TEST FACILITIES FOR ULTRA-HJGH-SPEED AERODYNAMICS By R Smelt; GDF> ARO Inc Hypersonic wind tunnels with test section temperatures approach-
Financial shared support is most critical right now for the facilities at Ames: the 12-Foot, the National Full-Scale Aerodynamics Complex (NFAC), and the 11-
wte larc nasa gov/facilities/aerodynamics/14X22 Northrop Grumman www is northropgrumman com/test/test_capabilities/wind_ tunnel/wind_ tunnel html
The CTF provides lethality, rocket propulsion, aero- thermal, supersonic, hypersonic, and space test and evaluation services The Complex coordinates testing
AEDC User Test Facilities (Nominal Values) Hypersonics/Aerodynamics Test Section Size, in Total Pressure, psia Total Temperature,
Icing Research Tunnel Propulsion Systems Laboratory Engine Component Facilities Aero-Acoustic Propulsion Lab NASA GRC Unique Aero Test Facilities
We're the most complete ground testing facility in the world core competencies in aero-sciences, acoustics, structures, and materials to identify and
![[PDF] test facilities for ultra-high-speed aerodynamics - DTIC [PDF] test facilities for ultra-high-speed aerodynamics - DTIC](https://pdfprof.com/EN_PDFV2/Docs/PDF_3/3813_3AD0064755.pdf.jpg)
3813_3AD0064755.pdf
AEDC- TR-55-6 ' : ',.,
® @
AIR TEST
FACILITIES FOR ULTRA-HIGH-SPEED
AERODYNAMICS @
By R. Smelt; 6DF, ARO, Inc. C.," ..', ,-,: I t" ( ; t" '" ,': ....... :
June 1955 ®
ARNOLD ENGINEERING
DEVELOPMENT CENTER
RESEARCH AND DEVELOPMENT
U S ~)~~A F COMMAND
,wo,, ,r a
AF-AEDC
Tullahoma,Tennessee
AEDC·TR·55·6
TESTFACILITIESFORULTRA-HJGH-SPEEDAERODYNAMICS
By
R.Smelt;GDF>ARO.Inc.
June1955
ContractNo.AF18(600)-1233
CONTENTS
SUMMARY
NOMENCLATURE
AEDC·TR·55·6
Page ---, I. II.
INTRODUCTION.
ENGINEERINGLIMITATIONSTOHYPERSONIC
WINDTUNNELS.. . . . . . . . . . . .
7 9
III.PHYSICALLIMITATIONSTOHYPERSONIC
WINDTUNNELS.. . . . . . . . . . . .
IV.FACILITIESFORMACHNUMBERSABOVE12
V.SEPARATIONOFINDIVIDUALPHENOMENA.
VI.SHORT-DURATIONHIGHMACHNUMBERFACILITIES.
VII.HEATADDITIONTOASUPERSONICSTREAM
REFERENCES.. . . . . . . . . . . . . . . .
15 19 23
24
30
32
AEOC·TR·55·6
ILLUSTRATIONS
Figure
1.ApproximateTemperatureRequiredtoAvoidLiquefaction
ofAirinHypersonicWindTunnels.. . . . . . . . . .9 2.
MaximumHeat-TransferRateatHypersonicTunnelThroat;
VariationwithMachNumberandReynoldsNumber. . ..12
3.LimitingMachNumber-ReynoldsNumberRelationinContinuous
HypersonicTunnels,fromCoolingConsiderations.....13 4. 5.
EnthalpyofAiratHighTemperatures
TestArea/ThroatAreaRatioinTemperature
SimulatingHypersonicTunnels. . . . . . .
15 20
6.MaximumHeat-TransferRateatThroatofTemperature
SimulatingTunnel;VariationwithMachNumberand
ReynoldsNumber. . . . . . . . . . . . . . . . .21
7.LimitingMachNumber-ReynoldsNumberinContinuous
Temperature-SimulatingTunnels,fromCooling
Conditions. . . . . . . . . . . . . . . . . .22
8.Operating-TimeLimitationsinIntermittentTemperature
SimulatingTunnelsDuetoOverheating(Steelwalls
assumed;multiplyby20fortungstenwalls).. . . . . .27
9.LimitingMachNumber-ReynoldsNumberinImpulseTunnels,
asDeterminedbyWall-Melting;OperatingTimeOne Millisecond. . . . . . . . . . . . . . . . . . . . . . .27
10.TunnelArea/DriverTubeAreaRatioinNonreflecting
ImpulseTunnels. . . . . . . . . . . . . . . . .28
AEDC·TR·55·6
SUMMARY
Hypersonicwindtunnelswithtestsectiontemperaturesapproach ingliquefactionareshowntohaveanupperlimitinMachnumberofabout
12withReynoldsnumbersinthegasdynamicflowregime.Thislimit
arisesbothfromtheproblemofcoolingthetunnelwalls,andfromthe requirementforcorrectsimulationofairtemperatureaboveMach number12,toreproducedissociationandionizationphenomena. ThreetypesoffacilitiesforhigherMachnumber,providingcor rectflighttemperatures,arediscussed:
1.Aconventionalwindtunnelwithsufficientlyhighsupplytem
perature.ThisappearsonlypracticableatReynoldsnumbers intheslip-flowregime.
