Flowregimeidentificationusingchaoticcharacteristicsoftwo-phaseflowOriginalResearchArticle

Chaoticnatureoftwo-phaseflowisinvestigatedforitsdependencyupontheflowregimebyconstructingthepseudophasespacewiththetimesequentialimpedancesignalsofthevoidfraction.Inordertoconstructthepseudophasespace,theautocorrelationfunction(ACF)andtheaveragemutualinformation(AMI)forthedelayedtime,andthefalsenearestneighborhood(FNN)forthedimensionsastheembeddingparameterswereemployedhere.Itwasfoundthatthedelayedtimeandembeddeddimensionarehighlydependentupontheflowregime.Tovisualizethetrajectoryofsignalsinthephasespace,adensitymapisproducedbytheprojectionofthetrajectoryintothetwo-dimensionalplanewhichiscorrespondentwiththetwo-dimensionalprobabilitydistributionfunctions(2D-PDF).Sincethedensitymapof2D-PDFshowedcleardistinctionofflowpatterns,wedevelopedamethodtoidentifyflowregimebyapplyingsimpleclassificationrulestothedensitymap.Theproposedmethodsuccessfullyidentifiestheflowregimesoftheexperimentaldataofimpedancesignalsforthevoidfractionproducedflowregimemapfortheverticalchannelwithdiameterof25.4and50.8mm.