AIAA99-4948 An Airbreathing Launch Vehicle Design with Turbine … · 2015. 9. 29. · Viable and practical airbreathing, hypersonic propulsion systems for these vehicles are a pri-mary

AIAA 99-4948An Airbreathing Launch Vehicle Designwith Turbine-Based Low-SpeedPropulsion and Dual Mode Scramjet High-Speed Propulsion

P. L. Moses, K. A. Bouchard, R. F. Vause, S. Z. PinckneyFDC / NYMA, Inc., Aerospace Sector, NASA Langley Research Center, Hampton, VAandS. M. Ferlemann, C. P. Leonard, L. W. Taylor lll, J. S. Robinson, J. G. Martin, D. H. Petley, J. L. HuntNASA Langley Research Center, Hampton, VA

9th International Space Planes and HypersonicSystems and Technologies Conference

and3rd Weakly Ionized Gases Workshop

November 1-5, 1999/Norfolk, VA

For permission to copy or republish, contact the American Institute of Aeronautics and Astronautics370 L'Enfant Promenade, SW • Washington, DC 20024

1American Institute of Aeronautics and Astronautics

AIAA-99-4948

AN AIRBREATHING LAUNCH VEHICLE DESIGN WITH TURBINE-BASED LOW-SPEED PROPULSION AND DUAL MODE SCRAMJET HIGH-SPEED PROPULSION

P. L. Moses, K. A. Bouchard, R. F. Vause, S. Z. PinckneyFDC / NYMA, Inc., Aerospace Sector, NASA Langley Research Center, Hampton, VA

L. W. Taylor III, S. M. Ferlemann, C. P. Leonard, J. S. Robinson, J. G. Martin, D. H. Petley, J. L. HuntNASA Langley Research Center, Hampton, VA

Abstract

Airbreathing launch vehicles continue to be a subject of great interest in the space access community.In particular, horizontal takeoff and horizontal landing vehicles are attractive with their airplane-like benefitsand flexibility for future space launch requirements. The most promising of these concepts involve airframeintegrated propulsion systems, in which the external undersurface of the vehicle forms part of the propul-sion flowpath. Combining of airframe and engine functions in this manner involves all of the design disci-plines interacting at once. Design and optimization of these configurations is a most difficult activity,requiring a multi-discipline process to analytically resolve the numerous interactions among the designvariables. This paper describes the design and optimization of one configuration in this vehicle class, a lift-ing body with turbine-based low-speed propulsion. The integration of propulsion and airframe, both from anaero-propulsive and mechanical perspective are addressed. This paper primarily focuses on the designdetails of the preferred configuration and the analyses performed to assess its performance. The integra-tion of both low-speed and high-speed propulsion is covered. Structural and mechanical designs aredescribed along with materials and technologies used. Propellant and systems packaging are shown andthe mission-sized vehicle weights are disclosed.

Nomenc lature

A/R Airbreather / RocketABLV Air Breathing Launch VehicleACE-TR Air Core Enhanced Turbine Ram-

jetAMHT All Moving Horizontal TailsAML Adaptive Modeling LanguageAOA Angle of AttackAPF Advanced Polyimide FoamATS Access to SpaceBSL Boeing Company, St. LouisCAD Computer Aided DesignCFD Computational Fluid DynamicsCG Center of GravityCTSC Combination Turbine Scramjet

CyclesDOE Design of ExperimentsDOF Degree of FreedomDMRS Dual Mode Ramjet/ScramjetDW Dry WeightGr/Ep Graphite EpoxyHTO Horizontal TakeoffHXRV Hyper-X Research VehicleIMI Internal Multiscreen InsulationIspeff Effective Specific ImpulseISS International Space Station

L/E Leading EdgeLOX Liquid OxygenOHR Oxygen-to-Hydrogen RatioP&W United Technologies Pratt & Whit-

ney DivisionPAI Propulsion-Airframe IntegrationPFA Propellant Fraction AvailablePFR Propellant Fraction RequiredPOST Program to Optimize Simulated

Trajectoriespsf Pounds per Square FootRBCC Rocket Based Combined CycleRCS Reaction Control SystemSCM Shape Control MembersSOL Shock on LipSSTO Single-Stage-to-OrbitTABI Tailorable Advanced Blanket Insu-

lationTOGW Takeoff Gross WeightTPS Thermal Protection SystemVTO Vertical Takeoff


Backgr ound

Single-stage-to-orbit (SSTO) space accessand atmospheric cruise missions are the subject ofexpanding activity within the hypersonic commu-nity. Viable and practical airbreathing, hypersonicpropulsion systems for these vehicles are a pri-mary focus of these efforts. Over the past 15 years,the U.S. Department of Defense and NASA hyper-sonic programs have spent over $3 billion advanc-ing the readiness of hypersonic technologies [1].New advances in high-speed propulsion, propul-sion-airframe integration (PAI), structures, materi-als, thermal management concepts, and advanceddesign methods have further contributed to readi-ness for design of such vehicles. This paperaddresses the integration of these technologiesinto a viable HTO SSTO configuration. This config-uration, as shown in Figure 1, is a lifting body with

an airframe-integrated, airbreathing propulsionsystem.

NASA’s Access-To-Space (ATS) Studyrevealed one promising configuration within theconfiguration matrix, for an HTO SSTO missionwith 25,000 lb. of payload [2]. Payloads of thisclass are required to service the InternationalSpace Station (ISS) in a 220 N.M. orbit at 51.6˚inclination. The configuration that appears mostpromising for HTO SSTO mission is the lifting bodywith an airframe-integrated, dual mode ramjet/scramjet (DMRS) high-speed propulsion. Furtherrefinement of this configuration and the exactnature of multi-mode propulsion required for thismission are objects of NASA’s current AirBreathingLaunch Vehicle (ABLV) Study. This configurationalso requires a low-speed propulsion system to getoff the ground and accelerate to the ramjet take-over speed, as well as some form of rocket and/orLOX augmented propulsion to transit from atmo-

spheric to exoatmospheric operation. Several con-figurations and propulsion systems were examinedfor HTO SSTO using the ISS mission in the currentstudy. These included winged-bodies, conical vehi-cles, high-fineness vehicles, podded engine con-cepts, and lifting bodies. From this effort, the liftingbody configuration stands out as having potentialfor development to meet the demanding HTOSSTO mission.

The single most critical technology for suchvehicles is the high-speed propulsion system,which is the dual mode ramjet/scramjet (DMRS).Recent progress in this area is illustrated byNASA’s Hyper-X Program, where autonomous,powered and unpowered flight at Mach 7 and 10will soon be demonstrated. The Hyper-X ResearchVehicle (HXRV) or X-43 will be boosted to the testconditions by a Pegasus based booster, separatefrom the launch vehicle, and perform the hydrogenfueled propulsion test in free flight [1]. Thisresearch vehicle is highly significant to the spaceaccess mission because the configuration, asshown in Figure 2, is very similar and employs the

same type of DMRS high-speed engine with similarairframe integration. Ground tests with a full-scale,X-43 flight engine have already demonstratedoperability and performance at or above predictedlevels, in the 8’ High Temperature Tunnel at NASALangley Research Center. Combining this efficienthigh-speed propulsion technology with appropriatechoices for the remaining propulsion modes is thekey to creating an efficient configuration for HTOSSTO.

Intr oduction

Reference Vehicle - ABLV Study

In the performance of a system study for con-figuration development and propulsion systemtrades, it is imperative to have a good baseline con-

Figure 1. Horizontal-Take-Off, Single-Stage-To-Orbit,Airbreathing Launch Vehicle with Airframe-Integrated

Propulsion Systems

Figure 2. Hyper-X Research Vehicle with Airframe-Integrated, Dual Mode Ramjet/Scramjet Propulsion


figuration or reference vehicle. This vehicle shouldbe relatively well characterized and be compatiblewith the configuration and propulsion systems tobe compared. The maturity of the reference vehicledesign and analyses must be such that the sensi-tivity to a given technology will be evident when it isinstalled. If the reference design is poorly substan-tiated, changes in propulsion technology may notshow significant impact to the vehicle. Likewise, thelevel of analytical effort applied to characterize anew propulsion system must be compatible withefforts applied to the reference vehicle. In the ABLVStudy, the reference vehicle chosen for comparisonof configurations and propulsion systems was anupdated version of the Airbreathing / Rocket SSTOconfiguration of the ATS Study [2]. This referencevehicle, ABLV-4, is a lifting body with a finenessratio of about 5.6, as shown in Figure 3. Through-

out this paper, configuration numbers will be usedto refer to various vehicles from the ABLV Study.They are simply a numerical designator with noother significance. The reference vehicle employsthe same propulsion system arrangement as usedin the ATS Study. ABLV-4, is a cold, integral-tankarchitecture, with advanced thermal protection sys-tems (TPS) and is fueled with slush hydrogen (50%solids / 50% liquid). The thick sidewalls, whichhoused the novel main landing gear installation,impose substantial drag penalties on the vehicleand as the size of the vehicle increases, the land-ing gear installation becomes a greater risk. Finally,there is a concern that the reference vehicle hastoo low a fineness ratio (relatively blunt configura-tion) and may be imposing higher drag losses thannecessary. The reference vehicle, ABLV-4, is cur-rently closed at a TOGW of about 1.07x106 lb.,without addressing the risks/issues explainedabove. The work reported in this paper seeks to

address these issues and mitigate the risksthrough appropriate design changes, during devel-opment of the preferred airbreathing HTO SSTOconfiguration, with turbine-based low-speed/DMRShigh-speed propulsion systems in an over/underintegration.

