Experimental Characterization of a Composite Morphing Radiator Prototype in a Relevant Thermal Environment ChristopherL.Bertagne ,JorgeB.Chong†,JohnD.Whitcomb‡,DarrenJ.Hartl§ ⇤ TexasA&MUniversity,CollegeStation,Texas,77843 LisaR.Erickson¶ NASAJohnsonSpaceCenter,Houston,Texas,77058 Abstract Forfuturelongdurationspacemissions,crewedvehicleswillrequireadvancedthermalcontrolsystemstomaintaina desiredinternalenvironmenttemperatureinspiteofalargerangeofinternalandexternalheatloads. Currentradiators areonlyabletoachieveturndownratios(i.e. theratiobetweentheradiator’smaximumandminimumheatrejection rates)ofapproximately3:1. Upcomingmissionswillrequireradiatorscapableof12:1turndownratios. Aradiatorwith theabilitytoaltershapecouldsignificantlyincreaseturndowncapacity. Shapememoryalloys(SMAs)offerpromising qualitiesforthisendeavor, namelytheirtemperature-dependentphasechangeandcapacityforwork. In2015, the firstevermorphingradiatorprototypewasconstructedinwhichSMAactuatorspassivelyalteredtheradiatorshapein responsetoathermalload. Thisworkdescribesafollow-onendeavortodemonstrateasimilarconceptusinghighly thermallyconductivecompositematerials. Numerousversionsofthisnewconceptweretestedinathermalvacuum environmentandsuccessfullydemonstratedmorphingbehaviorandvariableheatrejection,achievingaturndownratio of4.84:1. Asummaryofthesethermalexperimentsandtheirresultsareprovidedherein. I. Introduction ForfuturecrewedspacecrafttargetingexplorationbeyondLowEarthOrbit(LEO)difficultthermalcontrolrequirements willneedtobesatisfied. Thethermalcontrolsystem(TCS)willbeexpectedtomaintainpreciseinternaltemperatures despite large large variations in the external thermal environment and internal system heat loads. In many cases the TCS will be required to reject a high heat load to a warm orbital environment and a low heat load to colder transitenvironments.1 Thisinverserelationshipbetweenthethermalenvironmentandtheheatrejectionneedsofthe mission/spacecraftrequiresthermalcontroldeviceswithahighturndownratio. Theturndownratio,definedasthe ratiobetweenthemaximumandminimumheatrejectioncapabilitiesoftheTCS,isaprimarymeasurementofTCS performance.2 Moreprecisely,thisratioinvolvesthemaximumheatrejectionrateinthehottestenvironmentandthe miminimumheatrejectionrateinthecoldestenvironment. TheturndownratioofaTCSisthereforemeasuredrelative totheparticularthermalenvironmentsinwhichitisdesignedtooperate. FuturemissionsbeyondLEOarepredictedto requireturndownratiosbetween6:1and12:1.3 Thehighestturndownratioastate-of-the-artTCSachievesis3:1(in LEO).4Here,theTCSrequiresaheattransferfluidwithalowfreezingpoint.5 Suchfluids,however,aretypicallytoxic (e.g. ammoniaontheInternationalSpaceStation),whichposeahazardtothecrew. ThislimitstheTCSarchitectureto atwo-loop(multi-fluid)designthatisolatesthecrewfrompossibleexposuretotheharmfulfluid. Several efforts have been undertaken to improve TCS performance through the development of variable heat rejectionradiators.5 Examplesincluderoll-outfinradiators,6 freezableradiators,2 digitalradiators,7 andvariable- emissivityradiators.8–11 Thisworksupportsthedevelopmentofanoveltypeofradiator,knownasavariable-geometry ormorphingradiator,12 andprovidesdetailsregardingthethermalcharacterizationofvariousprototypesthatwere built. ⇤NASASpaceTechnologyResearchFellow,DepartmentofAerospaceEngineering,3141TAMU,AIAAStudentMember †UndergraduateResearcher,DepartmentofAerospaceEngineering,3141TAMU ‡Professor,DepartmentofAerospaceEngineering,3141TAMU,AIAAMember. §AssistantProfessor,DepartmentofAerospaceEngineering,3141TAMU,AIAAMember. ¶ThermalEngineer,CrewandThermalSystemsDivision,2101NASAParkway. 