowthof0.1in.Criterion Ji~Jo.i13FlawStability Criterion 23FlawGrowthof0.1in.Criterion Ji~o.i26FlawStability Criterion 3717

3.0 RESPONSES

TOENCLOSURE 2REQUESTSFORADDITIONAL INFORMATION

-SERVICELEVELSCANDD3.1Information Request1.-Temperature Dependencies "Thereportindicates inSection4.1thattemperature dependent properties wereusedinthethermalandstressanalyses.

Providethedetailsofthesetemperature dependencies."

RESPONSE

Table3.1-1showsthetemperature dependent properties referredtoinSection4.1ofReference

[MA93b].ThefiniteelementsoAwarePVELD3]useslinearinterpolation withinthematerialpropertytables.Thevolumetric heatcapacity(c)isrelatedtospecificheat(C,)anddensity(p)by:c=pC,Theinstantaneous coefficient ofthermalexpansion isdefinedintermsoftheslopeofthethermalstrainversustemperature curve:dera=-dTTheinstantaneous coefficient isdifferent fromtheaveragecoefficient whichisperhapsmorecommonly'sed.

Whiletheaveragecoefficient musthaveanassociated reference temperature (thetemperature atwhichthermalstrainiszero),theinstantaneous valuedoesnot.Table3.1-2showstheaveragecoefficient ofthermalexpansion thatwasautomatically generated bythefiniteelementsofbvarefromtheinputinstantaneous values.Thevaluesbasedonareference temperature of1150'Fwereusedincomputing theinitialresidualstressstateduetoslowcoolingfromastress-free condition at1150'Fto528'F.Thevaluesbasedona'reference temperature of528'Fwereusedforthetransient thermalanalysesassociated withLevelCandLevelDloadings.18

Table3.1-1Temperature Dependence ofMaterialProperties Temperature Conductivity Vol.HeatCapacityElasticModulusPoisson's RatioInst.Coef.Th.Exp.(T):(k):(G)(E):(v)-(a).OFBtu/in/sec/'F Btu/iq/'Flh/innondimensional 1/oFStainless steelTk50.0.000182300.0.000212550.0.000242750.1000.1300.claddingc0.03120.03460.0371(type304)E28700000.

27100000.

25800000.

24200000.

22500000.

20200000.

0.260.280.310.320.300.280.00000816

0.0 0000894

0.00000960

0.0 0001003

0.00001056

0.0 0001141

A302BhasemetalTkc50.300.550.750.1000.1300.0.0005340.02980.0005720.03410.0005530.037630000000.

29000000.

27700000.

26200000.

24500000.

22200000.

0.280.280.280.280.280.280.00000607 0.00000710=

0.0 0000816

0.00000894

0.0 0001000

0.00001100 NOTE:Dataforkandcattemperatures above550'Farenotprovidedsincethermaltransient analyseswereperformed attemperatures below550'F.19

Table3.1-2AverageCoefficients ofThermalExpansion forReference Temperatures of1150'Fand528'FStainless steelcladding(type304)a,(1/oF)50.300.550.750.1000.1300.A302Bhasemetal1150oF9.64330E-06 9.96485E-06 1.02544E-05 1.04741E-05 1.07725E-05 1.11975E-05 528oF8.87958E-06 9.24096E-06 9.57096E-06 9.79082E-06 1.00579E-05 1.04181E-05 6,(1/oF)50.300.550.750.1000.1300.1150~F8.33523E-06 8.85000E-06 9.35833E-06 9.76250E-06 1.02500E-05 1.07500E-05 528DF7.06121E-06 7.58336E-06 8.11336E-06 8.50673E-06 9.01694E-06 9.59326E-06 20

3.2Information Request2.-95%Confidence Properties "1'igure4-12inthereportdatedDecember17,1992,andinapreviousreportdatedOctober16,1992,indicates thattheMean-2oproperties andthe95%confidence properties (Mean'-1.645o)giv'ethesamelowerboundline.ClarifythisandconfirmthatMean-2aproperties havebeenusedforLevels2,8,andCanalyses."

