有关DPM的问题.docx

有关DPM的问题.docx

《有关DPM的问题.docx》由会员分享，可在线阅读，更多相关《有关DPM的问题.docx（34页珍藏版）》请在冰点文库上搜索。

有关DPM的问题

1.DPM

I'dliketoset3particleinjections（e.g,30particlesforeachinjections）fordifferentdiameterof20um,1um,0.01umandcomparethedepositionvelocity.iamusingtheunsteadytracking,theproblemiswhenistartingtheiteration,Fluentmixthe3injectionstogether,thatis,itwillshowontheconsolewindows:

numbertracked=90,trapped=15,escaped75,sthlikethis,soican'tgetthetrappednumberofparticlesforeachdiameter.Canyouhelpme?

thanks

Hi

Iwasthinkinginyourcase!

Whydoyouwanttorunthethreediametersinonlyoneiteration?

Ithinkyoucanruneachdiameterseparatelyinordertomeasurethevariablethatyourequest,afteralltheparticletrackingisindividualforeachparticleanditisnotaffectedbyotherparticles!

InadditionIthinkthat3pointofinjectionaretolow!

Perhapsmorepointperareacouldbemorerepresentativeforyourcasethan30particleslaunchedfromonepoint!

GiveyourcommentbackIaminterestedinyourcase!

Alex

thanksforyourresponding!

ImeanifIcansubmit3injections（justtheparticlediameterisdifferent,theotherparametersarethesame）inonetimestepsandgettheresultsseperately.butnowicanonlyputoneinjectionatonetime.forexample,ineedtheresultsduring20timesteps,onetimestepcosts1hour,thenforoneinjectionineed20hours,ifiputtheinjectiononebyone,thenihavetospendn*20hours!

soiwonderificanputalltheinjectionstogether,andthenthetimeneedisonly20hours.

hopeiexplainitclealy,thanks!

HICandy

YesIcanhelpyou!

Solution

MadeThreefileswiththefollowingformatandinformationdata

（（x,y,z,v,v,w,d,T,mass-flow）injectionID:

particleID）wherex,y,zarethecordinatesu,v,w,aretheabsolutevelocityoftheparticlesddiametermTtemperatureKelvinMass-flowKg/sInjectionID:

particleID=Injectioni:

jwhereithenumberofpointforinjection（inyourcase0to2）andjthenumberofparticleperinjectionpoint（inyourcase0to29）

Thenreadthatfilesinyourinjectionwindow（belowsingle,group,surfaceyouwillseefile）Thenatthebottomoftheinjectionwindowyouwillseaboxwiththenamefile..Clickonthefile..boxandselectthefile（namethefilecandy0.inj,candy1.injandcandy2.inj）Then,runtheparticletrackingforyousteadycaseMakesurethatyouaregettingreasonableresultsThen,writeaUDFdpmOutlettowriteafilewiththevariablethatyouwishThen,openreport>discretephase>sample>selectyousurface,injectionandoutputfilecompute,plotthehistogramreport>discretephase>histogrammakesurethatyourstatisticalmomentsforthevariablesofconcerndoesnotchangeafterincresingthenumberofparticlesThen,runtheunsteadycasewithallyouinjectionsImporttheoutput.dpmfiletoExcelThen,youcanseethevariableofconcernforeachinjectionID:

particleIDFinallyyoucanplaywithyourresultsasyouwant!

Havefun!

Ifimisssomethingthinksmart!

Regards

Alex

Hicandy

Now,Ineedyouhelp

Ihavenotrununsteadyparticletracking!

couldyoupleaseexplainmehowyoudoit

Thankyou

Alex

1.OpenDISCRETEPAHSEMODELpanel,selectUnsteadytracking.

2.GotoInjections-setinjectionproperty-creat,ifyouselectasurfaceasinjection,thenonthebelowofthepanel,therewillexist"pointproperties"forunsteadytracking,itwillappeartheitem“starttime"and"stoptime",particlewillbereleasedateverytimestepduringtheperiodoftimeyoudefined.Aftertheperiodoftimeyouset,particlewillstillbetrackeduntilalltheparticlearetrappedorescapedor....3.notethatunsteadytrackingiscalculatedtogetherwiththefluiditeration,thatis,open"iterate"panel,and"iterating",aftereachtimestepiteration,theparticlewillbereleasedandcaculated.

hopetheaboveisuseful,goodluck!

unsteadyparticletrackingisonlyactivewhenyouselecttheunsteadyoptioninthesolver.

helloCandyandHedonist

Thankyouforyourinformation!

Regards

AlexMUnoz

Hicandy

Iforgettoaskyouonedetail.

isyourDPMcouplewiththecontinuousfluid?

becuasemycaseisnotcouplewiththecontinuousflow,andIwonderifIneedtoiteteratetheflow.

Thankyouforanswermethesequestions!

Bestregards

AlexMunoz

Mycaseisalsonotcoupletoo,butitisalsoneededtoiteratetheflow,otherwisehowcanyoudotheunsteadytrackingsincethecaseanddatafileonlyrecordtheflowinadefinitetime?

HiCandy

Thanksalotforyourreply,Ithoughtaboutit!

JustIwanttobesure

Thankyou

AlexMunoz

Hicandy

Iwanttoaskyouanotherquestion!