Figure of Merit

The design of an airframe-integrated propul-sion system and hypersonic vehicle is a very com-plex process, involving numerous physicalinteractions and requiring a multi-disciplineapproach. How then, does the vehicle designerdetermine a figure of merit to measure propulsiveefficiency? The answer has both simple and com-plex parts. The simplest figure of merit to demon-strate propulsive impact on a vehicle is the “closed”vehicle’s takeoff gross weight (TOGW) or dryweight (DW). A closed vehicle is defined as havingthe exact propellant fraction available (PFA) todeliver the propellant fraction required (PFR) of themission, or simply PFA=PFR. The choice betweenTOGW and DW is complex because DW typicallydrives initial cost more heavily, while TOGW drivesoperational cost and life-cycle-cost (including infra-structure) heavily. At the conceptual level of thesestudies costs are not yet being addressed, so mini-mum TOGW was chosen as the figure of merit.

The complex part of this “propulsion assess-ment” comes from the level of design and analysiseffort required to credibly compare a series of pro-pulsion systems integrated to the same configura-tion. Characterization of airbreathing, hypersonicvehicle performance is a multi-discipline process,requiring interaction of the design and analysis dis-ciplines to an unparalleled extent. Figure 4 illus-

Figure 3. ABLV-4 Reference Vehicle - HTO SSTOMission to International Space Station

Figure 4. Design / Analysis Process for HypersonicAirbreathing Vehicles


trates the level and type of interactions required toresolve vehicle performance, with numerous com-peting design variables that must be evaluated.

Propulsion System Selection

With minimum TOGW as the goal, variousmulti-mode propulsion systems were examined byintegrating them to the reference configuration forthe HTO SSTO mission. The choices consideredfell into three potential arrangements of combinedcycles (one flowpath, multiple modes) and combi-nation engines (multiple flowpaths, multiplemodes). The reference arrangement was an air-breather/rocket system like the A/R SSTO vehicleof Option 3 from the ATS Study[2], which was usedfor reference only and not discussed here. Thesecond propulsion arrangement studied, wasrocket-based-combined-cycle (RBCC), in which asingle flowpath functions in all modes from low-speed, to high-speed, to pure rocket. The third pro-pulsion arrangement was a combination of enginesand cycles with turbine-based low-speed, DMRShigh-speed, and external tail rocket. The combina-tion of turbine-based and DMRS was designatedCombination Turbine Scramjet Cycles (CTSC) andis the principle focus of this paper.

Selection of propulsion cycles for integrationinto a vehicle requires consideration of all impactsto the vehicle. One aspect of these considerationsis the amount of liquid oxygen (LOX) required to becarried by the vehicle. LOX is a very dense fluidand will increase average propellant density rapidlyas increasing quantities of LOX are carried with thelow density hydrogen fuel. This can have theimpact of increasing structural weight to carrygreater loads and increasing the planform loading,which will increase takeoff speeds of the liftingbody configuration. These distinctions for vehiclescarrying higher LOX fractions led to the consider-ation of propulsion cycles that emphasized air-breathing and minimized LOX usage.

Among those cycles requiring less LOX are theturbine-based machines such as the Pratt &Whit-ney (P&W) Air-core-enhanced Turboramjet (Ace-TR), a proprietary, high thrust-to-weight, turbinebased engine. This engine is one example ofadvanced turbine-based systems that are pro-jected to deliver uninstalled thrust-to-weight (T/W)ratios between 16:1 and 24:1. The early goal foruninstalled T/W ratio of the turbine-based systemwas 24:1, but this was later dropped to 20:1 to

reflect more near term technology. The higher T/W was used to make initial predictions of this typeengine’s impact on the configurations studied, butthe final configuration closure evaluated reducingthe uninstalled T/W to 20:1. Combining this typeengine with the DMRS high-speed system in anover/under arrangement appeared to be the bestairframe integration to study. To gage the potentialof this low-speed system integration, a derivative ofthe reference vehicle was developed with the Ace-TR low-speed engine, designated ABLV-4B, asshown in Figure 5.

The Ace-TR low-speed integration was jointlydesigned with the Boeing Company - St. Louis(BSL), where the installed low-speed performancewas generated. This vehicle was closed at1.04x106 lb. TOGW, with the same landing gearintegration as the reference vehicle, but using tri-ple-point hydrogen fuel. Thus, the first comparisonto the reference vehicle was favorable, but theother issues with the reference vehicle were stillpresent. A decision was made to develop a secondCTSC configuration with the main landing gearintegrated into the fuselage and a somewhathigher fineness ratio to reduce drag.

Configuration De velopment

Concept Definition

The remainder of this paper will focus on thestatus of the design for a new CTSC vehicle andthe design/optimization process employed. Thisprocess will be illustrated with the selected configu-ration, but issues with prior configurations will beaddressed as they are resolved in the new vehicle.The prior CTSC configuration, ABLV-4B, had rela-tively high drag through the transonic region, whichdrove the decision to adopt a slightly higher fine-ness ratio relative to the reference vehicle. A simi-lar configuration was developed with a finenessratio of 6.9 as compared to the reference vehicle at

Figure 5. ABLV-4B, Airbreathing Launch Vehicle - CTSCand Rocket Propulsion


5.6. A detailed geometry sized to 120,000 ft3 wasprepared for analysis and designated ABLV-9, asshown in Figure 6.

Initial results for this configuration indicatedthat a shock-on-lip (SOL) Mach number of 12would improve performance. This resulted in a newconfiguration with Mach 12 SOL, designated ABLV-9B. Maintaining consistent forebody and nozzleshapes enabled the design team to move aheadwith the existing ABLV-9 geometry for developmentof structural concepts and packaging (internal sub-systems arrangement). When vehicle geometrybecame available for the ABLV-9B keel line, the vol-ume and surface area distributions of ABLV-9would be adjusted to reflect ABLV-9B values, sothat weights and propellant fraction available (PFA)from the synthesis model would be corrected to thenew geometry.

The cold integral-tank architecture usinggraphite/epoxy (Gr/Ep) composite material wasretained, due to its high efficiency. This efficiencyresults from two characteristics. The first is the veryhigh specific strength and specific stiffness of theGr/Ep material. The other is the high volumetricpackaging factor for propellants, due to the confor-mal nature of the tank. An added contribution to theefficiency comes from the relative thinness of theadvanced cryo-insulation and TPS. This is illus-trated in Figure 7, which shows that a hot, non-inte-gral-tank concept packages less fuel for the samecross-section. The nose of the vehicle is a carbon-carbon composite structure with an actively cooledleading edge (L/E). The control surfaces are twinvertical tails with rudders and twin all-moving-hori-zontal-tails (AMHT), which are a hot structure con-cept with passively cooled L/E. The structures willbe more fully described in a later section. The

ABLV-9 configuration, compared to the ABLV-4 ref-erence vehicle, has the increased fineness,reduced body width, and smaller sidewalls.

