Mechanical performance of integrally bonded copper coatings for the long term disposal of used nuclear fuel

The preferred method for disposal of used nuclear fuel is underground emplacement in a Deep Geological Repository (DGR). Many countries have light water reactor fuels which require large Used Fuel Container or Canister (UFC) designs weighing up to 25 ton for containment.

Mechanical performance of integrally bonded copper coatings for the

a Nuclear Waste Management Organization, 22 St Clair Ave East, Toronto, Canada

b Department of Mechanical and Industrial Engineering, University of Toronto, 5 King’s College Road, Toronto, Canada

•AnovelUsedFuelContainerwithanintegrallybondedcoppercoatingisproposed

•Twodevelopedcoatingprocessessuccessfullyproducedprototypecontainercomponents

•Wecreatedavalidatedfiniteelementmodeltopredictcoatingstructuralperformance

•Mechanicaltestingconfirmscoatingsuitablyforrepositoryuse

Article history:

Received 15 April 2015

Received in revised form 29 June 2015

Accepted 4 August 2015

ThepreferredmethodfordisposalofusednuclearfuelisundergroundemplacementinaDeep Geo-logicalRepository(DGR).ManycountrieshavelightwaterreactorfuelswhichrequirelargeUsedFuel ContainerorCanister(UFC)designsweighingupto25tonforcontainment.Incontrast,Canadaexclusively usesheavywaterreactorfuel,whichissubstantiallysmaller.ThishasledtheNuclearWaste Manage-mentOrganization(NWMO)tocreateanovelUFC,whichusesstandardpressurevesselgradesteelfor structuralcontainmentandathick,integrallybondedcoppercoatingappliedtotheexteriorsurfacefor corrosionprotection.Currently,thecoatingisappliedusingtwodifferentmethods:electrodeposition andgasdynamiccoldspray.Thisnovelcoppercoatingneedstobefullyvalidatedtoensureadequate mechanicalstrengthandchemicalresistanceforuseunderrepositoryconditions.Detailedmechanical andcorrosiontestingprogramswereundertaken.Mechanicaltestsindicatedthatadhesionstrengths exceeded45MPaandtensilepropertieswerecomparabletowroughtcopper.AFiniteElementModel (FEM)ofthecopper–steelcompositewascreatedandvalidatedusingthreepointbendtests.Thismodel accuratelypredictstheresponseofthecomposite,includinglargedeformationanddebondingfailure mechanisms.Nowvalidated,thismodelwillbeusedtoassesstheperformanceofthecoatingonthe full-scaleUFCundersimulatedDGRloadingconditions

©2015TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBY-NC-ND

license(http://creativecommons.org/licenses/by-nc-nd/4.0/)

1 Introduction

Theinternationallypreferredmethodforthelong-termdisposal

ofusednuclearfuelisaDeepGeologicalRepository(DGR).Many

countries,includingSweden,Switzerland,andCanada,beganDGR

researchanddevelopmentasearlyasthe1970s.Currently,there

areseveraladditionalcountriespursuingDGRs,includingFinland,

Japan,Korea,Belgium,France,andtheUnitedKingdom.The

long-termsafetyofaDGRreliesontheuseofmultipleengineeredbarrier

systems(EBS), which provide redundantcontainment, isolation,

∗ Corresponding author Tel.: +1 6472593736.

E-mail address: cboyle@nwmo.ca (C.H Boyle).

andretardationfunctions,asshowninFig.1.TheEBSconsistsof usedorspentfuelbundlespackagedintoalong-livedUsedFuel Con-tainerorCanister(UFC).Thecontainerissurroundedbybentonite clay,which retardstheflow ofwater and suppressesmicrobial growth(WolfaardtandKorber,2012;Stroes-Gascoyneetal.,2010) TheDGRisconstructedatadepthofover400m.Thegeosphereof denserock,whichhasnofreeflowingwater,limitsthemovement

ofradioactiveparticles.NaturalanaloguesofDGRs,suchastheCigar Lakeuraniumdeposit,haveeffectivelyisolatedhigh-gradeuranium oreformillionsofyears(Milleretal.,1994)

SincethebeginningofDGRresearchanddevelopmentinthe 1970s,copperhasbeenafavoredmaterialforcontainercorrosion prevention.Copperwasselectedduetoitsthermodynamicstability fromcorrosionunderDGRconditionsandseveralnaturalanalogues http://dx.doi.org/10.1016/j.nucengdes.2015.08.011

0029-5493/© 2015 The Authors Published by Elsevier B.V This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-nc-nd/4.

