AIAA 99-2104SCCREAM v.5 : A Web-Based AirbreathingPropulsion Analysis Tool(http://titan.cad.gatech.edu/~jebradfo)

J. BradfordJ. OldsGeorgia Institute of TechnologyAtlanta, GA

35th AIAA/ASME/SAE/ASEE Joint PropulsionConference and Exhibit

20-24 June 1999Los Angeles, California

For permission to copy or to republish, contact the American Institute of Aeronautics and Astronautics,1801 Alexander Bell Drive, Suite 500, Reston, VA, 20191-4344.

AIAA 99-2104

SCCREAM v.5 : A Web-Based Airbreathing Propulsion Analysis Tool

John E. Bradford†

Dr. John R. Olds*

Space Systems Design LaboratorySchool of Aerospace Engineering

Georgia Institute of Technology, Atlanta, GA 30332-0150

ABSTRACT

To properly assess the advantages anddisadvantages of various RBCC design options at theconceptual vehicle level, an engine performanceanalysis tool is required. This tool must be capableof modeling engine performance effects that willsubsequently be propagated throughout theconceptual design process via trajectory analysis,weight assessment, fuel balance calculations, thermalenvironment, life cycle cost, etc. For a given engineconfiguration, the tool will need to generate enginethrust and Isp as a function of altitude and Machnumber for each operating mode of an RBCC engine.

A project to create a new computer program forthe analysis of RBCC engines has already beeninitiated. Called SCCREAM, for SimulatedCombined-Cycle Rocket Engine Analysis Module, itis intended for use in the conceptual phase ofairbreathing launch vehicle design.

This paper will detail the capabilities of the latestversion of SCCREAM and present the results ofvalidation efforts. Combustor thermodynamicproperties and overall engine performance for asample engine will be compared with industrystandard codes. Results from the new scram-rocketmode will be discussed. Ejector mode performanceplots generated over the web will also be presented.

NOMENCLATURE

Ac normalizing area for thrust coefficient (ft2)Ai engine cross-sectional area at station i (ft2)Ct thrust coefficient (thrust/q*Ac)CO carbon monoxideCO2 carbon dioxideCPG calorically perfect gasD hydraulic diameterESJ ejector scramjet engineESR ejector scram-rocket engineEtake inlet kinetic energy efficiencyf friction coefficientH monatomic hydrogenH2 hydrogenIsp specific impulse (sec)Kp equilibrium constantLOX liquid oxygenMR propellant mixture ratioO monatomic oxygenO2 oxygenOH hydroxyl radicalPc rocket primary chamber pressure (psi)Pi partial pressure of species iPOST program to optimize simulated trajectoriesPt total pressure (psi)phi combustor equivalence ratioq freestream dynamic pressure (lb/ft2)RBCC rocket based combined-cycleSERJ supercharged ejector ramjet engineSESR supercharged ejector scram-rocket engineV velocityVf fuel injection velocityy axial fuel velocity componentγ ratio of specific heatsθι fuel injection angleτ shear stresssρ density∆GT standard gibbs function change

† - Graduate Research Assistant, NASA MSFC Graduate

Student Researchers Program, School of AerospaceEngineering, Student member AIAA.

* - Assistant Professor, School of Aerospace Engineering,Senior member AIAA.

Copyright © 1999 by John E. Bradford and John R. Olds.Published by the American Institute of Aeronautics andAstronautics, Inc. with permission.

AIAA 99-2104

2

RBCC BACKGROUND

Rocket-based combined-cycle engines areunique in that they combine the most desirablecharacteristics of airbreathing engines and rocketengines into a single, integrated engine. RBCCengines have the advantage of high average specificimpulse (Isp) in comparison to rockets, and highthrust-to-weight ratios in comparison to airbreathers.

Figure 1-Supercharged Ejector Ramjet Engine (ref.1)

The concept of combined-cycle engines hasexisted since the mid-60’s. During this inceptionphase, an extensive study was conducted by theMarquardt Corporation, Lockheed-California, and theU.S. Air Force on various ‘composite engine’designs, as they were formerly called [1]. This studyinitially analyzed 36 different variants of combined-cycle engines. At the study’s conclusion, two types ofRBCC engines were selected as the most interestingoptions — a near-term option and a far-term option.The decisions were made based on technologicalfeasibility and resulting performance on arepresentative two-stage-to-orbit launch vehicle. Thetwo final selections were the Supercharged EjectorRamjet (SERJ) configuration (figure 1), and the moretechnically challenging Supersonic CombustionRamjet with Liquid Air Cycle (ScramLACE)configuration. The SERJ engine configuration iscomposed of four operating modes: ejector, fan-ramjet, ramjet, and pure rocket. A derivative of theSERJ is the Supercharged Ejector Scramjet (SESJ).This configuration consists of five operating modes,the four from the SERJ and an additional scramjetmode.

