AIAA/CIRA 13th International Space Planes and Hypersonics Systems and Technologies AIAA 2005-3207

13th AIAA/CIRA International Space Planes and Hypersonic Systems

and Technologies Conference, 16-20, May, 2005, Capua, Italy

Application of Optimization Algorithms to Scramjet Inlet Design

Susumu Hasegawa*, Doyle Knight#

* Space Propulsion Research Center

Institute of Space Technology and Aeronautics
Japan Aerospace Exploration Agency
Kimigaya, Kakuda, Miyagi 981-1525, JAPAN

# Department of Mechanical and Aerospace Engineering

Rutgers University - The State University of New Jersey
98 Brett Road, Piscataway, NJ, 08854-8058

An automated design optimization process is applied to both single and multi ob-
jective optimization problems of scramjet inlet design. This optimization process inte-
grates together an optimizer with a mesh generator, a flow solver, and an objective
analysis tool into an automated optimization loops because the flow simulation is re-
quired for every step along the line search and finding the feasible direction. This paper
presents the implementation of these new design techniques by the gradient-based
optimizer Sequential Quadratic Programming (SQP) and their application to scramjet
inlet case in flight condition of Mach 8. The performance of scramjet inlets with uniform
inflow is improved, and the optimized functions, that is, the total pressure recovery co-
efficient increases by approximately 10%. The trade-off (also known as the ε-constraint)
method is applied and implemented to find the Pareto optimal set formed by the non-
dominated solutions of the feasible design. The objective functions are the total pressure
loss and the drag, and some solutions are obtained to analyze the relations between the
total pressure loss and the drag.

Introduction vehicle show its revolutionary engine

A scramjet engine is the key enabling
technology for sustained hypersonic flight,
and significant research is in progress
worldwide on scramjet engine design and
performance. New research developments
including flight tests of scramjet engines
were achieved in the U.S. in 2004.
NASA's X-43A research vehicle screamed
into the record books, demonstrating an
air-breathing engine can fly at nearly 10
times the speed of sound. Preliminary
data from the scramjet-powered research
*Researcher, Numerical Space Engine
Team, JAXA, AIAA member
# Professor, AIAA Associate Fellow.
Copyright©2005 by S. Hasegawa and D.
Knight. Published by the American Insti-
tute of Aeronautics and Astronautics, Inc.
with permission.
worked successfully at nearly Mach 9.8,
as it flew at about 110,000 feet.
Supersonic combustion ramjets
(scramjets) promise more airplane-like
operations for increased affordability,
flexibility and safety in ultra high-speed
flights within the atmosphere and for the
first stage to Earth orbit. The scramjet
advantage is once it is accelerated to
about Mach 4 by a conventional jet en-
gine or booster rocket, it can fly at hy-
personic speeds, possibly as fast as Mach
15, without carrying heavy oxygen tanks,
as rockets must.
Japan's new space agency officially
commenced its activities from 1, October,
2003. "Japan Aerospace Exploration
Agency (JAXA)" is Japan's largest space
agency with the Institute of Space and
Aeronautical Science (ISAS), National

