The aerodynamic design evaluation of a blended-wing-body configuration

Aerospace Science and Technology 43 (2015) 96–110

Contents lists available at ScienceDirect

Aerospace Science and Technology www.elsevier.com/locate/aescte

The aerodynamic design evaluation of a blended-wing-body configuration Payam Dehpanah a,1 , Amir Nejat b,∗,2 a b

K.N. Toosi University of Technology, P.O. Box 16765-3381, Tehran, Iran University of Tehran, P.O. Box 11155-4563, Tehran, Iran

a r t i c l e

i n f o

Article history: Received 11 May 2014 Received in revised form 18 February 2015 Accepted 19 February 2015 Available online 24 February 2015 Keywords: Blended wing body Future airliners Computational aerodynamics Aircraft design Sequential airframes

a b s t r a c t Inherent aerodynamic potential and environmental benefits of the blended-wing-body configuration make it an appropriate candidate for the future airliners. This article studies an initial scaled blendedwing-body airframe using computational analyses in early conceptual design stage. Then, a modified airframe is developed based on evaluation of the initial airframe. Eventually, a full-scaled high-capacity blended-wing-body configuration is proposed for a long-range mission. In assessment of the initial airframe, its aerodynamic coefficients are obtained for a range of angle of attacks based on ReynoldsAveraged Navier–Stokes simulations. The second airframe is designed using conceptual design approach with a typical mission profile, and it is modified based on evaluation of the first airframe. The sequential aerodynamic investigation of the airframes with emphasizing on geometric parameters facilitates the design methodology at its early stage. In the second airframe, the appropriate space for 800 passengers is provided, and geometric parameters are changed according to the mission profile. The current design philosophy allows utilization of maximum aerodynamic potential for designing a blended-wing-body configuration. © 2015 Elsevier Masson SAS. All rights reserved.

1. Introduction The first blended-wing-body airliner, called the Stout Batwing, was designed by William Bushnell Stout in 1926 [35]. He was promoting his design with an unorthodox configuration. Furthermore, the Junkers G38 super jumbo was flying with capacity of 34 passengers in its central body in 1926. Another example of such a configuration was the Ford Trimotor airliner which was flying with 9-passenger capacity at the same time [11]. In early 1940, the X Minor was designed as a research model for studying combination of wing and body in a large airliner [3]. Following this further, the Burnelli CBY-3 with its airfoil liked central body flied in 1944. It was designed with a twin boom for improving the stability in flight [30]. At the end of the World War II, Horton brothers designed the Ho 229, which was a true flying wing configuration [23]. Later, Jack Northrop developed the YB-49 [34]. Nowadays, NASA and the Boeing Company are developing the blended-wingbody configuration as a commercial transports for the future [16].

* 1 2

Corresponding author. E-mail address: [email protected] (A. Nejat). Aerospace Engineering Department. Assistant Professor, School of Mechanical Engineering.

http://dx.doi.org/10.1016/j.ast.2015.02.015 1270-9638/© 2015 Elsevier Masson SAS. All rights reserved.

After emergence of rectangular-shaped body and then tubeshaped body, wings and cylindrical body have become two main characters of commercial flights since early 20th century. Aircraft manufactures remained loyal to them, and passengers, more or less, entered the cylindrical body to travel around the world. At the time of designing the B747, it has been believed a typical configuration with cylindrical body has reached its maximum performance, and further development for commercial transport could be a challenge [15]. However, the Boeing Company came up with an innovative idea which was a practical substitute for addressing real requirements of the future commercial transport in 1998, in a conference in Reno, Nevada. Accordingly, the blended-wing-body configuration officially came into existence for the future generation [17]. In general, aircraft configurations are classified according to conventional, blended wing body, hybrid flying wing, and true flying wing. In comparison with flying wing configuration with no central body also known as tailless fixed wing, in the BWB configuration, passenger cabins, cargo, and equipment are located in central structure of the wings and body. In other words, the BWB configuration combines features of the conventional configuration with the flying wing configuration. It has advantages in terms of performance, and construction in comparison with the conventional configuration. This configuration exploits thick airfoil-liked body in the center, and it accommodates cargo and passengers

