0 Bewertungen0% fanden dieses Dokument nützlich (0 Abstimmungen)

72 Ansichten13 Seiten© Attribution Non-Commercial (BY-NC)

PDF, TXT oder online auf Scribd lesen

Attribution Non-Commercial (BY-NC)

Als PDF, TXT **herunterladen** oder online auf Scribd lesen

0 Bewertungen0% fanden dieses Dokument nützlich (0 Abstimmungen)

72 Ansichten13 SeitenAttribution Non-Commercial (BY-NC)

Als PDF, TXT **herunterladen** oder online auf Scribd lesen

Sie sind auf Seite 1von 13

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

October 19-21, 1999

San Francisco, CA

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.

Aeroplus Dispatch

1722 Gilbreth Road

Burlingame, California 94010-1305

Phone: (800) 662-2376 or (415) 259-6011

Fax: (415) 259-6047

or from:

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

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.

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

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 Table 1. Optimization constraints

partner. The primary role of the LMAS interactions was to

1. Aircraft Zero Fuel Weight Convergence

add practical industry experience to the vehicle study.

This was achieved by calibrating the Virginia Tech MDO 2. Range Calculated > Reference Range

code to the LMAS MDO code for 1995 and 2010 technol- 3. Initial Cruise Rate of Climb > 500 ft/min

ogy level cantilever wing transports. LMAS also reviewed

aspects of the Virginia Tech design methods specific to 4. Cruise Section CLmax < 0.7

the strut-braced wing [9]. One of the authors worked on 5. Fuel Weight < Fuel Capacity

location at LMAS to upgrade, calibrate and validate the

Virginia Tech MDO code before proceeding with optimi- 6. CN Available > CN Required

zations of cantilever and strut-braced wing aircraft. 7. Wing Tip Deflection < Max. Wing Tip Deflection

at Taxi Bump Condition

Several SBW concepts have been investigated within this

project. Design studies cover wingtip engines, under- 8. Wing Weight Convergence

wing engines, and fuselage-mounted engines with a T- 9. Max. Body and Contents Weight Convergence

tail. However, emphasis of this paper is placed on the

structural aspects of the optimization procedure for fuse- 10. Second Segment Climb Gradient > 2.4%

lage-mounted engine SBW configurations (Fig. 2). Since 11. Balanced Field Length < 11,000 ft

differences in T-tail fuselage-mounted and under-wing

12. Approach Velocity < 140 kts.

engine cantilever designs are small, this study uses can-

tilever optima with wing mounted engines, to make direct 13. Missed Approach Climb Gradient > 2.1%

comparisons with the SBW. 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

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).

Figure 2. SBW with Fuselage-Mounted Engines. Numerous differences between the analysis details of

cantilever and SBW configurations are present in the

DESIGN OPTIMIZATION design code, as is necessary for such dissimilar vehicles.

The primary difference is in the analysis of the wing

GENERAL ASPECTS – The Virginia Tech Truss-Braced bending material weight, as discussed in the structures

Wing (TBW) code models aerodynamics, structures, section. The strut has parasite drag and interference drag

weights, performance, and stability and control of both at its intersections with fuselage and wing. Some geome-

cantilever and strut-braced wing configurations. Design try differences are justified, such as setting the minimum

Optimization Tools (DOT) software by Vanderplatts R&D root chord for the cantilever wing to 52 feet to make room

[10] optimizes the vehicles with the method of feasible for wing-mounted landing gear and kick spar.

directions. Between 15 and 22 design variables are used

in a typical optimization. These include several geometric The SBW, without need for double taper, has the chord

variables such as wing span, chords, thickness to chord linearly interpolated from root to tip. The SBW has a high

ratios, strut geometry and engine location, plus additional wing and fuselage mounted gear. It is important to note

variables including engine maximum thrust and average that, even though the external geometry of the fuselage

cruising altitude. As many as 17 inequality constraints for all cases is identical, the fuselage weights will gener-

may be used (Table 1). ally be different.

