Engineering design and analysis of an ITER like first mirror test assembly on JET

Engineering design and analysis of an ITER like first mirror test assembly on JET F E a Z V a b c d e h • • • • a A R A A K I J A R D 1 c s s t f m c e t h 0 ARTICLE IN PRESSG Model USION 8982; No of[.]

Fusion Engineering and Design xxx (2017) xxx–xxx

ContentslistsavailableatScienceDirect

j ou rn a l h o m epa g e : w w w e l s e v i e r c o m / l o c a t e / f u s e n g d e s

Engineering design and analysis of an ITER-like first mirror test

assembly on JET

Z Vizvarya,∗, B Bourdelb, A Garcia-Carrascoe, N Lama, F Leipoldc, R.A Pittsd, R Reichled,

a CCFE, Culham Science Centre, Abingdon, Oxon OX14 3DB, UK

b Ecole Polytechnique, Route de Saclay, 91120 Palaiseau, France

c Technical University of Denmark, Department of Physics, DK-2800 Kgs Lyngby, Denmark

d ITER Organization, Route de Vinon-sur-Verdon-CS 90 046, 13067 St Paul Lez Durance Cedex, France

e Fusion Plasma Physics, Royal Institute of Technology (KTH), 100 44 Stockholm, Sweden

h i g h l i g h t s

•NewITERFirstMirrortestassemblyhasbeendesignedandinstalledintoJET

•Theassemblyhasbeenanalysedtocopewiththermalanddisruptionloads

•Themulti-coneapertureshavebeenproducedbyadditivemanufacturing

•MaterialqualificationprogramforInconel718producedbyselectivelayermelting

a r t i c l e i n f o

Article history:

Received 3 October 2016

Accepted 12 December 2016

Available online xxx

Keywords:

ITER-like first mirror

JET

Additive manufacturing

Remote handling

Disruption loads

a b s t r a c t

TheITERfirstmirrorsarethecomponentsofopticaldiagnosticsystemsclosesttotheplasma.Deposition maybuilduponthesurfacesofthemirroraffectingtheirabilitytofulfiltheirfunction.However,physics modellingofthislayergrowthisfraughtwithuncertainty.AnewexperimentisunderwayonJET,under contracttoITER,withprimaryobjectivetotestif,underrealisticplasmaandwallmaterialconditions andwithITER-likefirstmirroraperturegeometry,depositsdogrowonfirstmirrors.Thispaperdescribes theengineeringdesignandanalysisofthismirrortestassembly

Theassemblywasinstalledinthe2014–15shutdownandwillberemovedinthe2016–17shutdown

©2016PublishedbyElsevierB.V

Opticaldiagnosticsystemsrelyonfirstmirrorswhicharethe

componentsthatguide/directlighttothedetectorofthediagnostic

system.Assuchtheyareplasma-facingcomponents(PFCs)andare

subjecttodepositionand/orerosion.Theresultingmodificationsto

themirrorfrontsurfacescanhaveaprofoundimpactonthe

per-formanceoftheassociateddiagnostic.InadevicelikeITER,where

maintenance andcleaning of theseelementsisextremely

diffi-cult,itiscrucialtotryandpredicttheleveloferosion/deposition

expectedinadvanceofoperation.Unfortunately,physics

simula-tionsoftheseprocessesarefraughtwithuncertaintiesandsmall

∗ Corresponding author.

E-mail address: zsolt.vizvary@ukaea.uk (Z Vizvary).

adjustmentsininputparameterscanleadtopredictionsranging overordersofmagnitude.Inthiscase,theonlyoptionis“designby experiment”

First Mirror Testing (FMT) has been performed at JET for manyyears(seee.g.[1–3]),bothwithcarbonwalls(2004–2009) andintheITER-LikeWall(ILW)beryllium-tungstenenvironment (2011–present).Inthelattercase,mirrorsmountedontheoutboard mainchamberwallwereobserved,encouragingly,tobeveryclean afterexposuretoafullILWplasmacampaign[3].However,these mirrorsampleswherenotexposedunderITERrelevantgeometrical conditionsinthesensethatITERmirrorswillsitbehindapertures engineeredintotheneutronshieldingblocksofthediagnosticfirst wall.Anewexperimentwasthusproposedin2014bytheITER Organization(IO)toexposeanITER-likemirrorassemblyinJETto studywhetherunderexposuretorelevantplasmafluxes(eitherion fluxesduringglowdischargecleaningorcharge-exchangeneutral

http://dx.doi.org/10.1016/j.fusengdes.2016.12.016

0920-3796/© 2016 Published by Elsevier B.V.

