多相流模型经验谈修订稿

WEIHUAsystemofficeroom【WEIHUA16H-WEIHUAWEIHUA8Q8-WEIHUA1688】多相流模型经验谈多相流模型经验谈多相流的介绍：Currentlytherearetwoapproachesforthenumericalcalculationofmultiphaseflows:theEuler-LagrangeapproachandtheEuler-Eulerapproach.TheEuler-LagrangeApproach:TheLagrangiandiscretephasemodelinFLUENTfollowstheEuler-Lagrangeapproach,thisapproachisinappropriateforthemodelingofliquid-liquidmixtures,fluidizedbeds,oranyapplicationwherethevolumefractionofthesecondphaseisnotnegligible.TheEuler-EulerApproach:InFLUENT,threedifferentEuler-Eulermultiphasemodelsareavailable:thevolumeoffluid(VOF)model,themixturemodel,andtheEulerianmodel.1)TheVOFModel:itisdesignedfortwoormoreimmisciblefluidswherethepositionoftheinterfacebetweenthefluidsisofinterest.ApplicationsoftheVOFmodelincludestratifiedflows,free-surfaceflows,filling,sloshing,themotionoflargebubblesinaliquid,themotionofliquidafteradambreak,thepredictionofjetbreakup(surfacetension),andthesteadyortransienttrackingofanyliquid-gasinterface.2)Mixturemodel:Applicationsofthemixturemodelincludeparticle-ladenflowswithlowloading,bubblyflows,sedimentation,andcycloneseparators.Themixturemodelcanalsobeusedwithoutrelativevelocitiesforthedispersedphasestomodelhomogeneousmultiphaseflow.3)ApplicationsoftheEulerianmultiphasemodelincludebubblecolumns,risers,particlesuspension,andfluidizedbeds.离散相模型（离散相的装载率10～12%）求解参数的设定：OptionsforInteractionwithContinuousPhase：Forsteady-statesimulations,increasingtheNumberofContinuousPhaseIterationsperDPMIterationwillincreasestabilitybutrequiremoreiterationstoconverge.UpdateDPMSourcesEveryFlowIterationisrecommendedwhendoingunsteadysimulations;ateveryDPMIteration,theparticlesourcetermsarerecalculated.LengthScale:controlstheintegrationtimestepsizeusedtointegratetheequationsofmotionforthesmallervaluefortheLengthScaleincreasestheaccuracyofthetrajectoryandheat/masstransfercalculationsforthediscretephase.LengthScalefactor:AlargervaluefortheStepLengthFactordecreasesthediscretephaseintegrationtimestep.颗粒积分方法：numerics叶中trackingscheme选项1）implicitusesanimplicitEulerintegrationofEquationwhichisunconditionallystableforallparticlerelaxationtimes.2）trapezoidalusesasemi-implicittrapezoidalintegration.（梯形积分）3）analyticusesananalyticalintegrationofEquationwheretheforcesareheldconstantduringtheintegration.4）runge-kuttafacilitatesa5thorderRungeKuttaschemederivedbyCashandKarp[47].Youcaneitherchooseasingletrackingscheme,orswitchbetweenhigherorderandlowerordertrackingschemesusinganautomatedselectionbasedontheaccuracytobeachievedandthestabilityrangeofeachscheme.Max.Refinementsisthemaximumnumberofstepsizerefinementsinonesingleintegrationstep.Ifthisnumberisexceededtheintegrationwillbeconductedwiththelastrefinedintegrationstepsize.AutomatedTrackingSchemeSelectionprovidesamechanismtoswitchinanautomatedfashionbetweennumericallystablelowerorderschemesandhigherorderschemes,whicharestableonlyinalimitedrange.Insituationswheretheparticleisfarfromhydrodynamicequilibrium,anaccuratesolutioncanbeachievedveryquicklywithahigherorderscheme,sincetheseschemesneedlesssteprefinementsforacertaintolerance.Whentheparticlereacheshydrodynamicequilibrium,thehigherorderschemesbecomeinefficientsincetheirsteplengthislimitedtoastablerange.Inthiscase,themechanismswitchestoastablelowerorderschemeandfacilitateslargerintegrationsteps.IncludingaCoupledHeat-MassSolutionontheParticles：Thisincreasedaccuracy,however,comesattheexpenseofincreasedcomputationalexpense.非稳态跟踪1）连续相稳态离散相非稳态：yousimplyentertheParticleTimeStepSizeandtheNumberofTimeSteps,thustrackingparticleseverytimeaDPMiterationisconducted.WhenyouincreasetheNumberofTimeSteps,thedropletspenetratethedomainfaster.2）连续离散相都为非稳态：Whensolvingunsteadyequationsforthecontinuousphase,youmustdecidewhetheryouwanttoUseFluidFlowTimeSteptoinjecttheparticles,orwhetheryoupreferaParticleTimeStepSizeindependentofthefluidflowtimestep.Withthelatteroption,youcanusetheDiscretePhaseModelincombinationwithchangesinthetimestepforthecontinuousequations,asitisdonewhenusingadaptiveflowtimestepping.随机轨道模型的参数：numberoftries:AninputofzerotellsFLUENTtocomputetheparticletrajectorybasedonthemeancontinuousphasevelocityfield(Equation,ignoringtheeffectsofturbulenceontheparticletrajectories.Aninputof1orgreatertellsFLUENTtoincludeturbulentvelocityfluctuationsintheparticleforcebalanceasinEquation.Ifyouwantthecharacteristiclifetimeoftheeddytoberandom(Equation,enabletheRandomEddyLifetimeoption.YouwillgenerallynotneedtochangetheTimeScaleConstant(CLinEquationfromitsdefaultvalueof,unlessyouareusingtheReynoldsStressturbulencemodel(RSM),inwhichcaseavalueofisrecommended.液滴颗粒碰撞与破碎碰撞：破碎：有两种模型，TAB模型适合低韦伯数射流雾化以及低速射流进入标态空气中的情况。对韦伯数大于100的情况，波动模型适应性较好。在高速燃料射流雾化中，波动模型应用甚广。对于TAB模型，用户需要在y0文本框中设定y0的值。Thedefaultvalue(y0=0)isrecommended.对于Y波动模型，需要输入B0与B1，youwillgenerallynotneedtomodifythevalueofB0,asthedefaultvalueisacceptablefornearlyallcases.AvalueofisrecommendedforB1.颗粒类型中的燃烧类型燃烧（``combusting''）颗粒是一种固体颗粒，它遵从由方程所确定的受力平衡、由定律1所确定的加热冷却过程、由定律4所确定的挥发份析出过程（节）以及由定律5所确定的异相面反应机制（节）。最后，当颗粒的挥发份完全析出之后，非挥发份的运动、变化由定律6所确定。在SetInjectionPropertiespanel面板中选定WetCombustion选项，用户可以在燃烧颗粒中包含有可蒸发物质。这样，颗粒的可蒸发物质可在挥发份开始析出之前，经历由定律2、3所确定的蒸发与沸腾过程.若定义的是Combusting燃烧类型颗粒，可在DevolatilizingSpecies下拉列表框下选定由挥发份析出定律4确定的气相组分，参与焦炭表面燃烧反应（定律5）的气相组分列于OxidizingSpecies（氧化剂组分）列表中，有表面反应生成的气相组分则列于ProductSpecies（生成物组分）列表中。需要注意的是，对于选定的燃烧颗粒介质，如果燃烧模型为multiple-surface-reaction多表面异相反应模型，那么，由于化学反应计量比在混合介质中已经被确定，所以OxidizingSpecies与ProductSpecies列表将变灰（不可选）。液滴喷射类型平面雾化模型的输入l位置：在X-,Y-,andZ-Position文本框区可以设定射流的沿直角坐标的三向位置（在三维情况下才会有Z-Position出现）l速度：在X-,Y-,andZ-Velocity文本框区可以设定射流初始速度沿直角坐标的三向分量（在三维情况下才会有Z-Velocity出现）l轴的方向（仅适用于三维）：设定确定喷嘴轴线方向的三个分量，在X-Axis,Y-Axis,andZ-Axis区设定。l温度：在Temperature区可设定喷射颗粒流的初始颗粒（绝对）温度。l质量流率：可在FlowRate区设定喷嘴的的颗粒质量流量。l射流持续时间：对于非稳态颗粒跟踪计算（请参阅节），在StartTime和StopTime区设定喷射的开始于结束时间。