1999-01-5621

Multidisciplinary Design Optimization of a

Transonic Commercial Transport with a
Strut-Braced Wing
F. H. Gern, J. F. Gundlach, A. Ko, A. Naghshineh-Pour, E. Sulaeman, P. -A. Tetrault,
B. Grossman, R. K. Kapania, W. H. Mason and J. A. Schetz
Virginia Polytechnic Institute and State Univ.

R. T. Haftka
University of Florida

1999 World Aviation Conference

October 19-21, 1999
San Francisco, CA

American Institute of Aeronautics

SAE International and Astronautics
400 Commonwealth Drive 370 L’Enfant Promenade, S.W.
Warrendale, PA 15096-0001 U.S.A. Washington, D.C. 20024
For permission to copy or republish, contact the American Institute of Aeronautics and Astronautics or SAE International
Published by the American Institute of Aeronautics and Astronautics (AIAA) at 1801 Alexander Bell Drive,
Suite 500, Reston, VA 22091 U.S.A., and the Society of Automotive Engineers (SAE) at 400
Commonwealth Drive, Warrendale, PA 15096 U.S.A.

Produced in the U.S.A. Non-U.S. purchasers are responsible for payment of any taxes required by their
governments.

Reproduction of copies beyond that permitted by Sections 107 and 108 of the U.S. Copyright Law without
the permission of the copyright owner is unlawful. The appearance of the ISSN code at the bottom of this
page indicates SAE’s and AIAA’s consent that copies of the paper may be made for personal or internal
use of specific clients, on condition that the copier pay the per-copy fee through the Copyright Clearance
Center, Inc., 222 Rosewood Drive, Danvers, MA 01923. This consent does not extend to other kinds of
copying such as copying for general distribution, advertising or promotional purposes, creating new
collective works, or for resale. Permission requests for these kinds of copying should be addressed to
AIAA Aeroplus Access, 4th Floor, 85 John Street, New York, NY 10038 or to the SAE Publications Group,
400 Commonwealth Drive, Warrendale, PA 15096. Users should reference the title of this conference
when reporting copying to the Copyright Clearance Center.

ISSN #0148-7191
Copyright © 1999 by SAE International and the American Institute of Aeronautics and Astronautics, Inc.
All rights reserved.

All AIAA papers are abstracted and indexed in International Aerospace Abstracts and Aerospace
Database.

All SAE papers, standards and selected books are abstracted and indexed in the Global Mobility
Database.

Copies of this paper may be purchased from:

AIAA’s document delivery service

Aeroplus Dispatch
1722 Gilbreth Road
Burlingame, California 94010-1305
Phone: (800) 662-2376 or (415) 259-6011
Fax: (415) 259-6047

or from:

SAExpress Global Document Service

c/o SAE Customer Sales and Satisfaction
400 Commonwealth Drive
Warrendale, PA 15096
Phone: (724) 776-4970
Fax: (724) 776-0790

SAE routinely stocks printed papers for a period of three years following date of publication. Quantity
reprint rates are available.

No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise,
without the prior written permission of the publishers.

Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of
SAE or AIAA. The author is solely responsible for the content of the paper. A process is available by
which discussions will be printed with the paper if it is published in SAE Transactions.
 1999-01-5621

Multidisciplinary Design Optimization of a Transonic

Commercial Transport with a Strut-Braced Wing
F. H. Gern, J. F. Gundlach, A. Ko, A. Naghshineh-Pour, E. Sulaeman, P. -A. Tetrault,
B. Grossman, R. K. Kapania, W. H. Mason and J. A. Schetz
Virginia Polytechnic Institute and State Univ.

R. T. Haftka
University of Florida

Copyright © 1999 by SAE International and the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.

ABSTRACT wing, to name only a few. This study exclusively com-

pares the strut-braced wing concept (SBW) to the cantile-
This paper details the multidisciplinary design optimiza- ver wing configuration.
tion (MDO) of a strut-braced wing aircraft and its benefits
Favorable interactions between structures, aerodynamics
relative to the cantilever wing configuration. The multidis-
and propulsion give the SBW potential for higher aerody-
ciplinary design team is subdivided into aerodynamics,
namic efficiency and lower weight than a cantilever wing
structures, aeroelasticity and synthesis of the various dis-
(Fig. 2). The strut provides bending load alleviation for
ciplines. The aerodynamic analysis consists of simple
the wing, allowing the wing thickness to be reduced for a
models for induced drag, wave drag, parasite drag and
given wing load. Reduced wing thickness decreases tran-
interference drag. The interference drag model is based
sonic wave drag and parasite drag. This favorable drag
on detailed computational fluid dynamics (CFD) analyses
reduction allows the wing to unsweep for increased
of various wing-strut intersection flows. The wing struc-
regions of natural laminar flow and promotes further wing
tural weight is partially calculated using a newly devel-
structural weight savings. Decreased overall weight,
oped wing bending material weight routine that accounts
along with increased aerodynamic efficiency permits
for the special nature of strut-braced wings. The remain-
engine size reduction.
ing components of the aircraft weight are calculated
using a combination of NASA’s Flight Optimization Sys-
tem (FLOPS) and Lockheed Martin Aeronautical System
formulas. The strut-braced wing and cantilever wing con-
figurations are optimized using Design Optimization Tools
(DOT). Offline NASTRAN aerolasticity analysis prelimi-
nary results indicate that the flutter speed is higher than
the design requirement.

