10.0000@pubs - casi.CA@Generic A6504037A4BF

December 2005 dcembre
Volume 51, No. 4
ISSN 0008-2821
CANADIAN
Can. Aeronaut. Space J. Downloaded from pubs.casi.ca by UNIV MANITOBA on 06/22/15

For personal use only.
AERONAUTICS AND
SPACE JOURNAL
JOURNAL
ARONAUTIQUE ET
SPATIAL DU CANADA
Published by Publi par

the Canadian Aeronautics and Space Institute
lInstitut aronautique et spatial du Canada
105 1750, croissant Courtwood Crescent
Ottawa, Ontario, Canada K2C 2B5
Tel./Tl. : 613-234-0191, Fax/Tlc. : 613-234-9039
casi@casi.ca www.casi.ca
Canadian Aeronautics and Space Journal

EDITORIAL BOARD / CONSEIL DITORIAL

Editor / Directeur scientifique
Mr. Stewart Baillie
Director Flight Research Laboratory
IAR/NRC
Directeur, laboratoire de recherches en vol
IRA/CNRC
Associate Editors / Directeurs

scientifiques associs
Dr. Susan Skone
Associate Professor
Department of Geomatics Engineering
University of Calgary
Dr. David Zingg
Canada Research Chair in
Computational Aerodynamics
Institute of Aerospace Studies
University of Toronto
Dr. Nashed Youssef
Chair, CASI Propulsion Section
Pratt & Whitney, Canada
Dr. Alex Jablonski
Chair, CASI Astronautics Section
Canadian Space Agency / Agence spatiale
canadienne
Mr. Graeme Eastaugh
Research Officer, Structure and Materials
Performance Laboratory IAR/NRC /
Laboratoire des structures, des matriaux et
de la propulsion IRA/CNRC
SMPL/IAR/NRC LSMP/IRA/CNRC
Assistant to the Editor / Adjoint au

directeur scientifique
Angelo Fatoric
FRL/IAR/NRC LRV/IRA/CNRC
CASJ@nrc-cnrc.gc.ca
PRODUCTION
Publishing Services, NRC Research Press
National Research Council of Canada
Services ddition
Presses scientifiques du CNRC
Conseil national de recherche du Canada
PUBLISHED / PUBLI PAR

Canadian Aeronautics and Space Institute
Institut aronautique et spatiale du Canada
1750 Courtwood Cr., Suite 105
Ottawa, ON K2C 2B5
(613) 234-0191
publications@casi.ca
Journal aronautique et spatial du Canada
Scope
The Canadian Aeronautics and Space Journal, the
official publication of the Canadian Aeronautics
and Space Institute, is devoted to timely reporting
on recent archival research and development of
aerospace sciences to the international community.
Contents include a wide spectrum of technical
materials, ranging from advanced theoretical
treatises to dissemination of recent experimental
results, as well as book reviews and abstracts of
recent reports and publications of Canadian
aerospace research organizations.
Submission of manuscripts
The Institute invites the submission of new and
unpublished technical papers, articles, and notes for
publication. Instructions to authors can be found
elsewhere in this Journal or at: http://www.casi.ca/
index.php?pg=casj. Manuscripts should be submitted
in electronic format at http://mc.manuscriptcentral.
com/casj.
Publication charges
The organization sponsoring the work reported
upon in any paper or article published will be asked
to pay publication charges of $125 per typeset page.
Upon payment, the lead author will receive access
to an electronic copy of the Journal issue in which
the paper appears in Portable Document Format
(pdf) that will allow the production of an unlimited
number of hard copies of the paper.
Annual subscriptions
Annual subscriptions for CASI members is
complimentary. Annual subscriptions for non-CASI
members $88, includes access to electronic version
of CASJ plus delivery by post of each issue on CD
(taxes and postage extra). For more information
contact CASI.
Porte
Le Journal aronautique et spatial du Canada,
publication officielle de 1Institut aronautique et
spatial du Canada, se consacre au reportage
information opportune pour la communaut
internationale sur les recherches rcentes dans des
documents darchives et le dveloppement des
sciences arospatiales. Cette publication prsente
une grande diversit de documents techniques, allant
de traits thoriques approfondis aux rsultats de
rcents essais exprimentaux, en passant par des
comptes-rendus de livres et des rsums de
publications et de rapports rcents manant
dorganismes canadiens vous la recherche
arospatiale.
Prsentation de manuscrits
LInstitut encourage la soumission de rapports,
darticles et de notes techniques indits des fins de
publication. Les instructions aux auteurs se trouvent
dans le prsent journal ou ladresse Internet
suivante : http://www.casi.ca/index.php?pg=casj. La
soumission des manuscrits devrait seffectuer par
voie lectronique ladresse suivante : http://mc.
manuscriptcentral.com/casj.
Frais de publication
Lorganisme responsable de la publication dun texte
ou dun article doit payer des frais de publication de
lordre de 125 $ par page imprime. Aprs avoir
effectu son paiement, lauteur principal pourra
accder la version lectronique du numro du
Journal dans lequel a paru son article en format PDF,
qui lui permettra de produire un nombre illimit de
copies papier de larticle.
Abonnement annuel
Labonnement annuel pour les membres de lIASC
est gratuit. Labonnement annuel pour les
non-membres de lIASC est de $88, incluant laccs
la version lectronique du JASC et les frais denvoi
par la poste de chacun des numros sur CD (taxes et
frais de poste en sus). Pour plus dinformation,
contacter lIASC.
Volume 51, No. 4

CANADIAN AERONAUTICS AND SPACE JOURNAL

JOURNAL ARONAUTIQUE ET SPATIAL DU CANADA
Turbulence Model Studies to Investigate the Aerodynamic Performance of a NASA

Dual Control Missile at Supersonic Mach Numbers
M. Khalid, A. Dujardin, P. Hennig, L. Leavitt, F. Leopold, M. Mendenhall, S. Prince . 153
Aerodynamic Forces Approximations using the Chebyshev Method for Closed-Loop
Aero-servoelasticity Studies
Alin Dorian Dinu, Ruxandra Mihaela Botez, Iulian Cotoi . . . . . . . . . . . . . . . . . . . . 167
Applying Fast Fourier Transform Analysis and Data Window in Software Global
Positioning System Receivers to Mitigate Continuous Wave Interference under
Dynamic Conditions
Z. Jiang, G. Lachapelle, C. Ma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Adjoint-Based Sonic Boom Reduction for Wing-Body Configurations in Supersonic
Flow
Siva K. Nadarajah, Antony Jameson, Juan Alonso. . . . . . . . . . . . . . . . . . . . . . . . . . 187
Cover Image: The ability to manage the shock waves and the resulting sonic boom from a supersonic business jet will be a
determining factor in the overall acceptance of this aircraft type by the general public and its ability to over-fly populated areas.
Nadarajah et al. (pp. 187199) presents a numerical method for determining how the aircraft geometry alters these far-field pressure
distributions from a supersonic aircraft, as shown above.

Sur la couverture : La matrise des ondes de choc et du bang supersonique produits par un ract daffaires constitue un critre
crucial pour lacceptation gnrale de ce type davion par le grand public et lobtention de la permission de survol des zones
habites. Nadarajah et coll. (pages 187199) prsentent une mthode numrique permettant dtablir de quelle faon la gomtrie
dun avion supersonique modifie les distributions de pression dans le champ lointain, tel que montr ci-dessus.
This journal is indexed or abstracted in / Ce journal est signal ou rsum dans AIAA Aerospace and High Technology database,
Cambridge Scientific Abstracts Mechanical Engineering and Transportation Abstracts database, Cambridge Scientific Abstracts Civil
Engineering Abstracts, the Engineering Information Inc. Compendex, the National Research Council of Canada CISTI Source.
Canadian Aeronautics and Space Institute Institut aronautique et spatial du Canada

150 - 1750, croissant Courtwood Crescent, Ottawa, ON, Canada K2C 2B5
Tel./Tl. : 613-234-0191, Fax/Tlc. : 613-234-9039, E-mail/Courriel : casi@casi.ca
Publication Mail Registration No. 231452
ISSN 0008-2821
i

Vol. 51, No. 4, December 2005
Vol. 51, no 4, dcembre 2005
Turbulence Model Studies to Investigate the

Aerodynamic Performance of a NASA Dual
Control Missile at Supersonic Mach Numbers
M. Khalid * A. Dujardin ** P. Hennig *** L. Leavitt **** F. Leopold *****

M. Mendenhall ****** S. Prince *******
NOMENCLATURE
Abstract
This paper contains an investigation of the ability of some
of the current turbulence models to predict viscous effects
in supersonic Mach number regimes. In most cases, these
turbulence models have been derived and validated in
traditional subsonic to transonic Mach number regimes and
their application to supersonic and hypersonic regimes is
assumed valid without a proper recourse to understanding
the wider implications of the viscous flow field at high
Mach numbers. In such flow regimes, such characteristics
as the strong shock interactions with control surfaces and
evolving vortices present in the flow, or other large entropy
gradient effects brought about by the forebody or other fin
surfaces, produce non-trivial challenges for the turbulence
models. This study was aimed at interrogating the strength
of the turbulence models to model such physics. The
Applied Vehicle Technology Panel Group 082 of the
Cl, CL
lift coefficient (L/qS)
Cm
pitching moment M/qSl
CN
normal force coefficient
Cd
drag coeffiecient
CA(p)
axial force coefficient based on pressure
CA(f)
axial force coefficient based on friction

forces
missile diameter
lift
moment reference length (2.6 in) (1 in =

2.54 cm).
dynamic pressure(U2)
q
X
ratio of horizontal component of velocity

with respect to free stream
v/U, vz/V, v/V
ratio of normal velocity with respect to the

free stream
reference area (D2/4 = 5.31 in2)
continued on page 154
* Institute for Aerospace Research

National Research Council Canada
Ottawa, ON K1A 0R6, Canada.
E-mail: mahmood.khalid@nrc-cnrc.gc.ca
** Deutsches Zentrum fr Luft- und Raumfahrt
(DLR)
Cologne, Germany.
*** EADS
Lenkflugkrpersysteme GmbH (LFK)
Germany.
**** NASA
Langley, Virginia, USA.
***** ISL
France.
****** Nielsen Engineering
USA.
******* QinetiQ
UK.
Received 12 December 2005.
2005 CASI
axial distance
u/U, vx/V, u/V
INTRODUCTION
At supersonic Mach numbers, the classic problems
associated with drag prediction and such characteristics as
shock shape and shock boundary layer interactions together
with transition and heat transfer effects continue to take the
central stage. At higher angles of attack, the flow separation
problems become unavoidable and the long slender missile
geometry gives rise to the presence of unsteady vortex regions,
which evolve into Karman vortex sheets pervading over the aft
regions. Here, the wing CLMAX is often compromised as the
separated flow interacts with the developing vortices. At even
higher Mach numbers M > 5, other non-equilibrium and real
gas effects become important. Higher heat transfer rates lead to
surface ablation and other associated unsteady flow
complexities. Leading-edge regions on wings and (or) control
surfaces bear the brunt of higher temperature effects, which
often lead to loss of control.
From a design viewpoint the resulting optimized
configuration, in many ways would be a blend of excellence
from many disciplines. The best design would have conformed
153

suite de la page 153

Research Technology Organisation selected a dual control
NASA missile to investigate the capabilities of
Computational Fluid Dynamics (CFD) to predict the flow
field and performance characteristics of complex-shaped
projectiles at high Mach numbers.
There are two types of difficulties that are encountered
when computing such problems. The first type are
geometry based, and are related to the shape of the nose
and forebody, number and types of strakes or forward
control mechanisms, fin geometry deployment, and shape
and design of the cowls and intakes. The second type of
difficulty deals with flow complexities such as the heat
transfer (M > 4) implications; the vortices shed from the
forebody and other control surfaces; the interaction of
these vortices with the body structures, the free stream, and
with each other; the base flow interaction with the free
stream; the boundary-layer development; and, in the case
of supersonic flows, the shock boundary-layer interactions.
Air-breathing missiles with intakes or missiles with jetcontrolled guidance systems add to the flow complexities.
Various turbulence models employed in current CFD codes
are tested for their ability to address these viscous issues.
Rsum
Cet article prsente notre recherche sur la capacit de
certains modles actuels de turbulence prdire les effets
visqueux, aux nombres de Mach correspondant aux
rgimes supersoniques. La plupart de ces modles de
turbulence ont t drivs et valids pour une gamme de
rgimes allant du subsonique au transsonique donc pour
les nombres de Mach peu levs. On a prsum quils
sappliquaient
aux
rgimes
supersoniques
et
hypersoniques, mais sans tenir compte des consquences
plus larges pour le champ dcoulement fluide, aux grands
nombres de Mach. Certaines caractristiques de tels
rgimes dcoulement, notamment les interactions
intenses entre le choc et les gouvernes et lvolution des
tourbillons prsents dans lcoulement, ou les effets
considrables sur le gradient dentropie causs par le
fuselage avant ou les autres surfaces de gouverne
constituent des difficults dont la rsolution est ardue pour
les modles de turbulence. Cette tude visait examiner la
capacit des modles de turbulence reproduire ces
phnomnes physiques. Le Groupe 082 sur la technologie
applique aux vhicules, de lOrganisation pour la
recherche et la technologie, a choisi un missile double
commande de la NASA pour dterminer si la dynamique
numrique des fluides pouvait prdire le champ
dcoulement et les caractristiques comportementales des
projectiles aux formes complexes, se dplaant des
nombres de Mach levs.
Le calcul de tels problmes comporte deux types de
difficults. Le premier relve de la gomtrie et dcoule de
154
to the weight/structural and aero elastic issues. Its structural

components would be well suited for higher heat-transfer rates.
Its intake would be well embedded into the geometry and
plume and other vectored jet emissions controlled for minimum
radar cross section. Owing to advances in modeling techniques
and increased computer power, there is a strong compulsion in
missile design to rely more and more on numerical simulation
rather than opt for more exhaustive and expensive wind tunnel
testing. Knowledge gained per design cycle and faster iteration
times make strong arguments for using mathematical-based
models.
It is equally well recognized that the quality of turbulence
models in NavierStokes codes is pivotal to efficient and
accurate design of missiles. Towards this end, The Applied
Vehicle Technology (AVT) Panel of the Research Technology
Organisation (RTO) tasked Technical Group 082 to investigate
the applicability of various turbulence models to address the
complex viscous problems associated with the design of
modern missile configurations. The main objective of the
present research was to provide an assessment of the relative
cost and accuracy of the turbulence models used in modeling
viscous terms in NavierStokes solutions when predicting
missile aerodynamics at high subsonic and supersonic speeds.
The suggested Mach number range 0.8 < M < 4 made the scope
of the activity somewhat unique as it covered the challenging
transonic regimes where the flow is not so easy to resolve and
extended to upper supersonic regimes, where the
compressibility effects, shock wave/turbulent boundary layer
interactions and separation effects become important. It was
also recommended that during the course of the group activity,
a number of test-cases be defined and that various turbulence
models be applied to the selected set of test-cases to gauge the
accuracy and the relevant costs associated with different
closure schemes. The group was to study the predicted balance,
pressure, and other flow field data in light of the measured
quantities and arrive at some definitive conclusions. In the first
phase the group surveyed the literature for the previous
research in this area and researched into previous experimental
and modeling exercises carried out to study some of the
aforesaid problems. The team members made documented
presentations to the group, which are contained in the final
AVT/RTO Report (2004) of Task Group 082.
A more defined purpose of this particular study was to
investigate the capability of the turbulence models to accurately
perform CFD simulations of the flow past the NASA dual
control missile at two angles of attack, = 6 and 24. The
turbulence modeling studied during the course of this
investigation included the one equation SpalartAllmaras, the
two-equation standard k , and the k turbulence models as
well as the BaldwinLomax turbulence model enhanced to
account for the cross-flow separation due to DeganiSchiff
(DS) or other curvature effects from QinJayatunga. Some pure
laminar (lam) computations were also carried out to isolate the
viscous effects from complete computations. The group used
both structured and unstructured type grids to identify any
strengths or advantages of a particular method. The evolution
2005 CASI


la forme de la pointe et de lavant du fuselage, du nombre
et des types de listons ou de dispositifs de gouverne avant,
de la gomtrie de lempennage, et de la forme et la
conception des capots et des admissions. Le second type
provient des complexits de lcoulement, telles celles
dcoulant du transfert de chaleur (pour M > 4); les
tourbillons dtachs du fuselage avant et des autres plans
de gouverne; linteraction de ces tourbillons avec les
structures du fuselage et lcoulement libre, et entre eux;
linteraction de lcoulement de culot avec lcoulement
libre; la formation dune couche limite; et, en prsence
dcoulements supersoniques, les interactions entre la
couche limite et le choc. Les admissions dair des missiles
arobies ou les commandes dorientation par jets ajoutent
la complexit de lcoulement. Nous avons test la
capacit des diffrents modles de turbulences utilizs par
les logiciels actuels de dynamique numrique des fluides
traiter ces problmes de viscosit.
of the vortex formed by the front fins was of particular interest

as the vortices will undoubtedly interact very strongly with the
forward strakes at higher angles of attack and affect the
controllability of the projectile. Also, the flow deflected by the
forward strakes will subsequently impinge upon the aft control
surfaces, which could possibly lead to control failures if the
missile was not designed appropriately.
In the final analysis, based on the extent of wind tunnel data
available the group decided to use the NASA Tandem Control
Missile shown in Figure 1 as the main model for the group
study. Most balance and pressure data were to come from
NASA Langley, US, whereas other flow field data were
generated by ISL, France. During the test campaign at NASA,
two Mach numbers, M = 1.75 and M = 2.5 were investigated for
angles of attack varying from 4 to 28. The Reynolds number
for both velocities was set at 6.56 106/m (2 106/ft). For the
numerical study the first Mach number M = 1.75 was utilized
for most computations with two angles of attack settings: = 6
and 24. To provide greater confidence, at least one participant
completed a full drag and lift polar using the less timeconsuming coarser grid and appropriate turbulence model. In
terms of grid type studies, members agreed to work with both
structured and unstructured type grids. The original
unstructured grid was supplied by NASA, whereas QinetiQ in
UK, NLR from the Netherlands, and IAR from the National
Research Council of Canada supplied the structured grids that
were individually further adapted by various other members of
the team. While a more detailed analysis of the results are
reserved for a AVT/RTO report (2004), a synopsis of the
important findings are included in this paper. Except for the
first-order methods (DATCOM, MISL3), which make use of
the complete missile geometry and were only included to
provide a rough estimate of the global forces, most detailed
2005 CASI
computations in the absence of the yaw angle had adopted a

reflection plane model for NavierStokes computations. In
broader terms, the paper includes a comparison of various
viscous flow-field characteristics as computed using different
turbulence models. In some cases, specific velocity profiles at
some critical junctions were also compared against measured
data. Most computations recovered the force and moment
values quite successfully and these were also compared against
Euler and other empirical-based algorithms.
To measure the full impact of the turbulence modeling on
simulation of flows at high Mach number, the group decided to
look at results from lower order methods and quantify the
influence of turbulence modeling in higher order Navier
Stokes solutions. Towards this end, some members opted for
applying empirical-based and Euler codes to obtain the base
line results. Parabolized NavierStokes (PNS) codes and
Reynolds Averaged NavierStokes (RANS) codes with
different turbulence models were expected to provide most
accurate flow field and force and moment predictions.
The group mandate had asked for turbulence modeling
assessment studies in the Mach number range 0.8 < M < 4.
Owing to the limited resources available for such group
activities, it was decided to conduct CFD-based computations
at a free stream Mach number of M = 1.75, with unit Reynolds
number of Re/ft = 2 106 and two settings of angle of attack,
= 6 and 24. At the lower angle of attack = 6, the flow
should largely be attached on majority portions of the missile
and most CFD codes should have little problem in reaching a
satisfactory solution. At the higher angle of attack, there will
undoubtedly be regions of separated flows with well-developed
vortex structures invoking various control problems as they
interact with missile fins. Prediction of such flows will test the
accuracy, the economy, and the reliability of most turbulence
models. Viscous flows not withstanding, where possible,
results from simpler empirical based methods from DATCOM
or MISL3 would be introduced when comparing ordinary force
and moment predictions, as these tools provide a quick and
ready method of evaluating design performance under most
cruise conditions.
Finally, it was concluded that for the flow cases studied, the
SpalartAllmaras turbulence model was able to resolve the
physics as well as the two-equation turbulence model. Away
from the wall, most turbulence models captured such viscous
features as the velocity profiles quite well, however, such
viscous characteristics near the wall region were not as
accurately predicted.
SOLVERS
WIND Code
The computations at IAR were carried out using the WIND
CFD code version 2.0, supplied by the National Project for
Applications-oriented Research in CFD (NPARC) Alliance
(Bush et al., 1998). This code evaluates second-order accurate
finite differences of the governing NavierStokes partial
155

