NASA Technical Memorandum 107274 ICOMP-96--6; AIAA-96-3145 AJ '_ • --_ _3 Rocket-Based Combined Cycle Engine Technology Development_Inlet CFD Validation and Application J.R. DeBonis Lewis Research Center Cleveland, Ohio and S. Yungster Institute for Computational Mechanics in Propulsion Cleveland, Ohio Prepared for the 32nd Joint Propulsion Conference cosponsored by AIAA, ASME, SAE, and ASEE Lake Buena Vista, Florida, July 1-3, 1996 National Aeronautics and Space Administration I_( ICOMP _ _..._\ _ Rueavch Centor https://ntrs.nasa.gov/search.jsp?R=19960033187 2020-04-10T04:11:15+00:00Z
Transcript
NASA Technical Memorandum 107274
ICOMP-96--6; AIAA-96-3145
AJ '_ • --_
_3
Rocket-Based Combined Cycle Engine
Technology Development_Inlet CFDValidation and Application
J.R. DeBonis
Lewis Research Center
Cleveland, Ohio
and
S. YungsterInstitute for Computational Mechanics in Propulsion
Cleveland, Ohio
Prepared for the32nd Joint Propulsion Conference
cosponsored by AIAA, ASME, SAE, and ASEELake Buena Vista, Florida, July 1-3, 1996
Rocket-Based Combined Cycle Engine Technology Development -Inlet CFD Validation and Application
J. R. DeBonisNASA Lewis Research Center
Cleveland, Ohio
and
S. YungsterInstitute for Computational Mechanics in Propulsion
Cleveland, Ohio
Abstract M
PS
A CFD methodology has been V
developed for inlet analyses of Rocket-Based pCombined Cycle (RBCC) Engines. A fullNavier-Stokes analysis code, NPARC, was Xw
used in conjunction with pre- and post-processing tools to obtain a complete subscriptsdescription of the flow field and integrated inletperformance. This methodology was closeddeveloped and validated using results from a expsubscale test of the inlet to a RBCC "Strut-Jet" ext
engine performed in the NASA Lewis 1x 1 ft. intsupersonic wind tunnel. Results obtained from mornthis study include analyses at flight Mach offnumbers of 5 and 6 for super-critical operating onconditions. These results showed excellent open
agreement with experimental data. The p/sf
analysis tools were also used to obtain pre-test Pperformance and operability predictions for the viscousRBCC demonstrator engine planned for testing xin the NASA Lewis Hypersonic Test Facility. oo
This analysis calculated the baseline fuel-offinternal force of the engine which is needed todetermine the net thrust with fuel on.
Mach number
pressuresurface
velocity
density
wall shear stress
closed or solid surface
experimentally obtained forceexternal forceinternal forcemomentum balance forcefuel off conditionfuel on condition
open or permeable surfacepressure/skin friction forcepitotviscous forceaxial distancefreestream
Introduction
Ao
AiF
Nomenclature
freestream area of capturedstreamtube
inlet geometric projected areaforce
Currently, interest in hypersonic
propulsion systems is on the rise. Thesesystems have a variety of applications includingmissile, reconnaissance and single stage to
orbit vehicles. The propulsion system mustoperate efficiently throughout the flight regimefrom takeoff to hypersonic cruise. Toaccomplish this, combined cycle engines are
being studied. Because different propulsioncycles operate more efficiently at different flightconditions, these engines combine two or more
different propulsion cycles into one integratedsystem for better performance. One suchpropulsion system being developed at theNASA Lewis Research Center is the Rocket-
Based Combined Cycle Engine (RBCC)IThis engine integrates a high specific impulselow thrust-to-weight airbreathing engine with alow specific impulse high thrust-to-weightrocket. From takeoff to Mach 2.5 the engineoperates as an air-augmented rocket. At Mach2.5 the rockets are turned off and the engine istransitioned to a dual-mode ramjet. BeyondMach 8 the rocket would be turned back on.
One RBCC engine concept underdevelopment at Lewis is the "Strut-Jet" concept(Fig. 1). This work is a joint effort betweenNASA Lewis, the United States Air Force,
Gencorp Aerojet, GASL and Lockheed Martin.This engine contains two struts in the inlet toprovide compression. The struts divide theflow into three separate flow paths. Therockets are embedded in the base of the struts.
The fuel injectors needed for the ramjet modeare also housed in the struts.