INTRODUCTION

Very few recent transonic transport aircraft designs divert

from a low cantilever wing with either wing or fuselage
mounted engines. Within that arrangement, few visual
Figure 1. Conventional Cantilever Configuration
dissimilarities allow one to discern the various models
(Fig. 1). It is unlikely that large strides in performance will
This strong synergism yields significant increases in per-
be possible without a significant departure in vehicle con-
formance over the cantilever wing. A Multidisciplinary
figuration.
Design Optimization (MDO) approach is necessary to
Numerous alternative configuration concepts have been fully exploit the interdependencies of various design dis-
introduced over the years to challenge the cantilever wing ciplines. Several SBW design studies have been per-
design paradigm. These include the joined wing, formed in the past ([1]-[6]), though not with a full MDO
blended-wing-body, twin-fuselage and the strut-braced approach until quite recently ([7]-[9]).

1
This study was funded by NASA Langley with Lockheed
Martin Aeronautical Systems (LMAS) as an industrial
partner. The primary role of the LMAS interactions was to
add practical industry experience to the vehicle study.
This was achieved by calibrating the Virginia Tech MDO
code to the LMAS MDO code for 1995 and 2010 technol-
ogy level cantilever wing transports. LMAS also reviewed
aspects of the Virginia Tech design methods specific to
the strut-braced wing [9]. One of the authors worked on
location at LMAS to upgrade, calibrate and validate the
Virginia Tech MDO code before proceeding with optimi-
zations of cantilever and strut-braced wing aircraft.
Several SBW concepts have been investigated within this
project. Design studies cover wingtip engines, under-
wing engines, and fuselage-mounted engines with a T-
tail. However, emphasis of this paper is placed on the
structural aspects of the optimization procedure for fuse-
lage-mounted engine SBW configurations (Fig. 2). Since
differences in T-tail fuselage-mounted and under-wing
engine cantilever designs are small, this study uses can-
tilever optima with wing mounted engines, to make direct
comparisons with the SBW.
Figure 2. SBW with Fuselage-Mounted Engines.
DESIGN OPTIMIZATION
GENERAL ASPECTS – The Virginia Tech Truss-Braced
Wing (TBW) code models aerodynamics, structures,
weights, performance, and stability and control of both
cantilever and strut-braced wing configurations. Design
Optimization Tools (DOT) software by Vanderplatts R&D
[10] optimizes the vehicles with the method of feasible
directions. Between 15 and 22 design variables are used
in a typical optimization. These include several geometric
variables such as wing span, chords, thickness to chord
ratios, strut geometry and engine location, plus additional
variables including engine maximum thrust and average
cruising altitude. As many as 17 inequality constraints
may be used (Table 1).
There are two side constraints to bound each design vari-
able. Each design variable is scaled to have a value
between 0 and 1 at the lower and upper limits, respec-
tively. Take-off gross-weight, economic mission take-off
gross weight, and fuel weight are important examples for
possible objective functions that can be minimized.
2
Table 1. Optimization constraints
1. Aircraft Zero Fuel Weight Convergence
2. Range Calculated > Reference Range
3. Initial Cruise Rate of Climb > 500 ft/min
4. Cruise Section CLmax < 0.7
5. Fuel Weight < Fuel Capacity
6. CN Available > CN Required
7. Wing Tip Deflection < Max. Wing Tip Deflection
at Taxi Bump Condition
8. Wing Weight Convergence
9. Max. Body and Contents Weight Convergence
10. Second Segment Climb Gradient > 2.4%
11. Balanced Field Length < 11,000 ft
12. Approach Velocity < 140 kts.
13. Missed Approach Climb Gradient > 2.1%
14. Landing Distance < 11,000 ft
15. Econ. Mission Range Calculated > 4000 nmi
16. Econ. Mission Section CLmax < 0.7
17. Thrust at Altitude > Drag at Altitude
The MDO code architecture is configured in a modular
way such that the analysis consists of subroutines repre-
senting various design disciplines. The primary analysis
modules include: aerodynamics, wing bending material
weight, total aircraft weight, stability and control, propul-
sion, flight performance and field performance (Fig. 3).
Numerous differences between the analysis details of
cantilever and SBW configurations are present in the
design code, as is necessary for such dissimilar vehicles.
The primary difference is in the analysis of the wing
bending material weight, as discussed in the structures
section. The strut has parasite drag and interference drag
at its intersections with fuselage and wing. Some geome-
try differences are justified, such as setting the minimum
root chord for the cantilever wing to 52 feet to make room
for wing-mounted landing gear and kick spar.
The SBW, without need for double taper, has the chord
linearly interpolated from root to tip. The SBW has a high
wing and fuselage mounted gear. It is important to note
that, even though the external geometry of the fuselage
for all cases is identical, the fuselage weights will gener-
ally be different.