l蒸气压：设定控制通过喷嘴内部流动的蒸气压（表中的pv），在VaporPressure区设定。l直径：设定喷嘴直径（表中的d），在InjectorInnerDiam.区设定。l喷嘴长度：设定喷嘴的长度（表中的L），在OrificeLength区设定。l内台阶角半径（导角半径）：设定喷嘴内台阶处的导角半径（表中的r），在CornerRadiusofCurv.区设定。l喷嘴参数：设定射流角修正系数（方程中的CA），在ConstantA区设定。{CA=3+L/d,喷射角度的大小强烈依赖于喷嘴的内部流动。因此，对于空穴喷嘴，用户设定的CA值应该比单相流的要小才可以。CA的常见取值范围为～。返流喷嘴的喷射角度更小}l方位角：设定三维情况下的喷嘴方位开始角与结束角，在AzimuthalStartAngleandAzimuthalStopAngle区设定。压力－旋流雾化喷嘴的点属性设定(气体透平工业的人把它称作单相喷嘴（simplexatomizer）。这种喷嘴，然后流体通过一个称作旋流片的喷头被加速后，进入中心旋流室。在旋流室内，旋转的液体被挤压到固壁，在流体中心形成空气柱，然后，液体以不稳定的薄膜状态从喷口喷出，破碎成丝状物及液滴。)l射流角：在SprayHalfAngle区下设定射流喷射半角（方程中的θ）。l压力：在UpstreamPressure区下设定喷嘴上游压力（表中的p1）。l液膜破碎常数：设定确定液膜破碎时形成的线状液膜长度的一个经验常数（方程中的ln（ηb/η0），在SheetConstant设定。{ln（ηb/η0）为3～12的经验常数。这个值必须由用户设定，其缺省值为12withexperimentalsheetbreakuplengthsoverarangeofWebernumbersfrom2to200.}l线状液膜直径：对于短波，确定液膜破碎波长与线状液膜半径之间的线形比例关系的比例常数，在LigamentConstant区设定。{whereCL,ortheligamentconstant,isequaltobydefault.}空气辅助雾化喷嘴的点属性设定(为了加速液膜的破碎，喷嘴经常会添加上辅助空气。液体通过喷座的作用形成液膜,空气则直接冲击液膜以加速液膜的破碎。)l喷嘴外半径：在InjectorOuterDiam.区下设定射流的外部半径。此数值与喷嘴内部半径共同确定了液膜厚度（方程中的t）。l射流角：设定射流离开喷口时的液膜初始轨道（方程中的θ），在SprayHalfAngle区设定。l相对速度：设定液膜与空气之间的最大相对速度，在RelativeVelocity区设定。l液膜破碎常数：设定确定液膜破碎时形成的线状液膜长度的一个经验常数（方程中的ln（ηb/η0）），在SheetConstant区设定。l线状液膜直径：对于短波，确定液膜破碎波长与线状液膜半径之间的线形比例关系的比例常数，在LigamentConstant区设定。{whereCL,ortheligamentconstant,isequaltobydefault.}平板扇叶雾化喷嘴的点属性设定(液体从宽而薄的喷口出来后形成平面液膜，继而破碎成液滴。只有在三维的情况下才可以使用这个模型)l扇叶中心点：设定射流源起始位置的三向坐标值（请参阅图），在X-Center,Y-Center,andZ-Center区设定。l虚点位置：设定喷嘴扇叶的各边的虚拟交叉点（请参阅图），在X-VirtualOrigin,Y-VirtualOrigin,andZ-VirtualOrigin区设定。l垂直方向：设定垂直扇叶的向量的各个分量，在X-FanNormalVector,Y-FanNormalVector,andZ-FanNormalVector区设定。l温度：设定颗粒流的温度，在Temperature区设定。l质量流量：设定喷嘴的质量流量，在FlowRate区设定。l射流持续时间：对于非稳态颗粒跟踪计算（请参阅节），在StartTime和StopTime区设定喷射的开始于结束时间。l射流角：在SprayHalfAngle区下设定射流喷射半角。l喷口宽度：设定喷口垂直方向的宽度，在OrificeWidth区设定。l液膜破碎常数：设定确定液膜破碎时形成的线状液膜长度的一个经验常数（请参阅方程的ln（ηb/η0）），在FlatFanSheetConstant区设定。气泡雾化喷嘴的点属性设定(，液体中混合了过热液体或者类似的介质。当挥发性液体从喷口喷出时，迅速发生相变。相变使流体迅速以很大的分散角破碎成小液滴。此模型也适用于热流体射流。)混合情况参数：设定射流中液－气混合物中已蒸发的液滴质量分数（方程中的x），在MixtureQuality区设定。l饱和温度：设定可挥发成分的饱和温度，在SaturationTemp.区设定。l液滴扩散系数：设定控制液滴在空间扩散性能的扩散系数（方程中的Ceff），在DispersionConstant区设定。l射流角：设定液膜离开喷口时的初始轨道方向角，在MaximumHalfAngle区设定。通用多相流模型mixturemodelVSeulermodel1)Ifthereisawidedistributionofthedispersedphases.,iftheparticlesvaryinsizeandthelargestparticlesdonotseparatefromtheprimaryflowfield),themixturemodelmaybepreferable.,lesscomputationallyexpensive).Ifthedispersedphasesareconcentratedjustinportionsofthedomain,youshouldusetheEulerianmodelinstead.Ifinterphasedraglawsthatareapplicabletoyoursystemareavailable(eitherwithinFLUENTorthroughauser-definedfunction),theEulerianmodelcanusuallyprovidemoreaccurateresultsthanthemixturemodel.Eventhoughyoucanapplythesamedraglawstothemixturemodel,asyoucanforanon-granularEuleriansimulation,iftheinterphasedraglawsareunknownortheirapplicabilitytoyoursystemisquestionable,themixturemodelmaybeabetterchoice.Formostcaseswithsphericalparticles,thentheSchiller-Naumannlawismorethanadequate.Forcaseswithnon-sphericalparticles,thenauser-definedfunctioncanbeused.