Figure 2 - Rocket Primary (ref.1)

During ascent phase, the RBCC engine initiallyoperates in ejector mode. The ejector mode utilizesthe rocket primaries (figure 2) as the main source ofthrust. Entrained air from the inlet and fuel from thesecondary fuel injectors is also burned in thecombustor to provide additional thrust. A low-pressure ratio fan, located between the inlet andprimary, may also be used. Once significant rampressure is achieved from the surrounding air,typically occurring around Mach 2 to 3, the rocketprimaries are shut off. The fan remains functioningup to about Mach 3, constituting the fan-ramjetmode. At Mach 3, the fan is removed from the flowpath or perhaps windmilled in place to as high asMach 6. The engine operates in pure ramjet mode upto around Mach 6. At Mach 6, depending upon theengine type (SESJ or SERJ), the engine willtransition either to scramjet mode or directly torocket mode. If scramjet mode is available, theengine will continue operating as an airbreather withsupersonic combustion up to an optimal transitionMach number. Recent conceptual vehicle designshave suggested transition to pure rocket mode mightoptimally occur between Mach 10 and Mach 15.While transitioning to rocket mode, the inlet face isclosed and the rocket primaries are restarted. VacuumIsp’s in the range of 410-470 seconds are typicalvalues during rocket mode.

AIAA 99-2104

3

SCCREAM BACKGROUND

SCCREAM has the capability to model theperformance of six types of RBCC engines. One ofthese configurations is the one identified in theMarquardt study — the supercharged ejector ramjet(SERJ). The other five are the (non-supercharged)ejector ramjet (ERJ), the ejector scramjet (ESJ),supercharged ejector scramjet (SESJ), ejector scram-rocket (ESR), and supercharged ejector scram-rocket(SESR). Additionally, SCCREAM can model pureramjet and pure scramjet configurations.

SCCREAM operates by solving for the fluidflow properties (velocity, temperature, pressure, massflow rate, gamma, specific heat capacity, etc.)through the various engine stations for each of theengine operating modes. Equations for conservationof mass, momentum, and energy are used. Thisprocess is often iterative at a given engine station orbetween a downstream and an upstream station. Theflow properties are calculated using quasi-1D flowequations. Engine cross-sectional area is the onlygeometry variable along the stream direction.Component efficiencies are used to simulate losses oftotal pressure in the mixer and nozzle, and reducedenthalpy in both the rocket primary and maincombustor. A started inlet is simulated by a simpletotal pressure recovery schedule. If the inlet is notstarted, SCCREAM places a normal shock in front ofthe cowl. Mass capture for the started inlet isdetermined by the flow conditions at the cowl leadingedge. Mass capture for the unstarted inlet is based onthe maximum allowable mass flow at the inlet throat.Thrust and Isp are determined using a control volumeanalysis of the entering and exiting fluid momentumand the static pressures at the inlet and exit planes.

Most internal areas of the engine are determinedin SCCREAM based on ratios of the inlet/cowl cross-sectional area. Default area ratios are supplied, sotypically a user enters only the inlet area. The size ofthe rocket primary unit is primarily based on a user-entered propellant mass flow rate for the rocketprimary. These two independent variables can bevaried to produce an engine with a desired sea-levelstatic thrust and secondary-to-primary mass flowratio. In practice, however, the inlet area is often

limited by overall vehicle geometry or shock-on-lipconditions. Optionally, the user can enter a desiredsea-level static thrust and inlet area, and SCCREAMwill iterate to determine the primary mass flow raterequired.

In order to generate a POST engine table, acandidate engine’s performance is evaluated over arange of altitudes and Mach numbers [2]. TheseMach number and altitude ranges can be set by theuser. For example, a ramjet’s operational Machnumbers might be set from 2 to 5.5, with altituderanges from 30,000 feet to 150,000 feet. OverlappingMach numbers and altitudes between variousoperating modes allows POST to select optimumengine mode transition points if desired. DefaultMach number and velocity ranges are provided foreach mode.

Performance in pure rocket mode is determinedby analyzing a high expansion ratio rocket engineoperating in a vacuum. A user-entered nozzleefficiency is used to account for losses associatedwith the expansion of the primary exhaust throughthe engine and then onto the aftbody.

X f t

Figure 3 - Axisymetric Engine Station Locations

X f t

Figure 4 - 2-D Engine Station Locations

Figure 3 shows the station numbers andreference areas used by SCCREAM for anaxisymetric RBCC engine configuration. Figure 4shows station locations for a 2-D engineconfiguration. The 2-D engine layout is morecommon for vehicles with scramjet capability.

AIAA 99-2104

4

Station 1 is at the inlet plane of the engine.Freestream flow conditions at station ‘infinity’ aremodified by a single shock wave to simulate anyprecompression effects of the vehicle forebody on theengine. The forebody shape (wedge or cone) and theforebody angle are entered by the user. Therefore theflow conditions at station 1 are typically not the sameas the freestream flight conditions.

The started inlet performance is modeled by acurve fit of the total pressure (Pt) recovery forsubsonic combustion and a kinetic energy efficiency(Etake) for supersonic combustion. Both are functionsof the Mach number at the inlet face. Variablegeometry at the inlet throat is assumed.

Station 2 is at the location of the rocket primary.For ejector mode, station 2 to 3 is a constant areamixing process between the entrained air stream andprimary exhaust.

From station 3 to 4, the fuel is injected at aspecified position, velocity, angle, and equivalenceratio. A heat release profile provided by the usercontrols the reaction rate of the fuel. Upon exitingthe combustor, the flow is passed through aconverging-diverging nozzle to the exit plane of theengine (station e or e’).

For a more complete description of the flowprocess, the reader is referred to references 3 and 4.

SCCREAM v.5

The following is a list of the improvementsmade to SCCREAM that will be discussed in thefollowing sections.

1. Improved Combustor Modeling2. Hydrocarbon Rocket Primary Propellants3. Hydrocarbon Afterburner Fuels4. Scram-Rocket Mode Analysis5. On-line Plotting Capability6. Primary and Afterburner Phi Throttle7. Expert and Novice Versions

Each of these improvements and the method used toimplement them will be discussed in detail.