2.Anintermittentfacilityoperatingforafewmilliseconds,using
shocktubetechniques.MaximumReynoldsnumbersareabout
100timesgreaterthanintheprecedingtype,
3.Asupersonictunnelinwhichtheairisheatedafterreaching
highsupersonicspeed.Techniquesforachievingthisaredis cussed. 5
AEDC·TR·55·6
NOMENCLATURE
ACross-sectionalarea
A*Throatarea
AtCross-sectionalareaofthetestsection
cSpecificheatofthewallmaterial c p
Specificheat,constantpressure
HHeattransferperunitareaofwindtunnelwallpersecond
kThermalconductivityofthewallmaterial khHeattransfercoefficient tv!Machnumber
PoSupplypressure
TTemperature
TmLiquefactiontemperature?fthewallmaterial
ToSupplytemperature
T r
Recoverytemperature
T w
Walltemperature
Time
RReynoldsnumber
rTest-sectionReynoldsnumberperfootofmodellength vVelocity ySpecificheatratio flCoefficientofviscosityatthetestsection
PDensity
6
AEDC-TR-55-6
I.INTRODUCTION
Itiswellknownthatordinarysupersonicwindtunnels,operatingfromanair supplyatapproximatelyroomtemperature,haveanupperlimitinMachnumber whichisdeterminedbythecommencementofliquefactionoftheairaroundthe modelinthetestsection.Forawindtunnelwithasupplypressureofoneat mosphereandasupplytemperatureof80 o
F,theairinthetestsectionreaches
theliquefactionpointataMachnumberof5.Thisdoesnotnecessarilyimply thatthewindtunnelceasestogiveusefulresultsexactlyataMachnumberof5; theexactlimitisdependentuponthemagnitudeofthelocaladditionalexpansion aroundthetestobject,andtheextenttowhichtemperatureslowerthantheliq uefactiontemperaturearepermissible,eitherbecauseoflocalsupersaturation duringtheextremelyshorttimeatlowertemperature,orbecausetheheatre leaseisnotsignificantwhenthedegreeofliquefactionissmalLThesefactors, primarilydependentuponthetypeandsizeofmodelandtherequiredprecision ofmeasurement,donotchangethelimitMachnumbergreatlyfromthevaluede duceddirectlyfrom"equilibrium"liquefactiondata. Inthehypersonictunnelsnowoperatinginafewlaboratoriesthroughoutthe country(Refs.I,2,3),theonsetofliquefactionhasbeendelayedbyheatingthe supplyair.Bythismeans,anextensionofthemaximumMachnumbertoabout
10or11hasbeenobtainedattheexpenseofaconsiderableincreaseinthecom
plexityofthewindtunneLFurtherextensionoftheMachnumberrangeofhyper sonicwindtunnelsbymeansofheatingispossible,buttherearepracticalengi neeringlimitstothisprocess.Furthermore,thereisalsoaphysicallimitto thesimulationofflightcharacteristicsathighMachnumberinwindtunnelsof thistype. '7
AEDC·TR·55·6
Inthefollowingparagraphstheseengineeringandphysicallimitationswhich preventfurtherincreaseinMachnumberofhypersonicwindtunnelsaredis cussedinaroughquantitativemanner.ExtensionoftheMachnumberrangeof testfacilitiesbeyondtheselimitscallsforthedevelopmentofnoveltypesof equipment,capableofhandlingtemperaturesandpressureswellbeyondcurrent wind-tunnelexperience.Someoftheproposalswhichhavebeenmadefortest facilitlesofthistypearediscussedlaterinthereport. 8
AEDC-TR·55-6
II.ENGINEERINGLIMITATIONSTOHYPERSONICWINDTUNNELS
Asstatedabove,thetemperaturerequiredtoavoidsignificanteffectsofli- quefactioninahypersonicwindtunnelisdependentupontheintendedapplication ofthewindtunnel,i.e.,uponthetypeofmodelbeingtestedandtherequired precisionofthetestresults.Afairlyrepresentativepictureisobtainedbyas- sumingthattheairatambientpressureinthetestareashouldbeatatempera- tureequaltotheequilibriumdewpoint.Thisimpliesthatthesupplytemperature ToshouldincreasewithincreasingMachnumberM,andalsowithincreasing supplypressurePo,inaccordancewiththefollowingrelation:':< (1+=4.7+3.5log,o(1+-log,oPo (1)
Therelationofsupplytempera-
turewithMachnumberandsupply pressureisshowngraphicallyinFig. evidentthattheupperlimitofMach engineeringabilitytohandlehigh temperaturewithincreasingMach
1.Therapidityoftheriseinsupply
numbershowninthefiguremakesit numberwillbedependentuponthe temperaturegases,i.e.,uponthe
I,OOOI---------:H777"----------I
4 6 81012 1416 1820
MACHNUMBER
5'0001--------------f--1--+1
13 w ffi4'0001--------,---,-,--------,,==-:c-:=----+---f--.f----f---T--j wSUPPLYPRESSURE o1,000ATM l ~ ~ ~ ~ ~ ~ J 10 wI QIII
I-::3,000
w a.. :E w f--2,oool------------:hLA.£-r'-------I a..a.. :::l (f)
Fig.LApproximateTemperatureRequiredtoAvoid
LiquefactionofAirinHypersonicWind
Tunnels
effectivenessofthecoolingarrange- mentswhichcanbemadeinwind expressioncombinesanapproximateClausius-Clapyronrelationbetween liquefactiontemperatureandpressure,basedontheexperimentsofDodgeand Dunbar(Ref.4),withexpansioncharacteristicsforaperfectgaswiththe specific-heatratioY=1.4.Thissecondassumptionisofcourseincreasingly inaccuratewithincreasingsupplytemperature. 9