Figure 1. Dual control missile (all dimensions are in mm).
differential equations in conservative form. The explicit nonviscous terms were calculated using Roes second-order fluxdifference splitting algorithm; smoothing was used to dampen
numerical instabilities. Certain types of explicit operators for
example, the central difference operator for structured grids
may require the addition of numerical smoothing to dampen
instabilities that are a natural part of the scheme. Smoothing
must also be added explicitly when utilizing Winds
convergence acceleration capability. Values for various
smoothing parameters are specified in the input data file. The
default smoothing values for Jameson-type Euler solutions are
1/4, 1/256, and 0. For viscous solutions these values are as
much as 1, 1/64, and 2, respectively. The implicit viscous terms
were evaluated using either a full block implicit operator or a
parabolized NavierStokes (PNS) operator, depending on the
simulation. The SpalartAllmaras (1992) model was used to
calculate the turbulent viscous terms. For economy, the grid
was constrained to remain within y+ ~ 5 near the wall regions
and was thus thought to provide adequate sublayer resolution
for such viscous characteriztics as shear stress, skin friction,
and velocity profiles u/U and v/U. For a more accurate solution
based on the SpalartAllmaras model the wall distance y+ is
recommended to be less than 1.
CFD-FASTRAN Code
LFK (Lenkflugkrpersysteme GmbH) made use of
commercial code CFD-FASTRAN as well as the DLR FLOWer
code. CFD-FASTRAN is a density based NavierStokes solver
for compressible flows using Van Leers FVS scheme (Van
Leer, 1982). Among several turbulence models provided as
options in CFD-FASTRAN the standard k turbulence model
for high Reynolds number flows (Launder and Spalding, 1974)
was selected. Therefore, logarithmic wall functions were used
to avoid the necessity of resolving the laminar sublayer thus
permitting considerably fewer grid points in the near-wall
156
region. Strictly speaking, the assumptions underlying the

logarithmic law of the wall are compromised wherever
boundary layer separation occurs. Therefore, the results might
provide useful hints regarding a possible loss in overall
accuracy for the missile configuration considered here caused
by the much more economic wall-function approach when
compared to turbulence models requiring the resolution of the
laminar sublayer.
TAU Code
The TAU code as used by DLR is a finite volume
Euler/NavierStokes solver that can use structured,
unstructured, and hybrid meshes and has already been applied
to the study of various configurations as reported in Hanneman
and Mack (2002) and Dujardin et al. (2002). The Reynoldsaveraged NavierStokes equations are discretised by a finite
volume technique using tetrahedral, pyramids, prisms, and
hexahedral elements. Prismatic elements are used for the
boundary-layer region while the tetrahedral ones are used in
inviscid flow regions. The internal data storage is based on an
efficient edge-data structure of Galle (1999), using the
NETCDF format. Having a fine mesh y+ < 1, the DLR were
able to resolve the flow accurately in the viscous sublayer.
Beside the central scheme, different upwind solvers such as
Van Leer, AUSM/Van Leer, and AUSMDV are implemented.
Some acceleration procedures such as multi-grid or residual
smoothing are available and have been modified for hypersonic
flows. Also, modifications on the limiter have been done to
achieve a higher accuracy and better convergence. Shock and
leeside treatments especially for Euler computations have been
added. A three-stage RungeKutta scheme is used for time
discretisation including local time-stepping. Due to the
implemented explicit residual smoothing, CFL number up to 4
can be used. To reduce the computational time for twodimensional flows, a special two-dimensional model is
2005 CASI

activated that suppresses the flux calculation in the span wise

direction. Beside the common boundary conditions such as
constant wall temperature and adiabatic wall, radiative
equilibrium is available. Laminar flow as well as turbulent flow
can be considered, as various turbulence models have been
implemented. The DLR had used both the k and the Spalart
Allmaras turbulence models during their studies. Concerning
the turbulence parameters used for the computation one should
be careful in the choice of the t / l coefficient as well as the
turbulence intensity. Indeed the k is relatively freestream
dependant (Menter, 1994), and the influence of these factors
can be important. For this study both factors have been chosen
around 1%. Higher values somehow exaggerate the resulting
viscosity and the turbulent behaviour of the flow (thicker
boundary layer and shorter separation length, high turbulent
energy), but two low values lead usually to numerical problems
as poor convergence and shock resolution.
UMS3DNS Code
NASA Langley had used USM3Dns for their computations
on an unstructured grid. USM3Dns is a tetrahedral cellcentered, finite-volume Euler and NavierStokes solver (Frink,
1998). The inviscid flux quantities are computed across the cell
faces using Roes flux-difference splitting scheme and the
spatial discretization is accomplished by a novel analytical
reconstruction process. The solution is advanced in time to
steady state using an implicit backward-Euler time-stepping
scheme. Flow turbulence is modeled by the SpalartAllmaras
one-equation model, which is optionally coupled with a wall
function to reduce solution stiffness and the number of cells in
the sublayer of the boundary layer. All computations presented
in this paper are performed using the wall-function feature of
USM3Dns in a fully turbulent mode. Using a fine mesh with
wall function may not be justified, and thus viscous solutions
from NASA Langley were not included in meticulous
comparisons in the wall regions.
USM3Dns runs on massively parallel computers and
clusters of personal computers (PCs). A single processor
version is available for vector processors such as the Cray
super-computers (with multi-tasking) and single processor
workstations. The parallel version (Bhat and Parikh, 1999;
Parikh, 2001) is the version of choice because of its rapid
turnaround time for large problems. The code requires
175 eight-bit words of memory per tetrahedron. It runs with a
speed of 34 s/cell/cycle/processor on a Cray C90 and
230 s/cell/cycle/processor on the SGI Origin 2000 parallel
computer.
USM3Dns supports standard boundary conditions such as
the flow tangency, no-slip solid surface, characteriztic
inflow/outflow (for subsonic flows), and freestream
inflow/extrapolationoutflow (for supersonic flows). In
addition, some special boundary conditions including wall
functions, wake flow, jet engine, and propeller are available in
the code, and are reviewed in Frink et al. (2000).
2005 CASI
IMPNS Code
QinetiQ in the UK had used an iterative PNS solver
IMPNS for its computations (Birch et al., 2000). The
governing equations, solved by IMPNS, are the steady-state
compressible NavierStokes equations. The perfect gas law
and Sutherlands law for laminar viscosity are employed to
close the system. When necessary, turbulence is handled using
the Favre mass-averaged form of the NavierStokes equations
with either a BaldwinLomax (1978) algebraic turbulence
model, enhanced to account for crossflow separation (by
modifications due to either DeganiSchiff (1986) or the
curvature method of Qin and Jayatunga (1998)), or the oneequation model of Spalart and Allmaras (1992). For cases
where viscous effects are unimportant, the viscosity may be
switched off and the corresponding Euler equations may be
solved with appropriate boundary conditions.
The governing equations are discretised using a cell-centred
finite-volume scheme in a generalized co-ordinate system for
use with structured, body-conforming grids. To enable a stable
well-defined space-marching solution, the discretised flow
equations are parabolized by neglecting viscous terms in the
streamwise direction. In addition, for single-sweep calculations
the flow outside the boundary layer must be supersonic in the
stream wise direction and a portion of the stream wise pressure
gradient within subsonic regions is assumed negligible. Several
interpretations of the stabilizing approximation by Vigneron et
al. (1978), in which a portion of the stream wise pressure
gradient is neglected, have been investigated and implemented
in the solver.
FLUENT 5 Code
The QintiQ team had also made use of the FLUENT 5
commercial flow solver to compute the test cases to compare
the accuracy and relative efficiency of the PNS space-marching
technique. The FLUENT flow solver is based on the finitevolume method and can utilize structured, unstructured, or
hybrid grids. The coupled solver, which simultaneously
computes both the continuity and momentum equations, was
employed in the present study. FLUENT uses an upwind, fluxdifference splitting and can operate using either implicit or
explicit time-marching schemes. For the present investigation
the spatial accuracy was set to second order, and the Spalart
Allmaras turbulence model was used for comparison with the
SpalartAllmaras results from IMPNS.
STRUCTURED AND UNSTRUCTURED MESHES

The computational domain for the structured grid is shown
in Figure 2. The inlet was located 0.1 in downstream from the
tip of the missile to avoid the grid quality problems associated
with modeling flow in this region. The fact that the actual tip
was not included in the computational domain should not have
a large effect on the results. The outlet was located 2.0 in
downstream from the end of the actual missile shown in
Figure 1; the body of the missile model used in the
157

Figure 3. Side view of the computational structured grid near the

missile body.
grids. The coarse grid is composed of about 700 000 nodes

while the fine grid get about 1 500 000 points. A view of the
fine grid is shown in Figure 4.
Figure 2. The computational domain.
computations was extended downstream from the rear control

fins to the outlet. The azimuth free stream boundary was
located approximately 4.75D from the centreline at the front of
the missile, and approximately 22D from the centreline at the
rear of the missile.
Figure 3 shows the side view of a typical structured mesh
used in the computations. The grid lines were continuous
through all the blocks of the computational domain. The mesh
around the front fins was similar to the mesh around the rear
fins. The density of the grid lines was high near the missile
surfaces where the viscous resolution of the flow had a direct
bearing on the accuracy of the computations. The grid lines
were farther apart in outer regions of the domain where the flow
was not affected as much by viscosity. The average y+ value at
the first grid point above the missile surface was approximately
5. A total of 4.0 million control volumes were used for each
simulation.
A specimen of (coarse and fine) unstructured grids was
supplied by NASA Langley. They are completely unstructured,
even if an anisotropic distribution resulting from the splitting of
prismatic cells into tetrahedral cells is used in the boundary
layer. These shapes are, however, more adapted to viscous
computation and allow better control of the near-wall normal
distribution and particularly the normal first spacing. For this
reason y+ 1 has been kept almost everywhere, a y+ peak
occurring inevitably at some edges, and for both coarse and fine
158
Figure 4. General views of the fine unstructured grid.
RESULTS
As a first test, the aforementioned codes were tested for their
ability to compute ordinary lift force and moment coefficients
for free stream Mach number of M = 1.75, with unit Reynolds
number of Re/ft = 2 106 and two settings of angle of attack,
= 6 and 24. Within various degrees of accuracy almost all
the codes were able to resolve these coefficients to a
satisfactory value. In some cases, see Figure 5, the participants
2005 CASI


were able to produce a complete polar ranging from = 0 to
24. Table 1 provides a detailed study of the exact force and
moment coefficients as computed by various participants.
While it was satisfying to note that at 6, the normal force
seems to be well predicted by all computations the differences
in axial drag results varied from as low as 0.4296, as predicted
by Price and Moule (QINETIQ) to as much as 0.731 by Hennig
(LFK) when compared to an experimental value of 0.5594. It is
plainly obvious that the turbulence models are unable to cope
with complex vortex-induced separated flow that evolves along
the length of the missile. At higher angle of attack 23.98,
the computations seem to match the normal force reasonable
well, but the predictions for axial drag are no better. It was
equally interesting to note from computed results from

DATCOM and MISL3 as published by Lesieutre et al. (2002)
that for lower angles of attack even empirical-based codes are
very economic design tools.
The predicted Mach number contours as obtained from the
WIND Code computations using the SpalartAllmaras
turbulence model by McIlwain and Khalid (2004), on the
vertical (a) and horizontal (b) planes bisecting the missile
centreline, are shown in Figure 6 for = 24. These planes lie
midway between the fins. The bow shock wave is clearly
visible in both plots. The flow between the top two front fins
was accelerated, creating a region of higher Mach number flow
trailing out into the free stream. There were regions of slower
Figure 5. Evolution of the body force and moment coefficients versus the angle of attack for M = 1.75 as obtained using the TAU code.
2005 CASI
159
Table 1. Computed force and moment coefficients.

160
2005 CASI

Figure 6. Mach number contours for the = 24 simulation; (a) z = 0 (vertical) plane and (b) y = 0 (horizontal) plane (McIlwain and Khalid, 2004).
flow underneath the missile, especially beneath the tip and

between the two sets of fins. Between the front and rear fins
there was also a region of low-speed flow near the body on the
top and bottom of the missile, and a region of high-speed flow
near the body along the sides of the missile. This flow remained
close to the body and did not trail into the free stream like the
shock waves extending from the tip and the front fins. Note that
the same flow computed by Pirzadeh and Leavitt (2004) using
the SpalartAllmaras turbulence model on an unstructured grid
shows very similar flow patterns in Figure 7.
Figure 8. Shadowgraphs at 100 ms (Leopold et al., 2004).
Figure 7. Mach contours for tandem controls missile at = 24, M =

1.75 (from Pirzadeh and Leavitt, 2004).
When comparing the computed results against the velocity

field as depicted in the measured shadowgraph (Figure 8)
obtained by Leopold et al. (2003) the flow patterns are in good
agreement with experiment. It is useful to note that the
shadowgraph has picked up the complex shock interaction
features near the forward control surface and the distinct bow
line, which runs from the oblique shock at the forebody
2005 CASI
indicates the interaction between the shock and the outer edge
of the vortex. These features have been well resolved by most
computations.
It is apparent that the flow on top surface and on the sides
evolves into some very distinct vortices. Figures 9 and 10 taken
from Dujardin (2004) and Nding and Hennig (2004),
respectively, show the total pressure contours at a number of
station cuts along the axial length from different computations
using three different turbulence models. While the flow
features are discussed in greater detail in the AVT/RTO Report
(2004), it is apparent from the reflection plane flow field as
depicted in Figure 11 that the vortices start evolving from the
region of the missile. They seem to be energized and
strengthened somewhat from the presence of the forward
control surfaces. Although the computations were performed to
represent a steady-state case, there appears to be a numerical
periodicity in solution showing a vortex pattern on the top
surface, which appears to stretch outwards and give birth to
161

Figure 9. Total pressure contours at = 24 and M = 1.75 obtained with the k model (left) and the SpalartAllmaras model (right).
Figure 10. Total pressure contours at M = 1.75 and = 24 obtained using the k turbulence model. Left: x/D = 6; right: x/D = 10.
further vortices closer to the surface, mimicking the classic

von-Karman vortex behaviour. It must also be recognized that
even though, time-marching iterative schemes may be
employed in some of the CFD codes, the solutions remain
steady in nature and can only be an approximate reflection of
the true vortex flows at high angle of attack. The vortex
growing on the side of the missile becomes increasingly
distinct towards the middle length of the missile and stands
proud from the system developing on the centre upper surface.
The vortices coalesce together near the trailing edge. The
complete vortex pattern appears to maintain its main features
for some distance past the aft trailing surfaces and merges into
the mainstream flow. These distinct vortex patterns will
decidedly impact upon the aerodynamic performance of the
missile at higher angles of attack.
Another aspect of this study addresses the question of
viscous resolution of solvers and the included turbulence
162
models to predict velocity profiles near the missile surfaces.

Profiles of the x and y components of velocity for the = 24
case as obtained from the WIND code are compared to
experimental data in Figures 12 and 13. These profiles are
located at x/D = 5.8 (slightly downstream from the front set of
fins), 10.3 (between the two sets of fins), and 13.8 (in front of
the rear set of fins). There is considerable disagreement
between the experimental and numerical velocity profiles next
to the missile surface. The relatively coarse numerical grid that
was used next to the missile surface (y+ 5) and the simple oneequation turbulence model (SpalartAllmaras) that was used in
the simulations were likely contributing factors towards this
discrepancy. There was better agreement between the two sets
of results farther away from the surface of the missile, at y/D >
1.75, especially for those stations located far downstream from
the front set of fins.
2005 CASI

Figure 11. Distribution of the flow field Mach number (left) and corresponding total pressure(right) (Dujardin, 2004) on several cross plane.
Visualisation of the vortices for M = 1.75, = 24, k model.
Figure 12. Comparison of numerical and experimental (Leopold et al.,

2003) U-velocity profiles for = 24 at x/D = 5.8 (red), 10.3 (green), and
13.8 (blue).
Figure 13. Comparison of numerical and experimental (Leopold et al.,

2003) V-velocity profiles for = 24 at x/D = 5.8 (red), 10.3 (green), and
13.8 (blue).
Figure 14 shows another set of comparisons, which had

used the DLR TAU code (Dujardin, 2004) on an unstructured
grid using two different turbulence models to compute the axial
and normal components of the velocity profiles. From the
velocity profiles shown in Figure 14 and indeed those studies
earlier in Figures 12 and 13, it is obvious that U components of
velocity are better resolved away from the surface where the
viscous effects tend to be weaker. For the V component of
velocity it can be observed that the SpalartAllmaras model
does a far better job as employed in WIND code to predict the

velocity profiles in the near regions at stations x/D = 10.3 and
x/D = 13.8. The corresponding predictions from the TAU code
using k or the SpalartAllmaras model did not do as well at
both stations. The station at x/D = 5.8 is immediately
downstream of the control surfaces and suffers from possible
vortex-induced flow separations and complex interactions with
the shock. The V component of velocity at this station is,
therefore, not as accurately resolved. The fact that
2005 CASI
163

Figure 14. (a) Velocity profiles for the vz component along the z axis at
two positions of the symmetry plane. (b) Velocity profiles for the vx
component along the z axis at two positions of the symmetry plane.
measurements were recorded at M = 2.0 rather than the

computed free-stream input of M = 1.75 must also contribute to
the discrepancy between the predicted and experimental
values.
The authors from QinetiQ, UK (Prince and Birch, 2004)
further investigated the question of velocity profile
comparisons at different Mach numbers. To resolve the issue of
how the ISL experimental results, obtained for Mach 2.0, and
0.43 million per D Reynolds number flow, compares with the
NASA results from a Mach 1.75, Re = 0.63 million per D, flow,
two additional PNS calculations were performed. Solutions
were obtained for = 6 and 24 using the SpalartAllmaras
turbulence model on the finest PNS grid such that the results
164
could be compared with the ISL velocity measurements at the

equivalent conditions. These calculations also allowed for a
comparison of the PNS results at the two conditions to assess
the differences in the flowfields for the two cases.
As expected, the solutions performed by QinetiQ (see
IMPNS solutions Figures 15 and 16) at the ISL wind-tunnel
test conditions best matched the experimental data, and the
differences between the computed profiles for the two different
conditions are seen to be significant. The implications of this
are that the ISL data, while useful for qualitative analysis of the
flow physics, are not good for comparison with CFD for
validation of the various turbulence models employed in this
NATORTO study.
Figure 15 presents a cross plot of earlier axial and normal
velocity component profiles when compared against the results
from the UK. The axial velocity profiles at stations x/D = 5.8,
10.3, and 13.8, as obtained from the DLR TAU code and the
WIND code are compared against the UK IMPNS code and the
measurement in Figure 15. In most cases the comparisons were
carried with the unified use of the SpalartAllmaras model as it
seemed to perform better than the two-equation model.
Individual attempts (AVT/RTO Report, 2004) at using other
two-equation models had shown that the SpalartAllmaras
model had performed better, and thus the cross plots in this
paper compare results from this turbulence model. For the axial
velocity comparisons it is noted that DLR and QinetiQ have
done very well in being able to resolve the flow close to the
wall. The results from the WIND code near the wall region
especially at the aft stations are not as accurate. The normal
components of the velocity ratio v/U, from various codes at the
same axial stations of x/D = 5.8, 10.3, and 13.8 are shown in
Figure 16. The computed and experimental velocity profiles
capture the highly turbulent and vortical nature of the flow. The
normal components of velocity show a large range in v/U
values extending from v/U = 0.3 to 0.4 with at least two
inflexion regions in between.
It is thus convenient to deduce that IMPNS computations
with SpalartAllmaras model and TAU based computations
from DLR with fine grids (y+ ~ 1), provide the best resolution
of viscous flows at high supersonic Mach numbers. The WIND
solutions from IAR/NRC had used a relatively coarser grid y+ =
5, and did not resolve the axial viscous components close to the
wall as accurately. The IMPNS-based solutions are the more
welcome as the parabolised version of the NavierStokes
equations tend to be less stiff and provide faster convergence.
CONCLUSION
Most CFD computations were able to capture the global
large-scale features of the flow; the turbulence models as yet,
are not able to accurately resolve the complexity of the viscous
flow in regions of high shockshock, shockvortex, and other
shockboundary-layer interactions; or other regions of largescale separation at high angle of attack. While the economy of
various computational methods has not been a subject of this
2005 CASI

Figure 15. Comparison of measured and predicted profiles of axial

velocity ratio, u/V, at x/D = 5.8, 10.3, and 13.8.
paper, it was found during various deliberations of this group

that the PNS-based NavierStokes computations are far more
efficient in producing satisfactory results than those based on
the full NavierStokes equations.
There is evidence from the literature at large (Lesieutre et
al., 2002) that for most design purposes it would suffice to
conduct a majority of preliminary design exercises using the
DATCOM-type algorithms and resorting to higher order
computations where the flow becomes increasingly challenging
for the lower order methods.
Of the various turbulence models investigated, it appears
that despite the empiricism of the single equation approach, the
SpalartAllmaras model appears to resolve the physics as well
2005 CASI
Figure 16. Comparison of measured and predicted profiles of normal

velocity ratio, v/V, at x/D = 5.8, 10.3, and 13.8.
as the two-equation k model at higher Mach numbers.

Almost all the turbulence models fared well when computing
velocity profiles away from the region of strong flow
interactions from the control surfaces or the principal shocks
triggered from the forebody and the control surfaces. On the
same token the axial component of velocity seemed to be better
modelled than the normal component.
For this particular missile, the study found the existence of
complex vortex systems that would impact upon the missile
performance at higher angles of attack.
165
ACKNOWLEDGEMENT
Menter, F.R. (1994). Two-Equation Eddy-Viscosity Turbulence Models for

Engineering Applications. AIAA J. Vol. 32, No. 8, pp. 15981605.
The authors are grateful to Peter Tonn and Sandra Cheyne of

NATOs AVT/RTO panel for facilitating meetings and
discussion sessions, which led to the completion of this work.
Nding, P., and Hennig, P. (2004). LFK Results for the Dual Control
Missile. In Assessment of Turbulence Modeling for High Speed Air Vehicles.
AVT/RTO Report.
Parikh, P.C. (2001). Application of a Scalable, Parallel, Unstructured-GridBased Navier-Stokes Solver. AIAA Paper 2001-2584.
REFERENCES

AVT/RTO Report. (2004). Assessment of Turbulence Modeling for High

Speed Air Vehicles.
Baldwin, B., and Lomax, H. (1978). Thin-Layer Approximation and
Algebraic Model for Separated Turbulent Flows. Proceedings of the AIAA
16th Aerospace Sciences Meeting, Huntsville, Alabama, 1618 January 1978.
Bhat, M.K., and Parikh, P.C. (1999). Parallel Implementation of an
Unstructured Grid-Based Navier-Stokes Solver. AIAA Paper 99-0663.
Birch, T.J., Ludlow, D.K., and Qin, N. (2000). Towards an Efficient,
Robust and Accurate Solver for Supersonic Viscous Flows. Proceedings of
the ICAS 2000 Congress, Harrogate, United Kingdom, 27 August
1 September 2000. CD-ROM. International Council of the Aeronautical
Sciences, Reston, Virginia. pp. 242.
Bush, R.H., Power, G.D., and Towne, C.E. (1998). WIND The
Production Flow Solver of the NPARC Alliance. AIAA Paper 98-0935.
Degani, D., and Schiff, L.B. (1986). Computation of Turbulent Supersonic
Flows Around Pointed Bodies Having Crossflow Separation. J. Comput.
Phys. Vol. 66, pp. 173196.
Pirzadeh, S., and Leavitt, L. (2004). Performance Studies on a Dual

Control Missile Using Unstructured Grids. In Assessment of Turbulence
Modeling for High Speed Air Vehicles. AVT/RTO Report.
Prince, S., and Birch, T. (2004). Application and Validation of the
Parabolized Navier-Stokes Method for the Investigation of the NASA Dual
Control Missile Configuration in Supersonic Flow. In Assessment of
Turbulence Modeling for High Speed Air Vehicles. AVT/RTO Report.
Qin, N., and Jayatunga, C. (1998). Algebraic Turbulence Modeling for
Vortical Flows Around Slender Bodies. In Missile Aerodynamics. NATO
RTO-MP-5, Paper 20.
Spalart, P.R., and Allmaras, S.R. (1992). A One-Equation Turbulence
Model for Aerodynamic Flows. AIAA Paper 92-0439.
Van Leer, B. (1982). Flux Vector Splitting for the Euler Equations. In 8th
International Conference on Numerical Methods in Fluid Dynamics. Edited by
E. Krause. Lect. Notes Phys. Vol. 264, pp. 667683.
Vigneron, Y.C., Rakich, J.V., and Tannehill, J.C. (1978). Calculation of
Supersonic Viscous Flows over Delta Wings with Sharp Subsonic Leading
Edges. AIAA Paper 78-1137.
Dujardin, A. (2004). Flow Computation on a Dual Control Missile Using

the DLR Unstructured TAU Solver. In Assessment of Turbulence Modeling
for High Speed Air Vehicles. AVT/RTO Rep.
Dujardin, A., Glhan, A., Longo, J.M.A., and Mack, A. (2002).
Numerical/Experimental Investigation of a Wedge-Compression Corner with
Gap-Flow under Cold-Hypersonic Gas Condition. Proceedings of the 4th
European Symposium on Aerothermodynamics for Space Applications, Capua,
Italy, pp. 717723.
Frink, N.T. (1998). Tetrahedral Unstructured Navier-Stokes Method for
Turbulent Flows. AIAA J. Vol. 36, No. 11, pp. 19751982.
Frink, N.T., Pirzadeh, S.Z., Parikh, P.C., and Pandya, M.J. (2000). The
NASA Tetrahedral Unstructured Software System (TetrUSS). Aeronaut.
J. Vol. 104, No. 1040, pp. 491499.
Galle, M. (1999). Ein Verfahren zur Numerischen Simulation
Kompressibler, Reibungsbehafteter Strmungen auf Hybriden Netzen.
Deutsches Zentrum fr Luft- und Raumfahrt, Cologne, Germany. Report
DLR-FB-1999-04.
Hanneman, V., and Mack, A. (2002). Validation of the Unstructured DLR
TAU-Code for Hypersonic Flows. AIAA Paper 20023111.
Launder, B.E., and Spalding, D.B. (1974). The Numerical Computation of
Turbulent Flows. Comput. Methods Appl. Mech. Eng. Vol. 3, pp. 269289.
Lesieutre, D., Love, J., Dillenius, M., and Blair, A.B., Jr. (2002). Recent
Applications and Improvements to the Engineering Level Aerodynamic
Prediction Software MISL3. Proceedings of the 40th AIAA Aerospace
Sciences Meeting and Exhibit, Reno, Nevada, 1417 January 2002. American
Institute of Aeronautics and Astronautics, Reston, Virginia. AIAA Paper
2002-0275.
Leopold, F., Demeautis, C., and Faderl, N. (2003). Experimental
Investigations of the RTO Missile Configuration at High Angles of Attack.
ISL PU 657/2003.
McIlwain, S., and Khalid, M. (2004). WIND CFD-Based Performance
Study of a Dual Control Missile. In Assessment of Turbulence Modeling for
High Speed Air Vehicles. AVT/RTO Report.
166
2005 CASI
Aerodynamic Forces Approximations using

the Chebyshev Method for Closed-Loop Aeroservoelasticity Studies
Alin Dorian Dinu * Ruxandra Mihaela Botez * and Iulian Cotoi *
NOMENCLATURE