IMPROVED COMBUSTOR MODELING

A 1-D combustor model that accounts for theeffects of mass addition, heat addition, friction, andvariable area has been implemented in SCCREAMv.5. SCCREAM was previously limited to a constantarea, frictionless combustion processes. This newmodel has resulted in only a minor increase inSCCREAM's total runtime.

The combustor model uses a common analysismethod known as 'Influence Coefficients'. Thistechnique allows for all the flow properties to bedetermined by simply solving for the change in Machnumber throughout the combustor. The solution forthe Mach number variation involves solving anordinary differential equation of the form:

dM

MC

dA

A

M fdx

DC

dT

TC y M

dm

m

= − + + + −[ ]

1

2

2

42 3

2γ γ.

.(1)

where,

CM

M1

11

22

1 2=+ −

−

( )γ

(1a)

CM

2

21

2=

+( )γ(1b)

C M321= +( )γ (1c)

The derivation of Eqn. (1) will not be presented (theauthors will refer the interested reader to reference 5),but the significance of each term in the equation willbe discussed.

The first term in Eqn. (1) is the impact on theMach number due to area changes in the combustor.With the user specified area ratios and combustorlengths, this value can easily be computed. After thegeometry is defined, the combustor is discretizedinto 1,000 axial steps.

The second term accounts for the effect of thewall shear stresses on the flow. The user provides anaverage skin friction coefficient, defined as:

AIAA 99-2104

5

fV

= τρ 2

2

(2)

where τ is a shear stress. Common values for f rangefrom 0.001 to 0.002. The variable 'D' is the flows'hydraulic diameter' and is the mean diameter theflow experiences at each combustor step. Note thatthis method assumes a circular combustor crosssection.

The third term is for the heat release due tochemical reactions. The 'influence coefficient'method requires the heat release to be specified interms of a total temperature. The procedure used fordetermining this will be discussed later.

The last term in Eqn. (1) is the contribution fromthe injected fuel's mass and velocity. This term isonly nonzero at the location of the fuel injection. The'y' term has the simple form shown in Eqn. (3).

yV

Vi

stream

=( )f cos θ

(3)

where Vf is the velocity of the fuel, θi is the injectionangle of the fuel, and Vstream is the velocity of the flowjust upstream of the fuel injection position. If thefuel is injected normal to the flow, the injection angleis 90o and the value for 'y' will be zero. If theinjection angle is parallel to the flow (θi=0), 'y' willhave its maximum value and contribution to the flowmomentum. If the fuel injection occurs at any otherangle, the 'y' term accounts for the axial contributionof the fuel injection.

The conditions at the entrance to the combustorare known based on the outflow conditions from themixer section (station 2 to station 3) of the engine.To begin the analysis, the required flow propertiesare: the initial total enthalpy, static pressure, statictemperature, Mach number, mass flow rates, and therate of change in mass flow rate composition throughthe combustor. Linear profiles for the species ratesof change are assumed.

The user is required to specify the frictioncoefficient, injection location of the fuel, start of theheat release, end of the heat release, fuel injection

angle (θi), and fuel injection velocity (Vf).SCCREAM allows the user to specify different fuelinjection and heat release parameters for subsonicand supersonic combustion cases.

Returning now to the calculation of the heatrelease profile. With the initial enthalpy andcomposition known, the total temperature can beevaluated. This is an iterative procedure whichinvolves guessing the total temperature, evaluatingeach species enthalpy, mass averaging theirenthalpies, and comparing the new total enthalpywith the specified initial enthalpy. Until the point offuel injection and chemical reations start to occur, thetotal temperature will remain unchanged and at thisvalue in the combustor. Thus, the 3rd term in Eqn. (1)will be zero up to the point of fuel injection.

At the location of the fuel injection, a new totaltemperature must be evaluated. The mass flow rateof the fuel is added to the flow composition and anyenthalpy contribution from the fuel is added to theflow. The same iterative procedure performed for theinitial temperature is repeated for the new massinjection total temperature.

Now, the final total temperature must bedetermined. This is easily accomplished since thefinal mass flow rates at the end of the heat releasehave been specified beforehand. The same procedureis then executed, using the final mass flow rates andtotal enthalpy, to determine the final totaltemperature value.

With a specified fuel injection location, startingpoint and ending points of heat release, a lineardistribution based on the change in total temperaturecan be established. Note that the heat release in a realengine is most likely not a linear profile. The authorshave assumed a linear profile due to its simplicity andSCCREAM can easily be modified to accommodateother profiles.

The main limitation of the influence coefficientmethod is due to an assumption made in itsderivation. The influence coefficients assume acalorically perfect gas (CPG). Without thisassumption, the relatively simple form presentedabove could not be arrived at . Instead of Eqn. 1, a

AIAA 99-2104

6

set of partial differential equations, the Eulerequations, for a 1-D flow would be arrived at. TheCPG assumption typically does not hold above 2,000R. But, by updating the specific heat and specific heatratio at each spatial step, the usefulness of thetechnique has been expanded. This will bedemonstrated and become evident in the subsequentverification process.

A fourth order Runge-Kutta method is used tosolve Equation (1) for the variation in Mach number.The general procedure is outlined next.

All conditions at the upstream step are knownbeforehand. The differentials of total temperature,area, and mass flow rate are determined for thecurrent position in the combustor. Using the Runge-Kutta solver, the local change in the Mach number isevaluated. The Mach number at the downstreampoint is then determined from Equation (4).

