A Multidisciplinary Performance Analysis of a Lifting-Body Single-Stage-to-Orbit Vehicle

14 Pages·2000·1 MB·English

by  Tartabini, Paul V.

AIAA 2000-1045
A Multidisciplinary Performance
Analysis of a Lifting-Body
Single-Stage-to-Orbit Vehicle
Paul V. Tartabini
Roger A. Lepsch
J. J. Korte
Kathryn E. Wurster
NASA Langley Research Center
Hampton, VA 23681-2199
38th Aerospace Sciences
Meeting & Exhibit
10-13 January 2000 / Reno, NV
For permission to copy or republish, contact the American Institue of Aeronautics and Astronautics
1801 Alexander Bell Drive, Suite 500, Reston, VA 20191

AIAA-2000-1045
A MULTIDISCIPLINARY PERFORMANCE ANALYSIS OF A LIFTING-BODY
SINGLE-STAGE-TO-ORBIT VEHICLE
Paul V. Tartabini*, Roger A. Lepsch '_,J. J. Korte* and Kathryn E. Wurster _
NASA Langley Research Center
Hampton, Virginia 23681
ABSTRACT GLOW Gross Lift-off Weight, klbs
Isp Specific Impulse, sec
Lockheed Martin Skunk Works (LMSW) iscurrently LaRC NASA Langley Research Center
developing a single-stage-to-orbit reusable launch vehi- LH 2 Liquid Hydrogen
cle called VentureStarT_'t A team at NASA Langley Re- LMSW Lockheed Martin Skunk Works
search Center participated with LMSW in the screening LOX Liquid Oxygen
and evaluation of a number of early VentureStar TM con- MECO Main Engine Cut Off
figurations. The performance analyses that supported Me Edge Mach Number
these initial studies were conducted to assess the effect M Freestream Mach Number
of a lifting body shape, linear aerospike engine and me- O/F Total oxidizer-to-fuel ratio
tailic thermal protection system (TPS) on the weight and q Dynamic pressure, psf
performance of the vehicle. These performance studies q-(x Dynamic pressure times angle-of-attack,
were performed in a multidisciplinary fashion that indi- psf-deg
rectly linked the trajectory optimization with weight es- Re 0 Momentum Thickness Reynolds Number
timation and aerotherma] analysis tools. This approach RLV Reusable Launch Vehicle
was necessary todevelop optimized ascent and entry tra- S ' Aerodynamic reference area, ft2
.jectories that met all vehicle design constraints. T/W Thrust-to-weight ratio
Engine thrust-to-weight ratio
(T/_r)eng
Significant improvements in ascent performance TVC Thrust Vector Control
were achieved when the vehicle flew a lifting trajectory W Entry weight, Ibs
and varied the engine mixture ratio during flight. Also, a Empty weight, lbs
Wemply
considerable reduction in empty weight was possible by Wins inserted weight, lbs
adjusting the total oxidizer-to-fuel and lifloff thrust-to- X/L Body position over vehicle length
weight ratios. However, the optimal ascent flight profile (x Angle-of-attack, deg
had to be altered to ensure that the vehicle could be ,5payload Change in payload from baseline
trimmed in pitch using only the flow diverting capability
of the aerospike engine. Likewise, the optimal entry tra- INTRODUCTION
jectory had to be tailored to meet TPS heating rate and
transition constraints while satisfying a crossrange re- Many papers have been written describing the diffi-
quirement. culty of designing a fully reusable single-stage-to-orbit
launch vehicleJ _ The physical difficulty of this prob-
NOMENCLATURE lena is further exacerbated by the large degree of cou-
pling between the various design disciplines. Nearly ev-
cg Center-of-gravity ery subsystem design decision has far reaching
CL Lift coefficient consequences that must be evaluated in a multidisci-
*Research Engineel. Mail Stop 365, Vehicle Analysis Branch.
_Senior Research Enghwel. Mail Stop 365, Vehicle Analysis Branch, Member AIAA.
*Senior Research Engineet. Mail Stop 159, Muhidisciplinal T Optimization Branch, Senior Member AIAA.
_Senior Research Engineer, Mail StoI,365, Vehicle Analysis Branch, Associate Fellow AIAA.