Updated Design Variables
Propulsion
SFC
Field
Performance
Stability and
Control
Figure 3. Description of the MDO Process
MISSION PROFILE – The primary mission of interest is
a 325-passenger, 7500 nautical mile range, Mach 0.85
transport with a 500 nautical mile fuel reserve (Fig. 4).
Range effects on take-off gross weight and required fuel
weight are investigated. A minimum fuel design is also
considered.
Several technology groups distinguish the 1995 and 2010
technology level aircraft. A 1995 technology aircraft rep-
resents an all-metallic benchmark similar to the Boeing
777. The other aerodynamics grouping includes the
effects of riblets on the fuselage and nacelles, supercriti-
cal airfoils, active load management for induced drag
reduction and all moving control surfaces. Systems tech-
nologies include integrated modular flight controls, fly-by-
light and power-by-light, simple high-lift devices, and
advanced flight management systems. Airframe technol-
ogies represent weight savings from composite wing and
tails and integrally stiffened fuselage skins. The propul-
sion technology is reflected in reduced specific fuel con-
sumption.
Mach 0.85 Cruise
Climb 140 Knot
Approach
Speed
11,000 ft 7500 Nmi Range 11,000 ft
T/O Field Length LDG Field Length
Figure 4. Mission Profile
AERODYNAMICS – Numerous iterations of both the Vir-
ginia Tech TBW code and Lockheed’s version of NASA’s
Flight Optimization System (FLOPS) [16] were made so
that drag polars produced by each code are consistent at
reference design conditions. The drag components con-
Baseline
Design
Induced
Initial Design Variables Drag
Geometry
Definition Friction and
Form Drag
Structural
Aerodynamics Wave Drag
Optimization/
Weight
L/D Interference
Weight
Drag
Range Offline CFD
Performance
Analysis
Objective Function,
Constraints
Optimizer
sidered in the Virginia Tech TBW code are parasite,
induced, interference and wave drag. Unless specified
otherwise, the drag model is identical to previous Virginia
Tech SBW studies [8]. A detailed description of the drag
calculations can be found in [11].
Parasite Drag – To calculate the parasite drag, form fac-
tors are applied to the equivalent flat plate skin friction
drag of all exposed surfaces on the aircraft. The amounts
of laminar flow on the wing and tails are estimated by
interpolating Reynolds number vs. sweep data for F-14
and 757 glove experiments. Fuselage, nacelles, and
pylon transition locations are estimated by an input tran-
sition Reynolds number. Laminar and turbulent flat-plate
skin friction form factors are calculated with LMAS formu-
las in the Virginia Tech MDO tool. LMAS form factors for
wing, tails, fuselage, and nacelles are applied to the skin
friction drag to obtain the parasite drag.
Induced Drag – The induced drag module uses a dis-
crete vortex method to calculate the induced drag in the
Trefftz plane [8]. Given an arbitrary, non-coplanar wing/
truss configuration, it provides the optimum load distribu-
Mach 0.85
tion corresponding to the minimum induced drag. This
load distribution is passed to the wing sizing subroutine.
An additional lift-dependent parasite drag component
was added to correlate with LMAS drag polars at off-
design conditions.
500 Nmi Reserve
Wave Drag – The wave drag is approximated with the
Korn equation, modified to include sweep using simple
sweep theory [7], [8]. This model estimates the drag
divergence Mach number as a function of airfoil technol-
ogy factor, thickness to chord ratio, section lift coefficient,
and sweep angle.
The airfoil technology factor was selected by Lockheed to
agree with the LMAS wave drag. Finally, the wave drag
coefficient of a wing strip is calculated from the critical
3
Mach number. The total wave drag is found by integrating
the wave drag of the strips along the wing.
Interference Drag – The benefits of a strut-braced wing
configuration are accompanied by a potential interfer-
ence drag penalty at the junction of the strut with the
fuselage and the wing. The interference drag between
the wing-fuselage and strut-fuselage intersections are
estimated using Hoerner equations based on subsonic
wind tunnel tests [12].
The drag of wing-strut junctions can be important in tran-
sonic flow because of the presence of shock waves and
separated flow regions. In order to alleviate the problem
associated with a sharp wing-strut angle, the strut
employed here is given the shape of an arch and inter-
sects the wing perpendicularly. Analyses for an arch
radius ranging from 1 ft to 4 ft were performed with Com-
putational Fluid Dynamics (CFD) tools. Unstructured
grids were obtained with the advancing-front methodol-
ogy implemented in the code VGRIDns [13], [ 14]. The
Euler equations were solved using the CFD code USM3D
[14], [15] at the cruise Mach number of 0.85.
A very convenient way to extract the interference drag
penalty from a CFD calculation consists in subtracting
the drag of the wing alone from the drag of the strut-
braced wing design obtained with CFD. The resulting
number is a DCD penalty associated with the presence of
the strut. As the arch radius is increased, the drag pen-
alty decreases almost exponentially. From these results,
a curve fit is produced and used in the present analysis to
account for the drag of the wing-strut junction.
The drag polars output from the Virginia Tech MDO tool
and LMAS modified FLOPS agree within 1% on average
for cantilever wing designs.
STRUCTURES – Due to the unconventional nature of
the proposed concept, commonly available weight calcu-
lation models for transport aircraft (such as the NASA
Langley developed FLOPS) are not accurate enough. A
special bending weight calculation procedure was thus
developed, taking into account the influence of the strut
upon the structural wing design. In addition to the strut
design, a vertical strut offset was considered as to
achieve a significant reduction in wing/strut interference
drag.
Load Cases – To determine the bending material weight
of the strut-braced wing, two maneuver load conditions
(2.5g maneuver, -1.0g pushover) and a taxi bump (-2.0g)
are considered to be design critical. For the -1.0g push-
over and for the -2.0g taxi bump, the strut is not active
and the wing acts like a cantilever beam. Since the strut
is not supporting the wing in these cases, very high
deflections of the wing are expected for the -2.0g taxi
bump. As a result, an optimization procedure is imple-
mented to distribute the bending material to prevent wing
ground strikes. To maximize the beneficial influence of
the strut upon the wing structure, strut force and span-
4
wise position of the wing-strut intersection are optimized
by the MDO code for the 2.5g maneuver load case.
In order to attain acceptable aerodynamic characteristics
of the strut, an airfoil cross section is considered. The
strut is designed the way that it will not carry aerody-
namic forces during the cruise condition.
Structural Assumptions – Preliminary studies have
shown buckling of the strut under the –1.0g load condi-
tion to be the critical structural design requirement in the
single-strut configuration, resulting in high strut weights
[8]. To address this issue, an innovative design strategy
employs a telescoping sleeve mechanism to allow the
strut to be inactive during negative g maneuvers and
active during positive g maneuvers. Thus, under the –
1.0g case, the wing acts like a cantilever beam and for
the positive g maneuvers, the wing is a strut-braced
beam.
Even more wing weight reduction can be obtained by
optimizing the strut force and wing-strut junction location.
On a typical optimum single-strut design, this means that
the strut would first engage in tension at some positive
load factor. This can be achieved by assuming a slack in
the wing-strut mechanism. The optimum strut force at
2.5g is different from the strut force that would be
obtained at 2.5g if the strut were engaged for all positive
values of the load factor. Therefore, the slack load factor
is defined as the load factor at which the strut engages
for the first time. It is important to have the slack load fac-
tor always positive, otherwise the strut would be pre-
loaded at the jig shape to achieve the optimum strut
force.
Double Plate Model – For calculating the wing-bending
weight of single strut configurations, a piecewise linear
beam model, representing the wing structure as an ideal-
ized double plate model, was used (Fig. 5).
t
d
Cb
Figure 5. Double plate model for bending weight
calculation
This model is made of upper and lower skin panels,
which are assumed to carry the bending moment. The
double-plate model offers the possibility to extract the
material thickness distribution by a closed-form equation.
The cross-sectional moment of inertia of the wing-box
can be expressed as:
t ( y ) cb ( y ) d 2 ( y )
I ( y) =
2 (1)
where t(y) is the wing skin thickness cb(y) is the wing- Wing Neutral Axis
box chord, and d(y) is the wing airfoil thickness. To obtain
the bending material weight, the corresponding bending
stress in the wing is calculated from: Wing Lower Surface