RESPONSE

TheOctober16,1992,reportisbasedon95%lowerboundconfidence limits.Inparticular, the95%lowerboundJ<<valuesshowninFigure4-12werecalculated using:Jic=3.1(USE),USE(75ft-lbsJ<<=-363.4+7.93295(USE),USE>75ft-lbswhere,J<<=in-ib/in'SE

=ft-IbTheportionofthemodelbetweentheoriginand75ft-1bswasdetermined basedonconservative

~~engineering judgement.

Theportionofthemodelabove75ft-lbscomesfromtheregression analysisandrepresents the95%confidence lowerbound.InresponsetotheNRC'srequest,the95%confidence lowerboundwas.replaced byatwosigmalowerboundconfidence intervalandthismodelwasdescribed intheDecember17,1992,submittal.

Thetwosigmalowerboundmodelisgivenby:Jic=31(USE)~USE(75ft-lbsJic=-363.4+7.915(USE),USE>75ft-lbsTheportionofthemodelabove75ft-lbscomesfromtheregression analysisandrepresents thetwosigmalowerbound.Theportionofthemodelbelow75ft-lbsisbasedonengineering judgement andisidentical tothemodelusedintheOctober16,1992report.ItisNMPC'spositionthatthemodelusedbelow75ft-lbsismoreconservative thanatwosigmalowerboundlevel.SincetheJ-g.curvemodelbelow75ft-ibsusedintheOctober16,1992,reportisthesameasthatusedintheDecember17,1992,report,andtheminimumallowable USEisbelow75ft-Ib(calculations yielded23ft-lbs),theminimumallowable USEwhichwascalculated didnotchangewhenthetwosigmamodelwasused.Insummary,mean-2oproperties havebeenusedforServiceLevelA,B,andCanalyses.

21

3.3Information Request3.-J-Material ValuesIITheJmatenalvaluesat0.1inchlistedinTable5-3arelowerthanthecorresponding valuesinFigures5-1to5-4and5-7to5-10intheLevelsAdcBreportbyapproximately 6lbs.Explainthisdifference."

RESPONSE

Asdescribed inReference

[MA93b],pointwise experimental data,scaledtoaccountforthetoughness level,wereusedintheanalysis.

TheUSE(3.0)codeusesamulti-linear representation withinterpolation whenthepointwise inputoptionisused.Asanexample,thematerialJp,datuminTable5-3ofReference

[MA93b]at30ft-ibs(J=261in-lb/in')

wasdetermined byinterpolating thepointwise J-Rdata.ThematerialmodelinputforthiscaseisshowninTable3.3-1.ThedatainTable3.3-1showsthattheplateaubeginsat4a=0.112in.withJ=267.4in-Ib/in'.

Thus,theapparentdiscrepancy isanartifactofthepointwise model.Carefulexamination ofFigures5-1to5-4and5-7to5-10oftheReference IMA92]reportshowsthattheinterpolated J-material valuesat0.1inchhave'been correctly calculated andtheJ-Rcurvesarecorrectly plotted.22

Table3.3-1USE3.0OutputListingShowingJ-RCurvePointwise Input02/15/1993 15'-NPIP-1PLATEG"8"1A302BMATERIALNODELANALYSISCURVE¹JIC=92.4k'44$g4444$)kff4$)k$)gg)kgb)kg)kgb)}'4$

$$$)kgffffgggggg)gg)l(4444)kg$

$$44$44444441.Corresoondina uooershelfenerovUSE=30(ft-Ibs)¹1¹2¹3:¹¹5:¹6:¹7:¹8:¹9:¹10¹11:¹12¹13:¹14¹15:¹16:¹17¹18:19:¹20:Deltaa0.0020.0040.0050.0060.0100.0170.0170'220.0230.0250.0300.0320.0360.0430.0480.0560.0680.0730.0830.098JDelta.a21.600¹21:0.112~33.400¹22:3.00055.30075.00095.000109;-400116.400136.400144.400154.400165.400183'00191.400201.400210.400218.400225.400240.400247.400260.400J267.400268.00023

3.4Information Request4.-Transient Duration"LevelsCandDtransients mustbeanalyzedfromthebeginning ofthetransient tothetimeatwhichthemetalatthetipoftheJlawbeinganalyzedreachesatemperature equivalent totheadjustedRT>>rplus50'F.Confirmthatthispracticehasbeenadoptedorproviderevisedanalyses.