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

Baseline

Design

Induced

Initial Design Variables Drag

Updated Design Variables

Geometry

Definition Friction and

Form Drag

Structural

Propulsion Aerodynamics Wave Drag

Optimization/

Weight

SFC L/D Interference

Weight

Drag

Performance Performance

Analysis

Control Constraints

Optimizer

MISSION PROFILE – The primary mission of interest is sidered in the Virginia Tech TBW code are parasite,

a 325-passenger, 7500 nautical mile range, Mach 0.85 induced, interference and wave drag. Unless specified

transport with a 500 nautical mile fuel reserve (Fig. 4). otherwise, the drag model is identical to previous Virginia

Range effects on take-off gross weight and required fuel Tech SBW studies [8]. A detailed description of the drag

weight are investigated. A minimum fuel design is also calculations can be found in [11].

considered.

Parasite Drag – To calculate the parasite drag, form fac-

Several technology groups distinguish the 1995 and 2010

tors are applied to the equivalent flat plate skin friction

technology level aircraft. A 1995 technology aircraft rep-

drag of all exposed surfaces on the aircraft. The amounts

resents an all-metallic benchmark similar to the Boeing

of laminar flow on the wing and tails are estimated by

777. The other aerodynamics grouping includes the

interpolating Reynolds number vs. sweep data for F-14

effects of riblets on the fuselage and nacelles, supercriti-

and 757 glove experiments. Fuselage, nacelles, and

cal airfoils, active load management for induced drag

pylon transition locations are estimated by an input tran-

reduction and all moving control surfaces. Systems tech-

sition Reynolds number. Laminar and turbulent flat-plate

nologies include integrated modular flight controls, fly-by-

skin friction form factors are calculated with LMAS formu-

light and power-by-light, simple high-lift devices, and

las in the Virginia Tech MDO tool. LMAS form factors for

advanced flight management systems. Airframe technol-

wing, tails, fuselage, and nacelles are applied to the skin

ogies represent weight savings from composite wing and

friction drag to obtain the parasite drag.

tails and integrally stiffened fuselage skins. The propul-

sion technology is reflected in reduced specific fuel con-

Induced Drag – The induced drag module uses a dis-

sumption.

crete vortex method to calculate the induced drag in the

Trefftz plane [8]. Given an arbitrary, non-coplanar wing/

Mach 0.85 Cruise 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.

Climb 140 Knot An additional lift-dependent parasite drag component

Approach

Speed

was added to correlate with LMAS drag polars at off-

design conditions.

Wave Drag – The wave drag is approximated with the

T/O Field Length LDG Field Length

Korn equation, modified to include sweep using simple

Figure 4. Mission Profile sweep theory [7], [8]. This model estimates the drag

divergence Mach number as a function of airfoil technol-

AERODYNAMICS – Numerous iterations of both the Vir- ogy factor, thickness to chord ratio, section lift coefficient,

ginia Tech TBW code and Lockheed’s version of NASA’s and sweep angle.

Flight Optimization System (FLOPS) [16] were made so The airfoil technology factor was selected by Lockheed to

that drag polars produced by each code are consistent at agree with the LMAS wave drag. Finally, the wave drag

reference design conditions. The drag components con- coefficient of a wing strip is calculated from the critical

3

Mach number. The total wave drag is found by integrating wise position of the wing-strut intersection are optimized

the wave drag of the strips along the wing. by the MDO code for the 2.5g maneuver load case.