'RTheuse of trademarks or names of manufacturers in this report is for accurate reporling and does not constitute an official
endorsement, either expressed orimplied, ofsuch products ormanufacturers bythe National Aeronautics and Space Administration.
Copyright ©2000 American Institute ofAeronautics and Astronautics, Inc. No copyright isasserted intheUnited States underTitle
17, U.S. Code. The U.S. Government has a royahy-free license to exercise all rights under the copyright claimed herein for
Governmental purposes. All other rights are reserved bythe copyright owner.
1
American Institute of Aeronautics and Astronautics
plinaryfashion
inorder to assess the impact on the weight Throughout the duration of the LaRC study, the fea-
and performance of the entire vehicle. This paper dis- sibility of numerous configurations was evaluated in
cusses this process as it relates to the conceptual design terms of vehicle mass and payload capability. These mass
and analysis of Lockheed Marlin Skunk Works' (LMSW) properties are directly related to Ihe overall performance
proposed single-stage-to-orbit commerciaI reusable ofthc vehicle. Since many design parameters affect both
launch vehicle (RLV), VenturcStar TM. weight and performance, accurate determination of ve-
hicle sizing information requires a multidisciplinary ap-
LMSW iscurrently studying the VentureStar'" RLV proach to performance analysis. The approach utilized
which has a number of unique features that differentiate in this study indirectly coupled trajectory optimization,
it from other SSTO concepts evaluated in the past. Chief weight estimation, and heating analysis tools to ascer-
among these differences is the lifting body shape of the tain the impact of various design options on the payload
vehicle and the utilization of the linear aerospike engine. capability of each configuration. Using this approach,
These design features and others, such as the use of a trade studies were conducted to maximize vehicle per-
metallic thermal protection system (TPS), provide nu- formance and cost effectiveness. The primary objective
merous benefits that may ultimately contribute to a ful- of these studies was to address design issues that pre-
ly capable vehicle that approaches NASA's goal of en- sented opportunities and challenges that were unique to
abling commercial access to space at an order of VentureStaF". Specifically, this paper addresses issues
magnitude less than today's cost/A representative Ven- that pertain to 1) the effect of the lifting body shape and
tureStar'" configuration is shown in Fig. I. aerospike engine on ascent performance, 2) the ability
to influence vehicle sizing by varying the engine size
and mixture ratio, and 3) the ramifications of the metal-
lic thermal protection system on the entry trajectory de-
sign. A multidisciplinary approach was the only way to
ensure that the system fully exploited the performance
benefits offered by the unique VentureStar design while
TM
staying within the operational limits imposed by its cost-
saving elements.
APPROACH
Many of the trades discussed in this paper required
the calculation of various physical characteristics of the
vehicle including payload capability, empty weight and
gross lift-offweight (GLOW). These quantities were pre-
dicted using a multi-disciplinary analysis that included
trajectory optimization, weights and sizing estimation
Figure 1.A representative VentureStar TM and engine performance prediction.
Col!fig uration.
Trajectory optimization, which formed the core of
The decision to proceed with the full-scale devel- this analysis method, was performed using the three-de-
opment of VentureStar 'Mwill be made upon the conclu- gree-of-freedom version of the Program to Optimize
sion of the X-33 program. This program is ajoint ven- Simulated Trajectories (POST). 6This program has been
ture between NASA and LMSW and includes the design, developed as a joint government/contractor effort, and
construction, and flight testing of thc X-33 in order to it isavailable and widely used within the aerospace com-
demonstrate technologies critical to the development of munity. Inputs to this code include Earth atmospheric
VentureStaF". In 1997, after a year of dedicated effort and gravitational models, system mass properties, en-
in thc design of the X-33, LMSW began to use the les- gine performance, and vehicle aerodynamics.
sons learned toimprove the conceptual design of the full-
scale VentureStar During this phase of the program, Mass property estinaation was conducted using the
TM.
