Experimental Study of Turbulent Flow Heat Transfer

A.Muley1Assoc.Mem.ASMER.M.Manglik2Mem.ASMEDepartmentofMechanical,IndustrialandNuclearEngineering,UniversityofCincinnati,Cincinnati,OH45221-0072ExperimentalStudyofTurbulentFlowHeatTransferandPressureDropinaPlateHeatExchangerWithChevronPlatesExperimentalheattransferandisothermalpressuredropdataforsingle-phasewaterflowsinaplateheatexchanger(PHE)withchevronplatesarepresented.Inasingle-passU-typecounterflowPHE,threedifferentchevronplatearrangementsareconsidered:twosymmetricplatearrangementswithj3=30deg/30degand60deg/60deg,andonemixed-platearrangementwithft=30deg/60deg.Forwater(2Pr6)flowratesinthe600Re104regime,dataforNuandfarepresented.Theresultsshowsignificanteffectsofboththechevronangle/3andsurfaceareaenlargementfactor/.As/3increases,andcomparedtoaflat-platepack,uptotwotofivetimeshigherNuareobtained;theconcomitantf,however,are13to44timeshigher.Increasingfalsohasasimilar,thoughsmallereffect.BasedonexperimentaldataforRea7000and30degf3=s60deg,predictivecorrelationsoftheformNu=C,(/3)Dj(t)Rem)Prln'(pjJaMandf=C2(/3)D2(j)Re2'8'aredevised.Finally,atconstantpumpingpower,anddependinguponRe,(3,and4,theheattransferisfoundtobeenhancedbyupto2.8timesthatinanequivalentflat-platechannel.IntroductionTheplateheatexchanger(PHE)iscommonlyusedforpro­cessheatingandcoolingapplicationsinchemical,foodpro­cessing,andpharmaceuticalindustries,amongmanyothers.Ithasfoundincreasingusageinsituationswhereclosethermalqualitycontrolandmitigationofthermaldegradationofendproductsareprimaryobjectives.ThereviewsbyShahandFocke(1988),ManglikandMuley(1993),andManglik(1996)haveaddressedsomeaspectsofthethermal-hydraulicperformance,productdevelopment,anddesignapplicationsofPHEs.Ofthemanydifferenttypesofplatecorrugationsavailable(ShahandFocke,1988),themorecommonlyusedchevronplatepatternisillustratedinFig.1.Plateswith(3=30degor60degareusuallystackedtogetherineitherasymmetricormixedarrangementasshowninFig.1(a).Theplatesurfacegeometryischaracterizedbythecorrugationinclinationangle(3,itswavelength\,amplitudeb,andprofile(Fig.1(b)),andthesurfaceenlargementfactor/(ratioofeffectivecorrugatedsurfaceareatotheprojectedareaoftheplate).Theenhancedheattransferisdirectlyrelatedtothesefeatures,whichprovideincreasedeffectiveheattransferarea,disruptionandreattach­mentofboundarylayers,swirlorvortexflowgeneration,andsmallhydraulicdiameterflowchannels.Generallybothheattransfercoefficientsandflowfrictionlossesincreasewithhigher0andjchevronplates(Manglik,1996).Severalinvestigatorshavereportedthermal-hydrauliccharac­teristicsofchevronplatePHEs(Okadaetal.,1972;Marriott,1977;Fockeetal.,1985;Taliketal.,1995a,b;MuleyandManglik,1995,1997,1998;Thononetal„1995).Therecent1Currentaddress:AlliedSignal,Inc.,AerospaceEquipmentSystems,Torrance,CA.2Authortowhomallcorrespondenceshouldbeaddressed,e-mail:[e-mail]ransferDivisionforpublicationintheJOURNALOPHEATTRANSFERandpresentedatthe'97ASMENHTC,Baltimore.ManuscriptreceivedbytheHeatTransferDivisionSept.12,1997;revisionreceivedOct.7,1998.Keywords:Augmentation,Enhancement,ForcedConvection,HeatTrans­fer,HeatExchangers.AssociateTechnicalEditor:J.C.Han.surveybyManglik(1996)providesadetailedsummaryoftheavailableNuand/correlationsforchevronplates.Thepredic­tionsfrommostoftheseequationsareobservedtodisagreeconsiderablywitheachotherandpresentaratherwideperfor­manceenvelope(Manglik,1996).Invirtuallyallstudies,sepa­ratepower-lawtypecurve-fitequationsaregivenforeachplatesurfacewithdifferent(3thatcoveraratherlimiteddatasetandrangeofflowconditions.SavostinandTikhonov(1970),Tovazhnyanskietal.(1980)andWanniarachchietal.(1995)haveattemptedtoincorporatej3effectsintosingleequationsforNuand/,respectively.However,theirequationshavedifferentfunctionalforms,withlittleagreementbetweentheirpredictionsfortypicalflowconditionsand(3(Manglik,1996).Theeffectofsurfaceenlargementfactorjhasbeenlargelyignored.ThislackofgeneralizedpredictivetoolsinhibitseffectiveusageofchevronplatesinmanyPHEapplications.Anexperimentalinvestigationoftheeffectsofchevronangle13onthethermal-hydraulicperformanceofaPHEisreportedinthispaper.Single-phaseturbulentflow3ofwaterundercoolingconditionsisconsidered.Twosymmetricandonemixedchev­ronplatearrangements,with0=30deg/30deg,60deg/60deg,and30deg/60deg,respectively,areused.Basedontheexperimentaldataandsomepreviouslyreportedresults,correla­tionsforNuandisothermal/aredevelopedthatincorporatetheinfluenceof0and£alongwiththeRe,Pr,and(p/p,w)dependence.Finally,theheattransferenhancementintheseplatearrangementsisevaluated.ExperimentalApparatusandProcedureTheexperimentalapparatus,shownschematicallyinFig.2,consistsoftwoPHEs(aheatsourceandaheatsink)arrangedinseriesinaclosedloop.Avariablespeedmoynopumpcircu­latestheprocesswaterstream,whichwasfirstheatedinone3Inmuchoftheliterature,ithasbeenestablishedthatchevroncorrugationspromoteearlytransitiontoturbulentflow(400Re800),andthisissuehasbeendiscussedatsomelengthelsewhere(ShahandFocke,1988;Manglik,1996;MuleyandManglik,1997,1998).110/Vol.121,FEBRUARY1999Copyright©1999byA