-1-
American Institute of Aeronautics and Astronautics Paper 2005-3207

Copyright © 2005 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.
 13th AIAA/CIRA International Space Planes and Hypersonic Systems
and Technologies Conference, 16-20, May, 2005, Capua, Italy
Aerospace Laboratory (NAL) and Na-
tional Space Development Agency
(NASDA) consolidated. JAXA has been
conducting scramjet engine research at
the flight conditions of Mach 4, 6 and 8 by
using the RamJet engine Test Facility
(RJTF)1-5 at Space Propulsion Research
Center. A free-piston high-enthalpy shock
tunnel (HIEST)1,6, the largest in the
world, was completed in November 1997.
Scramjet Engine Testing in the range of
Mach 8 to 15 has been conducted. In ad-
dition to the large scale experimental
research, numerical computations of
scramjet engines are conducted on vector
parallel supercomputers7-10. The com-
prehensive research activities are sum-
marized in Ref.1.
Design optimization in aerodynamics
and propulsion has always been of great
interest and has, therefore, received
much attention11-18. The direct result of
improvement in computer technology is
the growing interest in automated design
techniques of both single and multi ob-
jective optimization19-22 aimed at achiev-
ing better design in a shorter period of
time than classical methods, i.e., by hand
or extensive analysis.
The design of an efficient compression
system for an airframe-integrated
scramjet engine is essential for accept-
able performance of a hypersonic air-
breathing vehicle. The goal in the design
of a scramjet inlet is to define a minimum
weight geometry that provides an effi-
cient compression process, and provides
these characteristics over a wide range of
flight and engine operating conditions.
The design of a scramjet inlet is compli-
cated by the many constraints, both
aerodynamic and mechanical, that are
imposed on the inlet. Examples of aero-
dynamic constraints include starting
limits, boundary layer separation limits,
and constraints on combustor entrance
flow profiles. Examples of mechanical
constraints include variable geometry
flexibility and cooling system limits. The
design of scramjet inlets is influenced
greatly by vehicle and flight constraints.
This paper focuses on the implemen-
-2-
American Institute of Aeronautics and Astronautics Paper 2005-3207
tation and application of an efficient and
powerful tool to optimize the aerody-
namic performance of a two-dimensional
scramjet inlet. The results for improved
scramjet inlets are described using the
single objective optimization. In addition,
multi objective optimizations are con-
structed, and some solutions are obtained
by applying the trade-off method.
Geometry Model
A set of parameters is needed to de-
fine the two-dimensional scramjet inlet
geometry so that a design space can be
formed for the optimization. Fig.1 shows
the model of the two-dimensional scram-
jet inlet geometry which is composed of 6
variables and 5 fixed parameters given in
Table 1 and 2. The geometrical model has
two parts23.
The first part is the inlet (compres-
sion system). The inlet shown in Fig.1
has three ramps for multiple compres-
sions. The angle of first ramp, second
ramp, and third ramp are indicated by
θ1, θ2, and θ3 , respectively.
The second is the isolator, constant
section area part. Since the flow speed
decreases and the pressure rises during
combustion at supersonic speeds - this
phenomenon is reversed at subsonic
speeds - a scramjet is fitted with an ad-
ditional element, the isolator, between
the inlet and the combustion chamber.
The isolator's function is to isolate the
combustion chamber before injection. It is
supposed to prevent the pressure force,
which rises during supersonic combus-
tion, from affecting the inlet flow via the
bordering walls. This could lead to
blocking the inlet. A phenomenon called
"a shock train" forms within the isolator,
consisting of a changing succession of
compression shock waves and expansions.
It is caused by the interaction of shock
waves and the boundary layer of the iso-
lator's wall and causes a (beneficial) fur-
ther rise in flow pressure.
The geometrical constraints are lin-
ear inequality constraints, and a nonlin-
 13th AIAA/CIRA International Space Planes and Hypersonic Systems
and Technologies Conference, 16-20, May, 2005, Capua, Italy

ear equality constraint. Besides the ob- of Wilcox [Ref.23] to generate the inflow
vious geometrical constraints such as, profile. Program EDDYBL is a program
suitable for computing properties of
0 ≤ θ1 ≤ θ2 ≤ θ3 (radians) laminar, transitional and turbulent
boundary layers. The program embodies
0.2 ≤ X1 ≤ X 2 ≤ 1.25 (m) a wide variety of turbulence models
0.2 ≤ Xe ≤ 1.25 (m) ranging from mixing-length oriented al-
gebraic models to a complete stress
the following constraint is required: transport model. The freestream condi-
tions at the entrance to the inlet are de-
(X1 − 0.2) tan θ1 + (X 2 − X1 ) tan θ2 scribed in Table.3. The condition corre-
sponds to the RJTF experimental condi-
+ (1.25 − X 2 ) tan θ3 = 0.15 (m) tion in the flight condition of Mach 8.

