ppt1、接触CONT-L03-SurfaceBasedContact

Thefollowingtopicsarecoveredinthislesson.Lessoncontent:Surface-basedContactWorkshop2:LapJointAnalysis(IA)Workshop2:LapJointAnalysis(KW),Lesson3:Surface-basedContact,2hours,Bothinteractive(IA)andkeywords(KW)versionsoftheworkshopareprovided.Completeonlyone.,Surface-basedContact,ContactFormulationsContactDiscretizationContactEnforcementMethodsRelativeSlidingBetweenBodiesOutputofContactResultsSummary,ContactFormulations(1/2),BasicfeaturesofacontactformulationContactdiscretizationConstraintenforcementContacttracking(relativesliding),Theexampleshownherecomesfrom“Submodelingofastackedsheetmetalassembly,”Section1.1.19oftheAbaqusExampleProblemsManual.,ContactFormulations(2/2),ContactdiscretizationWhereistheconstraintapplied?Node-to-surfaceSurface-to-surfaceConstraintenforcementHowistheconstraintenforced?Direct(Lagrangemultipliers)PenaltymethodAugmentedLagrange(Lagrangemultiplierscombinedwithpenaltymethod)Contacttracking(relativesliding)Howdoestheconstraintevolve?FiniteslidingSmallsliding,ContactDiscretization(1/17),Thelocationsofthecontactconstraintsdependonthecontactdiscretizationtechniqueused:Node-to-surfacetechnique:DefaultmethodforcontactpairsNotavailableforgeneralcontactNodesononesurface(theslavesurface)contactthesegmentsontheothersurface(themastersurface).Contactisenforcedatdiscretepoints(slavenodes)Surface-to-surfacetechniqueOnlymethodforgeneralcontactAlternativemethodforcontactpairsThemethodconsiderstheshapeofboththemasterandslavesurfaces.Contactisenforcedinanaveragesenseovertheslavesurface.,ContactDiscretization(2/17),Whenthesurface-to-surfacediscretizationtechniqueisusedwiththecontactpairapproach:ManyofthedistinctionsbetweencontactpairsandgeneralcontactdisappearTheprimarydifferencebetweenthemreducestoeaseofusevs.performance.Abriefdiscussionofthediscretizationmethodsfollows.Unlessotherwisenoted,thediscussionregardingthesurface-to-surfacediscretizationthroughoutthiscourseisindependentofthechoiceofcontactpairsorgeneralcontact.,ContactDiscretization(3/17),Node-to-surface(N-to-S)contactdiscretizationTraditional“point-against-surface”methodContactisenforcedbetweenanodeandsurfacefacetslocaltothenodeThenodeisreferredtoasa“slave”node;theopposingsurfaceiscalledthe“master”surfaceThisformulationisalsocalledastrictmaster/slaveformulation,ContactDiscretization(4/17),Theslavesurfaceisessentiallytreatedasacollectionofpoints(thesecorrespondtothenodesoftheslavesurface)Theopeningorpenetrationdistanceismeasuredalongthenormaltothemastersurface(i.e.,contactnormalisalongmastersurfacenormal)Anodalareaisassignedtoeachslavenodetoconvertcontactforcestocontactstresses(tractions)Foranelement-basedslavesurface,thisareaisbasedonthesizesofadjacentfacesForanode-basedslavesurface,theuserassignsthisarea,ContactDiscretization(5/17),Keyimplicationsofthestrictmaster/slaveformulation:Slavenodescannotpenetratemastersurfacesegments.Nodesonthemastersurfacecanpenetrateslavesurfacesegments.AgrosspenetrationofamastersurfaceintoacoarselymeshedslavesurfacecausessolutioninaccuracyRefinementoftheslavesurfaceimprovescontactresolutionAssignmentofmasterandslaverolesThemore-refinedsurfaceshouldactastheslavesurfaceThestifferbodyshouldbethemasterTheactivecontactregionshouldchangemostrapidlyonthemastersurface(minimizescontactstatuschanges),ContactDiscretization(6/17),Whilerefinementoftheslavesurfaceleadstoglobalaccuracy,localcontactstressoscillationsmaystillbeobservedwithN-to-SUseofa“matchedmesh”acrossthecontactinterfacewilleliminatethissolutionnoise,ContactDiscretization(7/17),Surface-to-surface(S-to-S)contactdiscretizationContactisenforcedinaweightedsense,betweentheslavenodeandalargernumberofmastersurfacefacetssurroundingit.Ineffect,contactissmearedoveralargernumberoffacets.Contactisenforcedinanaveragesense.,ContactDiscretization(8/17),Opening/penetrationandsliparecomputedinanaveragesenseoverafiniteregionperconstraintOpening/penetrationmeasuredalongtheslavesurfacenormalSlipsaremeasuredperpendiculartotheslavenormalTheslaveandmastersurfacesarebothtreatedascontinuoussurfacesTheslavesurfacemustbeelement-basedRecallN-to-Scomputespenetrationandslipatadiscretepoint(slavenode),ContactDiscretization(9/17),Thesurface-to-surfacediscretizationoffersmanyadvantagesoverthenode-to-surfacediscretization,including:IncreasedcontactstressaccuracyReducedsnaggingofsurfacesReducedsurfacepenetrationsReducedsensitivitywithrespecttochoiceofmasterandslaverolesTheseareillustratednext.,ContactDiscretization(10/17),AdvantagesofS-to-SoverN-to-SS-to-Softenimprovestheaccuracyofthecontactstresses.Contactissmearedoveralargerareaofthemastersurface.Example:classicalHertzcontactproblem:Thecontactpressurecontoursaremuchsmootherandthepeakcontactstressisinverycloseagreementwiththeanalyticalsolutionusingthesurface-to-surfaceapproach.,Node-to-surface,AnalyticalCPRESSmax=3.01e+05,Surface-to-surface,CPRESSmax=3.425e+05,CPRESSmax=3.008e+05,ContactDiscretization(11/17),S-to-Sreduceslikelihoodofsnagging,Node-to-surface,Surface-to-surface,Computingaveragepenetrationsandslipsoverfiniteregionshasasmoothingeffectthatavoidssnagging,Treatingtheslavesurfaceasacollectionofpointscantriggersnaggingasslavenodestraverseacorner,ContactDiscretization(12/17),S-to-Sreducesthelikelihoodofmasternodespenetratingaslavesurface,Somepenetrationmaybeobservedatindividualnodes;however,large,undetectedpenetrationsofmasternodesintotheslavesurfacedonotoccur.,ContactDiscretization(13/17),S-to-Sismuchlesssensitivetothechoiceofmasterandslavesurfaceswhenthemeshdensitiesaresimilar.,Choosingtheslavesurfacetobethefinermeshwillstillyieldbetterresults;choosingthemastersurfacetobethemorerefinedsurfacewilltendtoincreaseanalysiscost,ContactDiscretization(14/17),S-to-Stakesintoconsiderationshellandmembranethicknesseswhenperformingcontactcalculations.N-to-Sonlyconsidersthiseffectforthesmallslidingformulation.,Thicknessignored,Thicknesstakenintoaccount,ContactDiscretization(15/17),Effectonsolutioncost:S-to-Scanincreaseper-iterationsolutioncostReason:morenodesperconstraintthanN-to-SHowever,S-to-SoftenreducesthenumberofNewtoniterationsrequiredtoobtainaconvergedsolutionThishelpsoffsetthehigherper-iterationcost.,ContactDiscretization(16/17),Summary:N-to-Svs.S-to-S,ContactDiscretization(17/17),Summary(contd),ContactEnforcementMethods(1/13),InAbaqus/Standard,thedefaultcontactbehavioris“hard”contact“Hard”contactisdepictedbelow;thebehaviorisdescribedbyacontactpropertyknownasthepressure-penetrationcurve(alternativebehaviorcanbespecified;discussedlater).Thedesiredbehavior(nopenetration)isachievedusinganenforcementmethod,ContactEnforcementMethods(2/13),ThreenumericalmethodsareavailableinAbaqus/Standardtoachieveorapproximate“hard”contactconditions:DirectenforcementmethodStrictenforcementofpressure-penetrationrelationshipusingtheLagrangemultipliermethodPenaltymethodApproximateenforcementusingpenaltystiffnessAugmentedLagrangemethodApproximateenforcementusingpenaltymethodwithaugmentationiterations;notdiscussedfurtherhere,ContactEnforcementMethods(3/13),Thepenaltymethodisusedbydefaultforeachofthefollowing:GeneralcontactContactpairswiththefinite-sliding,surface-to-surfaceformulationAutomaticcontactpairdetectioncapabilityinAbaqus/CAETheaugmentedLagrangeisusedbydefaultfor:3Dselfcontactwithnode-to-surfacediscretizationOtherwise,directenforcementisthedefault,MutuallyexclusiveparametersDIRECT,PENALTYitsmagnitudeishigherinAbaqus/StandardthaninAbaqus/Explicit.,ContactEnforcementMethods(10/13),Forcasesinwhichthedefaultpenaltystiffnessisnotsuitable,optionstoscalethepenaltystiffnessareavailableOrder-of-magnitudechangesarerecommendedIfthescalefactor100,AbaquswillautomaticallyinvokeavariantofthemethodthatusesLagrangemultiplierstoavoidill-conditioningissues,Multiplicative!,Stepdependent(careful!),Keywordinterface*SURFACEINTERACTION*SURFACEBEHAVIOR,PENALTYpenaltystiffness,clearanceoffset,scalefactor(alloptional):*STEP:*CONTACTCONTROLS,STIFFNESSSCALEFACTOR=value,ContactEnforcementMethods(11/13),PenaltystiffnessmagnitudeStifforblockyproblemsThedefaultpenaltystiffnessgenerallyproducesresultsthatarecomparableinaccuracywiththoseobtainedwiththedirectmethodUsuallyrequireslessmemoryandCPUtimeBending-dominatedproblemsThedefaultpenaltystiffnesscanoftenbescaledbackbytwoordersofmagnitudewithoutanysignificantlossofaccuracyScalingbackthepenaltystiffnessforbending-dominatedproblemssometimesdramaticallyincreasestheconvergencerate,ContactEnforcementMethods(12/13),Example:HertzcontactExample:Three-pointbendingspecimen,1.5%difference,47%difference,Nearlyidenticalresults,14%reduction,ContactEnforcementMethods(13/13),Summary:directenforcementvs.penaltymethod,RelativeSlidingBetweenBodies(1/9),Abaqusoffersfinite-andsmall-slidingversionsofS-to-SandN-to-ScontactFinite-slidingformulation:generalapplicabilityPointofinteractiononmastersurfaceisupdatedusingthetruerepresentationofthemastersurfaceSmall-slidingformulation:approximationintendedtoreducesolutioncost;limitedapplicabilityPlanarrepresentationofmastersurfaceperslavenodebasedontheinitialconfigurationOnlyavailableforcontactpairs,RelativeSlidingBetweenBodies(2/9),FiniteslidingKinematically,finiteslidingofdeformablebodiesagainsteachotherallowsforthemostgeneralrelativemotionbetweenthebodies:Arbitrarilylargeslidingisallowed.Arbitrarilylargerotationsanddeformationsofthesurfacesareallowed.,Node102cancomeintocontactanywherealongthemastersurface.Whileincontact,node102isconstrainedtoslidealongthemastersurface,regardlessofitsorientationordeformedshape.,RelativeSlidingBetweenBodies(3/9),Afinite-slidingformulationisusedbydefaultOnlyfiniteslidingcanbeinvokedwithgeneralcontactAnalternativesmall-slidingformulationisavailableforcontactpairs(discussedlater),*contactpair,interaction=.bearing,shaft,RelativeSlidingBetweenBodies(4/9),Self-contactofasinglebodySelf-contactismeantforcasesinwhichitisdifficulttopredictbeforehandwhichpartsofabodywillcomeintocontactwithitself.Bydefinition,finiteslidingisrequired.Generalcontact:Automaticallyconsideredbydefault(self-contactofanall-inclusivesurface).*CONTACT*CONTACTINCLUSIONS,ALLEXTERIORContactpairs:Usejustasinglesurface:*CONTACTPAIR,INTERACTION=SEALOUTER,orusethesamesurfacetwice:*CONTACTPAIR,INTERACTION=SEALOUTER,OUTER,RelativeSlidingBetweenBodies(5/9),OverconstraintissuesrelatedtoselfcontactAsymmetricmaster-slaveformulationisusedforself-contact:Everysinglenodeofthesurfaceactsasaslavenode,andeveryelementfaceformingthesurfaceactsasamastersurface.Thismaycauseanoverconstraintinthesystemofequations.Thiscanleadtonumericaldifficulties,suchassingularitiesandzeropivots.Penaltycontactenforcementisusedbydefaultforself-contacttomitigateoverconstraintandredundantconstraintissues.Anexceptionismadein2-Dwhenasurfacefoldsontoitselfformingasharpcorner;inthatcasethenodesadjacenttothevertexfollowastrictmaster-slavealgorithm.,RelativeSlidingBetweenBodies(6/9),SmallslidingSmallslidingassumesthattherelativemotionperslavenoderemainssmallcomparedtothelocalcurvatureofthemastersurfaceEveryslavenodeinteractswithitsownlocalslideplane(in2D/AXIitisdepictedasaline);thisslideplaneisaplanarapproximationofthemastersurfaceandisassumedtobeareasonablyaccuraterepresentationofthemastersurfacethroughouttheanalysis.Advantage:SimplifiesthenumericalmethodPotentialforreducedcostperiterationandfindingaconvergedsolutioninfeweriterationsDisadvantage:Theresultscanbenon-physicaliftherelativetangentialmotiondoesnotremainsmallItistheusersresponsibilitytoensurethattheassumptionisnotviolated,RelativeSlidingBetweenBodies(7/9),CharacteristicsofsmallslidingAbaqushasafairlycomplexalgorithmtodeterminetheslideplane.Theslideplaneischaracterizedby:LocationofanchorpointEachslideplanetranslateswithitsanchorpointSlideplanenormaldirectionsIngeometricallynonlinearanalyses,thesewillrotateWhichmasternodesparticipateintheconstraintsThesenodesparticipateinloadtransferTheN-to-SandS-to-Scontactformulationsdifferinhowtheabovearedetermined,RelativeSlidingBetweenBodies(8/9),Invokingsmall-sliding(contactpairsonly):,*CONTACTPAIR,SMALLSLIDING,RelativeSlidingBetweenBodies(9/9),Exampleofnon-physicalbehaviorwithasmall-slidingformulationTherectangularbodyactsastheslaveTheslideplanesarerepresentedbywhitelinesintheanimationTheslideplanestranslatewiththepunchasitmovestotherightThehighlightedslavenodesestablishcontactwiththeirrespectiveslideplanes,beforeanydownwardmotionofthepunchhasoccurred(i.e.,whileaphysicalclearanceexistsbetweentheparts)!Lesson:Usethefinite-slidingformulationifyoudontwanttoworryaboutwhethersmall-slidingassumptionsareappropriate!,OutputofContactResults(1/13),PreliminariesOutputfilesOutputdatabase(.odb)fileThisfileisusedforpostprocessingwithAbaqus/CAE.Bydefault,ODBoutputincludespreselectedvariablesData(.dat)filePrintedoutput;nooutputbydefaultResults(.fil)fileUsedforpostprocessingwiththird-partypostprocessors;nooutputbydefaultOutputvariabletypesNodalvariablesWholesurfacevariables,OutputofContactResults(2/13),NodaloutputtotheODBfileThedefaultnodalcontactoutputtotheODBfileincludesthefollowingvariables:Contactstresses:ContactpressureCPRESSFrictionalshearstressesCSHEAR1andCSHEAR2ContactdisplacementsContactopenings:COPENAccumulatedrelativetangentialmotions:CSLIP1,CSLIP2CSHEAR2andCSLIP2areprintedonlyinthree-dimensionalproblems.Aboveoutputavailableasbothfieldandhistorydata.,OutputofContactResults(3/13),AdditionalfieldoutputincludesContactnodalforcevectors(CFORCE)Contacterrorindicators(CSTRESSERI)Nodalareasassociatedwithactivecontactconstraints(CNAREA)Contactstatus(CSTATUS)Enablescontourplotsofsticking/slipping/openstatus.,OutputofContactResults(4/13),Contactstress(andalsocontactstatus)outputisavailableonbothmasterandslavesurfaces.Exceptions:Cannotviewcontactstressesonsurfacesbasedonrigidelementstypes(whenusedaspartofacontactpair)oranalyticalrigidsurfaces,Rotatedinordertoseecontactpressureonbothsidesofcontactinterface,Uniaxialcompressionloading,Resultsobtainedwithsurface-to-surfacecontactformulation,OutputofContactResults(5/13),ErrorindicatorsProvideanindicationofcontactstressaccuracyNotconservativeorpreciseestimateoferrorNotnormalized:sameunitsasbasevariableUsetoguidemeshrefinementIfforagivenlocationinthemodeltheerrorindicatorisonthesameorderofmagnitudeasthemaximumcontactstress,thenmayneedtorefinethemeshinthatarea.Meshrefinementnotautomatic;usermustremeshaccordingly,OutputofContactResults(6/13),Errorindicators(contd),Errorindicator,Errorindicator,AccuratepredictionofmaximumCPRESSSomeuncertaintywheregradientislargebutpressureislow,Needfinermeshtopredictmaximumcontactpressure,OutputofContactResults(7/13),ContactareaSmallsliding:Thecontactareaisalwaysbasedonthereferenceconfiguration(regardlessofwhetherornotgeometricallynonlineareffectsareconsidered).Finitesliding:Thecontactareaisalwaysbasedonthecurrentconfiguration(regardlessofwhetherornotgeometricallynonlineareffectsareconsidered).UnitsUnitsofcontactstressesdependontheelementsunderneaththesurface.Mostelementsandnode-basedsurfaces:forceperactualunitarea(stress).Beams(two-dimensionalorthree-dimensional):forceperunitlength.,OutputofContactResults(8/13),Contactopening(COPEN)AbaqusoftendoesnotprovideCOPENvaluesforregionswithsignificantgapsYoucanextendtherangeforwhichAbaqus/StandardprovidesCOPENoutputAvailableforgapdistancesuptoauser-specified“trackingthickness”via:*SurfaceInteraction,TrackingThickness=valueWarning:candegradeperformance,Sphere-on-plateexample,OutputofContactResults(9/13),SelfcontactresultsOutputforCPRESSandCSHEARatanoderepresentanaveragewhileitactsasaslavenodeplusanaveragewhileitactsasamasternode.Contourplotsreflectthetotalinterfacepressure.,Displacementmagnificatio