Geometry Definition

Geometry definition starts with the missionrequirements. Horizontal or vertical orientation fortakeoff and landing are major considerations thatdrive shaping and landing gear location. Shape isimpacted by many competing considerations, suchas propulsion capture area, payload packaging,and minimized aerodynamic drag. Thus, a clearand well defined mission is required to start config-uration development. The mission, cost targets,and technology goals help to set ground rules forthe design, including minimum included angles,thermal protection systems, structural concepts,fuel and oxidizer, crew, payload, gear, reusability,materials, maximum takeoff speed, cruise range,and others. The HTO SSTO mission with maximumairbreathing propulsion is a demanding designchallenge. Relatively slender lifting bodies with air-frame-integrated propulsion have proven an effec-tive choice for hypersonic missions. The lifting bodyconfiguration is primarily driven by design of thehigh-speed airbreathing flowpath. The flowpath iscomposed of the entire undersurface of the vehicle,including forebody inlet ramps, centerbody enginewith matching cowl, and aftbody nozzle surfaces.The first step is definition of a 2D keel line (longitu-dinal shear view). This design process is iterativein nature, requiring performance computations todetermine when acceptable 2D operability and per-formance have been achieved. This process isaided by a modeling tool generated with AdaptiveModeling Language (AML) software [3]. The AML

Figure 6. ABLV-9 CTSC Configuration - 20% HigherFineness Ratio than ABLV-4

Integral Concept - 94% Used for Fuel

Hot Non-Integral - 89% Used for FuelFigure 7. Comparison of Cold Integral Tank Concept

with Hot Non-Integral Tank Concept


tool is used for initial geometry construction and 2Dpropulsion analyis input deck generation. After anacceptable flowpath is defined, a three-dimen-sional vehicle geometry is wrapped around theextruded and trimmed 2D flowpath, using the Pro/ENGINEER computer-aided-design (CAD) tool [4].This configuration forms the basis for packaging,thermal management, and structures disciplineefforts.

The new configuration will be ABLV-9B and willfollow the same geometry design process as theoriginal configuration. In general, the fuselage mustbe shaped and sized to enclose required payloadsand fuel/oxidizer volumes. The objective is to cre-ate, as closely as possible, a two-dimensional inletfor uniform flow into the engine and uniform flowonto the external nozzle from the engine exhaust.Design rules are followed for minimum includedangles for cowl, sidewalls, and fuselage nose.Sidewall thickness for the ABLV-4 series of configu-rations was driven by the novel main landing gearinstallation, but for the ABLV-9 series the gear wasmoved into the fuselage volume. The aftbody sidesurfaces provide a flat wiping plane for the AMHT.Vertical control surfaces are fixed with trailing edgerudders. Linear aerospike rockets are integratedinto a bump on the aftbody upper surface. Thisrearward bump location is shadowed from forebodyflow and provides volume to package rocket sys-tems. The lower aftbody nozzle surface is trimmedfor the rocket nozzles. This first geometry isreferred to as the “as-drawn” vehicle.

Vehicle Configuration Control

When the as-drawn vehicle is complete, aero-dynamics, packaging, computational fluid dynam-ics (CFD), thermal management, structures, andother disciplines begin their analyses using modelsderived from these master surfaces. It is imperativethat all disciplines are working the same configura-tion geometry (configuration control) or at least,fully understand any disconnects accepted in theprocess. If great care is not exercised in configu-ration control, closures can have significant errors.With a common configuration database, data canbe shared confidently between disciplines, usingcommon coordinate systems and known assump-tions that are referenced to a single vehicle. Afterthe as-drawn vehicle analysis is completed and tra-jectory simulation performed, closures may be per-formed to produce an “as-flown” vehicle. If the as-flown vehicle does not meet mission requirements,

revisions may be made to address performanceproblems. This is a negotiated process among thedesign disciplines that requires compromise andconcession to yield an improved configuration.Revised configurations are analyzed and the pro-cess is repeated until a satisfactory closure isobtained and the as-flown vehicle meets allrequirements.

Trajector y Anal ysis

Analysis and Modeling

For this study, the vehicle mission is the inser-tion of a 25000 lb. payload into a ISS orbit (220N.M. circular orbit, 51.6 degrees inclination). Thetrajectory and vehicle optimization were completedusing the 3-Degree of Freedom (DOF) Program toOptimize Simulated Trajectories (POST) [5]. Trajec-tory simulations were performed with the vehicleaerodynamically trimmed in the pitch plane. Thegoal of the trajectory optimization is to perform themission for the minimum vehicle TOGW. A closedvehicle’s TOGW is determined by inputting the mis-sion PFR and propellant LOX fraction, from the tra-jectory analysis, into the vehicle closure model.The closure model is more fully described in a fol-lowing section. The closure model then scales theas-drawn vehicle to a size and weight that matchthe trajectory requirements. Using the vehicle clo-sure model, a table of TOGW as a function of PFRand LOX fraction was generated and imported intothe POST simulation as the optimization function.This provides the optimizer with a direct impact onTOGW as it trades thrust and Isp levels for the var-ious propulsion systems, as well as varying otherparameters. Figure 8 describes the trajectory interms of altitude versus Mach number and illus-trates the various segments of the trajectory char-acterized by each propulsion mode.

Propulsion Modes

A typical airbreathing hypersonic vehicle oper-ates in multiple engine cycles and the CTSC inte-gration is no different in that respect. In low-speed(Mach 0-4) the vehicle is powered by a bank ofAce-TR engines, which themselves have multiplecycles/modes (subsonic turbine operation, super-sonic turbine/ramjet operation). This mode transi-tion is transparent to the trajectory design becausethe engine operation and performance are repre-sented purely by a single matrix of thrust and pro-pellant flowrate, as functions of Mach number andangle of attack (AOA). Also in this Mach range, the


optimizer is given the option of using the linearaerospike tail rocket for additional thrust. This isgenerally the case at takeoff and in the transonicregion, where the vehicle drag increases signifi-cantly and additional thrust is usually required toaccelerate through and beyond Mach 1.

The next propulsion mode is main engine ram-jet/scramjet operation. Again, the transition fromramjet to scramjet mode is accounted for in thedata table and is transparent to the trajectory opti-mization. Following pure scramjet operation, a mixof LOX augmented scramjet and tail rocket opera-tion is employed. Here, the optimizer trades thehigh Isp of the pure scramjet for a higher thrust ofthe internal LH2/LOX rocket motors firing into themain flowpath in the LOX augmented mode. Theoptimizer again has the option of using the tailrocket as necessary. It is during this mode that thepull-up maneuver usually occurs, requiring highthrust and necessitating the use of the tail rocket.During the pull-up, when the dynamic pressure fallsbelow the level that will sustain main engine com-bustion, the main engine is shut down and the vehi-cle relies solely on the tail rocket for the remainderof the flight.

Although there are several propulsion modetransitions that are transparent to the trajectoryoptimization, there are several that play key roles inreducing the vehicle’s TOGW. The Ace-TR systemrequires a substantial amount of variable geometry,especially in the Mach 3.5 and higher range. Therequired amount of variable geometry directlyimpacts the weight and volume of the system,

effecting the entire vehicle’s size and weight. Thus,the transition Mach number from Ace-TR to high-speed engine operation plays an important role inthe overall vehicle optimization. The trade-off forlowering this transition Mach number comes in apotentially larger ramjet to deliver added low-speedthrust and reduced performance by the scramjetengine in the higher Mach range (above Mach 12).This optimization is not addressed by the currenteffort. The second transition that the optimizeruses is the switch from pure scramjet to LOX aug-mented scramjet operation. The optimizer is alsogiven control over the oxygen-to-hydrogen ratio(OHR) of flowpath rocket motors. In addition, theoptimizer may again use tail rocket as required. Italso has indirect control of when to shut off themain engine by moving the pull-up Mach number(recall that the main engine is shut off at a limitingdynamic pressure).

Guidance

In addition to the aerodynamic and propulsivecharacteristics of the vehicle, the simulation toolrequires a guidance scheme(s) in order to simulatethe trajectory. The guidance schemes employedmust be compatible with the vehicle design andmission requirements. Figure 9 shows trajectory

dynamic pressure versus Mach number and illus-trates where various guidance schemes are opera-tive, as described below. In the first phase of theflight, the guidance routine follows a flight pathangle derivative profile (with respect to time), con-trolled by the optimizer. After the vehicle has estab-lished supersonic flight, the guidance scheme

Figure 8. Trajectory Altitude vs. Mach Number - Rangesof Propulsion Mode Operation

Figure 9. Trajectory Dynamic Pressure vs. MachNumber - Ranges of Guidance Schemes


switches to a dynamic pressure following routine.The cruise dynamic pressure for this vehicle isgenerally around 2000 psf. Again, this profile iscontrolled by the optimizer. This guidance modetakes the vehicle up to the point where the opti-mizer decides to augment the main scramjet withLOX augmented scramjet performance or switch topure tail-rocket thrust. From this point forward (untilorbit insertion), the vehicle is controlled by an AOAprofile as a function of Mach number, again con-trolled by the optimizer.