"RESPONSE:

ForserviceLevelsCandD,theARTNDTforplateG-307-4rangesbetween144'Fand163'Ffromthe1/4TpositiontotheIDsurfaceat18EFPY.Therefore, theARTNDrplus50'Fwouldrangefrom199'Fto210'F.Theblowdowntransients areterminated whenthepressurereaches35psigtoaccountforthecontainment pressurelevelatthattimeinthetransient.

IntheReference PdA93b]thermalstresscalculations, thesetransients wereextendedtolongertimes,conservatively assuminga300'Fperhourcooldowntoa212'FvesselIDtemperature.

Thus,theLevelCandDtransients werenotanalyzedtoatemperature equivalent totheARTNDrplus50'Fattheflawtip.However,asdiscussed inReference IMA93b],thelimitingtransients experienced peakthermalandmechanical loadspriortothepointwhenthetransient analysiswasterminated.

Thecooldownfromthefinaltransient conditions toARTNTplus50'Fisacontrolled evolution whichisnotincludedinthetransient definition andisproperlyconsidered asarecoveryaction.Thecooldownfrom212'Fwouldbeboundedbytheemergency cooldowneventandinmostcaseswouldbeboundedbythenormaloperation cooldownanalysis.

ThestandardGEthermalcycletransient definition usedforthedesignbasisemergency andfaultedstressanalysisdoesnotincludeacooldowntoARTNDTplus50'F.ThestandardGEthermalcyclediagramisthebasisforthelimitingLevelC(emergency) andlimitingLevelD(faulted) thermaltransient usedfortheReference PvtA93b]analyses.

ThestandardLevelCandDtemperature andpressuretransient aredefinedbasedonthedesignbasiseventandareterminated whentheeventisstabilized.

Thecooldownfromthefinalstabilized transient condition totheARTNDTplus50'Fiscontrolled byoperatoractionsandemergency operating procedure guidelines.

Ingeneral,theoperatorguidelines includemaintaining thecooldownwithinthe100'Fperhournormalguideline.

ForalltheLevelCtransient conditions, theoperatorcanbeassumedtohavetheabilitytocontroltherecoverycooldownratewithinthenormaloperating guidelines aAertheeventhasstabilized.

ForthelimitingdesignbasisLevelDrecirculation linebreakevent,theemergency operating procedure guidelines includeacontainment floodupwhichoccursovera6to12hourperiod.Containment floodupiscompleted usinglakewaterassumedtobeatthemaximumof81'Fandaminimumofapproximately 35'F.Thelimitingassumption wouldbethatthevesselwalltemperature israpidlycooledfrom212to100'F(ambientcontainment temperature andpressureisapproximated toremaingreaterthan100'Fduetodecayheat).Thislimitingcondition iscloselyapproximated bythenormalcooldownrateassumptions.

24

AssumingtheNMP-1designbasisLOCAscenariowherethereactorisnotreflooded, theultimatecooldownfromsaturated conditions iscontrolled bythecontainment accidenttemperature.

Theprimarycontainment wetwellanddrywelltemperature profileresultsinthedrywellairspacetemperature remaining greaterthat175'Fforapproximately 4hourswithasubsequent slowcooldownrate(muchlessthan100'Fperhourcooldown) linkedtothecontainment heatremovalsystems.Insummary,theLevelCandDtransients werenotanalyzedtothetimeatwhichthemetalatthetipoftheflawreachesatemperature equivalent totheadjustedRT~rplus50'F.However,thelimitingtransients reachedpeakthermalandmechanical loadspriortothepointwherethe.transient analysiswasterminated.

Therefore, theresultsreportedinReference

[MA93b]arethemostconservative resultsforanyoftheServiceLevelCandDtransients.

25 0

3.5Information Request5.-ThermalTransient Parameters "Supplyacompletelistofinputparameters andconditions forthetransient thermalanalysis, including specificheat',thermalconductivity,,density, theresulting valueofthermaldiffusivity, coefficient ofthermalexpansion, elasticmodulusandPoisson's ratio(forbothcladdingandbasemetal);alsotherelationships neededtodetermine theinsidesurfaceheattransfercoeJJicient."