M MdM

dxdx2 1= + (4)

A check on the value of this Mach number isperformed next. A comparison against a criticalMach number is done. The critical Mach number iseither 0.95 or 1.05 depending if subsonic orsupersonic combustion is occuring. If thecombustion is subsonic and the new Mach numberexceeds the critical Mach number, or in thesupersonic case, the Mach number is less than thecritical Mach number, then the combustor analysisstops and a reduction in the equivalence ratio (phi) isrequired to prevent choking. SCCREAMautomatically reduces phi and restarts the combustoranalysis. If a phi reduction is not required, then theaxial position, area, and total temperature are updatedto the downstream position and become the upstreamconditions for the next starting point.

Next, the flow composition and molecularweights are updated at the new position. A new statictemperature based on the new Mach number and totaltemperature is then computed. With this statictemperature, a new specific heat and gamma can bedetermined. The static pressure is then determinedbased on conservation of mass, using the new massflow rate.

This entire procedure is repeated until the end ofthe combustor is reached, or the flow violates thecritical Mach number condition. After the flowconditions at the exit of the combustor have beenobtained, they are passed into the nozzle routine asentrance conditions.

Validation:

To verify the new combustor model, comparionscases were run for a scramjet combustor with theindustry accepted code, SRGULL. SRGULL uses a1-D Euler routine for its combustor analysis. Thismethod is similar to the influence coefficient method,but without the assumption of a CPG at each localstep. Similar to SCCREAM inputs, the SRGULLuser can establish the engine geometry, fuel injectionposition, heat release profile, and combustorefficiency. One of the main differences between theSRGULL and SCCREAM model is that SRGULLuses an equilibrium chemistry routine, as opposed toSCCREAM's complete-combustion model. Properlymodeling the chemistry becomes more important athigher flight Mach numbers. But, the SCCREAMmodel has proven adequate up to Mach 12, withengine performance numbers becoming progressivelymore conservative at higher Mach numbers [6].

For the test cases, a hydrogen fueled scramjetengine flying at a freestream Mach number of 6.5 and8 were examined. The vehicle flying this engine hada 9o 2-D wedge forebody and flew along a constantdynamic pressure boundary of 2,000 psf. Thecombustor geometry modeled in SCCREAM andSRGULL were identical and corresponded to a ESJengine design. Figure 5 provides the area ratio'sversus axial position for the test case.

After SRGULL completed its analysis, theconditions at the entrance to its combustor were usedas the entrance conditions to the combustor inSCCREAM. These conditions included mass flowrate, static pressure, static temperature, Machnumber, and gamma. A constant friction coefficientof 0.0018 was specified. The fuel injection occuredat an X/C value of 0.55, which also corresponded tothe start of the heat release. A linear profile wasestablished in SRGULL, which is the standard profileused in SCCREAM, and the end of the heat release

AIAA 99-2104

7

was at an X/C value of 0.95. Parallel fuel injection ata velocity of 6,000 ft/s was specified for both models.At both the Mach 6.5 and 8 conditions, anequivalence ratio of 1.0 was allowed without causinga thermal choke in the combustor.

X/C

A/A

ref

2.5

2.3

2.0

1.8

1.5

1.3

1.00.0 0.2 0.4 0.6 0.8 1.0

Figure 5 - Combustor Geometry

Figure 6 shows the Mach number distributiongenerated by SCCREAM and SRGULL for the Mach8 flight conditions. The Mach 6.5 results were verysimilar and have not been included for brevity.

SRGULLSCCREAM

Ma

ch N

um

be

r

X/C

0.0 0.2 0.4 0.6 0.8 1.01.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

reaction zone

Figure 6 - Mach Number Distribution

In the non-reacting region of the combustor (upto X/C=0.55), SCCREAM and SRGULL have nearlyidentical profiles. At the location of the fuelinjection, the sudden drop in Mach number is due tothe addition of the fuel. SCCREAM appears toslightly underpredict the strength of this drop, but theeffect is clearly captured by SCCREAM. Theremaining portion of the combustor is the chemicallyreacting region. SCCREAM and SRGULL bothdisplay very similar trends and profile shapes overthe entire heat release process. Differences between

the two curves appear to be caused by the initialdifferences from the fuel injection. If the magnitudeof these effects agreed, it is believed that the heatrelease profiles would agree almost exactly.

SRGULLSCCREAM

2.5

2.0

1.5

1.0

0.5

0.00.0 0.2 0.4 0.6 0.8 1.0

X/C

Ps/

Pre

f

reaction zone

Figure 7 - Static Pressure Distribution

Figure 7 shows the static pressure distribution inthe engine. Once again, excellent agreement isobtained between the two codes. The Mach 6.5 staticpressure distibution displayed similar trends and willnot be presented again for brevity.

HYDROCARBON ROCKET PRIMARY

Two new propellant combinations have beenmade available to the SCCREAM user for analysis.Either a methane (CH4) and oxygen or a JP-5 (C10H19)and oxygen engine can be selected. These propellantcombinations are in addition to the previoushydrogen (H2) and oxygen or mono-propellanthydrogen-peroxide (H2O2) modeling capabilities.