Optimization

In summary, the optimizer has control over thefollowing parameter profiles: low speed and highspeed tail rocket throttling tables, a flight path anglederivative table, a dynamic pressure table, and apull-up AOA table, all as functions of Mach number.This optimization process typically results in 30-40independent variables for the optimizer to control,depending on the size of the tables. Additionally,constraints are added from the various disciplines.Structural constraints impose a normal accelera-tion limit of 2.5 g’s and a total acceleration limit of4.0 g’s. Thermal restrictions limit the maximumdynamic pressure to 2200 psf., as well as limitingthe AOA to less than 5 below Mach 18 and lessthan 8 above Mach 18. In the LOX augmentedscramjet mode, the OHR is limited to between 1and 6. Also, as stated earlier, propulsive operabilitydictates that the scramjet engine be cut off at 200psf. dynamic pressure.

Aerodynamic and Aer othermod ynamicPerformance

The aerodynamic database can be split intotwo distinct pieces: low speed and high speed. Thereason for the division lies in the methods used tocalculate the aerodynamic phenomena. The highspeed portion of the database corresponds roughlyto speeds greater than Mach 3. At these velocities,lift and drag are modeled successfully by engineer-ing codes such as APAS [6] and SHABP [7] forarbitrary configurations. For these programs, theinviscid pressure forces are predicted by impactand shadow methods such as modified newtonian,tangent cone, tangent wedge, and Prandtl-Meyerexpansions. However, for compression surfacescareful attention must be paid to the initial flow con-ditions for increasing angle surfaces as flow movesdownstream. Skin friction results are generatedthrough reference temperature and referenceenthalpy methods. Engineering codes are powerful

design tools in the conceptual design stage. Theiruse in the early stages of design enables higherlevel CFD and wind tunnel analyses to be per-formed on more mature vehicle concepts. CFD andexperimental work can also be used to capture flowfeatures, such as flow separation regions, whichmay not be predicted by lower order methods.

Low speed aerodynamics are more difficult toobtain in part due to the strong nonlinear aerody-namics associated with hypersonic lifting body con-figurations. Solutions for arbitrary shapes are notreadily available and higher order methods mustoften be used to achieve adequate fidelity. For thisreason, the low speed aerodynamic database isanchored to an historical database for a similarconfiguration. The historical database was createdfrom a combination of engineering code, CFD, andwind tunnel testing for a very similar configuration.A limited number of higher order analyses can thenbe used to tailor the historical database to thepresent configuration.

Because of the nature of an airframe-inte-grated scramjet powered vehicle, the distinctionsbetween traditional aerodynamics and propulsionbecome blurred. The entire lower surface of thevehicle becomes part of the engine flowpath.Determining which discipline should be responsiblefor which areas on the vehicle is critical. Great caremust be taken when accounting for forces andmoments on the vehicle. For the class of vehicledescribed in this paper, cowl-to-tail accounting isused. Cowl-to-tail accounting gives the forebody,upper surface, chines, external engine sidewalls,wings, tails, and wiping planes to the aerodynam-ics discipline, while the external cowl, engine inter-nals, and nozzle are totaled in the propulsionforces and moments.

Aerothermal load generation on vehicle acre-age is well modeled through Reynold’s analogiesfor ideal or real gases using codes such as APASand SHABP. Blunt body heating can be solvedthrough the use of other codes such as StagHeat,an in-house engineering program, which employsadjusted Fay-Riddell correlations for ideal or realgases. While these methods cover the majority ofthe vehicle surface, airbreathing configurationshave particular thermal challenges. Shock-shockinteractions on the engine cowl leading edge, cor-ner flows such as the vertical tail/AMHT junction,and gap heating between the AMHT and verticaltail are phenomena which can easily increase the


predicted heating by an order of magnitude. Quan-tifying the heat augmentation is difficult because itis heavily configuration and attitude dependent. Forthis reason, multipliers can be obtained from litera-ture surveys for similar configurations and appliedto heat loads for initial thermal analyses, but shouldbe verified through wind tunnel testing or othermeans.

Propulsion System

Low-Speed System

The low-speed system selected for initial studywas the P&W Ace-TR, a proprietary developmentof the Pratt & Whitney Aircraft Engine Company.For this reason, the specific design details and per-formance parameters are not disclosed in thispaper. Because it provides a good example ofadvanced turbine engines, the P&W engine cycleperformance was used in computing the installedperformance of the system. However, other tur-bine-based engines will be considered in thefuture. Figure 10 shows the installation of this

engine, as applied to the ABLV-9 configuration. Sixof these engines were installed in the ABLV-9 con-figuration, with three 2-D common inlet and nozzlestructures serving each pair of Ace-TR engines, asshown in Figure 11. Each engine is served by itsown rectangular-to-circular transition section that issplit off the common 2D inlet or nozzle. Thisarrangement was chosen to integrate compatiblywith the three high-speed engines, which eachhave two side-by-side flowpaths, for a total of six.Because of structural integration requirements withthe airframe, the six Ace-TR engines are evenlyspaced across the vehicle, on-center with the high-

speed flowpaths. This will allow for similar servicingand installation procedures for both low-speed andhigh-speed engines.

Development of this installation was a jointeffort between NASA LaRC, P&W, and BSL. P&Wsupplied BSL with the Ace-TR engine airflowdemand curve and Boeing designed the inlet tomeet that demand. The engine analysis performedat BSL accounted for inlet operability and perfor-mance using 2D methods. Operability issues are akey element of the design, since the variablegeometry inlet must receive supersonic flow andsupply subsonic flow to the Ace-TR, with relativelylow distortion. For this reason, an isolator/subsonicdiffuser must be designed that is long enough toproduce subsonic flow across the entire Machnumber range of the low-speed system. The inletrecovery developed at BSL was used in a P&WAce-TR cycle deck (computer code) to get installedengine performance. The nozzle characteristicsare relatively well known, so the installed perfor-mance can be computed using an estimated noz-zle gross thrust coefficient (CFG).

High-Speed System

As previously described, ABLV-9B is a Mach12 SOL version of the ABLV-9 configuration. Devel-opment of the Mach 12 keel line geometry startedwith an extension of the existing, final forebodyramp to make the engine deeper. Several valuesfor the inlet final ramp length were investigated toproduce a cowl inlet length that was somewhatshorter than the ABLV-4 cowl inlet. A matrix of pro-pulsion performance computations was made andtabulated as functions of Mach number, angle ofattack, and dynamic pressure. Also included werecases for LOX augmented scramjet performance.These data were then utilized in the POST trajec-tory optimization and vehicle performance assess-

Figure 10. Ace-TR Low-Speed System Integration -Shear View

Figure 11. Ace-TR Low-Speed System Integration - 2DInlet and Nozzle with Transitions

Inlet

Inlet Turbine

Nozzle

Nozzle

VariableGeometry

Transition Core

Transition

VariableGeometry


ment. Thus, the closures generated for the ABLV-9B vehicle will be made with appropriately com-puted data, from the 2D performance code. While3D effects on propulsion are a concern, theassessment would be premature at this stage ofdesign. When making multiple trades of propulsionsystems in a study, an exhaustive CFD study isunwarranted in terms of cost and time. As long asthe configurations are similar, as they are for thecurrent study, the 3D impact will not be a discrimi-nator. If two promising systems were to be nearlyidentical in performance, then an assessment for3D effects may be in order, but only if the 2D per-formance showed potential in the vehicles.

External rocket

The external rocket system utilized on theseconfigurations was a linearized version of the aero-spike rocket engine concept. This engine conceptintegrates small rocket combustors around anannulus exhausting over an axi-symmetric nozzle.The linearized version of this technology unwrapsthe annular combustor arrangement into a 2Darrangement, with symmetric upper and lowercombustors exhausting over a symmetric 2D noz-zle, as shown in Figure 12. The upper and lower

sections are broken into a number of independentsegments that are arranged across the width of theengine. The installation of the tail rocket is orientedto place its thrust vector through the CG of thevehicle. In addition, each segment is independentlyoperable and “throttleable”, giving the ability to pro-duce asymmetric thrust for trimming or control ofthe vehicle. This capability is in addition to the vehi-cle’s reaction control system, which is designed pri-marily for exoatmospheric operation. With all

segments in operation, the gross thrust is270,000lb at 458 sec. Isp (vacuum). This rockettechnology was predicted to deliver a goal thrust-to-weight ratio of 77:1, which is used to obtain thescaled rocket weight in the synthesis model.