Fig 1. Canada’s Deep Geological Repository (DGR) concept.

proving its performance These analogues include

archaeologi-calartifacts(King,1995)(i.e.coins,cannons,etc.),whichcontain

metalliccopper,aswellas,mineraldepositsthatcontainnaturally

occurringmetalliccopper(Chastainetal.,2011).Inaddition,

cor-rosionprocessesthatimpactthelifespanofcopperwithinaDGR

havebeenextensivelystudiedforover30yearsbytheinternational

community.MostrecenteffortsbyCanada’sNuclearWaste

Man-agementOrganization(NWMO)(Kwong,2011),aswellas

indepen-dentcorrosionexperts(ScullyandEdwards,2013)havefocusedon

developingandreviewingcorrosionallowancestoaccountforall

theprocessesthatsignificantlyaffectcoppermaterials.Fromthese

reviews,acoppercorrosionallowanceoflessthan1.3mmhasbeen

deemedappropriateforonemillionyearsstorageinaCanadianDGR

Thiscorrosionallowanceisveryconservative;itisexpectedthat

muchlessthan1.3mmofcopperwillcorrodeoverthattimeperiod

Swedenand Finlandhaveproposeda“dual-vessel”container

designconsisting ofa largecast-iron innervesselfor structural

strength with a 50mm thick copper overpack vessel for

cor-rosionprotection Thisconcept is knownas theKBS-3 (Svensk

KärnbränslehanteringAB,2010).NWMOalsohasareference

dual-vesselcontainerdesign;however,akeydesigndifferenceistheuse

ofahollowinnersteelshellforthecontainmentvesselinsteadofa

honeycombcast-ironinsert.Canada’sheavywaterCANDUreactors

usesmall,naturaluraniumfuelbundles,whichcanbepackagedas

denselyaspossiblewithnegligibleriskofcriticalityinwaterorair

Asaresult,ashelldesignallowsmoreefficientstorageofCANDU

bundlesbytheuseofinternalbaskets.Whilethedual-vesseldesign

istechnicallyfeasible,thereare severalpotentialchallenges for

implementation.Fromafunctionalperspective,anominalradial

gapoflessthan2mmbetweentheinnerandoutervesselhasbeen

identifiedasarequirementtolimitthecreepstrain(Raikoetal.,

2010)and preventrupturefromlow creepductility (Petterson,

2012).Thisrequiresmanufacturinga9to14tsteelorcast-iron

ves-selanda7.5tcoppervesselalmost5minlengthwithtightradial

fit-uptolerances,followedbyprecisionassembly.Theassembled

UFCisthenhandledfromthecoppervesselandneedstosupport

theentire∼25tloadedcontainerweight.Consequently,the

thick-nessofthecoppershellmustbemuchgreaterthanwhatisrequired

forcorrosionprotection

Adaptivephasedmanagement(APM)is theNWMO’stechnical

methodandmanagementsystemforimplementingCanada’sDGR

(NWMO,2005).APMemphasizesadaptabilityandincorporationof

evolvingknowledgeandtechnology.Thisphilosophyhasdrivenan

initiativetodevelopanalternativeUFCforCanada’suniqueCANDU

fuelandgeosphere,whichovercomessomeofthepotentialissues

inherenttothedual-vesseldesign.ThisUFC,knownastheMark

IIandshowninFig.2,isunderdevelopmentwithseveralnovel

designconcepts:

Fig 2. NWMO’s Mark II Used Fuel Container for CANDU bundles (cut-away shown for clarity) Approximate dimensions 562 mm (∼22  ) diameter, 2514 mm (∼99  ) length.

• Copper coating: is integrallysupported bythe steelstructural substrate.The thicknessis driven bythecorrosion allowance requirementandcanbetailoredtosite-specificrequirements

• Hemi-sphericalheads:betterdistributionoftheexternalpressure loadresultinginbiaxialcompressivestresses.Flatheaddesigns canproducethetensilestressesduetobending;tensionis unde-sirableinthecontainerasitisakeycomponentincrackgrowth mechanisms,suchas,StressCorrosionCracking(SCC)andfatigue

• Pressurevesselmaterials:Theproposeddesignusescommon,weld understoodnuclearpressurevesselgradematerialsandsizes.For example,theshellismanufacturedfromstandardsizedextruded steelpipeorsmallforgings,approvedforusebyASMESection3 forstoragecontainments.Abenefitofusingthesematerialsis easeofavailability(comparedtolargesizedcustomcastingor forgings)

• Manageablesize:Manyinternationalnuclearwastemanagement organizationshaveverylargesteelorcastironUFCs,weighing

upto25tonceloaded.Handlingandundergroundemplacement

ofsuchheavycontainersrequireslarge,customequipment.The sizeofthesecontainersisdrivenbythelightwaterreactorfuel, whichcanexceed4minlength.Incontrast,Canada’sCANDUfuel

isonlyahalfmetreinlength.ThisallowedtheNWMOtooptimize theUFCdimensionsforthissmallerfueltype.TheresultingMark

IIUFCweighslessthan3tandcouldpotentiallybehandledusing radiationshieldedconventionalsizedforklifttrucks

Themostnovelaspectofthiscontaineristhecoppercoating Thisconceptallowsfordirectdepositionofthecoppercorrosion barrier layeronto thesteelor cast-iron structuralcomponents, formingarobustmetallurgicalormechanicalbondresultingina single,unifiedUFCcompositestructure.Assemblytolerancesand creepruptureissuesareresolved.Additionally,thethicknessofthe coppercanbetailoredtothesitespecificgeosphereand environ-ment.Currently,NWMOisproposinga3mmcopperlayerbased

onthepreviouslystatedcorrosionrequirement;however,various thicknessesupto10mmare beinginvestigated withinongoing workprograms