NASA Langley Research Center (LaRC) participated in Configuration Sizing program (CONSIZ)? CONSIZ uti-
the design, analysis and screening of a number of differ- lizes parametric mass-estimating relationships based on
ent vehicle concepts and configurations? historical rcgrcssion, finite element analysis and tech-
nology readiness level. With knowledge of the vehicle
layout, CONSIZ can also be used to estimate the Ioca-
American Institute of Aeronautics and Astronautics
lion of the vehicle center-of-gravity for trim calculations. toapogee where the Orbital Maneuvering System (OMS)
In this paper, mass property data was generated by LaRC engines were used to circularize at the ISS altitude.
using CONSIZ and was calibrated to weight statements
released by LMSW. Another required input to POST was The aerospike engine performance reflected the
a propulsion data model thai was computed using a suite baseline propulsion system at the conclusion of this study
of computer codes that simulated linear aerospike en- with 8engines operating at achamber pressure of 2,500
= gine performance. These codes were able to generate an psia and an area ratio of 196. A LaRC version of this
aerospikc engine database as afunction of altitude, mix- database was created that provided realistic engine per-
ture ratio, power level and thrust vectoring for use with- formance across the mixture ratio/power level envelope,
in the POST trajectory code. The capabilities and devel- thereby minimizing the amount of interpolation error.
opment of these tools are discussed in Reference 8. A
final input to the trajectory optimization was an aerody- The nominal VentureStar ascent trajectory was
TM
namic database that included coefficients for lift, drag determined by maximizing the weight inserted into the
and pitching moment as a function of Mach number, target orbit. The trajectory was optimized by adjusting
angle-of-attack and control surface deflection. This da- the pitch attitude, engine power level and engine mix-
tabase was obtained through the blending of solutions ture ratio flight profiles. These control variables were
from computational methods and wind tunnel data. Up- constrained by angle-of-attack and engine operating lim-
dales to the aerodynamic coefficients were made peri- its. An additional constraint was imposed on the mix-
odically during the conceptual design phase to reflect ture ratio profile which had to be varied such that the
configuration shape changes. ratio of oxidizer to fuel was consistent with the vehicle
tank design value of 6.0. The optimized trajectory had
All of the entry performance trades discussed inthis to meet anumber of inflight constraints including limits
paper were performed using POST and MINIVERY on axial acceleration, dynamic pressure, q-m, and angle
POST was the ideal tool to perform entry trajectory op- of attack. The parameter q-_ isproportional to the struc-
timization, but was limited in its ability to provide heat- tural loading of the vehicle during flight and is the prod-
ing environments. An aerothermal analysis tool, MINI- uct of dynamic pressure and angle-of-attack. The nomi-
VER, was chosen to complement POST and to provide nal trajectory was untrimmed. A subsequent section will
a reliable measure of the heating levels during trajecto- discuss the trim capability of the aerospike thrust vector
ry development. MINIVER is an engineering code that control (TVC) system and the effect of a trim constraint
uses approximate heating methods with simple flowfietd on vehicle performance. The values of all flight con-
and geometric shapes to model heating on critical re- straints and control limits are listed in Table I.
gions of the vehicle. MINIVER has been used exten-
sively as a preliminary design tool and has demonstrat- The significant flight parameter profiles for the nom-
ed excellent agreement with more detailed heating inal VentureStar ascent trajectory are shown in Figs.
TM
solutions for stagnation and windward acreage areas on 2a and 2b. Main engine cut off (MECO) took place near
awide variety of vehicle configurations. __-" In addition perigee of the transfer orbit, at an altitude of 57 nmi and
to providing sufficiently accurate heating levels, it can a flight path angle near 0.1°. The peak dynamic pressure
be used to predict the onset of transition to turbulent flow.
