精选-AbaqusCAE-CONT-WK0S-DiskForging

QuestionW4Answer:QuestionW4Answer:QuestionW4Answer:QuestionW4Answer:QuestionW4Answer:T:Whichcontactformulation,smallorfinitesliding,doyouthinkisappropriateforthisanalysis?Why?Usefiniteslidingbecausetheblankisgoingtomoveanddeformagreatdeal.2:WhatiscausingthenegativeeigenvaluesinthisanalysistheLagrangemultipliersassociatedwiththecontactconstraintsoraninstabilityinthemodel?Instabilitiesduetosevereplasticstrainsarecausingthenegativeeigenvaluesinthismodel.Thenegativeeigenvaluemessagesalwaysappearwithwarningsaboutexcessiveplasticdeformation.3:Wheredoesitappearthatrefinementofthemodelisneeded?Alongthetopoftheblank,wherethereisalargecurvaturetothedie,andalongthebottomright-handside,whereagainthecurvatureofthedieislarge.4:Isthelocationofthelargestplasticstrains-thatis,thepeakvalueofPEEQ—whereyouwouldhaveexpectedittobe?Thelocationisrightbelowthetransitionregionofthedie,whichiswherethematerialhashadtoflowplasticallythemost,squeezingthesteelintothelargeropeninginthedie.5:ArethecontoursofCPRESSdifficulttovisualize?WhatcanbedonetomaketheCPRESScontourseasiertovisualize?Withlargedeformationsandrotationsintwodimensions,thecontoursofCPRESScanbehardtovisualize.TobettervisualizeCPRESS,trysweepingtheaxisymmetricmodelandplottingthecontoursontheundeformedblank.Fromthemainmenubar,selecViewODBDisplayOptions.IntheSweep/Extrudetabbedpage,toggleorSweepelements.ClickOK.ClicktheReplacetoolWinthetoolbar.InthepromptareaselectSectionsastheentitytypethenclickontheblank.ClickDonetocompletetheoperation.Fromthemainmenubar,selecPlotContoursOnDeformedShape.Usetheviewmanipulationtoolstoresizeandrepositionthemodelasnecessary.QuestionW4-6:Whyisthereasuddenincreaseintheforcerequiredtomovethedielateintheanalysis?Answer:Theincreaseinforceistheresultofthesuddenconstraintoftheflowofthematerialbytheedgeofthedie.Thesteelneedstodeformplastically,butmetalplasticitytheoryassumesincompressibledeformation.However,incompressibledeformationisnolongerpossibleoncethematerialisconstrainedbytheedgeofthedie.Thereactionforceincreasesaccordingly.QuestionW4^7:Wherewouldyouexpectthepeakplasticstrainstobelocated?Answer:Awayfromthecontactingsurface.QuestionW4-8:Canyouexplainwhytheregionofhighestplasticstrainhasshifted?Answer:Thelocationoftheplasticstrainschangesbecausethematerialalongthetopoftheblankcannotflowbecauseofthefrictionalongtheinterface.Thus,thematerialmustflowfrombelowthesurface,andthecenteroftheblankishighlystrained.QuestionW4-9:Intermsoftheglobalcoordinatesintheplaneofdefinition,inwhichdirectiondoesapositivevalueofCSHEAR1actalongthetopedgeoftheblank?Inwhichdirectionalongthesideoftheblank?Answer:RecallfromLecture7thatfortwo-dimensionalcontactsurfaces,theslipdirectiontisobtainedbya90clockwiserotationofthecontactdirectionn.Inthisexamplethecontactdirectionisthemastersurfacenormal.Youcanplotsurfacenormalsontheundeformedmodelshape.Alongmostofthetopoftheblank,apositiveCSHEAR1isinthenegativer-direction;thatis,thenegativeX-directionintheplaneofdefinition.Alongthesideoftheblank,trunsinthez-direction;thatis,theY-directionintheplaneofdefinition.QuestionW410:HowmuchmoreexpensivewastheanalysiswithLagrangefrictioncomparedtotheanalysiswiththepenaltymethodfrictionintermsofruntime?Intermsofrequiredmemory?Answer:TheruntimewithLagrangefrictionisabout75%longerthantheruntimeusingpenaltyfriction.TheLagrangefrictionmodelrequiredcontactstabilizationandnearlytwiceasmanyincrementsasthepenaltyfrictionmodel.ThesameamountofRAMwasneeded(comparethememoryestimatesinthEatfile).QuestionW4T1:Wastheadditionalcomputationalexpensejustifiedforthismodel?Answer:Theanswerswerethesameforeitherfrictionmodel(althoughthemagnitudeofthediereactionforcefortheLagrangefrictionmodelisslightlyhigherduringthemiddleportionoftheanalysisduetothecontactstabilization).Theadditionalexpensewasnotjustified.QuestionW4Answer:T2:Arethereanyadditionalchangestotheinputthataresuggestedformodelsthathavehighcoefficientsoffriction(>0.2).Withhighfrictioncoefficientstheunsymmetricsolvershouldbeused.Abaqus/Standardwillautomaticallyusetheunsymmetricsolverwhenever>0.2.QuestionW4Answer:T3:HowmuchCPUtimewassavedbyusingtheunsymmetricsolver?Morememoryisneededwiththissolver.Howmuchmorememorywasneededfortheanalysiswiththeunsymmetricsolver?NosignificantamountofCPUtimewassavedbyusingtheunsymmetricsolver.Thisisasmallmodel,andeachiterationtakesverylittletime.Theunsymmetricsolverneeded337iterations,andthesymmetricsolverneeded395.ThistinydifferenceisnotenoughtocauseanysignificantsavingsinCPUtime.Thememoryneededwasidenticalbecauseofthesmallsizeoftheproblem.QuestionW4Answer:T4:Didyouexpectthereactionforcesforthetwoanalysestobesimilarordifferent?Why?Reactionforcesshouldbethesamebecausetheunsymmetricsolvershouldnotaffectthesolutionobtained,justthetimeneededtogetit.QuestionW4Answer:T5:Isthisinterfaceconditionrealisticforthistypeofproblem?No.QuestionW4T6：Whattypesofproblemswouldyouexpectinthisanalysis?Why?Doyouthinktheanalysiswillbeabletocompletethefullloading?Answer:Expectseverelydistortedelementsand/orexcessiveplasticstrains.Theanalysiswillprobablyfailbeforethefullloadisapplied.QuestionW4T7:Aretheresultsconsistentwithyouranswerstotheearlierquestions?Answer:Theyshouldbeconsistent.