6 some factors of hyper sonic in let airplane interactions

8/9/2019 6 Some Factors of Hyper Sonic In Let Airplane Interactions

1/14

JOURNAL OF AIRCRAFTVol. 39, No. 1, January February 2002

Some Factors of Hypersonic Inlet/Airplane Interactions

Y. P. Goonko and I. I. Mazhul

Russian Academy of Sciences, Novosibirsk, 630090, Russia

Results o f an experimental study a re presented on a erodynamics of a hypersonic vehicle whose air-breathing

engine nacelle, including a two-dimensional inlet, is located under the lifting body. Experiments with a ducted

model were performed in a blow-down wind tunnel at M1 = 4 and 6 . Aerodynamic forces acting on the model

were measured with a balance; aerometry of the inlet-captured airstream was carried out; parameters of the

boundary layer in front of the inlet were also measured. Problems of inlet/airplane interaction such as wedge-

aided diversion of the boundary layer developed upstream of the inlet, lateral ow spillage over the wedge ramp

of the inlet with and without side cheeks, and forebody nose bluntness were studied. Effects of these factors on

the inlet ow-rate, the total aerodynamic and performance characteristics of the vehicle as a whole were revealed.

The resultant aeropropulsive characteristics, which showed ight properties of the vehicle, were obtained by

combining experimental aerodynamic data and calculated estimates of the thrust of the engine in a scramjet

mode. Aeropropulsive force polars demonstrated some special path instability of the vehicle and an extraordinary

behavior of these polars in the case of the blunted forebody.

Nomenclature

A = cross-sectionalarea, m2

A0 = frontal area of the inlet, m2

A1 = cross-sectionalarea of the free airstream capturedby the inlet, m2

NAch = relative cross-sectionalarea of the plenumchamber in the model duct, Ach=AthNAen = relative cross-sectionalentrance area of theinlet, Aen =A0NAnz = relative cross-sectionalarea of the meteringnozzlesat the duct exit,

NnzX

1

Anz

Ath

NAth = relative cross-sectionalthroat area of theinlet, Ath =A0NA0 = relative frontal area of the inlet, A0=Spl

A = cross-sectionalarea covered by boundary-layerdisplacement thickness , m2

b, h = width and height of a componentof theaerodynamicconguration, mm

CD = aerodynamicdrag coefcient, D=q1SplCD0 = drag coefcient at CL D 0CL = aerodynamiclift coefcient, L=q1SplC

L

= CL derivative with respect to the angleof attack, deg1

CRx = coefcient of resultant aeropropulsivetangentialforce, Rx =q1Spl,

CRy = coefcient of resultant aeropropulsiveliftforce, Ry =q1Spl

CT = engine thrust coefcient, T=q1A0D = aerodynamic drag force or x -component of the

resultant aerodynamicforce,Raxdnz = diameter of the metering nozzle, mm

Received 12 February2001; revision received 5 September 2001;acceptedfor publication 26 September 2001. Copyright c 2001 by the AmericanInstitute of Aeronautics and Astronautics, Inc. All rights reserved. Copiesof this paper may be made for personal or internal use, on condition that the

copier pay the $10.00 per-copy fee to t he Copyright Clearance Center, Inc.,222 Rosewood Drive, Danvers, MA 01923; include the code 0021 -8669/02$10.00 in correspondence with the CCC.

Head of Research Group, Institute of Theoretical and Applied Mechan-ics, Siberian Branch. Experimental Aerodynamics Laboratory.

Senior Research Scientist, Instituteo f Theoretical and Applied Mechan-ics, Siberian Branch. Experimental Aerodynamics Laboratory.

f = ow-rate factor of the inlet, A1=A0Hbl = boundary-layershape factor,

=

Hu = fuel caloric value, J/kghblw = height of wedges for boundary-layerdiversion

(diversion wedges), mmI = total momentum of the owstream, N; for the

one-dimensionalowstream, m VC .p p1/AL = aerodynamic lift force or y-component of the

resultant aerodynamicforce,RayL tr = length of boundary-layertransition, mmL =D, = lift-to-drag ratio, maximal lift-to-dragratio(L=D/maxL 0 = stoichiometric mass air-to-fuel ratiol = span of the lifting conguration, mM = ow Mach numberMd = inlet design Mach numberM1 = freestream Mach numberm = mass ow rate, kg/m2s; for the one-dimensional

owstream, VANnz = number of metering nozzlesn = unit-directingvectorp = pressure, N/m2

q1 = freestream dynamic pressure, N/m2

R = vector of the resultant aeropropulsiveforceRa ; Rax ; = vector of the resultant aerodynamicforce and its xRa

y

and y componentsRair ; R f ; = specic constants for air, fuel, and combustionRcp products, respectivelyRb = radius of bluntnessof the lifting-bodynose or the

leading edge of the inlet, mmRg = specic gas constant, J/kg/KRx = tangential component (x component) of the

resultant aeropropulsiveforceRy = lift component (y component) of the resultant

aeropropulsiveforceRe11 = freestream Reynolds number per meter, 1/mRetr = Reynolds number of boundary-layertransitionR z = roughness parameter, mS = plan area of the lifting conguration or wing, m2

Spl = reference plan area of the model conguration, m2

T = vector of engine thrustt = temperature,KV = velocity, m/s

x ; y;z = aerodynamic coordinateaxes accepted in Russia,see Fig. 1

= angle of attack, deg

37


2/14

38 GOONKO AND MAZHUL

af = air-to-fuel ratio = adiabatic exponent (or specic heat ratio), , = boundary-layerthickness, displacementthickness,

and momentum thickness, mm = wedge angle, deg1, 2, 2 = wedge angles of the triple-shockinlet ramp, deg = aspect ratio of the lifting conguration, l2=Spl = density, kg/m3

= total-pressure recovery factor

= sweep angle, deg

Subscripts

a = aerodynamic forcesb = bluntnessb al = b alan ce- measu re d for ce sb ase = b ase fac e o f t he bo dybl = boundary layerblw = wedge for boundary-layerdiversionch = plenumchamber of the model duct; engine

combustion chambercp = combustion p roductsduct = m odel duct f orces

en = entranceof the internal duct of the inleteng = airbreathing engineext = forces over externalsurfacesof the airplane

including the engine nacellefuel = fuel parametersinl = inlet parametersint = internalsection of the inlet from the entrance

to the throatlb = lifting bodym = midsectionnz = metering nozzle of the model ductnze = engine nozzless = forces over surface of the inlet-capturedairstreamsupp = support o f the model

sw = shock wave parametersth = inlet throatw = wing, also wall0 = total (stagnation) ow parameters1 = freestreamparameters

Introduction

A DISTINCTIVE feature of hypersonicvehiclespoweredby air-breathingenginesis a highdegreeof interactionandintegrationof the propulsion unit and the airplane. Problems of this kind aretypical, in particular, of congurations with the engine nacelle lo-cated under the lifting body or the wing, which are conventionalforhypersonicscramjet-poweredvehicles.One of theseproblemsis as-sociatedwith the liftingsurfacesof the vehicleforebodyor the wing

simultaneously acting as precompression surfaces of the airstreamcaptured by the inlet. The boundarylayer developingon such a sur-face and entering the inlet may have an adverse effect on the inletand engine performance. The latter may be improved by placingthe inlet outsidethe boundary layer, for example, by installingit onwedgesfor boundary-layerdiversion. Obviously, these wedgeswillincrease somewhat the drag of the vehicle. Thus, the efciency ofsuch boundary-layer diversion will be determined by the compro-mise between the improvement of the internal characteristicsof theinlet and engine and the increasein the externaldrag of the vehicle.Therefore, investigations of this kind o f wedge-aided diversion ofthe boundary layer should include an integrated analysis of its ef-fects not only on the characteristicsof the inlet itself but also on theoverall aerodynamic and aeropropulsive efciency of the vehicle.

Presently, particular questions are generally considered that arerelated to the effectsof the boundary layer and its diversion at highsupersonicvelocities on the characteristicsof the inlets themselves,outsideof their combining with an airplane. Some results of an ex-perimental study (within the range M1D 2 5) of these effects fora two-dimensional (at) inlet mounted under a delta wing are pre-

sented in Refs. 1 and 2. It is shown that the adverse effect of theboundarylayer developingon the wing surface is almost eliminatedby shiftingtheinletfromthe wing to a relativeheight h= D 0:50.6,where is the boundary-layerthickness ahead of the inlet. In thiscase, the ow-rate and the pressure recovery factors of the inlet are7595%of those valuesthatcan be reachedwith complete diversionof the boundarylayer.Measurementsof the boundarylayer develop-ing on forebody precompression surfaces upstream of an inlet, forhypersonic vehicles of various congurations, were performed in

Ref. 3 for M1 D 4 and in Refs. 4 and 5 for M1D 6. It was found4;5that the boundary-layerthickness in the plane of symmetry couldreach 20 100% of the height of the inlet entrance cross section, de-pending on the vehicle conguration. The state and characteristicsof the boundary layer developing on the compression wedge rampof at inlets mounted on diverter wedges was experimentally stud-ied in Ref. 1 for M1 D 25 and in Ref. 3 for M1 D 2 4. The owaroundvarioussuch wedges withdummy boxlike inlets mountedonthemand the drag of these wedges were studied in Refs. 6 and 7 forsupersonic velocities (M1 D 1:0 2.5). In particular, it was shown7that, for M1 D 2:25, diverter wedges with a swept leading edge ofblw 46 deg have a lower drag (by 15 20%) as compared to thewedge variant with blw D 0.

There is almost no literature dealing with effects of wedge-aided

diversion of the boundary layer on the total aerodynamic and aero-propulsive characteristicsof scramjet-poweredhypersonic vehiclesas a whole. Only Refs. 8 and 9 are recalled. In Ref. 8, estimates ofthe resultant aeropropulsive characteristics of a scramjet-poweredvehicle were obtained using approximate calculation methods. Itwas shown that boundary-layer diversion with the use of wedgesmay improve the resultant characteristics of the vehicle within thefreestream Mach number range from M1 4 to M1 D 7 10. Notethat the efciency of wedge-aided diversion of the boun dary layerdecreaseswith increasingM1. Goonkoet al.9 givesome experimen-talsupportto the integralapproachproposedin Ref. 8 for evaluationof the effect of boundary-layerdiversionensured by wedges on theperformanceof a hypersonic scramjet-powered vehicle as a whole.The vehicle conguration considered here was studied in Ref. 9,

and the main resultsobtainedare outlinedin the presentpaper in as-sociation with other factors of hypersonic inlet/airplane interactioninvestigatedb y the authors more recently.