Abstract
The approximation of unsteady generalized aerodynamic
forces from the frequency domain into the Laplace domain
acting on a Fly-By-Wire aircraft presents an important
challenge in the aero-servoelasticity area. The aerodynamic
forces in the reduced-frequency domain are approximated
in the Laplace domain, to be able to study the effects of the
control laws on the flexible aircraft structure. In this paper,
we present a new method for the approximation of the
generalized aerodynamic forces by use of Chebyshev
polynomials and their orthogonality properties. A
comparison of this new method with the Pad method used
to calculate an approximation of the generalized
aerodynamic forces from the frequency domain into the
Laplace domain is presented. This comparison shows that
this new method gives excellent results with respect to the
Pad method and is applied on the Aircraft Test Model
from NASA Dryden Flight Research Center.
Rsum
Lapproximation de forces arodynamiques gnralises
non-stationnaires du domaine de la frquence dans le
domaine du Laplace, forces qui actionnent sur un avion a
commandes lectriques, reprsente une importante
challenge pour le domaine de laroservolasticit. Les
forces arodynamiques du domaine de la frquence rduite
sont approximes dans le domaine du Laplace pour tudier
les effets des lois de contrle sur la structure flexible de
lavion. Dans cet article nous prsentons une nouvelle
mthode pour lapproximation de forces arodynamiques
modal damping matrix
wing chord length
modal elastic stiffness matrix
reduced frequency
modal inertia or mass matrix
Mach number
Chebyshev polynomial
modal generalized aerodynamic forces matrix

imaginary part of the modal generalized
aerodynamic forces matrix
QI
QR
real part of the modal generalized aerodynamic

forces matrix
non-dimensional generalized
respect to time t)
dynamic pressure
qdyn
V
(with
true airspeed
VE
equivalent airspeed
Vp
matrix of eigenvectors
V0
reference true airspeed

generalized coordinates
coordinates
vector of eigenvalues
airspeed ratio
true air density
reference air density
air density ratio
modal transformation matrix
natural frequency
INTRODUCTION
* Department of Automated Production Engineering

cole de technologie suprieure
1100 Notre Dame West
Montreal, QC H3C 1K3, Canada.
E-mail: ruxandra@gpa.etsmtl.ca
Received 6 May 2005.
2005 CASI
The aero-servoelasticity represents the combination of

several theories regarding different aspects of aircraft
dynamics. Studies of aero-servoelastic interactions on an
aircraft are very complex problems to solve, but essential for
aircraft certification. The instabilities issued by the adverse
interactions among the flexible structure, the aerodynamic
forces, and the control laws acting on it could appear at any
time inside the flight envelope, so we can state that the aero167


gnralises en utilizant les polynmes de Chebyshev et
leurs proprits dorthogonalit. Une comparaison de cette
nouvelle mthode avec la mthode de Pad qui est
habituellement utilize pour calculer lapproximation de
forces arodynamiques gnralises du domaine de la
frquence dans le domaine du Laplace est aussi prsente.
Cette comparaison nous montre que cette nouvelle
mthode donne des rsultats excellents par rapport la
mthode Pad en lappliquant sur lAircraft Test Model
(lATM) de NASA DFRC (Dryden Flight Research
Center).
servoelastic interactions concern mainly the research field

located at the intersection of the following three disciplines:
aerodynamics, aeroelasticity, and servo-controls. One main
aspect of the aero-servoelasticity is the conversion of the
unsteady generalized aerodynamic forces Q(k, Mach) from the
frequency domain into the Laplace domain Q(s), where k
represents the reduced frequency, Mach is the Mach number,
and s is the Laplace variable. There are mainly three classical
methods for approximating the unsteady generalized forces by
rational functions from the frequency domain to the Laplace
domain (Tiffany and Adams, 1984, 1988; Edwards, 1977;
Roger, 1977; Karpel, 1982): Least Square (LS), Matrix Pad
(MP), and Minimum State (MS). All three methods use rational
functions under Pad form.
Several aero-servoelastic analysis software codes are
developed for the aerospace industry. One of the computer
programs used for aero-servoelasticity analyses is the Analog
and Digital Aeroservoelasticity Method (ADAM) that was
developed at The Flight Dynamics Laboratory (Noll et al.,
1986). ISAC (The Interaction of Structures, Aerodynamics, and
Controls) was developed at NASA Langley Research Center
(Adamsand Hoadley, 1993). At the Boeing Company, the aeroservoelastic software FAMUSS was used (Pitt, 1992). An
aeroelastic code, ZAERO, has been developed at Zona
Technology, which has been used for aero-servoelastic studies
(Chen and Sulaeman, 2003). The STARS code was developed
at NASA Dryden Flight Research Center (Gupta, 1997).
Among all these computer codes, we have chosen to work
with the STARS code. The STARS program is an efficient tool
for aero-servoelastic interactions studies and has an interface
with NASTRAN (Newsom et al., 1984; Rodden et al., 1979) a
computer program frequently used in the aeronautical industry.
In this paper, the lateral dynamics of a half aircraft test model
ATM is modeled in STARS. Following finite element modeling
and the doublet lattice method application on the ATM in
STARS, the unsteady aerodynamic forces are calculated as
functions of reduced frequencies k and Mach number M. We
have provided here a bibliographical research on the software
used in aero-servoelasticity. All these codes exploit two main
classical methods for aerodynamic-force approximations from
168
the frequency domain (aeroelasticity) into the Laplace domain

(aeroservoelasticity): Least Squares (LS) and Minimum State
(MS).
We will now present in detail bibliographical research on
other existing methods in the literature.
The approximation of the unsteady generalized
aerodynamic forces is a must for the control analysis of our
system. Due to the fact that Q(k, Mach) can only be tabulated
for a finite set of reduced frequencies, at a fixed Mach number
M, it must be interpolated in the s domain to obtain Q(s). In this
paper, we describe such an interpolation method that uses the
Chebyshev polynomials and its results. In the subsonic regime,
the unsteady generalized aerodynamic forces Q(k, Mach) are
calculated, using finite elements computer programs such as
STARS or NASTRAN, by the Doublet Lattice Method (DLM).
We further need to convert these forces into the Laplace domain
where they will be denoted as Q(s). The aerodynamic forces
dependence of s may be written as an irrational function even
for simple cases such as two-dimensional potential
incompressible flows on an airplane wing profile. Theodorsen
(1933) proved that Q(s) could be expressed by use of Hankels
functions. A few years later, Wagner found the first rational
approximation (Dowell, 1995) for Q(s). Another approach used
the approximations of unsteady aerodynamic forces by Pad
polynomials. This approach was based on a fractional
approximation of the form P(s)/R(s), where P and R are two
polynomials in s, for every term of the unsteady force matrix. In
this way, every pole of R(s) showed a new state called the
augmented state, in the final linear invariant aero-servoelastic
system. Thus, if the initial square matrix had N dimensions, and
if a Pad approximation of M order is used, then there will be
introduced N(N + M) augmented states. The number of
augmented states was reduced by Roger (1977). In his
formulation, only N M modes were introduced, where N is the
number of initial modes. Rogers method is based on the fact
that the aerodynamic lag terms remain the same for each
element of the unsteady aerodynamic force matrix. Thisis
called the LS method and is used in computer aero-servoelastic
codes such as STARS and ADAM.
Another method was derived from the LS method and was
proposed by Vepa (1977). This method used the same
denominators for every column of the aerodynamic matrix Q,
and was called MP method.
Various improvements were made to the two methods
presented above. One such type of improvement is that
different conditions (restrictions) were imposed on
approximations to pass through certain points. Generally, the
approximations were restricted to be exact approximations at
zero and at two other chosen points. Generally, the first point
was chosen to represent the estimated flutter frequency and the
second point to represent the gust frequency. Then, the
improved methods were renamed: ELS method (Extended
Least Squares) (Tiffany and Adams, 1984) and EMMP method
(Extended Modified Matrix Pad) (Dunn, 1980). Later, Karpel
(1998) proposed a completely different approach to solve the
above approximations. He knew from the beginning that the
2005 CASI

goal was to find a linear invariant system in the time domain

and he decided to integrate this information directly in the
equation giving the unsteady aerodynamic forces values, by
adding a term resembling a transfer function of a linear system.
Further, as he wanted to find a linear system of reasonable
dimensions, he wrote the approximation under the MS form.
The advantage of this method with respect to Rogers method
resided in the fact that it allowed an excellent approximation to
be obtained, but with a smaller number of augmented states. All
of the methods described above allow for the approximation of
unsteady aerodynamic forces for one Mach number at a time.
To obtain approximations for several Mach numbers, we
should perform the approximation approach given by the MS
method for each Mach number, which might be expensive in
terms of computing time. A valid approximation for a range of
Mach numbers could be useful for military Fly-By-Wire
aircraft, where the Mach number varies rapidly during highspeed manoeuvres, and where aero-servoelastic interactions are
extremely important. Poirion (1995, 1996) constructed an
approximation allowing for the calculation of unsteady
aerodynamic forces for Mach number values contained in a
specific interval and for a frequency domain. He used the MS
method and considered a regular dependence with the Mach
number. He used several MS approximations, obtained for
several fixed Mach numbers, and a spline interpolation method
for Mach number dependence. Thus, he obtained a formula that
allows for the computing of unsteady aerodynamic forces for
any couple (k, Mach), where k is the reduced frequency, the
equations remaining valid for a range of (k, Mach) [kmin,
kmax]*[Mmin, Mmax].
The approximation methods should satisfy two opposed
criteria simultaneously: an excellent (exact) approximation,
which could be obtained by increasing the number of lag terms
and a linear invariant system in the time domain of a very small
dimension (with the smallest possible number of lag terms).
There was no method satisfying both criteria until now. In two
recent papers, Cotoi and Botez (2001, 2002) have proposed a
new approach based again on a precise Pad approximation.
The two authors used order-reduction methods for the last term
of the approximation, which could be seen as a transfer
function of a linear system. The approximation error for this
new method is 1240 times lower than for the MS method for
the same number of augmented states and depends of the
choice made for the model reduction method. However, this
method remains very expensive in terms of computing time.
Dinu et al. (2005) presented open-loop flutter analysis results
using a new method based on Chebyshev polynomials theories
of the ATM (Aircraft Test Model). Control laws were not
considered in their previous paper (where open-loop flutter
analysis was performed), while in the present paper, this new
method is applied to the ATM on which ailerons and rudder
act so that closed-loop flutter analysis results are
determined.
Aircraft Equations of Motion

The flexible aircraft equations of motion, where no external
forces are included, may be written in the time-domain as
follows:
~ && ~ & ~
M
+ C + K + q dyn Q (k, Mach) = 0
(1)
where is the air density, V is the true airspeed, and qdyn =

0.5V2 is the dynamic pressure; is the generalized coordinates
variable defined as q = where q is the displacement vector
and is the matrix containing the eigenvectors of the following
free-vibration problem:
Mq&& + Kq = 0
(2)
The following transformations were used in Equation (1):

~
M = TM,
~
C = TC ,
~
K = TK
Q(k, Mach) = T Ae (k)
(3)
Here, M, K, and C are the generalized mass, stiffness, and

damping matrices; k, the reduced frequency is written as k =
b/V where is the natural frequency and b is the wing semichord length. Ae(k) is the aerodynamic influence coefficient
matrix for a given fixed Mach number M and a set of reduced
frequencies values k. The Laplace transformation is further
applied to Equation (1), and we obtain:
~
~
~
[M s 2 + C s + K] (s) + q dyn Q (s) (s) = 0
(4)
Q(s) are the unsteady aerodynamic force approximations of

Q(k, Mach) in the Laplace domain. In this paper, we describe a
new approximation method that uses Chebyshev polynomials
and its results.
Chebyshev Polynomial Theory

These polynomials (Rivlin, 1990; Weisstein, 19992005)
are a set of orthogonal polynomials defined as the solutions to
the Chebyshev differential Equation (10) and are denoted as
Tn(x). They are used as an approximation to a least squares fit,
and are closely connected to trigonometric multiple-angle
equations. The Chebyshev polynomials of the first kind
denoted by Tn(x) are implemented in Mathematica as
ChebyshevT [n, x], and are normalized so that Tn(1) = 1.
Any continuous function may be expressed by use of
Chebyshev polynomials using the following equation:
f (x) =
1
c0 + c j T j (x)
2
j =1
(5)
where Chebyshev polynomials used in Equation (5) are

expressed under the following form:
2005 CASI
169

T j (x) = cos( j arccos(x))
(6)
and the coefficients cj used in Equation (5) are expressed as

follows:
2 f (x)T j (x)
dx
1 1 x 2
and
j = 1, 2, K
each element of the unsteady aerodynamic force matrix, we

found a power series development under the following form, by
use of Maples chebyshev function:
Q ij (s) =
cj =
(7)
1 ( ij )
c0 + cn( ij ) Tn( ij ) (s)
2
n =1
(12)

where
The Chebyshev polynomials have orthogonality properties that
allow us to keep the approximations error within a
predetermined bandwidth.
The following recurrence relationships are used in the new
approximation method:
T0 (x) = 1
T1 (x) = x
T (x) = 2 x T (x) T (x)
r
r 1
r +1
(9)
(2 j + 1)
2r
(10)
Tr(x) is a function defined by cosines, which lets us conclude

that between two solutions of this function we find an extreme
of |1| amplitude in the middle of the interval, specifically at:
x = cos
j
,
r
j = 0,1, ..., r
(11)
Methodology for the Chebyshev

Approximation Method
To develop our approximation method, we used the
predefined functions for the Chebyshev polynomials expressed
in Equation (6) that have been implemented in the Maple
kernel, in MATLAB.
These functions (chebpade and chebyshev) allow the
construction of a polynomial interpolation for the unsteady
generalized aerodynamic forces, acting on the ATM for 14
values of reduced frequencies k and 9 values of Mach number.
The elements forming the matrices of the unsteady generalized
aerodynamic forces calculated by the Doublet Lattice Method
DLM in STARS are denoted by Q(i, j) with i = 18 and j =
18 for the first eight elastic modes.
The approximation by means of this method is obtained by
use of a similar path to the one used for the Pad method. For
170
for
n = 0,1, K
We found an approximation by rational fractions by use of the

chebpade function:
M
$ ij (s) =
Q
where r specifies the rank of the Chebyshev polynomial.

Equation (9) gives the following solution:
x = cos
( ij )
2 Q ij (s)Tn (s)
ds
1
(1 s 2)
(8)
and the following condition in the aim to find the Chebyshev

polynomials solution is imposed:
Tr (x) = 0
cn( ij ) =
a n(ij ) Tn(ij )(s)
n= 0
P
1+
(13)
bn( ij ) Tn( ij ) (s)
n =1
where M = P + 2.
This new form integrates the orthogonality properties of
Chebyshev polynomials and allows the variation of the degree
of the numerator M and the denominator P, to obtain a very
good approximation.
In Equation (13), an approximation order [M, P] = [16, 14]
gives M = 16 and P = 14 where M is the maximum rank of
Chebyshev polynomials at the numerator and P is the
maximum rank of Chebyshev polynomials at the denominator.
We compared the results found by means of our Chebyshev
approximation method with the results given by the Pad
method. These results are expressed in terms of a total
normalized approximation error.
The Pad method uses a parameter identification solution to
determine a polynomial fractional form that identifies an
orthogonal polynomial interpolation. This fractional form is the
key aspect of this method, due to the fact that it allows the order
reduction system.
The Pad polynomials are used in the LS method of
implementation, which is considered the most classical and
most used method now for aero-servoelastic interactions
studies. The LS method was implemented in most aeroservoelasticity software ISAC, ADAM, and STARS.
For various aircraft types (such as CL-604 or F/A-18),
classical aero-servoelasticity studies by use of the LS or MS
methods were performed. Following an analysis of the results
obtained and the algorithms used, we found that the LS method
was easier to implement and execution time was faster than that
of the MS method. Our computer programs were written in
MATLAB. For this reason, in this paper, a comparison is
performed between flutter results obtained with the Pad
method and flutter results obtained with our method based on
Chebyshev polynomials properties.
2005 CASI

Figures 1 and 2 show the real and imaginary parts of

aerodynamic forces elements approximated by Pad and
Chebyshev methods versus their initial values calculated in
STARS in the frequency domain. Figures 3 and 4 show that our
new approximation method gives the best approximation error
on an interval chosen in the proximity of each approximation
point. The effect of the Chebyshev polynomials properties is
seen in these types of results. Due to these properties, we are
able to impose a bandwidth for the error convergence for each
element of the unsteady generalized aerodynamic force
matrices.
Figure 2. The imaginary part of the aerodynamic forces calculated by

Pad and Chebyshev methods versus their initial values in the
frequency domain.
Figure 1. The real part of the aerodynamic forces calculated by Pad

and Chebyshev methods versus their initial values in the frequency
domain.
Both figures show the overall normalized approximation

error by Chebyshev and Pad methods. The differences
between Figures 3 and 4 are based on the model order:
Figure 3 shows the results for the [16, 14] polynomial model
order while Figure 4 shows the results obtained for the [15, 13]
polynomial model order. In these examples, the results were
obtained using the ATM data generated in the STARS code at
Mach number M = 0.5 and for 14 reduced frequencies k =
[0.0100 0.1000 0.2000 0.3030 0.4000 0.5000 0.5882 0.6250
0.6667 0.7143 0.7692 0.8333 0.9091 1.0000].
The Pad method gives a small error near the middle of the
approximation interval and an increased error towards each end
of it. The Chebyshev approximation method demonstrates an
almost constant value of the error all along the approximation
interval. The total normalized approximation errors differences
at both ends of the approximation interval are noticed in both
figures. Those differences are higher for k14 = 1 than for k1 =
0.01.
2005 CASI
Figure 3. The total normalized approximation errors for the [16, 14]
model order.
The total normalized approximation errors were calculated

for different values of the polynomial approximation order
using the Pad and the Chebyshev polynomial fraction methods
(polynomial model order should be equivalent for both
methods) for all the Mach numbers and the same differences
were observed. The total normalized approximation error
obtained with the Chebyshev method was found to be much
171
Table 1. Total normalized approximation errors by Chebyshev and

Pad methods.
Order
Method
JQ_REAL
JQ_IMAG
[16, 14]
Chebyshev
Pad
Chebyshev
Pad
Chebyshev
Pad
0.065065
0.065271
0.055391
0.260147
0.061466
0.136338
0.037611
0.287657
0.040053
0.775989
0.035109
0.055638
[15, 13]

[10, 8]
Chebyshev or by Pad theories. Nmodes is the total number of

modes (also the dimension of Q), k is the index of the reduced
frequency, and J is the total normalized error.
Open-Loop Flutter Results Obtained

using the Chebyshev Method
Figure 4. The total normalized approximation errors for the [15, 13]
model order.
smaller with respect to the total normalized approximation

error given by the Pad method.
In Table 1, a few examples regarding the values of the total
normalized approximation errors provided by the Chebyshev
method and by the Pad method are given. In Table 1, three
different orders of approximation were used: [16, 14], [15, 13],
[10, 8] for the same fight condition and the results are presented
for the real aerodynamic forces part errors denoted by JQ_REAL
and for the imaginary aerodynamic forces part errors denoted
by JQ_IMAG. It can be clearly seen that no matter the order of the
polynomial approximation, the total normalized error for the
Chebyshev method is lower than the total normalized error
given by Pad method. These errors were calculated using the
following formula:
N modes N modes Q ij Rnew Q ij Rold

=
2
k =1
Q ij
i =1 j =1
14
JQ _ REAL
JQ _ IMAG
14 N modes N modes Q
ij Inew Q ij Iold
=
2
k =1
Q ij
i =1 j =1
100%(14)

100%

To validate our method, we used the STARS ATM developed

by the NASA Dryden Flight Research Center. This lateral
model (only anti-symmetric modes are provided) includes
aero-structural elements (flexible aircraft) and control surfaces
(ailerons and elevator). First, a free vibration analysis was
performed in the absence of aerodynamics to obtain the free
modes of vibration. We obtained the same frequencies and
modes of vibration using our MATLAB method as with
STARS.
Then to calculate aerodynamic forces in the frequency
domain by the DLM, the same simulation parameters were
considered as the ones considered in the STARS computer
program: reference semi-chord length b = 38.89 in (1 in =
2.54 cm), reference air density at sea level 0 = 1.225 kg/m3,
altitude at sea level Z = 0 ft (1 ft = 0.3048 m), reference sound
airspeed at sea level a0 = 340.294 m/s.
In Table 2, the speeds and frequencies at which flutter
occurs are calculated by two methods, Pad and Chebyshev, for
three different types of approximation orders. The execution
speed is three times smaller for the Chebyshev polynomials
method than the execution speed used for the Pad method.
Table 2. Flutter results comparison for ATM in open loop.
Flutter results (fuselage first bending mode)

Method
Speed
(knots)
Frequency
(rad/s)
Computation
time (s)
Pk-Pad [8, 6]
Pk-Pad [9, 7]
Pk-Pad [10, 8]
Pk-Chebyshev [8, 6]
Pk-Chebyshev [9, 7]
Pk-Chebyshev [10, 8]
445.5
445.5
445.8
446.5
446.6
446.6
77.5
77.5
77.5
77.5
77.5
77.5
122
134
144
40
47
53
where QRold and QIold are the real and the imaginary parts of the
unsteady aerodynamic forces given by STARS for the ATM
model and QRnew and QInew are the real and the imaginary parts
of the unsteady aerodynamic forces approximated by
172
2005 CASI

Closed-Loop Flutter Results Obtained

using the Chebyshev Method
We applied our new approximation method for the closedloop aero-servoelastic analysis, using the transfer function
information provided by the aircrafts control laws (ailerons
and elevator are used for lateral aircraft control). The
conversion from the frequency domain into the Laplace domain
(as closed-loop calculations should be realized in the Laplace
domain) was done this time by use of the LS method for the
approximations obtained with Chebyshev polynomials. To
achieve this type of conversion, we used the following variable
change:
s = jk k =
s
= js
j
(15)
where j is the complex number j = 1.

We rewrote the approximation of the unsteady generalized
force matrix in the Laplace domain as follows:
Q(s) = A0 + A1( j s) + A2 ( j s) 2 + A3
Figure 5. Frequency versus damping calculated by the P-Chebyshev

method for the closed loop ATM analysis.
js
j s + b1
+ A4
js
+L
j s + b2
(16)
and we took into account that Q(s) has a real part QR(s) and an
imaginary part QI(s), such as:
Q(s) = Q R (s) + j Q I (s)
(17)
We obtained then the following expressions for these

aerodynamic forces:
s2
A
Q R (s) = A0 + s 2 A2 + 2
2 n+ 2
n =1 s + bn
sbn
Q (s) = A s
A
I
1
2
2 n+ 2
n =1 s + bn
(18)
where A0, A1,, An+2 are the LS decomposition matrices and bn

are the lag terms introduced by the LS method.
Figures 5, 6, and 7 show the flutter closed-loop results. In
Figure 5, the frequency is plotted versus the damping, in
Figure 6, the frequency is plotted versus the Equivalent
Airspeed and in Figure 7, and damping is plotted versus the
Equivalent Airspeed. From these three figures, the flutter
frequency, damping, and equivalent airspeed are calculated
(flutter occurs where damping is zero). Their numerical values
are given in Table 3.
In Table 3, a comparison of the results obtained with our
new method (a combination of the P flutter method with the
Chebyshev approximations) with the results obtained by a
2005 CASI
Figure 6. Frequency versus equivalent airspeed calculated using the PChebyshev method for the closed loop ATM analysis.
combination of the P flutter method with the Pad

approximations is provided for 2 to 6 lag terms.
Our new approximation method provides the same values
for the flutter speeds and frequencies as the classical methods
(which takes Pad approximations into account, because these
methods are actually based on Pad) no matter the number of
lag terms considered. These good results were expected to
occur, following the properties of the Chebyshev polynomials.
In addition, the execution time required by the P-Chebyshev
method is three times smaller than the execution time required
by the P-Pad method.
It is well known that a larger number of lags imply on one
hand an increased number of calculations and on the other hand
it introduces new poles, thus modifying the initial system and
173
most important, is the fact that the computation time for the
open-loop case is up to 3 times smaller than in the PkPad
method and up to 30 times smaller than in our PkLS case, even
for an increased approximation order.
As for the closed-loop case, we observed that no matter the
initial Chebyshev approximation order, it would be enough to
make use of only 2 lags when converting to the Laplace domain
to obtain excellent approximation results.