The above geometrical contraction is

a nonlinear equality contraction in order Optimization Algorithm
to fix the geometrical contraction ratio.
We consider the general nonlinear
problem
Numerical Simulations
Minimize F(x) (or Maximize F(x))
Reynolds Averaged Navier-Stokes s.t. G(x) ≦ 0
equations with a turbulence model are
used as governing equations to investi- where x =（x1, x2, …, xn）is the vector
gate influence of shock waves, expansion of design variables, F(x) is a scalar func-
waves, boundary layers, and compression tion and G(x) is a vector-valued set of
waves on flowfield inside a scramjet en-
nonlinear constraints. F(x) is the fitness
gine.
function characterizing how much the
It utilizes a finite volume spatial dis-
shape suits the optimization problem.
cretization in which the state variables
G(x) represents all the constraints of the
are stored at the cell centers. The accu-
problem. The design variables xi are
racy of reconstructing the primitive
typically geometric parameters defining
variable field at the cell faces determines
the inlet shape（e.g., lengths and angles).
the spatial accuracy of the solution. The
There are two families of optimization
MUSCL approach is performed for vari-
algorithms. One family is local algo-
able extrapolation.
rithms such as gradient methods19. An-
In our computations, the inviscid flux
other family is constituted by global al-
scheme is Roe’s Method with third-order
gorithms such as Genetic Algorithm (GA),
spatial accuracy restruction using the
and Simulated Annealing (SA).20-22 In
Min Mod limiter. Turbulence model is
our research, the optimization of the
represented by the Wilcox k-omega model.
two-dimensional scramjet inlet began to
A steady state solution is obtained by
be performed by the gradient-based algo-
applying implicit scheme.
rithms. From a basic shape, the gradi-
In this computation, the grid is gen-
ent-based algorithms are able to quickly
erated by a Fortran program. A
find an optimum. However, they have
two-dimensional, 4 zone grid is used for
some drawbacks that make their use dif-
the computation as shown in Fig.2. The
ficult for complex industrial problems,
grid dimensions are 82×70 (Zone 1),
(e.g., they can be trapped by local min-
108×70 (Zone 2), 82×18 (Zone 3), 108×18
ima.) Another drawback of the gradi-
(Zone 4).
ent-based algorithms is their sensitivity
We use the k-omega model in the
to design spaces where a large amount of
EDDYBL turbulent boundary layer code