AEDC·TR·SS·6
tunnelsofthistype. Thereisnogreatengineeringdifficultyinobtainingtherequiredsupplytem perature.Whentherequiredtemperatureexceedsthemaximumwhichcanbe achievedbyconventionalconvectiveheating,additionaltemperaturerisecanbe obtainedbysuchmethodsasadiabaticcompressionorarcheating.Thereisalso noseriousengineeringdifficultyinobtainingtherequiredsupplypressuresfor hypersonicwindtunnels;commercialequipmentandtechniquesareavailablefor pressurestentimesgreaterthanthemaximumnowbeingemployedinsuch tunnels.Theengineeringlimitisdefinedbytheproblemofcoolingthewallsof thewindtunnelexposedtohigh-velocityairstreamsatthesehightemperatures andpressures. Toexaminethisupperlimitmoreclosely,itisnecessarytoobtainarough measureofthecoolingrequirementsasafunctionofMachnumber.Theheat transferliperunitareaofthewindtunnelwallpersecondmaybewritteninthe form: (2)
InthisexpressionthequantitieskhandT
r varysomewhatwiththelocalMach numberandReynoldsnumberoftheair,Le.,withpositionalongthetunnel;but theirvariationisnotsignificantcomparedwiththechangesinPandv.Inprac tice,therefore,thecoolingprobleminahypersonicwindtunnelisatitsworst inareasnearwhichthequantityP•visamaximum,i.e.,nearthethroatofthe wind-tunnelnozzle.Theheat-transferrateisinfactroughlyinverselypropor tionaltothecross-sectionalareaofthewindtunnelateverypoint. Themagnitudeofthecoolingproblemisevidentlyincreasedwithincreasing supplypressure,andtherewouldinfactbenodifficultyinoperatingatunnelat extremelyhighMachnumberifitwerepermissibletouseasufficientlylow 10
AEDC·TR·55·6
supplypressure.Thereareseveralreasonswhythisisnotpossible.Oneob- viousproblemwouldbetoobserveormeasureaerodynamiccharacteristicsin thetestareaatextremelylowdensity. Afurtherrequirementforhighsupplypressureisinherentintheaerody namicapplicationsofhypersonicwindtunnels.Veryfewproblemsathypersonic speedsareassociatedwithperfect-gasaer0dynamics;themostimportantques tionsrelatetotherateofheatingorthedragofahypersonicvehicle,andthese requireclosesimulationofboundary-layercharacteristicsonatestmodel.In otherwords,itwillalwaysbeimportantinahypersonictunneltoobtainReynolds numbersasclosetoflightvaluesaspossible.Inpractice,ofcourse,asin lower-speedwindtunnels,exactsimulationofflightReynoldsnumberswillfre quentlybeimpracticable,andtherequirementwillbereducedtoaneedforsim ulationofthegeneralcharacteroftheboundary-layerflow.Asexamples,a turbulentboundarylayerinaflightshouldofcoursebesimulatedbyaturbulent layeronthetestobject,andflightinthegasdynamicsregionwillnotbeade quatelyrepresentedifwindtunnelReynoldsnumbersareintheslip-flowregion (definedasR<10 4 M 2 ). Anapproximateassessmentofthecoolingproblemasinfluencedbytest Reynoldsnumbercanbemadeverysimply.Ifthetest-sectionReynoldsnumber perfootofmodellengthisdenotedbyr,then: r'"pV/flattestsection.
FromEquation(2)above:
H'"khc
p r/l-(T r -T w)( 3) ThisequationhasbeenemployedtoconstructFig.2,whichshowsthevari ationinmaximumheat-transferrateatthethroatwithMachnumberandReyn oldsnumberperfoot,assumingthatT r isequaltothesupplytemperatureas 11
AEDC-TR-55·6
showninFig.1.andthatanacceptablewalltemperatureatthethroatis1000 0 R (540 0 F).Thereisconsiderableuncertaintyinassessingavalueofkhappro- priatetoboundary-layerconditionsatthethroat.inviewofthehighpressure gradientinthisregion.theundefinedReynoldsnumber.andthelackofknowledge oftransitionphenomenainsucharegion.Basedonhypersonictunneldesignex- perience.aconstantvalueof0.0014hasbeenchosen;thisisequivalenttoan assumptionthattransitiontoturbulentflowtakesplaceinthevicinityofthe throat.andisprobablysomewhatconservative.Amoreexactvaluecouldbe obtainedbymakingastep-by-stepcalculationofthegrowthoftheboundarylayer fromthesubsonicsectionofthenozzlethroughthethroat;aconsiderableamount ofanalysisofthistype.asyetunpublished.hasbeenmadebyProfessorPaulA. LibbyofPolytechnicInstituteofBrooklyn.Fortheverygeneralpurposeofthe presentsurvey.thesimplerapproachofassumingaconstantheat-transferco- efficientisadequate. ~ a: :::> (f) lJ. o t:: CJ (f) " a: :::> a: lJJ a.. :::> f- l1J a: a: lJJ lJ. (f) Z f- lJJ I o a: I f-
REYNOLDSNUMBER
PERFT:
107
10 6
J-----------jL--f-+---L--+--------i
o24'6 810121416 1820
MACHNUMBER
Thethroatheat-transferrate
showninFig.2isgivenoverarange ofReynoldsnumberfrom5millionsto
0.1millionperfoot.whichcoversthe
rangeoftypicalhypersonictunnels.
TheReynoldsnumberboundarybelow
whichslip-flowphenomenabeginto appearisshowninthefigureforamo- dellengthof5ft.;equivalentcurves forotherlengthsareeasilydeter- mined.