Structure

Structural Concepts

The structural concept used for the ABLV vehi-cle is shown in Figure 19. Surface panels transferairload and propulsion pressures to an underlyinggrid of longitudinal keel beams and transversebulkhead beams. Longitudinal bending moment iscarried primarily by the top and bottom skins withthe keel beam webs reacting the associated trans-verse shear. In areas where the skin panels arediscontinuous (payload and engine compart-ments), longitudinal bending is resisted by coupleforces in the top and bottom caps of the keelbeams. The transverse bulkhead beams carryshear and bending in the inboard-outboard direc-tion and also function to reduce the buckling lengthof compression-loaded skin panels. Bulkheads arealso important in distributing concentrated loadssuch as landing gear forces. Between bulkheadsare panels designated as shape control members(SCM) which function in pressurized compart-ments to limit skin lateral displacements. In addi-tion to aerodynamic concerns, lateral skindisplacements must be minimized to prevent in-plane loads from generating additional skin bend-ing moments. Surface loads on the horizontal arecarried by a skin-spar system with the taperedspars originating at the spindle attachment. Thevertical control surface is also formed with a skin-stiffener system. This structural concept is exceed-ingly stiff and can carry axial load, bi-directionalbending, and torsion very efficiently.

The planned upgrade to structural sizing meth-odology is described in the following sections, butis not yet fully functional. The major advantage tothis process is that it is a “bottoms-up” analysis andsizing for the actual airframe loads. It will notrequire a vehicle density correction factor forincreased LOX load. Unfortunately, other demandshave prevented completion of this effort. Therefore,the unit weights and structural methodology usedin the closure model remain the same as for thereference vehicle, ABLV-4. However for these vehi-cles, the LOX fraction stayed near the referencevalue and planform loading never exceeded thetakeoff limit. Thus, the density correction factor on

Figure 12. Linear AeroSpike Rocket Engine Integrationin Tail Section of Vehicle

Thrust

NozzleFlow

Cells

RampFence


structure and takeoff penalty were never an issuefor these vehicles. While the absolute values ofweight may have some uncertainty, the relativetrends will be correct.

Structural Geometry and Elements

Initial ABLV surface geometry was verydetailed in order to be suitable for aerodynamicmodeling. In order to predict structural unit weights,a simpler geometry definition was needed to gen-erate an internal structure and to obtain quick esti-mates of lengths, surface area, volume, and massproperties. The base structural geometry was cre-ated by passing planes through the original modelat selected bulkhead stations (Figure 13). The

planes are located where the geometry changesabruptly or where concentrated loads will beapplied to the vehicle structure. A simplified cross-section provides reference points for calculatingcross-section width, depth, and center of gravity.All external and internal structural elements aregenerated from the base cross-section geometry.Bulkheads are located at the base cross-sectionstations while the spacing of intermediate shapecontrol members and keel beams can be varied inthe geometry generating code. The initial ABLVgeometry and a simplified structural geometry areshown in Figures 14 and 15 . The elements shownin Figure 15 are actually a NASTRAN model writ-ten by the geometry generator.

The structural elements which are tracked inthe sizing and weights procedure are shown in Fig-ure 16. Keel beams, bulkhead beams, and sidebeams are all formed from webs and flange caps.

The beam caps carry axial and bending loads(couple force) while the webs mainly resist shear-ing loads. The skin panels are shown as a simpleplanar structure but these elements are assigned askin-stiffener arrangement in the sizing procedureto generate appropriate membrane, bending, and

Figure 13. Structural Arrangement

Figure 14. Simplified Geometry for Structural Models

Figure 15. Structural Geometry

Figure 16. Section Structural Members


shear stiffnesses at the modeled reference surface.A single solid element is generated within eachcompartment to simplify mass distribution for themodel.

Structural Loads and analysis

Structural load cases for the ABLV HTO SSTOvehicles are tied closely to points on the trajectoryanalysis. Initial load cases are based on trajectorypoints for maximum axial acceleration, maximumnormal acceleration, and the maximum vectorcombination of axial and normal accelerations.Also associated with the trajectory load case pointsare integrated airloads, integrated propulsionloads, accelerations, and control surface trimloads. Pressure distributions for high-speed air-loads and propulsion loads are defined by engi-neering aerodynamic codes and engineperformance codes. However, maximum vehiclebending loads often occur during a pull-up maneu-ver at low speed (less than Mach 3). Airload andpropulsion pressure distributions at low speeds areusually not well defined during the early stages ofsystems analysis and integration. Pressures on theengine ramps and nozzle are important to paneldesign, as well as to overall shear and bending ofthe fuselage structure. For low speed cases,approximate pressure distributions on the lowerforebody and nozzle areas were generated fromtrajectory forces and moments. Aerodynamic nor-mal force and reference pitching moment wereused to solve for a pressure distribution with con-stant and quadratic terms. Axial aerodynamic forcewas matched using the force component from theramp pressures and a drag contribution from thevehicle upper and side surfaces. Nozzle pressuresalso use constant and quadratic terms based onpropulsion moment and thrust requirements. Thevehicle free body with relevant pressure distribu-tions is illustrated in Figure 17. The distributedpressures are applied to the ramp and nozzle pan-els and then combined with other forces to checkfor dynamic equilibrium. The accelerations solvedfor using the distributed pressures are compared toaccelerations from the trajectory analysis. Goodagreement between the two measures of accelera-tion are required before the load case definitionproceeds.

Loads on the surface panels are mapped to aline model at the center of a station segment asseen in Figure 18. This procedure replaces all sur-face loads with an equivalent set of loads on a line

model which extends from nose to tail. A similarprocedure is used for mapping the distributed iner-tial loads to the line model. Checks are made toverify that the distributed loads on the line modelproduce the same set of integrated forces thatwere used in the trajectory analysis. The linemodel should be in dynamic equilibrium under acombined set of airloads, propulsion loads, inertialloads, and control surface loads. Internal axialforce, shear force, and bending moment at any sta-tion point can be determined from standard equa-tions. Repeating this procedure for the chosen loadcases produces a useful set of internal force enve-lopes. Panel pressures and forces can also be writ-ten to the NASTRAN model shown in Figure15 fora finite element analysis. The inertia relief capabil-ity of NASTRAN is used to solve for inertial loadswhich will be in equilibrium under applied airloads,propulsion loads, and control surface loads. Loadsdevelopment was tightly integrated through allthree levels of structural modeling for ABLV. Con-sistency was maintained when moving from a

Figure 17. Vehicle Free Body for Flight Loads

Figure 18. Equivalent Segment Loads for Line Model


lumped mass and force trajectory model to a dis-tributed mass and force line model or three-dimen-sional panel and beam model.

Structural Sizing

Internal forces from the line models describedabove are used for first order structural sizing ofthe central tube formed by the skins and keelbeams. Graphite-epoxy stiffness values and strainallowables are used with the applied forces to cal-culate required section thicknesses, using classicalengineering methods. Beyond the first order pre-diction described above, formal sizing of the struc-tural elements may be carried out with theHyperSizer code from Collier Research Corpora-tion [8]. Specific HyperSizer cross-sections andmaterials are assigned by the user to all panel andbeam element groups. HyperSizer’s materiallibrary includes composites and metals and therange of panel cross-section types includes sand-wich sections, hat-stiffened sections, biaxial bladestiffened sections, and grid-stiffened sections.Beam cross-section types include most open andclosed sections common to aircraft construction.Within each cross-section description are a groupof variables such as stiffener width, depth, andspacing. The user assigns realistic ranges to eachof the design variables. HyperSizer submits the jobto NASTRAN for the load cases described above.Internal membrane and bending forces are thenpassed back to HyperSizer, which chooses thecross-section variables that produce the lowestsection weight and satisfy all strength and stiffnesscriteria. Updated NASTRAN stiffnesses can bewritten out to see the influence of the new sectionstiffnesses on load distribution. If there are signifi-cant changes in load distribution, then additionaliterations of the analysis cycle are performed withthe new load distribution and revised stiffnessesuntil the process converges (output weights don’tchange).

Structural Weights

The structural weights determined by analysisare summed up for each section of the structuralmodel. (Recall that the structural geometry is aclose approximation of the actual vehicle geome-try.) Structural unit weights for each section of thestructural model are generated using lengths andsurface areas of the model. These unit weights arepassed to the synthesis tool for integration over theactual geometry of the configuration, insuring that

all parts of the real geometry are assigned realisticunit weights. Recall that the status results are stillfor the ABLV-4 structural unit weights.