A demonstration of this engine conceptat Mach 6 and 7 flight conditions is scheduledfor NASA Lewis' Hypersonic Tunnel Facility
(HTF)2. Critical to the success of thisdemonstration is the performance of the inlet.It must: provide sufficient pressure recoveryover the entire Mach number range,demonstrate reasonable starting and unstartcharacteristics, provide a mass distributionadequate for combustion and operate with alow internal drag. Because of the inlet's
importance, a significant effort has beenfocused on providing a good understanding ofits behavior. This effort included a subscale
experiment in the NASA Lewis lxl ft.
supersonic wind tunnel3 and the developmentof a computational fluid dynamics (CFD)capability for RBCC inlet analysis. This paperfocuses on the development, validation andapplication of the CFD capability for RBCCcombined cycle inlets.
Computational A_nproach
The methodology developed for theStrut-Jet analyses utilized a combination ofgovernment developed codes (grid generator,flow solver and post-processors) already in use
at NASA Lewis. The codes' capabilities andthe existing experience base were the reasonsfor their selection.
Geometry_ Definition and Grid Generation
The Strut-Jet geometry can be definedas a series of flat planes with no curvature orirregularities of the surfaces. The coordinateswhich defined these planes were taken from aset of engineering drawings for the subscaleinlet model. A short FORTRAN code was
written to generate a set of points which lay onthe strut-jet surfaces. This surface databasewas then input into the grid generator.
The grid generator used for this workwas GRIDGEN version 9.64. It is an
interactive graphical interface driven codewhich runs on UNIX workstations. The gridwas subdivided into several domains in order
to simplify grid generation. In order to keepthe number of mesh points defining the strut toa reasonable level, the break lines defining thechanges in strut width were not explicitlymodeled. However, all grid points wereconstrained to lie on the database surface. This
was deemed to be a reasonable compromisebetween geometric accuracy and computationalefficiency.
Two grids were used for this study(Fig. 2 and 3). The first grid models thesubscale inlet experiment and was used foranalysis at the super-critical conditions. The
second grid models the demonstrator engineflow-path for the HTF test. It is based on thefirst grid, but the last block was modified torepresent the exhaust nozzle on the demo-engine. Both grids model one-half of thesymmetric inlet flow path, including the pre-compression plate, diverter, center duct, sideduct and combustor. Table 1 gives detailedinformation on the grids generated for thisstudy.
FlOW Solution
Code- The NPARC codes was used asthe flow solver. It is a general purpose
computational fluid dynamics code which iswidely used in government, industry andacademia for fluid flow simulations,
particularly of aircraft propulsion systems
components.NPARC solves the Reynolds averaged
Navier-Stokes equations in strong conservationlaw form using the Beam-Warming
approximate factorization algorithm6. Spatialdiscretization is performed using a centraldifference scheme. Jameson-style artificial
dissipation is added for stability and to smoothshock oscillations and odd-even grid point
decouplingT. The code uses a perfect gasequation of state.
NPARC is very flexible in handling
computational grids. The code allows the userto specify a boundary condition on any portionof any grid surface. This allows a complexgeometry to be handled using one grid block.It is also capable of using grids subdivided intomultiple grid blocks. The grid points may ormay not be contiguous across the blockinterface. In the case of a noncontiguousinterface, the code uses a trilinear interpolation
scheme to pass data through the boundary.The code has several options for
modeling turbulence varying from algebraic(zero transport equations) to one and twoequation models. The Chien low-Reynolds
number k-_ model8 was used for this work. It
has been shown to give good results for a wide
range of flow problems.Boundary Conditions.- A fixed
supersonic inflow was specified upstream ofthe pre-compression plate. Flow conditionswere extrapolated to the boundaries at regionsof supersonic outflow, including the diverterand exit planes. Solid boundaries were treatedas no slip adiabatic walls. Symmetry about theinlet centerline was assumed and a slip wall
was specified on this plane.
Post Processing
The flow solution was processed todetermine, wall static pressure distributions,massflow distributions, pitot pressuredistributions and internal drag. A separateFORTRAN code was written to perform thesecalculations. Pressure distributions, on the
cowl and body along the centerline, and on thesidewall are output for comparison to
experimental data. For the sidewall pressuredistribution, the location of the experimentaldata did not lie on a computational grid line.Therefore, the computational pressures had to
be interpolated onto the experimental locationsfor comparison. Mass continuity was checkedby integrating the massflow over each axialgrid plane through the duct. Pitot pressuredistributions at the strut base were used to
compute massflow, distortion and pressurerecovery for comparison to experiment. Thecomputational pressures were interpolated ontothe experimental rake positions in order toeliminate any differences that would be causedby lower resolution of the experimental data.The experimental data reduction program wasthen used to process the data for comparison.