Table 1. Constraints imposed upon nominal
RESULTS AND DISCUSSION ascent trajectory
Ascent Performance
Constraint Name Constraint Value
The figure of merit used for the ascent trade studies Final Orbit 50 × 248 nmi
was payload capability. This paranaeter was determined Final Inclination 51.6 deg
for each vehicle assuming aconstant propellant volume, Axial Acceleration Limit a <3g's
which ensured that the vehicles being compared were Dynamic Pressure Limit q <600 psf
similar in size (dry weight), thus minimizing errors due q-c_Limits lq-o_l< 1500 psf
to scaling. The design reference mission for the Ventur- Angle-of-attack Limits -2 deg < c__<12 dog
cStar RLV was delivery of a 25-klb payload to the In- Lift-off Thrust to Weight Ratio 1.35
TM
ternational Space Station (ISS). The nominal ascent tra- Overall Tank Ratio 6.0
jectory began with launch at Kennedy Space Center and Engine Power Level 0.2 < power level _<1.0
ended with the insertion of the payload into a 50 × 248 Engine Mixture Ratio 5.5 -<mixture ratio < 7.0
nmi orbit inclined 51.6 °. At this point the RLV coasted
American Institute of Aeronautics and Astronautics
Structural Constraint
,00 -
350 I- 281- 15 / One feature of the VentureStar _' design that could
I be exploited during ascent was its lifting body shape.
300. 24 F" 12 /11 I
By flying a lilting trajectory, it was possible to signifi-
25O 20 _- 9 it Altitu cantly decrease the amount of gravity losses, thereby
5t_! / ,, Meo. improving vehicle performance and payload capability.
_£200 t,
Yet increasing the amount of lift during ascent general-
150 - . ////
lyrequired flight at higher angles-of-attack and resulted
100 - 8}- 0| "- --"" /' "'-. in greater stress on the vehicle structure. Accordingly,
the nominal trajectory was constrained to keep the pa-
50-
4}- -3} / i'" rameter q-_ below a 1500 psf-deg structural design lim-
O- it to ensure that the aerodynamic loads did not exceed
50 100 150 200 250 300 350 400
Altitude, kft the structural capability of the vehicle. The effect of this
trajectory constraint on vehicle performance is shown
a. Altitude, Mac& and angle-of-attack profiles. in Fig. 3.There was a substantial benefit associated with
using lift during ascent since flying a non-lifting trajec-
tory resulted in apayload penalty of over 1000 lbs com-
5 - 600 _ 2000_
pared to the nominal case. As the q-o_ limit was raised
Ic\_,Dynamic pressure the payload increased because the vehicle could use more
i
4 - 480 I- 150C
lift to further reduce gravity losses. Eventually, the ad-
I i
ditional payload capability flattened out at a q-c_ limit
1oio! , I Acceleration
3 - _ 360 I- near 5000 psf-deg because higher drag losses, which are
also incurred by fying alifting trajectory, began toover-
2- _ 24o__ 5oo come the corresponding decrease in gravity losses. The
I results shown inFig. 3only reflected the effect of chang-
1
1 - 120 !t- ° i J• __"_. _-_._._ q._ ........ ing the structural limit on vehicle performance and did
not account for the impact of structural weight increases
/!
that would be required to design a vehicle that could
I
0 - 0L -500 • , I endure higher aerodynamic loading.
0 5O 100 150 200 250 300 350 400
Altitude, kfl
1000
b. Acceleration, dynamic pressure, and q-o_profiles.
Figure 2. Important flight parameter profiles fi_r the 500
nominal ascent traject<n T.
Baseline
/
of 540 psf occurred at a Mach number of I.I and the q- A
limit of 1500 psf-deg was held for roughly 25 sec. from
Mach 0.4 to 0.6. Although these two structural parame-
ters were kept within their required limits throughout the -500
trajectory, the peak normal force was nearly 2.3 times
the empty weight of the vehicle and may present struc-
tural problems since itis near the normal load factor limit -1000 I i t i I _ I i I
0 1000 2000 3000 4000 5000
of 2.5 used by the Space Shuttle. The engine was flown Peak q-m psf-deg
at 100q power level from ]iftoff until the 3-g axial ac-
celeration limit was reached at approximately 125 scc into Figure 3. Effect of the q-o_structural design constraint
flight (h = 112 kft, M = 4.2). At this point the engine was on vehicle payload capability.
gradually throttled down to nearly 20G at MECO to
maintain the 3-g limit. The engine mixture ratio was Axial Accck;ration Limit
varied continuously throughout ascent. The manner in
which this parameter is varied during flight has a signif- A series of ascent trajectories was optimized at a
icant effect and will be discussed in more detail later. range of axial acceleration constraints from 3 to 5-g's.
