物流中心的消防研究(英文版)(1)

DepartmentofFireSafetyEngineeringLundInstituteofTechnologyLundUniversityReport1013ExperimentalandTheoreticalStudyofRackStorageFiresHaukurIngasonLund19963ISSN1102-8246ISRNLUTVDG/TVBB--1013--SEKeywords:Rackstoragefire,firegrowthrate,in-rackfireplume,flameheight,gastemperature,gasvelocity,flameheattransfer,in-racksprinkler,responsetimeAbstract:Atheoreticalandexperimentalstudyofrackstoragefiresandresponsivenessofsprinklersispresented.Free-burntestswithnon-combustibleandcombustiblematerialwerecarriedoutinreducedscalewithverificationinlargescale.Formulasforin-rackflameheight,excessgastemperature,gasvelocityandheatfluxtostoragewallsareprovided.Theformulasincludeoverallheatreleaserate,verticalfluewidth,heightabovethefloor,heightofvirtualoriginandsootinessoffuel.Theycanbeusedtopredictactivationtimesofin-racksprinklersanditispossibletodirectlyincorporatethemintoengineeringmodelsdesignedtopredictfiregrowthinstoragegeometries.Thestoragearrangementisimportantfortheinitialflamespreadandfiregrowthrate.Thereducedscalestudyshowsthattheinitialfiregrowthratedecreaseswithincreasingverticalgaps(flues)andthattheverticalandlateralflamespreadrateincreasewhenthelateralflueheightincreases.Thefiregrowthrateofrackstoragesisusuallydescribedbyapowerlawdependenceontimetothethirdpower.Thelargescaletestshows,however,thattheinitialfiregrowthrateisbetterdescribedbyanexponentialfunction.Thepresentworkprovidesmeasurementsoftheheatfluxdistributionatthesurfaceoffoursquaresteeltowersrepresentinganidealisationofarackstorageatreducedscale.Threegaseousfuels,carbonmonoxide(CO),propane(C3H8),andpropylene(C3H6)weresuppliedfromacirculargasburneratthefloor.Thefuelswerechosentocoverawiderangeofsootingtendenciesleadingtodistinctlydifferentflameheatfluxes.Thedifferencesaresurprisinglylarge.ForthesameoverallfireheatreleaseratethepeakheatfluxfromC3H8flamesistwicethatfromCOflames,whereasthepeakheatfluxfromC3H6is2.8timesgreaterthanfromCOflames.Theheatfluxesweremeasuredbythermocouplesspot-weldedontothebacksideoftheexposedsteeltowersheets.Themeasuringtechniquewasfoundtobesimple,accurateandruggedinadditiontobeinginexpensive.©CopyrightInstitutionenförbrandteknikLundstekniskahögskola,Lundsuniversitet,Lund1996FrontPage:MariaAndersenFrontpagephoto:Largescalefiretestswithsprinklerscarriedoutin1989attheSwedishNationalTestingandResearchInstitute(SP)inBorås.ThetestswerecommissionedbythefurnituremanufacturerIKEAAB.Layout:HaukurIngasonFigures/Diagrams:HaukurIngasonDepartmentofFireSafetyEngineering⋅TheLundInstituteofTechnology⋅LundUniversityAdress/AddressTelefon/TelephoneTelefaxE-post/E-mailBox118/JohnEricssonsväg1046-222[e-mail]4NomenclatureA1definedinequation(16).a,b,cdefinedbyequation(18)and(19)(m/kW2/5)thermalplumewidth(m)bwnon-dimensionalwidthcpspecificheat(kJ/kgK)Cconductionparameter(m1/2/s1/2)Cb,CT,CunondimensionalconstantsCHPChangeofPhaseParameter()Deffectivediameter(m)Ffunctiondefinedinreference52FrFroudenumbergaccelerationofgravity(m/s2)GrGrashofnumberHheightofrackstorage(m)∆Hcheatofcombustionperunitmassofthefuel(J/kg)Lfmeanflameheight(m)mairmassflowrate(kg/s)metotalmass-entrainmentrateofairbelowthemeanflametip(kg/s)mffuelmassflowrate(kg/s)Qchemicalheatreleaserate(kW)Qcconvectiveheatreleaserate(kW)Qc,0centrelineconvectiveheatreleaserate(kW)Qc,expconvectiveheatreleaseratebasedonmeasurements(kW)ReReynoldsnumberRTIResponseTimeIndex(m1/2s1/2)rstoichiometricmassair-to-fuelratio5ttime(s)t0timeatt=0(s)toptimeofsprinkleractivation(s)Taambientconditionsofsprinkler(K)Tggastemperatureinwindtunnel(K)airambienttemperature(K)T0centrelinegastemperature(K)∆Ttemperatureriseaboveambient(K)∆Teoptemperaturedifferenceatsprinklerresponse(K)∆Tgtemperaturedifference(K)∆T0centrelineexcessgastemperature(K)∆Tlfcentrelineexcesstemperatureatmeanflameheight(K)uvelocity(m/s)u0centrelinevelocity(m/s)wwidthofaverticalflue(m)zheightfromfloorlevel(m)z0heightofvirtualorigin(m)Greeksymbolsentrainmentcoefficientair-to-fuelstoichiometricfraction6densityatambienttemperature(kg/m3)timeconstant(s)convectivefractionofchemicalheatreleaserate,Qc/Qnondimensionalparameterdefinedbyequation(22)7TableofcontentAbstract3Nomenclature4Tableofcontent6Listofpapers71Overview81.1Sprinklerresponse81.2In-rackfireplumes82Introduction102.1Aim102.2Relatedwork113Rackstoragefireprotection144Responsivenessofsprinklers194.1Responsemodels194.1.1Heatlossestosprinklerfittings204.1.2Timedelayinglassbulbsprinkles204.1.3Heatreleaserateatsprinklerresponse224.2Responsivenessofin-racksprinklers225In-rackfireplumes245.1Twodimensional(2D)tests,non-combustiblematerial245.2Threedimensional(3D)tests,combustiblematerial285.2.1Modelling305.2.2In-rackflameheight325.2.3In-racktemperatureandvelocity375.3Threedimensional(3D)tests,non-combustiblematerial456SummaryandConcludingRemarks487FutureWork518Acknowledgements529References538ListofpapersThisthesisisbasedonthefollowingpapers,publishedorsubmittedforpublicationininternationalsymposiumsorfiresafetyjournals.PaperIInvestigationofTherma