-3-
American Institute of Aeronautics and Astronautics Paper 2005-3207
 13th AIAA/CIRA International Space Planes and Hypersonic Systems
and Technologies Conference, 16-20, May, 2005, Capua, Italy
individuals are not valuable. The objec-
tive function is then discontinuous so
that the computation of gradient has no
real significance. The method used in this
research is C code for Feasible Sequential
Quadratic Programming (CFSQP)19.
CFSQP is a set of C functions for the
minimization of the maximum of a set of
smooth objective functions (possibly a
single one) subject to nonlinear equality
and inequality constraints, linear equal-
ity and inequality constraints, and simple
bounds on the variables. In addition,
CFSQP contains special provisions for
efficiently handling problems with many
sequentially related objec-
tives/constraints, for example discretized
Semi-Infinite Programming (SIP) prob-
lems. CFSQP is an implementation of two
algorithms based on Sequential Quad-
ratic Programming (SQP), modified so as
to generate feasible iterates. In the first
one (monotone line search), a certain
Armijo type arc search is used with the
property that the step of one is eventually
accepted, a requirement for superlinear
convergence. In the second one the same
effect is achieved by means of a non-
monotone search along a straight line.
The merit function used in both searches
is the maximum of the objective functions
if there is no nonlinear equality con-
straints, or an exact penalty function if
nonlinear equality constraints are pre-
sent.
Automatic Optimization Methodology
The automated optimization process
is shown in Figure 3. The optimization
software which has been developed and
used for the present research is the gra-
dient-based method, that is CFSQP. The
simulation part has implied the devel-
opment of a solver fast enough to allow a
large number of calls and accurate
enough so as to correctly predict the
trends between investigated configura-
tions. Outside of the loop, some verifica-
tion are made using Reynolds-averaged
Navier-Stokes simulations.
-4-
American Institute of Aeronautics and Astronautics Paper 2005-3207
The control of the optimization loop
implemented for designing the scramjet
inlet is provided by the optimizer. The
optimizer generates candidate inlet de-
signs described by five geometrical pa-
rameters. These parameters are passed
to the analysis module which returns the
aerodynamic performance of this candi-
date inlet. Based on these simulation
results, the objective function and some
aerodynamic constraints are calculated.
Single Objective Optimization
In this single optimization program,
the objective function is the total pres-
sure recovery coefficient η. The total
pressure recovery coefficient η is repre-
sentative of the efficiency of the flow de-
celeration performed by the inlet. To
maximize the total pressure recovery is
as same as to maximize the kinetic en-
ergy efficiency. The relation between the
total pressure recovery and the kinetic
energy efficiency is given in Ref.25-26.
The total pressure recovery coefficient η
is obtained numerically as follows:
Pt exit
η=
Pt 0
where Ptexit and Pto are respectively the
averaged total pressure in the exit plane
of the inlet and the free stream total
pressure. Therefore, the problem ad-
dressed in the following can be expressed:
Maximize η
s.t. Geometrical Constraints
Several optimizations are performed
with the gradient-based optimizer
CFSQP. Different settings of CFSQP have
been tried. CFSQP, as a gradient-based
optimizer, requires a starting point to
initialize the search process. The initial
set of geometric parameters is the case of
single ramp (θ=θ2=θ3).
The uniform inflow condition is assumed
and the numerical simulation of flow-
 13th AIAA/CIRA International Space Planes and Hypersonic Systems
and Technologies Conference, 16-20, May, 2005, Capua, Italy
fields are conducted under the condition
of Table 3. The initial total pressure re-
covery coefficient η0 is obtained as 0.545.
The flowfield is shown in Fig. 4(a). The
shock waves from the ramp collides with
the cowl, and then the reflected shock
from the cowl collide around the corner
between the inlet part and the isolator
part. The separation region appears due
to the shock collision.
The perturbation used by CFSQP is
approximating the gradients by finite
difference is called udelta. Different val-