In studying the characteristics and preliminary designing ofscramjet-powered hypersonic vehicles, two-dimensional or quasi-two-dimensionalsupersonic inlets with a multiwedge compressionramp are widely spread. For two-dimensional inlets, the designregime is usually identied with M1 D Md for which the planeoblique shock waves forming on the ramp wedges are focusing onthe leading edge of the inlet cowl lip. Such a design ow regimefor a at inlet really may be achieved if the shock wave forming onthe rst wedge of the ramp is aligned with the plane of the sweptleadingedgesof the side cheeksrestrictingthe design ow regioninthe lateral directions. In off-design regimes for the inlets designed

in such a manner, for instance, at Mach numbers M1 < Md, theshocks come out above these leading edges of the cheeks, and theow over the compression ramp is no longer strictly two dimen-sional because of lateral ow spillage. Similar lateral spillage alsoarises if the cheeks restrict the compression ow region near theramp only partly (short cheeks) or if there are no cheeks.

The effects of three dimensionality of the ow around a multi-wedge compression ramp on the characteristics of at inlets, withcheeks of various shapes and without them, were studied, for ex-ample, in Refs. 2 and 10 12. It was noted2 that, for at inlets withthe width-to-height ratio binl= h inl 4, the lateral spillage over thecompressionramp weakly affects the ow-ratefactor depending onthe conguration of the cheeks. At the same time, the numericalcomputations10;11 show that, in ow around the inlet with shortside

cheeks, internal shock waves form on the latter because of slip de-ectionof the ow near the compression ramp. Investigationshavenot as yet been carried outof possible effects of lateral owspillagenear the compression ramp on the characteristics of the at inletscombined with the airplane and on the characteristicsof the vehicleas a whole.


3/14

GOONKO AND MAZHUL 39

The ight of vehicles at high hypersonic speeds is accompaniedby intense aerodynamic heating. Therefore, the nose tips and theleading edges of the forebody and other elements of these vehi-cles should be blunted. The bluntness dimensions are usually smallrelative to the vehicle dimensions, but strong normal-curvedshockwaves formovera bluntednose or leadingedge.The inletowstruc-ture can signicantly changethe overall pattern of the external owaround the vehicle, especiallyfor vehicles powered by airbreathingengines, as compared to equivalent pointed bodies. A high-entropy

layer can form near the surface of such a blunted body, which ischaracterized by high nonuniformity and low Reynolds numbersfor the layer ow. Under these conditions, the development of theboundarylayer, its integral characteristics,and its effect on the inletperformance may also signicantly differ from those for a sharpconguration.

The effects of blunted leading edges of the compression wedgeramp, side cheeks,andcowllip onthe performanceof a atinletwasexperimentally studied for M1D 4 5.5 in Ref. 13. It is shown thatthe small bluntnessof the cowl lip and side cheekshas a weak effecton the ow-rate factor of the inlet. In contrast to that, the bluntnessof the leading edge of the ramp wedge has a signicant effect. For arather large bluntness (with a radius Rb 0:025 hinl /, the operatingregime of a started inlet may be destroyed. The shape of leading-

edgebluntnessof theramp wedgealso exertsa signicanteffect.Theeffectof bluntedleadingedgesof a delta wing on the characteristicsof an inlet mounted under this wing was studied experimentallyin Ref. 13 and numerically, for an inviscid ow, in Ref. 14. Anincrease in the wing bluntnessradius is accompaniedby a decreasein the inlet ow-rate factor13;14 and can also lead to ow stallingahead of the inlet entrance.13 The adverse effect of wing bluntnessdecreases13 with increasing angle of attack and sweep angle of thewing or with moving the inlet away from the wing surface, thatis, in the case where the boundary layer is diverted partially orcompletely.An analysis of the drag producedby the bluntedleadingedges themselvesas applied to supersonicand hypersonicinlets canbe found in Ref. 15. The effects of nose bluntness of the forebodyon the characteristics of the inlet, engine, and vehicle as a whole

have not been adequatelystudiedyet; this is particularlytrue for thetotal aeropropulsivecharacteristics.

The objectiveof the presentedwork was an experimentalstudy ofthe inuence of the mentioned factors of inlet/airplane interaction(boundary-layer diversion, lateral ow spillage over the compres-sion wedge ramp of the inlet, and nose bluntness of the forebody )on the characteristics of scramjet-powered hypersonic vehicles. Incontrast to many cited papers dealing with particular aspects, anattempt was made to perform an integrated analysis of these fac-tors extending the works.8;9 We consider their effects, rst, on thecharacteristics of the inlet itself, taking into account its integrationwith the lifting body and, second, on the total aerodynamic andaeropropulsivecharacteristicsof the vehicle as a whole.

Fig. 1 General view of the tested model of a ramjet-powered vehicle.

Model and Test ConditionsThe experimentalstudy was performedwith a schematizedmodel

representinga scramjet-poweredvehicle,which consists of a liftingbody (fuselage), a delta wing, and a scramjet engine nacelle locatedin a ventral position under the lower surface of the body (Fig. 1).The body and the wing form a total lifting delta conguration withleading edges of sweep D 80 deg. The lifting body has a trape-zoidalcross section of its forebody transformedto a rectangularoneat the tail of its rearbody. A lower triangular plane surface of the

forebody is a precompression surface of an airstream captured bythe inlet. The wing has a lower plane surface and a wedge-shapedairfoil of angle w D 3 deg. The overall dimensions (in millimeters)of the model are given in Fig. 1. The model conguration consid-ered is characterized by the following parameters: reference planarea of the model Spl D 0:0936 m2, aspect ratio of the overall lift-ing planform surface D 0:708, relative planform area of the wingSw =Spl D 0:305, relative area of the base cross section of the liftingbody together with the engine nacelle Alb =Spl D 0:084, and relativearea of the midsection of the airplane including the wing and theengine nacelle Am =Spl D 0:09. The body has a replaceable nose tip,which can be pointed or three dimensionally blunted; the blunt-ness radii are Rb D 7 mm in the plan view (2Rb =blb D 0:122, whereblb

D115 mm is the lifting body width) and Rb

D2:5 mm in the

vertical plane of symmetry (2Rb = h lb D 0:124, where hlb D 40 mmis the lifting body height).

The overall dimensionsof the model are made with tolerancesnomore than 0.7 mm. The inlet tolerances are 0.1 mm in length,0.07mminwidth,and0.03mm in throatheight;deviationsof theramp wedgesare no more than17 minute of an arc.The roughnessof external aerodynamic surfaces of the model and compressionsurfaces of the inlet is not worse than R zD 20 m and, apparently,R zD 2:5 m.

The general arrangement of the engine nacelle and the modelduct is shown in Fig. 2. The inlet has a triple-shock wedge rampof external compression with wedge angles 1 D 7:5, 2 D 15, and3 D 22:5deg.ForadesignMachnumberMdD 6,whichreferstotheinlet proper, plane shock waves arisingon the ramp wedges should

intersecton theleadingedgeof theinletcowl.The inlet cowlis madewithout bending its lip. The frontal area of the inlet is A0 D binlh inl,where binl = hinl 1:91 and binl D 68:3 mm and hinl D 35:6 mm arethe height and width of the inlet by its leading edges includingthose of the wedge ramp, the side cheeks, and the cowl. Its relativevalueis NA0 D0:026. Theinlethas thesimplestgeometry of the inputsection of the internal duct with a relative entrance area and throatarea NAen NAth 0:2. The inlet throat is rather short with a relativelength of about twice its height and with a small are of an angleabout 2 deg, downstream of which there is a 17-deg diffuser. Thegeneral arrangement of the model duct is made by requirementsfor aerodynamic testing the ducted models representing vehiclespowered by airbreathingengines.These requirementsare presented


4/14


Fig. 2 General arrangement of the enginenacelle andthe model duct.

in Ref. 16. Damping screensare mountedin the plenumchamber ofthe model duct with a relative area NAch 7:2. Four narrow meteringnozzles (plugs) are mounted at the model duct exit (N

nz D4). Their

diameteris chosenunderthe conditionthatthe inletshouldbe startedwithin the entire range of freestreamparametersexamined and thatthe rear normal shock arising in the internal duct ow should belocated in the inlet diffuser. The experimental data presented wereobtained for dnz D 15:6 mm, and the total relative exit area of themetering nozzles is NAnz D 1:62. A pitot rake for aerometry of owparameters in the metering nozzles is installed at the nacelle ductexit (Fig. 2).

The model is equipped with two replaceable engine nacelles,which differs by the presence or absence of the side cheeks (walls)

along the inlet wedge ramp, but the geometric parameters of theducts and the main lines of the external contour were identical(Fig. 1). In tests with boundary-layerdiversion,one or the other na-

cellecan be mounted on two symmetricallylocateddiverterwedgesof height hblw. The geometry of these wedges is shown in Fig. 2.The wedges of height hblw D 2 and 3.7 mm and hblw D 3 and 5 mmwere used in experiments for M1D 4 and 6, respectively.

The model was tested in the supersonic blowdown wind tun-nel T-313 (Ref. 17) based at Institute of Theoretical and AppliedMechanics (Novosibirsk,Russia). This wind tunnel has a rectangu-lar test section of 0:6 0:6 m transverse dimensions and of 2.5 mlength; interchangeable nozzle inserts provide ow Mach numbersin the range M1 D 26. It is equipped with a staff four-componentmechanical-typebalance,multichanneltransducerforpressuremea-surement, and schlieren shadowgraph device. The instrumental er-ror of balance measurement is less than 0.1% of the upper limit ofthe balance ranges. Respectively, in testing the model, the e rror of

onefold measurement for the drag was 0.7

0.2% for M1 D 4 and1.1 0.4% for M1 D 6; the error of liftforce measurementwas evensmaller. The pressure transducer has the ranges of 0 105 N/m2 and06 105 N/m2 with a measurement error of less than 0.3% of theupper limit of these ranges.

For the modelconsidered,the testconditionsincludedthe follow-ing freestream parameters: M1 D 4:05, Re 11 D 50 106 1/m, andq1 D 7:5 104 N/m2 and M1 D 6:06, Re11 D 17:7 106 1/m, andq1 D 1:2 104 N/m2 . The stagnation temperature in the plenumchamber of the wind tunnel was t01 286 K for the both Machnumbers.