Perhaps,youwonderWhyIdon'trunFluentandIcheckformyself,Currently,wearemakingmodificationandIcannotrunaniterationformorethan5minutesusingtheGUI,ifIruniterationformorethan5minutesIhavetodoitunderbatchmode.Idon'tknowifyouhaveworkingunderthismodebutIhavetobe100%surewhatIamdoing.

Thequestionisthefollowing

IhavetoreportparticletrackingunderReport>DPM>SampleandIhavetospeficytheStarttimeandStoptime.Fluentdocumentationsaysthe

"Forunsteadyparticletracking,clickingontheStartbuttonwillopenthefilesandwritethefileheadersections.Ifthesolutionisadvancedintimebycomputingsometimesteps,theparticletrajectorieswillbeupdatedandtheparticlestateswillbewrittentothefilesastheycrosstheselectedplanesorboundaries.ClickingontheStopbuttonwillclosethefilesandendthesampling"

Iamveryconfusewiththisinformationperhapsyoucanhelptounderstandthisissue.

Thankyouinadvance!

AlexMunoz

2.ProblemofDPM

Deareveryone:

IamusingDPMmethodtosimulatedropsdistributioninacompartment.Theresidualsarealwaysoscillatoryanddon'tconverge.Couldyoupleasetellmehowtomakemysolutionconvergence?

Thankyouverymuchforyourhelp.

Iamlookingforwardtoyourreply.

Withbestregards,

You

Ihaveexactlythesameproblemifitisanyconsolation!

Hiyou,

canyougivemoredetailonyourcase:

coupledorsegregatedsolver?

twowaycoupleddpmcalculationornot?

withstochasticparticletrack?

doesthecalculationforthecontinuousphaseconvergewithoutparticle,...

bestregards

Alain

HiAlain,

Thankyouverymuchforyourreply.Iamusingsegregatedsolverandcoupleddpmcalculationwithstochasticparticletrack.Thecalculationforthecontinuousphaseconvergeswithoutparticle.Thedropdiameteris1.2mm.

Doyouprovideenoughinformation?

Bestregards,You

Canithavesomethingtodowithhowoftenyouinjectparticles?

Christian

Dearyou

1）DropdiameterisfairlybigcomparedtothegeneralDPMapplication.So,Ithink,it'smomentumsourceisfairlylarge.Becauseyouusestochasticparticletracking,smallchangeinmomentumsourceinlocalarea（localcell）cancausefairlybigeffect.So,howabouttotrymeanparticletracking（withoutstochasticoption）?

2）WhenItryDPMandconvergenceisdifficult,Iincreasegasphaseiteration,thatis,50~100iterationpereachDPMiteration.Atthattime,Isetresidualconvergencecriteriaverylowvalue,e.g.,1.e-05.Then,iterationiscontinued.Aftermanyiterations（atleast50~100DPMinjectioniteration）,Icheckeveryresidualwhethertheyareenoughlowornot.Withthisapproach,ingeneral（noteverycase）,Iwassuccessful.

Hi,

Thankyouverymuchforyourreply.Ihavetriedinjectingparticlesevery5to20iterations.

Withbestregards,You

Ihopethatitmightbehelpfultoyou.

Sincerely,Jinwook

Dearyou

Iagreewithjimwook.

Yourlackofconvergenceiscertainlydotostochastictracking.

Notethatthiscanalsobetheresultofaphysicalphenomenon!

Stochastictrackingintroduceimplicitlyatemporalterminthemomentumequationduetofluctuationofthetermofexchange.

Thiscanbeobservedexperimentalyindensetwophaseflowsuchaspneumatictransportorfluidizedbed.

Takealookatyourmomemtumsourcetermandyourflowpressureandvelocityfield.Yoursourcetermcanfluctuatebutyourfieldcanbesteady.Inthispeculiarcaseyourflowis"converged"evenifyourresidualsremainshigh.

Atlast,ifyourflowdiverge,youcanalsounderrelaxthesourcetermcomingfromDPM.

Bestregards

Alain

Addtopreviouscomment:

Increasinginjectionnumbermightbehelpfulforconvergencebecausemomentumsourceofeachinjectionisrelativelysmallwithmanyinjectionnumber.

Sincerely,Jinwook

3.Usingfiletodefineinjectiondistribution!

!

（DPM）

Hello,

InfluentDPM,particleinitialconditionscanbereadfromanexternalfile.Thefilehasfollowingform:

（（xyzuvwdiametertemperaturemass-flow）name）

Butwhat'sthefile'sformat?

?

Thankyouforyourhelp!

!

winnie

Hiwiinie

Thefilehastheformatthatyouwrite（（xyzuvwdiametertemperaturemass-flow）injection0:

ID）（（###########）injection0:

0）（（###########）injection0:

1）（（###########）injection0:

2）.........（（###########）injection0:

n）

where#isanumberundertheunitthatyouareusing.

Regards

AlexMunoz

Hi,AlexMunoz

Thankyouforyouranswer.Iwouldalsoliketoknowthefile'stype,anotherword,whatkindoffilescanbeaccessedbyfluent?

Ihavetried.txt,.candsoon,buttheycan'tbeinputtothefluent.

Waitingforyouranswer!

Bytheway,ifIhavefourinjections,canIusefoursingleinjections?

Doesithaveanydifferencewithafileinjection?

Regards

winnie

Hiwinnie

Thefileisatextfile,ForinstanceIgenerateafileusingaCcodethatIwrote,thenInameit"bach.inj"

YOucancreatefourfilesorIguessyoucanwritejustonefilecontaining4injecti