M ( y )d ( y )
σ max = Structural Strut Offset
2I ( y) (2) Aerodynamic
Strut Offset
where σmax denotes the maximum stress, M(y) is the
bending moment of the wing, and I(y) denotes the cross-
Horizontal Strut Force
sectional moment of inertia.
If the wing is designed according to the fully-stressed cri-
terion, the allowable stress σall can be substituted into
Eq. (2) for σmax. Substituting I(y) into equation (2), the Vertical Strut Force
wing panel thickness can be specified as: Figure 6. Vertical strut offset and applied loads

M ( y) Figure 7 depicts the bending moment distributions on the

t( y) =
cb ( y )d ( y )σ all wing for the design critical load cases of the fuselage
(3) mounted engine SBW design. Due to the vertical strut
This skin thickness is modified by the results obtained offset, an additional bending moment is induced into the
from the tip displacement constraint optimization. The wing at the wing/strut breakpoint, leading to a discontinu-
bending material weight of the half-wing therefore is: ity in the bending moment distribution. Since the strut is
inactive in compression, the bending moment distribu-
tions for the -1.0g pushover as well as for the 2.0g taxi
Wwb = 2 ∫
bs / 2
t ( y )cb ( y ) ρdy bump do not exhibit this discontinuity.
0 (4)
bs = b cos .Λ
2.5E+07
where b
iss the structural span with
2.0E+07 2.5G Maneuver
-1.0G Pushover
Vertical Strut Offset – To reduce the wing/strut interfer-
Bending Moment (Ft-Lb)

1.5E+07 -2.0G Taxi Bump

ence drag, a vertical offset between strut and wing is
1.0E+07
implemented. The vertical offset member is designed for
a combined bending/tension loading. In this context, the 5.0E+06
horizontal component of the strut force is of special con-
0.0E+00
cern (Fig.6). Since this horizontal force results in a con- 0 20 40 60 80 100 120
siderable bending load on the offset piece, its weight -5.0E+06

increases dramatically with increasing strut force and off- -1.0E+07

set length.
As a result, it is imperative to employ MDO tools to obtain
optimum values for vertical offset, strut force, and span-
wise wing/strut breakpoint. By this way, it is possible to
trade off the two contrary design requirements: (i) a
reduced offset length to reduce strut loading, (ii) an
increased offset length to reduce the wing/strut interfer-
ence drag. After a complete design optimization with the
vertical strut offset as an active design variable, the influ-
ence of the offset weight on the total strut weight
becomes comparably small. For the wing bending weight
and especially for the TOGW it is almost immaterial.
-1.5E+07
-2.0E+07
Wing Half Span (Ft)
Figure 7. Bending moment distributions for the design
critical load cases of the fuselage mounted
engine SBW
AEROELASTICITY
Hexagonal Wing-Box Model – Although the double plate
model renders very accurate estimates for the wing
bending material weight, it is not suitable for calculation
of the wing-box torsional stiffness. Nevertheless, tor-
sional stiffness becomes essential when calculating wing
twist and flexible wing spanload, as well as for the incor-
poration of aeroelastic constraints and design variables
into the MDO optimization.

5
Therefore, a hexagonal wing-box model provided by Computational Aeroelasticity – Beyond rendering accu-
LMAS was implemented into the code (Fig. 8). In contrast rate quantities for bending and torsional stiffness, the
to the double plate model, the hexagonal wing-box allows hexagonal wing-box is suitable to create input data and
computation of bending and torsional stiffness with a high realistic sizing for detailed finite element analyses. Cur-
degree of accuracy. Based upon Lockheed Martin’s expe- rently, the panel thickness distributions from the double
rience in wing sizing, the wing-box geometry varies in the plate model are used to create a hexagonal wing-box
spanwise direction with optimized area and thickness according to the spanwise variation of the respective
ratios for spar webs, spar caps, stringers, and skins. Fur- cross sectional data.
thermore, minimum gauges and maximum stress cutoffs
To obtain the spanwise distribution of the moments of
can be accurately applied.
inertia, the overall cross sectional area of stringers, spar
Hexagonal Wing-Bo
caps, and skins are matched with the respective cross
sections of the double plate model. With this data, a
L detailed finite element model of the structural wing-box is
0.04
computed and analyzed using NASTRAN. It consists of
0.03 630 grid points, 1239 rod elements, and 3232 plate ele-
Airfoil ments. The structural material is an equivalent isotropic
0.02