The characteristicsof thestream capturedby the inlet andpassingthrough the nacelle duct were measured at the duct exit using themetering nozzles and the pitot rake mentioned earlier. This aerom-

etry was carriedout simultaneouslywith the balancemeasurement;thus, the pitot rake did not affect the latter. These aerometric datawere usedto determinetheow-rate factorof theinletand theforcesover the internal surfaces of the duct. The latter were not simulatedfor a full-scale vehicle and were excluded from the resultant aero-dynamic forces measured with the balance. Based on the data ofRefs. 16 and 18, the total instrumental errors occurred in tests of

ducted models in T-313 with the aerometric technique consideredare as follows: the inlet ow-ratefactor 14%, the drag at near-zeroangles of attack about3:0% for M1 D 4 and8:7% for M1 D6,and the maximal lift-to-drag ratio 1:3% for M1 D 4 and 3:8%for M1 D 6.

Notethat the aerodynamiccharacteristicspresentedfor the modelconsidered includealso the base drag over that part of the rear faceof the lifting body and engine nacelle that is not occupied by themetering nozzles. There is some base drag of the wing, which also

has a rear face because of a wedge-shapedprole. Estimates for thebase drag of the airplane and engine nacelle yield6% for M1D 4and 3% for M1 D 6.

All of the tests were performed with natural development of theboundarylayer on the model surfaces. As evidenced by an analysisof velocity proles in the boundary layer measured on the modelconguration without nose bluntness for M1 D 4 and calculatedestimates of the transition length for M1 D 6, the boundary layerimmediately ahead of the inlet was turbulent.

Representation of Aerodynamic Forces of the Model

A technique for aerodynamic testing of ducted models of vehi-cles poweredby airbreathingengines,whichhas been usedbefore inRussia (see Ref. 16), was used for experimentation with the studied

model. The essential feature of this technique is that the externalowaroundthe vehicleand over the inlet is simulated,but theoper-ation of the engine and the exhaust jet are not considered. Internalnonsimulated forces acting in the model duct are determined andexcluded from the balance measurement results in this case; theyare found from the change in the momentum of the inlet airstreamcaptured by the inlet and passing through the duct. The ow-rateand momentum of this stream, which are required for this purpose,are determinedby measuringow parametersat the exit duct. Note,as this takes place, the ow-rate factor of the inlet integrated withthe airplane is determined.

Theow ratem1 oftheairstreamcapturedbytheinletandpassingthroughthe model duct is determinedfromthe mass owcontinuityandby theowratemeasuredattheductexitusingmeteringnozzles:

m1 DNnzX

1

mnz; mnz D km q .Mnz/ p0nz Anzp

t0nz(1)

Here

km Dq

[2=. C 1/].C 1/=. 1/ Rgis a dimensional coefcient, K1=2 s/m,

q.M/ D M..2=. C 1/f1 C [. 1/=2]M2g//.C 1/=2. 1/

is a gasdynamicfunctionof the reducedow rate, Anz D d2nz=4,andMnz , p0nz, and t0nz are the average values of the stream parametersin the metering nozzle. In testing the model under the freestream

parametersconsidered,a critical chokedow occurred at the modelduct exit, and the ow in a narrow metering nozzle was close tosonic. In accordance with this, the local total pressure in the nozzlecan be taken equal to the pitot pressure measured.Note, in a generalcase of a supersonicow, the totalpressurebehind the normal shockis measured with a pitot tube. The average total pressure p0nz wasdetermined by averaging over the area the pitot pressures measuredwith the rake in every nozzle. It was also assumed that Mnz = 1.The ow in the model duct was assumed to be adiabatic, and thethermophysical properties of air were assumed to be constant, thatis, t0nz D constD t01 and D constD 1:4.

The ow-rate factor of the inlet is found from the ow-rateequation (1) as

f D A1A0

DPNnz

1 p0nz q.Mnz/ Anzp01 q.M1/ A0

(2)

The total aerodynamicforces obtained in testinga ductedmodel aredetermined in the form

Ra D Rbal Rduct Rsupp (3)


5/14


whereRa is the vector of the resultant aerodynamic force to be de-termined;Rbal is the vectorof theresultant forceactingon themodeland measured by the balance;Rduct is the vector of the excluded re-sultantforce of the duct, the componentsofRduct are usuallycalledduct-effectcorrections;and Rsupp is thevector of theforcerepresent-ing some methodical correction related to the manner of installingthe model on a support of the aerodynamic balance. The last termwas determined by a staff procedure accepted for the wind tunnelT-313 (Ref. 17) and is not considered here.

The forces acting in the model duct are found from the changeinthe momentum of the inlet-capturedairstream with considerationofthis stream portion between the cross section in the freestreamowto the duct-exitcross section. The excluded forces contain a part ofduct forces that are not simulated for a full-scale vehicle and a partof modeled forces. Differentforms of decompositionof theseforcesand, correspondingly,determinationof internal nonsimulatedforcesare possible.16 Hence, the aerodynamic characteristics of the sameducted model obtained in various ways may differ. In the presentedwork, we used the approach of Ref. 19 (discussed also in Ref. 16).According to Ref. 19, the forces involved by the deection of themomentum vector of the inlet-captured airstream passing throughthe model duct refer to modeled forces and are not excluded. Theforces arising as a result of the change in the magnitude of this

momentum refer to nonsimulated forces and should be excludedfrom the balance measurementresults. In accordancewith Ref. 19,the duct-effectcorrectionshould be written as

Rduct D .Inz I1/ nnz

Inz DNnzX

1

[m nzVnz C .pnz p1/Anz]

DNnzX

1

pnzM

2nzC .pnz p1/Anz

I1

Dm1V1

Dp1M

2

1f A0 (4)

where nnz is the unit-directing vector of the axis of the meteringnozzles, which is assumed to coincide with the vector neng of theengine axis, and Vnz and pnz are the velocity and static pressure ofthe duct stream in a metering nozzle.

The aerodynamic characteristics of the model, which were ob-tained from the test results taking into account the duct-effect cor-rection in the form of Eq. (4), have the following expansion:

Ra D Rext CRss CI1 .nnz n1/ (5)whereRext is the vector of the resultant force over the external sur-faces of theairplaneand the engine nacelle,whichare not wettedbythe airstream captured by the inlet. As stated earlier, this resultantforce,in the aerodynamicdata presented,includesthe forceRbase of

base pressureactingover that part of the rear face of the liftingbodyandenginenacellethatis notoccupiedby the metering nozzles.Thereciprocalof the x componentRbase is usually called base drag. Thevector Rss is the resultant force over a so-called nonsolid stream-surface of the inlet. This streamsurface is formed by streamlines ofthe inlet-captured airstream passing through the leading edges ofthe inlet and corresponds to a portion of this stream between thefreestream cross section and the inlet edges. The drag componentof the Rss force is also known as ram or preentry drag. The vectorI1.nnz n1/ refers to deection-causedforces mentioned earlier;they are a result of deection of the airstream captured by the inletand passing through the duct from the freestream direction n1 tothe direction nnz . Note, the resultant aerodynamic force [Eq. (5)]does not include forces acting on the forebody surface wetted by

the inlet-capturedairstream, that is, the resultant aerodynamicdragof the ducted model does not include viscous and wave drag fromthe forebody surface compressing the inlet ow.

Thus, the characteristics of the inlet integrated in the vehiclesystem, Eqs. (1) and (2), and the total aerodynamic characteris-tics, Eqs. (3 5), were obtained from testing the ducted model underconsideration.

Determination of the ResultantAeropropulsive Characteristics

In this work, apart from consideration of the aerodynamic char-acteristics already mentioned, we performed a more detailed anal-ysis of the properties of the examined scramjet-powered vehicle asa whole, extending the approach,9 which implies consideration ofall forces acting on the vehicle including the propulsion unit. Forthis purpose, estimates of the resultantaeropropulsiveforces for thetest Mach numbers were obtained using experimentallydeterminedaerodynamic characteristics and ow-rate factor, as well as enginethrust characteristics calculated on the basis of one-dimensionaltreatment of the ow with heat addition from fuel burning in theengine combustion chamber.

The aerodynamic characteristics of ducted models obtained ex-perimentally should be incorporated into the total forces acting onthe vehicle,including all of the externaland internal surfaces of theairplane and propulsion unit, in the form

R D Ra C T (6)

where R is the vector of the resultant aeropropulsiveforce over allexternal and internal surfaces of the airplane and engine nacelle.The lift and drag components of the resultant aerodynamic force

Ra for a full-scale vehicle should be determined through nondi-mensional aerodynamic coefcients obtained experimentally. Theengine thrust vector T should be determined in a different manner,dependingon the method of duct force decompositionas mentionedearlier. When experimental determination [Eq. (5)] of aerodynamicforces correspondingto Ref. 19 are taken into account, the thrustvector in Eq. (6) should be represented in the following form:

TD .Inze I1/ nengInze D [mnzeVnze C .pnze p1/Anze m1V1] (7)

where mnze , Vnze , and pnze are the mass ow rate, velocity, andstatic pressure of the exhausted jet and Anze is the exit area of the

nozzle of a full-scale engine whose thrust vector is directed alongthe axis neng common to the nozzle and engine. Note that the exitarea of the engine nozzle of the vehicle considered refers to thetotal area of the base face of the lifting body and engine nacelle;thus, this relative area is Anze=A0 D 3:78 4.02 depending on theheight of the diverter wedges. The ow-rate characteristics of theengine, that is, the quantity m1 in Eq. (7), should be determinedtakinginto accounttheow-ratefactorof theinlet[Eq.(2)] obtainedexperimentally;the quantity m nze follows from adding to m1 somemass of fuel supplied into the engine combustor.

The denitions and decomposition of forces used are shown inFig. 3, which shows schematics of a ducted model in a longitudinal

a)

b)

Fig. 3 Schematics of forces acting on the ducted modeland airbreath-ing vehicle.


6/14


section (Fig. 3a) and an appropriate vector diagram (Fig. 3b). Thesameconsiderationscould be appliedto aerodynamicand thrustmo-ments of the model and full-scale vehicle, but they are not involvedhere.