The internal drag of the inlet was
computed using a control volume method9.The momentum equation written in controlvolume form is
_ (p V . dS) V = -_ pdS + F_._co_,S S
where,
The control volume for the Strut-Jet is shown
in Figure 4. It is drawn such that no portion ofthe control volume passes through a solidsurface. The control volume can be separated
into the open or permeable surfaces and theclosed or solid surfaces. For the closed
surfaces, the momentum flux terms are zero.
For permeable surfaces the viscous forces arezero. This yields the following equation.
The fight hand side of this equation representsthe forces on the solid surfaces and is the
internal force generated on the inlet. It is
equivalent to the integral of the net rate flow ofmomentum and pressure over the open controlvolume surfaces as represented by the left hand
side of the equation. Therefore the internalforce on the inlet can be calculated by either a
pressure skin friction integration or amomentum flux integration.
Fpl_ = - _ pdS + Fv_cou,Screwed
p.. : (pv.ds)v+ pctS
Typically, accurate calculation of skin frictionfrom a CFD solution is difficult, due to gridskewness at the wall and boundary layerresolution. The momentum flux integration
offers a simpler less computationally intensiveway to determine the drag force. Both methodswere examined in this study to determine the
best method for application to this flow
problem.
_x_rimcntal Pro m'ams
Subscale Inlet Test
Results from the subscale inlet test were
used for the code validation portion of this
study. This test was performed in the NASALewis lxl ft. supersonic wind tunnel. It is a
one-pass continuous flow facility that canprovide Math numbers from 1.3 to 6.0 in thetest section. These Math numbers are achieved
by changing two-dimensional fixed geometrynozzle blocks. The air is heated to preventcondensation and liquification of the flow. The
test program ran at Math 4.0, 5.0, 5.5, and6.0. CFD comparisons were made for theMach 5.0 and Mach 6.0 conditions.
The geometry is a 40 percent scalemodel of the demonstrator engine inlet. Themodel uses a pre-compression plate to generatethe initial shock wave. The model was
mounted 0.4 in. below this plate in order todivert the pre-compression plate boundarylayer. Static pressure taps were located on themodel centerline on both the cowl and bodysurfaces and on the model sidewalls. A
traversing rake of 12 pitot pressure tubes (3 ineach side duct and 6 in the center duct) waslocated at the base of struts. The rake took
measurements at 20 equally spaced intervals.A wedged shaped massflow plug was used tosimulate combustion back-pressure. The plugsimulates several back pressured conditionsbetween super-critical (no back pressure) andcritical (near inlet unstart). The CFDcomparisons presented here were made atsuper-critical conditions. Code validation atback pressured conditions is ongoing.
Demonstrator Engine Test
The complete Strut-Jet engine will bedemonstrated at NASA Lewis' HypersonicTunnel Facility (HIT) in the Summer of 1996.HTF, located at Lewis' Plum Brook Station, is
a blow-down non-vitiated free-jet facility
capable of providing true air composition fortesting up to Mach 7 flight conditions.
In this experiment the engine will betested in both the air-augmented rocket andramjet modes. The rockets operate onmonomethyl hydrazine (MMH) and inhibitedred fuming nitric acid (IRFNA). In ramjetmode the engine is fueled by Liquid JP-10
injected from the struts. The JP-10 is pilotedby MMH-IRFNA pilots. HTF will be operatedat test section Mach numbers of 5 and 6, with
enthalpy levels corresponding to Mach 6 and 7.This is done to simulate some vehicle forebodycompression. The engine will be instrumentedwith static pressure taps and a limited numberof thermocouples and dynamic pressuretransducers. The engine will be mounted in thetest section on a force balance. The balance
will measure the total force on the engine
including internal and external components.
4
From this force data one can obtain the changein internal force on the engine between fuel-offand fuel-on conditions, assuming that theexternal forces are not affected by fueling the
engine.
aF= Fo.-
= +
= - F O,o.
However because the external drag of the
engine cannot be separated out of the fuel-onmeasurement, the true thrust produced by the
engine cannot be obtained directly from theexperiment. The CFD analysis for this testprogram focused on obtaining the intemal forceon the engine at fuel-off conditions. By addingthe change in thrust between fuel-on and fuel-off conditions to the fuel-off internal force
obtained by CFD, an estimate of the net internal
engine thrust can be made.