ACKNOWLEDGEMENTS
Figure 7. Damping versus equivalent airspeed calculated using the PChebyshev method for the closed loop ATM analysis.
Table 3. Flutter results comparison for ATM in closed loop.
Flutter results (control mode 2)

Method
Speed
(knots)
Frequency
(rad/s)
Computation
time (s)
P-Pad [4, 2]
P-Pad [5, 3]
P-Pad [8, 6]
P-Chebyshev [4, 2]
P-Chebyshev [5, 3]
P-Chebyshev [8, 6]
287.7
287.6
287.6
287.6
287.6
286.7
51.15
51.13
51.10
50.92
50.92
50.72
354
361
392
97
102
126
generating approximation errors. For this reason, the advantage

of using this new method with respect to classical methods is
that using a small number of lags (2 lag terms) the same results
are obtained as if using a large number of lags (6 lag terms).
CONCLUSION
The Chebyshev approximation method provides excellent
flutter results for a small number of lag terms. However, due to
the fact that the Chebyshev polynomials were generated by use
of the ATM data, there are quite large differences between the
values of the elements contained in the unsteady generalized
aerodynamic force matrices (1e +10). Restraints regarding the
threshold of the approximation error had to be imposed, i.e., for
smaller elements we imposed an error value of 1e 4 and for
larger elements an error value of 1e 2. Without these
restraints, the Chebyshev polynomials cannot be generated.
We could see that by use of the Chebyshev method in open
loop, we were able to find very good values for the flutter
speeds and the frequencies at which flutter occurs. One of the
most important achievements of our new method, if not the
174
The authors would like to thank Dr. Gupta from NASA

Dryden Flight Research Center for allowing us to use the ATM
on STARS. We would also like to thank the other members of
the STARS engineering group: Mr. Tim Doyle, Mr. Marty
Brenner, and Dr. Shun Lung for their precious assistance and
collaboration.
This work was made possible due to funds received from the
Natural Sciences and Engineering Research Council of Canada
(NSERC), and from the Ministre du Dveloppement
conomique, de linnovation et de lexportation (MDEIE).
References
Adams, W.M., Jr., and Hoadley, S.T. (1993). ISAC: A Tool for
Aeroservoelastic Modeling and Analysis. Collection of Technical Papers
AIAA/ASME Structures, Structural Dynamics and Materials Conference, La
Jolla, California, 1922 April 1993. Vol. 2. American Institute of Aeronautics
and Astronautics, Washington, D.C. pp. 10101018.
Chen, P.C., and Sulaeman, E. (2003). Nonlinear Response of
Aeroservoelastic Systems Using Discrete State-Space Approach. AIAA J.
Vol. 41, No. 9, pp. 16581666.
Cotoi, I., and Botez, R.M. (2001). Optimization of Unsteady Aerodynamic
Forces for Aeroservoelastic Analysis. Proceedings of the IASTED
International Conference on Control and Applications CA2001, Banff,
Alberta, 2729 June 2001. Edited by M.H. Hamza. IASTED/ACTA Press,
Anaheim, California. pp. 105108.
Cotoi, I., and Botez, R.M. (2002). Method of Unsteady Aerodynamic
Forces Approximation for Aeroservoelastic Interactions. J. Guid. Control
Dyn. Vol. 25, No. 5, pp. 985987.
Dinu, A., Botez, R.M., and Cotoi, I. (2006). Chebyshev Polynomials for
Unsteady Aerodynamic Calculations in Aeroservoelasticity. J. Aircr. Vol. 43,
No. 1, pp. 165171.
Dowell, E.H. (1995). A Modern Course in Aeroelasticity, Kluwer
Academic, Dordrecht, The Netherlands.
Dunn, H.J. (1980). An Analytical Technique for Approximating Unsteady
Aerodynamics in the Time Domain. NASA Tech. Pap. TP-1738.
Edwards, J.W. (1977). Unsteady Aerodynamic Modeling and Active
Aeroelastic Control. Department of Aeronautics and Astronautics, Stanford
University, Stanford, California. SUDAAR Rep. 504, February.
Gupta, K.K. (1997). STARS An Integrated, Multidisciplinary, FiniteElement, Structural, Fluids, Aeroelastic, and Aeroservoelastic Analysis
Computer Program. NASA Tech. Memo. TM-4795, pp. 1285.
Karpel, M. (1982). Design for Flutter Suppression and Gust Alleviation
Using State-Space Modeling. J. Aircr. Vol. 19, No. 3, pp. 221227.
Karpel, M. (1998). Reduced-Order Models for Integrated Aeroservoelastic
Optimization. American Institute of Aeronautics and Astronautics, Inc.
2005 CASI
Newsom, J.R., Adams, W.M., Jr., Mukhopadhyay, V., Tiffany, S.H., and
Abel, I. (1984). Active Controls: A Look at Analytical Methods and
Associated Tools. In Proceedings of the 14th ICAS Congress, ICAS 84-4.2.3,
Toulouse, France, 914 September 1984. Edited by B. Laschka and R.
Staufenbiel. International Council of the Aeronautical Sciences and American
Institute of Aeronautics and Astronautics, New York, N.Y. pp. 230242.
Noll, T., Blair, M., and Cerra, J. (1986). ADAM, An Aeroservoelastic
Analysis Method for Analog or Digital Systems. J. Aircr. Vol. 23, No. 11,
pp. 852858.

Pitt, D.M. (1992). FAMUSS: A New Aeroservoelastic Modeling Tool.

AIAA Pap. 92-2395-CP.
Poirion, F. (1995). Modlisation Temporelle des Systmes
Aroservolastiques. Application ltude des Effets des Retards. La
Recherche Arospatiale, No. 2, pp. 103114.
Poirion, F. (1996). Multi-Mach Rational Approximation to Generalized
Aerodynamic Forces. J. Aircr. Vol. 33, No. 6, pp. 11991201.
Rivlin, Th.J. (1990). Chebyshev Polynomials: From Approximation
Theory to Algebra and Number Theory. 2nd ed. Wiley, New York, N.Y.
Rodden, W.P., Harder, R.L., and Bellinger, E.D. (1979). Aeroelastic
Addition to NASTRAN. NASA Rep. CR-3094.
Roger, K.L. (1977). Airplane Math Modeling Methods for Active Control
Design. In Conference Proceedings on Structural Aspects of Active Controls.
AGARD, Neuilly-sur-Seine, France. CP-228, pp. 4.14.11.
Theodorsen, T. (1933). Interference on an Airplane of Finite Span in an
Open Rectangular Wind Tunnel. NASA Tech. Rep. TR 461.
Tiffany, S.H., and Adams, W.M., Jr. (1984). Fitting Aerodynamic Forces in
the Laplace Domain: An Application of a Nonlinear Nongradient Technique to
Multilevel Constrained Optimization. NASA Tech. Memo. TM 86317.
Tiffany, S.H., and Adams, W.M. (1988). Nonlinear Programming
Extensions to Rational Function Approximation of Unsteady Aerodynamics.
NASA Tech. Pap. TP-2776, July.
Vepa, R. (1977). Finite State Modeling of Aeroelastic System. NASA
Rep. CR-2779, February.
Weisstein, E.W. (19992005). Chebyshev Polynomial of the First Kind.
MathWorld A Wolfram Web Resource [On-line]. Available from: [retrieved
March 2004].
2005 CASI
175


Applying Fast Fourier Transform Analysis and

Data Window in Software Global Positioning
System Receivers to Mitigate Continuous
Wave Interference under Dynamic Conditions
Z. Jiang * ** G. Lachapelle * and C. Ma *
INTRODUCTION
Abstract
Traditionally, most Radio Frequency Interference
mitigation methods have been implemented and tested
using conventional hardware receivers. With the rapid
development of computer technologies, the signalprocessing computational load is becoming less of a
concern, and thus it becomes feasible to develop and test
new interference mitigation methods based on software
receivers together with modern digital signal-processing
techniques. This paper describes an investigation into the
development of a narrow-band Continuous Wave (CW)
interference mitigation algorithm and performance testing
using a software Global Positioning System receiver based
on Fast Fourier Transform analysis and data windowing.
In this paper, some strategies were proposed to improve
the anti-jamming performance. First, for high-level
interference, a fixed detection threshold suggested in
previous literature is not optimal. An adaptive detection
threshold that is of better performance and is a function of
the standard deviation of the normalized spectrum and the
correlator power output was used in lieu. Second, a priori
information-based soft thresholding techniques for both bit
synchronization and enhanced signal acquisition were
applied under high dynamic conditions and high-level
interference environments. The factors that are crucial for
weak-signal detection and tracking, namely, coherent
* Position, Location and Navigation (PLAN) Group

Department of Geomatics Engineering
The University of Calgary
Calgary, AB T2N 1N4, Canada.
** Present address: NovAtel Inc.
1120 68th Avenue NE,
Calgary, AB T2E 8S5, Canada.
E-mail: zjiang@novatel.ca
The Global Positioning System (GPS) uses a direct sequence

spread spectrum (DS-SS) signal that incorporates some degree
of anti-jamming capability in the signal structure itself.
However, GPS signals, normally in the range of 160 to
156 dBW for the C/A-code, are weak and well below the
background RF noise level and thus make it easy for an
interference signal to overcome the inherent anti-jamming
capability of the DS-SS signal owing to the fact that the
processing gain is not high enough to compensate the
interference.
For narrow-band interference, which usually arises from
spurious signals generated by inadequately shielded electrical
equipment or from narrow-band radio links adjacent to GPS
frequencies, a so-called frequency-domain interference
excision algorithm based on Fast Fourier Transform (FFT)
analysis has been proved to be a good approach to mitigate such
susceptibility. Several techniques based on this concept have
been proposed, including adaptive transversal filters (ATF)
(Przyjemski et al., 1993), FFTs, and filter banks (FB) (Rifkin
and Vaccaro, 2000).
Generally speaking, the objective of this FFT-based narrowband interference mitigation algorithm is to reduce the level of
the interference at the expense of introducing some degree of
distortion on the desired signal before signal correlation is
conducted. The estimation and the suppression of interference
are in fact performed in the frequency domain. So, three
operations are obviously necessary herein. The first one
consists of transforming the time domain signal into the
frequency domain by fixed-length Discrete Fourier
Transform/Fast Fourier Transform (DFT/FFT), the second one
is the actual interference identification and suppression, and the
third one consists of transforming signal from the frequency
domain back to time domain for conventional GPS signal
processing.
The power distribution of CW interference suggests that
only a small number of frequency domain cells contain nearly
all of the interference power within the band (Dipietro, 1989).
Based on this conclusion, one possible strategy could be to set
the weights on all cells containing large interference to zero,
Received 9 March 2005.

2005 CASI
177

suite de la page 177s

integration time, tracking loop bandwidth, and integration
time in the loop filter were evaluated to assess the
effectiveness of this algorithm. Some interference
suppression strategies for spread spectrum systems,
namely, windowing and overlap processing, were also
investigated. Test results proved that the proposed
algorithm is effective to mitigate narrow-band CW
interference under certain power level. Windowing and
overlapped processing are good strategies to further
improve the anti-interfernce capability by 2 dB.
Rsum
Traditionnellement, la plupart des mthodes de rduction
du brouillage RF ont t mises en uvre et testes au
moyen de rcepteurs matriels conventionnels. Grce
lvolution rapide des technologies de linformatique, la
charge de traitement du signal devient de moins en moins
problmatique, et par consquent, il est devenu ralisable
de mettre au point et de tester de nouvelles mthodes de
rduction du brouillage bases sur des rcepteurs logiciels,
allis des techniques modernes de traitement du signal
numrique. Cet article dcrit une tude qui vise la mise au
point dun algorithme de rduction du brouillage associ
un signal continu bande troite ainsi que de tests de
performances au moyen dun rcepteur GPS logiciel bas
sur lanalyse FFT et le fentrage des donnes.
Dans cet article, nous proposons un certain nombre de
stratgies qui visent amliorer le rendement des
mthodes antibrouillage. Tout dabord, en ce qui concerne
le brouillage de haut niveau, le seuil de dtection fixe
suggr dans la littrature antrieure nest pas optimal.
Nous avons utilis la place un seuil de dtection adaptatif
qui offre un meilleur rendement et est fonction de lcart
type du spectre normalis et de la puissance de sortie du
corrlateur. Ensuite, nous avons appliqu, dans des
conditions hautement dynamiques fort niveau de
brouillage, des techniques de seuillage logiciel bases sur
des informations a priori pour la synchronisation des bits et
lacquisition dun signal amlior. Afin dvaluer
lefficacit de cet algorithme, nous avons valu les
facteurs qui sont essentiels dans la dtection et le suivi
dun signal faible, savoir un temps dintgration
cohrent, la bande passante de la boucle de poursuite et le
temps dintgration dans la boucle de filtrage. Nous avons
galement tudi un certain nombre de stratgies de
suppression du brouillage dans les systmes talement
du spectre, savoir le fentrage et le traitement simultan.
Les rsultats des essais effectus dmontrent que
lalgorithme propos se rvle efficace pour rduire le
brouillage associ un signal continu bande troite, pour
certains niveaux de puissance. Le fentrage et le traitement
simultan constituent de bonnes stratgies pour amliorer
de 2 dB supplmentaires les capacits antibrouillage.
178
while leaving the others at unity. The computational challenge

in this case lies in determining which cells constitute large
interferences. One straightforward way is establishing a
threshold, and then comparing it to signal components in all
cells; when the magnitude of a signal component exceeds the
threshold, interference is assumed and the component is
removed by setting its corresponding weighting function to
zero. The threshold can be established on the basis of
knowledge of the interference distribution or on the basis of
heuristic experience, such as setting the threshold to excise a
fixed percentage of cells or total interference power. Another
strategy for the weighting function derivation could be to set
any cell value exceeding the threshold to the background noise
level, and thus whiten the interference spectrum. This approach
may yield improved performance, because the cell value
containing interference will be reduced only to the background
level, hence retaining most of the signal power. The drawback
of this approach, however, is the requirement of the
background noise level estimation, thus further increasing the
computational burden.
Applying FFT analysis and data windowing to narrow-band
CW interference mitigation is actually not a brand new
technique. Since the early 1980s, quite a lot of research has
been conducted, e.g., Li and Milstein (1982), Dipietro (1989),
Young and Lehnert (1994), Wang and Amin (1998). However,
these previous investigations have focused mainly on spread
spectrum communications systems, and only in recent years
has such a frequency domain analysis-based interference
mitigation method been applied to GPS by some researchers,
namely, Peterson et al. (1996), Badke and Spanias (2002), and
Cutright et al. (2003). The studies are far from completed and
many implementation issues, such as the determination of the
detection threshold for different kinds of interference and
different interference levels, the impact of narrow-band
interference, the effectiveness of the mitigation algorithm in a
software receiver for signal acquisition, tracking and position
fixing, have not been fully addressed. Thus, to enhance the
application of the frequency excision algorithm in GPS, further
research into this algorithm, aided by the flexibility that a
software GPS receiver provides, becomes necessary.
THEORETICAL BACKGROUND
The received baseband data stream has three components,
namely, the signal samples sk, the broadband noise samples k ,
and the narrow-band interference sample k . Each component
is assumed to be uncorrelated with the other two and
characterized by a zero mean. The individual component
correlations are (Dipietro, 1989)
[ ]
0
E sk s *m =
S
k m
k =m
(1)
2005 CASI
[ ]
0
E k *m = 2
k m
k =m
[ ]

(3)
where E[] is the expectation operator and * denotes the

complex conjugate. Obviously, the signal and noise samples are
each uncorrelated, while the narrow-band interferences are
correlated. Based on this characteristic, the spectral peaks of
the interference can,therefore, be discriminated and suppressed
from the GPS signal and the thermal noise (Gaussian
distribution) through an adaptive threshold power level and an
adaptive notch filter.
The first stage of the FFT-based interference mitigation
algorithm is interference detection The correlation power,
which indicates the average post-correlation signal-to-noise
ratio (SNR), is used in this paper to detect the existence of
interference. It is computed as (Ndili and Enge, 1997):
I 2 + Q2
Expected Noise Floor
(4)
where I and Q are the prompt 1 ms in-phase and quadrature

correlation signals. The mean and the variance of the SNRpc are
functions of noise and interference in the signal, and, therefore,
are candidates for interference detection.
The change in the SNR before and after interference
mitigation is often used to show the performance improvement
since the objective of the algorithm is to identify and mitigate
narrow band interference based on its statistical properties. The
SNR of the output of a correlator is given by (Dipietro, 1989)
SNR =
( E[skH X$ ]) 2
Var (skH X$ )
(5)
where Var () is the variance operator and H is the conjugate

transpose. Replacing the corresponding components in
Equation (6) with the assumed statistical properties of the
signal components and the characteristics of the DFT
transformation matrix, the SNR of the correlator output
becomes
2
N S
k =1 k
SNR =
d1 + d 2 + d3
(6)
with
N 2
N
k =1 k
2
d1 = S k =1 k
2005 CASI
d 2 = 2 k =1 k2
N
(2)
d3 =
E k*m rm k
SNRpc =
(7)
1
N
k =1 (k k)2
N
(8)
(9)
where d1 is a self noise term that is analogous to that seen in

linear prediction suppression processors (Ketchum and
Proakis, 1982), a term that vanishes in the case of no filtering,
d2 is the residual broadband noise, d3 is the narrow-band
interference power, and k is the kth FFT bin component of the
narrow-band interference signal.
If the value of k is set to unity, the pre-suppression SNR is
SNR =
NS
N
1
2 + k =1 (k ) 2
N
(10)
where N is the FFT length for suppression processing and S is

equal to E[sk s*k ].
As mentioned before, the ratio of Equations (6) to (10)
yields the improvement factor used to assess interference
mitigation performance.
The influence of CW interference in the frequency domain is
not a single line because interference spreads out through the
entire spectrum due to the finite FFT analysis period. This
phenomenon is called spectral leakage. Thus, simply removing
one frequency component with the largest power spectrum line
is not sufficient, and a suitable detection method is needed to
identify which frequency component contains strong
interference and should be removed. The judging criterion is
based on the statistical analysis of the input signal.
Traditionally, the standard deviation of the resulting
normalized spectrum is multiplied by a fixed value to set a
detection threshold for determining the presence of Radio
Frequency Interference (RFI) (Cutright et al., 2003), and the
estimate of this fixed value is determined empirically.
The traditional FFT-based interference mitigation algorithm
can remove most of the interference energy in the frequency
domain. But it has the drawback that it can suppress energy
only in the main-lobe spectral bins, which is not adequate due
to the residual interference distributed throughout the spectrum
by these side-lobes, especially when the interference power
level is high.
FFT operations assume the signal processed to be a periodic
extension of a finite sequence. For each block of Np samples of
the input sequence that is used to do the FFT analysis, a
discontinuity will occur at the FFT block boundary if the Np
samples are not periodic in the FFT window of observation.
Therefore, the FFT of a signal often exhibits high spectral sidelobes due to the finite processing interval. The behaviour of
these side-lobes is very similar to a broadband interference
source in terms of contributing energy to all FFT spectral bins.
As a result, no threshold can be selected that will allow the
removal of the interference energy without also removing
179

excessive signal spectrum. This phenomenon also constitutes

spectral leakage.
Spectral leakage that survives the frequency domain filtering
operation may severely degrade the performance of the
mitigation algorithm. To mitigate the effect of spectral leakage,
a data window is often applied prior to FFT computation.
Windowing smoothes the discontinuities at the block boundary
and therefore lessens the effect of spectral leakage. The
windowed FFT can be expressed as:
X (k) =
N p 1
w(n) x(n) e
j 2 kn
Np
(11)
n= 0
k = 0, 1, ..., N p 1
The window function w(n) determines the amount of spectral
leakage in the DFT output. If w(n) is set to identity, namely,
w(n) = 1,
n = 0, 1, ..., N p 1
(12)
then Equation (12) reverts to non-windowed processing, which

is the same as the traditional FFT output. In this situation, it is
equivalent to using a rectangular window. The Fourier
transform of the rectangular window is a Sinc function with the
first side-lobe reduced by 13 dB relative to the main lobe.
When the frequency of a signal is not exactly one of the DFT
frequencies, the signal energy will be spread across the
spectrum proportional to the width of the main lobe and the
height of the side-lobes of the window (Capozza et al., 2000).
In the traditional FFT, which is equivalent to a rectangle
windowed FFT, the first side-lobe attenuation is only 13 dB.
The interference frequency is much easier to mix with the
signal frequency, causing degradation of the mitigation
performance. So selecting a window with lower side-lobes will
reduce the amount of spectral leakage. However, a lower sidelobe usually results in a wider main lobe, causing reduced
spectral resolution.
The objective of using a data window in an interference
mitigation algorithm is to minimize the spectrum spreading of
the interference frequency component to minimize the number
of frequency bins that need to be excised. At the same time,
when no interference is presented, the windowing operation
should minimize the SNR loss of the GPS signal. Thus, window
selection requires a trade-off between the reduction in SNR due
to the signal attenuation caused by introducing window
operation and the effectiveness of the spectral containment for
interference.
Overlapped processing techniques can be used to minimize
the degradation of the SNR due to data windowing while
maintaining frequency content. The overlapped processing
allows the tails of the window to be eliminated, thereby
substantially reducing the output SNR loss. With 50%
overlapped processing, the SNR loss can be reduced to as low
as 0.6 dB (Capozza et al., 1999).
180
PROPOSED METHOD
The fixed interference detection threshold as proposed in the
previous literature is not optimal in some case, especially when
the interference is strong. In this paper, an adaptive interference
detection threshold determination method that is a function of
the standard deviation of the normalized spectrum and the postcorrelation SNR is used since the latter is a good indicator of
the interference level. All of the results reported herein are
derived from this adaptive detection threshold. After the
detection threshold is determined, the normalized spectrum is
then compared against the threshold and bins exceeding the
detection level are identified. The bins containing RFI, along
with a variable number of surrounding bins, are then set to zero
in the original frequency domain spectrum. The effect of
removing the frequency bins is equal to applying band-pass
filters in the time domain. The Inverse Fast Fourier Transform
(IFFT) of this spectrum is finally taken to yield a new time
domain signal without RFI or with mitigated RFI. To obtain the
optimal anti-jamming performance, three parameters have to be
carefully chosen, namely:
the average interval to remove the bias
the detection threshold, and
the number of samples to be removed near the bin containing
the RFI.
In GPS applications, the interference levels are normally
well above the 13 dB side-lobes of the rectangular window.
The spectrum herein exhibits significant spectral leakage. The
windowing operation reduces the amount of the spectrum that
must be excised, thus preserving more of the desired signal
spectrum. In this paper, a four-term BlackmanHarris window
with a 92 dB side-lobe was carefully chosen. The equation for
computing the coefficients of this window is as follows (Harris,
1978):
k
w[k + 1] = 1 2 cos 2
n 1
k
k
+ 3 cos 4
4 6
n 1
n 1
(13)
where 0 k (n 1), 1 = 0.35875, 2 = 0.48829, and 4 =

0.01168
The performance of frequency containment for the interference
component is excellent with this window because of the very
low side-lobe that essentially restricts the spectral leakage.
However, high signal attenuation on the FFT block transitions
will be introduced as the same time that may result in a 3 dB
SNR loss. But this relatively high SNR loss can be
compensated for by an overlapped processing technique.
To reduce the bit synchronization error when using a FFTbased interference mitigation algorithm under high-dynamic
conditions and a high-level interference environment, some
changes to the tracking loop have been made, namely, soft
2005 CASI
thresholding in bit synchronization and improved acquisition

based on earlier information (details can be found in the results
and analysis section).