RESPONSE

Theinformation providedbelowdefinestheinputparameters andconditions forthetransient, thermalanalysis.

Thematerialproperties aregiveninTables3.5-1and3.5-2.Specificheats(C,)anddensities (p)werenotinputtothethermalanalysis.

Volumetric heatcapacity(c),theproductofthesetwoparameters, wasinputinstead.Thermaldiffusivity (x)wasalsonotadirectinputtotheanalyses.

However,itwascomputedfromtheconductivity (k)andheatcapacity(c)properties asfollows:x=k/cTable3.5-3summarizes thethermaldiffusivities resulting fromtheconductivities andheatcapacities listedinTable3.5-1.Thetimedependent internalpressureandfluidtemperature boundaryconditions fortheLevelCandDloadingsaregiveninTables3-7(LevelC)and3-8(LevelD)ofthereport[MA93b].Theoutersurfaceofthevesselisassumedtobeinsulated.

Thetimedependent heattransfercoefficient attheinnervesselsurfaceisalsogiveninthesetables.Thefiniteelementsofbvarelinearlyinterpolates (intime)betweentheinputvaluesofinternalpressureandfluidtemperature thatarespecified byTables3-7and3-8ofReference

[MA93b].Theheattransfercoefficients (h),however,arenotlinearlyinterpolated.

Theheattransfercoefficients arechangedinthemodelinastepwisemanner.Forexample,inTable3-7[MA93b],hisheldat10,000untilatimeof380seconds;thenhischangedinstantaneously tothenewvalueof164.Sincehneverincreases duringthecriticaltimesofthesetransients, thisprocedure resultsinlargerhvaluesbeingusedfurtherintothecoolingtransient.

Thisresultsinlargerthermalgradients beingcalculated andthusconservative thermalstresspredictions.

Theheattransfercoefficients ofTables3-7and3-8aregiveninunitsofBTU/hr/ft'/'F.

TheanalysisusedunitsofBTU/sec/in'/'F.

Table3.5-4providesthehvaluesofTables3-7and3-8[MA93b]intheunitsoftheanalysis.

26 0

Table3.5-1Temperature Dependence ofMaterialProperties Temperature Conductivity Vol.HeatCapacityElasticNoduluspoisson's RatioInst.Coef.Th.Exp.(T):(k):(c):(E):(v).(a)oFBtu/in/sec/'F Btu/ip/Flb/innondimensional 1/0F50.300.550.750.1000;1300.0.0001820.0002120.0002420.03120.03460.037128700000.

27100000.

25800000.

24200000.

22500000.

20200000.

A302BbasemetalTkStainless steelcladding(type304)TkcEV0.260.280.310.320.300.280.00000816

0.0 0000894

0.00000960

0.0 0001003

0.00001056

0.0 0001141

50.300.550.750.1000.1300.0.0005340.0005720.0005530.02980.03410.037630000000.

29000000.

27700000.

26200000.

24500000.

22200000.

0.280.280.280.280.280.280.00000607

0.0 0000710

0.00000816

0.0 0000894

0.00001000

0.0 0001100

Dataforkandcattemperatures above550'Farenotprovidedsincethermaltransient analyseswereperformed attemperatures below550'F.27

Table3.5-2AverageCoefficients ofThermalExpansion forReference Temperatures of1150'Fand528'FStainless steelcladding(type304)0,(1/F)aveA302Bbasemetal50.300.550.750.1000.1300.11500F9.64330E-06 9.96485E-06 1.02544E-05 1.04741E-OS 1.07725E-05 1.11975E-OS S28'F8.87958E-06 9.24096E-OG 9.57096E-OG 9.79082E-OG 1.00579E-05 1.04181E-05 0,(1/F)50.300.550.750.1000.1300.11500F8.33523E-06 8.85000E-06 9.3S833E-06 9.76250E-OG 1.02500E-05 1.07500E-05 528oF7.06121E-06 7.58336E-OG 8.11336E-06 8.50673E-06 9.01694E-06 9.59326E-OG

Table3.5-3ThermalDiffusivity Diffusivity (K):in/secStainless steelcladding(type304)TK=50.5.83E-03300.6.13E-03550.6.52E-03A302BbasemetalTK50.1.79E-02300.1.68E-02550.1.47E-0229