Theheat-transferratehasbeen
Fig.2.MaximumHeat-TransferRateatHypersonic
TunnelThroat;VariationwithMachNum
berandReynoldsNumber 12 expressedinBtuperhourpersquare
AEDC·TR·55·6
footofsurface,sincethisistheformmostfamiliartoheatengineers.Forcom- parison,typicalheat-transferratesinhigh-capacity,high-temperaturesteam boilersareintherangefrom45,000to80,000Btu/hr/sqft,>:
Thehighestheat-transferrates knowledgeoftheauthor,areencoun- incurrentengineeringpractice,tothe teredinrocketmotors;andthecooling nelproblem.Itisreasonabletoas- isverysimilartothehypersonictun- problematthethroatofsuchmotors SLIPFLOW
10'1--------"---
2:<10 6 MAXIMUMHEAT
TRANSFERRATE/SOFT:-
2xI0 5 1------------------".___-
5 0: g; 0: 5:<10 5 ::;: => z If) o <5 z )0- w 0: 5:<10 6 1-------'<----'--------------1
IOxI061-----\----':---------------1
sumethereforethatifalltherocket 101214161820
MACHNUMBER
Fig.3.LimitingMachNumber-ReynoldsNumber
RelationinContinuousHypersonicTunnels,
fromCoolingConsiderations developmentsinliquidjacketcooling, boilingfluidcooling,orfilmcooling areadoptedforuseinhypersonictunnels,similarmaximumheat-flowrates,of theorderof10 7 Btu/hr/sqft,shouldbepossible.Thecorrespondingmaximum Machnumber,asafunctionofReynoldsnumberperfoot,isshownasthefull curveinFig.3.Thisfigurealsoincludesasimilarcurve,showndotted,il- lustratingtheeffectofdoublingthemaximumheat-transferrate.Itisevident thatsuchadevelopment,ifpossible,wouldonlypermitthelimitMachnumberto beincreasedbyabout1.O. Thisassumptionthatheat-transferratesequaltothoseofrocketmotors shouldbeattainableinhypersonictunnelpracticeisanoptimisticone,sincethe hypersonictunnelthroatpresentsanadditionalproblem,inaccuracyofprofile, Kent'sMechanicalEngineersIHandbook,Power,12thEdition,Section7, p.16. 13 AEDC·TR·55·6
notpresentinrocketmotors.Toillustratethisproblem,considerahypersonic tunnelwithatestsection1-ftsquare,utilizingatwo-dimensionalnozzletoobtain aMachnumberof10.Theordinateatthethroatisthenonly0.022in.,anda changeofonly0.001in.inthisordinatewillchangethetest-sectionMachnumber by0.1andthetest···sectionstaticpressureby7percent.Anidenticalchangea crossthewholewidthofthisthroatcanofcoursebecorrected;butifthelarge heatflowproducesuneventemperaturesanddistortioninthethroatwall,there sultingtransversenonuniformitiesintheflowcannotbeeliminated. Itisevidentthatahypersonictunnelaimedatreachinglimitheat-transfer conditionsatthethroatshouldpreferablyhaveanaxiallysymmetricalnozzle, notonlytoreducedistortionbutalsotominimizethesurfaceareainthevicinity ofthethroat.Theaxiallysymmetricalnozzlehasnotbeenadoptedgenerallyin hypersonicwindtunnelshowever,fortworeasons:(1)afearthattheaxialpres suredistributionmaybenonuniformbecauseoffocussingofwalldisturbances, and(2)arequirementforvariationinMachnumberofthenozzle,whichismore difficulttoachieveinanaxiallysymmetricaldesign. 14 A EDC.TR·55·6
III.PHYSICALLIMITATIONSTOHYPERSONICWINDTUNNELS
Hypersonicwindtunnelsofthetypeconsideredinthelastparagraphdonot simulatethetemperatureswhichexistaroundanobjectinflight,butoperateat ambienttemperaturesclosetotheliquefactiontemperatureofair.Ifairwerea perfectgaswithconstantspecificheats,overthewholetemperaturerangeex- periencedintunnelandflight,alltemperatureswouldbeproportional,andanac- curatepictureofflighttemperaturedistributionswouldbeobtainedfromtunnel testsprovidedthatsurfacetemperaturesweresimulated.Unfortunatelyairis notaperfectgas,anditsdeviationsbecomesignificantatthehightemperatures generatedinveryhighspeedflight. i= a:: r--IN "1.6f----------+---+----j'-----j-----j c. u g I-1.51---------+-+-_+_-{-----j
a:: lJ.J a- LLlitI----------,---+-+-+-+-{------j
o ....I " 1.3f---------/--/t--cH7----------j
z lJ.J ....I 1 . 2 f - - - - - - - - f h ~ - - - - - - - - ; I z lJ.J 1.11------,£----1---------\
1.0o!:-..L--;2::-:,obo:::::
o --'I;;:::jO,oOO TEMPERATURE(DEGREESR)
Fig.4.EnthalpyofAiratHighTemperatures
Thesedeviationsarepresentedin
Fig.4,inwhichtheenthalpyathigh
temperaturesiscomparedwiththe valuewhichwouldbeexpectedifthe specificheatremainedconstantatthe lowtemperaturevalue.Thisfigure hasbeenconstructedfromthetables ofRef.5.Itshowshowtheenthalpy increaseswithincreasingtempera- ture,astheadditionalenergyisfirst absorbedinexcitingthevibrational degreeoffreedomofthemolecule,and thenindissociationofthemolecule. Theabsorptionofenergyinvibration
15 AEDC-TR-55-6