0.01
M
z/c 0
composite model same as the one used in wing-bending
-0.01 weight calculations. The fuel load is distributed into 47
-0.02

-0.03
mass elements. For unsteady aerodynamics, the Doublet
-0.04 Lattice method with compressibility correction for sub-
0 0.2 0.4 0.6 0.8 1
x/c sonic flight is employed. Aerodynamic loads are simu-
Center of Gravity lated using 300 box elements.
N·g·m
To calculate the flutter speed, 10 structural vibration
Shear Center (Elastic Axis) modes are considered. In this model, no strut structure is
included. Figure 9 illustrates the flutter boundary
Aerodynamic Center obtained using the PK method in terms of the true air
Figure 8. Hexagonal wing-box and applied sectional speed. At each altitude, flutter is related to the fundamen-
forces and moments tal wing bending and torsional modes. However, at
30,000 ft flight level, flutter occurs due to coupling of the
yawing mode with the first torsional mode.
50
The results should be corrected using a more accurate
transonic unsteady aerodynamics modeling to simulate
45
the transonic dip effect. Also, aeroelastic constraints
must be included into the optimization.
40

WEIGHTS – The aircraft weight is calculated by incorpo-

35
rating several different methods. The majority of the
weights equations come from NASA Langley’s Flight
Altitude (103 ft)

30
Optimization System (FLOPS) [16]. Many of the FLOPS
25
equations were replaced with those suggested by LMAS.
The LMAS and original FLOPS methods do not have the
20
option to analyze the strut-braced wing with the desired
1.15 Vd Flight Envelope
fidelity. Therefore, the bending material weight from the
15
FLOPS equations is replaced by the bending material
Isolated Wing, full fuel
weight obtained from the piecewise linear wing load mod-
Wing-Strut, full fuel
10 els described above [17].
Wing-Strut, zero fuel

Isolated Wing, zero fuel

The wing bending weight is calculated using the panel
5 thickness results or hexagonal wing-box cross sections
from the piecewise linear beam model for the different
0
load cases (Fig. 10). The overall panel thickness distribu-
0 200 400 600 800 1000 1200 1400 1600 1800
tion of the wing is obtained by considering the highest
Flutter Speed (fps) value of the panel thickness or cross section at each
spanwise position (envelope). To account for sudden
Figure 9. Flutter boundary vs. altitude for different flight
changes in the material distribution, an additional 1%
conditions of the fuselage mounted strut-
weight penalty is applied.
braced wing configuration