In order to attain acceptable aerodynamic characteristics

Interference Drag – The benefits of a strut-braced wing

of the strut, an airfoil cross section is considered. The

configuration are accompanied by a potential interfer-

strut is designed the way that it will not carry aerody-

ence drag penalty at the junction of the strut with the

namic forces during the cruise condition.

fuselage and the wing. The interference drag between

the wing-fuselage and strut-fuselage intersections are

Structural Assumptions – Preliminary studies have

estimated using Hoerner equations based on subsonic

shown buckling of the strut under the –1.0g load condi-

wind tunnel tests [12].

tion to be the critical structural design requirement in the

The drag of wing-strut junctions can be important in tran- single-strut configuration, resulting in high strut weights

sonic flow because of the presence of shock waves and [8]. To address this issue, an innovative design strategy

separated flow regions. In order to alleviate the problem employs a telescoping sleeve mechanism to allow the

associated with a sharp wing-strut angle, the strut strut to be inactive during negative g maneuvers and

employed here is given the shape of an arch and inter- active during positive g maneuvers. Thus, under the –

sects the wing perpendicularly. Analyses for an arch 1.0g case, the wing acts like a cantilever beam and for

radius ranging from 1 ft to 4 ft were performed with Com- the positive g maneuvers, the wing is a strut-braced

putational Fluid Dynamics (CFD) tools. Unstructured beam.

grids were obtained with the advancing-front methodol-

Even more wing weight reduction can be obtained by

ogy implemented in the code VGRIDns [13], [ 14]. The

optimizing the strut force and wing-strut junction location.

Euler equations were solved using the CFD code USM3D

On a typical optimum single-strut design, this means that

[14], [15] at the cruise Mach number of 0.85.

the strut would first engage in tension at some positive

A very convenient way to extract the interference drag load factor. This can be achieved by assuming a slack in

penalty from a CFD calculation consists in subtracting the wing-strut mechanism. The optimum strut force at

the drag of the wing alone from the drag of the strut- 2.5g is different from the strut force that would be

braced wing design obtained with CFD. The resulting obtained at 2.5g if the strut were engaged for all positive

number is a DCD penalty associated with the presence of values of the load factor. Therefore, the slack load factor

the strut. As the arch radius is increased, the drag pen- is defined as the load factor at which the strut engages

alty decreases almost exponentially. From these results, for the first time. It is important to have the slack load fac-

a curve fit is produced and used in the present analysis to tor always positive, otherwise the strut would be pre-

account for the drag of the wing-strut junction. loaded at the jig shape to achieve the optimum strut

force.

The drag polars output from the Virginia Tech MDO tool

and LMAS modified FLOPS agree within 1% on average

Double Plate Model – For calculating the wing-bending

for cantilever wing designs.

weight of single strut configurations, a piecewise linear

beam model, representing the wing structure as an ideal-

STRUCTURES – Due to the unconventional nature of

ized double plate model, was used (Fig. 5).

the proposed concept, commonly available weight calcu-

lation models for transport aircraft (such as the NASA

t

Langley developed FLOPS) are not accurate enough. A

special bending weight calculation procedure was thus d

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 Cb

achieve a significant reduction in wing/strut interference

drag. Figure 5. Double plate model for bending weight

calculation

Load Cases – To determine the bending material weight

of the strut-braced wing, two maneuver load conditions This model is made of upper and lower skin panels,

(2.5g maneuver, -1.0g pushover) and a taxi bump (-2.0g) which are assumed to carry the bending moment. The

are considered to be design critical. For the -1.0g push- double-plate model offers the possibility to extract the

over and for the -2.0g taxi bump, the strut is not active material thickness distribution by a closed-form equation.

and the wing acts like a cantilever beam. Since the strut The cross-sectional moment of inertia of the wing-box

is not supporting the wing in these cases, very high can be expressed as:

deflections of the wing are expected for the -2.0g taxi

bump. As a result, an optimization procedure is imple- t ( y ) cb ( y ) d 2 ( y )

mented to distribute the bending material to prevent wing I ( y) =

2 (1)

ground strikes. To maximize the beneficial influence of

the strut upon the wing structure, strut force and span-

4

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

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)

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

set length.