Theobjectiveofthisworkistodeveloparobust,coppercoating whichcanbeappliedtodisposalcontainersforthesafedisposalof usednuclearfuelinaDGR.Toaccomplishthis,twonovelcopper coatingprocesseswerestudied:gasdynamiccoldsprayandpulsed electrodeposition During process development, coating quality wasmeasuredagainsttwomajorfunctionalrequirements: corro-sionandmechanicalperformance.Chemically,thecoppercoating musthaveequivalentorexceedthecorrosionperformanceofthe referencewroughtcopperthatiscurrentlyproposed.Thisworkis ongoingandpreliminaryresultshavebeenpublishedelsewhere (Jakupi,2015;Keechetal.,2014).Mechanically,thecoppercoating musthavesufficientstrength,ductility,andadhesiontowithstand allloadingsunder DGRconditions Anexperimentallyvalidated mechanicalintegritymodelofthecoppercoatingwasdevelopedto predictitsbehaviourunderbeyonddesignbasisloading,including

2 Methods

2.1 Coatingprocessdevelopment

Thecoatingprocessdevelopmenthadthreemainobjectives:

1.Theperformance of thecoppercoating,both themechanical

structural response and chemical corrosion resistance, must

meetorexceedthatofthewroughtcopperasdeterminedby

experimentaltesting

2.Thecoatingprocess mustensurethat repeatable, fullydense

coatingsareproduced

3.Thecoatingprocessisfeasibleforlargescalecontainer

manu-facturing

TheNWMOhasinvestigatedseveraldifferentcoatingmethods

including:weldoverlay,gasdynamiccoldspray,and

electrode-position.Coldsprayandelectrodepositionprocessesfacilitatethe

production of highpurity coatings(i.e., noalloyadditions); on

thisbasis,thesemethodswereselectedforfurtherdevelopment

describedherein.Thetest programincorporatedtheapplication

of coatingsto steelsubstrates used in MarkII UFCfabrication,

includingplate,pipe,andhemi-sphereproductforms,rangingin

thicknessfrom12to46mm.Bothcoatingmethodsaredescribed

indetailelsewhere(Papyrinetal.,2006;Austetal.,2008)

Gasdynamiccoldsprayor“coldspray”involvesthe

accelera-tionofpowderswithinaninertcarriergastohighvelocities,at

whichtheyimpactasubstrateandformastrongmechanicalbond

(Irissouetal.,2008).Themethodissimilartothermalspraycoatings

butthetemperatureofthepowderdoesnotexceedthemelting

temperature-asolidstateprocess.Asaresult,thedepositedlayer

hasidenticalchemicalpropertiesastheinitialpowderfeed.While

veryhighdepositionratescanbeobtainedbya singlegun(i.e

upto 1kg/min), it hasbeenprimarilyused asa repair process

withinindustry.WithintheNWMOprogram,bothlowpressure

coldspray(LPCS)andhighpressurecoldspray(HPCS)havebeen

investigated,forcompleteUFCcoverage(i.e.factorysupplied

com-ponents),partialUFCcoverage(i.e.coatingweldclosurezoneafter

finalassembly),andcoatingrepair.TheuseofcoldsprayforUFC

manufactureisalsobeinginvestigatedbytheKoreannuclearwaste

managementprogram(Choietal.,2010)

Electrodepositioninvolvesimmersionoftwoelectrodesintoa

specializedchemicalbathsolution.Acurrentisappliedtothe

elec-trodes,oxidizingtheanodematerialproducingdissolvedcations

inthesolution,whicharethenreductivelyplatedatthecathode

Forthisapplication,ahighpurity,oxygenfreecopperanodewas

usedasthecoppersourceandpressurevesselgradesteelwasthe

substratecathode.Pyrophosphatewasusedastheprimarybath

solutiontominimizecarbonandoxygencontentwithinthecopper

coatingandpulsedpotentiometrywasusedtoapplysufficient

cur-rent.Inthisapplication,electrodepositionwouldbeusedtosupply

pre-coatedUFCcomponents

2.2 Coatingmechanicalperformance

Theprimaryfunctionofthecoppercoatingisacorrosionbarrier

Nonetheless,toremainaneffectivebarrieritmustbefullydense

2008a)togeneratestress–straincurves,asshowninFig.3.Five specimens ofeach materialwerepreparedbywireelectric dis-chargemachining.Theyieldstrength,ultimatetensilestrength,and overallstrainwerecalculatedfromthecurves.Inadditionaltothe coatings,theASTMA516Gr.70(ASTM,2010)carbonsteelsubstrate wasalsotestedintheas-receivedandannealedconditions Adhesionstrengthtestswereperformedtomeasurethebond strengthofthebimetalliccopper–steelinterface.ASTMC633-01 (ASTM,2008b)isthestandardtestingmethodforadhesionstrength

ofthermal/coldspraycoatings.Thetestingmethodologyinvolves applyingthecoatingtoa1diameterplugmanufacturedfromthe substratematerial,whichisbondedtoaseparateblankplugusing

astrongadhesivebondingagent(suchasepoxy).Forthiswork,the selectedbondingagentrequiredacuringheattreatmentof∼150◦C for1h