1of11 AmericanInstituteofAeronauticsandAstronautics Shapememoryalloysareideallysuitedforthedesignofmorphingradiators. SMAsundergochangesinshapein responsetochangesintemperature;areversibleprocessduetothestress-andtemperature-dependenttransformation between two solid material phases: austenite and martensite.13 Applied as mechanical actuators on a morphable radiatortheywouldexploitenergytransferredbetweentheenvironmentandworkingfluidtochangetheradiatorshape withnoneedforexternalpower,control,orsensinginstrumentation. Sucharadiatorhasthepotentialtoachievethe highturndownratiosnecessarytoenablesingle-loopthermalcontrolofavehicleusinganon-toxic,high-freezing-point workingfluid,suchasapropyleneglycol/watersolution(PGW).Tradestudieshaveshownthatasingle-loopTCSwith amorphingradiatorsystemwouldreducetheTCSmassbyapproximately25%comparedwithastandardtwo-loop design.3 Thus, the morphing SMA radiator concept has the potential to revolutionize current TCS technology by decreasingsystemmassandcomplexity,whileincreasingversatility. II. Description of Morphing Radiator Concept Theprimarygoalofthisworkwastodevelopandconductexperimentalstudiesonastructurallyandthermallyoptimized compositemorphingradiatorpanel,orfacesheet. Thispaperfocusesprimarilyonthethermalanalysisconsiderations; the mechanical design is discussed separately.14 Figure 1 shows a morphing radiator design which combines the transformationresponseofshapememoryalloyswithathermallyconductiveandlinearlyelasticbiasingstructure, creatingaradiatorpanelthatreconfigurespassivelyinresponsetochangesintemperature.12 Theradiatorconsists ofacircularcompositepanelfixedalongthepanellineofsymmetryattheroot. Hereaflowtubeisattached,which conductsheatfromtheTCSworkingfluidintotheradiator. Ahigh-emissivitycoatingisappliedontheinner(concave) surface,shownwithdarkshading,andalow-emissivitycoatingisappliedontheouter(convex)surface,shownwith lightshading. SMAsattachedtotheoutermostsurfaceofthepanelcausetheradiatortomorphbetweenvariousshapes, alteringitsviewfactortospace. Thepaneltemperaturedrivesthisprocess. Whensufficientlycold,theradiatortakes onthefullyclosedcircularshapeshowninFig.1a. Astheradiatortemperatureincreasesduetoawarmerambient environmentand/orincreaseintheheatload(fromthefluidintheflowtube),theSMAscontract,whichopensthe radiatortoawiderconfigurationasshowninFig.1b;themaximumheatrejectionshapeisdepictedinFigure1c. This morphingbehaviorisintendedtobefullyreversible. Together,thevariableviewfactorandselectivesurfaceemissivity increasetheturndownratiooftheradiator,aswillbeshown. (a)Closedshapeforminimum (b)Semi-openshapeforintermediateheat (c)Openshapeformaximum heatrejection. rejection. heatrejection. Figure1: Schematicrepresentationofaflexiblemorphingradiatorpanel. Lightanddarkshadingrepresentslow-and high-emissivitycoatings,respectively. 