Fi,,_" = Finto_ + AF
Results
Code Validation, Subscale Inlet Calculations
Code validation was done by analyzing
the subscale inlet geometry and comparing theresults to experimental data. The inlet wasanalyzed at super-critical conditions forfreestream Mach numbers of 5 and 6. TheMach 5 calculation was run at a freestream
static pressure of 25.50 psf and a freestreamstatic temperature of 118.5 R. The Math 6calculation was run at a freestream static
pressure of 14.98 psf and a freestream statictemperature of 93.13 R. Near critical analysesat Mach 5 and 6 are underway, but were not
completed at the time of this report.Mach 5, Super-Critical.- A contour
plot of static pressure on the inlet's centerlineplane (Fig. 5a) clearly shows a series of strongoblique shocks initiated by the pre-compressionplate. A series of shock waves in thetransverse direction is created by the strut
leading edge and are evidenced by the nearvertical pressure contours on the forward
portion of the inlet. Mach contours (Fig. 5b)indicate a large region of low speed flow along
the body side of the inlet. This low energyflow is initially caused by a boundary layerseparation induced by the transverse shock.The reflections of the cowl shock onto the body
side creates a series of adverse pressuregradients which continue to enlarge the lowspeed region.
Figure 6 compares the predicted staticpressure distributions to the experimental datafor the cowl, body and sidewall surfaces.
Agreement for the cowl and body pressures isvery good. On the cowl centerline the largefore and aft pressure rises are due to the cowlshock and its subsequent reflection. Thetransverse shock from the strut creates the
small pressure rise at x=18 inches. On thebody centerline pressure rises at x=21 inchesand x=26.5 inches are generated by thereflected cowl shock. The code overpredicts
the pressures along the sidewall. This could bedue an over-prediction of the side ducts'massflow, caused by the fact that the analysisassumed no spillage around the sidewalls.
A comparison of pitot pressuredistributions at the base of the struts is shown
in Figure 7. Overall the CFD predicts lowerlevels of pressure throughout the duct. The
predicted shapes of the contours arequalitatively correct.
Results of the internal drag calculationare shown in Table 2. The data shown is for
the computational domain which is one half ofthe actual inlet. Forces for each surface of the
integration are listed for both methods. For themomentum calculation positive values representmomentum flowing out the control volume and
negative values represent momentum flowingin. For the pressure/skin friction calculationpositive values represent thrust and negativevalues represent drag. At this condition themethods agree within 7 percent.
Mach 6, Super-Critical.- Qualitativelythe Mach 6 results are very similar to the Mach
5 results. The contour plot of static pressureon the inlet's centerline plane (Fig. 8a) againshows the cowl and transverse shocks and their
reflections. Mach contours (Fig. 8b) show the
large region of low speed flow along the bodyside of the inlet.
Figure 9 compares the predicted staticpressure distributions to the experimental datafor the cowl, body and sidewall surfaces.Agreement for the cowl and body pressures isvery good. On the cowl centerline the largefore and aft pressure rises are due to the cowlshock and its subsequent reflection. Thetransverse shock from the strut creates the
small pressure rise at x=18 inches. On thebody centerline pressure rises due to thereflected cowl shock, x=22 inches, is betterresolved than at the Mach 5 condition. The
sidewall pressure distribution is overpredictedand the discrepancy is slightly larger for thiscase.
A comparison of pitot pressuredistributions at the base of the struts is shown
in Figure 10. Agreement is good for the sideducts. In the center duct near the body side theregion of low energy flow is predicted verywell. Near the cowl surface, the analysis
predicts a large region of high pressure lowdistortion flow whereas the experimentmeasured a lower pressure more distortedflowfield.
Results of the internal drag calculationare shown in Table 3. At this condition the
methods agree within 1.5 percent. At bothMach numbers calculation of skin friction was
difficult and time consuming. Problems arosein determining the vector normal to the wall inskewed areas of the grid. Because of this andthe fact that agreement between the forcecalculations was good, the momentum flux
integration was used for the remainder of thecalculations to determine the internal drag.
Code Application.Calculations
Demonstrator Entzine
Once satisfactory results were obtainedon the subscale inlet model the code was then
applied to the demonstrator engine geometry inorder to obtain the fuel-off internal drag to
support the experimental program. There aretwo significant differences between subscaleinlet and demonstrator engine calculations.First the constant area combustor section of the
subscale test was replaced with the engine's
exhaust nozzle. Second, true high temperatureflight conditions consistent with the engine testwere used instead of the cold temperatures ofthe subscale experiment. At the fuel-off (nocombustion) condition there is no back-pressure on the inlet. Therefore these resultsare comparable to the super-critical subscaleinlet results. Both the Mach 5.2 and Mach 6calculations were run at a freestream static
pressure of 89.1374 psf and a freestream statictemperature of 514.8 R.