To calculatethe thrustcharacteristicsof the engine,we estimatedthe parameters of a one-dimensional ow in the inlet throat. TheMach number Mth was obtained from the equation of mass owcontinuity using experimental values of the ow-rate factor,

q .Mth / D q.M1/ fNAth th(8)

where th = p0th=p01 is the total-pressure recovery factor of thethroat ow, th Den int , where en D p0en=p01 is the total-pressurerecovery factor of the inlet-capturedairstream in the entrance crosssection and int D p0th=p0en is the total-pressurerecovery factor forthe internal ow in the inlet input section between the entrance andthroat cross sections.

The values of en were determined by calculation of the invis-cid ow over the compression surfaces of the forebody and theinlet. The ow parameters near the triangular precompression sur-face of the forebody were found using an approximate method,20

which presents an adaptationof the well-known tangent-wedge and

tangent-conemethods as applied to triangular wing-shapedbodies.In this method, the ow near a triangular at surface is assumed tobe uniform, and its parameters are determined by interpolation be-tweentwo values calculatedfor tangentwedgeand cone,with takinginto account sweep and incidence angles of the surface. Analyticalformulas were used: exact for the wedge21 and approximate for thecone.22 The ow over the inlet wedge ramp and shock waves on itswedgeswere assumedto be planar.The geometry of the model inletentrance is simple, so that the total pressure loss in the entrancesection is mainly produced by an oblique shock wave arising onthe cowl lip, where a supersonic ow at the entrance deects fromits direction along the wedge ramp (with an angle 3 D 22:5 deg)to the direction in the throat (th D 0 deg). Consequently, the valueint D sw D p0sw=p0en was taken.

The change in the total-pressure recovery factor of the inlet-captured airstream includes some losses causedthe boundarylayer.In accordancewith Ref. 1, if there is no shock-inducedseparationofthe boundarylayerat theinlet entrance,these losses canbe estimatedby the value bl D .1 Aen =A0/, where Aen is the area covered bythe boundary-layerdisplacement thickness in the inlet-entrancecross section. The value bl was not obtained in the experimentsand calculations, but these losses for a full-scale vehicle are smallin comparison with the total pressure losses in the shock waves es-timated by the value sw dened earlier, and they were neglectedin thrust estimations. The remaining parameters of the inlet throatow are evident from known Mth and th , and the engine thrust canbe calculated.

Thus, the change in the total-pressurerecovery factor of the inlet

caused by the examined factors (wedge-diversioneffects, presenceor absence of the side cheeks, and nose bluntness) is not taken intoaccount in estimating the throat ow parameters. Correspondingly,the calculated change in the thrust coefcient depending on thesefactorsis related mainly to thechange in theow-ratefactorobtainedexperimentally.

In accordancewith Eqs. (6) and (7), the coefcients of tangential(directed against the freestream ow) CRx and lifting CRy compo-nents of the resultant aeropropulsiveforceR are determined as

CRx D CT cos./ A0=Spl CDCRy D CT sin./ A0=Spl CL (9)

where CD and CL are coefcients of the drag D DRax and liftL D Ray components of the aerodynamic force Ra determined ex-perimentallyin theform of Eq. (3) CT is thecoefcient of theenginethrust value jTj determined in accordancewith Eq. (7).

Determining the aeropropulsive characteristics of the examinedvehicle conguration as a whole means actually going from themodel to a full-scale vehicle, and it should include extrapolation

of experimental aerodynamic characteristics to ight conditions.It could be suggested that the change in these characteristics ismainly determined by the change in friction drag with increas-ing the Reynolds numbers to ight values; thus, the overall dragcoefcient will decrease and improve the aerodynamic character-istics. It is difcult to separate the wave and friction drag in ex-perimental data obtained; therefore, the said extrapolation was notperformed.

A scramjet operation regime with a supersonic velocity in the

combustion chamber was examined for the propulsion unit of thefull-scalevehicle considered.In determiningthethermogasdynamicand thrust characteristics of the scramjet, we used a simplied ap-proach whose brief descriptionand main assumptions can be foundin Refs. 23 and 24. The approach was based on Refs. 25 28. FlightMach numbers corresponded to the test Mach numbers, and theother ightconditionswere consideredcorrespondingto a dynamicpressureq1 D 7 104 N/m2 (p1D 6:25 103 N/m2, t1 = 216.65Kfor M1 D 4 and p1 D 2:78 103 N/m2, t1 D 221 K for M1 D 6).The relative throat area NAth D 0:12 was preset for an inlet of thefull-scale vehicle, and the throat ow parameters were calculatedfrom Eq. (8). The scramjet operation mode with supersonic com-bustion and with fuel supply corresponding to the air-to-fuel ratioaf

D1 was assumed. An initial fuel portion was supplied in the

combustion section of constantcross-sectionalarea Ach =Ath D 1:5;the possible amount of injected fuel was determined from the con-dition of reaching the critical velocity Mch D 1:0 at the exit ofthis section. An additional fuel portion was supplied in the nextaredcombustion section to reachthe given air-to-fuel ratio afD1.The condition Mch D constD 1:0 was held for this process. The en-gine jet expanded adiabatically in an exhaust nozzle from a cross-sectional area corresponding to the fuel supply completion up tothe nozzle exitcross-sectionalarea Anze =A0 D 3:78 4.02 mentionedearlier.

A caloric value Hu D 1:3 108 J/kg and a stoichiometric massair-to-fuel ratio L 0 D 34:25 were taken for the fuel (gaseoushydro-gen). The specic gas constant of combustion products was de-termined as Rcp

D.Rair m

1 CRfuel m fuel/=.m

1 Cmfuel/. The heat

capacity of the combustion products was assumed to be constantand characterized by the effective ratio of specic heats cp = 1.26accepted by Ref. 25. The combustion efciency and other coef-cients that specify losses of different kinds in the combustor and inthe nozzle were assumed to be equal to unity, that is, the thrust esti-matesobtainedcorrespond,in a certainsense,to idealcharacteristicsof the scramjet.

Note that the contribution of the engine thrust to the lift forceof the vehicle is rather small despite its reasonably large motor-ization factor. The latter can be characterized by the relative cross-sectionalareasof theinputand outputunits of theengine,of theinletA0=Spl D 0:026 and ofthe nozzle Anze =Spl 0:1: This is appreciablyexplained by the engine thrust being aligned with the constructionaxis of the engine and the vehicle by presumption, that is, the lift

and tangential thrust components are merely CTy DCT sin./ andCTx DCT cos./.

Boundary-Layer Diversion

Effects of boundary-layerdiversion were experimentallystudiedfor a conguration of the model on which the pointed nose of thelifting body and the engine nacelle variant with the noncheekedinlet were mounted. The schlieren pictures of the ow pattern overthe external compressionsection of the inlet were obtainedin theseexperiments. They showed that, within almost the entire range ofthe angle of attack under study, a supersonic inow at the inletduct entrance occurs for the model variant without diverter wedgesboth for M1 D 6 and for M1 D 4. A shock wave detached fromthe leading edge of the inlet cowl was observed for angles of attack > 9 degin tests of themodelvariantwith wedges hblw D 3:7 m m a tM1 D 4. Thus,regimesof theowaroundthe inletwitha supersonicinowat theentranceanda supersonicow in thethroatwere mainlyobtained in the experiments.

To c hoose properly the required heights of diverter wedges, oneshould know the boundary-layerparameters ahead of the inlet. The


7/14


state of the boundary layer developedon the forebodycompressionsurfacewas estimatedon thebasisof transitiondataobtainedby V. I.Kornilov (unpublished,privatecommunication) foraatplatetestedin the T-313 wind tunnel for M1 D 2:56 and different Reynoldsnumbers per meter. The transition Reynoldsnumbers obtainedobeythe power dependence Re tr Ren11 with an exponent n 0:4 ex-tended for the T-313 wind tunnel in Ref. 29 by data for M1D 3and 4. The effect of the sweep angle was taken into account bydata30 for triangular at wings also tested in T-313 at M

1 D3 and

4. According to these data, the relationship .Re tr/ 6D 0=.Retr/ D 0changes little as compared between M1 D 3 and 4, and the saidratio determined for M1D 4 was extended to M1 D 6. For exper-imental conditions under consideration, D 0, the following val-ues of the Reynolds number and the length of the boundary-layertransition on the nose swept compression surface were estimated:M1 D 4, Re tr D 3:8 5.7 106, and L tr D 1929 mm, and M1 D 6,Re tr D 5:98.6 106, and L tr D 83 120 mm. Thus, a developedtur-bulentboundarylayerwas expectedimmediatelyin frontof the inletwedge,which correspondsto ight ow conditionsof full-scalehy-personic aircraft.

For the model under consideration, the characteristics of theboundarylayeron the precompressionsurface of the forebodywereexperimentally dertermined9 in the plane of symmetry z

D0 at a

distance x D 340 mm from the nose tip of the model. The staticpressure on the wall and the pitot-pressure prole in the bound-ary layer were measured with a small-size probe positioned on atraversing gear. The velocity distribution in the boundary-layer,theboundary-layerthickness , the displacement thickness , and themomentum thickness were obtained from these data. There-with, Croccos integral was assumed to be valid for the descriptionof temperature variation across the boundary layer. The boundarylayeron the modelsurfacesdevelopednaturallyundernear adiabatictemperatureof the model wall tw t01 295K. The freestreampa-rameterscorrespondedtoMach numbersM1D 2:03,3.03,and4.05(Re11 D 27 106, 3 7 106, and 56 106 1/m, respectively).

Some integral characteristics of the boundary layer on the fore-body precompression surface, which will be used subsequently,

were estimated under the assumptionthat this boundarylayeris twodimensional and fully turbulent. We used the method of Ref. 31developed as applied to the turbulent boundary layer of a non-thermoinsulated at plate. This method presents an analytical for-mulation based on the logarithmic law for the velocityprole of thecompressible turbulent boundary layer. Integral parameters of theboundary layer such as the thickness, displacement thickness, andmomentum thickness are determined. In due course, Kovalenko31

has demonstrated that the analyticalmethod yieldssomewhat betteraccuracyin comparisonwith thewell-knownsemi-empiricalpredic-tion procedure of Ref. 32. Note, the ow about the model forebodyis, in general, three dimensional. In addition, the boundary layerover the model is really transitional, but the lengths of laminar andtransitional sections of it were not determined. Thus, the said es-

timates cannot be rigorously correlated with the experimental dataobtained, but they give some reference estimates. In particular,theyallowed the choice of diverter wedge heights.