AlimitationofASTMC633-01testingstandardistheuseofa bondingagent.Commercialhighstrengthadhesivesprovidebond strengthupto60–70MPabeforefailingintheepoxy;asaresult,the testmayonlyidentifythatthecoatingadhesionexceedsthis mini-mumepoxystrength.Todeterminetheactualadhesionstrength,a modifiedversionoftheASTME08–04tensiletestisused.A bimetal-liccopper–steelmicrospecimen,similartothoseintheASTME8 tensilestandard, weremanufacturedandtested usingacustom fixture,asshowninFig.3.Fifteenspecimensforeachcoatingwere prepared.Thismethodallowsanaccuratemeasurementof adhe-sionstrength,asthegeometryensuresfailureinthebulkcopperor

atthebimetallicinterface

Three point bend tests were performed to assess ductility, resistancetocracking,and debondingofthecopper–steel com-posite.Thetestingand specimengeometryfollowedtheguided U-bendtestinaccordancewithASTME290-09(ASTM,2009).The thicknessesof thecopperand steelsubstrate were∼3mmand

∼6.5mm,respectively,foratotalspecimenthicknessof∼9.5mm (3/8 asperthestandard).Fivespecimensweretested foreach coatingtype.Thetestingapparatusmeasuredtheforce–deflection responsethroughoutthebend.Thespecimenswerefilmedwitha high-resolutioncameraduringtestingtodeterminetheonsetof surfacedefectsanddebonding

Itisimportanttonotethatthethreepointbendtestrepresents

anextremeloadingscenario,farexceedingthecontainer deflec-tionsand strainsresultingfrom theDGRloads Thecontainer’s steelsubstrateandcoppercoatingaredesignedtoremaininthe elasticrange duringnormal expectedloadings(i.e.groundwater hydrostaticheadandbentoniteswelling).Eveninextreme load-ingscenarios,includingthehydrostaticpressurefroma3000m thickglacierpositionedovertherepositorywouldinducestrains less than1%in thecoppercoating Thepurposeofthis beyond designbasistestistovalidatetheperformanceofthecoppercoating mechanicalintegritymodelandtoensureitcanaccuratelypredict thecoating’sbehaviourincludingpotentialfailuremechanisms

2.3 Coatingmechanicalintegritymodel

A coating mechanical integrity model, which canaccurately predict the behaviour of the copper–steel composite at the bimetallicinterfaceand inthebulkmaterials,ispresented.The

Fig 3.Copper coating material property testing: (A) ASTM E08-04 tensile testing, (B) ASTM C633-01 adhesion test plug specimen, (C) modified ASTM E08-04 adhesion “Dog Bone” specimen, (D) custom test fixture for modified E08 adhesion testing.

following methodology was used to develop and validate the

model:

1.Materialcharacterization:Theindividualtensilepropertiesofthe

coppercoating(s)andsteelsubstrate;aswellas,the

correspond-ingadhesionpropertieswereexperimentallydetermined

2.Developmentofthecoatingmechanicalintegritymodel:AFinite

ElementModel(FEM)ofthebimetalliccopper–steelcomposite

wasdeveloped.Thebimetallicinterfacebondisimplemented

using the numerical Cohesive Zone Model (CZM) for

con-tact/interfaceelements.Ifthefailurecriteriaaremet,theCZM

willinitiatethefractureanddebondingofthecoating.The

exper-imentallydeterminedtensileandadhesionpropertiesactasthe

inputstothemodel

3.Experimentalvalidation viathreepointbendtesting: Usingthe

developedmodel,simulationsofthethree-pointbendtestsfor

thevariouscopper–steelspecimenswerecompleted.The

com-putationalresults,includingtheforce–deflectionresponseand

onsetofdebonding,werecomparedtotheexperimentalbend

tests

CZMisanumericalfracturemechanicstechnique,whichwas

originallydevelopedtopredictcrackgrowthinconcretebuthas

since been applied to other materials and failure mechanisms

(Hillerborgetal.,1976).ThebilinearCZMformulation,asproposed

byAlfanoandCrisfield(2001),wasimplementedtomodel

debond-ingbetweenthecopper–steelinterfaceandcrackpropagationin

thebulkcoppercoating.Theexpectedcontainerloadsactnormal

tothecoatingsurfaceanddonotcreatesubstantialshearingloads

atthebimetallicinterface,thereforetangentialslipwillnot

signifi-cantlycontributetocoatingdebonding.Atthistime,thetangential

slipfailurecriteriaareassumedtobeidenticaltonormal

separa-tion(i.e.failureisModeIdominated).ThebilinearCZMconstitutive

modelemploysalinearsofteningrelationshipbetweenthe

nor-malcohesivecontactstressandtheinterfaceseparationdistance

(contactgap)tosimulatethedebondingprocess

Thefiniteelementmodelingofthethree-pointbendtest

speci-mens,asshowninshowninFig.4,wascompletedinANSYSV14.5

software(ANSYS).Non-linear,largedeformationformulationwas

used.Allmaterialpropertiesweretakenfromtheexperiments

dis-cussedabove.Isotropicstrainhardeningwithmaximumdistortion

energytheoryflowrulewasusedtomodeltheplasticdeformation

behaviourofboththecoppercoatingsandsteelsubstrate.TwoCZM

zoneswereimplemented:betweenthebimetallicinterfaceandin

thebulkcoppercoating.Failureinthebulkcoatingtheoretically

occursatthecentreofthespecimenduetothehightensileloads;