加快收敛求解策略Youcanincreasethesizeofthetimestepafterperformingafewtimesteps.Forsteadysolutionsitisrecommendedthatyoustartwithasmallunder-relaxationfactorforthevolumefraction,Anotheroptionistostartwithamixturemultiphasecalculation,andthenswitchtotheEulerianmultiphasemodel.VOF模型界面之间的scalar梯度不要太大界面插值：therearefourschemeforinterfaceinterpolation:geometricreconstruction,donor-acceceptor,eulerexplicit,inexplicit,Thegeometricreconstructionschemerepresentstheinterfacebetweenfluidsusingapiecewise-linearapproach.InFLUENTthisschemeisthemostaccurateandisapplicableforgeneralunstructuredmeshes.thedonor-acceptorschemecanbeusedonlywithquadrilateralorhexahedralimplicitschemecanbeusedforbothtime-dependentandsteady-statecalculations.Eulermodel中的附加作用力LiftForces:Inmostcases,theliftforceisinsignificantcomparedtothedragforce,sothereisnoreasontoincludethisextraterm.Iftheliftforceissignificant.,ifthephasesseparatequickly),itmaybeappropriatetoincludethisterm.Thevirtualmasseffectissignificantwhenthesecondaryphasedensityismuchsmallerthantheprimaryphasedensity.,foratransientbubblecolumn).Bydefault,virtualmasseffectisnotincluded.多相湍流模型k-emodelk-emixturemodel（default）itisapplicablewhenphasesseparate,forstratified(ornearlystratified)multiphaseflows,andwhenthedensityratiobetweenphasesiscloseto1.，它应用于相分离，分层（或接近分层）的多相流，和相之间的密度比接近1。Thedispersedturbulencemodelistheappropriatemodelwhentheconcentrationsofthesecondaryphasesaredilute.Inthiscase,interparticlecollisionsarenegligibleandthedominantprocessintherandommotionofthesecondaryphasesistheinfluenceoftheprimary-phaseturbulence.当明显地有一个主连续相和其它的是分散稀释的第二相时，这个模型是适用的。Thedriftvelocityresultsfromturbulentfluctuationsinthevolumecorrectionisnotincluded,bydefault,butyoucanenableitduringtheproblemsetup(define－－－model－－－multiphases-option).k-eTurbulenceModelforEachPhase:Thisturbulencemodelistheappropriatechoicewhentheturbulencetransferamongthephasesplaysadominantrole.RSMmodelMultiphaseturbulencemodelingtypicallyinvolvestwoequationmodelsthatarebasedonsingle-phasemodelsandoftencannotaccuratelycapturetheunderlyingflowaretwooptionsforRsmmodel,mixtureanddispersedturbulencemodel.WetSteamModel通用多相流模型的输入：1)vofmodelnumberofphases:VOFformulation:1)Time-dependentwiththegeometricreconstructioninterpolationscheme:Thisformulationshouldbeusedwheneveryouareinterestedinthetime-accuratetransientbehavioroftheVOFsolution.2)Time-dependentwiththedonor-acceptorinterpolationscheme:Thisformulationshouldbeusedinsteadofthetime-dependentformulationwiththegeometricreconstructionschemeifyourmeshcontainshighlytwistedhexahedralcells.Forsuchcases,thedonor-acceptorschememayprovidemoreaccurateresults.3)Time-dependentwiththeEulerexplicitinterpolationscheme:Sincethedonoracceptorschemeisavailableonlyforquadrilateralandhexahedralmeshes,itcannotbeusedforahybridmeshcontainingtwistedhexahedralcells.Forsuchcases,youshouldusethetime-dependentEulerexplicittheEulerexplicittime-dependentformulationislesscomputationallyexpensivethanthegeometricreconstructionscheme,theinterfacebetweenphaseswillnotbeassharpasthatpredictedwiththegeometricreconstructionscheme.Toreducethisdiffusivity,itisrecommendedthatyouusethesecond-orderdiscretizationschemeforthevolumefractionequations.4)Time-dependentwiththeimplicitinterpolationscheme:Thisformulationcanbeusedifyouarelookingforasteady-statesolutionandyouarenotinterestedintheintermediatetransientflowbehavior,butthefinalsteady-statesolutionisdependentontheinitialflowconditionsand/oryoudonothaveadistinctinflowboundaryforeachphase.5)Steady-statewiththeimplicitinterpolationscheme:Thisformulationcanbeusedifyouarelookingforasteady-statesolution,youarenotinterestedintheintermediatetransientflowbehavior,andthefinalsteady-statesolutionisnotaffectedbytheinitialflowconditionsandthereisadistinctinflowboundaryforeachphase.NotethattheimplicitmodifiedHRICschemecanbeusedasarobustalternativetotheexplicitgeometricreconstructionscheme.IncludingBodyForces:Inmanycases,themotionofthephasesisdue,inpart,togravitationaleffects.Toincludethisbodyforce,turnonGravityintheOperatingConditionspanelandspecifytheGravitationalAcceleration.ForVOFcalculations,youshouldalsoturnontheSpecifiedOperatingDensityoptionintheOperatingConditionspanel,andsettheOperatingDensitytobethedensityofthelightestphase.Ifanyofthephasesiscompressible,settheOperatingDensitytozero.ModelingOpenChannelFlows:FLUENTcanmodeltheeffectsofopenchannelflow.,rivers,dams,andsurfacepiercingstructuresinunboundedstream)usingtheVOFformulationandtheopenchannelboundarycondition.thestepstoopenopenchannelflowsare:1.Turnongravity2.EnablethevolumeoffluidmodelandselectOpenChannelFlow.boundaryconditionssettingforopenchannelflow:1)DeterminingtheFreeSurfaceLevel(ylocal)Wecansimplycalculatethefreesurfacelevelintwosteps:1.Determinetheabsolutevalueofheightfromthefreesurfacetotheorigininthedirectionofgravity.2.Applythecorrectsignbasedonwhetherthefreesurfacelevelisaboveorbelowtheorigin.Iftheliquid'sfreesurfacelevelliesabovetheorigin,thentheFreeSurfaceLevelispositive(seeFigureLikewise,iftheliquid'sfreesurfacelevelliesbelowtheorigin,thentheFreeSurfaceLevelisnegative.2)DeterminingtheBottomLevel(ybottom):Wecansimplycalculatethebottomlevelintwosteps:1.Determinetheabsolutevalueofdepthfromthebottomleveltotheorigininthedirectionofgravity.2.Applythecorrectsignbasedonwhetherthebottomlevelisaboveorbelowtheorigin.3)SpecifyingtheTotalHeightytot=ylocal+V*V/2/g4)DeterminingtheVelocityMagnitude(appearinthepressureinlet)Thisistobespecifiedasthemagnitudeoftheupstreaminletvelocityintheflow5)DeterminingtheSecondaryPhasefortheInletConsideraprobleminvolvingathree-phaseflowconsistingofairastheprimaryphase,andoilandwaterasthesecondaryphases.Consideralsothattherearetwoinletgroups:1.waterandair2.oilandair;Forthefirstinletgroup,youwouldchoosewaterasthesecondaryphase.Forthesecondinletgroup,youwouldchooseoilasthesecondaryphase.liminationsofchannelflow:Limitations:Thefollowinglistsummarizessomeissuesandlimitationsassociatedwiththeopenchannelboundarycondition.1.TheconservationoftheBernoulliintegraldoesnotprovidetheconservationofmassflowrateforthepressureboundary.Inthecaseofacoarsermesh,therecanbeasignificantdifferenceinmassflowratefromtheactualmassflowrate.Forfinermeshes,themassflowratecomesclosertotheactualvalue.So,forproblemshavingconstantmassflowrate,themassflowrateboundaryconditionisabetteroption.Thepressureboundaryshouldbeselectedwhensteadyandnon-oscillatingdragisthemainobjective.2.Specifyingthetopboundaryasthepressureoutletcansometimesleadtoadivergentsolution.Thismaybeduetothecornersingularityatthepressureboundaryintheairregionorduetotheinabilitytospecifylocalflowdirectioncorrectlyiftheairentersthroughthetoplocally.3.Onlytheheavierphaseshouldbeselectedasthesecondaryphase.4.Inthecaseofthree-phaseflows,onlyonesecondaryphaseisallowedtoenterthroughoneinletgroup.Thatmeans,themixedinflowofdifferentsecondaryphasesisnotallowed.DefiningPhasesfortheVOFModel:1)Ingeneral,youcanspecifytheprimaryandsecondaryphaseswhicheverwayyouprefer.Itisagoodidea,especiallyinmorecomplicatedproblems,toconsiderhowyourchoicewillaffecttheeaseofproblemsetup.Forexample,ifyouareplanningtopatchaninitialvolumefractionof1foronephaseinaportionofthedomain,itmaybemoreconvenienttomakethatphaseasecondaryphase.Also,ifoneofthephasesisacompressibleidealgas,itisrecommendedthatyouspecifyitastheprimaryphasetoimprovesolutionstability.IncludingSurfaceTensionandWallAdhesionEffects(SurfacetensioneffectscanbeneglectedifCa》1orWe》1.Forcalculationsinvolvingsurfacetension,itisrecommendedthatyoualsoturnontheImplicitBodyForcetreatmentfortheBodyForceFormulationintheMultiphaseModelpanel.Thistreatmentimprovessolutionconvergencebyaccountingforthepartialequilibriumofthepressuregradientandsurfacetensionforcesinthemomentumequations):youmustspecifysurfacetensioncoefficient.youcanconsideringwalladhesionbyturningonwalladhesioninphasesinteractionpanel.youmustinputthecontactangleinthewallboundarycondition.Thedefaultvalueforallpairsis90degrees,whichisequivalenttonowalladhesioneffects.,theinterfaceisnormaltotheadjacentwall).Acontactangleof45degree,forexample,correspondstowatercreepingupthesideofacontainer,asiscommonwithwaterinaglass.timedependentsolutionIfyouwantFLUENTtosolvethevolumefractionequation(s)ateveryiterationwithinatimestep,turnontheSolveVOFEveryIterationoptionunderVOFParameters.Thischoiceisthelessstableofthetwo,andrequiresmorecomputationaleffortpertimestepthanthedefaultchoice.