isnotgreatlyaffectedbypressure,butdissociationproceedsmuchmorerap- idlyatlowerpressures. Intheflowaroundamodelinahypersonicwindtunnel,theairisatlow temperatureandbehavesessentiallyasaperfectgas,withenthalpycorresponding tothevalueof1.0onFig.4.Figure4thusexpressestheratiobetweenthe actualenthalpyinflightandthevaluesimulatedinthewindtunnel,asafunction offlighttemperature.Theexactextenttowhichthisdifferencechangestheaero dynamicparametersofcoursedependsuponthedetailsoftheflow;atpresent comparisonscanonlybemadeinafewcaseswheretheflowoftherealgaswith vibrationanddissociationhasbeencomputedtheoretically.Thecharacteristics ofanormalshockinarealgashavebeencalculatedbyBetheandTeller(Ref.6) andlatercalculationsarealsogiveninRef.5.Thelaminarflowintheboundary layerhasbeencomputedbyMoore(Ref.7),Crown(Ref.8)andothers.Gen eralizingthesecalculations,itappearsthatchangesinthetemperatureandden sityofthesameorderasthechangeinenthalpyshowninFig.4maybeexpected, althoughthepressuresarenotgenerallymodifiedtothesameextent. Withthisbackground,Fig.4canbeusedtoestimateroughlythemaximum Machnumberatwhichhypersonicwind-tunneltestresultscanbeappliedtoflight conditions.ThereisevidentlyaMachnumberrangeoverwhichthechangesin airpropertiesinflightaresosmallthattheycanbeneglectedcompletely.The pointwheretheenthalpyhaschangedbyonepercenthasbeenarbitrarilyselected asdefiningtheupperlimitinthisrange.Thereisasecondregimeinwhich significantbutsmalldifferencesbetweenwindtunnelandflightcharacteristics aretobeexpected;thesedifferencesmightbetreatedassmallcorrectionstothe wind-tunneltestresults,thecorrectionsbeingbasedlargelyupontheoretical treatmentoftheconsequencesofairimperfection.Fromthepointofviewof 16 AEDC·TR·55·6
simplificationofthetheoreticaltreatment,itappearsadvantageoustodefinethe upperlimitofthisregimeasthepointatwhichdissociationbecomessignificant, therebyconfiningthecorrectionstovibrationaleffectsonly.Inthethirdand highestrangeofMachnumber,vibrationaleffectsarelargeandappreciable dissociationisalsopresent.Theoreticaltreatmentisthenmorecomplex,the differencesbetweentunnelandflightcharacteristicsarelarge,anditappears thatthehypersonicwindtunnelhaslostmuchofitsutility. Thetemperaturesdefiningthelimitsoftheseregimescanbeobtaineddi rectlyfromFig.4.Atfirstsight,thelimitMachnumbercouldbeobtainedby equatingthesetemperaturestothestagnationorrecoverytemperatureinflight, sincethesetemperaturesareattainedbehindanormalshockorattheinneredge oftheboundarylayeronaninsulatedwall.Fromthepracticalpointofview,this wouldbeover-conservative;avehicleflyingatveryhighMachnumbersisnot likelytohaveanyextensiveareasofstagnationconditions,anditswalltempera turemustbewellbelowtherecoverytemperature.Undertheseconditions,the maximumtemperatureoccursintheboundarylayerawayfromthewall;the temperatureriseisaboutone-fourthofthefullstagnationvalue(Ref.9).The Machnumberlimitsforthethreeregimesundertheseconditionsaregivenin Table1below.InpreparingthistablethedissociationcurveofFig.4fora pressureofO.01atmospherehasbeenused,sincethiscorrespondsapproxi matelytoboundarylayerconditionsatthetopofthestratosphere. 17 AEDC·TR-SS-6
TableI.RangesofApplicationofHypersonicWindTunnels
RangeIRangeIIRangeIII
ResultsApplicableCorrectionsforLargeDifferences
toFlightwithoutVibrationEffectsbetweenTunnels CorrectionRequiredandFlight
Maximumlocal
temperature(0R)Upto10001000-3400Above3400 Corresponding
Stagnationtem-
perature(0R)Upto28002800-12,400Above12,400 FlightMachnum-
berinstratosphere (T=400 0 R)Upto5.55.5-12Above12
Fromthistableitisevidentthatthehypersonictunnellosesmuchofitsvalue attheupperlimitofRangeII,i.e.,ataMachnumberofabout12.Itisinter- estingtocomparethisconclusionwiththecurveofFig.3,whichshowsthatthe practicalcoolingproblematthetunnelthroatbringstheReynoldsnumberalmost downtoslip-flowvaluesataMachnumberof12. ItwillbeobservedfromTableIthatcorrectionsforthevibrationaleffects inflightarerequiredoverquiteawiderangeofhypersonictunnelMachnumber, i.e.,aboveaMachnumberof5.5.Thereisatpresentverylittleexperimental ortheoreticaldatauponwhichtobasecorrectionsofthistype.Probablythe greatestneedatpresentisforanadequatetreatmentoftheturbulentboundary layerinarealgas,topermittheinterpretationofhypersonictunnelmeasure- mentsofheat-transferrateandskinfriction.Theproblemiscomplicatedby therelaxationtimeofthevibrationaldegreeoffreedom,butevenanequilibrium theorywouldbeofconsiderableassistance.ThenewfacilityattheFreeport LaboratoryofthePolytechnicInstituteofBrooklyn,withitsabilitytoduplicate flighttemperaturesatmoderatelyhighMachnumbers,shouldgivemuch-needed informationinthisarea. 18 AEDC·TR·55·6
IV.FACILITIESFORMACHNUMBERSABOVE12