Fig 4.Three-point bend specimen geometry and cohesive zone model (CZM) inter-face locations.

therefore,theCZMmodeltriggersfailureiftheexperimental ulti-matetensilestrengthisreached(99%oftheexperimentaltensile valueisusedtoavoidnumericalinstability).ThesecondCZMmodel

atthebimetallicinterfacetriggersdebondingiftheexperimental adhesionstrengthisexceeded.TheguidedU-bendsupportsand puncharemanufacturedfromhighstrength steeland assumed

toberigidinordertoreducecomputationaleffort.Thepunchis loadedincrementallytoatotaldeflectionof30mm,identicaltothe experiment,tomakeanapproximate90◦bendinthespecimen

3 Results and discussion

3.1 Coatingprocessdevelopment Gasdynamiccoldsprayprocessdevelopmentcommencedwith characterizing high purity, low-oxygen copper powders Fig 5 demonstratestypicalpowdershape/sizeusedinhighpressurecold spray(HPCS)coatings.Thenextphaseofdevelopmentoptimized various cold spray operating parameters, such asgas pressure, pre-heating,andfeedrate.Thesewereallexperimentallytested formechanical performanceand thetop performersselected.It wasdeterminedthatcosteffectivecoatingscouldbeproducedvia twostages:initiallya50–100␮mbondcoator“strikelayer”was depositedusingheliumasacarriergas,followedbyabulkcoating depositedusingnitrogen

Fig 5. Scanning election microscope analysis of cold spray copper characteristics (A) low-oxygen copper powder, (B) cross-section of fully dense test coating, (C) cross-section depicting “Jetting” Bond.

Fig 6.Cold spray coating on 20  diameter pipe segment (A) cold spray equipment and process, (B) machined coating, (C) section showing >3 mm fully dense copper coating.

Theconstanthighvelocityimpactofparticlesresultsina

homo-geneouscoating,asshowninFig.5,withnonoticeableindividual

particlegeometryremaining Theintimate mixingbetweenthe

coppercoatingandsteelsubstrate,aprocessknownas“jetting”,

isalsovisible.Jettingis avisualindicationthat thecoating has

goodadherenceontothesubstrate.Aftercoatingparameter

opti-mization,thefinaltaskwastoensurefeasibilityofcoatingactual

containergeometry.Thetechnologywasusedtosuccessfullycoat

thepressurevesselshellmaterialtoathicknessexceeding3mm,

asshowninFig.6

The deposited coating material strain hardens due to the

high impact velocity and bonding process This highly cold

worked structure exhibits decreased ductility and increased

yield strength However, material properties consistent with

polycrystalline wrought copper can be achieved by annealing

the as-deposited coating (Eason et al., 2012) Several differ-entannealingtemperaturesare beingevaluatedwithinongoing research

Electrodepositionprocessdevelopmentfocusedonoptimizing thebath solutionchemistry and thepulsedcurrent application

toensureauniform, finegrained,highpuritydeposited copper layer.Thedeveloped electrodepositedcoppersamplesexhibited hightensilestrength,ductility,andadhesion.Incontrasttocold spray,nopostdepositionannealingisrequired.Forthisinitialwork depositionratesperunitareaweregenerallyslow,withthe3mm coatingtakingapproximately72htoproduce.However,thenature

oftheprocessallowssimilardepositionratesregardlessofthe coat-ingarea(i.e.smallplatesor thecontainercanbecoatedin the same time).The processis also easilyscalable, makingparallel production ofmultiplecontainers possible.Thetechnologywas

Fig 7. Electrodeposited coating on 22  diameter mock-up Mark II container section (A) electrodeposition solution tanks, (B) steel mock-up prior to immersion, (C) as-deposited

Table 1

Tensile strength and ductility of used fuel container materials.

Copper electrodeposition—as-deposited 226.1 ± 4.7 312.1 ± 6.2 43.1 ± 5.6 Wrought SKB OFP-copper ( Sandström et al., 2009 ) ∼70 ∼194 ∼38

* NOTE: 2 specimens failed outside the gage length and were not considered.

usedtosuccessfullycoatamock-upMarkIIcontainersectionto

athicknessexceeding3mm,asshowninFig.7

3.2 Coatingmechanicalperformance

Tensile properties of the various copper coatings and steel

substrates, as summarized in Table 1, were comparable to or

exceededthereferenceSKBwroughtcopperwiththeexception

oftheas-deposited coldspray asexpected.Theductility ofthe

coldspraycoatingsvarieddependingonthedegreeofannealing

The as-deposited samples consistently had maximum strains

ofless than0.3% resultinginimmediate brittlefracture.Asthe

annealingtemperaturesincreased,ductilityincreasedwhileyield

strengths decreased Large variability in maximum strain at

fracturewasnotedfortheannealedspecimenswiththestandard

deviationrangingfrom13to22%ofthemean.Thevariabilityin

the600◦Cannealedspecimenswascompoundedsincetwotests

wereexcludedduetofailureoutsidethegagelength.Despitethe

variability,thesepreliminaryresultsindicatethatpost-deposition

annealing can achieve strengths and ductility suitable for the

container

Representativestress–straincurvesofthevariouscoatingsare

showninFig.8.Inordertodemonstratethevariabilityofthetensile

Table 2

Adhesion strength of copper coatings.