库兰数的设定：ifthemaximumallowedCourantnumberis(thedefault),thetimestepwillbechosentobeatmostone-fourththeminimumtransittimeforanycellneartheinterface.2)mixturemodel的输入numberofphases:asaboveIncludingBodyForces:asabovewhetherornottocomputetheslipvelocities：bydefault,fluentturnonthecomputetheslipvelocities,ifyouareDefiningaHomogeneousMultiphaseFlow,turnoffthecomputeslipvelocities.DefiningaGranularSecondaryPhase:PackingLimitspecifiesthemaximumvolumefractionforthegranularphase.Formonodispersedspheres,thepackinglimitisabout,whichisthedefaultvalueinFLUENT.Inpolydispersedcases,however,smallerspherescanfillthesmallgapsbetweenlargerspheres,soyoumayneedtoincreasethemaximumpackinglimit.IncludingCavitationEffects:Toenablethecavitationmodel,turnontheCavitationoptionintheMasstabofthePhaseInteractionpanel.youmustsettheVaporizationPressure,theSurfaceTensionCoefficient,andthemassfractionofNonCondensableGas.Whenmultiplespeciesareincludedinoneormoresecondaryphases,ortheheattransferduetophasechangeneedstobetakenintoaccount,themasstransfermechanismmustbedefinedbeforeturningontheCavitationoption.Itmaybenoted,however,thatforcavitationproblems,atleasttwomasstransfermechanismsaredefined:1.Masstransferfromliquidtovapor.2.Masstransferfromvaportoliquid.3)eulermodel输入numberofphases:asaboveIncludingBodyForces:asaboveDefiningaGranularSecondaryPhase:GranularTemperatureAlgebraicformulation(thedefault).Itisobtainedbyneglectingconvectionanddiffusioninthetransportequation,Equation[340].PartialDifferentialEquation.ThisisgivenbyEquationanditisallowedtochoosedifferentoptionsforitproperties.ConstantGranularTemperature.Thisisusefulinverydensesituationswheretherandomfluctuationsaresmall.PackingLimitspecifiesthemaximumvolumefractionforthegranularphase.Formonodispersedspheres,thepackinglimitisabout,whichisthedefaultvalueinFLUENT.Inpolydispersedcases,however,smallerspherescanfillthesmallgapsbetweenlargerspheres,soyoumayneedtoincreasethemaximumpackinglimit.DefiningtheInteractionBetweenPhasesdefinethedragfuction:Selectschiller-naumanntousethefluid-fluiddragfunction.TheSchillerandNaumannmodelisthedefaultmethod,anditisacceptableforgeneraluseinallfluid-fluidmultiphasecalculations.Selectmorsi-alexandertousethefluid-fluiddragfunction.TheMorsiandAlexandermodelisthemostcomplete,adjustingthefunctiondefinitionfrequentlyoveralargerangeofReynoldsnumbers,butcalculationswiththismodelmaybelessstablethanwiththeothermodels.Selectsymmetrictousethefluid-fluiddragfunction.Thesymmetricmodelisrecommendedforflowsinwhichthesecondary(dispersed)phaseinoneregionofthedomainbecomestheprimary(continuous)phaseinanother.Forexample,ifairisinjectedintothebottomofacontainerfilledhalfwaywithwater,theairisthedispersedphaseinthebottomhalfofthecontainer;inthetophalfofthecontainer,theairisthecontinuousphase.Selectwen-yutousethefluid-soliddragfunction.TheWenandYumodelisapplicablefordilutephaseflows,inwhichthetotalsecondaryphasevolumefractionissignificantlylowerthanthatoftheprimaryphase.Selectgidaspowtousethef