ues of udelta have been investigated,
showing its important influence on the
optimization behavior. Moreover, udelta
equally affects the six design variables
used to parameterize the inlet geometry
which are passed to CFSQP, where they
are angle or distance.
The description of the five CFSQP
optimizations is provided in Table 4. The
setting parameters and the total pressure
recovery coefficient are also described in
this Table. Depending on the runs, from 5
to 13 line searches are performed during
these optimizations. The maximum value
obtained by these computations is ηmax
=0.599.
The maximum value is higher than the
initial value by 9.9%. The flow field is
shown in Fig.4 (b). The improved con-
figuration is compared with the initial
configuration. Though the initial con-
figuration has single ramp, the improved
configuration has triple ramp. The sepa-
ration gets smaller and the flow around
the corner of the improved configuration
becomes closer to the shock-cancel flow
than the flow around the corner of the
initial configuration.
With this low value of udelta, CFSQP
has explored a very limited region of the
design space around the starting point,
which explains the poor result obtained
in terms of the total pressure recovery
coefficient.
Multi Objective Optimization
The purpose of this study is to apply
-5-
American Institute of Aeronautics and Astronautics Paper 2005-3207
multi objective optimization algorithms
to scramjet inlet design. The tools and the
algorithms of our optimization study are
summarized. Tools are the software
packages and the hardware resources
required for our analysis. Algorithms are
our overall design optimization method-
ology and the multi-objective optimiza-
tion methodologies applied in our study.
Having more than one objective function,
our problem is a multi-objective optimi-
zation problem.
Minimize fi (x) (or Maximiza fi (x) )
i=1,…,m
Optimization problems with more
than one objective are known as
multi-objective or multi-criteria deci-
sion-making problems. Most of the design
problems in engineering have several
objectives. These objectives can conflict
with one another. Therefore, rather than
finding one single solution, a set of good
compromises are sought. The decision
maker searches for the values of design
variables, x= (x1,…,xk ), which optimize
the objective functions, fi, i=1,…,m, si-
multaneously. Since a vector of design
variables that are acceptable by the deci-
sion maker are searched, this type of op-
timization problems is also known as
vector optimization. The set of solutions
from which decision maker selects the
desired values of design variables is
called the Pareto optimal set.
A set of points, x*, is Pareto Optimal
if there exists no feasible vector of deci-
sion variables, x∈F , which would de-
crease any one of the objective functions
without causing a simultaneous increase
in at least one other objective function.
The vector x* corresponding to the solu-
tions included in the Pareto optimal set
are called nondominated, A design xa is
dominated by a design xb if fi (xb)≦fi (xa),
for i=1, …, k, and for at least one value of
i, fi (xb)＜fi (xa), where k is the dimension
of vector x. The set of nondominated de-
signs in the Pareto set. The plot of the
objective functions whose nondominated
 13th AIAA/CIRA International Space Planes and Hypersonic Systems
and Technologies Conference, 16-20, May, 2005, Capua, Italy
vectors are in the Pareto set is called the
Pareto front.
Multi-objective optimization problems
can be treated as single-objective prob-
lems by scalarizing the objective vector
into a single objective. There are different
methods available in the literature for
scalarization of the objective functions.
The trade-off (also known as the
ε-constraint) method is applied to solve
multi objective optimization problem of
scramjet inlet design.
The trade-off (the ε-constraint)
method is not biased towards convex por-
tions of the Pareto set. It selects a single
objective function to minimize. All other
objective functions are restricted with
user-specified values. The problem with
many objectives is transferred into a sin-
gle objective problem with an additional
set of constraints. The user defined ε
values are varied to determine the com-
plete Pareto set. It is critical to specify a
valid range of the ε values to avoid empty