TEST SET-UP
To obtain repeatable and controllable GPS signals with
narrow-band CW interference, a hardware GPS simulator and a
hardware arbitrary signal generator were used to generate the
GPS signals and specified interference. These two signals were
then combined in an interference combiner unit. The output
was fed to a GPS front-end, namely, a Signal Tap, which downconverted the RF signals to intermediate frequency (IF) signals
and sampled them; the resulting data were used in a software
receiver where a mitigation algorithm was used to assess the
performance.
Two synchronous 12-channel L1-only hardware signal
simulation units (Spirent GSS STR 6560) were used. They can
reproduce the signal propagation environment of a receiver
installed on a dynamic platform, simulating the effects of highdynamic host vehicle motion, navigation satellite motion, and
ionospheric and tropospheric effects. An Agilent ESG E4431B
signal generator generated the narrow-band CW interference
simulated in the test. The GPS and interference signals were
combined in a GSS 4766 interference combiner unit that
facilitates the use of commercial-off-the-shelf (COTS) signal
generators as fully integrated interference sources with
hardware simulators, such as the Spirent GSS 6560. The COTS
signal generators were controlled through an IEEE-488compliant (GPIB) bus, via Spirent SimGEN for Windows,
hosted on the control PC. The interference signal is defined
along with all the other scenario parameters from within
SimGENs normal user GUI environment. The RF outputs are
combined with the satellite signal generators (SSG) in the GSS
4766 interference combiner unit (ICU). The system hardware
configuration is shown in Figure 1.
A hardware front-end, GPS Signal Tap, made by Accord
Software & Systems Private Limited, was used to downconvert, quantize, and log GPS signals for base-band signal
processing with the software GPS receiver. The software
receiver used herein was that described by Ma et al. (2004).
Owing to the limited capacity of the on-board RAM of the
Signal Tap, only 80 s of data could be collected per test, which
is nevertheless sufficient to test the above procedures because
in the worst case, only 60 s of data are needed to obtain the
ephemeris.
From a receiver design point of view, to maximize tolerance
to dynamic stress, the pre-detection integration time should be
short and the carrier-loop filter bandwidth should be wide;
however, to receive a weak signal or to combat interference, the
pre-detection integration time should be long and the carrierloop filter-noise bandwidth should be narrow. To evaluate the
performance of this algorithm under extreme conditions, the
test was designed to apply interference and dynamic stress to
the receiver at the same time.
2005 CASI
Figure 1. Test set up.
Two-dimensional constant speed circular trajectories were

simulated. The circle radius was set to 1000 m and a constant
velocity of 80 m/s was used to simulate high dynamics. The
interference power was changed from a J/S ratio of 18 dB to
28 dB and performance was compared. To ensure that the
results were stochastically repeatable, all the tests were
conducted six times.
RESULTS AND ANALYSIS

1. Traditional FFT-Based Mitigation Algorithm
To investigate the impact of integration time on the loop
filters in dynamics and under interference conditions, a
constant vehicle velocity of 80 m/s with a J/S of 21 dB were
used. Integration times of 1, 2, 4, and 6 ms were tested, and the
results are shown in Table 1.
As shown in Table 1, the use of a 6 ms integration time
could successfully mitigate the influence of interference and
yielded a reasonable solution. But only four satellites could be
tracked, constituting a low availability and poor reliability. This
confirmed that longer integration time results in lower satellite
availability in high-dynamic conditions due to the longer
response time in the tracking loop. When the integration time
was reduced to 2 ms or 4 ms, the dynamic capability was
improved and six satellites could be tracked. But a large
position error resulted in this case due to the impact of
interference. If the integration time further reduces to 1 ms, the
satellite availability goes back to 4 (under interference
conditions, the signal is weak and longer integration time is
actually needed). So for GPS receiver design, the compromise
must be made in choosing the integration time to get a balance
181

Table 1. Impact of integration time on loop filter.
Integration time (ms)
Horizontal RMS error (m)

Vertical RMS error (m)
Tracked PRN
1 066 097
222 133
1, 14, 16, 25
216 309
50 234
1, 14, 16, 20, 25, 30
278 293
888 585
11, 14, 20, 25, 30
15
6
14, 16, 25, 30
between dynamic stress and interference mitigation ability. In

the situation of 2 and 4 ms, if the problem of the large position
error can be solved, both high accuracy and an acceptable level
of availability can be achieved. The tracking of one or more
satellites that might have sizeable range errors caused the large
position error in the test. Figure 2 shows the pseudorange
errors of different satellites using a 2 ms integration time.
Figure 3. Bit synchronization result for PRN 20.
bit transition point was assumed to be at cell 17 which was 2

bins away from the true value of cell 15.
To improve bit synchronization performance, the following
two methods can be used:
Figure 2. Pseudorange errors compared with true value.
In the test, an integration time of 2 ms allowed the tracking

of six satellites. However, the coordinates could not converge to
their true values after a LS adjustment. One large pseudorange
error existed and thus caused divergence of the estimates. This
error could however be removed using a fault detection and
exclusion algorithm as sufficient measurement redundancy was
available in this case.
From Figure 2, it is obvious that PRN 20 was abnormal. The
pseudorange measurement is off from its true value by about
600 km, which represents a 2 ms transmission time error. This
error is an integer millisecond error caused by the bit
synchronization error. A bit synchronization error will not
affect signal tracking, especially when the integration time is
1 ms. That is to say, this error will continue to persist if the
signal is not re-acquired again. To eliminate this error, further
investigation into the bit synchronization process is needed.
The histogram method was used in the software receiver for
bit synchronization determination. In Figure 3, the histogram
of PRN20 is shown. It can be seen that only cell 17 exceeded
threshold 2, and no other cell reached threshold 1. Thus, the bit
synchronization process was declared successful and the data
182
Use a soft threshold for threshold 2. This threshold is

associated with the noise floor. If the noise level is high, the
threshold will be lower. There will be a high possibility of
two cells exceeding threshold 2 before one cell exceeds
threshold 1. The bit synchronization will restart again instead
of a bit being synchronized at an incorrect point.
If 2 cells exceed threshold 2, go back to acquisition instead of
starting bit synchronization again, because the required
acquisition C/N0 is about 6 dB higher than tracking. The bit
synchronization error is most probably caused by an
inaccurate acquisition result. If acquisition is re-started, the
acquisition detection threshold can be adjusted through the
newly calculated noise floor. The acquisition result will be
more precise and there is a high possibility that the bit
synchronization result will be true.
Figure 4 shows the bit synchronization result with
combination of the above two methods. The synchronized bit
transition point is at the correct place. Figure 5 shows the
pseudorange measurement errors when using the improved bit
synchronization procedure. After least-squares adjustment, the
errors of all the satellites converge to their zero mean.
Two scenarios, warm start and cold start, were defined and
used herein. The definition of warm start is to start acquisition
in a clean environment, and interference is applied after the
2005 CASI

Figure 4. Histogram of bit synchronization using the improved

procedure.
Figure 5. Pseudorange errors using the improved bit synchronization

procedure.
convergence of signal tracking. In contrast, cold start means

acquisition is started after the interference is applied. The
comparison of the position errors between warm-start and coldstart scenarios in kinematic mode with FFT-based interference
mitigation and improved bit synchronization are shown in
Figure 6.
It can be seen that, when the mitigation algorithm is applied,
the RMS position error increases. This is to be expected
because the mitigation algorithm also removes part of the
signal energy when the interference frequency component is
removed. It in turn causes a distortion of the correlation peaks
and larger pseudorange errors on all satellites and finally
increases the position error.
In a warm start environment, no improvement in the position
domain was observed when the mitigation algorithm was
applied. On the contrary, the RMS horizontal error worsened.
2005 CASI
This is because when the interference frequency component

was excised, some of the true signal components were removed
as well, since true signals always mix with interference to a
certain degree. If the interference power level is not high, the
inherent anti-jamming property of the spread spectrum system
can mitigate the interference without any extra processing. But
with mitigation, the resulting position solution may be slightly
degraded due to the signal loss upon frequency excision.
Under cold-start conditions, acquisition was performed in an
interference environment. Acquisition was more difficult than
tracking since the signal strength required in acquisition is 6 dB
higher than that required in tracking. Thus, under cold-start
conditions, the use of the mitigation algorithm is essential. It
can be seen in Figure 6 that the maximum tolerance in cold
start was 18 dB without interference mitigation while it
increased to 22 dB with interference mitigation, a 4 dB
improvement in this case.
To investigate the stability of the algorithm, six independent
tests were carried out under the same conditions. Figure 7
shows the rate of success of these tests in obtaining position
solutions under different interference levels when the
mitigation algorithm was applied. If no position could be
calculated, or if the position error exceeded 50 m (meaning that
the error is well above the normal value and the result is not
reliable), the algorithm declared a failure.
Under cold-start conditions, when the J/S was no greater
than 21 dB, the success rate in obtaining a position solution was
100%. If the J/S increased to 22 dB or 23 dB, the success rate
was reduced to 33.3% or 16.7%, respectively. If the J/S
continued to increase, no solution could be obtained. The
mitigation algorithm is statistically stable when the J/S is no
greater than 21 dB in a cold start condition.
Under warm-start conditions, when the J/S was no greater
than 26 dB, a 100% success rate in obtaining a solution could
be achieved. If the J/S increased to 27 dB, the success rate was
reduced to 33.3%. If the J/S continued to increase, it was not
possible to obtain a solution. The mitigation algorithm was
statistically stable when the J/S was no greater than 26 dB in
this case.
2. Implementation of Data Windowing Before FFT
Figure 8 presents a zoomed section of the spectra obtained
with the 4 ms FFT processing of the non-windowed and
windowed data, respectively. It is clear that the windowing
operation increased the concentration of the interference
frequency and reduced the number of frequency bins that need
to be excised. Windowing reduced the amount of spectrum that
must be excised and thus preserved more of the desired signal
spectrum. As a result, less signal energy loss and less distortion
in the correlation peak improved the accuracy of the position
estimation and increased the maximum tolerance of the
interference mitigation algorithm. The plots in Figure 9 show
the effect of windowing on position estimation.
From the position domain results shown in Figure 9, one
can say that the RMS position error decreases slightly when
using the windowing operation. The RMS vertical error
183

Figure 6. Comparison of mitigation results with CW interference.
Figure 7. Rate of successful position fixing.
decreased from 4.9 to 3.7 m and the RMS horizontal error

decreases slightly from 5.6 to 5.4 m. The number of satellites
that can be tracked increases from six to seven, which means
that employing a windowing operation allowed all of the
simulated satellites to be tracked.
To check the performance stability, six independent tests
were conducted under each J/S ratio condition, and the results
are shown in Figure 10. When the J/S ratio was below 23 dB,
the position error was within 10 m in all of the six tests. When
the J/S ratio increased to 23 dB, the position error also
increased. The position error varied between different tests due
to the random noise. In some tests, the RMS error reached
20 m; however, the success rate in terms of obtaining
184
Figure 8. Enlarged spectra of windowed and non-windowed data.
reasonable position solutions was still 100% with the window

operation. Thus, the interference mitigation algorithm of the
data window, combined with frequency excision, was
statistically stable when the J/S ratio was no greater than 23 dB.
2005 CASI

Figure 9. Effect of windowing on GPS position estimation.
Figure 10. Stochastic repeatability test results with data windowing.
When the J/S ratio increased to 24 and 25 dB, even with a

windowing operation, the success rate reduced to 16.7%, which
means that only one out of six tests could achieve a solution.
Under this circumstance, the mitigation result was not
stochastically repeatable. The maximum tolerance of this
mitigation algorithm with a data window is 23 dB under coldstart conditions. If the J/S ratio is greater than 25 dB, a position
cannot be obtained.
From the above analysis, the advantage of using a data
window before FFT transform is apparent. Although the
improvement in position accuracy is slight, the receiver
sensitivity is improved, since more satellites can be effectively
tracked. Using of a data window together with the frequency
excision algorithm can successfully mitigate interference of J/S
2005 CASI
ratios values no greater than 23 dB under cold-start conditions.

The maximum tolerance increases by 2 dB, as compared to a
traditional FFT based mitigation algorithm, thus improving the
sensitivity.
CONCLUSIONS
The FFT-based algorithm is effective to mitigate narrowband CW interference with a certain power level. An adaptive
detection threshold that is a function of the standard deviation
of the normalized spectrum and the correlator power output has
a better interference mitigation performance than that of the
fixed detection threshold, as was suggested previously in the
literature.
185

In high-dynamic mode (velocity of 80 m/s), by using a soft

threshold for bit synchronization, an enhanced acquisition
method, and a careful selection of the integration time in the
loop filter, the maximum mitigation capability for position
fixing can reach 21 dB in a cold-start environment. Windowing
and overlapped processing before FFT transform have been
shown to be a good strategy to improve the performance of the
FFT-based interference mitigation algorithm. Although the
improvement in position accuracy is slight, more satellites can
be effectively tracked. Data windowing along with FFT
analysis can successfully mitigate interference with a J/S ratio
up to 23 dB under cold-start conditions, a 2 dB improvement
compared with traditional FFT based algorithms.
To achieve optimal interference mitigation performance,
some key receiver parameters should be adjusted as well, such
as the adaptive selection of the coherent integration time, the
tracking loop bandwidth and the integration time in the loop
filter for different interference levels and receiver dynamics.
These parameters were chosen manually in this paper and the
adaptive selection of these parameters according to different
dynamic stress and interference power level is recommended
for future works.
REFERENCES
Badke, B., and Spanias, A.S. (2002). Partial Band Interference Excision
for GPS Using Frequency-Domain Exponents. Proceedings of the 2002 IEEE
International Conference on Acoustics, Speech, and Signal Processing,
(ICASSP 02), Orlando, Florida, 1317 May 2002. Vol. 4. Institute of
Electrical and Electronics Engineers, Piscataway, New Jersey. pp. 39363939.
Capozza, P.T., Holland, B.J., Li, C., Moulin, D., Pacheco, P., and Rifkin, R.
(1999). Measured Effects of a Narrowband Interference Suppressor on GPS
Receivers. Proceedings of the 55th Annual Meeting of the Institute of
Navigation on Navigational Technology for the 21st Century, 2830 June
1999, Cambridge, Massachusetts. CD-ROM. The Institute of Navigation,
Alexandria, Virginia. pp. 645651.
Capozza, P.T., Holland, B.J., Hopkinson, T.M., and Landrau, R.L. (2000).
A Single-Chip Narrow-Band Frequency-Domain Excisor for a Global
Positioning System (GPS) Receiver. IEEE J. Solid-State Circuits, Vol. 35,
No. 3, March, pp. 401411.
Li, L., and Milstein, L. (1982). Rejection of Narrow-Band Interference in

PN Spread-Spectrum Systems Using Transversal Filters. IEEE Trans.
Commun. Vol. COM-30, No. 5, May, pp. 925928.
Ma, C., Lachapelle, G., and Cannon, M.E. (2004). Implementation of a
Software GPS Receiver. Proceedings of GNSS 2004, Session A3, Long
Beach, California, 2124 September 2004. The Institute of Navigation,
Fairfax, Virginia.
Ndili, X. and Enge, P. (1997). Receiver Autonomous Interference
Detection. In Proceedings of 53rd Annual Meeting of the Institute of
Navigation on the Future of Navigation: Facing the Challenges, 30 June 2
July 1997, Albuquerque, New Mexico. The Institute of Navigation,
Alexandria, Virginia.
Peterson, B., Hartnett, J., Fiedler, R., and Nebrich, A. (1996). Frequency
Domain Techniques for Fast GPS Acquisition and Interference
Detection/Rejection. J. Inst. Navig. Vol. 43, No. 3.
Proakis, J.G. (1996). Interference Suppression in Spread Spectrum
Systems. Proceedings of the IEEE 4th International Symposium on Spread
Spectrum Techniques and Applications, Mainz, Germany, 2225 September
1996. Vol. 1. Institute of Electrical and Electronics Engineers (IEEE),
Piscataway, New Jersey. pp. 259266
Przyjemski, J., Balboni, E., and Dowdle, J. (1993). GPS Anti-Jam
Enhancement Techniques. Proceedings of the 49th ION Annual Meeting on
Future Global Navigation and Guidance, Cambridge, Massachusetts, 2123
June 1993. Institute of Navigation, Alexandria, Virginia. pp. 4150.
Rash, G.D. (1997). GPS Jamming in A Laboratory Environment.
Proceedings of the 10th International Technical Meeting of the Satellite
Division of the Institute of Navigation, ION GPS-97, 1619 September 1997,
Kansas City, Missouri. The Institute of Navigation, Alexandria, Virginia.
pp. 389398.
Rifkin, R., and Vaccaro, J.J. (2000). Comparison of Narrowband Adaptive
Filter Technologies for GPS. IEEE 2000 Position Location and Navigation
Symposium, San Diego, California, 1316 March 2000. Institute of Electrical
and Electronics Engineers (IEEE), Piscataway, New Jersey. pp. 125131.
Wang, C., and Amin, M.G. (1998). Performance Analysis of Instantaneous
Frequency-Based Interference Excision Techniques in Spread Spectrum
Communications. IEEE Trans. Signal Process. Vol. 46, No. 1, January,
pp. 7082.
Young, J.A., and Lehnert, J.S. (1994). Sensitivity Loss of Real-Time DFTBased Frequency Excision with Direct Sequence Spread-Spectrum
Communication. Proceedings of the 1994 Tactical Communications
Conference on Digital Technology for the Tactical Communicator, 1012 May
1994, Fort Wayne, Indiana. Vol. 1. Institute of Electrical and Electronics
Engineers (IEEE), Piscataway, New Jersey. pp. 409420.
Cutright, C., Burns, J.R., and Braasch, M. (2003). Characterization of

Narrow-Band Interference Mitigation Performance versus Quantization Error
in Software Radios. Proceedings of the ION 59th Annual Meeting, 2325
June 2003, Albuquerque, New Mexico. CD-ROM. The Institute of Navigation,
Fairfax, Virginia.
Dipietro, R.C. (1989). An FFT Based Technique for Suppressing NarrowBand Interference in PN Spread Spectrum Communications Systems.
Proceedings of the 1989 IEEE International Conference on Acoustics, Speech,
and Signal Processing (ICASSP-89), Glasgow, Scotland, 2326 May 1989.
Vol. 2. Institute of Electrical and Electronics Engineers, New York, New York.
pp. 13601363.
Harris, F.J. (1978). On the Use of Windows for Harmonic Analysis with the
Discrete Fourier Transform. Proc. IEEE, Vol. 66, No. 1, January, pp. 5183.
Ketchum, J.W., and Proakis, J.G. (1982). Adaptive Algorithms for
Estimating and Suppressing Narrow-Band Interference in PN SpreadSpectrum Systems. IEEE Trans. Commun. Vol. COM-30, No. 5, May,
pp. 913924.
186
2005 CASI
Adjoint-Based Sonic Boom Reduction for

Wing-Body Configurations in Supersonic
Flow
Siva K. Nadarajah * Antony Jameson ** Juan Alonso **
NOMENCLATURE

Abstract
flux Jacobian matrix
B
C
boundary
chord length
D
D
domain
discrete adjoint artificial dissipation flux
artificial dissipation flux
internal energy
F
F
design variable
FF
far field
Euler flux vector
Rsum
gradient
Dans cet article, nous prsentons une mthode adjointe de

calcul des sensibilits distance dans le flot supersonique.
Lobjectif est la mise au point dune srie dquations
adjointes et de leurs conditions aux limites
correspondantes, afin de quantifier linfluence des
modifications de la gomtrie sur la distribution de la
pression une position arbitraire dans le domaine dintrt,
smoothed gradient
numerical flux across cell interface
cost function
outward normal
pressure
flux velocity
residual
S
S
shape function
arc length
temperature
time
scaled contravariant velocity component
velocity (physical domain)
state vector
coordinates (physical domain)
altitude
angle of attack
adjustable constant for artificial dissipation scheme
numerical spectral radius of the flux Jacobian matrix
spectral radius of the flux Jacobian matrix
coordinates (computational domain)
This paper presents an adjoint method for the calculation of

remote sensitivities in supersonic flow. The goal is to
develop a set of adjoint equations and their corresponding
boundary conditions to quantify the influence of geometry
modifications on the pressure distribution at an arbitrary
location within the domain of interest, away from the
surface of the aircraft. First, this paper presents the
formulation and discretization of the adjoint equations. The
special treatment of the adjoint boundary condition to
obtain remote sensitivities is also discussed. Second, we
present results that demonstrate the application of the
theory to a three-dimensional remote inverse design
problem using a supersonic business jet wing-body
configuration.
* CFD Laboratory
Department of Mechanical Engineering
McGill University
688 Sherbrooke Street West, Room 711
Montreal, QC H3A 2S6, Canada.
** Department of Aeronautics and Astronautics
Stanford University
Durand Building, 496 Lomita Mall
Stanford, CA 94305, USA
E-mail: siva.nadarajah@mcgill.ca
non-dimensional coefficient
Euler numerical flux vector
face areas of computational cell
Received 12 April 2005

2005 CASI
187

distance de la surface de lavion. Cet article prsente,

premirement, la formulation des quations adjointes et
leur dicrtisation. Nous y discutons galement du
traitement spcial de la condition aux limites adjointe
permettant de driver les sensibilits distance. En second
lieu, nous montrons lapplication de la thorie un
problme inverse tridimensionnel de conception tenant
compte des effets distance, visant la configuration du
fuselage et des ailes dun ract daffaires supersonique.
density
weighting coefficients
Subscripts
contributions associated with variation of the state
vector
Lagrange multiplier
II
contributions associated with variation of the shape

function
drag
i, j, k
cell indices
contributions associated with variation of the state

vector
II
contributions associated with variation of the shape

functions
max
maximum index
NF
near field
target
wall
+,
cells across the cell face
INTRODUCTION
A 2001 US National Research Council study (Committee on
Breakthrough Technology for Commercial Supersonic Aircraft
2001) concluded that, the sonic boom is the major barrier to
the development of the supersonic business jet and a major, but
not the only, barrier to the development of supersonic transports
with overland capability. The Committee also determined that
there was a potential market for at least 200 supersonic business
jets over a 10-year period. The 815 passenger jets will
probably fly at approximately Mach 1.8 with a range of 3000
4000 nautical miles (1 nautical mile = 1.852 103 m).
Concorde, a commercial supersonic aircraft built by France
and Britain was in operation for more than 30 years. It cruised
at Mach 2.0 with a total range of 4090 nautical miles at
60 000 ft (1 ft = 3.048 10G1 m) and consumed 6200 gallons of
fuel per hour (1 gallon = 4.54609 10G3 m3). The aircraft
retired from commercial service on the 24 October 2003 due to
188
both commercial and technical factors. The new breed of

supersonic transports must possess superior performance
characteristics apart from low sonic boom capability, compared
to its predecessors to compete in the commercial jet industry.
These include improvements in structures, aerodynamics, and
propulsion. In particular, the experience of NASAs HSR
program suggests that it should be possible to improve the liftto-drag ratio from the value of 7.5 attained by the Concorde to
around 9 (Cliff et al., 2001). Designs of supersonic transports
of the future will benefit from multidisciplinary optimization
techniques that were not available during the design and
construction of aircraft like the Concorde.
Before the sonic boom reduction problem can be attempted,
it is important to have the capability of calculating the sonic
boom or ground pressure signature accurately. For typical
cruise altitudes required for aircraft efficiency, the distance
from the source of the acoustic disturbance to the ground is
typically greater than 50 000 ft. A reasonably accurate
propagation of the pressure signature can only be obtained with
small computational mesh spacings that would render the
analysis of the problem intractable for even the largest parallel
computers. An approach that has been used successfully in the
past is the use of near-to-far field extrapolation of pressure
signatures based on principles of geometrical acoustics and
non-linear wave propagation. These methods are based on the
solutions of simple ordinary differential equations for the
propagation of the near-field pressure signature to the ground.
The Whitham F-function (Whitham, 1952) and Thomas
equivalent wave-form parameter method (Thomas, 1972) are
common methods of choice.
Figure 1 is a schematic of the propagation of the aircraft
pressure signature. CFD Far Field indicates the far-field
boundary of the mesh. At a pre-specified distance below the
aircraft and still within the CFD mesh, the location of a nearfield plane can be seen. This plane is the effective interface
between the CFD solution and the wave propagation program.
At the near-field plane, the flow solution wo is represented
using a number of parameters, M, which can be taken, for
example, as the number of mesh points on which the pressure
wave-form has a value significantly different from the free
stream. The lower portion of the domain between the CFD near
field and the ground plane is where the pressure signature
propagation method will be active. Given the conditions, wo,
the propagation altitude, and the altitude-dependent
atmospheric properties (z), p(z), T(z), the propagation method
produces a pressure signature at the ground plane we are
interested in, which can be used to determine any of a variety of
measures of sonic boom impact such as overpressures, rise
time, impulse, etc.
Traditional methods to reduce the sonic boom signature
target aircraft weight reduction, increases in lift-to-drag ratio,
better specific fuel consumption, etc. Seebass and Argrow
(1998) revisited sonic boom minimization and provided a
detailed study of sonic boom theory and figures of merit for the
level of sonic booms. Diverse methods have been employed in
the design of low-boom aircraft configurations. The following
2005 CASI