2 Z Vizvary et al / Fusion Engineering and Design xxx (2017) xxx–xxx

Fig 1. Exploded view of ITER first mirror design.

fluxesduringplasmaoperation)wouldleadtoenhanceddeposition

asaresultoferosionofmaterialfromtheapertures.Thisworkwas

subsequentlyperformedunderIOContractandthispaperdescribes

theengineeringdesignofthisnew,ITER-likeFMT

The only available in-vessel support for this assembly is

a welded mounting bracket no longer used by other

deposi-tion/erosiondiagnostics.Testsonmock-upsandcalculationsdefine

themaximum loadforthis bracket Themirrors arevery close

totheplasma,resultinginconflictingelectromagneticand

ther-malrequirements.Thecomponentsneedtobesufficientlymassive

tocopewiththethermal loads(settinga minimumwall

thick-ness),butatthesametimeresistiveenoughtokeepthedisruption

loadswithinthoseallowedbythemountingbracket.Inaddition,

installationmust beperformedfully byRemote Handling only

Asaconsequence,thedesignevolvedintoafourpartstructure:

interface—support—housing—aperturecones(Fig.1).Wall

thick-nesseswereminimized,thehousingsurfacesareplasmasprayed

withalumina toinsulatethem and thesupportshape wasalso

designed minimizing theformation of current loops.The most

challengingcomponentstomanufacturewerethemulti-cone

aper-tures Thiswasnot suitable for conventional machining,hence

additivemanufacturingwasused

Theanalysiseffortwasfocusedonthestructuralintegrity of

thecomponentandespeciallyitsfixationtotheexistingunused

bracketintheJETvacuumvessel.Itisdrivenbythemassofthe

wholestructure and more importantly by the electromagnetic

loadswhichpeakduringdisruptions

Theeddycurrentloadsontheinitiallyproposeddesigncreated

momentsontherailwhichwerewellovertheallowablelimitsfor

thesupportbracket.Severaldesignchanges havebeenmadeto

reducetheseloads.Twoideasdrovethesechanges:

• Breakupcurrentloops:theresultingtorquesdependonthearea

enclosedbythecurrents

• Reducewallthicknessasmuchaspossiblethusincreasingthe

resistivityofthematerial

Thelatterismainlylimitedbythetemperaturein the

struc-tureduringplasmaoperation.Thestructuremusthavesufficient

thermalcapacitytoensurethatthepeaktemperaturestaysbelow

1200◦C(thelowerendofthemeltingtemperaturerangeofInconel

718),orevenlowerifthecomponenthasastructuralimportance

Electromagneticandthermal analyseshavebeencarried out

usingANSYStocheckthemechanicalloadsandthepeak

tempera-Table 1

Mechanical loads in toroidal, poloidal and normal directions.

tures.Theweldandboltstrengthwerethencheckedbyanalytical calculations

2.1 Transientthermalanalysis Transientthermalanalysishasbeenperformedinordertocheck themaximumtemperatureinthestructure.Theassumedheatload was300kW/m2,accordingtoJETdesigncriteriaformainchamber components.Theboundaryconditionsare200◦Catthebolt loca-tionsatthesupportbracketonthevacuumvesselwall;radiation

tothe200◦Cvacuumvesselwith0.5emissivityisalsoapplied.The heatloadisappliedfor20s.Althoughthissetupisquitesimplethe temperatureresultsshouldbeagoodindicationofwhetherthey areacceptable

Itwasfoundthatwallsoftheconescannotbereducedtoless than3mm,asthepeaktemperaturewiththiswallthicknessis alreadycloseto1000◦C.ThemeltingtemperatureofInconel718

isintherangeof1260–1336◦C,howevermechanicalproperties alreadybegindroppingovertherange650–700◦C.Sincethe aper-tureconeshavenootherstructuralrolethantosupporttheirown weight,thepeakcomputedtemperatureof∼1000◦Cisdeemed

acceptable

2.2 Electromagneticanalysis Thestructureisaffectedbyboththepoloidal(␪)andnormal (n)magneticfieldchangeduringdisruptions,thetoroidal()field variationisassumedtobezero.Theassumeddurationofdisruption

is10ms.Themagneticfieldandfieldvariationvaluesatthemirror locationare:

