Page 149 - konishipaper

κTakahashieta.lIProcedingsoftheCombustionInstitute30(205)271-27 275
circulationismuchsmaleratthecoresponding l七 utanol temperature.
3.Experimenialresults
3.1. Critical length scα le ratio
Figure4showsthecritical1engthscale1'atio， h/L，measuredbyano1'malSchlierenphotog1'aph technique[2，3，14Jasafunctionoftheinitiall'l-
、 propanolthicknes，H.Thec1'iticalengthscalera- tioh/Lisaboutaconstantvalueof5overtheini- tialfuellayerthicknesH= 5-25m m，whilethe 1'atiosha1'plyincreasedwithaninc1'easeintheini- tial fuel laye1' thicknes s ove1' H ニ 0-5m m.This indicatesthatthe1'earetwo1'egimessepa1'atedby H=.5mm:thinliquidpolsfo1'Hく 5l1nand t h i c k l i q u i d p o o l s f o 1' H > 5 m m . T h e f l a m e s p r e a d ove1' l-butanol using a thick liquid pol ( H = 2 5 m l n ) f o 1' m i c 1' o g r a v i t y w e l l c o r r e l a t e s t o athinliquidpol(H= 3nml)conditionfo1'p.or- lnalg1'avity，suporting.thefactthatthemic1'o- gravity 丑ame sp1'ead over thick liquid po ol becomes similar to the normal gravity flame sp1'eadoverthinliquidpol，despitethefueldiι ferencebetwennormalgravity(n-p1'opanol)and micro gravity (ト butanol). This finding，if vali- datedforwiderangeoffueltypes，couldhelp researche1'stosimulatethebehaviorofmicrog1'av- ityflamespreadwithnormalgravityexperiments.
Figure5showsthecha1'acteristiclength:L，and liquidcirculationdepth:h，undertheinitiall-bu- tanoltemperature，17oC，asafunctionofinitial fuelthicknes. Itrequires 5m mfuel depthfor thesub-su1'facecirculationtofulydevelop.When thefueldepthbecomeslesthan5m m，thesub- surfaceliquidcirculationreduceshandincreases L.ThehorizontalcharacteristiclengthLhasthe minimumvalueofH =4mm.Fortheinitialfuel thicknes，H>5mm，both the characte1'istic lengthLandthecirculationdepthhinc1'easewith aninc1'easeintheinitialfuelthicknesandreach itsownconstantvaluewhenthethickliquidpol
10 Thγinliquidpol r <=コ 1
Fig.4.Ratioofcharacteristiclengthandcharacteristic depthofsub-surfacelayercirculationversusinitialfuel layer thicknes (fuel is n-propanol and tray width= 20m).
s7.5
ヘL
今 Thは liq叫 ∞ 。 。 fc
5 10 15 20 25 Hmm
Fig. 5. Characteristic length and characteristic depth definedinnomenclatureversustheinitialfuellayer thicknes(fuelisn-propanolandtraywidth=20m).
conditionwassatisfied.Thesu1'facewavevelocity mayalsodecreasewithadecreaseintheinitial fuelthicknesbecausebuoyancyefectissup- presedbytheincreasedMαefect(seFig.3). As conlpared to normal gravity，however，the Maefectforthinliquidpolsisrelativelysmal inmicrog1'avity，indicatingthatasurfacewave canberelativelyeasilycreatedint~in liquidpols undermicrogravity.
3.2.Scaleanalysisonsα71.espread
Itisreasonabletoassumethatthe:flamespread ratecanbeuniquelycorelatedwiththeaverage velocity1ofliquidsurface.flow.Consideringthat FrisproportionaltouandMaisproportionalto u-O.5，andasumingthatWenumberdoesnotaf- fecttheflmnespread1'atebecauseitdoesnotin- clude h and the wavelength of surface wave is expectednottohavedi1'ectimpactontheflame spreadrate，theflames.Rr~ad ratecanbeexpresed asafunctionof(Fr/M(t.コ).Theflamespreadrates undersub-flashconditionsmeasuredinthisstudy (forfourdiferentfuels:nlethanol，ethanol，n-pro- panol，andl-butanol)andexperimentaldatafrom [15，}gJareplotedinFig.6asafunctionof(Fr/ MαU'J).Theinitialfuelthicknesofthesedatavaト iesfr01n1to20mm.A cleardi町erencebetwen thick liquid po ols (H>5mm) and thin liquid po ols ( H < 5 m m ) can be se en: for thick liquid
5 d
|必
'po)s，- u
:tl~me
spread rate corelates with (Fr/
2.5
コ)ーしコ， whilefC?r_th~I! lateswith(Fr/Mα0.5)-1.0.Microgravitydata[IJ belongtothethinliquidpolregion，butarecon- siderablyofline，probablybecauseittakeslonger timeunde1' nlicrogravitythannormalgravityto reachafulydevelopedcirculationnlodedueto thelackofg1'avity，whichhelpstocommunicate eachliquidelement.Thismeansthatthemicro gravity experiments require longer tray length thanthenormalgravityones(whichusualyem- ploy30-40cmlongtray)toachieveasteadystate condition.However，thedata[IJwereobtained witha40Clnlongtray.Ifalonger(1m orlonge1')
Ma
liquidfuelpolsitcore-
，
h