Mach 5.2 - Contour plots of pressureand Math number (Fig. 11) indicate that theflowfield is very similar to the cold subscaleresults. The shock system appears to be thesame as the subscale case. The boundarylayers on the pre-compression plate and in the
inlet appear to be thicker. The Math contoursin the exhaust nozzle show that the flow thereis distorted. Pressure distributions on the cowl
and body centerlines (Fig. 12) have the sameshape as the subscale inlet experiment. Thisimplies that the shock system is the same. Anydifferences in the pressure levels between thesubscale inlet and the demonstrator engine canbe attributed to viscous effects.
Fuel-off internal drag calculated using
the momentum flux integral is show in Table 4.As in the subscale results the values are for the
computational domain which represents onehalf of the inlet.
Mach 6 - Contour plots of pressure andMach number (Fig. 13) again indicate that theflowfield is very similar to the cold subscaleresults. The boundary layers on the pre-compression plate and in the inlet appear to bethicker. The Mach contours in the exhaustnozzle show that the flow there is distorted.
Again, the pressure distributions on the cowland body centerlines (Fig. 14) have the sameshape as the subscale results. At identicalfreestream Mach numbers the demonstrator
engine result has noticeably lower pressurelevels than the subscale result. This is further
evidence of a stronger viscous interaction in thedemonstrator engine.
Fuel-off internal drag calculated usingthe momentum flux integral is show in Table 4.The decreased drag at higher Mach number isconsistent with the subscale results.
Inlet Mass Capture
Using the computed massflow fromboth the subscale inlet and demonstrator enginecalculations the inlet massflow ratio was
determined. Figure 15 shows these results inrelation to the theoretical results. The
theoretical results were computed using two-dimensional inviscid streamlines. They are
shown for both the engine flow and the totalflow which includes the flow which passes
through the diverter. The theory neglects thecompression due to the struts upstream of thecowl lip. This compression creates additionalflow turning which will decrease the freestream
capture area. It also neglects the boundarylayer on the pre-compression plate. The theorypredicts that the pre-compression shock is onthe cowl lip at Mach 5 and all the flow iscaptured without spillage. The diverter reducesthe amount of flow into the inlet and theamount of diverted flow increases with Machnumber. The CFD results indicate that spillage
flow continues to exist beyond Mach 5 because
the capture area continues to increase withMach number. This is due to the additional
compression of the struts. The demonstratorengine results show a higher capture ratio thanthe subscale results, and than the theory atMach 6. This may be accounted for by the
large boundary layer on the pre-compressionplate which reduces the amount of flow that canenter the diverter.
Concluding Remarks
A CFD methodology for analyzinginlets for Rocket-Based Combined Cycle
propulsion systems was developed. Thismethodology was put in place to support an
upcoming demonstration of an RBCC engineconcept, the Strut-Jet. The resulting systemwas validated against data obtained from asubscale test of the Strut-Jet inlet. Very good
agreement between the analysis andexperimental data was obtained for surfacestatic pressure distributions, and pitot pressureprofiles. A means for determining the internaldrag of the system was also developed. Two
methods a momentum flux integral and a
pressure/skin friction integration wereexamined. Both methods yielded similarresults for the subscale inlet. However, the
momentum flux integral proved to be less
computationally intensive and easier toimplement. The validated analysis tools werethen applied to the demonstrator engine andinternal drag forces were obtained. Theseforces will be used in the upcoming test to
compute the net thrust of the engine.
References
1. Escher, W.J.D., Hyde, E.H. andAnderson, D.M., "A User's Primer for
Comparative Assessments of All-Rocket andRocket-Based Combined-Cycle Propulsion
Systems for Advanced Earth-to-Orbit SpaceTransport Applications," AIAA 95-2474, July1995.
2. Thomas, S.R., Trefny C.J. and Pack
W.D., "Operating Capability and CurrentStatus of the Reactivated NASA Lewis
Research Center Hypersonic Tunnel Facility,"AIAA 95-6146, NASA TM 106808, April1995.
3. Fernandez, R., Trefny, C.J., Thomas,S.R. and Bulman, M., "Parametric Data froma Wind Tunnel Test on a Rocket Based
Combined Cycle Engine Inlet," NASA TM107181, July 1996.