B=−3T,B=1.2T,Bn=0.4T

˙B=±120T/s, ˙Bn=±80T/s

Theeddycurrent analysishasbeencarried outusingANSYS [4].Tobeabletoobtainareasonablemeshthecadmodelofthe mirrorassembly had tobe simplified Since preliminary analy-sesshowed that there is a substantial contribution due tothe currentloopsfromboth thepoloidalandthenormal field vari-ation,itwasdecidedthatthesideplatesofthemirrorboxwill

beplasmasprayedandboltswillhavetophatstocuteddy cur-rentloopsandreducethetorquesactingonthemirrorbox.The absenceoftoroidalfieldvariationmeansthattheFEmodeldoes notevencontaintheseplates.Aseparateanalysisontheomitted platesshowedthattheelectromagnetictorquesareindeed negligi-ble(M=2.3·10−3Nm,M=7.8·10−3Nm,Mn=3.01·10−2Nm) AlthoughtheFEmodel isa muchsimplified versionof thereal structure,itisstillrepresentativefromtheelectromagneticpointof view.Evenwiththesimplificationsthegeometryiscomplicated;

itistherefore assumedthatthestructureis fullypenetratedby themagneticfield.Thiswillresultinanoverestimationandhence conservativeestimateoftheloads(Table1)

DuringtheFEanalysistheapertureconesandthebaseplate wereassumedtobestainlesssteel,followingtheoriginalmaterial choiceatthebeginningoftheproject.Subsequently,thedecision wastakentomanufacturetheminInconel718whichhasslightly higherresistivity.Asaresult,theinducededdycurrentsinduced willbeslightlylowerthanestimatedhere

Z Vizvary et al / Fusion Engineering and Design xxx (2017) xxx–xxx 3

Table 2

SLM Tensile Test Results (Batch no C1653D).

Testlog Sample ID E [GPa] 0.2% PS [MPa] UTS [MPa] Elon [%] R/A [%]

a Indicates if the specimen broke outside the middle 1/3 of the gauge length.

Thesupportbrackethasbeenweldedalongtwoedgestothe

ves-selwall.Theweldshavebeentestedbyaneccentricforce,whichis

usedasareferenceinouranalyticalcalculations.Thereservefactor

fortheweldwas1.4duetoelectromagneticloadforthefinaldesign

Thecalculatedstressfromthetestwasalsohigherthanthatofthe

combinedgravityandelectromagneticload.Thisgivesadditional

confidencethatthestrengthofthebracketweldsissufficient

Thesupportbrackethas4boltholesforM6bolts.Itwasdecided

thatall4willbeusedtowithstandtheelectromagneticloads

The aperture cones are made fromInconel 718using

addi-tivemanufacturingtechnology:selectivelayermelting(SLM).SLM

offerssignificantadvantages for JET in-vesselcomponents over

conventional machining including (a) more complex geometry

options,(b)rapidproductionofsmallbatchesand(c)littleorno

wastageofparentmaterial

AlthoughInconel718isawellknownmaterialinJET,duetothe

newmanufacturingtechnologyaqualificationprogramwasputin

place

Thequalificationprocesshasincluded:

• Mechanicaltests:

StatictensileatRT(RoomTemperature)andat450◦C

FatiguetestsatRT

• RGA(ResidualGasAnalysis)

• Porosityandchemicalanalysis

• MicrostructureusingSEM(ScanningElectronMicroscope)

• Mechanicalprooftestonaprototypeofadifferentcomponent(a

limiterassembly)

• Creeptesting(stillinprogress,theapertureconeswillnotoperate

inthecreepregime)

SLMpartsareproducedbylasermeltingapatternintoafine

layerofmetalpowderwhichislaidontoatable-mounted

base-plateinverythinlayers(about30␮mthick)whicharegradually

builtupintothefinishedcomponent.AnM270SLMmachinetable

(270mm×270mm)wasusedtoproducetestingsamplesandall

thepartsforthiswork

Thefirstbatchrequiredmorebuilds inordertodevelopthe

bestmethodforreducingdistortiononthefinishedparts,in

par-ticularforthemainbody.Eachbuildincludedfour10mmcubes

forchemical,porosityandmicrostructuretests,butthe

mechani-caltestpiecesweregeneratedinseparatebuildsasshown(Fig.2)

wherethepowderhadbeenremoved,priortoseparatingtheparts

fromthebase-plate

ThetensiletestresultsforthesamplesareinTable2.Thetable

includeswroughtInconel718propertiesforcomparison[5]

Fig 2.SLM Build C1653B.