Changing this constraint had asmall effect on the vehi-
American Institute of Aeronautics and Astronautics
clepayloacdapabilit(yseeFig.4).Increasinthgeaccel-
included TVC data was used to assess the impact of a
erationlimitfrom3.0to3.3-g'sresulteidnslightlylow-
trim constraint on the nominal ascent trajectory. The
ergravitylossesw,hichledtoanadditiona3l00Ibsof
trimmed trajectory was designed such that no more than
payloacdapabilityA.sthelimitwasincreasebdeyond
50U, of the control authority was used throughout the
3.3-g'st,helowergravitylossewsereeradicatebdyan ascent. The vehicle cg was varied linearly with weight
increasienthrusvtectorinlgosse(stheAVrequiretdo
from the lift-offposition of 39.9c_ of the reference length
turnthevelocityvector)T.oinseritntoa50x248nmi- (nose to cow[) to the MECO value of 77.6c_ ,.In order to
orbit,thevehiclemuspterformmuchofitspitch-oveart keep the amount of TVC used bclow 50c_, the angle-of-
altitudehsighenougthoavoidaccumulatinegxcessive
attack had to be kept within the envelope shown in Fig.
draglossesA.stheg-limitwasincreasetdh,evelocityat
5. Also shown in Fig. 5are the angle-of-attack profiles
whichthispitch-oveorccurrebdecamheighearndmore
for the trimmed and untrimmed nominal ascent trajecto-
energyhadtobeexpendetodturnthevelocityvector.
ries. At altitudes below 40-kft, the angle-of-attack had
Althoughchangintgheacceleratiolinmithadasmall
to bc kept lower than optimal in order to meet the trim
effecotnpayloacdapabilitiyncreasinthgelimitmaybe
constraint. This loss of optimality led to an I100-1b pen-
beneficiatoltheengindeevelopmesnintceahighelrim-
ally in vehicle payload capability. Note that the TVC
itenablehsigheprowelrevelsatMECOandshortebrurn effectiveness varied with altitude and was lowest at 35-
times(seeFig.4).Theformecrouldfacilitattehedesign
kft where the angle-of-attack had to be kept within a
oftheenginceontroslystemwhilethelattermayincrease 1.3°window.
theoveralelnginelife.
3 50 r 360 18
16 ................................... Upper limit
fl
14 ,p
40 _ 34G \'\°
12
\, Min. power level _"
10 ] -- Trimmed
t '_ 30_ _ 32G •% _ 8 // --- Untrimmed
_ '_,,_ Burn time deg 6
0 " _ 300
0
10 _ 280
-2
-4
0L- 260 L ---_,_ J J .1...... a_._._t_ =.. J -6 I _ [ = I i t t I _--._L _=L
30 3.5 4,0 4.5 5.0 0 50 1O0 150 200 250 300 350
Axial acceleration limit, g's Altitude, kft
Figure 4. Effect of the axial acceleration limit on Figure 5. Effect of trim constraint on angle-of-attack
vehicle payhmd capability. profile for nominal ascent trajectoo'.
Pilch Trim Capability The baseline configuration discussed in this paper
had the LOX tank positioned in the nose of the vehicle.
The VentureStar is intended to be trimmed during One trade that was considered by the LaRC team was
TM
ascent using only thrust vector control (TVC) supplied moving the LOX tank to the aft end of the vehicle which
by the linear aerospike engine. The aerospike could pro- could potentially decrease the weight of the LH_ tank
duce a thrust moment to counteract the aerodynamic and the intertank structure significantly. One concern
pitching moment of the vehicle by diverting up to 15% with moving the LOX tank aft, however, was the effect
of the outflow from the top engine banks to the bottom of the resulting rearward shift in cg location on thc abil-
engine banks, or vice-versa. Diverting the flow created ity of the vehicle to trim during ascent. The effect of
a couple from the difference in axial force between the pitch trim was computed for two LOX aft vehicles that
top and bottom of the engine (although the net axial force were modeled as having a constant cg position during
remained unchanged). This couple was the largest con- ascent at locations of 78_ and 82% of the reference body
tributor to the total thrust moment (there was an addi- length (nose to cowl), respectively. Figure 6 compares
tional smaller contributor due to differences in the nor- the trim capability of these LOX aft configurations with
mal force acting on the top and bottom engine ramps). the baseline LOX forward configuration. These results
An engine performance database created by LaRC that show that the TVC requirements actually decrcascd the
American Institute of Aeronautics and Astronautics
tio of total LOX weight to total LH_, weight) had to be
I00
kept consistent with the vehicle propellant tank design
(baseline O/F was 6.0). The mixture ratio profiles for
50% control authority each case are shown inFig. 7along with the correspond-
50
ing effect that each had on the Ist,profile. By continu-
25 ously varying the mixture ratio, the inserted weight could
TVC be increased by 1900 Ibs over thc constant case and 200
available, 0
%
lbs over the step case. In the step and variablc cases the
-25
mixture ratio was initially set to 6.5 because thc slight
-50 increase inthrust that could be gained by increasing the
-- LOX tank fwO mixture ratio to 7.0 was overpowered by a greater toss
-75 --- LOX tank aft -78% e.g.