P. Dehpanah, A. Nejat / Aerospace Science and Technology 43 (2015) 96–110

97

Nomenclature t /c C D0 C L max C α0 b AR w Y i CR x, y , z CL CD CM L/D S wet C Lα S ref S exposed F C D0 Cp K

thickness-to-chord ratio minimum drag coefficient maximum lift coefficient zero-angle-of-attack lift coefficient wing span . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . m wetted aspect ratio (S wet / S ref ) mean aerodynamic chord location . . . . . . . . . . . . . . . . . . m angle of incidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ◦ root chord . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . m streamwise, spanwise, and vertical coordinates lift coefficient drag coefficient pitching moment coefficient lift-to-drag ratio wetted area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . m2 lift coefficient curve slope . . . . . . . . . . . . . . . . . . . . . . . rad−1 reference wing area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . m2 exposed wing area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . m2 fuselage lift factor parasite drag coefficient pressure coefficient induced drag factor

Acronyms AEROPP CFD RANS SA AR MAC AoA

aerodynamic research on passenger plane computation fluid dynamics Reynolds-Averaged Navier–Stokes Spalart–Almaras aspect ratio mean aerodynamic chord angle of attack

in the center with low compressibility drag. Meanwhile, it reduces total drag comparing with the conventional configurations because its airfoil-liked body with no tail is blended smoothly with outboard wings. Consequently, it increases lift-to-drag ratio and decreases fuel consumption for a long-range high-capacity missions [17]. Moreover, those advantages are expanding on economical fuel consumption, reliability, maintenance period, and low cost for large-scale production [2]. There are several technical advantages in the BWB configuration. Among them, effective spanwise lift distribution is intended to be obtained by using a wide airfoil-liked body. Therefore, entire airframe in this configuration play an effective role in lift generation that improves economical fuel consumption. Meanwhile, this configuration decreases aerodynamic load on outboard wings because of big central chord that bears major part of the span loading [31]. In addition, because of the biggest chord in central body, it needs low lift coefficient to bear an elliptical spanwise load distribution. Therefore, central spanwise location can be thicken to acquire required space for accommodating passengers and cargo without large compressibility drag penalty. In this configuration, most trapezoidal area of planform is covered by the wings, which decreases wing area, and consequently the skin friction drag. Furthermore, shape of the airframe relatively weakens shock waves over the wings and body, and also subsonic flow region behind the shock waves provides appropriate area for engine installation. Besides, its low and effective load coefficient eliminates needs for complex high lift devices because of trim effect. Therefore, it only needs leading edge slots in outboard wings and simple fowler flap along with elevons, which combines functionalities of elevator and aileron.

Re GPS ANT c.g. FAR

Reynolds number global positioning system antenna centre of gravity federal aviation regulation

Greek symbols

α α0 L LE 0.25C 0.5C θ  λ maxle

η β

angle of attack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . zero-lift angle of attack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . leading-edge wing sweep angle . . . . . . . . . . . . . . . . . . . . . . quarter-chord wing sweep angle . . . . . . . . . . . . . . . . . . . . . half-chord wing sweep angle . . . . . . . . . . . . . . . . . . . . . . . . . twist angle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . dihedral angle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . taper ratio sweep angle in maximum t /c location . . . . . . . . . . . . . . airfoil efficiency Mach number parameter

◦ ◦ ◦ ◦ ◦ ◦ ◦ ◦

Super-/subscripts L D M 0 w R C ref LE maxle

lift drag pitching moment zero angle of attack wetted root chord reference leading edge maximum t /c location from leading edge

In central body of this configuration, usable space accommodates passenger cabins, galleys, equipped restrooms. The least possible wetted area for this volume is obtained in shape of sphere. However, the sphere is not aerodynamically appropriate. It is only usable when it flattens out to a disk. Therefore, disk-liked body decreases total wetted area in this configuration, which has low compressibility drag in cruise flight condition [15]. Further, blending the body with the wings in addition of adding an elliptical nose in front of the configuration completes a commercial transport BWB configuration. Meanwhile, engines are connected to the aft portion of central body. Therefore, because of their vertical distance from neutral point, they need to be considered in balancing the configuration around the lateral axis. Several researchers around the globe are investigating the blended-wing-body configuration from different points of view. Among them, Liebeck et al. introduced the BWB configuration as a subsonic commercial transport in 1998. They compared it with conventional configuration, studied its advantages as the future airliner, and performed a multidisciplinary planform optimization for improving its aerodynamic performance [14–17,25]. Roman et al. [31] aerodynamically studied the BWB configuration. They used a multidisciplinary design and optimization technique on its planform for increasing its cruise speed. Kuntawala et al. [13] performed a series of aerodynamic shape optimizations for improving spanwise lift distribution on a BWB configuration with a short range mission. In addition, Reist and Zingg [29] investigated a series of multipoint shape optimizations on a BWB configuration using Euler and RANS simulations. Wakayama et al. [37–41] reconfigured a BWB aircraft using a multidisciplinary design and optimization technique. Lyu and Martins [18,19] studied a BWB