As is shown in Ref. 9 by analysis of the measured velocity pro-les, the boundary layer ahead of the inlet is turbulent in the entirerange of experimental freestream parameters considered. The inte-gral characteristics of the boundary layer are plotted in Fig. 4 asfunctions ofM1 . It is evident from these data that the shape factorat M1 D 4 is about Hbl D 7:8 7.9, which corresponds to a devel-oped turbulent boundary layer. A calculated estimate for M1D 4yields Hbl 8:1, which conforms to the experimental value. Notethat the difference in the values of , , and measured foran angle of attack D 3 deg does not exceed 6% in comparisonwith

D0. However, for

D10 deg, this difference increases ap-

preciably and, for example, reaches 35% for the displacementthickness .

According to the measurement data, the boundary-layer thick-ness ahead of the inlet is 3:9 mm for M1 D 4 and D 0, therelative values are c orrespondingly hblw = 0:51 for hblw D 2 mmand hblw = 0:95 for hblw D 3:7 mm. For M1 D 6, the calculated

Fig. 4 Integral character-istics o f the boundary layermeasured on the forebody

precompression surface as

functions ofM1 .

Fig. 5 Flow-rate factor of the model inlet vs the angle of attack for aset of diverter wedge heights.

boundary-layerthickness ahead of the inlet could reach

5 mm,

which was conrmed by the schlieren pictures of the ow patternaroundthe model inlet.Hence, for the chosen heights of the diverterwedges, their expected relative heights for this Mach number areabout hblw= 0:6 for hblw = 3 mm and hblw = 1 for hblw D 5 mm.Taking into account the high-power character of velocity distribu-tion in the turbulentboundary layer, we may assume that the valueshblw= 0:50.6 obtained for the model will allow elimination ofthe inuence of its most stagnated part and a tangible increase inthe inlet efciency.

Letus considerthe effectof wedgediversionof theboundarylayeron the ow-rate characteristicsof the inlet. The experimentalvaluesof the ow-rate factor f are plotted in Fig. 5 for Mach numbersM1 D 4 and 6 as functions of the angle of attack. The data for theexamined values of the diverter wedge height hblw are presented.

As evident from these data, an increase in the relative height of thediverterwedgesabovehblw= 0:5 0.6 weaklyaffects the ow-ratefactor.The schlierenpicturesshow that,for M1 D4, hblw D3:7 mmand M1D 6, hblw D 5 mm, thenear-wallstagnatedpartofthe bound-ary layer developedon the forebody surface passes through the slotfor boundary-layerdiversion.That is, the ow aroundthe inletrampis affected only by the external, dynamic part of the viscous layer.Under these conditions, a new boundary layer is developed on theinlet wedge ramp. The thickness of this new boundary layer at theinlet entrance cross section is much smaller than that occurringforthe model variant without the diverter wedges. For instance, basedon the data of the schlieren pictures, we can say that the thicknessof the boundary layer in the inlet entrance cross section for a Machnumber M

1 D6 in the absenceof its diversionis more than 50% of

the entranceheight;with thediverterwedgesof height hblw D 5 mm,this thickness decreases to about 25%.

The specic behavior of the ow-rate factor as a function of theangle of attack for > 6 deg in model testing with M1 D 4 andthe wedge height hblw D 3:7 mm (hblw= 1) should be noted (seeFig. 5). Despite that the boundary layer ahead of the inlet is almost


8/14


completely cut off, the ow-rate factor has lower values as com-pared to the wedge with hblw D 2 mm (hblw= 0:51). For > 14deg, this factor becomes almost equal to that for the variant with-out boundary-layerdiversion hblw D 0. The schlieren pictures showthat, in this range of ow parametersat the inlet entrance, there is ashockwave detachedfromthe leadingedgeof thecowl.Notethatthewedgeswith hblw D 3:7 mm provide the passageof practicallyall ofthe boundary layer developed on the forebodyunder the inlet ramp(hblw=

0:95). The calculatedestimates for an inviscidow in this

case also indicate the possibility of this detachment due to the ex-cess of thecriticalanglesof ow deectionon the cowl.At the sametime, no detached shock wave was observed in the case of wedgeswith hblw 2 mm, apparently, because of a certain reconstructionof the ow pattern with a rather thick boundary layer. As a whole,we can note that, up to angles of attack corresponding to the max-imum lift-to-drag ratio .L =D/max 6 deg, an increase in the wedgeheight above hblw= > 0:50.6 does not lead to a further signicantincrease in the ow-rate factor and does not seem reasonable. Forangles of attack close to .L =D/max , the ow-rate factor increasesby 5 and 11% for M1 D 4 and 6, respectively, which, obvi-ously, leads to an increase in the thrust produced by the propulsionunit.

The drag of thevehicleas a whole, in a conguration with wedges

for boundary-layer diversion, should increase througth the wavedrag and friction drag of the wedges themselves, friction drag ofadditional wetted surfaces in the slot for boundary-layerdiversionbetween the enginenacelleand the body, andalso interferencedrag.Obviously, for a ducted model with such wedges tested with the useof the experimental technique discussed earlier, the drag over theinletstreamsurfaceshouldalso changeas a consequenceof changingthe ow-rate factor as compared to its variant without wedges. Thedrag coefcient CD0 of the model as a function of the wedge heightis shown in Fig. 6. An increase in the value ofCD0 by 20 26% fora Mach number M1 D 4 and by 2732% for M1 D 6 is observedas compared to the case hblw D 0. These data show that the relativecontribution of the wedges to the drag somewhat increases withincreasing freestream Mach number.

An analysisof the liftingpropertiesof the hypersonicvehiclecon-gurationconsideredshowed 9 that, within the rangeof theangles ofattackD 012 deg, the presence of the diverter wedgesinvolved asmall increase in the lift coefcient CL for M1 D 6 and had practi-cally no effecton it for M1 D 4. The dependencesof the lift-to-dragratio L=D vs the angle of attack for M1 D 6 and for the studied

Fig. 6 Drag coefcient of the model at zero lift vs the relative diverterwedge height.

Fig. 7 Lift-to-drag ratio as a function of the angle of attack for a setof diverter wedge heights.

Fig. 8 Maximal lift-to-dragratiovs therelativediverter wedge height.

Fig. 9 Engine thrust coefcient vs the angle of attack forM1 =6 andfor a set of diverter wedge heights.

set of values ofhblw and the maximal lift-to-dragratio (L =D/max vsthe wedge height hblw are plotted in Figs. 7 and 8, respectively.Thedecrease in (L=D/max caused by the presence of diverter wedgesamounts up to 12% as compared to the variant without boundary-layer diversion.

The discussedexperimentaldatademonstratethat,despitethe im-

provement of the ow-rate characteristics of the inlet, the presenceof wedges for boundary-layer diversion leads to a noticeable in-crease in the drag of thevehicle as a whole and, hence,to a decreasein the maximal lift-to-drag ratio. At the same time, it is obviousthatthesedata areinsufcient for making the nal conclusionaboutthe efciency of boundary-layer diversion. Analysis of all of theforces acting on the vehicle including the propulsionunit is neededto evaluate such a global efciency. As has been mentioned, suchan analysis for the hypersonic vehicle under study was carried outusing estimates of the resultant aeropropulsiveforces, which wereobtained by combining the aerodynamiccharacteristicsdeterminedexperimentally and the thrust characteristicsof the scramjet enginecalculated by approximate methods.

As has been described, calculation of the engine thrust proceeds

from ow parameters in the inlet throat, which mainly depend onthe ow-rate factor obtained experimentally. By calculation, thepressure recovery factor and Mach number in the throat are th D0:73 0.76 with Mth D 2:25 1.75 at M1 D 4 and th D 0:51 0.56with Mth D 3:252.52at M1 D 6. Thus,the pressurerecoveryfactorvaries little with the angle of attack.

The engine thrust coefcient for M1 D 6 is presented in Fig. 9vs the angle of attack CT(). For the conguration with diverterwedges hblw D 5 mm (correspondingto hblw= 1), the thrust co-efcient increases by 9 11% as compared to the variant hblw D0.This increase is caused by the increasein theow-rate factor, whichresults from the mentioned simplied determinationof the ow pa-rameters in the inlet throat and, respectively, the engine thrust. Theestimated effect of boundary-layerdiversion on the scramjet thrust

is in agreement with the data of Ref. 8, where the same effect wascalculatedtaking intoaccount the displacingactionof the boundarylayer.

The resultant aeropropulsive characteristics for M1 D 6 are il-lustrated in the form of polars: dependences CRy .CRx / plotted inFig. 10. The aerodynamic polars CL .CD / are also presented inFig. 10. The greatest relative increment in the drag caused by the


9/14


Fig. 10 Polars of aerodynamic forces CL(CD ) and resultant a ero-

propulsive forces CRy (CRx ) for M1 = 6 and for a set of diverter wedgeheights.

diverter wedges is observed near D 0; therewith the resultant tan-gential force remains virtually unchanged. A signicant incrementin the coefcientCRx of this forceoccurs for > 6deg (CRy > 0:06),and it is almost the same for h

blw D3 mm (h

blw=D

0:6) andhblw D 5 mm (hblw= D 1:0).

It is obvious that this positive effect can be enhanced by opti-mization of the wedges for boundary-layerdiversion. For instance,the experimental data presented allow one to estimate the resultantcharacteristics corresponding to zero drag of the diverter wedges(CD /blw D 0, which in a certain sense corresponds to bleeding theboundary layer developed upstream of the inlet. The values of theow-rate factor for the greatest height of the wedges (hblw D 5 mmfor M1 D 6) and aerodynamic characteristics for the model vari-ant without the wedges (hblw D 0) can be used for this purpose. Adashed curve connecting the solid circles in Fig. 10 shows theseestimates. For this case, there is an increase in the resultant tan-gential force over the entire range of angles of attack examined.

However, note that these estimates can be considered only as alimit for the aeropropulsive characteristics that could be reacheddue to boundary-layer suction upstream of the inlet. A more rig-orous estimate of the efciency of such boundary-layer bleedingrequires the energy losses on suction to be taken into account, in-cludingthe effect of dumping the bled air back into the engine ductstream.