To enable the analysis of these new primarysubsystems in SCCREAM, a procedure similar tothat used for the hydrogen and oxygen propellantswas used. This procedure involves running sweepsof chamber pressure (Pc) and mixture ratio (MR) inthe Chemical Equilibrium with Applications code,CEA, for each of the propellant combinations. Forthe methane fuel, MR’s from 2 through 6, Pc’s of 500to 4,500 psi were examined. For the JP-5 fuel, MR'sof 1.9 through 4.4, at Pc's of 500 to 4,500 psi wererun. After each analysis by CEA, the results for eachspecies mole fractions, chamber specific heat ratio,and chamber temperature were recorded. Using astatistical analysis software, called JMP, Response

AIAA 99-2104

8

Surface Equations (RSE's) were generated for eachspecies concentration, chamber specific heat ratio,and chamber temperature. For improved accuracy,separate RSE's were generated for fuel rich and fuellean mixture ratios [2,7,8].

With this model of the rocket primarycombustion process, the RSE's are used to determinethe mass flow rates of each species exiting the rocketprimary to be combined with the flow in the mixersection during ejector mode and scram-rocket modeoperation. The user is required to specify Pc, MR,and an expansion ratio. The exit conditions from theprimary thruster can then be determined fromisentropic 1-D flow equations. The exhaust from theprimary is a contribution to the main flow'smomentum in the mixer section. These same RSE'scan also be used for the all-rocket mode performanceanalysis.

HYDROCARBON AFTERBURNER FUELS

Three new afterburner fuels have been addedinto SCCREAM. These fuels are methane (CH4), JP-5 (C10H19), and JP-10 (C10H16). Previous versions ofSCCREAM only provided the user with a hydrogenfuel afterburner. Addition of these fuels requiredadding in each of the new fuel's properties,hydrocarbon product properties, and new chemistryroutines into SCCREAM.

Sensible enthalpy and specifc heats as a functionof temperature for methane, JP-5, CO2, and CO wereobtained from a variety of sources and polynomialcurve fits were generated. The current references forthe JP-5 only included data for temperatures up to4,000 R. If the temperatures exceed this range,values are linearly extrapolated to the requiredtemperature. Due to the small differences betweenthe fuels, the JP-10 fuel uses the same sensibleenthalpy and specific heat curve fits as the JP-5 fuel[9].

Two different chemistry routines were requiredto implement the new hydrocarbon analysis. Forfuel lean cases, a complete combustion with anefficiency factor can be used. For fuel rich cases, anequilibrium analysis is required.

For the fuel lean case, a simple atom balance forthe reaction can be applied. Equation (5) shows thechemical reaction for the fuel lean case:

C H O CO O H Ox y + → + +2 2 2 2 (5)

where x is the number of carbon atoms and y is thenumber of hydrogen atoms in the fuel chain. For thecomplete chemistry assumption, there will not be anyCO nor minor species (O, H, OH) formed in theproducts. The user-specified fuel efficiency is usedafterwards to compute additional concentrations ofthe O2, H2 and CO species. The efficiency is applieddirectly to the CO2 and H2O mass flow rates. Theinefficency values of the mass flows are thendistributed according to the reactions:

H O H O2 2 21

2→ + (6)

CO CO O2 21

2→ + (7)

For the fuel rich case, the number of product speciesincreases and the chemical reaction underconsideration becomes:

C H O CO CO H H Ox y + → + + +2 2 2 2 (8)

Due to the presence of two C-O molecules, thisreaction cannot be solved by simple atom balances.An additional equation is required and can beobtained from the water-gas shift equilibriumreaction:

CO H CO H O2 2 2+ → + (9)

This reaction indicates how the carbon andoxygen molecules will be distributed between the COand CO2 molecules. It can be written in a moreuseful form in terms of the equilibrium constant, Kp,shown by Equation (10).

KpG

R T

P

P

P

P

P

P

P

P

T

u

COo

Ho

COo

H Oo

= −

=

exp∆

2 2

2

(10)

AIAA 99-2104

9

Kp is significant because it is only a function oftemperature. Because this reaction is bimolecular,the pressure terms on the right hand side will cancelout, leaving only mole fractions. This eliminates thepressure dependance of the reaction and greatlysimplifies the problem. With this setup, we now onlyrequire knowledge of the static temperature at theexit of the combustor. After careful consideration, itwas decided that using the total temperature at theentrance to the combustor provided a good estimateof the static temperature at the exit of the combustor.Ideally, this would be an iterative procedureconsisting of making a guess, performing thecombustor analysis, then using the new statictemperature for the new Kp value. It can be shownthat at the elevated combustion temperatures, Kp isonly mildly affected by the temperature, thus theiterative process is unnecessary [10].

With the CO2 and CO species concentrationsdetermined, the concentration of the H2O moleculecan found and any remaining H-atoms are in the formof diatomic hydrogen. Similar to the fuel-lean case,the efficiency is applied directly to the CO2 and H2Oconcentrations.

It should be mentioned that for ejector mode andscram-rocket mode calculations, any hydrocarbonspecies generated by the rocket primary, with theexception of O2, are not considered in the afterburnerchemistry analysis. This includes the species CO2,CO, H2O, OH, O and H. For all cases, anydisassociation of nitrogen (N2) that might be occuringat elevated temperatures is not considered.

Validation:

In order to validate the new hydrocarbonanalysis capability, a sample engine configurationwas established and analyzed during ramjet andscramjet operation. SCCREAM performancenumbers (thrust and Isp) were generated andcompared with results from the Ramjet PropulsionAnalysis code, RJPA, for both the methane and JP-5fuels [11]. RJPA was selected instead of SRGULLbecause it is a more common analysis tool to use inthe conceptual design environment. SRGULL ismore commonly used in the preliminary design

stages of engine development and requiressignificantly more set-up time.