---- LOX tank aft - 82% c.g in Is_,.Thc variable case differed from the step case at
-100 _=a.=.=.=J==_-_a_.,_l I I _ t i . I low altitudes because the mixture ratio profile could be
50 100 150 200 250 300 350
Altitude, kft tailored to take advantage of alower Ist,caused by shock
interference with the nozzle wall (see Reference 8). The
Figure 6. Comparison of the thrust vector control profile differed at higher altitudes because it was more
profile required for pitch trim for the baseline and two efficient to lower the thrust to meet the g-limit by de-
LOX-aft configurations. creasing mixture ratio (variable ease) rather than power
level (step case). The added design complexity of acon-
more the longitudinal cg was moved aft. This behavior tinuously varying engine over one that performs a step
occurred because most of the thrust moment was due to change may not be worth the small accompanying per-
the couple created by the difference in axial force be- forrnance increase.
tween the top and bottom engine banks. Since this cou-
ple was independent of longitudinal cg position, the
maximum attainable thrust moment did not change much
6.5
as the cg was moved. The effect of the cg location was Mi.xtureo 7U.
much stronger on the aerodynamic moment, and a rear-
ward shift in cg generally resulted in smaller aerody-
5.0
namic moments for the low angles-of-attack seen dur-
ing ascent. Therefore, the net effect of moving the LOX
tank aft was to increase the pitch trim control margin.
Although moving the LOX tank aft may ease pitch trim _ VariaNe
concerns during ascent, more work must be done to ful- Isp /f ---- Step
ly understand the effect of such a move on pitch trim
during entry. J - Constant
i I J I J I _ I I J i I = I
Engine Performance Trades 0 50 i00 i50 200 250 300 350
Altitude, kfl
With the aerospike engine it was possible to vary Figure 7.Engine mixture ratio profiles and
the mixture ratio between values of 5.5 and 7.0 during corresponding effect on lsp during flight.
flight. In general, as the engine m!xture ratio was in-
creased, total thrust incrcased and the specific impulse Vehicle Sizin_ Studies
(Is_,)decreased, ldcally the mixture ratio should be set
high early in flight, where high thrust levels arc required The iifioff T/W and total O/F ratios were critical
to accelerate the fuel-heavy vehicle, and later transitioned parameters that influenced the weight and performance
to the lowest allowable value to maximize vacuum Isp. of the vehicle. Since both of these parameters affected
Generally engine development becomes more compli- weight and performance in opposite ways, it was impor-
cated and expensive as thc flexibility of thc engine to tant to link the weight estimation and traiectory optimi-
vary mixture ratio is increased. zation in order to capture the combined effect. An itera-
tive process was undertaken that coupled the results of
A performance assessment was made for three dif- POST and CONSIZ. This process began with an initial
ferent modes of mixture ratio adjustment during flight guess of the mass ratio (GLOW/W,,,) which is a mea-
(constant, step, and continuously varying). In cases where sure of vehicle performance. Next, values for the liftoff
the mixture ratio was varied, the total O/F ratio (the ra- T/W and O/F ratio were selected. All three of these pa-
6
American Institute of Aeronautics and Astronautics
rameterwsereinputintoCONSIZtodetermine
their ef-
1.20
0 Performance only - Isp effecl
fect on the weight of the VentureStar vehicle. The mass
TM [] Weights only - bulk density effect
ratio essentially determined the total propellant load of 1,15 .A. Weights and performance
the vehicle. The liftoffT/W was directly proportional to
1,10
the size of the engine and impacted the weight of the
propulsion system (engines, feed system, etc) and thrust 1.05
structure. The total OfF ratio determined the propellant Normalize0
empty 1.00
bulk density and consequently affected the weight of the weight
.95
tanks and propellants.