InthehighestrangeofTable1,aboveaMachnumberof12,itappearsnec essarythatthetestconditionsshouldsimulatetheactualtemperatureoftheair inflight,sincethechangesinairpropertiesathightemperaturewillhaveapre dominanteffectuponthewholeaerodynamicpicture. Tobestrictlycorrect,italsoappearsnecessarytohavecorrectsimulation ofpressuresaroundthetestobject,sincethedegreeofdissociationischanged quiteappreciablybychangeinpressure.ItisevidentfromFig.4howeverthat theeffectofareducedtestpressurecouldbecompensatedapproximatelybya reductioninthestagnationtemperature,atleastoveramoderaterangeinpressure. Theinfluenceofpressureontherelaxationeffectsassociatedwithdissociation andvibrationpresentsmoredifficultproblems.Iftheseeffectsareimportant inflight,theirsimulationonthemodelrequiresanequalnumberofmolecular collisionsinacomparablelengthoftestobject,andthisresolvesitselfintoa requirementforequalReynoldsnumbersinmodelandflight.Ofcourse,ifthe flightReynoldsnumberissohighthatrelaxationeffectsarenegligible,there quirementforequalmodelReynoldsnumberisreplacedbyarequirementthat themodelReynoldsnumbershouldbehighenoughtoavoidsignificantrelaxation distancesinthiscasealso. Itisevidentfromtheearlierexaminationofthecoolingproblemsofhyper sonicwindtunnels(SectionII)thatMachnumbersabove12willnotbeattainable exceptatverylowReynoldsnumbers.Therequirementforcorrecttest-section temperatureofcourseconsiderablyenlargesthecoolingproblem;thetest-section temperatureisincreasedbyafactorofabout4comparedwiththatofanordinary 19 AEDC-TR-55-6
hypersonictunnel,sothattheheattransferatallpointsismultipliedbyabout thesamefactorfromthiscausealone.Thereare,however,twootherstrong effectswhichfurtherincreasetheheatingrateatthethroatofsuchatunnel: 100,000r--.......,--,-----,----r---,.---r--,----,-,.,-...---;
Throughoutmostoftheexpansion
processinthenozzleofatem perature-simulatingtunn e1,the vibrationalmodeoftheairisal mostfullyexcited,andthereis appreciabledissociation.Under theseconditionsthespecificheat isgreatlyincreasedandthespe cificheatratioYhasfallentoa valuebetween1.2and1.3.This considerablyincreasestheratio oftest-sectionareatothroatarea foragivenMachnumber;anap proximateestimateofthiseffect hasbeenmade,andisshownin Fig.5.Boththisratioandthe
specificheatenterintotheex pression(Equation3)fortheheat transferrateatthethroat,with theresultthatvery1a r g ein creasesinthisheatratecanbe expectedwhentheimperfections oftheairaretakenintoaccount. 2. 20181614
.........IDEALGAS "1'.1.4 6 I Ia I / I / D. ;'/ Ia D./ I 4 ,'6 I 10 I I Theincreasedtest-sectiontemperatureincreasestheviscosity,andre ducestheReynoldsnumber.ToobtainthesameReynoldsnumberper footasinthehypersonictunnel consideredinearliersections,it isnecessarytomultiplytheden sitybyafactorofalmost4.This ofcourseincreas e stheheat transferrateatthethroatbythe samefactor. 2 1. o Il // , IDEALGAS..........Ii
7=1.29//'AIR
10,000'-----:'-------;r<---J
/ a // 1/ // D.a I / 1,000 " UJ a: " !;:( 0 a: :I: " " UJ 100a:
" z 0 i= u UJ en en UJ 10 81012.
MACHNUMBER
Fig.5.TestArea/ThroatAreaRatioinTemperature
SimulatingHypersonicTunnels
ThecurvesshowninFigs.6and7havebeenconstructedtoprovidearough illustrationofthethroat-coolingproblemwhenallthesefactorsaretakeninto account.TheyshouldbecontrastedwiththesimilarcurvesshowninFigs.2 and3fortheheat-transferconditionsinaconventionalhypersonictunnel.It shouldbeemphasizedthattheheat-transferphenomenaatthethroataresocom- plexunderthehightemperatureconditionsnowbeingconsidered,andthedataon 20 AEDC-TR-55-6
Q) u .E .... :::> I/) - REYNOLDS0
"'"' NUMBER
- PERFT:
r:i- I/) ...... .... 10" :::> 0 .<= .... a> a. :::> I- ID LiJ t: a:: a:: LiJ lJ.. 10 7 CJ) Z <{ a:: l- t: LiJ J: t: 0 a:: J: I- 10"0 24
airpropertiesaresosparse.thatFig. 6mustberegardedonlyasanorder-
of-magnitudeestimateoftheprobable heat-transferrateandisprobablyan underestimateofthetotalheatflow inthisarea.sinceimportanteffects suchasradiationhavebeencompletely neglected.Itservesthepurposehow- everofdemonstratingthatawindtun- nelsimulatingcorrectflighttemp- eraturesandReynoldsnumbersinthe gasdynamicregimeisquiteimprac- 161820
ticaLItappearspossibletomake Fig.6.MaximumHeat-TransferRateatThroatof
Temperature-Simulating
Tunnel;Variation
withMachNumberandReynoldsNumber suchawindtunneltogivelowslip- flowReynoldsnumbersatMachnum- bersaround20.andtoextendoutoftheslip-flowregionatthelowerMach numbersaround10.Whileafacilityofthistypewouldclearlyhavesomeutil- ity.thepracticaldesignerofvehiclesforflightattheseextremeMachnumbers willundoubtedlybemoreconcernedwithaerodynamicproblemsatthehigher Reynoldsnumberswhichcorrespondtoflightconditionsatloweraltitudes.Fig- ures6and7invitespeculationontheextenttowhichitmaybepossibleinthe futuretoimprovecoolingconditionsatthethroatofahigh-temperaturetunnel bydevelopmentofnewcoolingtechniques.Thereisnodoubtthatsomeimprove- mentwilloccur;intheopinionoftheauthor.thefullgaintobeofferedbythe useofcoldairinjectionaheadofthethroathasnotbeenrealizeduptothe 21
AEDC·TR·55·6
10' present.Theremayalsobedevel- \GASDYNAMICFLOW \ _SLIPFLOWLIMIT,5FTMODEL (R:I0 4M2) opedhigh-temperaturematerialsca- pableofprovidingsufficientstrength 10' \ 5\ i!''\ ffiMAXIMUMHEAT\.SLIPFLOW a..TRANSFERRATE/SQFT:\ 10 4 1--------->.;,-----\-'--------------1
10MILLIONBTU/HR/\..