6
Before linking the wing weight module to the MDO code,
it has been validated using the 747-100 wing. The results
obtained from the double plate model as well as the hex-
agonal wing-box show good agreement with the actual
747-100 and with the results obtained from FLOPS [16]
andTorenbeek [18].
0.7
tip constraint
2g taxi bump
0.6
Panel thickness in
-1g maneuver
0.5 2.5g maneuver
0.4
0.3
0.2
0.1
0 engine failure [11].
0 20 40 60 80 100
Wing half span ft
Figure 10. Panel thickness distributions for the different
load cases (fuselage mounted engine
configuration)
The total weights for the different components (strut, off-
set, wing) are obtained using the FLOPS equations.
Here, the wing bending material and strut tension
weights are being multiplied by a technology factor to
account for the weight reduction achieved by the employ-
ment of composite materials by the year 2010.
After computation of the load carrying weights, a 10%
non-optimum factor is applied to account for manufactur-
ing constraints. The total wing weight is calculated using
the FLOPS equations with the overall load carrying
weight, i.e. wing, strut, and offset. The total weights of the
different components are determined according to the
ratio of their contributions to the load carrying weight.
LMAS provided a weight estimate for the telescoping
sleeve mechanism based on landing gear component
data. Weights calculated in the Virginia Tech transport
optimization code are identical to FLOPS with the excep-
tion of nacelle, thrust reverser, passenger service, land-
ing gear, wing, fuselage and tail weights. The above
weights are now calculated from proprietary LMAS for-
mulas. Weight technology factors are applied to major
structural components and systems to reflect weight sav-
ings due to advances in technology levels from compos-
ite materials, advanced electronics and other
technologies described above.
Some aircraft weights are implicit functions, and internal
iteration loops are typically required for convergence.
However, utilizing the optimizer to converge the zero fuel
weight of the aircraft showed to be more efficient by pro-
viding smoother gradients. DOT also selects the fuel
weight so that the range constraint is not violated. Other
weights such as the maximum body and contents weight
and wing weight converge efficiently with the lagging vari-
able method [10].
STABILITY AND CONTROL – The horizontal and verti-
cal tail areas are first calculated with a tail volume coeffi-
cient sizing method. The tail volume coefficients were
determined based on Lockheed statistical data. A vertical
tail sizing routine was developed to account for the one
engine inoperative condition [8], [17]. The engine-out
constraint is met by constraining the maximum available
yawing moment coefficient to be greater than the
required yawing moment coefficient. As specified by FAR
requirements, the aircraft must be capable of maintaining
straight flight at 1.2 times the stalling speed with the
operable engine at its maximum available thrust. The lat-
eral force of the vertical tail provides most of the yawing
moment required to maintain straight flight after an
120
The maximum available yawing moment coefficient is
obtained at an equilibrium flight condition with a given
bank angle and a given maximum rudder deflection. FAR
25.149 limits the maximum bank angle to 5°, and some
sideslip angle is allowed. The stability and control deriva-
tives are calculated using empirical methods of DATCOM
as modified by Grasmeyer [8], [20]. In order to allow a 5°
aileron deflection margin for maneuvering, the calculated
deflection must be less than 20-25°. The calculated avail-
able yawing moment coefficient is constrained in the opti-
mization problem to be greater than the required yawing
moment coefficient. If the yawing moment constraint is
violated, a vertical tail area scaling factor is applied by the
optimizer.
PROPULSION – A GE-90 class high-bypass ratio turbo-
fan engine is used for this design study. An engine deck
was obtained from LMAS, and appropriate curves for
specific fuel consumption and maximum thrust as a func-
tion of altitude and Mach number were found through
regression analysis. The general forms of the equations
are identical to those found in Mattingly [21] for high-
bypass ratio turbofan engines, but the coefficients and
exponents are modified.
The engine size is determined by the maximum thrust
required to meet several constraints. These constraints
are thrust at average cruise altitude, available rate of
climb at initial cruise altitude, balanced field length, sec-
ond segment climb gradient, and missed approach climb
gradient. The dimensions of the engine nacelles vary as
the square root of required thrust, and the engine weight
is assumed to be linearly proportional to the engine
thrust. The specific fuel consumption model is indepen-
dent of engine scale. A specific fuel consumption tech-
nology factor is applied to reflect advances in engine
technology.
7
Table 2. Parametric properties of aircraft designs for Table 3. Minimum Fuel Optimum Designs
minimum take-off gross weight (TOGW)
Cantilever SBW
Cantilever SBW Wing
Wing Span (ft) 256.2 262.3
Span (ft) 223.2 227.0 Sw (ft^2) 5800 4694
Sw (ft2) 5120 4233 AR 11.32 14.65
AR 9.73 12.17 Root t/c 13.06% 12.37%
Root t/c 14.50% 14.28% Tip t/c 5.31E-02 5.29%
Tip t/c 7.80% 6.15% Wing Λ1/4 (deg) 32.3 28.3
Wing Λ1/4 (deg) 33.3 29.9 Strut Λ1/4 (deg) 21.2
Strut Λ1/4 (deg) 20.1 η Strut 66.6%
η Strut 68.9% Max Thrust (lbs) 70919 57129
η Engine 37.0% Cruise Altitude (ft) 43826 42248
Max Thrust (lbs) 75133 59572 L/D 26.13 29.08
Cruise Altitude (ft) 41160 40322 Wing Wt. (lbs) 89373 86260
L/D 23.34 25.40 Bending Matl (lbs) 74846 68543
Wing Wt. (lbs) 63774 60745 Fuel Wt. (lbs) 176646 150147
Bending Matl (lbs) 48076 43326 TOGW (lbs) 554963 509881
Fuel Wt. (lbs) 184948 159883 % TOGW Improve- 8.1%
TOGW (lbs) 535643 492332 ment
% TOGW Improv. 8.1% % Fuel Improvement 15.0%
% Fuel Improv. 13.6% Section Cl Limit ACTIVE ACTIVE
% Thrust Reduction 20.7% 2nd Segment Climb ACTIVE
Section Cl Limit ACTIVE ACTIVE Balanced Field Length ACTIVE
2nd Segment Climb ACTIVE ACTIVE
Balanced Field ACTIVE Take-off and landing performance utilizes methods found
Length in Roskam and Lan [22]. The field performance subrou-
Engine Out ACTIVE tine calculates the second segment climb gradient, bal-
anced field length, missed approach climb gradient, and
the landing distance. All calculations are done for hot day
PERFORMANCE – The range is calculated by the
conditions at sea level. Sample drag polars for the aircraft
Breguet range equation [11]. The L/D ratio, flight velocity,
at take-off and landing were provided by LMAS [11].
and specific fuel consumption are determined for the
Trends are the same for both SBW and cantilever config-
average cruising altitude and Mach number. The initial
urations. The actual drag polars use correction factors
weight is 95.6% of the take-off gross weight to account
based on total aircraft wetted area and wing aspect ratio.
for fuel burned during climb to the initial cruise altitude. A
The second segment climb gradient is the ratio of rate of
reserve range of 500 nautical miles allows for emergency
climb to the forward velocity at full throttle while one
airport re-routing, extra loiter time while waiting for land-
engine is inoperative and the gear is retracted.
ing clearance at the end of a maximum range mission
and strong headwinds. Roskam and Lan methods are also used to determine
the landing distance [22]. Three legs are defined: the air
distance from clearing the 50-foot object to the point of
Fuselage-Engine wheel touchdown including the flare distance, the free roll
SBW distance between touch-down and application of brakes,
and finally, the distance covered while braking. The lift
coefficient on landing approach is the minimum CL asso-
ciated with either V=1.3Vstall or the CL to meet the tail
scrape requirement. The drag coefficient is calculated
with gear down.
Cantilever Wing
The missed approach climb gradient is calculated in the
same way as the second segment climb gradient with a
few exceptions. First, the weight of the aircraft at landing
is assumed to be 73% of the take-off gross weight as
specified by LMAS. Second, all engines are operational.
Figure 11. Minimum TOGW Designs Third, a landing drag polar distinct from the take-off drag
polar is used. In the present study, the FAR minimum
missed approach climb gradient constraint is never vio-
lated.