.90
In this analysis, CONSIZ was used to size thc vehi-
.85
cle to deliver a 25,000-1b payload to the ISS. That is, as
changes were made to mass ratio, liftoff T/W and the .80 , I , t , f , I , I , I , I
5,6 5,8 6.0 6.2 64 6.6 6.8 70
total OfF, the vehicle was photographically scaled to Total O/F
maintain a fixed 25,000-1b payload. Favorable changes
to these three parameters resulted in a vehicle that was Figure 8. Variation of vehicle empO' weight with total
smaller than the baseline (in physical dimensions and O/F ratio (normalized with baseline).
empty weight) and could still deliver the required pay-
load. The vehicle was scaled under the assumption that lifloff T/W of 1.35 was approximately 6.5 and was char-
the volumetric efficiency (ratio of tank volume to total acterized by a reduction in empty weight of over 2.5%
volume) remained constant. The impact of this assump- from the baseline. Two additional curves are shown in
tion was not significant since vehicles were not scaled Fig. 8todemonstrate the importance of coupling the per-
by huge amounts (less than 15q_ throughout the study) formance and sizing analyses. When only performance
and the results were concerned primarily with changes changes due to varying the total OfF were considered
in empty weight rather than absolute values. Using this (i.e., weight changes due to varying the propellant bulk
technique the effect of the lifioff T/W and total OfF on density were neglected) the empty weight increased rap-
the empty weight was determined with the assumption idly as the total OfF was raised. This increase occurred
that a vehicle with a lower empty weight would ulti- because the engine had tooperate at high mixture ratios
mately have lower development costs. for alonger duration during the trajectory inorder to raise
the total OfF, thus decreasing the average ls_,and increas-
The other key component of this iterative sizing pro- ing the mass ratio. On the other hand, if the mass ratio
cess was the optimization of the ascent trajectory using was assumed to remain constant as the total OfF ratio
the weights calculated with CONSIZ. The same mission was changed (i.e., the effect of total OfF on performance
and constraints listed in Table 1were used in the trajec- was neglected) then the empty weight went down as the
tory calculations. For each optimized traiectory the en- total OfF was increased. This improvement in empty
gine performance model was scaled to meet the required weight occurred because of the increase in propellant
liftoffT/W and the engine mixture ratio was varied such bulk density resulting from the higher ratio of liquid
that the total O/F was consistent with what was used in oxygen to liquid hydrogen. With a larger bulk density,
the weight calculation. Trajectories were determined that more propellant could be held inagiven volume, so that,
maximized the weight inserted into orbit, (which was with no penalty in performance, a smaller vehicle could
equivalent to the lowest possible mass ratio). Once the carry the same amount of propellant and consequently
minimum mass ratio was determined from the trajecto- deliver the same payload to the target orbit. In reality,
ry optimization, it was fed back into CONSIZ and this both effects work against each other, resulting in the
whole process was repeated until convergence. This en- actual curve that was minimized around 6.5. This trend
tire multidisciplinary analysis was performed for enough was nearly independent of the liftoffT/W, with the opti-
values of T/W and OfF to demonstrate how the empty mal ratio of LOX to LH., weight being around 6.5 for
weight changed with respect to the baseline vehicle. values of liftoff T/W between I.15and 1.35 (sec Fig. 9).