E8-04) [MPa]

near steel interface

necking prior to fracture

data,twocurvesdepictingthelowestandhighestachievedstrains

atfracturearepresentedforeachprocess(withtheexceptionof thelow performance,as-deposited coldspray) TheA516Gr.70 steelsubstratewasalsotestedintheas-received(normalized)and annealedconditions.Themeasuredyieldstrength exceededthe minimum260MPaspecifiedbytheproductformstandardforall conditions.Theannealedspecimenshadslightlylowerstrengthbut increasedductility

TheadhesionstrengthtestingresultsaresummarizedinTable2 andtypicalspecimenfailuresareshowninFig.9.Thecoldspray specimens failed in thebulk coating and exhibited noyielding

0

50

100

150

200

250

300

350

Enginee ring Strain [%]

Wrought Copp er: Representave SKB Electrodeposion: As-deposited (Highest Strain) Electrodeposion: As-deposited ( Lowest Strain) Coldspray: Ann ealed @ 60 0°C ( Highest Strain) Cold spray: Ann ealed @ 60 0°C ( Lowest Strain) Coldspray: Ann ealed @ 35 0°C ( Highest Strain) Cold spray: Ann ealed @ 35 0°C ( Lowest Strain) Cold spray: As-deposited ( Typical)

Fig 8.Engineering stress–strain curves of various copper coatings versus wrought copper The results of the coating specimens with the lowest and highest strain at fracture

Fig 9.Typical copper coating adhesion test results: (A) cold spray as-deposited, (B) cold spray annealed (1 h@350 ◦ C), (C) cold spray annealed (1 h@600 ◦ C), (D) electrodepo-sition as-deposited.

beforefailure.The electrodepositedalsofailedin thebulk

cop-per; however, significant necking occurred This demonstrates

that the adhesion strength of the steel-copper interface likely

exceedstheultimatetensilestrengthofthebulkelectrodeposited

copper

Three-pointbendtestingresultsareshowninFigs.10–12.All

as-depositedcoldsprayspecimensexhibitedsurfacecracksinless

thanamillimetreofloading.Thecrackspropagatedquicklythrough

thebulkcoatingtothesubstrate,followedbyfulldebondingfailure

atthecopper–steelbimetallicinterfaceasobservedinFig.10.As

loadingcontinued,thecoppercoatingprogressivelypeeledaway

fromthesubstrateleavingnoresidualcopperattheinterface.The

annealedcoldspraycoatingsperformedmuchbetter,reaching50◦

to80◦ bendbeforecrack formation, correlatingto ∼15–28mm

of deflection.This highvariability ofinitialfailure deflection is

discussed inSection 3.3 Oncea crack developed it propagated

rapidlyanddebonding ensued,similartotheas-deposited cold

spray.Examinationofthebimetallicinterfacepost-failurerevealed

athinresidualcopperlayerthatremainedadheredtothesteel

sub-strate.Thisishypothesizedtobetheinitialheliumstrike layer

Fig.11showstheonsetofcrackformationforthe600◦Csamples;as

wellas,fulldebondingatmaximumbend.The350◦Csamples

per-formedsimilarly.Theelectrodepositedcoatingdemonstratedthe

bestperformancereachingthefull90◦+bendwithoutanycracking

ordebondingonallfivespecimens;atypicalresultsisshownin Fig.12

3.3 Coatingmechanicalintegritymodel The coating mechanical integrity model was compared to the experimental three-point bend tests for four coatings: as-depositedcoldspray,annealedcoldspray(1h@350◦C),annealed coldspray(1h@600◦C),andas-depositedelectrodepositionusing the force–deflection curves and onset of cracking Compari-son of the model and experimentalforce–deflection curves, as showninFigs.10–12,revealgoodagreement.Todemonstratethe influence of the tensile property inputs, two simulations were completed for each coating type, corresponding to the lowest andhigheststrainsatfractureforthetensiledataexperimentally obtained

Fig.13showsthegoodcorrelationbetweentheforcedeflection curvefortheas-depositedcoldspraysamplesandtheFEM.The peakforceatinitialcrackformationrangedfrom4857to5128N occurringat0.62–0.83mmofdeflection.Forthemodel,peakforce rangedfrom4623to4679Nat0.60mmofdeflectionwere pre-dictedcorrespondingthelowestandhighesttensileperformance,

Fig 10. Three-point bend results for as-deposited cold spray: (A and C) crack-initiation, (B and D) debonding at full-bend (A and B) Experimental results versus (C and D)

Fig 11.Three-point bend results for 600 ◦ C annealed cold spray: (A and C) crack-initiation, (B and D) debonding at full-bend (A and B) Experimental results versus (C and D) model results.

Fig 12. Three-point bend results for as-deposited electrodeposition: no debonding at full bend (A) experimental result versus (B) model results.