The comparison engine configuration wassimilar to the engine used for a Georgia Tech vehicledesign called 'Stargazer'. Figure 8 provides anexternal view of the Stargazer vehicle.

Figure 8 - Stargazer TSTO Bantam Concept

The Stargazer vehicle is a two stage conceptdesigned to deliver a small, Bantam class payload(300 lbs) to a 200 nmi. circular orbit. The first stageof Stargazer is a RBCC ESJ booster. The secondstage is an expendable, LOX-RP liquid rocket. Thebooster stage is unpiloted and has horizontal take offand landing capability. For the hydrogen version, thebooster accelerates up to Mach 3 in ejector mode,then transitions to ramjet mode until Mach 6. AtMach 6, the engines operate as scramjets until Mach10. After Mach 10, the inlets are closed-down andthe rocket primaries are re-ignited. The boostervehicle then accelerates to Mach 15 and separateswith the second stage at a dynamic pressure of 1 psf.The booster then performs a turnaround maneuverand cruises back to the launch site under ramjetpower at Mach 3.5.

Stargazer has a 7o 2-D wedge forebody. Table 1provides the engine geometry, fuel injectorparameters, and efficiency factors used for thecomparative analysis.

The RJPA model was set up with an identicalengine geometry, friction coefficient, injection angle,mass capture, and equivalence ratio. For ramjetoperation, the normal shock model was utilized in

AIAA 99-2104

10

RJPA. Because of the extreme sensitivity of the totalpressure recovery to the diffuser exit area, it wasdifficult to exactly match the inlet total pressurerecoveries with those used by SCCREAM. Inscramjet mode, RJPA allows specification of the inletefficiency (Etake) aand the same value used bySCCREAM was used in RJPA.

Table - 1 Stargazer ESJ Engine Data

inlet area, A1 20.0 ft2

mixer area, A3 10.0 ft2

combustor break, A3' 12.0 ft2

combustor exit, A4 16.8 ft2

maximum exit area, Ae' 65.0 ft2

combustor efficiency, ηc 95.0%

nozzle efficiency, ηnozz 98.5%

friction coefficient, f 0.001

fuel temperature, Tf 500.0 R

fuel injection velocity, Vf 2,000 ft/s

fuel injection angle, θi 0.0 deg

Figure 9a shows the thrust coefficient (Ct) versusMach number results for the methane fuel cases. Thephysical cowl area (A1) of the engine was used tonon-dimensionalize the thrust to obtain the Ct. AtMach 3, the maximum equivalence ratio allowed is0.3. At Mach 4, 5 and 6 an equivalence ratio of 1.0can be obtained. In ramjet mode, SCCREAMappears to underpredict the thrust levels produced byRJPA. Most of this discrepancy is being attributed tothe difference in the total pressure recovery. Forscramjet mode operation, SCCREAM and RJPAmatch very closely. All scramjet mode points are atan equivalence ratio of 1.0. The maximum differencein the thrust coefficient is only 5%, which occurs at aMach number of 8.

Figure 9a - Methane Thrust Coefficient

Figure 9b - Methane Specific Impulse

Figure 9b shows the Isp versus Mach numberresults for the methane case. The same trends shownin the thrust coefficient plots are also displayed in thespecific impulse charts.

Figure 10a - JP-5 Thrust Coefficient

Figure 10a shows the results for the same enginegeometry and flight conditions, but with JP-5 fuel.As with the methane fuel, SCCREAM and RJPAboth display similar trends. The maximum differencein scramjet mode thrust is 7% and occurs at Mach 7.This difference is being attributed to the heat releaseprofile. RJPA does not perform a marching solutionthrough the combustor, thus the effects of fuel

AIAA 99-2104

11

injector placement are not accounted for. Thesedifferences in the thrust values are within thetolerances of different heat release profiles andinjector locations.

Figure 10b - JP-5 Specific Impulse

Figure 10b shows the results for the specificimpulse for the JP-5 fuel case. Similar trends tothose in the thrust coefficient plots are also shown.

SCRAM-ROCKET MODE ANALYSIS

The additional operating mode of a rocket-augmented scramjet (ESR or SESR engine) has beenadded. At high Mach numbers, the scramjet thrustcan become low enough that the vehicle cannotaccelerate properly. The previous solution to thisproblem was to shut down the scramjet, quicklytransition the engine to all-rocket mode, and make anabrupt exit from the atmosphere. In addition towasting the oxygen still available in the atmosphere,this pullup maneuver can significantly increase thewing loading of the vehicle. The preferred method isto augment the scramjet mode with the rocketprimary and gradually transition to full-throttle all-rocket mode. This will allow the vehicle to continueaccelerating while still using the free oxygencontained in the air. This results in a higher specificimpulse than the all-rocket mode and significantlyincreased thrust compared to the scramjet mode.

Figure 11a provides the results fromSCCREAM's scram-rocket mode analysis casecompared with the scramjet mode performance andall-rocket mode performance. The scram-rocketresults presented are for the case of the rocketprimary being ramped up incrementally, starting at

Mach 9 and having full-throttle at Mach 12. Acommon technique used in trajectory simulation is tosimply ramp down the scramjet thrust and ramp upthe rocket thrust in order to model a gradual switchbetween the modes. The results of using thistechnique are also plotted. It can be seen that thesimple averaging technique does not capture all ofthe flow dynamics created by the rocket primaryexhausting inside of the scramjet engine. It is alsointeresting to note that the scram-rocket modeprovides 15% more thrust than the all-rocket mode atMach 12.