The effect of the total OfF ratio on the empty weight A similar trade was conducted todetermine the sen-
is shown inFig. 8. All values have been normalized with sitivity of empty weight to liftoff T/W. The results of
respect to the baseline vehicle (T/W = 1.35, OfF = 6.0). this trade assuming a total OfF ratio of 6.0 are presented
Changing the OfF ratio had a notable effect on empty in Figure 10. Each point on the curve was determined
weight and led to differences of more than 5_ between using the same iterative process that was used in the OfF
the best and worst values. The optimal OfF value for a trade. The empty weight was minimized when the lift-
American Institute of Aeronautics and Astronautics
actually two engines-out (an additional engine must be
1.10
powered down to eliminate thrust imbalances between
108 O TIVV =110
each side of thc linear aerospike). This would limit the
[] TIW =1.23
106 T/W =1.35 lower T/W bound to about 1.25 assuming a 10e_ throttle
104 up of the remaining engines.
1.02
Normalized The trend of empty weight with lifioffT/W was in-
empty 1.00
weight dependent of the total ratio of LOX to LH2, with the
.98
optimal T/W being about 1.23 for O/F ratios between
,96
6.0 and 6.5 (sec Fig. II). As shown in the figure, a 5c_
.94
decrease in empty weight is possible by changing the
.92 baseline values of liftoff T/W from 1.35 to 1.23 and to-
tal O/F from 6.0 to 6.5, abort concerns notwithstanding.
5.60 5.80 6.00 6.20 6.40 6,60 6,80 7.00
Total O.E Also shown in Fig. !I is the influence of the weight of
the engine per pound of thrust (engine T/W) on the opti-
Figure 9.Effect of liftoff T/W o11optimal O/F ratio. mal value oflifloffT/W. The engine T/W is a key input
to the weights model and directly influences the weight
offT/W was approximately 1.23. Changing the T/W from of a number of propulsion system elements. For a de-
1.35 (baseline) to 1.23resulted ina 2% decrease in empty crease in engine T/W of 10%, the trend of empty weight
weight. If gross lift-off weight was minimized instead with respect to lift-off T/W was unchanged, although
of empty weight (also shown in Fig. 10), the optimal the empty weight increased uniformly by over 6%.
T/W was somewhat higher (between 1.30 and 1.35). The
optimal T/W differed between the two curves because
the increase in engine weight that would accompany a 1'I0 F
higher liftoffT/W was amuch larger percentage of empty
1.08 F
weight than GLOW. Therefore the performance benefit
1.06 _- 10% lower (T/W}eng
from a higher T/W was overcome by the added engine 1.04 _-
weight sooner when empty weight was minimized. Since 1.02
cost is more closely related to the empty weight rather Normalized Baseline _ j,.,,,,,,,,_] O/F =6.0
empty 1.00
than GLOW, the lower T/W of 1.23 would likely lead to weight
.98
the lower cost configuration. However, the benefit in
,96 O/F 65
empty weight resulting from a lower T/W would have
to be weighed against the affect such a change might .94
have on the engine-out abort capability of the vehicle. .92
For VentureStar'" in parlicular, a single engine-out is .90 i 1 _ I f I i J I t i [ i I
1.10 1.15 1.20 1.25 1.30 1.35 t.40 1,45
Liflofl T[W
1.10
Figure 11. Effect of O/F ratio and engine T/W on
t.08 O GLOW
[] wo_p,y optimal liftoff T/W.
1.04 Entry_ Trajectory Performance
1.02
Normalized An aerospace vehicle that operates inthe hyperson-
weight 1,00
ic flight regime must be protected from the aerodynam-
.98
ic heating environment. The use of arelatively low-tem-
.96
perature (- 1800°F max) metallic thermal protection
.94
system has been proposed to meet this need for the Ven-
.92 tureStar RLV. In contrast, the current Shuttle orbiter
TM
[ t I I , I i I i. I l employs relatively high-temperature ceramic tiles
1.10 1.i5 1.20 125 1.30 1.35 1.40 1.45
Liftoff TNV (-2600°F max). Although the metallic system may of-
fer some benefit in reduced maintcnance requirements,
Figure 10. Variation of gross liftoff weight and its lower temperature capability necessarily constrains
vehicle empty weight with lifmff T/W ratio the vehicle flight envelope so that excessive heating lev-
(normalized with baseline). els are avoided. Thc lowest achicvable peak laminar
8
American Institute of Aeronautics and Astronautics