4. Steinbrenner, J.P. and Chawner, J.R.,"Automatic Structured Grid Generation Using
Gridgen (Some Restrictions Apply)," NASACP 3291, Surface Modeling, Grid Generation,and Related Issues in CFD Solutions, NASALewis Research Center, Cleveland, Ohio, May1995.
5. Cooper, K., "NPARC 2.0 - Features and
Capabilities," AIAA 95-2609, July 1995.
7
6. Beam, R. and Warming, R.F., "An Implicit
Finite Difference Algorithm for HyperbolicSystems in Conservation Law Form," J.Comp. Phys., vol. 22, no. 1, September 1976,pp. 87-110.
7. Jameson, A., Schmidt, W. and Turkel, E.,
"Numerical Solutions of the Euler Equations byFinite Volume Method for Solving the Elderand Navier-Stokes Equations for High SpeedFlows," AIAA 81-1259, June 1981.
8. Chicn, K.-Y., "Predictions of Channel andBoundary-Layer Flows With a Low-Reynolds-Number Turbulence Model," AIAA Journal,
vol. 20, no. 1, January 1982, pp. 33-38.
9. Voland, R.T., "Methods for Determiningthe Internal Thrust of Scramjet Engine Modulesfrom Experimental Data," AIAA 90-2340, July1990.
/ i " i I.... _'__, _,m_ _" I....................i,...:...............;.............Io _.,,l.,_o, I........
,I i I c_ _®'_'L_n°_ I....................i....................2.............i o _.._._o_o_ I.......
E _ E.......... I ' ' ' I ' ' ' I ' ' '
0 2 4 6 8 10 12
Mach Number
Figure 15. Inlet capture area ratio
18
Form ApprovedREPORT DOCUMENTATION PAGE OMBNo. 0704-0188
_a'_e_ am malnmn_g the data traded, md =ompa_ng am revw*_n0 .me.=?.X=oq.or ,,_=r,,_,=m_ ¢,?m _,=ff,_.?_ ,_r?,LW _',_? _='_"_'_'_'_Y'.'__';__"...%'_'_,_2"_"__"collection al Inlormmon, Includings_tions for reducingthis huron, to wa=n,ngton vm=Xl.U_aner==_v_m. uy_.e_==F=_' mr2_._._ n .moL_. ,=jy_._,m_.,Davis Higl_t_/. Suite 1204, A.-fmgton.VA 2220_-430_ and to the Office of Mana_ ano uuo_N. =..apetwomminuet=onP'rolec¢(um*.-oluu;, wmnr, gmn. L_ zu=_.
1. AGENCY USIE ONLY (Leave blank) 2. REPORTDATE 3. i_.PuHi sYPE AND DATES COVER _-_
Prepared for the 32nd Joint Propulsion Conference cosponsored by AIAA, ASM_, SAE, and ASEE, Lake Buena V'L_a,
Florida, July 1-3, 1996. J-R. DeBonis, NASA Lewis Research Center, and S. Yungster, Institute for ComputationalMechanics in Propulsion, NASA Lewis Research Center, Cleveland, Ohio (work performed under NASA Cooperative
Agreement NCC3-370). ICOMP Program Directm, Louis A. Povinelli, organization code 2600, (216) 433-5818.
12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE
Unclassified -Unlimited
Subject Categories 07, 02, and 20
This publication is available from the NASA Center for AeroSpace Information, (301) 621--0390.
13. ABSTRACT (Max/mum 200 words)
A CFD methodology has been developed for inlet analyses of Rocket-Based Combined Cycle (RBCC) Engines. A full
Navier-Stokes analysis code, NPARC, was used in conjunction with pre- and post-processing tools to obtain a complete
description of the flow field and integrated inlet performance. This methodology was developed and validated usingresults fi'om a subscale test of the inlet to a RBCC "Strut-Jet" engine performed in the NASA Lewis lxl ft. supersonic
wind tunnel. Results obtained from this study include analyses at flight Mach numbers of 5 and 6 for super-criticul
operating conditions. These results showed excellent agreement with experimental data. The analysis tools were also
used to obtain In'e-test performance and operability predictions for the RBCC demonstrator engine planned fox testing inthe NASA Lewis Hypersonic Test Facility. This analysis calculated the baseline fuel-off internal force of the engine which
is needed to determine the net thrust with fuel on.
14. SUBJECT TERMS
Rocket based combine cycle; Computational fluid dynamics; Inlet