-1.5E+07

As a result, it is imperative to employ MDO tools to obtain

-2.0E+07

optimum values for vertical offset, strut force, and span- Wing Half Span (Ft)

wise wing/strut breakpoint. By this way, it is possible to

trade off the two contrary design requirements: (i) a Figure 7. Bending moment distributions for the design

reduced offset length to reduce strut loading, (ii) an critical load cases of the fuselage mounted

increased offset length to reduce the wing/strut interfer- engine SBW

ence drag. After a complete design optimization with the

vertical strut offset as an active design variable, the influ- AEROELASTICITY

ence of the offset weight on the total strut weight

becomes comparably small. For the wing bending weight Hexagonal Wing-Box Model – Although the double plate

and especially for the TOGW it is almost immaterial. 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

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

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, weights such as the maximum body and contents weight

it has been validated using the 747-100 wing. The results and wing weight converge efficiently with the lagging vari-

obtained from the double plate model as well as the hex- able method [10].

agonal wing-box show good agreement with the actual

747-100 and with the results obtained from FLOPS [16] STABILITY AND CONTROL – The horizontal and verti-

andTorenbeek [18]. cal tail areas are first calculated with a tail volume coeffi-

cient sizing method. The tail volume coefficients were

0.7 determined based on Lockheed statistical data. A vertical

tip constraint

2g taxi bump tail sizing routine was developed to account for the one

0.6

engine inoperative condition [8], [17]. The engine-out

Panel thickness in

-1g maneuver

0.5 2.5g maneuver constraint is met by constraining the maximum available

yawing moment coefficient to be greater than the

0.4 required yawing moment coefficient. As specified by FAR

0.3 requirements, the aircraft must be capable of maintaining

straight flight at 1.2 times the stalling speed with the

0.2 operable engine at its maximum available thrust. The lat-

0.1 eral force of the vertical tail provides most of the yawing

moment required to maintain straight flight after an

0 engine failure [11].

0 20 40 60 80 100 120

Wing half span ft The maximum available yawing moment coefficient is

obtained at an equilibrium flight condition with a given

Figure 10. Panel thickness distributions for the different

bank angle and a given maximum rudder deflection. FAR

load cases (fuselage mounted engine

25.149 limits the maximum bank angle to 5°, and some

configuration)

sideslip angle is allowed. The stability and control deriva-

tives are calculated using empirical methods of DATCOM

The total weights for the different components (strut, off-

as modified by Grasmeyer [8], [20]. In order to allow a 5°

set, wing) are obtained using the FLOPS equations.

aileron deflection margin for maneuvering, the calculated

Here, the wing bending material and strut tension

deflection must be less than 20-25°. The calculated avail-

weights are being multiplied by a technology factor to

able yawing moment coefficient is constrained in the opti-

account for the weight reduction achieved by the employ-

mization problem to be greater than the required yawing

ment of composite materials by the year 2010.

moment coefficient. If the yawing moment constraint is

After computation of the load carrying weights, a 10% violated, a vertical tail area scaling factor is applied by the

non-optimum factor is applied to account for manufactur- optimizer.

ing constraints. The total wing weight is calculated using

the FLOPS equations with the overall load carrying PROPULSION – A GE-90 class high-bypass ratio turbo-

weight, i.e. wing, strut, and offset. The total weights of the fan engine is used for this design study. An engine deck

different components are determined according to the was obtained from LMAS, and appropriate curves for

ratio of their contributions to the load carrying weight. specific fuel consumption and maximum thrust as a func-

tion of altitude and Mach number were found through

LMAS provided a weight estimate for the telescoping

regression analysis. The general forms of the equations

sleeve mechanism based on landing gear component

are identical to those found in Mattingly [21] for high-

data. Weights calculated in the Virginia Tech transport

bypass ratio turbofan engines, but the coefficients and

optimization code are identical to FLOPS with the excep-

exponents are modified.