14 AnarrayofmorphingradiatorpanelsisshowninFig.2,illustratingtheirapplicationinaparallel-flowconfiguration. Hotfluidenterstheradiatormanifoldviatheinletheader,whichdistributestheworkingfluidamongmultipleparallel flow tubes. A series of individual morphing radiator panels are attached along each flow tube. The working fluid temperaturedecreasesalongthelengthofeachtubeasheatisrejectedviaradiation. Inawarmenvironment(e.g. LEO), theworkingfluidandpaneltemperaturesremainabovethemartensitestarttemperature,suchthattheSMAsarefully contractedandallpanelsarefullyopen. Thus,heatisrejectedthroughthegreatestpossibleradiatorareaviathehigh emissivitysurface. Inacoolerenvironment,thetemperaturenearthedownstreamendoftheflowtubedecreasesbelow themartensitetransformationtemperature,causingtheSMAstoexpand. Thisallowsthepanelstoclose,minimizing heatrejection. Asaresultofsuchbehavior,thisradiatordesignwilltendtomaintainaminimumfluidtemperatureinthe rangeofthetransformationtemperaturesoftheSMA(i.e.,betweentheausteniteandmartensitefinishtemperatures15), whichcanvastlysimplifytheoveralldesignoftheTCS. 2of11 AmericanInstituteofAeronauticsandAstronautics Figure2: Illustrationofanarrayofradiatorpanelsinaparallelflowconfiguration. Arrowsindicatefluidflowdirection. This concept featured a number of design challenges. In particular the face sheet required three contradictory material characteristics: adequate thermal conductivity to transport heat out of the fluid loop, ample flexibility to deformtoitsopenhotshape,andsufficientstiffnesstoprovidespringforceforreturningtheradiatortoitsclosedcold shape. Figure3showstheprototypemorphingradiatorusedintheexperimentalstudiesin2015.16 Thisprototypewas designedtousewire-typeSMAstoachievethetemperature-inducedmorphingbehaviordepictedabove. (Notethat Fig.3acorrespondstoFig.1c,withtheSMAsintheiraustenitic(high-temperature)state.) Theprimarycomponentin thisprototypeisacompliantandthermallyconductivecopperpanel(7inlong,3inwide,0.007inthick),whichwas rolledalongitslengthtoformthesemicircularshapeshowninFig.3. Tenshapememoryalloywiresrestagainstthe outersurfaceofthepanelandarefixedateachendtothestraightedgesofthepanelwithapairofterminalblocks fabricatedfrom0.25insquarealuminumstock. Thewiresareotherwiseunconstrained,allowingthemtoslidealongthe panelastheytransformlocally. Benchtoptestsindicatedtheneedforanadditionalbiasingforcebeyondthatprovided bythecoppertodrivethepaneltowardstheclosedcircularshapeundercooling. A1inwide, 0.007inthick1095 steelclosingspringwasattachedtotheconvexsideofthepanelforthispurpose. Inordertoincreasetherateofheat rejectionviaradiationintheopenshape,theinsidesurfaceofthecopperpanelwaspaintedwithAeroglazeZ307®,a flexible,high-emissivitypolyurethanecoating. Theoutsideofthepanelremainedunpainted. Theemissivitiesofthe AeroglazeZ307paint,unpaintedcopper,andunpaintedsteelweremeasuredtobe0.943,0.047,and0.143respectively, withaSurfaceOpticsCorp.ET-100®emissometer. Thecopperpanelwasattachedtoa0.375indiameterstainlesssteel flowtubeusingathermallyconductiveepoxywhichallowedtheradiatortobeintegratedintoapumpedfluidloop fortheexperiment. Theprototypewastestedinathermalvacuumenvironmentwhereitdemonstratedthemorphing behaviorunderarangeofheatloadsandachievedaturndownratioof6.4:1.16 Thegoaloftheeffortsdescribedhere(andintheaccompanyingwork14)wastoimprovebeyondthepreliminary prototypeandstudyof2015, producinganewdevicecomprisedofmoreadvancedcompositematerialsandSMA componentswithtransitiontemperaturesbettertunedtothethermalvacuumenvironment. Acarbon-fiberfacesheetwas developedforthispurpose. Studiesweredonetooptimizethedesign,takingintoconsiderationfiberorientationand SMAintegration. PrototypefacesheetswerefabricatedusingmultiplepliesofCSWT40-800/5320-1,aunidirectional prepregtapedesignedforout-of-autoclavecuring.17 Thefullprocessofmaterialselection,manufacturing,andfatigue testingoftheseadvancedmorphingpanelsisdescribedintheaccompanyingpaper.14 Thefollowingsectionsdescribea benchtopdemonstrationandthermalvacuumtestingoftheseprototypes. 