8
OPTIMIZATION RESULTS
Ta ke -Off Gros s W eight vs. R a n g e

MIMIMUM TAKE-OFF GROSS WEIGHT – Table 2 shows 900000

the parametric results for TOGW minimization and Fig.
11 gives an impression of the geometric differences of 800000
the investigated aircraft designs. Note that the cantilever

Take-Off Gross Weight (lbs.)

wing has a trailing edge break to permit landing gear 700000
stowage. A comparison of the cantilever and SBW Conventional

designs shows that in general, the SBW aircraft have less 600000

wing area, higher aspect ratio and a reduced wing sweep SBW
compared to their cantilever counterparts. 500000

400000
MINIMUM FUEL CONSUMPTION – Fuel burn is likely to
be an increasingly important factor in aircraft design from
300000
two perspectives. First, as the Earth’s petroleum 4000 6000 8000 10000 12000 14000
resources are depleted, the cost of aviation fuel will rise. Ra nge
Any reduction in fuel demand will be welcome if the fuel
price becomes a larger part of transport life cycle cost. Figure 12. Effect of range on take-off gross weight
Second, strict emissions regulations stemming from envi-
ronmental concerns will limit the amount of pollutant dis-
charge permitted by an aircraft. Beyond engine design, Fue l W e ight vs. Ra nge
reducing the overall amount of fuel consumed for a given
flight profile by improved configuration design will also 450000

reduce the total amount of emissions. Table 3 shows the 400000

minimum fuel weight results.
350000

MINIMUM TOGW VS. MINIMUM FUEL

Fuel Weight (lbs)
300000
CONSUMPTION – For minimum TOGW and minimum Conventional
250000
fuel cases, the SBW is superior for the selected objective
SB W
functions. While the SBW has an 8.1% decrease in 200000
TOGW, the savings in fuel consumption are even more
150000
impressive. A SBW has a 13.6% lower fuel burn than a
cantilever configuration when optimized for minimum 100000

TOGW, and a 15% lower fuel weight when both are opti-
mized for minimum fuel weight.
The minimum-fuel-SBW has a higher wingspan to
increase the L/D and fly at higher altitudes. The mini-
mum-fuel-SBW TOGW is 8.1% lower than an equivalent
cantilever design, and 3.6% higher than a minimum-
TOGW-SBW. The SBW L/D increases from 25.4 to 29.1
going from the minimum-TOGW to the minimum-fuel
case, and from 21.7 to 26.1 for the cantilever configura-
tion. This improved aerodynamic efficiency is achieved by
increasing the wing span, and comes at a cost in struc-
tural weight.
Airport noise pollution can limit the types of aircraft per-
mitted to use certain urban airfields and impose opera-
tional restrictions on those that do. Minimizing engine
size can also be expected to reduce the noise generated
if the engine is of similar design. Minimum TOGW SBW
engine thrust is reduced by 20.7% over the equivalent
cantilever design, probably reducing airport noise pollu-
tion by a similar amount.
50000
4000 6000 8000 10000 12000 14000
Ra nge
Figure 13. Effect of range on fuel weight
RANGE EFFECTS – The SBW becomes increasingly
desirable as the design range increases. Figures 12 and
13 show the effects of range on TOGW and fuel weight.
The TOGW reduction relative to the cantilever configura-
tion steadily improves from 5.3% at a 4,000 nautical mile
range up to 10.9% at 12,000 nautical miles. The fuel
weight savings fluctuates within about 11-16%, generally
improving with increasing design range. These results
are for minimum TOGW designs, however greater fuel
burn improvements are expected for SBW aircraft opti-
mized for minimum fuel weight. Maximum fuel weight is
set at 400,000 pounds.
At 12,000 nautical miles an aircraft can reach any desti-
nation on Earth. The SBW maximum range is 13,099
nautical miles at the maximum fuel weight, whereas the
cantilever configuration can only reach 11,998 nautical
miles, or the SBW has 8.4% greater maximum range.
Therefore, the SBW can either have a reduced fuel
weight for a given range or an increased range for a given
fuel weight relative to the cantilever configuration.

