预览加载中，请您耐心等待几秒...

格子Boltzmann方法在GPU平台下对多孔介质流动的模拟.docx

预览
1/3
2/3
3/3

在线预览结束，喜欢就下载吧，查找使用更方便

下载提示 文本预览

如果您无法下载资料，请参考说明：

1、部分资料下载需要金币，请确保您的账户上有足够的金币

2、已购买过的文档，再次下载不重复扣费

3、资料包下载后请先用软件解压，在使用对应软件打开

格子Boltzmann方法在GPU平台下对多孔介质流动的模拟 Grid-BasedBoltzmannMethodforSimulationofPorousMediaFlowonGPUPlatform Introduction Porousmediaflowisubiquitousinnatureandengineering,rangingfromoilproductiontogroundwaterremediation,frombloodflowtocarbonsequestration.Accuratelysimulatingfluidflowinporousmediaisofgreatimportanceinenhancingourunderstandingofsubsurfaceprocessesandoptimizingengineeringdesigns.Recently,thelatticeBoltzmannmethod(LBM)hasemergedasapromisingalternativetotraditionalnumericaltechniquesforporousmediaflowsimulationduetoitssimplicity,flexibility,andefficiency.Inparticular,thegrid-basedBoltzmannmethod(GBBM)hasbeendevelopedtoovercomethechallengeofthecomplexgeometryandheterogeneousmediuminporousmediaflow.Inthispaper,wefocusontheimplementationandoptimizationofGBBMonthegraphicsprocessingunit(GPU)platform,aimingatachievinghighperformanceandscalabilityforlarge-scalesimulations. OverviewofGBBM GBBMisagrid-basedversionofthestandardLBM,wherethefluiddomainisdiscretizedintoaregulargridofcells,andthedistributionfunctionsofthefluidparticlesareupdatedaccordingtoaseriesofcollisionandstreamingoperations.ThemainadvantageofGBBMliesinitsabilitytohandlearbitrarygeometriesandheterogeneousmediabyintroducingasetofeffectivepermeabilitytensorsthatcapturethemicroscopicpropertiesoftheporousmedium.Thepermeabilitytensoriscalculatedbasedonthelocalporosityandtortuosityofeachcell,whichcanbeprecomputedandstoredinalookuptable.ThevelocityandpressurefieldscanthenbeobtainedbysolvingthePoissonequationandcomputingthemomentsofthedistributionfunctions. GPUImplementation GPUhasbecomeanincreasinglypopularplatformforscientificcomputingduetoitshighparallelismandmemorybandwidth.However,developingefficientGPUcodesforlatticeBoltzmannalgorithmsischallengingduetothecomplexdatastructureandirregularaccesspattern.Inthiswork,weadopttheCUDAprogrammingmodeltoimplementGBBMonGPUs.Themainstepsareasfollows: 1.Dataorganization:Thefluiddomainispartitionedintoblocksofcells,eachofwhichcanbeprocessedindependentlybyathreadblockonthe