The following distinctive feature of the polars of the resultantaeropropulsiveforcespresentedin Fig. 10 shouldbe noted.The tan-gential force increaseswith increasingthe lift force or with chang-ing the angle of attack from D 0 to 6 deg, and thereafter itbegins to decrease. The initial increase in this force is a result ofthe signicant increase in the inlet ow-rate factor with increasingthe angle of attack; the engine thrust rises, correspondingly, ap-

proximately linearly(see Fig. 9

). This increase in the propulsiveforce exceeds that for aerodynamic drag. For > 6 deg, the fur-

ther increase in the inlet ow-rate factor and engine thrust doesnot compensate the drag rising approximately as the angle of at-tack squared. Corresponding to the behavior of CRx as a functionof CRy , two branches, respectively, the lower and the upper ones,can be distinguished,which differ by the aircraft response to a per-turbation of the angle of attack at a given ight Mach number. Forinstance, for ight values CRy corresponding to the upper branchof the polar, a perturbationincreasingthe angle of attack should beaccompanied by an ordinary response: transition to a deceleratedclimb. For the lower branch of the polar, such a perturbationshouldbe accompanied by the transitionfrom a steady horizontal ight ora steady climb ight to a spontaneouslyacceleratedclimb. This sort

of path instability was discussed in Refs. 23 and 24; obviously, itcorrelates with the conventional characteristics of aircraft such asthe speed stability and the stability related to the angle-of-attackload.

The calculated resultant aeropropulsive characteristics ofscramjet-powered aircraft presented in Refs. 23 and 24 show thatthe said feature of the resultant polar is representative of aircraft

Fig. 11 Polars of aerodynamic forces CL(CD) and resultant aero-propulsive forces CRy (CRx ) for M1 = 4 and for a set of diverter wedgeheights.

congurations with a scramjet module positioned under the liftingbody. Note that conventional ight conditions of hypersonic air-craft correspond to dynamic pressures q1

D5 8

104 N/m2 and

CRy D 0:05 0.1.In the case M1 D 6 considered,these ight param-eters will fall onto the lower unstable branch of the polar.

Related estimates of the resultant aeropropulsiveforces were ob-tained for M1 D 4 (see Fig. 11). They showed that there is scarcelyany effect of the diverter wedges on increasing the resultant tan-gential force in this case. Taking into accountthe approximationofenginethrustcalculations,one cansay thatthe wedges forboundary-layer diversion in this case at least do not worsen the aircraft per-formance. Also note that the resultant polars CRy .CRx / for M1D 4have no unstablebranch, that is, they are characterizedby the ordi-nary response to perturbations over the entire range of the angle ofattack considered.Such a change in the form of the resultantpolarsfor M1 D 4 as compared to M1 D 6 is also associated23 with pathinstabilityanalogous to that discussed earlier.

As a whole, the presented experimental results and calculationestimates on effectsof wedge-aided diversionof the boundary layershowed that mounting the inlet on such wedges of an appropriateheightallows reaching,at high Machnumbers, some increase in thetotal propulsive force of the vehicle despite the increase in the dragforce. This supports the conclusions8 obtained only on the basisof calculation estimates. To determine the benets of arrangementof diverter wedges on a certain hypersonicvehicle, obviously, it isnecessary to take into account the inuence of other factors, forexample, the angles of the wedges and their longitudinal contour,the sweep angles and the bluntness of their leading edge, etc. Someoptimization and improvement of the characteristics of the wedgesis alsopossible.Also bear in mind thatthe characteristicsofdevicesfor boundary-layerdiversion depend on the particular arrangement

of the propulsion unit on the vehicle. Therefore, it is desirable toperform more complete investigations of the inuence of variousfactors on the efciency of boundary-layer diversion for particularconditions.

Effect of Lateral Spillage over the CompressionWedge Ramp of the Inlet

A conguration of the model with a pointed nose and withoutdiverter wedges was used for experimentation to compare the en-gine nacelle variants with the cheeked and noncheeked inlets. Theeffects of lateral spillage for a two-dimensional inlet with or with-out side cheeks depending on M1 should be considered by takinginto account the design Mach number of the inlet Md. For Mach

numbers M1 < Md or for M1D Md and > 0, the cheeks onlypartially restrict the region of ow over the compression ramp be-hindthe shock wavesfrom the externalowon the sides of the inlet.As a consequence, lateral spillage is observed even in the presenceof the cheeks, which can lead to a change in the inlet characteris-tics, in particular, in the ow-rate factor. Obviously, the effects oflateral spillage should depend on the width-to-length ratio of the


10/14


inlet ramp related directly to the width-to-height ratio of the inletbinl= hinl.

Effects of lateral spillage were studied, for instance, in Ref. 2.A at inlet dened by MdD 5:3, a three-wedge ramp of 1 D 7:5,2 D 15, and 3 D 27:5 deg and binl= h inl 4 was considered; it waslocated under a delta wing. It was shownthat thepresenceof thesidecheeksincreasedthe ow-rate factor only by 1.5 2% for M1 Mdand by 2 3% for M1 > Md as compared with the noncheeked in-let variant. However, a more pronounc ed effect of the side cheeks

was revealed,10 for an inlet with MdD 4, a two-wedge ramp of1 D 12 and 2 D 22:8 deg, and binl =h inl D 1:33. For this inlet, thepresence of the side cheeks led to an increase in the ow-rate fac-tor, essentially at M1 close to the design Md. For M1 D 4D Mdand D 0, this increase accounted for about 22%. Similarly, anincrease of about 11% in the ow-rate factor caused by the pres-ence of the side cheeks was observed11 for an inlet with a one-wedge ramp of1 D 10 deg, binl=hinl D 1 in the design ow regimeM1 DMdD 6:1. Note that the data10;11 refer to two-dimensionalinlets nonintegrated with a forebody, that is, outside of their ar-rangement on an airplane. An inlet integrated with the airplane ex-periences the action of a three-dimensionalow from the forebody.For instance, despite that the lower surface of the model forebodyis freestream directed in the case

D0, the upper and side sur-

faces are inclined relative to the freestream, and the ow over themis compressed. There is some cross overow from the upper sideto the lower surface caused by this compression, which involves achange in ow parameters ahead of the inlet and leads, in partic-ular, to an off-design ow regime for it. The said cross overowacts in a direction opposite to the lateral spillage over the inletramp, so that one might expect that the effects of lateral spillagerelated to the presence or absence of the side cheeks in this casewill be manifested to a smaller extentas compared to nonintegratedinlets.

The experimental data obtained for M1 D 4 and 6 on the ow-rate factor of the inlet for the examined model conguration withpointed nose are plotted in Fig. 12, where the model variant withthe inlet cheeked engine nacelle is compared with the noncheeked

variant. The increment in the ow-rate factor 1f caused by theside cheeksdoes not exceed 1.2 2.8% for D 0 and decreases withincreasing angle of attack. Thus, the efciency of the side cheeksin restriction of lateral spillage of the ramp-compressedow corre-spondsto thatnotedpreviouslyin Ref. 2 for inletswith binl= h inl 4,although the relative width of the inlet for the model congura-tion tested is approximately two times smaller, binl= h inl D 1:91.The relative increment in the ow-rate factor 1f=f for our in-let even for M1 DMd is signicantly smaller than the valuesobtained10 for a nonintegratedinlet with binl= h inl D 1:33, for whicha substantial effect of the side cheeks in design ow regimes wasnoted.

In the contextof considerationof lateralspillage effectsat M1DMd, the question on the inlet design-ow pattern should be dis-

cussed. Note that this ow pattern is dened for the model inlet

Fig. 12 Flow-rate factor for cheeked and noncheeked inlets of the

model vs angle of attack.

proper under the assumption that the ow with the design Machnumber MdD 6 upstream of the inlet is uniform and the ow overthe inlet is two dimensional.It shouldoccur for the cheeked inlet ofthe model in the case ofM1D 6 and D 0, where the forebodypre-compressionsurface is located along the freestream ow. The inletow-rate factor should be unity f D 1 therewith. However, the ex-perimentalow-rate factor is signicantly smaller than unity (aboutf 0:65, Fig. 12).

As was noted earlier, there is some deviation of the ow pattern

from the design mode one for the inlet combined with the vehicleforebody. The ow over the precompressionsurface formed by theforebody is three dimensional in front of the inlet even for D0,and these ow parameters differ from those for the freestream owwith M1 D Md. It is oneof thefactorsaffectingthe said discrepancyof the design and real inlet ow-rate factors. However, numericalcomputations33 of the ow around the conguration considered,which were performed for M1 D 4 using an Euler code, show thatthe said ow three dimensionality rather weakly affects the owover the inlet itself and its ow-rate factor as compared to a purelytwo-dimensional ow. The values of the ow-rate factor computedfor a three -dimensional ow over the forebody/inlet congurationand for a purely two-dimensional one are virtually the same for

D0 and differ by no more than 3.7% for

D5 deg.

The second essential factor is the effect of boundary-layerdis-placement. As was already discussed for M1 D 6 and D 0, theboundary layer on the wedge ramp in the inlet entrance cross sec-tion occupies more than 50% (roughly 5070%) of the entranceheight. If the boundary layer at the entrance is turbulent, its dis-placementthickness,by estimation in accordancewith Ref. 31, willbe about13 18% of the entranceheight.A decreasein the ow-ratefactor caused by this displacing effect of the boundary layer is inproportion to the already mentioned area Aen as 1f=f Aen=Aen.The value Aen for the noncheeked inlet variant is determined onlyby the boundary layer on the wedge ramp; for the cheeked inlet itshould be greater because of the boundary layer developingon theside cheeks. The value of1f=f will be even somewhat greater fora certain slowdown of the inviscid ow out of the boundary layer

due to its displacement effect.The effects of ow three dimensionality and boundary-layerdis-

placement together explain to a large extent the already mentioneddiscrepancy in the ow-rate factor. The displacement effect of arather thick boundary layer may also enhance three dimensionalityof the ow over the inlet, notably for its noncheeked variants.