Table3.5-4HeatTransferCoefficient Conversion BTU/hr!ft

/'FBTU/sec'/in l'F69,18810,0005001641.33E-011.93E-029.65E-043.16E-0430 0

3.6Information Request6.-CladEquivalent Stress"Supplythedetailedcalculation procedure fordetermining thecladequivalent stressvalueslistedinTable5-1."The"Extrapolated SurfaceStress"columninTable5-1ofReference IMA93b]isthestressatthepressurevesselIDsurfaceobtainedbyfittingthebasemetalfiniteelementcalculated stressdistribution tothefollowing

equation, a=Ao+A,X+A2X'A,X'here, A;=regression constants X=distancethroughthewallandextrapolating totheIDsurface.The"CladStressMinusExtrapolated SurfaceStress"columnisthedifference betweenthediscontinuous cladstressduetocooldownfromreactoroperating temperature duringthetransient andtheextrapolated basemetalstressatthesurface.The"Residual Stress"columnisthetensilestressinthecladduetocooldownfrom1150'Ftoreactoroperating temperature duringfinalstressrelief.The"CladTotalStress"columnisthesumofthe"CladStressMinusExtrapolated SurfaceStress"dataandtheclad"Residual Stress"data.The"CrackSurfacePressure" columnisthestressonthecrackfacesduetocoolantpressure.

The"CladEquivalent LineStress"columnwasobtainedbymultiplying the"CladTotalStress"bythecladthickness (5/32in.)toobtaintheequivalent linestressforthestressintensity model,andaddingthe"CrackSurfacePressure" timesthemaximumanticipated flawdepth(1.0in).Itisrecognized thatthe"CrackSurfacePressure" maybeaddedtothebasemetalfiniteelementcalculated stressdistribution andthenfitasdescribed earlier.However,theabovedescribed procedure isconservative andcomputationally simpler.31 00 3.7Information Request7.-StressIntensity FactorEquation"Providethederivation orthereference (indicating thepagenumber)ofEquation(5-3)."RESPONSE:

Equation5-3ofReference

[MA93b]canbefoundinthefollowing reference TheStressAnalsisofCracksHandbook, Tada,H.,Paris,P.,Irwin,G.,DelResearchCorporation, June,1973,page2.27AcopyoftheTadamodelisshowninFigure3.7-1.32

-2.27-F(~cQ~)3.52('l-/~)<.3o-5.28+~

(I-N)s/-"(I-4)'2<-(Fa)*PCdt7IIIIZIOl~oOUi+JII02II0./0.6C/g0.8Method:Estimated byInterpolation Accuracy:

F(c/a,a/b)-foraula

$sexpectedtohave2Xaccuracyforanyvaluesofc/aanda/b

Reference:

Tada1974Figure3,7-1Equivalent LineLoadStressIntensity FactorEquation.33 C

3.8Information Request8.-SampleCalculation "Provideloadsandvaluesofdafortheresultslabelledunder"FlawStability Criterion" inTables5-3and5-4.Supplydetailsforonecalculation."

RESPONSE

TheappliedstressesforthelimitingLevelCandDtransients areprovidedinReference

[MA93b].TheappliedJandhavaluesforthelimitingpostulated fiawdepthundertheASMEAppendixXflawstability criterion forLevelCloadingconditions aregiveninTable3.8-1.SimilardataforLevelDloadingconditions aregiveninTable3.8-2.Theresultsshownareforthelargesth,awhichcorresponds tothedeepestpostulated initialflawanalyzed.

