简介：

简介：DistributedLagrangeMultiplier/FictitiousDomain(DLM/FD)methodwasemployedtosimulatethechannelflowwithpolymersuspension.ThepolymermoleculesweremodeledasFinitelyExtensibleNonlinearElastic(FENE)chainsandFENEchaincluster.Thecoiled-stretchedtransitionofFENEchainswasexaminedandthechangeinconfigurationofFENEchainswaspresented.Theaveragevelocityprofileofthefullydevelopedchannelflowwith64FENEchainswasgivenandfittedwellwiththepower-lawcurve.Thechangeofchainclusterconfigurationwasalsosimulated.ThesesimulationsshowthatDLM/FDmethodiscapableofsimulatingthemotionofnotonlyFENEchain,butalsoFENEchaincluster.

简介：ＮＵＭＥＲＩＣＡＬ　ＣＯＭＰＵＴＡＴＩＯＮ　ＯＦ　ＴＨＥ　ＤＩＦＦＵＳＩＯＮ　ＯＦ　ＴＨＥ　ＣＯＮＴＡＭＩＮＡＮＴＳ　ＡＮＤ　ＴＨＥ　ＦＬＯＷ　ＦＩＥＬＤ　ＯＦ　ＳＫＥＷ　ＤＩＳＣＨＡＲＧＥＳ　ＩＮＴＯ　Ａ　ＣＨＡＮＮＥＬＮＵＭＥＲＩＣＡＬＣＯＭＰＵ...

简介：Ageneralizedbendflowmodel,treatinga90°singlebendand60°continuousbends,wasdesignedtoquantitativelydescribe3-Dturbulencemechanismofcirculatingnot-fully-developedflowinopenchannelswithbends.The3-Dfluctuatingvelocitiesofturbulentflowweremeasuredandanalyzedwitha3-Dacoustic-Dopplervelocimeter.Formulafor3-Dturbulentintensitywasderivedusingthedimensionanalysisapproach.Expressionsofverticalturbulent-intensitydistributionswereobtainedwiththemultivariant-regressiontheory,whichagreewithexperimentdata.Distributionsofturbulentintensityandturbulentstresswerecharacterized,andtheirrelationshipswereconcluded.Inthebend-turbulent-flowcoreregion,longitudinalandlateralturbulent-intensitydistributionsarecoincidentwithlineardistribution,butinnear-wallregionarecoincidentwiththeGammadistribution.VerticalturbulentintensitydistributionsarecoincidentwiththeRayleighdistribution.Herein,itisconcludedthatthebendturbulenceisanisotropic.

简介：Channelfrictionisanimportantparameterinhydraulicanalysis.AchannelfrictionparameterinversionmethodbasedonKalmanFilterwithunknownparametervectorisproposed.Numericalsimulationsindicatethatwhenthenumberofmonitoringstationsexceedsacriticalvalue,thesolutionishardlyaffected.Inaddition,KalmanFilterwithunknownparametervectoriseffectiveonlyatunsteadystate.Forthenonlinearequations,computationsofsensitivitymatricesaretime-costly.Twosimplifiedmeasurescanreducecomputingtime,butnotinfluencetheresults.Oneistoreducesensitivitymatrixanalysistime,theotheristosubstituteforsensitivitymatrix.