solution sets. This method is able to as-
sociate with non-convex parts of the
Pareto front. The mathematical formula-
tion is as follows:
Minimize fi (x) (or Maximiza fi (x) )
s.t. fj (x) ≦ ε , j≠i
In order to implement this single ob-
jective design optimization methodology
to the multi-objective optimization prob-
lems, a modified form of the objective
function f(x) is used. In case of optimiza-
tion based on the trade-off (ε-constraint)
approach, one of the objective functions is
minimized while restricting the other.
Fig.5 displays a flow chart of the optimi-
zation algorithm based on the trade-off
method.
In this multi-optimization program,
the objective functions are the total
pressure loss ζ (=1−η= f1) and the drag D
(=f2). The total pressure loss is introduced
to solve the problem for convenience and
the relation between the total pressure
recovery and the total pressure loss is
described above.
-6-
American Institute of Aeronautics and Astronautics Paper 2005-3207
Therefore, the problem addressed in
the following can be expressed:
Minimize f1=ζ
Minimize f2=D
s.t. Geometrical Constraints
The trade-off (ε-constraint) method is
applied to solve above multi-objective
optimization problem, and the problem is
scalarized as follows.
Minimize f1= ζ
s.t. f2=D≦ε,
and Geometrical Constraints
The trade-off method is implemented and
several optimizations are performed with
the gradient-based optimizer CFSQP. The
processes to get optimum solutions are as
same as the process described in previous
section. Different settings of CFSQP have
been tried. CFSQP, as a gradient-based
optimizer, requires a starting point to
initialize the search process. The initial
set of geometric parameters is the case of
single ramp (θ1=θ2=θ3). The initial total
pressure loss is 0.455 and the initial drag
is 2985N. By repeating the processes to
get the optimum solution, some optimum
solutions are obtained and shown in Fig.
6.
Conclusions and Futrure Work
An automated design of a two dimen-
sional scramjet inlet has been achieved.
The objective of the single optimization is
to increase the total pressure recovery
coefficient while maintaining the geo-
metrical contraction ratio. Computations
of the scramjet inlet are performed and
an automated optimal design of the
scramjet inlet is performed using six de-
sign variables and five fixed parameters
by using the gradient-based optimized
method, SQP method. The scramjet inlets
with the uniform inflow are improved,
and the optimized functions, that is, the
total pressure recovery coefficient are
 13th AIAA/CIRA International Space Planes and Hypersonic Systems
and Technologies Conference, 16-20, May, 2005, Capua, Italy
improved by approximately 10%.
The trade-off method is applied to
solve multi-objective optimization prob-
lem of scramjet inlet design. Multi objec-
tive optimization algorithm is imple-
mented and some solutions are obtained
to analyze the relations between the total
pressure loss and the drag.
The optimized function will be veri-
fied by the global optimization method
such as Genetic Algorithms.
Several additional factors needed to
be considered in future research for prac-
tical application of the results obtained in
this study. These include 1) the effect of
three dimensional condition, 2) the sen-
sitivity of the inflow boundary layer, 3)
several flight Mach conditions.
Acknowledgement
One of the authors (S. Hasegawa)
stayed at Rutgers University as a visiting
scientist under the financial support by
JAXA (Japan Aerospace Exploration
Agency) and MEXT (Ministry of Educa-
tion, Culture, Sports, Science and Tech-
nology, Japan). The author would like to
thank all the people concerned for their
assistance.
References
1 Chinzei, N. Mitani, T., and Yatsuy-
anagi, Y., “Scramjet Engine Research at
the National Aerospace Laboratory in
Japan,” Scramjet Propulsion. Edited by E.
T. Curran and S.N.B, Murthy, Vol. 189,
Progress in Astronautics and Aeronautics,
AIAA, Reston, VA, 2001, pp.159-222.
2 Kanda,T., Hiraiwa,T., Mitani,T.,
Tomioka, S., and Chinzei, N., “Mach 6
Testing of a Scramjet Engine Model,”
Journal of Propulsion and Power, Vol.13,
No.4, 1997, pp.543-551.
3 Sato S., Izumikawa, M., Tomioka, S.
and Mitani, T., “Scramjet Engine Test at
the Mach 6 Flight Condition,” AIAA Pa-
per 97-3021, July 1997.
4 Kanda, T., Wakamatsu, Y., Ono, F.,
-7-
American Institute of Aeronautics and Astronautics Paper 2005-3207