Figure 1. Schematic of the propagation of the aircraft pressure signature.
are a selected number of papers on this topic. Orr et al. (2002)

proposed a rather exotic concept to reduce the sonic boom.
Their goal was to increase the apparent length of the aircraft by
off-axis volume addition. A swept forward keel placed normal
to the Mach plane increased the apparent length and proved to
be effective in reducing the sonic boom strength. Komadina et
al. (2002) evaluated twelve different configurations. The
ground sonic boom signature, aircraft aerodynamics, mass
properties, and flight performance were evaluated for all twelve
configurations using empirical methods. The two most
promising concepts were then chosen and higher fidelity
methods were used to compute the vehicle performance and
characteristics. Argrow et al. (2002) argued that most sonic
boom minimization techniques shape the aircraft equivalent
body of revolution in the vertical plane and not the true
geometry to reduce the far-field pressure signatures. The
authors used a combination of linearized and non-linear
aerodynamic theories, computational fluid dynamics based on
the Euler equations, and a gradient-based method to optimize
the shape of the aircraft. The design variables were nose tilt
angle and canard and wing dihedral, sweep, and twist angles.
Through the support of the DARPA QSP Program, advanced
algorithms for the design and optimization of quiet supersonic
platforms have been developed during the last several years.
DARPAs vision for the project was to develop conceptual
aircraft designs that produced initial overpressures of 0.3 psf
(1 psf = 4.88243 kg/m2), while cruising at Mach 2.5 with a
range of 6000 nautical miles and with a weight around 100 000
lbs (1 lb = 4.53592 10G1 kg). This is an ambitious reduction in
2005 CASI
the initial overpressure compared to the Concordes 1.5

2.0 psf. Our experience has indicated that large reductions in
the ground peak pressure cannot be achieved with minor shape
modifications of the baseline configuration. Alternative design
methods such as genetic algorithms have been used in a multilevel design environment to reach the neighborhood of the
optimum design before switching over to a gradient-based
method to refine the design. Promising results have been
achieved by using genetic algorithms in a linear method
environment. Non-linear methods are needed to meet several
goals: first, to verify if not improve the results of the linearbased method; second, to improve the design by using the
techniques of optimal control; lastly, to allow the introduction
of more objective functions to improve the final design.
The concept presented in this work proposes the idea that the
ground pressure signature could be adjusted by modifying the
aircraft surface geometry to control the near-field pressure
distribution. It is not at all clear what type of changes to the
surface geometry would produce near-field pressure
distributions whose propagation to the ground would generate
sonic booms with lower peaks. It appears, however, that the
problem might be separated into two parts: first, the
identification of near-field pressure distributions that are both
feasible and lead to acceptable ground signatures; and second,
the design of the surface geometry such that it will produce the
desired near-field pressure distribution.
Traditional adjoint implementations have been aimed at
reducing a cost function computed from the pressure
distribution on the surface that is being modified. In this case,
189

however, we would like to obtain the sensitivity of pressure

distributions at points located at a distance from the surface
where the geometry is being modified. This type of sensitivity
calculation has not been attempted before in aerodynamic
shape design, but is closely related to inverse-scattering
problems in acoustics and electromagnetics. In such an
approach, a target near-field pressure distribution must be
specified. The cost function may then be chosen as the integral
of the square of the difference between the current and target
near-field pressure distribution. The gradient of the cost
function with respect to the design variables such as the
position of the surface mesh points is calculated, and a
direction of improvement is obtained from an optimization
algorithm. The procedure is repeated until a new aircraft
surface geometry is produced that provides a near-field
pressure signature that approaches the specified target near
field pressure distribution, provided that it is actually
realizable. The design procedure should also include other
objective functions and constraints to maintain or improve
other aircraft performance parameters such as lift-to-drag ratio
and the total amount of lift. The possibility that the adjoint
method could be adapted to solve the remote inverse problem
was demonstrated by Nadarajah et al. (2001) for a twodimensional internal flow problem. The method was then
extended for three-dimensional wing and wingbody
configurations in supersonic flow by Nadarajah et al. (2002a,
2002b, 2002c).
The issue of choosing a near-field signature to produce a
desired ground signature was addressed by Alonso et al.
(2002). The work accomplished in this research focuses on
controlling the near field signature, and not the ground
signature. A future extension of the method would be to include
the wave propagation program in the design procedure, such
that the ground pressure signature is considered as the target
pressure distribution instead of the near-field pressure
distribution.
The aerodynamic properties that define the cost function are

functions of the flow field variables, w, and the physical
location of the boundary, which may be represented by the
function S.
Suppose that the performance is measured by a cost function
BW
M(w, S) d B + 2
NF
N (w, S) d B
I = 1
BW
M(w, S) d B + 2
NF
N (w, S) d B
(2)
with M = [Mw ]I w + MII , N = [N w ]I w + N II ,

where we use the subscripts I and II to distinguish between
the contributions associated with the variation of the flow
solution w and those associated with the metric variations
S. Thus, [Mw ]I and [N w ]I represent M/ w and N / w
with the metrics fixed, while MII and N II represent the
contribution of the metric variations S to Mand N with the
flow solution fixed. The weak form of the Euler equations for
steady flow is
T
Fi d D = n i TFi dB
B
i
where the test vector is an arbitrary differentiable function,

and ni is the outward normal at the boundary. If a differentiable
solution, w, for this equation is obtained, then it can be
integrated by parts to give
D T i Fi d D = 0
(3)
Since this is true for any , the differential form can be

recovered. Here, Fi can be split into contributions associated
with w and S using a similar notation
where
[ Fiw ]I = Sij
f j
w
The domain can then be split into two parts as shown in

Figure 2. First, the near-field domain (D1) whose boundaries
are the wing surface and the near-field boundary plane. Second,
the far-field domain (D2), which borders the near-field domain
along the near-field boundary plane and the far-field boundary.
Thus equation (3) can be written as
(1)
containing both wall boundary (BW) and near field boundary

(BNF ) contributions, where dB includes the surface and nearfield elements in the computational domain, while 1 and 2
are the weighting coefficients. The coordinates i that describe
the fixed computational domain are chosen so that each
boundary conforms to a constant value of one of these
190
coordinates. In general, M and N will depend on both the flow

variables w and the metrics S defining the computational space.
The design problem is now treated as a control problem
where the boundary shape represents the control function,
which is chosen to minimize I subject to the constraints defined
by the flow equations. A shape change produces a variation in
the flow solution, w, and the metrics, S, which in turn
produce a variation in the cost function
Fi = [ Fiw ]I w + Fi II
THE REMOTE INVERSE DESIGN PROBLEM

USING CONTROL THERAPY
I = 1
T
1
Fi d D + T
Fi d D = 0
D2
i
i
2005 CASI
The corresponding wall and near-field adjoint boundary

conditions are produced by equating the subscript I boundary
terms in equation (5) to produce
n i T[ Fiw ]I = 1 Mw
on
BW
n i ( + ) T [ Fiw ]I = 2 N w
(7)
on
BNF
(8)

The remaining terms from equation (5) then yield a simplified

expression for the variation of the cost function, which defines
the gradient
I =
BW
D1
B NF
D1 + D2
T
Fi d D
i
n i ( + ) T Fi d D
D2
T
Fi d D = 0
i
(4)
where + and are the values of the Lagrange Multiplier, ,

above and below the boundary. Since the left-hand expression
equals zero, it may be subtracted from the variation in the cost
function (2) to give
I =
BW
[1M n i TFi ] d B
[ 2N n i( + ) Fi ] d B
T
T
Fi d D +
Fi d D
D
D
2
1
i
i
B NF
D1
(5)
Since is an arbitrary differentiable function, it may be chosen

in such a way that I no longer depends explicitly on the
variation of the state vector w. The gradient of the cost
function can then be evaluated directly from the metric
variations without having to re-compute the variation w
resulting from the perturbation of each design variable.
Comparing equations (2) and (4), the variation w may be
eliminated from Equation (5) by equating all field terms with
subscript I to produce a differential adjoint system governing
T
= 0 in D
[ Fiw ]I
D1 + D2
i
[Fi ]II dD
This may be integrated by parts to give

n i T Fi d D
{ 2 N II n i( + ) [Fi ]II} dB
T
B NF
Figure 2. Near-field and far-field domains.
{1 MII n i T[Fi ]II } d B
(6)
(9)
The details of the formula for the gradient depend on the way in
which the boundary shape is parameterized as a function of the
design variables and the way in which the mesh is deformed as
the boundary is modified. The boundary conditions satisfied by
the flow equations restrict the form of the left-hand side of the
adjoint boundary conditions (7) and (8). Consequently, the
boundary contribution to the cost functions M and N cannot be
specified arbitrarily. Instead it must be chosen from the class of
functions that allow cancellation of all terms containing w in
the boundary integral of equation (5). In this research, the cost
function is the weighted sum of the drag coefficient and the
Sobolev norm of the difference between the current and target
remote pressure distributions. From equation (1), M and N can
be defined as
M(w, S) =
1 y
x
Cp cos
sin
c
and
N (w, S) =
1
( p pT) 2
2
The cost function can then be written as

I = 1
+ 2
y
x
1
sin d B
C p cos
B
c W

1
2
NF
( p pT) d B
and further simplified to
2005 CASI
191

I = 1 C D + 2
2
1
( p pT) d B
B
2 NF
(10)
The values of the weighting coefficients are selected based

on the relative magnitude of the gradients of the drag
minimization and the remote inverse cost functions. The remote
inverse gradient is typically an order of magnitude smaller than
the gradient due to drag minimization. Therefore, the weights
are chosen to increase the magnitude of the gradient from the
remote inverse cost function. In practice, larger weights are
used for the remote inverse gradient, since the primary design
objective is to reduce the near-field pressure signature. The
disadvantage of this approach is that the weights must be
chosen at the beginning of the design process and if the user
does not have prior knowledge of the magnitude of the
gradients, then generally an initial guess is taken. The weights
are altered for subsequent runs.
An alternative method for problems with more than one
objective function is to develop separate adjoint equations, one
for each objective function. Both gradients are then calculated
separately, multiplied by weights, and summed. A direction of
improvement is then based on the new gradient. This method
has the advantage that the user is better equipped with
knowledge regarding the difference in magnitude between the
two gradients. Appropriate weights are chosen to achieve the
desired compromise. A disadvantage is the need to calculate a
separate adjoint solution for each objective function.
In this work, we prefer to use a composite cost function,
since we had a priori knowledge regarding the magnitude of the
gradient contribution from the remote inverse and the drag
minimization cost functions.
REMOTE INVERSE DESIGN VIA THE DISCRETE

ADJOINT METHOD
The remote inverse adjoint method can be formulated either
by the continuous or discrete adjoint method. The previous

section offers an overview of the continuous adjoint approach.
The continuous remote adjoint boundary condition would
require an update of the adjoint variables at the near-field cells.
Since the near-field cells, do not lie in a specific row or column
of cells in the vicinity of the near field, but rather as a group of
cells scattered along the near-field plane, the implementation of
the continuous adjoint approach would prove to be far to
complex. However, the discrete adjoint approach proved to be
less complicated, and offered a manageable approach to the
implementation of the remote adjoint boundary condition
problem.
The discrete adjoint equation is obtained by applying control
theory directly to the set of discrete field equations. To
formulate the discrete adjoint equation, we first define the cost
function I as such,
I = 1 C D + 2
1
( p pT)2 s
2 NF
where CD is total wing drag coefficient, p is the current near

field pressure, and pT is the target near-field pressure. Next, we
take a variation of the residual term, which can be written as
R(w) i, j , k = h
1
i+ , j , k
2
1
i , j , k
2
1
i, j , k
2
+ h
+ h
1
i, j + , k
2
i, j , k +
1
2
i, j , k
1
2
(11)
with hi+1/2,j,k = fi+1/2,j,k di+1/2,j,k, where f is the convective

flux and d is the artificial dissipation flux. We then pre-multiply
the variation of the discrete residual by the Lagrange Multiplier
and sum the product over the computational domain to produce
the following:
imax j max kmax
Ti, j , k R(w)i, j , k
i= 2 j = 2 k = 2
The variation of the cost function, I, can then be augmented by the product of the variation of the discrete governing equation
and the Lagrange Multiplier i, j, k .
imax j max kmax
I = 1C D + 2 ( p pT) p s + Ti, j , k R(w) i, j , k

NF
(12)
i= 2 j = 2 k = 2
To eliminate w from equation (12), terms multiplied by the variation wi,j,k of the discrete flow variables are collected and
equated to zero. The following is the resulting discrete adjoint equation:
192
2005 CASI
V =
i, j
t
1
AT + S12 1 AT2i , j , k + S13 1 AT3i , j , k ( i, j , k i 1, j , k )
S11
i , j, k
i , j, k
2 i 12 , j , k 1i , j , k
2
2
+ S11 1 A1Ti , j , k + S12 1 AT2i , j , k + S13 1 AT3i , j , k ( i +1, j. k i, j , k )

i + , j, k
i + , j, k
i + , j, k
2
2
2

+ S21 1 A1Ti , j , k + S22 1 AT2i , j , k + S23 1 AT3i , j , k ( i, j +1, k i, j , k )

i, j + , k
i .j + , k
i .j + , k
2
2
2
+ S21 1 A1Ti , j , k + S22 1 AT2i , j , k + S23 1 AT3i , j , k ( i, j , k i, j 1, k )

i, j , k
i .j , k
i, j , k
2
2
2
+ S31
A1Ti , j , k + S32
AT2i , j , k + S33
AT3i , j , k ( i, j , k +1 i, j , k )
1
1
1
i, j, k +
i, j, k +
i, j, k +2
2
2
+ S31
A1Ti , j , k + S32
AT2i , j , k + S33
AT3i , j , k ( i, j , k i, j , k 1)
1
1
1
i, j, k
i, j, k
i, j, k 2
2
2
+D
1
i+ , j , k
2
+D
1
i , j , k
2
1
i,j + , k
2
1
i,j , k
2
+D
i,j,k +
1
2
i,j,k
1
2
(13)
Here, V is the cell area and

D
1
i+ , j , k
2
= v2 1
i+ j , k
2
+ 2 v ( 4)1
i+ j , k
2
( Ti +1, j , k
1
i+ , j , k
2
iT, j , k ) v ( 4)3
( Ti +1, j , k
1
i+ , j , k
2
i+ j , k
2
iT, j , k ) v ( 4)1
i j , k
2
( iT+ 2, j , k
3
i+ , j , k
2
iT+1, j , k )
iT1, j , k )
( Ti, j , k
1
i , j , k
2
is the discrete adjoint artificial dissipation term that

corresponds to the discretization of the inviscid flow equations
by the JamesonSchmidtTurkel (JST) scheme (1981). The
dissipation coefficients (2) and ( 4) are functions of the flow
variables, but, to reduce complexity, they are treated as
constants. The effect of this partial discretization has been
explored by Nadarajah and Jameson (2000, 2001).
Discrete Adjoint Boundary Condition
To develop the discrete adjoint boundary condition for the
calculation of remote sensitivities for supersonic flow, the w NF
term from the discrete cost function is added to the
corresponding term from equation (12). The discrete boundary
condition appears as a source term in the adjoint fluxes. At the
NF cell, the source term NF for inverse design is added to
equation (13) and can be written as,
NF = 1( p pT) sNF pNF
The wall boundary condition appears as source terms in the
adjoint fluxes along the cells above the wall. The derivation of
this boundary condition was explored by Nadarajah and
Jameson (2000). For a first-order dissipation scheme, the
2005 CASI
discrete adjoint equations are completely independent of the

co-state variables in the cells below the wall. However, if we
use the blended first-and-third-order scheme, these flow
variable values are required. A simple zeroth-order
extrapolation across the wall has produced good results, as
shown by Nadarajah and Jameson (2000).
OPTIMIZATION PROCEDURE
In this paper, the inverse design boundary condition is
applied to the near field, while sensitivity derivatives or the
gradient are calculated on the airfoil surface. The gradient for
the discrete adjoint is obtained by perturbing each point on the
lower wall. Once the gradient G has been determined, it can be
used to drive a variety of gradient-based search procedures. The
search procedure used in this work is a descent method in
which small steps are taken in the negative gradient direction.
Let F represent the design variable, and G the gradient. Then an
improvement can be made with a shape change
F = G
193
However, it is better to replace the gradient G by a smoothed

value G in the descent process. This acts as a preconditioner
that allows the use of much larger steps and ensures that each
new shape in the optimization sequence remains smooth. To
apply smoothing in the 1 direction, the smoothed gradient G
may be calculated from a discrete approximation to

= G,
1 1
G =0
at end points
where is the smoothing parameter. If the modification is

applied on the surface 2 = constant, then the first-order change
in the cost function is
I = G F d 1

= G
G G d 1
1 1
G
= G 2 +
d 1
Figure 3. Business jet wingbody configuration: biconvex wing,

1934933 CH grid, 8 blocks.
<0
again guaranteeing an improvement unless G = G = 0 and
assuring an improvement if is sufficiently small and positive.
In some problems, it turns out that the Hessian can be
represented as a second-order differential operator, so that with
a proper choice of the smoothing parameter, the method
becomes the Newton method. Search methods were intensively
evaluated in a recent study by Jameson and Vassberg (1999),
and it was verified that these sample problems (which may have
a high linear content) could be solved with a number of search
steps independent of the number of design variables.
RESULTS
This section presents the results of remote inverse and drag
minimization for wingbody configurations in supersonic flow.
The objective is to reduce the peak pressure at the near-field
plane and thus reduce the ground signature peak. Viscous
effects are likely to be very small in these examples, so it is
sufficient to use the Euler equations. The calculations were
performed with the new SYN88-MBC multiblock code that
takes advantage of the FORTRAN 90/95 derived data type
architecture. The flow solver is augmented with an adjoint
solver and shape modification routines to allow for automatic
shape optimization.
WingBody Configuration: Sonic Boom Reduction,
Without Lift Constraint, Wing Redesign Only
The wingbody supersonic business jet configuration was
sized to accommodate between 6 to 8 passengers with a gross
take-off weight of 100 000 lbs and a fuselage length of 100 ft.
The supersonic flight condition at which all designs were
194
Figure 4. Convergence history of the flow solver.
calculated is Mach 1.5. Figure 3 shows the wingbody

configuration. The fuselage is cylindrical and the maximum
diameter occurs at 31% of the fuselage length measured from
the nose. The wing is a biconvex wing with a 7.125 leading
edge sweep, an aspect ratio of 3.0, and a taper ratio of 0.218.
The root airfoil is a 3% thick biconvex airfoil and the tip is
1.5% thick. The biconvex profile in the center sections was
obtained by interpolating between the root and tip. The airfoils
were constructed to accommodate deep spars at the 10% and
80% chord locations. The baseline wing does not have
geometric twist. The computational mesh has eight blocks with
193 49 33 nodes on a CH grid. The fuselage has 25 points
in the cross-streamwise-direction and 144 points in the
streamwise-direction. The wing contains 97 points in the
streamwise-direction and 17 sectional cuts in the spanwise 2005 CASI

direction. The far-field boundary is located at approximately 5

root chord lengths. In Figure 4, convergence history of the
flow solver is illustrated for the Mach 1.5 test case.
To obtain an accurate representation of the far-field pressure
signature through a wave propagation method, a precise nearfield pressure distribution must be acquired. The location of the
near-field boundary is largely influenced by the number of
mesh points. A large number of mesh points would result in
smaller mesh spacings in the region of the near field and
produce a pressure distribution of greater accuracy, however,
the computational cost would increase dramatically. If located
too close to the aircraft, a proper far-field representation of the
aircraft pressure signature may not be feasible. In this work, the
near-field boundary is located at approximately one root chord
length below the aircraft on the 193 49 33 CH mesh.
To illustrate the possibility of sonic boom reduction, a target
pressure distribution was obtained by re-scaling the initial nearfield pressure distribution. Ultimately, this step will be replaced
by a method that produces a target near-field pressure based
upon the desired ground pressure signature. The target pressure
was obtained by taking the result of SYN88-MBC at a flight
condition of M = 1.5 and scaling the resulting pressure
distribution to 40% of its original value.
The objective function is the integral of the difference
between the current and target near-field pressures. The
minimum permissible thickness constraint is imposed at each
chordwise cut between the 10% and 80% chord locations at the
end of each design cycle. Points from the leading edge upto the
10% chord location and from the 80% chord location to the
trailing edge are not constrained and free to move in any
direction. The lift coefficient in this case is not constrained. In
this design, the design variables are the locations of all of the
points on the surface of the wing. Therefore, only the second
peak in the near-field pressure profile will be expected to
change. The flow is calculated at Mach 1.5 at a fixed angle of
attack of 1.62.
Figure 6. Target, initial, and final near-field pressure distribution after

50 design cycles.
Figure 7. Convergence of the objective function.
Figure 5. Initial and final root airfoils at M = 1.5.
2005 CASI
Figure 5 illustrates the initial and final root airfoil profiles.