98

P. Dehpanah, A. Nejat / Aerospace Science and Technology 43 (2015) 96–110

configuration using a robust high-fidelity aerodynamic shape optimization technique. Furthermore, Mader and Martins [20] investigated aerodynamic shape optimization with stability constraint for a similar configuration. Hileman et al. [7,8] studied a BWB configuration for achieving a silent fuel-efficient configuration. Diedrich et al. [4] investigated a multidisciplinary design and optimization technique of a silent BWB configuration. Moreover, Sargeant et al. [33] studied static stability and lift distribution of a BWB configuration with leading-edge curving. Bradley [2] investigated sizing methodology for a BWB configuration. Mukhopadhyay et al. [21,22] studied structural design of fuselage in a BWB configuration. In this article, a simple one-percent airframe model of BWB configuration is designed and then studied. The second airframe is developed as a full-scaled configuration for a typical mission profile. The first airframe is obtained through extruding an S-shaped airfoil along the span. Possible modifications are introduced for designing the second airframe after evaluation of the first airframe. In this paper, design methodology for both airframes are explained briefly at the first. Then, aerodynamic performance of the first airframe is studied. A possible approach for meeting pitch trim and obtaining proper static margin are described. Later, design constraints, like required space for 800 passengers, are studied according to a mission profile. The second airframe with its conceptual approach, baseline geometry, interior arrangement, and control surfaces are introduced. Eventually, the first and the second airframes are compared, and at the end, the conclusion is stated. 2. The design methodology The first BWB airframe is obtained simply through extruding an S-shaped airfoil along the span. Provisionally, for opening possible future research opportunity, one-percent scaled airframe is considered at initial conceptual design layout. Meanwhile, because aft portion of the central body is appropriate for engine installation and total pitching moment of the airframe needs to be trimmed in cruise condition, an airfoil with S-shaped chamber line is chosen. The second BWB airframe is designed based on computational aerodynamic analyses of the first airframe. Meanwhile, aircraft conceptual design approach is used for designing the full-scaled configuration. Accordingly, mission profile includes main and reserved sections for this airframe. In addition, the interior arrangement and control surfaces are being sized. The modified airframe is obtained with arranging S-shaped airfoils from central body toward the span, and then smoothly converting their camber line into form of supercritical curvature near the outboard wing. 3. The first airframe In this section, design procedure, including conceptual design approach and baseline geometry, for the first airframe are introduced. The first airframe is assessed using computational fluid dynamics. The computational aerodynamic assessment, grid over the baseline geometry, and implemented computational schemes are also discussed here. Meanwhile, longitudinal stability and pitching moments of the first airframe is investigated afterward. At the end, usable space is discussed as a major design constraint. 3.1. Conceptual design approach Typical mission profile includes takeoff, climb, cruise, loiter, approach and landing segments. The mission profile is schematically shown in Fig. 1. Furthermore, range, cruise speed, and altitude are stated in Table 1, and static weight estimation is demonstrated in Table 2. Additionally, Table 3 illustrates aerodynamic performance. For the weight estimation, airframe’s structure is assumed to be constructed from composite material. Meanwhile, geometric

Fig. 1. Mission profile of the first airframe.