Thermal Mana gement

Thermal Protection Systems

A thermal protection system (TPS) is used onthe vehicle at any location where the aerodynamic,propulsion or other heating source will cause thestructural temperature to be exceeded and/or thethermal stresses would be too high. A passive ther-mal protection system is used where possible,because it weighs less than active cooling panelswith associated system weight. All of the vehiclesin the current study have integral hydrogen tanks,which makes the design of the TPS especiallychallenging. A layer of cryogenic insulation is usedunder the TPS in areas directly over a cryogenichydrogen tank. These layers of insulation must sur-vive an extreme temperature range, providingenough insulation to protect the structure and con-trol heat gain to the tank to minimize boil-off ofhydrogen. The conditions in the tank are shown inFigure 19 for tank options of normal boiling point,

triple point, and slush hydrogen.

The disadvantage of normal boiling pointhydrogen is that any added heat to the tank imme-diately boils the liquid into a vapor. The excesspressure due to this boiling will either cause imme-diate venting of gaseous fuel overboard or drive uptank weights in order to contain higher pressure.Neither condition is desirable for achieving opti-mum vehicle performance. A super cooled fuelsuch as slush or triple point hydrogen is desirablefor improved vehicle performance due to thereduced vapor pressure and added heat sink,which enable a lower tank design pressure and

Figure 19. Tank Pressure for Hydrogen Options


delay the time when venting must begin. Anincrease in the TPS/cryogenic insulation isrequired to reduce the heat load to such tanks,which increases vehicle weight somewhat. How-ever, the net effect is an advantage for the vehicle,because the slightly higher density of these liquids,over normal boiling point hydrogen, enables asomewhat greater fuel load in a given volume. Ofthese two fuels, slush has the marginal advantagefor vehicle design due to having 50% solids in themix, but the disadvantage of slush is that it requiresa lot of ground support infrastructure and addedenergy costs to produce and maintain the fuel.Slush also ties flight operations to only those loca-tions where the infrastructure is established. Forthese reasons triple point hydrogen is the currentchoice for fuel.

TPS sizing was accomplished using heat loadsdeveloped with engineering computer codes. TheAPAS aerodynamic panel code was used to obtainthe heat loads for the airframe. Where passive TPSis not adequate, such as inside the engines, activecooling means were used. Internal engine heatloads are obtained from the 2D propulsion analysiscode. These heat loads are used to perform amore detailed analysis and sizing of the coolingsystem and to verify the energy balance computedby the 2D propulsion analysis code. Unit weightscomputed from such analyses are used whereactive cooling is required.

Passive TPS

An advanced TPS system concept using Inter-nal Multiscreen Insulation (IMI), shown in Figures20 and 21 , is used on the lower surface of these

configurations. A carbon silicon carbide (C/SiC)

heat shield is used on the hot face, which has atemperature capability of 3000F. IMI is packagedbeneath the heat shield and consists of layers ofceramic fiber insulation contained between layersof metallic screen. The screen resembles tissuepaper in appearance. Platinum is used on the hotside because of the higher temperature capability.Gold is used toward the cold side, it has better radi-ation blocking performance but lower temperaturecapability than the platinum. This concept has apurge system integrated into the heat shield sup-ports. The support tubes are attached directly tothe graphite epoxy tank and penetrate theAdvanced Polyimide Foam (APF) insulation, whichis chemically bonded to the graphite epoxy tank.The foam is the cold boundary for the IMI. Duringground hold, warm dry nitrogen gas is suppliedthrough a ground support umbilical and exhauststo the atmosphere at a minimum temperature of40F, which will prevent frost formation on the exter-nal surface. The flow rate of purge gas is sized tomaintain the foam-to-IMI interface location abovethe liquefaction temperature of air (-298F). Thecontinuous supply of dry nitrogen purge will alsokeep water vapor out of the insulation system whileon the ground. Any fuel boiled off due to this heat-ing is made up by GSE, prior to disconnect fortakeoff.

Tailorable Advanced Blanket Insulation (TABI),shown in Figure 22 , is used on the upper surfaceof the vehicles. The upper surface has significantlyless heat load than the lower, which is why TABI isthe best choice. Insulation concepts that have con-tainment systems prove to be heavy at small thick-ness because the heat shield and support systemweights are almost independent of thickness. TheTABI insulation system consists of a surfacecoated, fiber mat containment system with triangu-

Figure 20. Internal Multiscreen Insulation (IMI) -Integrated Hot Gas Purge

Figure 21. Advanced TPS


lar, alumina prisms inside. A polyimide coatingmust be applied to the TABI so that it can be chem-ically bonded to the APF insulation. The polyimidechemical bond is important because it has a highertemperature limit (up to about 550F) than an RTVbond (up to about 400F), which was used in priorTPS concepts. The APF is also chemically bondedto the graphite epoxy tank. The graphite epoxy islimited to a temperature of 250 F, but this limit wasnot a controlling factor in the design. A purge gassystem is integrated into the insulation to preventliquefaction of air and frost formation while on theground. Warm, dry nitrogen gas is circulatedthrough the insulation system in the same manneras described above.

Passive TPS Sizing

The analysis and sizing of passive TPS wasaccomplished using the SINDA-85 finite difference,thermal analysis code [9]. The process uses ahighly specialized set of input files containing logicfor sizing of insulation based on iterative transientanalysis. TPS was sized for each of the panels inthe APAS aerodynamic model through an auto-mated process that uses heat loads computed byAPAS. Geometric models containing more than1000 panels have been run for these configura-tions, with different analysis models used for sur-face locations where there is a cryogenic tankversus locations where there is no tank. The analy-sis model for passive TPS over a cryogenic hydro-gen tank is shown in Figure 23 . The aerodynamicheat load was applied to the C/SiC heat shield onthe outer mold line and the hydrogen in the tankprovides a heat sink on the inside surface of themodel. A transient analysis of the system was runfor the complete ground hold and flight mission.The analysis started with a minimum insulation

thickness on the first iteration and proceeded toincrease the insulation on subsequent iterations,until the temperature requirements of the designwere satisfied. TABI and IMI were sized for theHTO SSTO mission. These results were computedon the as-drawn vehicle geometry of ABLV-4 andthe unit weights applied in the vehicle closure pro-cess for all configurations. The unit weight resultsfor the upper surface averaged 0.818 psf for theTABI and APF foam system. On the lower surface,the unit weight averaged 1.593 psf for the IMI andAPF foam system. When area-averaged over allcovered surface the TPS unit weight was 1.117 psf.

Active Cooling

Active cooling is used on inlet ramps, insidesurfaces of the engine, external nozzle, and lead-ing edges of the fuselage and engine inlet cowl.For these vehicles, hydrogen was used as the cool-ant, which flows through the cooling panels andleading edges. Cooled leading edges and coolingpanels are typically made of copper alloy or hightemperature super alloys such as molybdenum-rhenium or Haynes 188 alloy. Using the hydrogenfuel as a coolant enables the heat load input to thevehicle to be used regeneratively for energy todrive fuel system turbopumps. The required energybalance must be iterated between the propulsionperformance code and the thermal analysis ofengine cooling. The weights for cooling panels andsystem components used for the high-speedengines are supported by prior analyses for thismission. Mostly passive TPS is assumed for thelow-speed system. However, fuel cooling will berequired for some low-speed system components.A complete cooling system analysis must be per-formed on the final low-speed system design, in

Figure 22. Tailorable Advanced Blanket Insulation(TABI) with Integrated Purge

Figure 23. Analysis Model for Advanced TPS


order to size the components and balance coolantflow rates. Estimates of these component weightswere used to determine the installed weight of thelow-speed systems in this study.

Airframe Subsystems

The subsystems used in current airbreathingvehicle analysis are based on previous resultsdone to a detail design level. Extensive designstudies for prior vehicles yielded detailed systemlayouts, schematics, and master equipment lists.This work was reduced to a set of airframe systemalgorithms for predicting system weight and vol-ume of the installed equipment. This data wasupdated during the ATS Study and scaled to opera-tional vehicle size (fuselage volume of 100,000 ft3).Except for some minor updates for the ABLV Study,these systems carry that same detailed design her-itage into the current effort. The system require-ments, vehicle sizes, and vehicle weights for thecurrent effort are assumed to be similar enough toprevious work to warrant using these data. This isfurther justified, based on continued similarity ofsystem content and functionality for the currentconfiguration.

The general arrangement of the ABLV-9 config-uration is shown in Figure 24. Note the low-speed

system inlets forward of the high-speed engine andthe main landing gear moved out of the sidewalls.Systems layout is described in Figure 25. Payloadintegration over the low-speed system is shown inFigure 26. The following sections briefly describesome of the more important airframe subsystems.