tion of nacelle, thrust reverser, passenger service, land-

ing gear, wing, fuselage and tail weights. The above The engine size is determined by the maximum thrust

weights are now calculated from proprietary LMAS for- required to meet several constraints. These constraints

mulas. Weight technology factors are applied to major are thrust at average cruise altitude, available rate of

structural components and systems to reflect weight sav- climb at initial cruise altitude, balanced field length, sec-

ings due to advances in technology levels from compos- ond segment climb gradient, and missed approach climb

ite materials, advanced electronics and other gradient. The dimensions of the engine nacelles vary as

technologies described above. the square root of required thrust, and the engine weight

is assumed to be linearly proportional to the engine

Some aircraft weights are implicit functions, and internal

thrust. The specific fuel consumption model is indepen-

iteration loops are typically required for convergence.

dent of engine scale. A specific fuel consumption tech-

However, utilizing the optimizer to converge the zero fuel

nology factor is applied to reflect advances in engine

weight of the aircraft showed to be more efficient by pro-

technology.

viding smoother gradients. DOT also selects the fuel

weight so that the range constraint is not violated. Other

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

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

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

minimum fuel weight results.

350000

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- 50000

mized for minimum fuel weight. 4000 6000 8000 10000 12000 14000

Ra nge

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 Figure 13. Effect of range on fuel weight

cantilever design, and 3.6% higher than a minimum-

TOGW-SBW. The SBW L/D increases from 25.4 to 29.1 RANGE EFFECTS – The SBW becomes increasingly

going from the minimum-TOGW to the minimum-fuel desirable as the design range increases. Figures 12 and

case, and from 21.7 to 26.1 for the cantilever configura- 13 show the effects of range on TOGW and fuel weight.

tion. This improved aerodynamic efficiency is achieved by The TOGW reduction relative to the cantilever configura-

increasing the wing span, and comes at a cost in struc- tion steadily improves from 5.3% at a 4,000 nautical mile

tural weight. range up to 10.9% at 12,000 nautical miles. The fuel

weight savings fluctuates within about 11-16%, generally

Airport noise pollution can limit the types of aircraft per- improving with increasing design range. These results

mitted to use certain urban airfields and impose opera- are for minimum TOGW designs, however greater fuel

tional restrictions on those that do. Minimizing engine burn improvements are expected for SBW aircraft opti-

size can also be expected to reduce the noise generated mized for minimum fuel weight. Maximum fuel weight is

if the engine is of similar design. Minimum TOGW SBW set at 400,000 pounds.

engine thrust is reduced by 20.7% over the equivalent

cantilever design, probably reducing airport noise pollu- At 12,000 nautical miles an aircraft can reach any desti-

tion by a similar amount. 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 ACKNOWLEDGMENTS

Virginia Tech transport studies have shown the potential This project is funded by NASA Langley Research Grant

of the SBW over the traditional cantilever configuration. NAG 1-1852. Part of the work was done under subcon-

After much added realism by a major airframe manufac- tract from Lockheed Martin Aeronautical Systems in

turer, the MDO analysis shows that the SBW still demon- Marietta, Georgia. NASA deserves much credit for hav-

strates major improvements over the cantilever wing ing the vision to pursue bold yet promising technologies

configuration. A significant reduction in TOGW was with the hope of revolutionizing air transportation. Lock-

found, but the greatest virtue of the SBW is the improved heed Martin Aeronautical Systems provided valuable

fuel consumption and smaller engine size. These results contributions in data, design methods and advice.

indicate that the SBW will cost less, limit pollutant dis-

charge and reduce noise pollution for urban airports. REFERENCES

Advantages of the SBW increase with range, suggesting

that this configuration may be ideal for larger, long-range 1. Pfenninger, W., “Design Considerations of Large

transports. Subsonic Long Range Transport Airplanes with Low

Drag Boundary Layer Suction,” Northrop Aircraft,

The special design of the strut-braced wing necessitated Inc., Report NAI-54-800 (BLC-67), November 1954.