9
CONCLUSIONS
Virginia Tech transport studies have shown the potential
of the SBW over the traditional cantilever configuration.
After much added realism by a major airframe manufac-
turer, the MDO analysis shows that the SBW still demon-
strates major improvements over the cantilever wing
configuration. A significant reduction in TOGW was
found, but the greatest virtue of the SBW is the improved
fuel consumption and smaller engine size. These results
indicate that the SBW will cost less, limit pollutant dis-
charge and reduce noise pollution for urban airports.
Advantages of the SBW increase with range, suggesting
that this configuration may be ideal for larger, long-range
transports.
The special design of the strut-braced wing necessitated
the development of a wing sizing module suitable to fully
exploit the benefits of this structural configuration. After
validation with existing aircraft like the 747-100, the mod-
ule was used for wing sizing and structural weight com-
putation of the SBW. The great potential of the outlined
wing sizing procedure lies in the consideration of jig twist,
strut moment, wing flexibility, and flexible wing spanload.
Consideration of the actual in-flight maneuver loads not
only increases the accuracy in wing sizing but also gives
the potential for further weight savings. This is especially
important within a multidisciplinary design environment
where due to synergistic interaction even small weight
savings in one component are very likely to result in fur-
ther weight reductions for other components.
Further benefits of the SBW are expected to become
apparent in future studies. The cooperative relationship
with LMAS focused on adding realism to the SBW design
effort for direct comparisons with the cantilever design.
Realism often takes the form of weight penalties and
expanded performance analysis, which inevitably
detracts from SBW theoretical potential. Presently, efforts
are underway to identify technologies and strut/truss
arrangements to exploit the strengths of the strut. Limit-
ing the SBW design arrangements so that the aircraft
takes the appearance of a cantilever wing may not be the
most appropriate approach to realize the full potential of
the SBW.
The SBW is likely to have a more favorable reaction from
the public than other configurations, especially for those
who suffer from a fear of flying. Affirmative passenger
and aircrew acceptance is probable because other than
the addition of a visually innocuous strut and a high wing,
there is little to distinguish the SBW from the existing air-
liner fleet. Radical appearances of the blended-wing-
body, joined wing, or other candidate configurations may
cause apprehension in many flying patrons.
10
ACKNOWLEDGMENTS
This project is funded by NASA Langley Research Grant
NAG 1-1852. Part of the work was done under subcon-
tract from Lockheed Martin Aeronautical Systems in
Marietta, Georgia. NASA deserves much credit for hav-
ing the vision to pursue bold yet promising technologies
with the hope of revolutionizing air transportation. Lock-
heed Martin Aeronautical Systems provided valuable
contributions in data, design methods and advice.
REFERENCES
1. Pfenninger, W., “Design Considerations of Large
Subsonic Long Range Transport Airplanes with Low
Drag Boundary Layer Suction,” Northrop Aircraft,
Inc., Report NAI-54-800 (BLC-67), November 1954.
2. Gunston, B., Giants of the Sky: The Biggest
Aeroplanes of All Time, Patrick Stephens Limited,
Wellingborough, UK, pp. 240-250.
3. Kulfan, R.M., and Vachal, J.D., “Wing Planform
Geometry Effects on Large Subsonic Military Trans-
port Airplanes,” Boeing Commercial Airplane Com-
pany, AFFDL-TR-78-16, February 1978.
4. Jobe, C.E., Kulfan, R.M., and Vachal, J.D., “Wing
Planforms for Large Military Transports,” AIAA-78-
1470, 1978.
5. Turriziani, R.V., Lovell, W.A., Martin, G.L., Price, J.E.,
Swanson, E.E., and Washburn, G.F., “Preliminary
Design Characteristics of a Subsonic Business Jet
Consept Employing an Aspect Ratio 25 Strut-Braced
Wing,” NASA CR-159361, October 1980.
6. Smith, P.M., DeYoung, J., Lovell, W.A., Price, J.E.,
and Washburn, G.F., “A Study of High-Altitude
Manned Research Aircraft Employing Strut-Braced
Wings of High-Aspect Ratio,” NASA CR-159262,
February, 1981.
7. Grasmeyer, J.M., Naghshineh_Pour, A., Tetrault, P.-
A., Grossman, B., Haftka, R.T., Kapania, R.K.,
Mason, W.H., Schetz, J.A., Multidisciplinary Design
Optimization of a Strut-Braced Wing Aircraft with Tip-
Mounted Engines, MAD 98-01-01, 1998.
8. Grasmeyer, J.M., Multidisciplinary Design Optimiza-
tion of a Strut-Braced Wing Aircraft, MS Thesis, Vir-
ginia Polytechnic Institute & State University, April
1998.
9. Martin, K.C., Kopec, B.A., “A Structural and Aerody-
namic Investigation of a Strut-Braced Wing Transport
Aircraft Concept”, NAS1-96014, November 1998.
10. Vanderplaats Research & Development, Inc., DOT
User’s Manual, Version 4.20, Colorado Springs, CO,
1995.
11. Gundlach, J.F., Multidisciplinary Design Optimization
and Industry Review of a 2010 Strut-Braced Wing
Transonic Transport, MAD 99-06-03, 1999
12. Hoerner, S.F., Fluid Dynamic Drag: Practical Informa-
tion on Aerodynamic Drag and Hydrodynamic Resis-
tance, published by Mrs. Hoerner, 1965. Current
address: P.O. Box 65283, Vancouver, WA 98665.
13. Pirzadeh, S., “Structured Background Grids for Gen-
eration of Unstructured Grids by Advancing-Front
Method”, AIAA Journal, Vol. 31, February 1993, pp.
257-265.
14. Frink, N.T., Pirzadeh, S., and Parikh, P., “An Unstruc-
tured-Grid Software System for Solving Complex
Aerodynamic Problems”, NASA CP-3291, May 1995.
15. Frink, N.T., Parikh, P., and Pirzadeh, S., “A Fast
Upwind Solver for the Euler Equations on Three-
Dimensional Unstructured Meshes”, AIAA-91-0102,
1991.
16. McCullers, L.A., FLOPS User’s Guide, Release 5.81.
Text file included with the FLOPS code.
17. Naghshineh-Pour, A.H., Kapania, R., Haftka, R., Pre-
liminary Structural Analysis of a Strut-Braced Wing,
VPI-AOE-256, June 1998.
18. Torenbeek, E., "Development and Application of a
Comprehensive, Design Sensitive Weight Prediction
Method for Wing Structures of Transport Category
Aircraft," Delft University of Technology, Report LR-
693, Sept. 1992.
19. Roskam, J., Methods for Estimating Stability and
Control Derivatives of Conventional Subsonic Air-
planes, Roskam Aviation and Engineering Corp.,
Lawrence, KS, 1971.
20. Grasmeyer, J.M., Stability and Control Derivative
Estimation and Engine-Out Analysis, VPI-AOE-254,
January 1998.
21. Mattingly, J.D., Heiser, W.H., and Daley, D.H., Aircraft
Engine Design, AIAA, Washington, D.C., 1987.
22. Roskam, J., and Lan, C.-T. E., Airplane Aerodynam-
ics and Performance, DARCorporation, Lawrence,
KS, 1997.

CONTACT

Dr. Frank H. Gern

Multidisciplinary Analysis and Design (MAD) Center for
Advanced Vehicles
Department of Aerospace and Ocean Engineering
Virginia Polytechnic Institute and State University
Blacksburg, VA 24061-0203, USA
Phone: (540) 231-4860
Fax: (540) 231-9632
Email: fgern@aoe.vt.edu
http://www.aoe.vt.edu

11