The change in the drag coefcient CD0 at a zero lifting force ac-cording to whether the side cheeks are present or absent has ratheran ambiguous character depending on the freestream Mach num-bers M1 under study. For example, for M1D 4, the drag coef-cient for the conguration variant with the inlet-cheeked enginenacelle is CD0 D 0:0129 as compared to the noncheeked variantCD0 D 0:0117, that is, the drag increases by 10%. For M1 D6,we have CD0

D0:0065 and 0.0073, that is, the drag coefcient de-

creases by roughly the same value. This behavior of CD0 may berelatedto thecorrespondingchangein thedrag over theinletstream-surface CDss . The latter drag for two-dimensional inlets was nu-merically studied in Ref. 10, and the changes in CDss of the samecharacter were obtained for the cases in the presence or absenceof the side cheeks. This effect can be explained by the following.For a cheeked inlet, the area of the inlet streamsurface on the sidesdecreases as compared to the noncheeked variant, which leads tothe corresponding decrease in CDss , notably in the case of owregimes close to M1 Md. Despite an additional drag of exter-nal surfaces of the side cheeks, the total drag of the congurationmay become lower. For Mach numbers M1 < Md, the arrangementof the side cheeks, that is, the restriction of lateral spillage, addi-tionally leads to an increase in the slope of the shock waves arising

on the inlet ramp. This, consequently, leads to an increase in thelength and area of the inlet streamsurface and to an increase in thedrag over this streamsurface. The experimentaldata on CD0 for thevehicle conguration consideredconrm the conclusion10 made onthe basis of numerical calculationsof an inviscidow that, in termsof the smaller drag, a better inlet is the inlet with complete side


11/14


cheeks for M1 Md and the inlet with short cheeks or withoutthem for M1 < Md.

An analysis of the aerodynamic characteristics of the modelshowed that the lifting properties of the conguration and, corre-spondingly, the inductive drag did not practically depend on therestriction of lateral spillage on the compression ramp of the inlet.Therefore, the change in the lift-to-drag ratio L=D associated withthe presence or absence of the side cheeks as a function of M1 isdeterminedby the correspondingchangein CD0 .M

1/. The maximal

lift-to-drag ratio of the conguration with the inlet-cheeked enginenacelle variant as compared to the noncheeked one for M1D 4 is(L=D/max D 4:03 and4.28, respectively,thatis, it is smallerby4%,whereas for M1D 6, it is greaterby 8%, (L=D/max D 5:4 and 5:0.For both variants of the engine nacelle, the maximal lift-to-dragratio of the model increases with increasing the freestream Machnumber.

Thus, in ow regimes with M1 Md, both the ow-rate charac-teristics of the inlet itself and the total aerodynamic characteristicsare better for the model variant with the inlet-cheeked engine na-celle than for the noncheeked variant. The relative increase in theow-rate factor and the decrease in drag are not large, but bothcharacteristics are improved simultaneously, and a more noticeableincrease in efciency of the engine and the vehicle as a whole may

be expected.To evaluatetheeffectof thepresenceor absenceofthe side cheeks

on the overall efciency of the vehicle, the integral aeropropulsivecharacteristics of the conguration were considered for test Machnumbers.It was made inthesamemanner asfor the caseof thestudyon wedge-aided diversion of the boundary layer. The polars of theresultant aeropropulsiveforces CRy .CRx / together with the aerody-namic polars CL .CD / for the model tested with different variantsof the engine nacelle are shown in Fig. 13 for M1 D 6. Note thatthe difference in the aerodynamic polars CL .CD / for these variantsis mainly manifested in a shift of one polar relative to the otherbecause of the different values of CD0 . This is because the pres-ence or absence of the side cheeks has an insignicant effect onthe lift force and lift-dependent drag of the conguration. The po-

lars presented show that, at freestream Mach numbers close to theinlet design Mach number M1 Md, the inlet with complete sidecheeks, which is better, in terms of the smaller drag, as comparedto the noncheeked inlet, is also better in terms of aeropropulsiveperformance.

The propulsive force for the inlet-cheeked model congurationis larger in comparison to the noncheekedvariant within the entirerange of angles of attack considered. The increment 1CRx in theresultant aeropropulsive force caused by the presence of the sidecheeksremains almostconstant within the entire range of angles ofattack under study. Note that the said total increment 1CRx can bedecomposedinto componentscausedby the increasein the ow-ratefactor and by the decrease in the drag CD0 , respectively. For esti-mation of the rst effects, it is necessary to determine the resultant

characteristics CRy .CRx / using experimental data on the ow-rate

Fig. 13 Polars of aerodynamic forces CL(CD ) and resultant a ero-propulsive forces CRy (CRx ) for M1 = 6; model variants with cheeked

and noncheeked inlets.

factor for the inlet-cheeked model variant and on the aerodynamiccharacteristicsfor the noncheekedvariant.The correspondingpolarCRy .CRx / is also plotted in Fig. 13 (combined data, solid circles);it differs little from the polar for conguration with the noncheekedinlet. One can see that the increase in the resultant tangential force,in this context, is predominantly associated with the decrease in thedrag CD0 .

The polars of the resultant aeropropulsive forces obtained forM

1 D4 and for different variants of the engine nacelle revealed

that these characteristics for the cheeked inlet change very littleas compared to that for the noncheeked inlet, which are presentedin Fig. 11 for hblw D 0. The resultant polars for the cheeked inletvariant, both for M1 D6 and 4, have the same features of the pathinstability, which were discussed earlier for the noncheeked inlet inanalyzing the effects of the diverter wedges.

Effect of Nose Bluntness of the Lifting Body

The presented results on the effect of wedge-aided diversion ofthe boundary layer demonstratethat the boundarylayer developingon the precompression surface of the forebody and on the exter-nal compression ramp of the inlet itself is one of the main factorsdetermining the efciency of inlet operation for the vehicle con-

guration considered. In the case of the forebody with the bluntednose that forms a strong bow shock wave, a high-entropy layerdevelops near the body surface, which complicates developmentof the boundary layer. The thickness of this entropy layer may becomparable with or greater than the thickness of the viscous layer.Numerical computations13 of the inviscid ow around a delta wingwith blunted leading edges show that the thickness of the entropylayer on the lower surface related to the wing length increases withincreasing Mach number M1 and decreases with increasing angleof attack. Note that, in most cases, it is impractical to distinguishin the near-wall ow the viscous nonuniformity related to energydissipation in the boundary layer from the inviscid nonuniformitycaused by energy dissipation in the detached curved strong shockwave. Therefore, in our work we understand the near-wall ow as

the viscous

entropy layer, includingboth effects,under the assump-tionthat static pressureis constantacrossthis layer, which holds forthe boundary layer.

To reveal the bluntness effect on the aerodynamic efciency ofthe considered hypersonic scramjet-powered vehicle as a whole,a comparative study of the boundary layer on the precompressionsurface of the forebody was performed at M1 =4.05 for modelconguration variants with a pointed and blunted nose. A congu-ration of the model with the inletnoncheeked engine nacelle andwithout wedge-aideddiversionof the boundary layerwas tested. Asalready indicated,the boundary layerwas measured on the forebody

Fig. 14 Integral characteristics of the boundary layer measured intesting the model with different nose bluntness atM1 = 4.


12/14


pre-compressionsurfacein the planeof symmetryzD 0atadistancex D 340 mm from the model nose tip.

An example of the integral characteristics of the boundary layermeasured in testing the model with different nose bluntness forM1 = 4.05 and Re11 D 56 106 1/m is shown in Fig. 14. Note thatthe nose bluntness leads to a signicant increase in all thicknessdimensions of the boundary layer. For instance, within the range ofangles of attackD 0 3 deg, the values of , , and increaseby a factor of 2.5 2.7 and for

D10 deg by a factor of 1.4 1.8

as compared to the pointedconguration.An analysis of boundary-layervelocityprolesobtainedforthecongurationwiththebluntednose showed that the entropy layer seemed to be absorbed by theboundary layer upstream of the measurement cross section only forangles of attack 10 deg.

Based on the measurement data for M1 D 4 and D 0, theboundary-layerthicknessahead of the inlet for the pointed congu-rationis D 3:9 mm, and thedisplacementthicknessis D 1:7mm.In the case with nose bluntness, the corresponding values are D10:1 mm and D 4:3 mm. This signicant increase in the dis-placement thickness should be accompanied by noticeable lossesof the ow rate of air through the inlet. The experimental val-ues of the ow-rate factor obtained for the blunted congurationat M

1D4 and 6 are plotted in Fig. 15 as a function of the an-

gle of attack. For comparison, Fig. 15 also shows the data for thepointed conguration discussed earlier. For angles of attack nearD 0, the nose bluntness decreases the ow-rate factor by 23%for M1 D 4 and by 42% for M1D 6, that is, the losses of theow rate caused by the nose bluntness signicantly increase withincreasing the freestream Mach number. Nevertheless,the effect ofnosebluntnessdecreaseswith increasingangleof attack,as couldbeexpectedfrom evidenceof the measured values of. For 9 deg,the ow-rate factors for the blunted and pointed congurations areclose.

The character of variation of the ow-rate factor as a functionof the angle of attack for the blunted conguration determines to alarge extent the behavior of the total aerodynamiccharacteristicsofthe conguration as a whole, which include the forcesover the inlet

streamsurface.In particular, note that, for a Mach number M1 D 6,the model drag signicantly increases within the range of angles ofattack 6 deg,and the aerodynamicpolarCL .CD / for thebluntedconguration is no longer parabolic (Fig. 16). The increase in thezero-liftdrag coefcient CD0 reaches12% for M1 D 4 and21%for M1 D 6. Thus, the effects of nose bluntness on the vehicle dragand on the inlet ow-rate factor increase with increasingfreestreamMach number.

The experimental data showed that the lifting properties of theconguration depended weakly on nosebluntness,though some in-crease in the values CL and C

L

was observed within the range ofangles of attackD 06 deg for M1 D 6. As a whole, nose blunt-ness decreasedthe maximal lift-to-dragratio (L=D/max by 3% forM

1 D4 and by

7% for M

1 D6. At the same time, because of

the already noted deformation of the aerodynamic polars CL .CD /,

Fig. 15 Inlet ow-rate factor vs the angle of attack for the model vari-ants with blunted and pointed nose tips.

Fig. 16 Polars of aerodynamic forces CL(CD) and resultant aero-propulsive forces CRy (CRx ) for M1 = 6; model variants with blunted

and pointed nose tips.

the maximum of the lift-to-drag ratio for M1 D 6 for the bluntedconguration was reached at signicantly greater angles of attack.L =D/max D 1011 deg as compared to .L =D/max D 67 deg for thepointed conguration.

