Page 112 - konishipaper
P. 112
一巴0cmフmく L一一一一一~ Highsped
IRcamera
Videorecorder ved巴ocamera
Fig.1 Experimentalaparatus01theparticle-tracklaser-shet(LSPT)system
Videorecorder
(1972).Tomakesureoftheacuracyofthedata，werepeatedthe
experiments severa1 times and found the reproducibility to be
wi由主190percen.tBasedonourexperi皿 ents，weproposeanew
rnechanism:伽 t出efiamespreadintheuniformspreadregionis
governedbytheliquidconveptiQn(Ito，Masuda，andSaito，191). verysrnalcircu1ationinthegasphaseaheadofthefl.ame'sleading However，at 白 at time we did not 白 l ly understand why Akita's edge. They sug gested 出 at the gas-phase circulation，which is shadowgraphresu1tsweresodiferent企omourHIresu1ts. 田註quein血epu1sa也19spreadandapearswiththeliquid-phase
Later，HowardRos'groupatNASALewisResearchCenter circulation，mayplay叩 importantroleinfl.釘nepulsation(Schiler apliedRainbowSchlierenDefl.ectome句 (RSD)tomeasureliquid andSirignano，1992，196).Therearefl.owmeasurementdata convectiongeneratedbythespreadingfiame(Miler姐 dRos， obtainedwithLaser-DoplerVelocirne句r(LDV)bySantoroetal.
192).Theyfoundlit1eliquidconvectionaheadof出efiame's
1eadingedgeintheuniformspreadregion.TheirRSDdatawere
自 国1紅 toAkita'sshadowgraphdataand也eydisagredwithour
E召 results.Overthepastthreyears，bothNASAandourgroup
位iedtounderstandthereasonsforthedisagrement，butwere
unsucesfu1unti1aseriesofp紅 釘 即 位icexperimentswereunder-
taken.Inthoseexperiments，theefectsofsixdiferentp紅 むneters
wereinvestigated:sensitivityofbothRSDandHIsystems，puri句r
of出e1-prop姐 01，re1ativehumidityof也e出I，ambientairtem-
pera加re，typeofignitionmethod，andfinal1ydiferenttraywidths.
Wefoundthat出efirstfivep紅 むnetershadlitleinfl.uenceonthe
disagrement.However，whenwecheckedthee百'ectof甘aywidth
(0，5. 1，姐d2cm)ontheliquidconvection，the1eastsuspected
pむ むneter，asurprisingresu1temerged:The0.5cmwide仕ayhad
alargeliquidconvectionintheuniformspreadregion，whilethe1
crnand2cmwidetrayshada出inlayerofliquidconvection.It measurefl.owprofilesintwo-dimensionalwithaspatialresolution
wasso也in.andsrnalthatonlyaverysensitiverneasurernent teclmiquecoulddetectit.Theresultforthe2-crntraywasvery s也吐 l紅 to NASA's result for the 2-cm tray. A c10ser lo ok at their RSD.resu1talsorevealedasrna1and白血 layerofliquidconvec- tionthatwasalwaysaheadofthe自叩le'sleadingedge.Cornpar- isonofthesefindingsledustoagre出atintheuniformspread regionthereisalwaysasrna1liquidconvectionindicating也at出e rp.ajorheat仕 組sferisby出atmode.Italsobecarnec1earthat出e shadowgraph(Akita，1972)didnotdetecttheliquidconvection becausethesizeoftheliquidconvectionwastosrnalandbeyond thelimitationofthespatialresolutionoftheshadowgraph.Later weapliedtheshadowgraphteclmiqueandproved白isexplana- tlOn.
Thevalidi匂'of出isnewmechanismhasa1readybenprovenfor 出e位aythatis0
414IVo.l121，MAY1999
ontheorderofamilimeterintheprimaryanchoringregionofa
He-NeGasLas巴E
AIgon~ion Laser
Fig.2 AschematicoftheDWHIsystem TransactionsoftheASME
Ni仕ogen
deppolexperiments(Ros，194);insteadthefl.amewasextin- guished，indicatingtheirnportanceofthebuoy姐 cyefectin出e gasphaseonthernechanisrnofpulsation.SchilerandSirignano， usingnumericalcalculations(192)，predictedtheexistenceofa
(1978) showing a gas-phase circulation of ap proximately 1-crn diameterinthepu1satingspreadregion.However，theirLDVresult isratheraqua1itativeindicationofthefor:rp.ationofcirculation.
Toconflrm也eprediction(Schil1er叩 dSirignano，192，196)
andunderstandtheroleofthesrna1gas-phasecirculati ∞
in出e h皿 epulsation，transientvelocityprofi1es出 thegasphasejust abovetheliquidslIfaceandjustaheadofthe丑ame'sleadingedge mustbemeasured.Thesemeasv，rementsarenoteasy，because出e
rateoffl.amespreadvariesfromafewcentimeterstoafewtensof、 centimeterspersecondatthetargetregion，whiletheregionwhere detailedvelociザ profi1esarene ededisap proxirnate1yafewrnm泊 di包 neter.
Previous1ywelearned血atLDVwasnotsuficientlyacurate forourmeasurements(Venkatesh，Ito，Saito，andWichman，196). InsteadweusedLaserShetPartic1eTracking(LSPT)techr吐queto