The lower surface of the final airfoil contains a slightly larger
expansion region when compared with the original biconvex
airfoil. It is this modification that allows the near-field wing
peak pressure (second peak) to be reduced. The larger
expansion region weakens the strength of the leading edge
attached shock in the near-field region. Figure 6 shows the
initial near-field pressure in blue () and the target pressure in
black (+). After 50 design cycles, the final near-field pressure
distribution is obtained and illustrated as the red (*) line. The
wing peak pressure has been reduced by 40%. The large
modifications on the wing upper surface are a result of the
thickness constraint. Since the lift coefficient was not
constrained, CL reduced from 0.1 to 0.073. The baseline wing
drag coefficient is 0.00568 and the final wing drag increased
slightly to 0.00582. Even if drag due to lift has decreased due to
195

arbitrarily scaled near-field signatures may not result in more

desirable behavior at the ground.
Figure 8. Initial and final ground signatures after 50 design cycles.
the decrease in the lift coefficient, the stronger attached leading

edge shock has increased the wing wave drag.
In Figure 7, the convergence of the objective function is
shown. Here, the objective function as defined in equation (10)
is computed at each design cycle. For this test case, 1 is set to
zero and thus the objective function only represents the norm of
the difference between the current and target near-field
pressure. The near-field pressure quickly reaches the vicinity of
the final result within 20 design cycles. A total of 50 design
cycles are computed to ensure that the objective function has
reached its minimum value.
The complete shape optimization procedure for sonic boom
reduction requires the determination of desirable ground boom
signatures. In Figure 8, we show the initial and final ground
signature profiles. The PC Boom software for far-field
propagation developed by Wyle Associates was used to
calculate the ground signatures. The results clearly indicate that
WingBody Configuration: Sonic Boom Reduction, Lift

Constraint
We now repeat the same design case but with the following
three changes: first, the lift coefficient is constrained at 0.1.
Second, gradients are calculated for points on the surface of the
fuselage and thus allowed to be modified. Third, the objective
function is the weighted sum of the drag coefficient and integral
of the difference between the current and target near-field
pressures, where 1 is the weight on the drag coefficient and
2 is the weight on the remote inverse cost function. In this
case, the drag coefficient weight is, 1 = 0.005 and the remote
inverse cost function weight is set to 2 = 1.
The value of the lift coefficient is maintained by adjusting
the angle of attack to attain the desired lift coefficient of 0.1.
The wing thickness constraint is imposed in the same manner
as the previous case.
Figure 9 illustrates the baseline and optimized airfoil.
Figure 10 shows the target, initial, and final near-field pressure
distributions. The desired target pressure distribution is not
achieved in contrast with the unconstrained case illustrated in
Figure 6. In this case, there is a struggle between the near-field
peak pressure reduction versus maintenance of constant lift.
Each design cycle, produces a shape modification that shifts the
near-field pressure distribution towards the target pressure.
Unfortunately, this also causes a reduction in the lift
coefficient. This must be compensated by an increase in the
angle of attack to maintain the total lift coefficient, which in
turn leads to an increase in the near-field peak pressure. After
50 design cycles, the solution converges to the (*) line in
Figure 10. The final fuselage peak pressure has been reduced
to almost 18% its original value and the wing peak pressure is
reduced by 22%.
To maintain the lift coefficient, the angle of attack was
increased from 1.62 to 2.39. The wing drag increased slightly
Figure 9. Initial and final root airfoils at M = 1.5. (a) Wing root section and (b) mid-span section.
196
2005 CASI

Figure 10. Target, initial, and final near-field pressure distribution after 50 design cycles. (a) Wing root section and (b) mid-span section.
Figure 11. Sonic boom reduction: initial and final ground signatures
after 50 design cycles. M = 1.5, = 2.39, CL = 0.1.
Table 1. Near field peak pressure reduction and wing drag coefficient
for various design cases.
Case
Baseline
Remote inverse
Drag and remote inverse
Fuselage peak
reduction
18%
18%
Wing peak
reduction
Wing,
CD
25%
22%
0.00568
0.00610
0.00574
from 0.00568 to 0.00574. In an alternate test case, where the

drag coefficient weight was set to zero, the near-field peak
pressure for both the fuselage and wing were reduced by 18%
and 25%, however, the wing drag coefficient increased to
0.00610. Table 1 contains a comparison of the two design
cases. Table 1 clearly shows that a cost function that does not
include the drag coefficient will result in larger reductions in
the wing near-field peak pressure. However, in a multi 2005 CASI
disciplinary design environment, it is critical that other

important parameters are kept within acceptable amounts and a
trade-off between the various design goals are met. Table 1
clearly shows that a composite cost function that includes the
drag coefficient is unable to reduce both the near-field peak
pressure and drag coefficient but it is able to reduce the peak
pressure while maintaining the wing drag coefficient. A more
detailed study of the effect of the weights on the cost functions
is presented by Nadarajah et al. (2002).
In Figure 12, both the initial and final pressure contours are
plotted. The majority of the changes in the shape are localized
around the lower surface wingfuselage intersection. The
larger expansion regions on the lower surface of the wing are
illustrated in these plots by the shorter red region (compression)
and the longer green-orange region. This has the effect of
weakening the strength of the shock, and thus reducing the peak
of the near-field pressure. Figure 13 illustrates the initial and
final fuselage mesh. The larger expansion region on the
underside of the fuselage around the wingfuselage
intersection is clearly due to the increase in the fuselage
curvature.
CONCLUSIONS
The results presented in this paper demonstrate the
feasibility of remote inverse calculations using the adjoint
method. An application to the sonic boom minimization
resulted in a 40% reduction in the near-field peak pressure for
the unconstrained biconvex wing. In the constrained problem,
the fuselage peak pressure was reduced by 18% and the wing
peak by 22%. It proved highly beneficial to use a composite
cost function consisting of the sum of the weighted remote
inverse and drag minimization cost functions, resulting in final
designs that had a reduction in the peak pressure while
maintaining constant inviscid drag. Cases with no drag
coefficient added to the integral of the near-field pressure
197

Figure 12. Pressure contour at M = 1.5. (a) Initial and (b) final.
Figure 13. Fuselage mesh. (a) Initial and (b) final.
difference in the objective function saw an increase in the drag

coefficient.
Acknowledgments
This research has benefited greatly from the generous
support of the AFOSR under grant number AF F49620-98-1022 and DARPA QSP Program under grant number MDA97201-2-0003.
References
Alonso, J.J., Jameson, A., and Kroo, I. (2002). Advanced Algorithms for
Design and Optimization of Quiet Supersonic Platforms. Proceeedings of the
40th AIAA Aerospace Sciences Meeting and Exhibit, Reno, Nevada, 1417
January 2002. American Institute of Aeronautics and Astronautics, Reston,
Virginia. AIAA Pap. 02-0144.
198
Argrow, B., Farhat, C., Maute, K., and Nikbay, M. (2002). A Shape
Optimization Methodology for Reducing the Sonic Boom Initial Pressure
Rise. Proceedings of the 40th AIAA Aerospace Sciences Meeting and Exhibit,
Reno, Nevada, 1417 January 2002. American Institute of Aeronautics and
Astronautics, Reston, Virginia. AIAA Pap. 02-0145.
Cliff, S.E., Reuther, J.J., Saunders, D.A., and Hicks, R.M. (2001). Singlepoint and Multipoint Aerodynamic Shape Optimization of High-Speed Civil
Transport. J. Aircr. Vol. 38, No. 6, November-December, pp. 9971005.
Committee on Breakthrough Technology for Commercial Supersonic
Aircraft. (2001). Commercial Supersonic Technology: The Way Ahead.
National Research Council (US), National Academy Press, Washington, D.C.
Tech. Rep., March.
Jameson, A., and Vassberg, J.C. (1999). Studies of Alternative Numerical
Optimization Methods Applied to the Brachistochrone Problem. OptiCON
99.
Jameson, A., Schmidt, W., and Turkel, E. (1981). Numerical Solutions of
the Euler Equations by Finite Volume Methods with Runge-Kutta Time
Stepping Schemes. AIAA Pap. 81-1259, January.
2005 CASI
Komadina, S., Drake, A., and Bruner, S. (2002). Development of a Quiet

Supersonic Aircraft with Technology Applications to Military and Civil
Aircraft. Proceedings of the 40th AIAA Aerospace Sciences Meeting and
Exhibit, Reno, Nevada, 1417 January 2002. American Institute of
Aeronautics and Astronautics, Reston, Virginia. AIAA Pap. 02-0519.

Nadarajah, S., and Jameson, A. (2000). A Comparison of the Continuous

and Discrete Adjoint Approach to Automatic Aerodynamic Optimization.
Proceedings of the 38th AIAA Aerospace Sciences Meeting and Exhibit, Reno,
Nevada, 1013 January 2000. American Institute of Aeronautics and
Astronautics, Washington, D.C. AIAA Pap. 2000-0667.
Nadarajah, S., and Jameson, A. (2001). Studies of the Continuous and
Discrete Adjoint Approaches to Viscous Automatic Aerodynamic Shape
Optimization. A Collection of Technical Papers: 15th AIAA Computational
Fluid Dynamics Conference and Exhibit, Anaheim, California, 1114 June
2001. American Institute of Aeronautics and Astronautics, Reston, Virginia.
AIAA Pap. 2001-2530.
Nadarajah, S., Jameson, A., and Alonso, J.J. (2001). An Adjoint Method
for the Calculation of Non-collocated Sensitivities in Supersonic Flow.
Proceedings of the 1st MIT Conference on Computational Fluid and Solid
Mechanics, Cambridge, Massachusetts, 1215 June 2001. Vol. 2. Edited by
K.J. Bathe. Elsevier, New York, New York. pp. 921925.
Nadarajah, S., Jameson, A., and Alonso, J.J. (2002a). An Adjoint Method
for the Calculation of Remote Sensitivities in Supersonic Flow. Proceedings
of the 40th AIAA Aerospace Sciences Meething and Exhibit, Reno, Nevada,
1417 January 2002. American Institute of Aeronautics and Astronautics,
Reston, Virginia. AIAA Pap. 2002-0261.
Nadarajah, S., Jameson, A., and Alonso, J.J. (2002b). Sonic Boom
Reduction using an Adjoint Method for Wing-Body Configurations in
Supersonic Flow. Proceedings of the 9th AIAA/ISSMO Symposium on
Multidisciplinary Analysis and Optimization Conference, Atlanta, Georgia, 4
6 September 2002. Vol. 2. American Institute of Aeronautics and Astronautics,
Reston, Virginia. AIAA Pap. 2002-5547.
Nadarajah, S.K., Kim, S., Jameson, A., and Alonso, J.J. (2002c). Sonic
Boom Reduction using an Adjoint Method for Supersonic Transport Aircraft
Configuration. Proceedings of the IUTAM Symposium Transsonicum IV,
Gttingen, Germany, 26 September 2002. Edited by H. Sobieczky. Kluwer
Academic Publishers, Dordrecht, The Netherlands. p. 355.
Orr, M., Mozingo, J., Marconi, F., Bowersox, R., and Schetz, J. (2002).
Sonic Boom Alleviation using Keel Configurations. Proceedings of the 40th
AIAA Aerospace Sciences Meeting and Exhibit, Reno, Nevada, 1417 January
2002. American Institute of Aeronautics and Astronautics, Reston, Virginia.
AIAA Pap. 02-0149.
Seebass, R., and Argrow, B. (1998). Sonic Boom Minimization Revisited.
Proceedings of the 2nd AIAA Theoretical Fluid Mechanics Meeting,
Albuquerque, New Mexico, June 1998. AIAA Pap. 98-2956.
Thomas, C. (1971). Extrapolation of Sonic Boom Pressure Signatures by
the Waveform Parameter Method. NASA TN D-6832.
Whitham, G.B. (1952). The Flow Pattern of a Supersonic Projectile.
Commun. Pure Appl. Math. Vol. 3, pp. 301348.
2005 CASI
199


Instructions to Contributors
he Canadian Aeronautics and Space Institute invites the
submission of technical papers and technical notes for
publication in the Canadian Aeronautics and Space Journal.
Following the practice of other societies, contributors are expected
to cover page charges, which are currently set at $125 per typeset
page.
Papers are published in either English or French, whichever is
the language of the submission. Abstracts are in both languages.
INITIAL SUBMISSIONS
FOR
CONSIDERATION
Initial submissions should be made to the following website

http://mc.manuscriptcnetral.com/casj.
REFEREES
Authors will be requested during the submission process to
provide names, addresses, phone numbers and e-mail addresses of
four persons qualified to review their papers.
REFERENCES
References should be cited by the author and should be treated as
follows:
(a) references in the text should be included in parentheses and
shown as (James, 1997) or (James, 1997, 1999);
(b) direct citation of a reference in the text should be written in full,
e.g. As shown in James (1997) ...; and
(c) references should be grouped together in alphabetical order at
the end of the paper, each showing:
first, authors name, last name first, e.g. James, T.T.;
second, year of publication, e.g. (1997);
third, the title of his work, e.g. Aerodynamics and Ballistics; and
fourth, name, volume, issue number, and date identifying the
publication in which it appeared, e.g. R.B.S. Journal, Vol. 7,
No. 77, pp. 234240. Thus,
James, T.T. (1997). Aerodynamics and Ballistics. R.B.S.
Journal, Vol. 7, No. 77, pp. 234240.
Smith, R.A. (1995).
Systems analysis and design.
Prentice-Hall, Inc., Englewood Cliffs, N.J.
STYLE GUIDELINES
TITLES
The following form should invariably be adopted:

(a) titles should be brief;
(b) in upper and lower case, not in uppercase only;
(c) the names and initials of the authors should be written as they
prefer;
(d) the name of the organizations with which the authors are
associated should be shown under their names; and
(e) the authors positions in the organizations, referred to in (d)
above, should be shown as a footnote to the first page.
SUB-HEADINGS AND PARAGRAPH NUMBERING
Sub-headings should be inserted by the author at frequent
intervals. Sections and paragraphs should not be numbered.
ABSTRACTS
Each paper should be preceded by an Abstract:
(a) of 100 to 300 words (10 to 35 lines, double-spaced);
(b) in non-specialist language, as much as possible;
(c) stating the main objectives; and
(d) stating the main conclusions of the paper.
As we publish the abstract of each paper in both official languages,
we ask that you supply a translation of your abstract, if possible.
2005 CASI
FOOTNOTES
Comments on or amplification of the text should be given in
footnotes, appearing at the bottom of the appropriate pages.
Footnotes should be designated numerically and consecutively
throughout the paper; and a reference to a footnote in the text should
be indicated by a small superscript number, e.g. omitting
considerations of the third power1....
FIGURES, TABLES AND EQUATIONS
Figures, Tables and Equations should be numbered, each in its
own series, and each Figure and Table should also be identified by a
caption, e.g. Figure 7. Theoretical Lift Distribution. Figures are
preferably provided as separate files in JPEG or TIFF format, with a
minimum resolution of 300 dpi.
LIST OF SYMBOLS NOTATION
A list of symbols should always be provided in a paper,
preferably immediately following the Abstract.
201

Instructions lintention des collaborateurs

Institut aronautique et spatial du Canada sollicite la
L
soumission de documents et de notes techniques en vue de leur
publication dans le Journal aronautique et spatial du Canada.
Conformment dautres socits, nous demandons nos
collaborateurs dassumer les frais de production exigs par page,
couramment fixs 125,00 $ par page compose.
Les documents sont publis soit en anglais ou en franais,
suivant la langue de la soumission initiale. Les rsums sont
prsents dans les deux langues.
SOUMISSIONS INITIALES
POUR
TUDE
Les soumission initiale doivent tre envoyes au site Web

suivant: http://mc.manuscriptcentral.com/casj.
EXAMEN DES DOCUMENTS
On demande aux auteurs de donner, au cours du processus de
soumission, le nom, ladresse, le numro de tlphone et ladresse
lectronique de quatre personnes ayant les qualifications
ncssaires pour examiner leurs documents.
STYLE MARCHE SUIVRE

TITRES
Il faut observer les rgles suivantes dans tous les cas :
(a) Les titres doivent tre concis.
(b) Le texte ne doit pas figurer seulement en majuscules.
(c) Le nom et les initiales des auteurs doivent tre inscrits de la
faon quils prfrent.
(d) Le nom des organisations auxquelles les auteurs sont associs
doit figurer sous leur nom.
(e) Le poste quoccupent les auteurs au sein des organisations
mentiones en (d) doit tre annot au bas de la premire page.
SOUS-TITRES ET NUMROTATION DES PARAGRAPHES
Lauteur doit insrer de frquents sous-titres. Les sections et les
paragraphes ne doivent pas tre numrots.
RSUMS
Chaque texte doit tre prcd dun rsum qui:
(a) comporte de 100 300 mots (de 10 35 lignes), dactylographi
double interligne;
(b) est rdig pour les non-initis, autant que possible;
(c) expose les principaux objectifs du document;
(d) expose les principales conclusions du document.
Dans la mesure du possible, veuillez nous fournir une traduction des
rsums, car la publication des rsums est faite dans les deux
langues.
202
RFRENCES
Les rfrences doivent toujours se faire partir du nom de
lauteur, et elles doivent observer les rgles suivantes :
(a) Dans le texte, les rfrences doivent tre entre parenthses.
Exemples: (James, 1997) ou (James, 1997, 1999).
(b) Les rfrences faites directement dans le corps du texte doivent
tre inscrites au long. Exemple: Comme lindique le renvoi de
James (1997) ... .
(c) Les rfrences doivent tre regroupes par ordre alphabtique
la fin du document. Y sont indiqus:
Premirement, le nom de lauteur, dernier nom en premier.
Exemple : James, T.T. .
Deuximement, lanne de la publication du document.
Exemple : (1997) .
Troisimement, le titre du document. Exemple : Arodynamique
et balistique .
Quatrimement, le titre, le volume, le numro et la date de la
publication dans laquelle le document est paru. Exemple: R.B.S.
Journal , vol. 7, no 77, pp. 234240 .
Ce qui donne :
James, T.T. (1997). Arodynamique et balistique , R.B.S.
Journal, vol. 7, no 77, pp. 234240.
Smith, R.A. (1995).
Systems analysis and design.
Prentice-Hall, Inc., Englewood Cliffs, N.J.
NOTES EN BAS DE PAGE
Les commentaires sur le texte ou le dveloppement dides
doivent tre prsents dans des notes figurant au bas des pages
appropries.
Les notes en bas de page doivent tre dsignes par un chiffre et
prsentes en ordre numrique dans tout le document. Le renvoi une
note en bas de page dans le corps du texte doit tre indiqu par un
chiffre en indice suprieur. Exemple: en omettant les considrations
de la troisime puissance1... .
FIGURES, TABLEAUX ET QUATIONS
Les figures, les tableaux et les quations doivent tre numrots
suivant lordre de parution dans leur groupe respectif, et chaque
figure et tableau doit tre accompagn dune lgende. Exemple:
Figure 7. Rpartition thorique de la portance . De prfrence,
les figures sont fournies sous la forme de fichiers spars en format
JPEG ou TIFF, avec une rsolution minimale de 300 dpi.
LISTE DES SYMBOLES NOTATION:
Tout document doit comporter une liste des symboles, liste qui,
de prfrence, suit immdiatement le rsum.
2005 CASI

AeroVations Associates
39 Westpark Drive, Gloucester, ON K1B 3G6, Canada
Contact: Gerry Marsters
Tel: (613) 837-9326, Fax: (613) 837-9326
E-mail: marsters@celeris.ca
www.aerovations.com
BELL HELICOPTER TEXTRON

DIVISION OF TEXTRON CANADA LTD.
12 800 rue de lAvenir Mirabel, QC J7J 1R4
Contact: Dennis Lacroix, Marketing Director, Canada
Tel: (450) 437-3400, Fax: (450) 437-2006
E-mail: dlacroix@bellhelicopter.textron.com
BOMBARDIER AEROSPACE
P.O. Box 6087, Station Centre-Ville, Montral, QC H3C 3G9
Contact: Mr. Pierre Beaudoin, President and COO
Tel: (514) 855-7900, Fax: (514) 855-7903
COM DEV INTERNATIONAL

155 Sheldon Drive, Cambridge, ON N1R 7H6
Contact: Tony Stajcer, Vice President, Research and Development
Tel: (519) 633-2300 Ext. 2231, Fax: (519) 622-1691,
E-mail: tony.stajcer@comdev.ca
CONCORDIA UNIVERSITY
Concordia Institute for Aerospace
Design and Innovation (CIADI)
1455 de Maisonneuve Blvd. W.
Montreal, QC H3G 1M8
Contact: Dr. Hany Moustapha
Director
Tel: (514) 848-7931
Fax: (514) 848-7890
E-mail: Hany.Moustapha@pwc.ca
CRESTech
4850 Keele Street, Toronto, ON M3J 3K1
Contact: Mr. Richard Worsfold
Director, Business Development
Tel: (416) 665-3311, Fax: (416) 665-2032
E-mail: enquiries@crestech.ca
CANADIAN ARMED FORCES AIR FORCE

National Defence Headquarters
MGen George R. Pearkes Building
101 Colonel By Drive, Ottawa, ON K1A 0K2
Contact: Col R.M. Williams, Director, Air Strategic Planning
Tel: (613) 995-9565, Fax: (613) 995-0063
CANADIAN SPACE AGENCY

6767 route de lAroport, Saint-Hubert, QC J3Y 8Y9
Contact: CSA Communications Directorate
Tel: (450) 926-4800, Fax: (450) 926-4352
www.space.gc.ca
CARLETON UNIVERSITY
Mechanical and Aerospace Engineering
1125 Colonel By Drive, Ottawa, ON K1S 5B6
Contact: Professor Paul Straznicky
Tel: (613) 520-5684, Fax: (613) 520-5715
E-mail: pstrazni@mae.carleton.ca
2005 CASI
DYNACON INC.
3565 Nashua Drive
Mississauga, ON L4V 1R1
Contact: Mr. Stephen J. Sorocky, CEO
Tel: (905) 672-8828, Fax: (905) 672-8829
E-mail: sjs@dynacon.ca
EMS TECHNOLOGIES
21025 Trans Canada, Ste-Anne de Bellevue, QC H9X 3R2
Contact: Mr. J. Gareth Lewis, Vice President, Sales and Marketing
Tel: (514) 425-3077, Fax: (514) 425-3005
E-mail: lewis.g@ems-t.ca
203

GASTOPS LTD.
1011 Polytek Street, Ottawa, ON K1J 9J3
Contact: Mr. Dave Muir, V.P. Marketing and Sales
Tel: (613) 744-3530, Fax: (613) 744-8846
IMP GROUP INTERNATIONAL INC.

Aerospace Division, P.O. Box 970, Enfield, NS B2T 1L5
Contact: Mr. Carl F. Kumpic, VP International Marketing
Tel: (902) 873-2250, Fax: (902) 873-2290
E-mail: carl.kumpic@impaerospace.com
www.impgroup.com/impaero
MACDONALD DETTWILER AND ASSOCIATES LTD.

13800 Commerce Parkway, Richmond, BC V6V 2J3
Contact: Wendy Keyzer, Manager, Marketing Support Dept.
Tel: (604) 278-3411, Fax: (604) 278-2936
www.mda.ca
MANNARINO SYSTEMS & SOFTWARE INC.