It would appear reasonable that the decrease in the ow-rate fac-torof theinletand theincreasein thedragcoefcient causedby nosebluntness will lead to deterioration of the resultant aeropropulsivecharacteristics of the vehicle. For example, the thrust coefcientCT calculated for the blunted model conguration, as compared tothe pointed variant, is very low within the range of angles of at-tackD 0 8 deg, with a corresponding signicant decrease in theexperimental ow-rate factor (see Fig. 15). As a consequence, theengine thrust obtained does not compensate for the elevated dragof the conguration caused by nose bluntness and does not ensurethe resultant propulsive tangential force for these angles of attack(Fig. 16). However, a positive tangential force occurs for larger an-gles of attack > 6 deg. An extraordinaryS-shaped resultant-forcepolarfor the bluntedmodel congurationis obtained.One wouldex-pect somespecialinstabilityproperties,which would be reectedbythe said extraordinary S-shaped resultant-force polar, for a vehicleconguration with a blunted nose of the lifting body. For M1 D4,the changesin aerodynamicand resultantaeropropulsivecharacter-istics caused by nose bluntness are the same in the qualitative sensebut are less signicant in magnitude.

The noted extraordinary behavior of the resultant aeropropulsiveforces, by the data obtained, is a consequence of unusual varyingof the inlet ow rate. Because of scanty measurement on inlet owpatterns,one couldonly conjecturefrom general considerationsthatthis peculiarity is a summed effect of realignment of ow aroundthe blunted forebody in comparison to the pointed one, includingboth the inviscid ow and the boundary layer or a more extendedviscidow near itssurface.All of these effectsinvite a more detailedinvestigation.

Conclusions

Some factors of inlet/airplane interaction were experimentallystudied for a hypersonic scramjet-powered vehicle with a ventralposition of the engine nacelle. Experimental aerodynamic charac-teristics are supplemented by calculated estimates of the enginethrust. Thus, the effects of the considered factors on the total aero-propulsivecharacteristicsof the vehicleare determined.The resultsobtained allow us to make the following conclusions.

1) Diversion of the boundarylayer developingahead of the inlet,with the use of diverterwedges, improves the ow-rate characteris-tics of the inlet and increases the engine thrust at high ight Mach

numbers. Despite the increase in the drag force in this case, an in-crease in the resultantpropulsiveforce of the vehicleas a whole canbe reached with a certain height of the diverter wedges.

2) The arrangementof the side cheeks restrictinglateral spillageof ow over the compressionwedge ramp of the inlet leads to rathersmall increments in the ow-rate factor both for Mach numbersM1 < Md and M1 MdD 6. This may be explained by effects of


13/14


three dimensionalityof the ow aroundthe inlet integratedwith theforebody, which violate the design ow regime for the inlet itself.For freestream ow conditions with M1 Md, the inlet with theside cheeks ensures a lower drag of the model as compared to thenoncheeked inlet variant.Correspondingly,the resultant propulsiveforce of the vehicleincreasesin the rst case.For M1 D 4 < Md, bycontrast, the noncheekedinlet provides lower drag, therewith thereis a minor effect of increasing of the resultant propulsive force ofthe vehicle.

3) Nose bluntness of the lifting body is responsible for devel-opment of a thick near-wall viscous-entropy layer on the fore-body precompression surface, especially for small angles of at-tack. This leads to an increase in the vehicle drag and to sig-nicant losses in the ow-rate factor and engine thrust at near-zero angles of attack, and, as a consequence, the resultant aero-propulsive characteristics of the vehicle as a whole become muchworse.

The approach used in the presented study on the problem ofinlet/airplane interaction is based on experimentsand calculations.An important merit of such an approach is its comprehensive-ness and the possibility to consider ight properties of the ve-hicle as a whole, including the airplane and the engine. At thesame time, in this study, the extrapolation of the experimental

aerodynamic characteristics to a full-scale vehicle and determi-nation of the engine thrust are rather simplied. The inlet is notfully represented, the parameters of its internal section down-stream of the entrance are assumed, and the engine thrust is de-termined on the basis of a one-dimensional scramjet model. Thisdoes not allow one to consider in detail operability issues and per-formance quantities of the engine, such as inlet starting, engineow stability, distortion, fueling, effects of engine cross-sectionalareas, etc. That is why the results presented are not absolute,but they give a general idea about possible aeropropulsive prop-erties of the vehicle under consideration, which call for furtherinvestigation.

References1Starukhin,V. P., and Taryshkin,A. G., Investigationof Parameters of the

BoundaryLayerUpstreamoftheEntranceofaSupersonicTwo-DimensionalInlet Mounted Under the Plane Surface of a Delta Wing, Uchjonye ZapiskiTsAGI(Scientic Notes of Central AeroGydrodynamic Institute), Moscow,Vol. 13, No. 2, 1982, pp. 6977 (in Russian).

2Bosnyakov, S. M., Starukhin, V. P., and Chevagin, A. F., Effect of aDelta Wing on Characteristics of a Two-Dimensional Inlet Mounted Un-der the Wing, Uchjonye Zapiski TsAGI (Scientic Notes of Central Aer-oGydrodynamic Institute), Moscow, Vol. 25, No. 1 2, 1994, pp. 6775(in R ussian).

3Goldfeld, M. A., Lisenkov, I. G., and Shcherbik, D. V., ExperimentalStudy of Three-Dimensional Boundary Layer on a SupersonicAircraft Sur-face, Proceedings of InternationalConferenceo n theM ethodsof Aerophysi-cal Research, Inst. of Theoretical and AppliedMechanics, Russian Academyof Sciences Press, Siberian Branch, Novosibirsk, Pt. 2, 1994, pp. 96

101.4Lawing, P. L., and Johnson, C. B., Inlet Boundary-Layer Shapes onFour Aircraft Forebodies at Mach 6, Journal of Aircraft, Vol. 15, No. 1,1978, pp. 62, 63.

5Johnson, C. B., and Lawing, P. L., Mach 6 Floweld Survey at theEngine Inlet of a Research Airplane, Journal of Aircraft, Vol. 14, No. 4,1977, pp. 412 414.

6Vasilev, V. I., Problems of Designing Inlets for Supersonic PassengerAircraft, Trudy TsAGI (Transactions of Central AeroGydrodynamic Insti-tute), No. 1501, 1973, pp.1 18 (in Russian).

7Zheltovodov, A. A., Rafaelyants, A. A., and Filatov, V. M., Investiga-tion of Drag and Special Features of Flows Around Wedges for BoundaryLayer Diversion, Aerodynamic Interference in the Flows Around Three-

Dimensional Bodies, Inst. of Theoretical and Applied Mechanics, Academyof Sciences of the USSRPress, Siberian Branch, Novosibirsk, 1980,pp. 63

82 (in Russian).8Goonko, Y. P., Mazhul, I. I., and Shcherbik, D. V., Effect of Boundary-

Layer Bleeding on the Aeropropulsive Efciency of a Hypersonic Vehi-cle, Izvestiya VUZov, Aviatsionnaya Tekhnika ( News of Institutions of

Higher Education, Aviation Technology Series ), No. 1, 1993, pp. 89 92 (inRussian).

9Goonko, Y. P., Mazhul, I. I., and Shcherbik, D. V., Effect of Boundary-Layer Bleeding on Performance of an Inlet and the Hypersonic Scramjet-

Powered Vehicle as a Whole, Thermophysics and Aeromechanics, Vol. 7,No. 2, 2000, pp. 173 185.

10Bosnyakov, S. M., and Remeev, N. K., Investigation of Three-Dimensional Supersonic Flow Around a Two-Dimensional Inlet withSide Cheeks, Uchjonye Zapiski TsAGI (Scientic Notes of Central

AeroGydrodynamic Institute), Vol. 11, No. 5, 1980, pp. 40 45 (inRussian).

11Bosnyakov, S. M., Bykova, S. A., and Remeev, N. K., Inves-tigation of Three-Dimensional Flow and Aerodynamic Characteristicsof Two-Dimensional Inlets with Different Entrance Shape and Size

of Side Cheeks, Uchjonye Zapiski T sAGI (Scientic Notes of Cen-tral AeroGydrodynamic Institute), Vol. 14, No. 3, 1983, pp. 107 113(in Russian).

12Vinogradov, V. A., Duganov,V. V., Zakharov, N. N., Ivanov, M. Y., andIvanov, O. K., Experimental and Numerical Studies of Supersonic Flowover the Wedge Ramp of a Two-Dimensional Inlet, TR of CIAM (CentralInst. of AviationMotor Building), CIAM TR 8198, Moscow, 1977,pp. 169(in Russian).

13Starukhin, V. P., and Chevagin, A. F., Effect of Bluntness of EntranceEdges on Characteristicsof InletsMountedUnder a Wing,Uchjonye ZapiskiTsAGI (Scientic Notes o f Central AeroGydrodynamic Institute), Vol. 25,No. 1 2, 1994, pp. 89100 (in Russian).

14Bakharev, S. A., Gurylev, V. G., and Kosykh, A. P., Aerodynamic Char-acteristics of a Blunted Delta Wing with Simulation of Air Flow Throughan Inlet at Supersonic and Hypersonic Velocities, Uchjonye Zapiski TsAGI(Scientic Notes of Central AeroGydrodynamic Institute), Pt. 2, Vol. 22,

No. 1, 1991, pp. 3947 (in Russian).15Aleksandrovich, E. V., Drag of Blunted Edges of Supersonic and

Hypersonic Inlets, Tekhnika Vozdyshnogo Flota ( Aireet MachineryJournal of Aviation Science and Technology), No. 6, 1997, pp. 918 (inRussian).

16Goonko, Y. P., Aerodynamic Experimentation with Ducted Modelsas Applied to Hypersonic Air-Breathing Vehicles, Experiments in Fluids,Vol. 27, No. 3, 1999, pp. 219 234.

17Volonikhin, I. I., Grigorev, V. D., Demjanenko, V. S., Pisarenko, K.I., and Kharitonov, A. M., Supersonic Wind Tunnel T-313, Aerophysical

Research, Inst. of Theoretical and Applied Mechanics, Academy of Sci-ences of the USSR Press, Siberian Branch, Novosibirsk, 1972, pp. 811(in Russian).

18Goonko, Y. P., Experimental Comparison of Aerometric Techniquesfor Determination of the Total Flow-Rate of Supersonic Non-UniformStreams, Preprint,Inst. of Theoreticaland AppliedMechanics,Academy of

Sciences of the USSR, Siberian Branch, Novosibirsk, No. 2, 1981, pp. 127(in Russian).

19Blisch,V. G.,Externaland Internal AerodynamicForces and MomentsActing on Air-Breathing Vehicles and Their Models at Non-Zero Anglesof Attacks,

6 some factors of hyper sonic in let airplane interactions

Documents