Figure 11a - Scram-Rocket Mode Thrust

Figure 11b shows the Isp values for the scram-rocket mode performance. Also included are the all-rocket mode Isp adjusted for the altitude, scramjetmode Isp values, and the 'averaged' Isp values.

Figure 11b - Scram-Rocket Mode Isp

AIAA 99-2104

12

ON-LINE PLOTTING CAPABILITY

The SCCREAM web site at:

http://titan.cad.gatech.edu/~jebradfo

is continuously being improved. The latestcapabilities in dynamic HTML and JavaScript arebeing used to increase the capabilities and improvethe work environment for the SCCREAM user.Previously, the output returned by SCCREAM waslimited to a POST engine deck and tables consistingof detailed results at each design point. In addition tothese, the web interface now returns performanceplots for ejector mode engine performance.

To allow plotting of the ejector mode data,additional operations must be performed by the CGIscript. These operations involve the use of freewareplotting packages and a graphics interpreter. Thissoftware must be accessible to the web CGI script inorder to generate the plots.

After the SCCREAM executable hassuccessfully finished analyzing the engine'sperformance, a PERL script executes a 'C' programthat parses through the POST engine deck and createsa set of arrays that can be read by the plottingpackage.

Figure 12 - SCCREAM v.5 Output Page

After the 'C' script has finished parsing throughthe POST deck, the freeware plotting package called'GNUPLOT' is executed. A script with commandsfor GNUPLOT is piped into the executable from acommand line call executed by the PERL script.GNUPLOT then reads the arrays generated by the 'C'script and creates two different 2-D plots of thrustand Isp versus Mach number at every altitudeanalyzed. The first plot set are color postscript filesand the second are black and white postscript files(PS). The color postscript file is later converted to acolor JPEG image using the freeware program'ghostscript'. This JPEG image can then be displayedto the user over the web. The PS file can also bedownloaded by the user and provides a crisp, blackand white image for the user to send to their localprinter [12,13].

Figure 13a - Ejector Mode Thrust Plot from Web

Figure 12 is a snapshot from the SCCREAMweb site showing the various links to the engine dataafter a successful analysis. Figures 13a and 13b areJPEG images taken directly off the SCCREAM webpage. They were generated for a SuperchargedEjector Ramjet (SERJ) engine with a sea-level staticthrust of 75,000 lbs and fan pressure ratio of 1.2. Therocket primary used LOX/JP-5 propellants and theafterburner used JP-5.

AIAA 99-2104

13

Figure 13b - Ejector Mode Isp Plot from Web

The new plotting capability allows the designerto quickly assess their engine’s performance andprovides an easy method of comparing differentengines.

ROCKET PRIMARY AND AFTERBURNERPHI THROTTLE

SCCREAM now has the ability to create POSTdecks for throttled engines. This will allow thetrajectory analyst more control over optimizing theflight path of the vehicle. The rocket primary isthrottled in ejector mode and the remaining modesthrottle the afterburner equivalence ratio. Figure 14provides a sample POST deck for a throttled engine'sramjet mode Isp. The throtle is represented by thevariable 'genv5'. For example, at an altitude of 50,000feet, a speed of Mach 3.0, and a phi throttle of 0.667,the engine produces an Isp of 1612.06 seconds.

c R a m j e t M o d e I s p V a l u e s

l $ t a b t a b l e = 5 h i s p 3 t , 3 , 5 h g e n v 5 ,

4 h m a c h , 5 h g d a l t , 4 , 4 , 5 , 2 7 * 1 ,

5 0 0 0 0 ,

3 . 0 ,

1 , 1 4 1 3 . 3 2 ,

0 . 6 6 7 , 1 6 1 2 . 0 6 ,0 . 3 3 3 , 1 8 4 5 . 8 8 ,

0 , 0 ,4 . 0 ,

1 , 1 4 6 0 . 4 7 ,

0 . 6 6 7 , 1 5 7 2 . 5 3 ,0 . 3 3 3 , 1 7 3 8 . 7 3 ,

0 , 0 ,

5 . 0 ,

1 , 1 3 3 8 . 8 7 ,

0 . 6 6 7 , 1 3 7 2 . 4 3 ,0 . 3 3 3 , 1 2 2 1 . 6 5 ,

Figure 14 - Sample Throttled POST Deck

EXPERT AND NOVICE VERSIONS

Two different web interfaces for SCCREAM arenow available to the user, an 'expert' mode and a'novice' mode. The latter is a 'scaled-down' interfacewhich should facilitate and encourage the use ofSCCREAM in classroom environments.

The original SCCREAM interface is now the'expert' interface. This allows the user full access toall of SCCREAM's modeling options and designvariables. For the engineer who is new to the RBCCengine design process, the 'novice' interface isrecommended. This version has significantly fewernumber of design variables than the 'expert' mode.All second-order engine parameters like fuelinjection velocity, heat release profile, stationefficiencies, etc.. have been set to nominal values.The novice user is first encouraged to understand thebasic principles of designing an engine. Thisincludes thrust level to inlet area matching at sea-level static conditions, engine geometry effects onsubsonic and supersonic combustion, andpropellant/fuel type analysis.