10 345 Papineau, Suite 250, Montreal, QC H2B 2A3
Contact: John Mannarino
E-mail: info@mss.ca
Tel: (514) 381-1360 Ext. 31, Fax: (514) 381-7511
MARINVENT CORPORATION
50 Rabastaliere East, Suite 23
St-Bruno, QC J3V 2A5, Canada
Contact: John Maris
Tel: (450) 441-6464, Fax: (450) 441-2411
E-mail: info@marinvent.com
MESSIER-DOWTY INC.
574 Monarch Avenue, Ajax, ON L1S 2G8
Contact: Mr. Tim Whittier, Director
Marketing and Business Development
Tel: (905) 683-3100, Fax: (905) 686-2914
INSTITUTE FOR AEROSPACE RESEARCH

NATIONAL RESEARCH COUNCIL CANADA
Montreal Road, Ottawa, ON K1A 0R6
Contact: Mr. J. Mackwood, Marketing Manager
Tel: (613) 990-0765, Fax: (613) 952-7214
E-mail: jeff.mackwood@nrc-cnrc.gc.ca
www.nrcaerospace.com
CANADA CENTRE FOR REMOTE SENSING

588 Booth Street, Ottawa, ON K1A 0Y7
Contact: Willy Bruce
Tel: (613) 996-2648, Fax: (613) 996-9843
E-mail: Willy.Bruce@NRCan.gc.ca
www.ccrs.nrcan.gc.ca
PRATT & WHITNEY CANADA

1000 Marie Victorin Blvd, Longueuil, QC J4G 1A1
Contact: Dr. Hany Moustapha, Manager
P&WC Technology Programs
Tel: (450) 647-7593, Fax: (450) 647-3336
E-mail: Hany.Moustapha@pwc.ca
QUEENS UNIVERSITY
99 University Avenue
Kingston, ON K7L 3N6, Canada
Contact: Mr. Thomas J. Harris
Dean, Faculty of Applied Science
Professor of Chemical Engineering
Tel: (613) 533-2000, Fax: (613) 533-6500
E-mail: harris@post.queensu.ca
www.queensu.ca
ROYAL MILITARY COLLEGE OF CANADA

P.O. Box 1700, Station Forces, Kingston, ON
K7K 7B4
Contact: LCdr Douglas Burrell, Ph.D.
Tel: (613) 541-6000 Ext. 6340
E-mail: Douglas.Burrell@rmc.ca
204
2005 CASI

RYERSON UNIVERSITY
SCHOOL OF AEROSPACE ENGINEERING
Department of Mechanical, Aerospace and Industrial Engineering
350 Victoria St., Toronto, ON M5B 2K3, Canada
Contact: Kamran Behdinan, Ph.D., P.Eng.
Tel: (416) 979-5000 Ext. 6414
Fax: (416) 979-5265
E-mail: kbehdina@ryerson.ca
Telesat Canada
1601 Telesat Court, Gloucester, ON K1B 5P4
Contact: Roger Tinley, Vice-President
Space Systems
Tel: (613) 748-0123 Ext. 2339
Fax: (613) 748-8712
E-mail: info@telesat.ca
TRANSPORT CANADA
Safety and Security Group, 12th Tower C, Place de Ville
330 Sparks Street, Ottawa, ON K1A 0N8
Contact: Mr. William Elliott, Assistant Deputy Minister
Tel: (613) 990-3838, Fax: (613) 990-2947
E-mail: elliowj@tc.gc.ca
UNIVERSITY OF ALBERTA
FACULTY OF ENGINEERING
E6-050 Engineering Teaching and
Learning Complex
Edmonton, AB T6G 2V4
Contact: Dr. David Lynch
Tel: (780) 492-3596, Fax: (780) 492-0500
E-mail: david.lynch@ualberta.ca
UNIVERSITY OF CALGARY
DEPARTMENT OF GEOMATICS ENGINEERING
2500 University Dr. NW, Calgary, AB T2N 1N4
Contact: Marguerite Anderson
Tel: (403) 220-5834, Fax: (403) 284-1980
E-mail: geomatics@geomatics.ucalgary.ca
www.geomatics.ucalgary.ca
UNIVERSITY OF TORONTO
INSTITUTE FOR AEROSPACE STUDIES
4925 Dufferin Street, Toronto, ON M3H 5T6
Contact: Dr. Tony Haasz
Tel: (416) 667-7717, Fax: (416) 667-7743
E-mail: aahaasz@utias.utoronto.ca
TRANSPORTATION SAFETY BOARD OF CANADA (TSB)

1901 Research Road, Bldg. U-100, NRC Uplands
Ottawa, ON K1A 1K8
Contact: Mr. Jim Foot, Acting Director, Engineering Branch
Tel: (613) 990-0913, Fax: (613) 998-5572
2005 CASI
205

vol. 51, no 4, dcembre 2004
Vol. 51, No. 4, december 2004


Published by / Publi par

Canadian Aeronautics and Space Institute / Institut aronautique et spatial du Canada
150 - 1750, croissant Courtwood Crescent
Ottawa, ON, Canada K2C 2B5
Tel./Tl. : 613-234-0191, Fax/Tlc. : 613-234-9039
E-mail/Courriel : casi@casi.ca, Website: 247.casi.ca
Index to Volume 51, 2005

Number 1 - March/mars
Number 2 - June/juin
pp. 145
pp. 4793
Number 3 - September/septembre
Number 4 - Decembre/dcembre
pp. 95151
pp. 153209
Adjoint-Based Sonic Boom Reduction for Wing-Body

Configurations in Supersonic Flow (No. 4, December/dcembre
2005): S.K. Nadarajah, A. Jameson, J. Alonso . . . . . . . . . . . . . . 187
Aerodynamic Forces Approximations using the Chebyshev Method
for Closed-Loop Aero-servoelasticity Studies (No. 4,
December/dcembre 2005): A.D. Dinu, R.M. Botez, J. Cotoi . . . . 167
Alonso, J., S.K. Nadarajah, A. Jameson: Adjoint-Based Sonic
Boom Reduction for Wing-Body Configurations in Supersonic
Flow (No. 4, December/dcembre 2005). . . . . . . . . . . . . . . . . . . 187
Application of Three Combustion Models to a Model Combustor
(No. 1, March/mars 2005): L.-Y. Jiang, I. Campbell . . . . . . . . . . . . 1
Applying Fast Fourier Transform Analysis and Data Window in
Software Global Positioning System Receivers to Mitigate
Continuous Wave Interference under Dynamic Conditions (No. 4,
December/dcembre 2005): Z. Jiang, G. Lachapelle, C. Ma . . . . . 177
Assessing the Global Availability and Reliability of the Mars
Network, a Proposed Global Navigation Satellite System for
Mars (No. 1, March/mars 2005): K. OKeefe, G. Lachapelle,
S. Skone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Campbell, I., L.-Y. Jiang: Application of Three Combustion Models

to a Model Combustor (No. 1, March/mars 2005) . . . . . . . . . . . . . . 1
Chesser, H., Y. Soucy: Force Limited Vibration Testing Applied to
the MOST Spacecraft (No. 2, June/juin 2005) . . . . . . . . . . . . . . . 47
Contantinescu, I., I. Ilie, R. Jr. Landry: Simulation of GPS and
Galileo Architectures for Anti-jamming and Multipath Analysis
(No. 1, March/mars 2005) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Cotoi, I., A.D. Dinu, R.M. Botez: Aerodynamic Forces
Approximations using the Chebyshev Method for Closed-Loop
Aero-servoelasticity Studies (No. 4, December/dcembre 2005) . . . 167
Cronier, Y., S. Bisson, L. Del Ciotto, C. Legare, S. Rutherford,
O. Underhill, C. Mathias, D. Lee, L. Smyth, V.D. Nguyen,
B. Tanguay, Y. Mbarki, M. Deslauriers, A. Rebaine: Recent
Improvements to the NRC 9 m 9 m Wind Tunnel (No. 3,
September/septembre 2005) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
B
Backman, D., N. Bellinger, G. Shi, G. Li: Numerical Modeling of
a Single Aluminum Sheet Containing an Interference Fit
Fastener (No. 3, September/septembre 2005). . . . . . . . . . . . . . . . 107
Bellinger, N., G. Shi, G. Li, D. Backman: Numerical Modeling of
Benay, R., A. Bourgoing: Investigation of an Asymmetrical
Shock-Wave Boundary-Layer Interaction in a Supersonic Planar
Nozzle (No. 2, June/juin 2005) . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Bisson, S., L. Del Ciotto, C. Legare, S. Rutherford, O. Underhill,
C. Mathias, D. Lee, L. Smyth, V.D. Nguyen, B. Tanguay,
Y. Mbarki, M. Deslauriers, A. Rebaine, Y. Cronier: Recent
Book Review: Orbital Mechanics for Engineering Students written
by Howard D. Curtis (No. 2, June/juin 2005): A.M. Jablonski . . . . 87
Book Review: Stress, Strain, and Structural Dynamics An
Interactive Handbook of Formulas, Solutions, and MATLAB
Toolboxes written by Bngen Lang (No. 3,
September/septembre 2005): V.K. Wickramasinghe . . . . . . . . . . . 145
Botez, R.M., I. Cotoi, A.D. Dinu: Aerodynamic Forces
Bourgoing, A., R. Benay: Investigation of an Asymmetrical
Shock-Wave Boundary-Layer Interaction in a Supersonic Planar
Nozzle (No. 2, June/juin 2005) . . . . . . . . . . . . . . . . . . . . . . . . . . 71
D
Dujardin, A., P. Hennig, L. Leavitt, F. Leopold, M. Mendenhall,
S. Prince, M. Khalid: Turbulence Model Studies to Investigate
the Aerodynamic Performance of a NASA Dual Control Missile
at Supersonic Mach Numbers (No. 4, December/dcembre 2005) . . 153
Del Ciotto, L., C. Legare, S. Rutherford, O. Underhill, C. Mathias,
D. Lee, L. Smyth, V.D. Nguyen, B. Tanguay, Y. Mbarki,
M. Deslauriers, A. Rebaine, Y. Cronier, S. Bisson: Recent
Deslauriers, M., A. Rebaine, Y. Cronier, S. Bisson, L. Del Ciotto,
C. Legare, S. Rutherford, O. Underhill, C. Mathias, D. Lee,
L. Smyth, V.D. Nguyen, B. Tanguay, Y. Mbarki: Recent
Dinu, A.D., R.M. Botez, J. Cotoi: Aerodynamic Forces
F
Floryan, J.M., P. Luchini, M. Quadrio: Modification of Turbulent
Flow Using Distributed Transpiration (No. 2, June/juin 2005) . . . . . 61
Fluid Flow and Thermodynamic Analysis of a Wing Anti-Icing
System (No. 1, March/mars 2005): J. Hua, H.H.T. Liu. . . . . . . . . . 35
Force Limited Vibration Testing Applied to the MOST Spacecraft
(No. 2, June/juin 2005): Y. Soucy, H. Chesser . . . . . . . . . . . . . . . 47
G
Giroux, R., R. Gourdeau, R. Jr. Landry: Inertial Navigation
System/Global Positioning System Fusion Algorithm Design in
a Fast Prototyping Environment: Towards a Real-Time
Implementation (No. 3, September/septembre 2005) . . . . . . . . . . . 133
207
Gourdeau, R., R. Jr. Landry, R. Giroux: Inertial Navigation

Greatrix, D.R., J. Karpynczyk: Rocket Vehicle Design for
Small-Payload Delivery to Orbit (No. 3, September/septembre
2005) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

H
Hennig, P., L. Leavitt, F. Leopold, M. Mendenhall, S. Prince,
M. Khalid, A. Dujardin: Turbulence Model Studies to
Investigate the Aerodynamic Performance of a NASA Dual
Control Missile at Supersonic Mach Numbers (No. 4,
December/dcembre 2005) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Hua, J., H.H.T. Liu: Fluid Flow and Thermodynamic Analysis of a
Wing Anti-Icing System (No. 1, March/mars 2005) . . . . . . . . . . . . 35
I
Ilie, I., R. Jr. Landry, A. Constantinescu: Simulation of GPS and
(No. 1, March/mars 2005) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Inertial Navigation System/Global Positioning System Fusion
Algorithm Design in a Fast Prototyping Environment: Towards
a Real-Time Implementation (No. 3, September/septembre
2005): R. Giroux, R. Gourdeau, R. Jr. Landry. . . . . . . . . . . . . . . 133
Investigation of an Asymmetrical Shock-Wave Boundary-Layer
Interaction in a Supersonic Planar Nozzle (No. 2, June/juin
2005): R. Benay, A. Bourgoing . . . . . . . . . . . . . . . . . . . . . . . . . 71
J
Jablonski, A.M.: Book Review: Orbital Mechanics for Engineering
Students written by Howard D. Curtis (No. 2, June/juin 2005) . . . . 87
Jameson, A., J. Alonso, S.K. Nadarajah: Adjoint-Based Sonic
Jiang, L.-Y., I. Campbell: Application of Three Combustion Models
to a Model Combustor (No. 1, March/mars 2005) . . . . . . . . . . . . . . 1
Jiang, Z., G. Lachapelle, C. Ma: Applying Fast Fourier Transform
Analysis and Data Window in Software Global Positioning
System Receivers to Mitigate Continuous Wave Interference
under Dynamic Conditions (No. 4, December/dcembre 2005). . . . 177
K
Khalid, M., A. Dujardin, P. Hennig, L. Leavitt, F. Leopold,
M. Mendenhall, S. Prince: Turbulence Model Studies to
Karpynczyk, J., D.R. Greatrix: Rocket Vehicle Design for
Small-Payload Delivery to Orbit (No. 3, September/septembre
2005) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
L
Lachapelle, G., C. Ma, Z. Jiang: Applying Fast Fourier Transform
Lachapelle, G., S. Skone, K. OKeefe: Assessing the Global
Availability and Reliability of the Mars Network, a Proposed
Global Navigation Satellite System for Mars (No. 1,
March/mars 2005). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Landry, R. Jr., A. Constantinescu, I. Ilie: Simulation of GPS and
(No. 1, March/mars 2005) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Landry, R. Jr., R. Giroux, R. Gourdeau: Inertial Navigation
Leavitt, L., F. Leopold, M. Mendenhall, S. Prince,
M. Khalid, A. Dujardin, P. Hennig: Turbulence Model Studies
to Investigate the Aerodynamic Performance of a NASA Dual

Leopold, F., M. Mendenhall, S. Prince, M. Khalid, A. Dujardin,
P. Hennig, L. Leavitt: Turbulence Model Studies to Investigate
at Supersonic Mach Numbers (No. 4, December/dcembre
2005) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Lee, D., L. Smyth, V.D. Nguyen, B. Tanguay, Y. Mbarki, M.
Deslauriers, A. Rebaine, Y. Cronier, S. Bisson, L. Del Ciotto,
C. Legare, S. Rutherford, O. Underhill, C. Mathias: Recent
Legare, C., S. Rutherford, O. Underhill, C. Mathias, D. Lee,
L. Smyth, V.D. Nguyen, B. Tanguay, Y. Mbarki,
M. Deslauriers, A. Rebaine, Y. Cronier, S. Bisson,
L. Del Ciotto: Recent Improvements to the NRC 9 m 9 m
Wind Tunnel (No. 3, September/septembre 2005) . . . . . . . . . . . . . 95
Li, G., D. Backman, N. Bellinger, G. Shi: Numerical Modeling of
Liu, H.H.T., J. Hua: Fluid Flow and Thermodynamic Analysis of a
Wing Anti-Icing System (No. 1, March/mars 2005) . . . . . . . . . . . . 35
Luchini, P., M. Quadrio, J.M. Floryan: Modification of Turbulent
M
Ma, C., Z. Jiang, G. Lachapelle: Applying Fast Fourier Transform
Mathias, C., D. Lee, L. Smyth, V.D. Nguyen, B. Tanguay, Y.
Mbarki, M. Deslauriers, A. Rebaine, Y. Cronier, S. Bisson,
L. Del Ciotto, C. Legare, S. Rutherford, O. Underhill: Recent
Mbarki, Y., M. Deslauriers, A. Rebaine, Y. Cronier, S. Bisson,
L. Del Ciotto, C. Legare, S. Rutherford, O. Underhill,
C. Mathias, D. Lee, L. Smyth, V.D. Nguyen, B. Tanguay:
Recent Improvements to the NRC 9 m 9 m Wind Tunnel
(No. 3, September/septembre 2005) . . . . . . . . . . . . . . . . . . . . . . . 95
Mendenhall, M., S. Prince, M. Khalid, A. Dujardin, P. Hennig,
L. Leavitt, F. Leopold: Turbulence Model Studies to Investigate
at Supersonic Mach Numbers (No. 4, December/dcembre
2005) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Modification of Turbulent Flow Using Distributed Transpiration
(No. 2, June/juin 2005): M. Quadrio, J.M. Floryan, P. Luchini . . . . 61
N
Nadarajah, S.K., A. Jameson, J. Alonso: Adjoint-Based Sonic
Nguyen, V.D., B. Tanguay, Y. Mbarki, M. Deslauriers,
A. Rebaine, Y. Cronier, S. Bisson, L. Del Ciotto, C. Legare,
S. Rutherford, O. Underhill, C. Mathias, D. Lee, L. Smyth:
Numerical Modeling of a Single Aluminum Sheet Containing an
Interference Fit Fastener (No. 3, September/septembre 2005):
G. Li, D. Backman, N. Bellinger, G. Shi . . . . . . . . . . . . . . . . . . 107
O
OKeefe, K., G. Lachapelle, S. Skone: Assessing the Global
March/mars 2005). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
P
Prince, S., M. Khalid, A. Dujardin, P. Hennig, L. Leavitt,
F. Leopold, M. Mendenhall: Turbulence Model Studies to
208

Q
Quadrio, M., J.M. Floryan, P. Luchini: Modification of Turbulent

R
Rebaine, A., Y. Cronier, S. Bisson, L. Del Ciotto, C. Legare,
S. Rutherford, O. Underhill, C. Mathias, D. Lee, L. Smyth,
V.D. Nguyen, B. Tanguay, Y. Mbarki, M. Deslauriers: Recent
Recent Improvements to the NRC 9 m 9 m Wind Tunnel (No. 3,
September/septembre 2005): V.D. Nguyen, B. Tanguay,
Y. Mbarki, M. Deslauriers, A. Rebaine, Y. Cronier, S. Bisson,
L. Del Ciotto, C. Legare, S. Rutherford, O. Underhill,
C. Mathias, D. Lee, L. Smyth . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Rocket Vehicle Design for Small-Payload Delivery to Orbit (No. 3,
September/septembre 2005): D.R. Greatrix, J. Karpynczyk . . . . . . 123
Rutherford, S., O. Underhill, C. Mathias, D. Lee, L. Smyth,
V.D. Nguyen, B. Tanguay, Y. Mbarki, M. Deslauriers,
A. Rebaine, Y. Cronier, S. Bisson, L. Del Ciotto, C. Legare:
S
Shi, G., G. Li, D. Backman, N. Bellinger: Numerical Modeling of
Simulation of GPS and Galileo Architectures for Anti-jamming and
Multipath Analysis (No. 1, March/mars 2005): I. Ilie,
R. Jr. Landry, A. Constantinescu. . . . . . . . . . . . . . . . . . . . . . . . . 13
Skone, S., K. OKeefe, G. Lachapelle: Assessing the Global
March/mars 2005). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Smyth, L., V.D. Nguyen, B. Tanguay, Y. Mbarki, M. Deslauriers,
A. Rebaine, Y. Cronier, S. Bisson, L. Del Ciotto, C. Legare,
S. Rutherford, O. Underhill, C. Mathias, D. Lee: Recent
Soucy, Y., H. Chesser: Force Limited Vibration Testing Applied to
the MOST Spacecraft (No. 2, June/juin 2005) . . . . . . . . . . . . . . . 47
T
Tanguay, B., Y. Mbarki, M. Deslauriers, A. Rebaine, Y. Cronier,
S. Bisson, L. Del Ciotto, C. Legare, S. Rutherford,
O. Underhill, C. Mathias, D. Lee, L. Smyth, V.D. Nguyen:
Turbulence Model Studies to Investigate the Aerodynamic
Performance of a NASA Dual Control Missile at Supersonic
Mach Numbers (No. 4, December/dcembre 2005): M. Khalid,
A. Dujardin, P. Hennig, L. Leavitt, F. Leopold, M. Mendenhall,
S. Prince . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
U
Underhill, O., C. Mathias, D. Lee, L. Smyth, V.D. Nguyen,
B. Tanguay, Y. Mbarki, M. Deslauriers, A. Rebaine,
Y. Cronier, S. Bisson, L. Del Ciotto, C. Legare, S. Rutherford:
W
Wickramasinghe, V.K.: Book Review: Stress, Strain, and Structural
Dynamics An Interactive Handbook of Formulas, Solutions,
and MATLAB Toolboxes written by Bngen Lang (No. 3,
September/septembre 2005) . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
209

Postage paid at Ottawa

Publications mail
Registration No. 40069141
Port pay Ottawa

Poste-publication
Enregistrement no 40069141
Volume 51, No. 4


JOURNAL ARONAUTIQUE ET SPATIAL DU CANADA
Turbulence Model Studies to Investigate the Aerodynamic Performance of a NASA

Dual Control Missile at Supersonic Mach Numbers
M. Khalid, A. Dujardin, P. Hennig, L. Leavitt, F. Leopold, M. Mendenhall, S. Prince . 153
Aerodynamic Forces Approximations using the Chebyshev Method for Closed-Loop
Aero-servoelasticity Studies
Alin Dorian Dinu, Ruxandra Mihaela Botez, Iulian Cotoi . . . . . . . . . . . . . . . . . . . . 167
Applying Fast Fourier Transform Analysis and Data Window in Software Global
Positioning System Receivers to Mitigate Continuous Wave Interference under
Dynamic Conditions
Z. Jiang, G. Lachapelle, C. Ma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Adjoint-Based Sonic Boom Reduction for Wing-Body Configurations in Supersonic
Flow
Siva K. Nadarajah, Antony Jameson, Juan Alonso. . . . . . . . . . . . . . . . . . . . . . . . . . 187
Canadian Aeronautics and Space Institute Institut aronautique et spatial du Canada

150 - 1750, croissant Courtwood Crescent, Ottawa, ON, Canada K2C 2B5
Tel./Tl. : 613-234-0191, Fax/Tlc. : 613-234-9039, E-mail/Courriel : casi@casi.ca
Publication Mail Registration No. 231452
ISSN 0008-2821

10.0000@pubs - casi.CA@Generic A6504037A4BF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

10.0000@pubs - casi.CA@Generic A6504037A4BF

Uploaded by

Copyright:

Available Formats

December 2005 dcembre

Volume 51, No. 4

Can. Aeronaut. Space J. Downloaded from pubs.casi.ca by UNIV MANITOBA on 06/22/15

Published by Publi par

Canadian Aeronautics and Space Journal

Can. Aeronaut. Space J. Downloaded from pubs.casi.ca by UNIV MANITOBA on 06/22/15

EDITORIAL BOARD / CONSEIL DITORIAL

Associate Editors / Directeurs

Assistant to the Editor / Adjoint au

PUBLISHED / PUBLI PAR

Journal aronautique et spatial du Canada

December 2005 dcembre

Volume 51, No. 4

Can. Aeronaut. Space J. Downloaded from pubs.casi.ca by UNIV MANITOBA on 06/22/15

CANADIAN AERONAUTICS AND SPACE JOURNAL

Turbulence Model Studies to Investigate the Aerodynamic Performance of a NASA

Published by Publi par

Canadian Aeronautics and Space Institute Institut aronautique et spatial du Canada

Can. Aeronaut. Space J. Downloaded from pubs.casi.ca by UNIV MANITOBA on 06/22/15

Vol. 51, No. 4, December 2005

Vol. 51, no 4, dcembre 2005

Turbulence Model Studies to Investigate the

M. Khalid * A. Dujardin ** P. Hennig *** L. Leavitt **** F. Leopold *****

lift coefficient (L/qS)

pitching moment M/qSl

normal force coefficient

axial force coefficient based on pressure

axial force coefficient based on friction

moment reference length (2.6 in) (1 in =

ratio of horizontal component of velocity

v/U, vz/V, v/V

ratio of normal velocity with respect to the

continued on page 154

* Institute for Aerospace Research

u/U, vx/V, u/V

Canadian Aeronautics and Space Journal

Can. Aeronaut. Space J. Downloaded from pubs.casi.ca by UNIV MANITOBA on 06/22/15

suite de la page 153

Journal aronautique et spatial du Canada

to the weight/structural and aero elastic issues. Its structural

Vol. 51, No. 4, December 2005

Can. Aeronaut. Space J. Downloaded from pubs.casi.ca by UNIV MANITOBA on 06/22/15

suite de la page 154

of the vortex formed by the front fins was of particular interest

Vol. 51, no 4, dcembre 2005

computations in the absence of the yaw angle had adopted a

Can. Aeronaut. Space J. Downloaded from pubs.casi.ca by UNIV MANITOBA on 06/22/15

Canadian Aeronautics and Space Journal

Journal aronautique et spatial du Canada

Figure 1. Dual control missile (all dimensions are in mm).

region. Strictly speaking, the assumptions underlying the

Can. Aeronaut. Space J. Downloaded from pubs.casi.ca by UNIV MANITOBA on 06/22/15

Vol. 51, No. 4, December 2005

activated that suppresses the flux calculation in the span wise

Vol. 51, no 4, dcembre 2005

STRUCTURED AND UNSTRUCTURED MESHES

Can. Aeronaut. Space J. Downloaded from pubs.casi.ca by UNIV MANITOBA on 06/22/15

Canadian Aeronautics and Space Journal

Journal aronautique et spatial du Canada

Figure 3. Side view of the computational structured grid near the

grids. The coarse grid is composed of about 700 000 nodes

computations was extended downstream from the rear control

Figure 4. General views of the fine unstructured grid.

M. Khalid * A. Dujardin P. Hennig * L. Leavitt F. Leopold *