Propellant and Pressurization Systems

The fuel system utilizes triple point hydrogenfor all engine modes. A schematic of the system isshown in Figure 27 . Each tank has a pressure

transducer, a temperature transducer, a liquid leveltransducer and a liquid point transducer. Each tankis also connected to a vent system manifold thatcontrols ullage pressure. The mixer and spray sys-

Figure 24. ABLV-9 General Arrangement

Figure 25. Systems Layout

Figure 26. Payload Integration

Figure 27. Propellant System

LiquidNose Landing

Payload

Triple Point Main Landing

RocketOxygenGear Bay

Hydrogen Gear Bay

Actuator s

Payload

Low SpeedSystem

Main LandingGear

Ramjet /Scramjet

ScramjetSystems


tem are required to maintain a constant tempera-ture in the tanks, in order to prevent a pressurecollapse that would occur from splashing of coldliquid into a warmer ullage area. The tank tur-bopumps provide working net suction pressure tothe main engine turbopumps, which in turn powerthe tank turbopumps after the system is started.The exhaust from the tank turbines is returned tothe supply manifolds, along with the hydrogen sup-plied by the pumps.

Atmospheric oxygen is used as the oxidizer forboth the low-speed engine, as well as the DMRShigh-speed engine. Separate liquid oxygen (LOX)tanks are integrated into the vehicle to operate theLOX augmented scramjet mode, the linear aero-spike tail rocket, and the reaction control system(RCS). The oxygen tanks are a multi-lobe, alumi-num-lithium structure designed to contain liquidoxygen. The placement of the oxygen tanks isdriven by trajectory requirements for CG locationand control.

Pressurization and purge functions are accom-plished with gaseous helium. The spherical heliumtanks are located inside of the liquid hydrogentanks to keep the stored helium at the highest pos-sible density. The hydrogen tanks are pressurizedby helium to only 5 psig at the beginning of the mis-sion, while the LOX tanks are run at about 35 psig.Additional helium is admitted to the tanks as liquidis removed, to maintain pressurization. However,as the tanks are heated the vapor pressure of thehydrogen or LOX rises and less and less helium isrequired for pressurization. Careful control of theheat load on the tanks is required to limit boiloff,but increased thickness of TPS must be carefullyevaluated as it reduces overall volume availablewithin the vehicle.

Actuation and Hydraulics

The hydraulics system controls both discreteand continuously varying operations on the air-frame. The discrete operation includes the actua-tion of the landing gear, while continuously varyingoperations include actuation of the horizontal andvertical control surfaces. Hydraulics also controlwheel brakes and nosewheel steering. The actua-tors for the control surfaces are shown in Figure28.

Landing Gear

The design drivers for the landing gear are thevolume required in the vehicle, the takeoff speed,the vehicle weight, and the tipback angles. Thegeneral arrangement of the landing gear isdepicted in Figure 29.

Tires and landing gear carriages occupy alarge volume that must be placed inside the fuse-lage near the gear attachment points. The takeoffspeed of SSTO vehicles is extremely high,approaching three hundred knots. This requires astructure and a system that can withstand thesehigh speeds and drag forces, while loaded at maxi-mum TOGW. Landing at vehicle empty weight (withpayload) is not considered a design driver duringconceptual design, because other conditions, suchas taxi bumps at maximum TOGW, have beenshown to be a greater structural challenge. The tip-back angle typically chosen for these configura-tions is fifteen degrees, allowing a 14 degree AOAwith some margin at takeoff.

Vehic le Synthesis

It is highly unlikely that the vehicle design teamwill pick a vehicle size that provides the precisePFA to perform a required mission. Therefore, ameans is needed to resize the vehicle to the mis-sion required size. This process must be some-thing other than redesigning and reanalyzing thevehicle over and over (trial and error). The idealprocess would take into account complex multi-dis-cipline interactions and non-linear effects of chang-

Figure 28. Control Surface Actuators

Figure 29. Landing Gear Arrangement


ing propulsion and internal system sizes. It wouldalso accommodate changes in mission require-ments and trajectory refinement, as well as modelall sorts of physical relations and accept a varietyof multi-discipline information.

This ideal process does not exist, but theauthors have evolved a process that synthesizes avehicle design from an assembly of its constituentparts and systems. Vehicle synthesis connotes amodeling process that collects individual mathmodels of the vehicle components, each modeledas a function of size or other physical quantitiesrelated to size, such as thrust or power. All compo-nents are reduced to an algorithm or series orequations to describe its weight and volume occu-pied. Each of the individual models is linked to thegeometry model of the vehicle so that changes insurface areas and volumes are known to the col-lective process. The entire model is implementedon a multi-page workbook (assembly of spread-sheets) for ease of communication between sec-tions. The set of equations is highly interactive andmust be solved by iteration, so the spreadsheet for-mat is ideal to track nearly a hundred individualitems.

The assembly of this model involves expertinput from the multi-discipline team. The spread-sheet format allows ready access to all models andmakes updates simple and easy, as each disciplinematures its design input. Another interesting fea-ture of this format is the ease of performing tradestudies. Quick changes of models or technologyfeatures can easily be implemented for comparisonof impact on the closure weight. The interaction ofthe team members during its construction is one ofthe keys to an accurate vehicle assessment.Changes in sizing are driven by photographic scal-ing from a reference size. From the trajectory anal-ysis, a PFR and average propellant density arecomputed. The average propellant density isdirectly related to LOX fraction and was easier toimplement in the synthesis. These values are inputto the tool and it iterates until a closure is obtained.For the work in this paper, the synthesis model wasindexed to a fuselage volume of 100,000 ft3, whichnecessitated small adjustments in input parame-ters from those of the as-drawn vehicles, with fuse-lage volumes around 120,000 ft3. These minoradjustments were likewise made by photographicscaling. Thus, all sub-models or algorithms in thesynthesis tool were accurately indexed to a com-mon fuselage volume of 100,000 ft3. The synthesis

tool also provides useful information and guidanceto the design team in performance of the analysesand design task. The tool keeps track of propellantfraction, LOX fraction, mass properties, CG loca-tion, and load distribution as a function of fuselagestation. This information is used in trajectory simu-lation, structural analysis, and weight prediction.

The synthesis process works very well for pho-tographic scaling because small changes in sizedon’t affect aerodynamic coefficients and propul-sion efficiencies. Thus, gross thrust, lift, drag, andpitching moment will scale accurately to the sec-ond power of photographic scale factor, for smallchanges. The size of the as-drawn vehicle must becarefully estimated because all aerodynamic andpropulsion analysis will be performed on geometricmodels derived from the original as-drawn configu-ration. A critical issue for closure accuracybecomes the magnitude of the closure scale factor.Every model in the synthesis process was basedon point-design information at the as-drawn size. Ifthe models had been based on a very small fuse-lage volume the uncertainty (error) in scaling to amuch larger volume would be greater. The goal forthis effort was to keep closure scale factors within10% of the reference size. Thus, the choice of anas-drawn size for the point design work is impor-tant, both to performance analyses and designwork. If the propulsion integration performs sopoorly that the vehicle PFR drives the closure scalefactor out of range, then there are several potentialareas of concern. The least of these is the errorcaused by scaling. Although the error will increaseas scale factor moves out of the desired range, thetrends will still hold. The more relevant concern isto understand the fundamental problem with theconfiguration. Immediate areas to focus on wouldbe cycle performance, installation penalties, aero-dynamics, and vehicle drag including trim effects.

Closure

The closure process brings all relevant designand analysis data together in the synthesis model.For a particular LOX fraction, the PFA may bedetermined from the synthesis model indepen-dently of mission performance information. Thismay be represented as a continuous curve of PFAvs. TOGW as shown in Figure 30 for the statusABLV-9 vehicle. Every different LOX fraction willproduce a different curve, so in reality there will bea family of curves. From the trajectory simulation aPFR and LOX fraction are determined to fly the


mission. The closure process may be visualized asselecting the correct LOX fraction curve and pick-ing the TOGW that corresponds to the requiredPFR. Note that the LOX fraction and PFR arelinked through the trajectory simulation, so vehiclecomparisons on the basis of PFR alone can bequite misleading. Recall that the figure-of-meritselected for this study was minimum TOGW. Vehi-cle DW may be treated in the same manner andwill exhibit similar trends.