AIAA 99-2104

14

CONCLUSIONS

SCCREAM's analysis capability has beensignificantly expanded since version 4. As the toolcontinues to develop, it is expected that its usage willbecome even more commonplace. Over the pastyear, web access to SCCREAM has increasedsignificantly. A number of universities have usedSCCREAM in student engine and vehicle designcompetitions and are now hosting their ownSCCREAM web sites. Web statistics indicate theGeorgia Tech SCCREAM site has been accessed byusers from 8 different countries. In the past year,SCCREAM has analyzed over 1,500 different enginedesigns (over 500,000 flight conditions)

Among the conclusions drawn from thispaper are the following:

1. The new combustor model has been shown tocompare very well with the preliminary designtool, SRGULL's modeling capability. Thisincreased level of fidelity has not significantlyincreased the run-time for SCCREAM.Hundred's of flight conditions can still beanalyzed in under 2 minutes.

2. SCCREAM is now capable of modeling 4different rocket primary propellant combinationsand 3 different afterburner fuels. These newcapabilities have been validated with the industryaccepted code RJPA. SCCREAM is flexibleenough to model any combination of thesepropellants and fuels.

3. The new scram-rocket capability will allow forimproved modeling accuracy of the scramjet toall-rocket mode transition. It also provides awider flight envelope for the trajectoryoptimization process.

4. The web-based plotting capability will enableSCCREAM users to quickly and convenientlyview and interpret their engines performance.

FUTURE WORK

SCCREAM will continue to be improved withthe intent of increasing modeling accuracy andcapabilities without sacrificing speed, ease of use,and flexibility. Among many near-termimprovements being considered for version 6.0 arethe following:

1. Improving the inlet pressure recovery model byadding geometry dependent shock systemanalysis. This will replace the current curve fitmodels. A method for incorporating a normalshock model for subsonic combustion modes hasbeen devised but not implemented. A techniquefor supersonic combustion cases is still underdevelopment.

2. A propane rocket primary subsystem andafterburner fuel option will be added. Theprocedure for implementing this will beconsistent with current methods of analysis forhydrocarbon fuels.

3. The post-SCCREAM analysis plotting capabilitywill be expanded to include all engine modes, aswell as 3-D contour plots.

4. A method will be established for determiningangle of attack effects. This is a fairly simpleprocedure for the wedge configuration, but theredoes not appear to be a quick solution for conicalflows at an angle of attack. Once generated,these effects will be added to the POST deck forincorporation into the trajectory analysis.

ACKNOWLEDGMENTS

Current development of SCCREAM issponsored by the NASA - Marshall Space FlightCenter under the Graduate Student ResearchersProgram, grant number NGT8-52854 to the GeorgiaInstitute of Technology.

The authors would like to thank Mr. D.R. Komarat NASA MSFC for his assistance with obtaining theSRGULL results for the combustor comparisons.

AIAA 99-2104

15

The authors would also like to thank members ofthe Georgia Tech Stargazer design team for providingreference concept data - Laura Ledsinger, JohnBradford, David McCormick, Kirk Sorenson, AshrafCharania, Becca Cutri-Kohart, Kris Cowart, andDavid Way.

REFERENCES

1. Escher, William J. D. and B. J. Flornes, A Studyof Composite Propulsion Systems For AdvancedLaunch Vehicle Application. Contract NAS7-377. The Marquardt Corporation: Van Nuys,California 1966. Vol 1-7.

2. Brauer, G. L., et al., “Program to OptimizeSimulated Trajectories (POST).” Final report forNASA contract NAS1-18147, Martin-MariettaCorp., September 1989.

3. Bradford, J. E. and J. Olds, “Improvements andEnhancements to SCCEAM, A ConceptualRBCC Engine Analysis Tool.” AIAA-98-3775.1998. Conference Proceeding of the 34thAIAA/ASME/SAE/ASEE Joint PropulsionConference, Cleveland, OH.

4. Olds, J.R. and J.E. Bradford, “SCCREAM-Simulated Combined Cycle Rocket EngineAnalysis Module”. AIAA-97-2760. 1997.Conference Proceeding of the 32nd

AIAA/ASME/SAE Joint Propulsion Conferencein Seattle, WA.

5. Zucrow, M. J., J. Hoffman Gas Dynamics,Volume 1, Wiley, New York, NY, 1976.

6. Pinckney, S. Z. and Walton, J. T., "ProgramSRGULL: An Advanced Engineering Model forthe Prediction of Airframe-InegratedSubsonic/Supersonic Hydrogen CombustionRamjet Cycle Performance". NASP TM-1120,1991.

7. Gordon, S. and B. J. McBride, "ComputerProgram for Calculation of Complex ChemicalEquilibrium Compositions and Applications",

NASA Reference Publication 1311, October1994.

8. JMP Statistical Software Package, Version 5.0,SAS Institute. Cary, NC.

9. Rossini, F., et al. Selected Values of Physicaland Thermodynamic Properties of Hydrocarbonsand Related Compounds, Carnegie Press,Pittsburgh, PA, 1953.

10. Turns, S. R. An Introduction to Combustion,Concepts and Applications, McGraw-Hill, Inc.New York, NY, 1996.

11. Pandolfini, P. “Instructions for Using RamjetPerformance Analysis (RJPA) IBM-PC Version1.24”, JHU/APL AL-92-P175, June 1992.

12. GNUPLOT Plotting Package, Unix Version 3.7,Williams, T., et al., (c)1986.

13. Ghostscript Postscript Viewer, Version 5.03,Aladdin Enterprises, 1997, Menlo Park, CA.