z 20MILLIONBTU/HR'"
"\ '" tocontainthethroatpressureattem- peraturesofseveralthousandde- greesRankine.Butinfact,suchim- provementsincoolingsurfacetem- peraturebecomeinsignificantincom- ratesshowninFig.6areseveral ordersofmagnitudehigherthanthe temperature;andtheheattransfer parisonwiththerequiredair-supply presentpracticalmaximum,sothata 20181614
MACHNUMBER
1210
10' 10'L-__L-_----JL-_----"__-----l__....J.__.-I
Fig.7.LimitingMachNumber-ReynoldsNumberin
ContinuousTemperature-SimulatingTunnels,
fromCoolingConditions verymajorengineeringdevelopment mustbepostulatedtomakefacilitiesofthistypepossible.Researchworkersin thefieldhaverealizedthislimitation,andhavesoughtalternativeapproachesto obtainaerodynamicdataatveryhighMachnumbers.Someoftheseapproaches aredescribedinthefollowingthreesections. 22
AEDC.TR·55·6
V.SEPARATIONOFINDIVIDUALPHENOMENA
Intheprecedingsections,theaimhasbeentoprovideameansofsimu latingalltheparameterswhichenterintotheaerodynamicbehaviorofairin highMachnumberflight,includingMachnumberitself,viscosity,vibration, anddissociation.Thisisveryevidentlyadifficultobjectivetoattain,andsome ofthepresentapproachestotheproblemhaveconfinedthemselvestothesimpler problemofsimulationofonlyafewoftheparameters.Assoonasthissimpli ficationisaccepted,itisnolongernecessarytouseairastheworkingfluid, andothergaseswhichpermitthepracticalvaluesofthecriticalparametersto beobtainedmorereadilycanbeemployed. Forexample,apartialsolutionoftheprobleminvolvingonlytheeffectsof Machnumberandviscosity,andneglectingthevariationinairpropertiesdueto vibrationanddissociation,canbeobtainedbytheuseofagaswithalowlique factionpointsuchashelium.TheworkofProfessorBogdonoffofPrinceton University(Ref.10)isagoodexampleofthisapproach.Thelowtemperatures towhichgaseousheliumcanbereducedinthetestsectionpermitsratherhigh Reynoldsnumberstobeobtained,sothathypersonicviscousproblemsinthe gasdynamicregimecanbestudied.Thespecificheatratioyisnotcorrectly simulated,butitisevidentthatinsistenceoncorrectsimulationwouldleadtoa requirementforsimulationofthevariationwhichoccursinairinflight,andto thesamepracticalproblemasdiscussedinthelastsection. Anexperimentalapproachtotheeffectsofdissociationontheaerodynamic characteristicsisalsopossiblebychangeoftheworkingfluid.Thisapproach hasbeeninvestigatedattheNavalOrdnanceLaboratorybyDr0Slawskyandhis associates.Theyemploygasessuchasbromineandchlorinewhichdissociate 23
AEDC·TR·55·6
atverylowtemperatures;thispermitsthephenomenaassociatedwithdissocia- tiontobeexaminedwithoutseriouscoolingproblems.Atpresenttheworkin thisareahasbeenlargelyamatteroffree-flightinvestigationsusingballistic rangetechniques;thepossibilityofawindtunnelemployingreadily-dissociating gasesshouldalsobeexplored. VI.SHORT-DURATIONHIGHMACHNUMBERFACILITIES
Ifairistobeusedastheworkingfluid,itisevidentthatsomemeansmust befoundtocircumventthecoolingproblematthetunnelthroat.Onepossibility istoreducethetimeofoperationofthehigh-temperatureflowtosuchanextent thattheheattransfertothecriticalpartsofthewallisnotlargeenoughtodo damage.Thelimitingdurationofafacilityofthistypeisquitedifficulttoesti- matebecausetheextremelyhighheat-transferrate,operatingforaveryshort time,canproducetwotypesoffailure.Itnecessarilyresultsinextremedif- ferencesintemperaturebetweendifferentportionsofthetunnelwalls,giving risetothermalstresseswhichmaybeexcessive;oralternatively,themaximum temperatureontheinnerwallmaybesufficienttoliquefythesurfaceatthepoints ofmaximumheat-transferrate.Aquantitativepictureofthistransientheating processcanbeobtainedbyapplicationofclassicalheat-conductiontheory.The temperatureTatadistancexfromtheairsurfaceinthemetalaftertimet,is givenby: where n=k/cp xerfc(_x)] 2yni (4) Thetimetakenforthetemperatureatthesurface(x0)toreachthelique- factionvalueTmisthengivenby 24
t=ITkpc 4 (5) AEDC.TR·55·6
Ifthewallismadeofsteel,insertionofnumericalvaluesleadstotherough resultt=whereHismeasuredinBtuperhourpersquarefoot.However, steelisnotthebestmaterialforthisapplication;evidentlythelongestoperating timeisobtainedwithamaterialgivingthemaximumvalueofkpcT m 2; thissug geststheuseoftungstenoraceramicmaterial.Tungsten,forexample,gives twentytimesthedurationofsteel,beforeitsmeltingpointisreached. Itappearsatfirstsightthattheconditionoftheinnersurfaceofthewall reachingitsmeltingpointmightberegardedasanupperlimittotheoperating timeinintermittentfacilities.Thereishowevernostructuralproblempre sentedbythiscondition;thehightemperaturesareconfinedtoonlyafewthou sandthsofaninchindepthsothatthemajorityofthewallstillmaintainsitsorig inalstrength.Furthermore,thequantityofwallmaterialwhichismeltedis notverylarge.Inpointoffact,someoftheexperimentalfacilitiesnowinoper