Iterative calculations wereperformed whichallowthecracktoextendtoitsequilibrium lengthforcaseswheretheinitialJisgreaterthanJ<<.Aspectrumofinitialflaws,upto1/10ofthebasemetalwallthickness, wereassumed.Thesmallestpostulated flawis0.05in.andtheinitialflawsizeswereincremented by0.05in.uptoamaximuminitialflawdepthof0.75in.AsshowninTables3.8-1and3.8-2,forUSElevelsabove20Mbs,theflawgrowthislessth'an0.08in.Therefore theJ-Rcurveplateauisnotreachedandstabletearingoccursuntiltheequilibrium flawdepthisreached.Asampleflawstability calculation fortheLevelCloadingisprovidedinAttachment 1.34

Table3.8-1AppliedLoadsandCrackExtension forVariousUSELevelsAnalyzedUndertheASMEAppendixXFlawStability Criterion forLevelCLoadingConditions andanAxialFlawOrientation'SE Level102030405060708090100FinalAppliedJ'in-ib/in~

182.2181.5180.9180.7180.4179.8179.8.179.8179.8179.8h,aPhysical~in.0.07930.05080.03240.02460.01800.00.00.00.00.0AppliedT0.0960.1070.1140.1170.1200.1270.1270.1270.1270.127Criterion Satisfied yesyesyesyesyesyes,J<Jrcyes,J<Jrcyes,J<J<<yes,J<J<<yes,J<J<<E'esultsshownarefor'thelargesth,awhichoccursforthedeepestpostulated basemetalflaw(a.=0.75in)'hefinalappliedJisiteratively calculated andrepresents theappliedJaAerthecrackreachesitsequilibrium length35 I'

Table3.8-2AppliedLoadsandCrackExtension forVariousUSELevelsAnalyzedUndertheASMEAppendixXFlawStability Criterion forLevelDLoadingConditions andanAxialFlawOrientation'SE Level10FinalAppliedJ'in-ib/in

~haPhysical~in.AppliedTCriterion Satisfied no20299.50.07300.129yes30405060708090.100297.6296.4296.4296.4296.4296.4296.4296.40.02550.00.00.00.00.00.00.00.1580.1740.1740.1740.1740.1740.1740.174yesyes,JCircyes,J'Jrcyes,J<Jrcyes,J<Jrcyes,JNrcyes,J<Jrcyes,J<J<<'esultsshownareforthelargesthawhichoccursforthedeepestpostulated basemetalflaw(a.=0.75in)'hefinalappliedJisiteratively calculated andrepresents theappliedJaAerthecrackreachesitsequilibrium length'6 0

4.0REFERENCES

[ASME92]ASMEDraftCodeCaseN-XXX,"Assessment ofReactorVesselswithLowUpperShelfCharpyEnergyLevels",Revision11,May27,1992.[HI89][MA92]Hiser,A.L.,Terrell,J.B.,"SizeEffectsonJ-RCurvesforA302BPlate",NUI&G/CR-5265, January,1989.ENMPCLetterfromC.D.TerrytoNRC,datedOctober16,1992,"Elastic-Plastic FractureMechanics Assessment ofNineMilePointUnit1BeltlinePlatesforServiceLevelAandBLoadings".

[MA92b]NMPCLetterfromC.D.TerrytoNRC,datedDecember17,1992,"Elastic-Plastic FractureMechanics Assessment ofNineMilePointUnit1BeltlinePlatesforServiceLevelAandBLoadings".

[MA93]Manahari, M.P.Sr.,"Elastic-Plastic FractureMechanics Assessment ofNineMilePointUnit1BeltlinePlatesforServiceLevelAandBLoadings",

FinalreportpreparedforNMPC,MPM-USE-293215,

February, 1993.[MA93b]NMPCLetterfromC.D.TerrytoNRC,datedFebruaiy26,1993"Elastic-Plastic FractureMechanics Assessment ofNineMilePointUnit1BeltlinePlatesforServiceLevelCandDLoadings".

[MEA83]Materials Engineering Associates, Inc.,Lanham,MD(Hiser,A.L.,andFishman,D.B.),"J-RCurveDataBaseAnalysisofIrradiated ReactorPressureVesselSteels",FinalreportpreparedforEPIU,December, 1983.[MEA90]Materials Engineering Associates, Inc.,Lanham,MD,"Influence ofFluenceRateonRadiation-Induced Mechanical PropertyChangesinReactorPressureVesselSteelsFinalReportonExploratoiy Experiments",

preparedforNRC,NUIT/CR-5493,March,1990.[WELD3]"WELD3ComputerCodeVerification",

MPMResearch&Consulting, Calculation No.MPM-NMPC-99205, Rev.0,January21,1993.37 0

Appendix-ExampleLevelCFlawStability Calculation 38