|
Abstract
|
The objectiveofthisstudyistoproposeanovelapproachforpredictingthestructuralparametersof
integrallyskinnedporousmembranesusinggaspermeationdata.Itisintendedtoovercomethe
limitation oftheconventionalgaspermeationtesting(GPT)method,sincethelattermethodseemsto
suffer fromseveralconceptualdrawbacks.Inparticular,acomparisonismadebetweenthetheoretical
calculation andtheexperimentaldatatoshowthesuperiorityofthenewlyproposedmodel.Thenew
model isamodification ofWakaoetal.0s model,inwhich,unliketheconventionalGPTmethod,the
contribution oftheslip flow isconsidered.AlthoughWakaoetal.0s modelwasfoundsuperiortothe
conventionalGPTmethod,themodel fitting totheexperimentaldatawasnotcompletelysatisfactory.It
waslikelythattheslip flow,althoughitseffectcannotbeneglected,isnotfullydeveloped.Therefore,a
factor ? is introducedtoshowtheextentofthecontributionoftheslip flowmechanismtothetotalgas
permeation rate.Asaresult,thenewmethodcanovercometheshortcomingsoftheconventionalGPT
method bymanifestingthefollowingadvantages:(i)itcancovertheentirerangeof J versus P diagram,
(ii) itcanspecifythecontributionoftheindividualmechanismsinvolvedinthetotalgaspermeationand
(iii) unliketheconventionalGPTmethod,itisnotlimitedbyanyconstraintsorconditionsofdata
acquisition.Insummary,themodelcanpredictporesizeandeffectiveporosity,andalsosimulatethe
experimental J versus P trends withsufficient accuracy(within 2% overthepressurerangestudied)for
all typesofmembranes,i.e.NF,UF,MF,MDandmembranecontactors.Inviewofthisfact,theproposed
model issimplertoapplythanRangarajanetal.0s modelandmoreaccuratethantheconventionalGPT
method.
|