Before moving to the closure details, a discus-sion of design maturity and analysis fidelity is inorder. All of the vehicle designs in this paper arerelatively mature and the analysis methods are at aconsistent level. All trajectory simulations were3DOF with aerodynamic trim in the pitch plane.Aerodynamic databases were based on consistentengineering methods, CFD, and wind tunnel test-ing. Propulsion performance was based on 2Dmethods for inlets and nozzles, with 1D combustorcycle codes, both for low-speed and high-speedengines. The low-speed cycle code was P&W pro-prietary and the high-speed combustor model isextensively validated against actual wind tunneldata for numerous engine models. Recall that no3D effects were taken into account in the propul-sion performance for any vehicle. The impact of 3Deffects on the propulsion is more relevant in thescramjet mode and must eventually be assessed inthe configuration development. However, the sub-stantial CFD effort required to assess inlet andnozzle effects at multiple Mach numbers precludeits use in multiple configuration studies. Internalpackaging and subsystems layout were performedwith 3D CAD tools. Structures and TPS werebased on prior analysis and were consistent for allconfigurations. Installation weights for the turbine

engines were estimated based on previous designstudies, but were increased by 20% for the ABLV-9B vehicles because of the added variable geome-try. All CTSC configurations used triple point hydro-gen fuel. Based on the relative scale illustrated inTable1, the results presented herein are at a

design maturity level of 5 or 5+. An important con-sideration in vehicle development is weight growthas the design matures. To provide added margin inthe TOGW predictions made by the closure pro-cess, the synthesis tool includes a 15% DW growthallowance imposed above the predicted values ofvehicle DW. This penalty is imposed on all individ-ual components of the vehicle in a continuousmanner to assure that all closure TOGWs allow forcontinued dry weight growth at lower design matu-rities. The growth factor should be reduced as thevehicle reaches final design.

Figure 31 shows the status TOGW across the

configurations analyzed in the current study. Onthe left is the reference vehicle ABLV-4 at 1.07x106

Figure 30. PFA vs. TOGW Closure Curve

Table 1: Design Maturity Level

Figure 31. TOGW of Configurations


lb., followed by the first Ace-TR derivative, ABLV-4B at 1.043x106 lb.(old MLG installation) and theABLV-4B* at 1.063x106 lb. (*MLG in fuselage). Thenext vehicle is ABLV-9B** at 1.096x106lb. (Ace-TRT/W=24 and updated installation weights), followedby the final configuration, ABLV-9B*** at 1.134x106

lb. (Ace-TR T/W=20 and updated installationweights). An important criteria for evaluation ofthese results is the current runway weight limit foran HTO configuration. This is estimated to beabout 1.25x106 lb. for the current runway construc-tion. With the 15% dry weight growth factorincluded in these results, an acceptable marginexists for this limit.

Figure 32 depicts DW for the same configura-

tions. Note that the DW of the ABLV 4B* configura-tion is higher than the ABLV-9B**, but recall that thebody width of ABLV-4 allowed integration of eight(8) Ace-TR engines, while the more slender ABLV-9B** housed only six (6) Ace-TR engines. Withonly six engines, the ABLV-9B** configuration hadless thrust and required more rocket augmentationin the low-speed range, which slightly increasedLOX use and TOGW, but still had lower DW due toonly six Ace-TR engines. The ABLV-4B configura-tions, with eight Ace-TR engines, had more thrustand required less rocket augmentation in low-speed, delivering slightly lower LOX use andTOGW. However, the eight Ace-TR engines droveup the DW. In this case, the slightly higher TOGWmay be desirable, in order to reduce cost of thelow-speed engine installation (six engines vs.eight).

Summar y and Conc lusions

The development of a promising HTO SSTOconfiguration was traced through the multi-disci-pline design and analysis process. The missionand trajectory development were described alongwith the aerodynamic and propulsion predictionssupporting the 3DOF simulation. The propulsion-airframe integration of both low-speed and high-speed propulsion systems was described, fromboth aero-propulsive and mechanical perspectives.Design details of the preferred configuration weredescribed, along with the supporting analyses per-formed to substantiate the status results. Vehiclepackaging was described and mechanical designswere disclosed, with structures, materials, andtechnologies employed. Finally, the closed vehicleweights were provided.

Particular attention was focused on describingthe complex multi-discipline process as the pre-ferred configuration was developed. The maturityof the design and analysis was discussed andassessed on a relative scale, which was providedto illustrate the differing levels of credibilityachieved with various design/analysis approaches.With reference to Figure 40, the final ABLV-9B***configuration is assessed a Design Maturity Levelof 5+. The CTSC concept appears to be a robustpropulsion system with good potential for real vehi-cle development. There was some weight growth,relative to the ABLV-4 reference vehicle, as the newconfiguration was moved through this study. How-ever, there was a substantial payoff in design cred-ibility with that growth. Slush hydrogen wasremoved from the vehicle to reduce infrastructurecosts. The main landing gear were moved into amore conventional fuselage integration, substan-tially reducing risk from the reference installation.An updated weight estimate was imposed on theAce-TR installation and the uninstalled T/W wasreduced from 24 to 20.

However, there are many additional areas thatshould be worked to further mature this design.The updated structural analysis capability needs tobe completed and factored into the closure. Thesynthesis tool update needs to be enhanced totake advantage of the new structural capability. Ahigher fidelity design study of the low-speed inletand nozzle should be performed to better predictinstalled weight and volume for optimization of theramjet takeover Mach number. Fuel cooling of low-speed inlet flow should be explored to see if the

Figure 32. Dry Weights for Configurations


engines could be downsized to save weight andvolume. The impact of fineness on low-speed sys-tem installation should be explored to exploit anypossible cost advantages. Additional turbine-basedengines should be assessed against the Ace-TR tosee if one may have an advantage. When a moreoptimum configuration is ready, the 3D propulsioneffects should be assessed. Low-speed inlet andnozzle performance should be reviewed to assesswhether they might be improved.

In conclusion, the current study has identifiedand assessed a promising configuration for theHTO SSTO mission. The lifting body with CTSCairframe-integrated propulsion is supported bymany years of research, design, analysis, andexperimental results for the key technologies. Thecurrent configuration offers potential for furtherdevelopment and optimization. The propulsion andairframe technology incorporated in this vehicleshould be supported with continued research andhardware demonstration, to provide the enablingtechnology for the next century’s revolutionaryflight vehicles.

Ackno wledgments

The authors wish to express their sincereappreciation to A. M. (Jose) Espinosa of The Boe-ing Company in St. Louis, MO, for his substantialcontribution in supplying the installed low-speedsystem performance for these studies. The authorsalso wish to thank Hilmi Kamhawi for his AML pro-gramming support, which enabled much improvedproductivity for 2D propulsion analyses. Finally, theauthors must acknowledge Charles R. McClintonof NASA Langley Research Center for his contin-ued support of the tool and process developmentnecessary to perform this complex work.

References

1). McClinton, C.R., Hunt, J.L., Ricketts, R.H.,NASA Langley Research Center; Reukauf, P.,NASA Dryden Flight Research Center; Peddie,C.L., NASA John Glenn Research Center;“Airbreathing Hypersonic Technology VisionVehicle and Development Dreams”, AIAA 99-4978, November 1999.

2). Hunt, J.L., “Airbreathing / Rocket Single-Stage-to-Orbit Design Matrix”, AIAA 95-6011, April1995.

3). Adaptive Modeling Language, Version 3.1.2,TechnoSoft, Inc. Cincinnati, OH. Copyright 1999.

4). Pro/ENGINEER, Release 9839, Version 20,Parametric Technology Corporation, Waltham,MA, Copyright 1988-1998.

5). Powell, R. W. et al, NASA LaRC, Brauer, G. L. etal, Lockheed Martin Corp., “Program to OptimizeSimulated Trajectories (POST) Version 1.03”,February 1998

6). Sova, G., Divan, P., “Aerodynamic PreliminarySystem II”, North American Aircraft Operations,Rockwell International, Los Angeles, CA.

7). Gentry, A. E., Smyth, D. N., Oliver, W. R., “TheMark IV Supersonic-Hypersonic Arbitrary-BodyProgram”, McDonnell-Douglas Corporation,Technical Report AFFDL-TR-73-159, November1973.

8). Collier, C.S., “HyperSizer Analytical Methods &Verification Examples”, Collier ResearchCorporation, 1998.

9). Cullimore, B. A., Ring, S. G., Johnson, D. A.,“SINDA/FLUINT General Purpose Fluid NetworkAnalyzer, Version 4.1”, Cullimore and RingTechnologies

Documents

AIAA99-4948 An Airbreathing Launch Vehicle Design with Turbine … · 2015. 9. 29. · Viable and practical airbreathing, hypersonic propulsion systems for these vehicles are a pri-mary