Fig 13.Three-point bend force–deflection response: as deposited cold spray.

respectively.Afterdebondingbegan,themodel’spredictedforce

responsefallswithintherangeofexperimentalresults,asshown

inFig.13.Atfullbend,theexperimentalandmodelpeakforces

rangedfrom7408to7741N

Aspreviouslymentioned,theannealedcoldsprayspecimens showedthelargestvariabilitytermsinperformance,asreflected

inFigs.14and15.Forthe350◦Cannealedspecimens,allyielded

at approximately ∼4500N and 1mm of deflection, then had

a similar response up to 15mm of deflection Specimens then failedat16–24mmofdeflection,withfinalpeakforcesbetween

8172and8745N.Themodelpredictedcrack initiationat∼13.5 and20.4mmfor lowestand higheststraintensile data, respec-tively

Forthe600◦Cannealedspecimens,allyieldedatapproximately

∼4100Nand1mmofdeflection,thenhadasimilarresponseupto

15mmofdeflection.Atotalofthreespecimensfailedbetween15 and18mmandthefinaltwofailedat20mmand27mm,ascan

beobservedinFig.15.Forthefourspecimensthatfailedbetween

15and20mm,thefinalpeakforcewas7149–7268N.Themodel predictedcrackinitiationat∼16.5mmfortheloweststrain ten-siledata,followedbyrapidpropagationandcoatingdebonding Thefinalpeakforcewas7156N,whichiswithintherangeofthe experimentalresults.Forthehigheststraintensiledata,themodel predictednocoatingfailure;however,theresultingpeakstrainis within5%oftheultimatestrainandisclosetofailure.Itis hypoth-esizedthatinhomogeneitiesinherenttothecoldsprayprocessact

1000

2000

3000

Fig 14.Three-point bend force–deflection response: annealed (1 h@350 ◦ C) cold

spray.

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

Fig 15.Three-point bend force–deflection response: annealed (1 h@600◦C) cold

spray.

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

Fig 16.Three-point bend force–deflection response: as deposited

electrodeposi-tion.

asstressrisersandareenoughtoinitiatelocalizedfailurezonesnot

capturedinthemodel

For the electrodeposited specimens, yielding occurred at

∼6000Nandsubsequentloadingproducednofailuresofthe

coat-ingforallspecimens,asshowninFig.16.Themodelalsopredicted

nofailureusingboth thelowest andhighest straintensiledata

andfollowedtheexperimentalforce–deflectionresponsewithin4%

upto27mmofdeflection.Althoughnocrackingoccurs,thepeak

curveswereconsistentwiththeexperimentalresults,asshown

inFigs.13–16.Theannealedcoldsprayresultsshowedhigh vari-abilityinthetensiletesting;asaresult,thethree-pointbendtests producedawiderangeoffailures.Coldsprayprocessoptimization

isstillongoingandfabricationvariabilitycontinuestobereduced

Aspreviouslymentioned,itisimportanttonotethatpreliminary UFCdesignanalysishasshownthecopperstrainswouldbemuch lessthan1%evenundertheglacialloadingscenario.Eventheworst performingannealedcoldspraycoatingtestspecimensexceeded 20%strainandwouldbeatnoriskoffailure.Thethreepointbend loadingsrepresentabeyonddesignbasisscenariowithinduced strainsapproaching28%atfullbend

4 Conclusion

The experimentaldevelopmentand mechanical modeling of

arobustcoppercoatingforuseasa UsedFuelContainer corro-sionbarrierhasbeenpresented.Coldsprayandelectrodeposition coatings withcomparable mechanical performance to wrought copperhavebeenfabricatedonthefull-scalecontainermaterials Themechanicalperformanceoftheannealedcoldsprayandas depositedelectrodepositioncoatingswerecomparableorexceeded thatofthereferencewroughtcopperandaresuitableforthe con-tainerdesign.Variabilityintheperformanceofcoldspraycoatings wasnoted.Thisworkrepresentsonlytheinitial“proofofconcept” results;aspartoffuturework,additionalprocessrefinementsand researchintoalternativeannealingscheduleswillbecompletedto reducethisvariabilityandimproveoverallperformance

A mechanical integrity model for the copper–steel compos-ite was developed and experimentally validated It accurately predicted the various copper coating responses, including the bimetallic interfacefailure Themodel’s averageforceresponse deviated less than 4% from the experiments, with localized maximumsofapproximately10–15%.Forfuturework,thismodel willbeusedtoevaluatetheperformanceofthecoatingsonthe MarkIIUFCunderrepositoryloadingconditions;aswellas,beyond designbasisanalysestodemonstratetheconservativenessofthe design

In conclusion, this work demonstrates that copper coatings can be reliably fabricated on container materials and geome-tries These coatings have beenextensively tested and confirm amplemechanicalperformanceforcontainerdesign.Wecan accu-ratelymodeltheirresponseunderexpectedrepositoryconditions andbeyond Productionoffull-scalecontainers,additional opti-mizationofcoatingparameters,andapplicationofthemodelto containergeometriesarecurrentlyunderway

Acknowledgments

Coatingprocessdevelopment,optimization,andexperimental workhasbeenconductedincollaborationwithIntegran Technolo-giesInc.andtheNationalResearchCouncilCanadafacilities

References

Alfano, G., Crisfield, M.A., 2001 Finite element interface models for the debonding analysis of laminated composites: mechanical and computational issues Int J.

ANSYS ® Mechanical, Release 14.5 Help System, Workbench ANSYS Inc.,

Canons-burg, PA http://www.ansys.com/Industries/Academic/Tools/Citations

ASTM, 2008a Standard test methods for tension testing of metallic materials In:

ASTM E8/E8M-08 ASTM International, West Conshohocken, PA.