3of11 AmericanInstituteofAeronauticsandAstronautics Steel closing spring Flow tube SMA wires 33 High-emissivity 22 coating 11 Copper panel 99 10 88 66 77 55 Flow tube 44 Set screws Aluminum terminal block (b) Inside of prototype showing high-emissivity (a)Outsideofprototypeshowingprimarycomponents. polyurethanecoating. Figure 3: Prototype morphing radiator test article from previous 2015 studies.16 Panel is at room temperature demonstratingopenshape. Opencirclesshowlocationswherethermocoupleswereattachedfortesting. III. Benchtop Prototype Afterfabricationandpriortothermalvacuumtesting,abenchtopprototypewasfabricatedasaproof-of-concept. This prototype(showninFig.4)featuredanumberofSMAwiresinthemartensitephaseatroomtemperature(i.e. thepanel wasintheclosedconfigurationatroomtemperature(Fig.4a)). Thepanelwasheatedusingtwoheatgunstosimulate theheatloadfromthefluidlooponthespacecraft. Theincreaseintemperaturecausedthephasetransformationinthe SMAwires,drivingthepaneltoitsopenconfiguration(Fig.4b). Whentheheatwasremovedthecompositecooled convectivelyinthelaboratoryenvironmentandreturnedtotheinitialclosedconfiguration. Inthismanner,theshapeof theradiatorwascontrolledbythermalloadingexclusively,andthemorphingbehaviorwasdemonstratedinanambient roomtemperaturesetting. (a)Closedconfigurationcorrespondingtominimumtemper- (b)Openconfigurationcorrespondingtomaximumtemper- ature. ature,resultingfromheatprovidedbytwoheatguns. Figure4: Ademonstrationofthebenchtopprototype. 4of11 AmericanInstituteofAeronauticsandAstronautics IV. Thermal Vacuum Chamber Tests A. TestArticles Followingthebenchtopproof-of-concept,threedifferenttestarticleswerefabricatedfortestinginathermalvacuum environment. Table1givesanoverviewofthecharacteristicsofeachtestarticle. Thefirstrowdescribesthecarbonfiber plylayering(i.e. thelayup)asasequenceofplyorientationangleswithineachcompositepanel. Theplyorientation anglesweredefinedwithrespecttothecircumferentialdirection(i.e. thefibersofa0 plyarealongthecircumference � ofthepanelandthefibersofa90 plyareorientedparalleltotheflowtube). Thekeymechanicalandthermalproperties � ofthepanel,includingbendingstiffnessandeffectiveconductioncoefficientperunitlength(denotedbyK),arealso t listedforeachpanel,alongwiththeprimarydimensions. Moredetailsoftheseaspectsofpaneldesignareprovidedin theaccompanyingwork.14 Table1: Overviewofthethreetestarticlesfabricatedforthethermalvacuumchambertests. TestArticleA TestArticleB TestArticleC Layup [90/45/0/45/90] [90/45/0/0/45/90] [90/45/0/0/0/45/90] Bendingstiffness 0.44Pam3 0.82Pam3 1.42Pam3 · · · Cond. coeff. perunitwidth(K) 0.091W/K 0.121W/K 0.152W/K t Closeddiameter 3.0in 3.0in 3.5in Width 3.0in 3.0in 4.0in SMAtype&qty. Wires(x18) Wires(x36) Strips(x8) SMAphaseatroomtemp. Austenite Austenite Martensite Thefirsttwotestarticles,AandB,werenearlyidenticalapartfromthelayup. TestArticleAincorporatedalayup withasingle0 ply,whilethelayupforTestArticleBincludedtwo0 plies. BothofthesetestarticlesutilizedSMA � � wiresintheaustenite(high-temperature)phaseatroomtemperature,requiringTestArticlesAandBtobeassembled intheiropenconfigurations. ForTestArticleA,asingleSMAwirewasthreadedthrougheachholeintheterminal