Kudo, K., Murakami, A., and Izumikawa,
M., “Mach 8 Testing of a Scramjet Engine
Model,” AIAA Paper 99-0617, Jan. 1999.
5 Hiraiwa, T. Kanda, T., Mitani, T.,
“Experiments on a Scramjet Engine with
Ramp-Compression Inlet at Mach 8 Con-
dition,” AIAA Paper 2002-4129, July
2002.
6 Itoh, K., Ueda, S., Komuro, T., Sato,
K., Tanno, H., and Takahashi, M., “Hy-
pervelocity Aerothermodynamic and
Propulsion Research Using a High En-
thalpy Shock Tunnel HIEST,” AIAA Pa-
per 99-4960, Nov. 1999.
7 Hasegawa, S., Sunami, T., Takaha-
shi, M., Mochizuki, M., Miyajima, H.,
Itoh, K., Vogels, M.E.S., “The Virtual Test
Bed Environment at NAL-Kakuda Re-
search Center,” Parallel Computational
Fluid Dynamics 1999, pp.233-240, El-
sevier Science, Amsterdam, 2000.
8 Hasegawa, S., Tani, K. and Sato, S.,
“Numerical Simulation of Scramjet Inlets
on a Vector-Parallel Supercomputer,”
Parallel Computational Fluid Dynamics
2001, pp.159-166, Elsevier Science, Am-
sterdam, 2002.
9 Hasegawa, S., Tani, K. and Sato,
S.,“Aerodynamic Computation of a
Scramjet Engine on Vector-Parallel Su-
percomputers,” Parallel Computational
Fluid Dynamics 2002, Elsevier Science,
Amsterdam, 2003.
10 Hasegawa, S., Tani, K. and Sato,
S.,“Aerodynamic Analysis of Scramjet
Engines under the Flight Conditions of
Mach 6,”AIAA Paper 2002-5128, Oct.
2002.
11 Shukla, V., Gelsey, A., Schwabacher,
M., Smith, D., and Knight, D., `` Auto-
mated Redesign of the NASA P8 Hyper-
sonic Inlet Using Numerical Optimiza-
tion'', AIAA Paper No. 96-2549, 1996.
12 Zha, G.-C., Smith, D., Schwabacher,
M., Rasheed, K., Gelsey, A., and Knight,
D., ``High Performance Supersonic Mis-
sile Inlet Design Using Automated Opti-
mization'', AIAA Paper No. 96-4142,
1996.
13 Shukla, V., Gelsey, A., Schwabacher,
M., Smith, D., and Knight, D., ``Auto-
mated Design Optimization for the P2
 13th AIAA/CIRA International Space Planes and Hypersonic Systems
and Technologies Conference, 16-20, May, 2005, Capua, Italy
and P8 Hypersonic Inlets'', Journal of
Aircraft, Vol. 34, No. 2, March-April 1997,
pp. 228-235.
14 Zha, G.-C., Smith, D., Schwabacher,
M., Rasheed, K., Gelsey, A., Knight, D.,
and Haas, M., ``High Performance Su-
personic Missile Inlet Design Using
Automated Optimization'', Journal of
Aircraft, Vol. 34, No. 6, Novem-
ber-December 1997, pp. 697-705.
15 Blaize, M., Knight, D., and Rasheed,
K., ``Automated Optimal Design of Two
Dimensional High Speed Missile Inlets'',
AIAA Paper No. 98-0950, 1998.
16 Knight, D., ``Automated Optimal
Design of Supersonic and Subsonic Dif-
fusers Using CFD'', ECCOMAS 2000,
Barcelona, Spain, September 2000.
17 Knight, D., ``Application of Genetic
Algorithms to High Speed Air Intake De-
sign'', VKI Lecture Series Course on Ge-
netic Algorithms for Optimisation in
Aeronautics and Turbomachinery, May
2000.
18 Taskinoglu, E., Jovanovic, V.,
Knight, D. and Elliott, G.,
"Multi-objective Design Optimization and
Experimental Measurements for a Sub-
merged Inlet", AIAA Paper No. 2004-0025,
2004.
19 Lawrence, C., Zhou, J., and Tits, A.,
User's Guide for CFSQP Version 2.1: A C
Code for Solving (Large Scale) Con-
-8-
American Institute of Aeronautics and Astronautics Paper 2005-3207
strained Nonlinear (Minimax) Optimiza-
tion Problems, Generating Iterates Sat-
isfying All Inequality Constraints, Elec-
trical Engineering Department and In-
stitute for Systems Research, University
of Maryland, College Park, MD 20742
20 Michalewicz, Z., `` Genetic Algo-
rithms + Data Structures = Evolution
Programs'', Springer Verlag, 3rd Revision
edition, Oct., 1997.
21 Michalewicz, Z., Fogel, D., `` How to
Solve It: Modern Heuristics'', Springer
Verlag, Dec., 1999.
22 Deb, K., `` Multi-Objective Optimi-
zation using Evolutionary Algorithms'',
John Wiley & Sons, 2001.
23 Wilcox, D., “Turbulence Modeling
for CFD”, DCW Industries, 2nd edition,
Dec., 2002.
24 Kanda, T., Kudo, K., “A Conceptual
Study of a Combined Cycle Engine for an
Aerospace Plane”, AIAA Paper No.
2002-5146, Oct. 2002.
25 Curran, E.T. Leingang, J., Carreiro,
L., and Petters, D., “Further Studies of
Kinetic Energy Methods in High Speed
Ramjet Cycle Analysis,” AIAA Paper
92-3805, July. 1992.
26 Heiser, W.H., Pratt, D.T., “Hyper-
sonic Airbreathing Propulsion”, American
Institute of Aeronautics and Astronautics,
June, 1994.
13th AIAA/CIRA International Space Planes and Hypersonic Systems
and Technologies Conference, 16-20, May, 2005, Capua, Italy