WhilstnotstrictlynecessaryinordertoqualifytheSLMprocess forJET,itwasdecidedtoperformsomeadditionalmetallurgical examinationsinsupportoftheadoptionofSLMasasuitable man-ufacturingprocessforJETin-vesselcomponents

Theresultsofthesetestsallowthefollowingconclusionstobe drawn:

• AnearlybatchofSLMmaterialproducedpoorductilitybutthe reasonsfortheproblemwereunderstoodbythesupplieranda secondbatchwassuccessfullyproducedwithgoodductility

• TheuseofSA(SolutionAnnealed)ratherthanPH(Precipitation Hardened)materialisrecommendedasitoffersmechanical prop-erties(sufficientstrengthandductility)thataresuitableforthis application.Thisdoesnot,however,ruleouttheuseofPH mate-rialinSLMforotherapplications

• TestshavebeensuccessfullycompletedtoshowthattheSLM material haslow porosity and a sound micro-structure Out-gassingtestshavealsobeensuccessfullycompleted

• Aprototype(foradifferent,structurallyloaded,component)has successfullypassedmechanicalteststhatexceedtheexpected maximumoperationalloadsbyafactorof1.25:thisprototype wasmanufacturedusingSLMintheSAcondition

• AcostcomparisonhasshownthatSLMiscompetitivecompared withconventionalmachining

• ThisworkhasconfirmedthatSLMofferskeyadvantagesforJET in-vesselcomponents:

Flexibilitytomakepartswithcomplexgeometry

Rapidproductionofsmallbatches

4 Z Vizvary et al / Fusion Engineering and Design xxx (2017) xxx–xxx

Fig 3. Reflectivity of one of the mirror samples.

Fig 4. ITER First Mirror installed in JET.

Allmirrorswere pre-characterizedbeforeinstallationin the

ITER-likeholder.Themirrorsweremadeofpolycrystalline

molyb-denum.Totalanddiffusereflectivitiesweremeasuredinthevisible

andnearinfraredrange(400–1600nm).Themeasurementswere

performedusingatungstenhalogenlamp,aCCDspectrometerfor

thevisiblerange,anInGaAsphotodiodespectrometerforthenear

infraredrangeandanintegrating sphereof 80mmofdiameter

Fig.3showsthereflectivity tracesforone ofthemirrors.Total

reflectivityisabout55%inthevisiblerangeanditincreasesover

80%inthenearinfraredrange,whereasdiffusereflectivityis

main-tainedbelow4%acrossthestudiedspectralrange.Theothermirrors

presentedverysimilarresults,withadifferenceoflessthan2%

betweentraces

AnewITERFirstMirrortestassemblyhasbeendesigned,

ana-lysedandinstalledintotheJETvacuumvessel.Thestructurewas

installedremotelyonanexistingunusedbracketneartheoutboard

midplane, which imposed strong limitations on the combined weightandelectromagneticloadsinducedduringdisruptions.The mirrorsareveryclosetotheplasmaresultinginconflicting elec-tromagneticandthermalrequirements.Thecomponentsneeded

tobesufficientlymassivetocopewiththethermalloads(setting

aminimumwallthickness),butatthesametimeresistiveenough

tokeepthedisruptionloadswithinthoseallowedbythemounting brackets

Thefinaldesignincludedcomponentsthathavebeenproduced

byadditivemanufacturing,whosematerialqualificationprogramis alsopresented.Thisshowedthatthechosenmanufacturingprocess (selectivelayermelting)canbeadoptedasasuitablecandidatefor manufactureofcomponentsforuseintheJETvacuumvessel Theassemblywasinstalledinthe2014–15shutdown(Fig.4) andwillberemovedinthe2016–17shutdown

Acknowledgments

Thedesignandmanufactureofthemirrorassemblywasfunded

bytheITEROrganisationandtheinstallationwascarriedoutwithin theframeworkoftheContractfortheOperationoftheJET Facili-tiesandhasreceivedfundingfromtheEuropeanUnion’sHorizon

2020researchandinnovationprogramme.Theviewsandopinions expressedhereindonotnecessarilyreflectthoseoftheEuropean CommissionoroftheITEROrganisation.Toobtainfurther informa-tiononthedataandmodelsunderlyingthispaperpleasecontact publicationsmanager@ccfe.ac.uk

Theauthorsalsowouldliketoacknowledgethatthe success-ful completionof thisworkrelied onthededicatedinputfrom manypeopleincluding,inparticular,DanKirkfromCRDM(High Wycombe)andfromCCFE:RobLobel,JohnWilliams,NickPace, KevinCullandPaddyDoyle

References

[5] Special Metals Inconel 718 datasheet (Publication Number SMC-045) Table 19,