blocksforatotalof18wires. SinceTestArticleBfeaturedtwo0 pliesthebendingstiffnesswasapproximatelytwice � thatofArticleA(seeTable1),thusrequiringapproximatelytwicetheforcetoopentothesameradius. Tomaintaina comparablelevelofstressinthewiresbetweenthetwotestarticles,twoSMAwireswerethreadedthrougheachholein theterminalblockonTestArticleB.Figure5showsTestArticleApriortothermalvacuumtesting. Afterattachingthewirestothepanels,theflowtubeswereattachedusingthermaladhesiveandsecuredwitha thermallyinsulatedtape. Thermocoupleswereattachedtothefacesheetatvariouslocationstorecordtemperature acrossthepanel. Figure5: TestArticleAfeaturingSMAwiresinstalledinastressed(open)configuration. Test Article C was slightly larger than Test Articles A and B and used SMA strips instead of wires. Whereas theSMAwireswereintheaustenitephaseatroomtemperature,theSMAstripsusedinTestArticleCwereinthe martensite(low-temperature)phaseatroomtemperature. Thus,ArticleCwasassembledintheclosedconfiguration. Priortoassembly,eachstripwasdetwinnedat300MPatogeneratethetransformationstrainneededforausteniteshape 5of11 AmericanInstituteofAeronauticsandAstronautics recoveryduringheating. Thisprocessinvolvedsubmergingthestripsinliquidnitrogentoreachthemartensitephase throughoutthematerial,andthenapplying300MPaofaxialtensiontoeachstrip. Figure6showsTestArticleCfully assembled. Figure6: TestArticleCfeaturingSMAstripsinstalledinadetwinnedandstressfree(closed)configuration. B. TestSetup Thevacuumchamberusedintheexperimentswasasmall,high-vacuumthermalenvironmentchamberlocatedatNASA JohnsonSpaceCenter. Figure7showsaschematicdiagramoftheflowloopusedinthethermalvacuumchambertests. TheprimarycomponentwasanSPScientificRC211®,arecirculatingchillershownontheleftsideofFigure7. The chillercontainsapumpandintegratedheaterandchillermodules,allowingthetemperatureoftheworkingfluidtobe controlledbetween-45 Cand100 C.ThefluidloopusedDynaleneHC-50® —anontoxicwater-basedcoolantwitha � � freezingpointbelow-50 C—astheworkingfluid. Thetemperatureboundsforthetestwereselectedtoensurethe � fluiddidnotfreezeduringtesting. Beginningatthechiller,thefluidfirstpassedthroughtwoparallelflowlines,onewithaninlineheater,andtheother thatbypassedtheheater. Theselineswereopenedandclosedviaballvalvesduringthedifferentphasesofexperiment toadjusttheflowtemperatureasdesired. DownstreamofthesevalvesthefluidpassedthroughaCoriolisflowmeter manufacturedbyMicroMotion®,whichwasusedtomeasuretheinstantaneousflowratethroughouttheexperiments. Thefluidthenenteredthevacuumchamberandpassedthroughtheflowtubecontactingthetestarticle. Twoimmersion thermocouples,shownwithnumberedcirclesinFig.7,wereusedtomeasurethefluidtemperaturesattheinletand outletoftheradiatorflowtube. Finally,thefluidexitedthevacuumchamber,passedthroughacheckvalve,andreturned tothechiller. Figure8showsthethermalvacuumchamber,chillercart,andfluidloop. Figure9showsaclose-upofthetest sectioninsidethechamberwiththeradiatorinstalledintothefluidloop. Figure7: Schematicdiagramshowingexperimentalsetupforthethermalvacuumchambertest. 