Table 1 Design Variables

No. Parameter Definition
1 X1 Abscissa of end of first ramp
2 X2 Abscissa of end of second ramp
3 Xe Abscissa of cowl leading edge
4 θ Angle of first ramp
5 θ2 Angle of second ramp
6 θ3 Angle of third ramp

Table 2 Fixed Parameters

No. Parameter
1 R=150mm
2 R2=50mm
3 R3=20mm
4 L1=1250mm
5 L2=1450mm
Definition
Ordinate of ramp on the isolator
Width of the isolator
Width of the cowl
Abscissa of end of third ramp
Length of the inlet and the isolator

Table 3 Flight Condition

Flight Mach Number 8
Altitude 35 km
Inflow Mach Number 6.73
Inflow Velocity 2546 m/sec
Static Temperature 324 K
Static Pressure 1.6 kPa

Table 4 CFSQP optimizations in the case of uniform inflow condition

Run udelta ⊗ηmax
CFSQP 1 δ 0.2 %
CFSQP 2 2δ 9.9 %
CFSQP 3 3δ 4.2 %
CFSQP 4 4δ 1.1 %
CFSQP 5 5δ 1.6 %

-9-
American Institute of Aeronautics and Astronautics Paper 2005-3207
 13th AIAA/CIRA International Space Planes and Hypersonic Systems
and Technologies Conference, 16-20, May, 2005, Capua, Italy

Inlet Isolator

200mm

Xe

Air Flow
R3=20mm

R2=50mm

Y R1 =150mm
θ2 θ3
δ θ1
Origin X

X1

X2
L1=1250mm

L2=1250+200=1450mm

Figure 1 2D parameters of the inlet

Figure 2 Two-dimensional grid for RANS computations

Generates Returns
Optimizer
Candidate inlet Aerodynamic Performance

Mesh Generation Simulation Objective Analysis

Performance Analysis Module

Figure 3 Automated Optimization Loop

-10-
American Institute of Aeronautics and Astronautics Paper 2005-3207
13th AIAA/CIRA International Space Planes and Hypersonic Systems
and Technologies Conference, 16-20, May, 2005, Capua, Italy

(a) Initial Configuration with a single-ramp

(b) Improved Configuration with triple-ramp

Figure 4 Calculated Mach number distribution of the inlet

with uniform inflow condition of Mach 6.73

Figure 5 Flow Chart of Optimization based on the trade-off method

-11-
American Institute of Aeronautics and Astronautics Paper 2005-3207
13th AIAA/CIRA International Space Planes and Hypersonic Systems
and Technologies Conference, 16-20, May, 2005, Capua, Italy

3100

Solutions

3050

3000

2950

2900
0.4 0.42 0.44 0.46 0.48 0.5

Total Pressure Loss

Figure 6 Some solutions obtained by optimization based on the trade-off method

-12-
American Institute of Aeronautics and Astronautics Paper 2005-3207