ationutilizingtheshocktubeprinciple,describedinRef.11,alreadyoperatefor adurationsufficienttoproducemeltingandevenevaporation.Themainproblem undertheseconditionsisthepollutionoftheair,andeventuallyofthewallsof theshocktubeandtunnel,whichmakesitnecessarytoprovideforfrequent cleaningandreplacementofcriticalareas. Wecanassume,however,thattheupperlimitinoperatingtimewillbear somerelationtotheliquefactiontimeasobtainedfromEquation5.Thisequa tionhasthereforebeenemployed,inconjunctionwiththemaximumheat-flow ratesfromFig.6,toestimatethetimerequiredtoliquefythesurfaceinan intermittenthypersonicfacility.Theresultsofthisestimate,assumingasteel wall,aregiveninFig.8(seepage27).Itisevidentfromthisfigurethata reasonableReynoldsnumbercanonlybeobtainedifsurfacemeltingafterafew millisecondsisaccepted. 25
AEDC·TR·55·6
Thisresultleadsnaturallytotheuseofashocktubeasthedrivingelement ofahypersonicwindtunnel.Thecurrentpositionofdevelopmentof"impulse" windtunnelsofthistypehasbeensummarizedinRef.11.Itisthereforesuf- ficientinthepresentpapertopointoutsomeoftheadvantagesoftheshocktube astheyrelatetotheproblemsdiscussedhere: 1.Theshocktubeisprobablythesimplestequipmentpermittinghigh
pressure,high-temperatureflowstobegeneratedandterminatedwithin afewmilliseconds.Iteliminatesthepracticalproblemsofrapidly operatingvalves. 2.Theshocktubeconsiderablysimplifiestheproblemofobtainingthehigh
stagnationtemperaturesrequiredintheoperationofh ypersoni c temperature-simulatingfacilities.Thetemperatureoftheflowbe hindtheshockcanbemanytimesgreaterthanthetemperaturegenerated inthehigh-pressuredrivingchamber.Itshouldbeobserved,however, thatsufficientlyhightemperaturescancertainlybeobtainedbythedirect useofelectricarcheating. Twotypesofimpulsetunnelsutilizingshocktubeshavebeenconsidered.In thereflectingtype,theshockisdrivenalongthetubeandreflectedfromthewall atthefurtherend.Theflowbehindthereflectedshockisstationaryandatex- tremelyhightemperatureandpressure,sothatitcanbeusedastheairsupply forawindtunnel.Thereis,however,analternativeoperationinwhichthe shockisnotreflected,butinwhichtheendofthetubeisexpandeddirectlyinto thetestarea.Thishastheattractionthatthereisnosonicthroat,andthethroat heat-transferproblemthendoesnotexist.Initsplace,themaximumheat-flow rateoccursatthewallbehindthedrivingshock,whichissubjecttoahigh- pressure,high-temperaturestreamataMachnumberofabout2.Themaximum heat-transferrateundertheseconditionsisabout60percentofthethroatvalue. Takingintoaccountthissmallalleviationofthewallheatingproblemwith nonreflectedoperationofashock-tubetunnel,andassuminganoperatingtimeof onemillisecondbeforeliquefaction,thelimitingMachnumber/Reynoldsnumber 26
AEDC·TR·55·6
201816141210
0.0001L-_--'__..L.L_-'---'-----'-_..!....-__'--'--_---J
8 (ij o z § II) '.J w :E0.11----\--\--------\------\-------1 ...J...J 0.001!----\---T---\------\----___i 8,andisshowninFig.9.Theas- beenderivedfromthecurvesofFig. andthetotaloperatingtimeshouldbe relationforanimpulsetunnelhas pressureshavearesponsetimeofthe MACHNUMBER
thantheresponsetimeoftheinstru- ments.Thereis,ofcourse,scope Fig.8.Operating-TimeLimitationsinIntermittent
Temperature-SimulatingTunnelsDueto
Overheating(Steelwallsassumed;multiply
by20for tungstenwalls) shouldbeobservedthatareductionin onlyincreasethemaximumReynolds withshorterresponsetime;butit correspondstoadistance0fonly more,thetimeofonemillisecond operatingtimebyafactorof10will forfuturedevelopmentofinstruments numberbyafactorof10.Further- twentyfeetinflightataMachnum- STEELWALL
10'!--------------"""""------->,j
(f) o ...J o Z a:: b fr ffi a. ffi 1 0 5 1 - - - - - - - - - ~ - - - - - "---__j ::> z berof20,sothatfurtherreduction 10 3 MACHNUMBER
intimepresumablywillproducein- Fig.9.LimitingMachNumber-ReynoldsNumberin
Impulse
Tunnels,asDeterminedbyWall
Melting;OperatingTimeOneMillisecond
strumentationproblemscomparable withthoseoffree-flighttechniques. 27
AEDC·TR-55-6
AlthoughtheseconsiderationsmayindicatethatthecurvesofFig.9are somewhatconservative,itisprobablethatthefigurerepresentsveryapproxi- matelytheupperlimitatthepresenttime.Ifthedataarecomparedwithcorre- spondingdataonFig.7forthecontinuoustemperature-simulatingtunnel,it willbeobservedthatthemaximumReynoldsnumberhasbeenmultipliedbya factorofalmost100bytheuseofshock-tubetechniques. Atthispoint,itisappropriatetodiscussthemeasurementofmodeltemper- ature,whichconstitutesthechiefprobleminintermittenthypersonictunnels. Sinceheatingislikelytobethemostimportantproblemforthedesignerofhigh Machnumbervehicles,itisessentialtobeabletomeasureratesofheatflow intothemodelinanyhigh-speedtestfacility.Itisobviousthatthispresentsa difficultprobleminimpulsetunnelsoperatingonlyfortimesoftheorderofa millisecond. 10,0001--------------j'-----7I
n:: f- e]1,000 > "