6of11 AmericanInstituteofAeronauticsandAstronautics (a)Testsetupshowingthermalvacuumchamber,chiller (b)Closerviewofthermalvacuumchambertestsection cart,andexteriorfluidlines. showinginteriorfluidlinesandLEDlamps. Figure8: Photographsofthevacuumchambertestsetup. Figure9: Close-upofmorphingradiatortestarticleafterbeinginstalledinthefluidloop(TestArticleBshown). C. TestProcedure Priortothestartofeachtest,theappropriatetestarticlewasmountedtotheflowlineinsidethechamber(seeFig.7). Thechamberwasthensealed,avacuumestablished,andthefluidlinesopened. Thefluidtemperaturewascycled overa 3hourperiod,duringwhichtimethemorphingbehaviorwasobserved. Temperaturedatawasrecordedat ⇠ variouslocationsonthepanelandflowtubebythethermocouplesthroughoutthetest. Additionally,acameramounted outsideawindowonthevacuumchambercapturedimagesofthepanelat2secondintervalsthroughouttheheatcycling process. Figures 10and 11showTestArticleAinsidethethermalvacuumchamberduringtesting. (a) Open configuration corresponding to maximum fluid (b)Closedconfigurationcorrespondingtominimumfluid temperature. temperature. Figure10: ImagesofTestArticleAinthevacuumchamberphotographedbythecameraoutsidethechamber. 7of11 AmericanInstituteofAeronauticsandAstronautics (a) Open configuration corresponding to maximum fluid (b)Closedconfigurationcorrespondingtominimumfluid temperature. temperature. Figure11: ImagesofTestArticleAinthevacuumchambercapturedbythein-chambercamera. V. Analysis of Experimental Data AcustomMATLAB® scriptprocessedtheimagesofthetestarticlescapturedduringthetest,computingthepanel radiusovertimeviaaleastsquaresregressiontechnique. TestArticlesAandBachievedarangeinpanelradiusof1 inchand1.1inchesrespectively. TestArticleConlyachievedaradiusrangeof0.4inches. As the radius increased with fluid temperature, the view factor of the internal panel surface increased, thereby increasingthetotalradiativeheatrejectionofthepanel. Thus,therangeofradiidirectlydrovetherangeofachievable heatrejectionrates,establishingaturndownratioforthepanel. Tothisend,theanalysisofheatrejectioncentered aroundTestArticleB,whichachievedthewidestrangeinradius. Thetemperaturehistoriesandpanelradiusovertime areshownforTestArticleBinFigure 12. AthermalfiniteelementmodelofTestArticleBwasdevelopedinABAQUS®. Byspecifyingthepanelradiusand boundaryconditionsknownatvariousinstancesinthethermalvacuumtest,athermalanalysiswasperformedwhich providedthesteady-statetemperaturedistributionandheatrejectionrateofArticleBateachinstance. Fromthese simulationsthemaximumandminimumheatrejectionsofTestArticleBweredetermined;thusallowingcalculationof theturndownratio. Thesimulationalsodirectlycalculatedtheminimumtemperatureatthefreeedgesofthepanel(i.e. attheterminalblocks)tocomparewiththeminimumpaneltemperaturerecordedinthethermalchambertest. The time history of heat rejection throughout the testing process for Test Article B is shown in Fig. 13a. The comparison of minimum panel temperature between these predictions and the actual data from Test 2 is shown in Fig. 13b. The overall accuracy of the minimum temperature predictions serves as a meaningful validation of the thermalmodelandtherelatedresultsthatfollow. Intheopenconfiguration(hotphase,maximumradius)themodel calculatedaheatrejectionrateof9.97W,withadiscrepancyinminimumtemperatureresultsofonly 7 C.Inthe ⇠ � closedconfiguration(coldphase,minimumradius)aheatrejectionrateof2.06Wwascalculated,withatemperature discrepancyof 3 C.Thus,theoverallturndownratiowascalculatedat4.84:1. Thelargertemperaturediscrepancy ⇠ � forthehotphasewaslikelyduetothelargetemperaturegradientexistingbetweentheradiatorroot(pointofcontact withtheflowtube)andtheterminalblocks( 26 C).Thisgradientismuchlesssignificantinthecoldphase,witha ⇠ � temperaturevariationof 4 C,explainingthesmallertemperaturediscrepancyinthecoldphase. ⇠ � Asacomparison,theturndownratiowascomputedforArticleBwithoutincludingchangesduetomorphing. The radiatormodelwasgiventhesameradiusandsinktemperatureforwhichthemaximumheatrejectionwascomputedin Fig.13a,whiletheroottemperature(thepaneltemperatureboundarycondition)wasthatoftheminimumheatrejection case. Fromthissimulationtheturndownratiowithoutmorphingwasfoundtobe4.32:1. Thisdemonstratedanincrease inturndowncapacitythroughthemorphingbehavior. Earlierinthispaperitwasstatedthatthehighestturndownratioofastate-of-the-artTCSwas3:1. Itisimportant toreiteratethatthisisrelativetotheparticularthermalenvironmentinwhichsaidthermalcontrolsystemoperates (indicated above to beLow Earth Orbit). The turndown ratiosachieved by themorphing radiator areparticular to thethermalenvironmentsexperiencedinthethermalvacuumchamber. Thoughsimilar,theenvironmentsofthetwo systemsarenotidentical,whichshouldbenotedwhencomparingtheirturndownratios. Regardingtheturndownratioachievedbythemorphingradiator,4.84islowerthanrequiredforanactualproduction system(targetedbetween6:1and12:1asmentionedearlier). Thisturndownratiowaslargelyimpactedbyheatthat escapedthroughthecircularopenendsoftheradiatorduringthetestingprocess;asourceofuncontrolledheatrejection. Previousdesignstudieshaveshownthattheadditionoflow-emissivitysidecapswouldblockheatrejectionthroughthe 8of11 AmericanInstituteofAeronauticsandAstronautics openendsandsignificantlyreduceheatlosses,andthatlongcylindricalradiatorswithsidecapsareinfactneededto ensureoptimalperformance.18 Futureworkwillinvolvemodelingtheradiatorwiththesefeaturestoexploretheeffects onturndowncapacity. (a)TimehistoryofTestArticleBpaneltemperatureduringtesting.Thermocoupleswereattachedtotheradiator atvariouslocations,resultingintherangeoftemperaturesshown. (b)TimehistoryofTestArticleBpanelradiusduringtesting.Theradiusofcurvatureofeachsideofthepanelis shown. Figure12: ExperimentalresultsfromthetestofArticleB(theprototypewiththestrongestmorphingresponse). 9of11 AmericanInstituteofAeronauticsandAstronautics (a)RateofheatrejectionofTestArticleBthroughoutthermalvacuumtest. (b)ComparisonofminimumpaneltemperaturecalculatedinAbaqussimulationwiththatmeasured inthethermalvacuumexperiment. Figure13: ResultsfromthermalanalysisofTestArticleBmodel. VI. Summary & Future Work With the increasing need for advanced thermal control systems a variable heat rejection radiator offers a unique solution. Acompositemorphingradiatoractuatedwithshapememoryalloyswasdesignedtoachievethispurpose. Multipleprototypesweredevelopedandtestedinathermalvacuumenvironmentwheretheyexhibitedthedesired temperature-inducedactuationbehavior. TestArticleB,featuringtwocircumferentiallyorientedcarbonfiberpliesand 36SMAwiresachievedthewidestrangeinradiusduringthetestingprocessandofferedaturndownratioof4.84:1. Themorphingbehaviorwasshowntoaidtheturndowncapacity,whichwas4.32:1withoutmorphing. TheseeffortshaveshownthatSMAmaterialsarecapableofmorphingacompositeradiatorpanelthrougharange ofshapeconfigurations,allowingforvariableheatrejectionabovethatachievedbymodernthermalcontrolsystems. Movingforward,theprojectwillfocusondesignimprovementsinvolvingSMAandflowtubeintegrationtoimprove heattransferbetweentheworkingfluidandthepanel,thusimprovingthemorphingresponse. Furthermore,theradiator willbemodeledwithsidecapstoreduceuncontrolledheatrejectionandincreasetheoverallturndowncapacityofthe 10of11 AmericanInstituteofAeronauticsandAstronautics