the development of a wing sizing module suitable to fully

2. Gunston, B., Giants of the Sky: The Biggest

exploit the benefits of this structural configuration. After

Aeroplanes of All Time, Patrick Stephens Limited,

validation with existing aircraft like the 747-100, the mod- Wellingborough, UK, pp. 240-250.

ule was used for wing sizing and structural weight com-

putation of the SBW. The great potential of the outlined 3. Kulfan, R.M., and Vachal, J.D., “Wing Planform

Geometry Effects on Large Subsonic Military Trans-

wing sizing procedure lies in the consideration of jig twist,

port Airplanes,” Boeing Commercial Airplane Com-

strut moment, wing flexibility, and flexible wing spanload. pany, AFFDL-TR-78-16, February 1978.

Consideration of the actual in-flight maneuver loads not

4. Jobe, C.E., Kulfan, R.M., and Vachal, J.D., “Wing

only increases the accuracy in wing sizing but also gives

Planforms for Large Military Transports,” AIAA-78-

the potential for further weight savings. This is especially 1470, 1978.

important within a multidisciplinary design environment

where due to synergistic interaction even small weight 5. Turriziani, R.V., Lovell, W.A., Martin, G.L., Price, J.E.,

Swanson, E.E., and Washburn, G.F., “Preliminary

savings in one component are very likely to result in fur-

Design Characteristics of a Subsonic Business Jet

ther weight reductions for other components. Consept Employing an Aspect Ratio 25 Strut-Braced

Further benefits of the SBW are expected to become Wing,” NASA CR-159361, October 1980.

apparent in future studies. The cooperative relationship 6. Smith, P.M., DeYoung, J., Lovell, W.A., Price, J.E.,

with LMAS focused on adding realism to the SBW design and Washburn, G.F., “A Study of High-Altitude

effort for direct comparisons with the cantilever design. Manned Research Aircraft Employing Strut-Braced

Realism often takes the form of weight penalties and Wings of High-Aspect Ratio,” NASA CR-159262,

expanded performance analysis, which inevitably February, 1981.

detracts from SBW theoretical potential. Presently, efforts 7. Grasmeyer, J.M., Naghshineh_Pour, A., Tetrault, P.-

are underway to identify technologies and strut/truss A., Grossman, B., Haftka, R.T., Kapania, R.K.,

arrangements to exploit the strengths of the strut. Limit- Mason, W.H., Schetz, J.A., Multidisciplinary Design

ing the SBW design arrangements so that the aircraft Optimization of a Strut-Braced Wing Aircraft with Tip-

Mounted Engines, MAD 98-01-01, 1998.

takes the appearance of a cantilever wing may not be the

most appropriate approach to realize the full potential of 8. Grasmeyer, J.M., Multidisciplinary Design Optimiza-

the SBW. tion of a Strut-Braced Wing Aircraft, MS Thesis, Vir-

ginia Polytechnic Institute & State University, April

The SBW is likely to have a more favorable reaction from 1998.

the public than other configurations, especially for those 9. Martin, K.C., Kopec, B.A., “A Structural and Aerody-

who suffer from a fear of flying. Affirmative passenger namic Investigation of a Strut-Braced Wing Transport

and aircrew acceptance is probable because other than Aircraft Concept”, NAS1-96014, November 1998.

the addition of a visually innocuous strut and a high wing, 10. Vanderplaats Research & Development, Inc., DOT

there is little to distinguish the SBW from the existing air- User’s Manual, Version 4.20, Colorado Springs, CO,

liner fleet. Radical appearances of the blended-wing- 1995.

body, joined wing, or other candidate configurations may 11. Gundlach, J.F., Multidisciplinary Design Optimization

cause apprehension in many flying patrons. and Industry Review of a 2010 Strut-Braced Wing

Transonic Transport, MAD 99-06-03, 1999

10

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

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