ASTM, 2008b Standard test method for adhesion or cohesion strength of thermal

spray coatings In: ASTM C633-01 ASTM International, West Conshohocken, PA.

ASTM, 2009 Standard test methods for bend testing of material for ductility In:

ASTM E290-09 ASTM International, West Conshohocken, PA.

ASTM, 2010 Standard specification for pressure vessel plates, carbon steel, for

moderate- and lower-temperature service In: ASTM A516/516M-10 ASTM

International, West Conshohocken, PA.

Aust, K., Brooks, I., Gonzalez, F., Lin, P., Palumbo, G., Tomantschger, K.,

Nagara-jan, N., 2008 Method for preparing polycrystalline structures having improved

mechanical and physical properties Canadian Patent No 2674403.

Chastain, M.L., Deymier-Black, A.C., Kelly, J.E., Brown, J.A., Dunand, D.C., 2011

Met-allurgical analysis of copper artifacts from Cahokia J Archit Sci 38, 1727–1736.

Choi, H.J., Lee, M., Lee, J.Y., 2010 Application of a cold spray technique to the

fab-rication of a copper canister for the geological disposal of CANDU spent fuels.

Nucl Eng Des 240 (10), 2714–2720.

Eason, P.D., Kennett, S.C., Eden, T.J., Krull, I., Kowalski, B., Jones, J.L., 2012 In situ

observation of microstrain relief in cold-sprayed bulk copper during thermal

annealing Scr Mater 67, 791–794.

Hillerborg, A., Modeer, M., Petersson, P-E., 1976 Analysis of crack formation and

crack growth in concrete by means of fracture mechanics and finite elements.

Cem Concr Res 6, 773–782.

Irissou, E., Legoux, J.-G., Ryabinin, A., Jodoin, B., Moreau, C., 2008 Review on cold

spray process and technology: Part I Intellectual property J Therm Spray

Tech-nol 17, 495–516.

Jakupi, P., Keech, P.G., Barker, I., Ramamurthy, S., Jacklin, R.L., Shoesmith, D.W.,

Moser, D.E., 2015 Characterization of commercially cold-sprayed and

electrode-posited copper grains on mild steel J Nucl Mater 466, 1–11.

Keech, P.G., Vo, P., Ramamurthy, S., Chen, J., Jacklin, R., Shoesmith, D.W., 2014 Design

and development of copper coatings for long term storage of used nuclear fuel.

Corros Eng Sci Technol 49, 425 (Special Issue on Corrosion in Nuclear Waste

Systems).

King, F., 1995 A natural analogue for the long-term corrosion of copper nuclear waste containers—reanalysis of a study of a bronze cannon Appl Geochem 10, 477–487.

Kwong, G.M., 2011 Status of corrosion studies for copper used fuel containers under low salinity conditions In: NWMO Technical Report, NWMO-TR-2011-14 Nuclear Waste Management Organization, Toronto, Canada.

Miller, W.M., Chapman, N., McKinley, I., Alexander, R., Smellie, J.A.T., 1994 Natu-ral Analogue Studies in the Geological Disposal of Radioactive Wastes Elsevier Science, Amsterdam, The Netherlands.

NWMO, 2005 Choosing a Way Forward: The Future Management of Canada’s Used Nuclear Fuel Nuclear Waste Management, Toronto, Canada.

Papyrin, A., Kosarev, V., Klinkov, S., Alkhimov, A., Fomin, V.M., 2006 Cold Spray Technology Elsevier, Amsterdam, The Netherlands.

Petterson, K., 2012 A review of the creep ductility of copper for nuclear waste can-ister application In: SSM Technical Note 2012:13 Strålsäkerhetsmyndigheten, Stockholm, Sweden.

Raiko, H., Sandström, R., Rydén, H., Johansson, M., 2010 Design analysis report for the canister In: SKB Technical Report TR-10-28 Svensk Kärnbränslehantering

AB, Stockholm, Sweden.

Sandström, R., Hallgren, J., Burman, G., 2009 Stress Strain flow curves for Cu-OFP In: SKB Rapport R-09-14 Svensk Kärnbränslehantering AB, Stockholm, Sweden.

Scully, J.R., Edwards, M., 2013 Review of NWMO copper corrosion allowance In: NWMO Technical Report, NWMO-TR-2013-04 Nuclear Waste Management Organization, Toronto, Canada.

Stroes-Gascoyne, S., Hamon, C.J., Maak, P., Russell, S., 2010 The effects of the physical properties of highly compacted smectitic clay (bentonite) on the culturability of indigenous microorganisms Appl Clay Sci 47, 155–162.

Svensk Kärnbränslehantering AB, 2010 Design and production of the KBS-3 repos-itory In: SKB Technical Report TR-10-12 Svensk Kärnbränslehantering AB, Stockholm, Sweden (updated 2013-10).

Wolfaardt, G.M., Korber, D.R., 2012 Near-field microbiological considerations rele-vant to a deep geological repository for used nuclear fuel-state of science review In: NWMO TR-2012-02 Nuclear Waste Management Organization, Toronto, Canada.