Table 1 Cruise flight condition of the first AEROPP airframe. Cruise conditions

Unit

Value

Range Altitude Speed

km m km/h

90 1800 100

Table 2 Weight estimations of the first AEROPP airframe. Components

Weight (kg)

Control surfaces, jacks and accessories Structure and frame including ribs and spars Servo motor, propeller and speed controller Drive system Battery (Li Polymer) Avionic including transmitter, data link, GPS, ANT

0.35 0.65 0.15 0.04 0.17 0.14

Total weight

1.50

Table 3 Aerodynamic performance of the first AEROPP airframe. Parameters

Unit

Value

∂ C L /∂ α

rad−1

0.0304 −4.5 0.0148 0.801 0.167

α0L

◦

C D0 C L max C Lα

– – –

0

Table 4 Geometric parameters of the first AEROPP airframe. Parameters

Unit

Value

(t /c )airfoil

– m m2 – – m m

0.1 0.93 0.0493 8.35 3.1553 0.14 0.13 33.42 30.01 26.6 2.5 2.5 2 0.18 0.227

b S ref AR AR w MAC Y

LE 0.25C 0.5C

◦ ◦ ◦

i

◦

θ 

◦

CR

λ

◦

m –

parameters of the airframe are demonstrated in Table 4. In this airframe, Selig S5010 airfoil has been chosen for extrusion along the span. This airfoil is designed for low Reynolds number. Its zero pitching moment coefficient is close to zero, which is appropriate for longitudinal stability of the airframe.

P. Dehpanah, A. Nejat / Aerospace Science and Technology 43 (2015) 96–110

99

Fig. 2. The first baseline airframe.

3.2. Baseline geometry The first baseline geometry, as already mentioned, is obtained through extruding the Selig S5010 airfoil along the span with respect to a planform. The first baseline airframe is illustrated in Fig. 2, and also technical drawing of the planform is shown in three views in Fig. 3. The airfoil-liked body is connected smoothly with the wings, and they are connected to a winglet in wingtips. The winglet improves lift distribution by weakening the wingtip vortices [5]. They are important in reducing induced drag. For this reason, they are included in the first baseline airframe. For sizing the winglets, vertical tail volume coefficient is used. 3.3. Computational grid A hybrid grid is used for RANS simulations. This grid is combining a structured boundary layer block with an unstructured block surrounding the boundary layer. The unstructured block includes tetrahedral and pyramid cells. The pyramid cells are connecting hexahedral cells in the structured boundary layer block to the tetrahedral cells in the unstructured block. The boundary layer block is shown in Fig. 4. For the structured boundary layer block, nodes over the surfaces are distributed in streamwise direction, and then, they are extruded in normal direction. For improving orthogonality of the hexahedral cells, from 95% of the chord toward sharp trailing edge of the airframe, surface mesh is extruded separately. The cylindrical unstructured block is illustrated in Fig. 5. The pyramid cells are the interface between the hexahedral and the tetrahedral cells. However, in the trailing edge, due to highaspect-ratio hexahedral cells in initial layout of the boundary layer block, non-conformal grid interface is used instead of the pyramid interface. In case of pyramid interface, the high-aspect-ratio hexahedral cells generates high-aspect-ratio pyramid interface in

Fig. 3. Technical drawing of the first AEROPP airframe.

trailing edge of the boundary layer block. Consequently, the highaspect-ratio pyramid interface reduces the mesh quality affecting the grid convergence. On the contrary, the non-conformal grid interface uses hanging nodes, and connects the hexahedral and the tetrahedral cells without damaging the mesh quality. The nonconformal grid interface is shown in Fig. 5. The hybrid grid has 3.046M cells including 1.6M tetrahedral cells, 1.4M hexahedral cells, and 46K pyramid cells. 3.4. CFD solver The Reynolds-Averaged Navier–Stokes equations with Spalart– Almaras turbulence model is used for the simulations. The oneequation Spalart–Almaras turbulence model is chosen because of anticipated class of flow regime and shape of the airframe [9]. The simulations are iterated in 200,000 Reynolds number, in which the incompressible solver includes the SIMPLEC algorithm for pressurevelocity coupling with second-order scheme and the Green–Gauss gradient evaluation. Meanwhile, with respect to different flight conditions, simulations are performed in different angles of attacks ranging from −16◦ to 25◦ . 3.5. Aerodynamic performance The computational aerodynamic analysis includes variation of lift coefficient, drag coefficient, pitching moment coefficient, lift-

100

P. Dehpanah, A. Nejat / Aerospace Science and Technology 43 (2015) 96–110

Fig. 5. Non-conformal grid interface illustration in (a) winglet, (b) trailing edge. Fig. 4. Structured boundary layer block illustration in (a) nose, (b) winglets, (c) winglet tip, (d) trailing edge.

to-drag ratio, and center of pressure location with angle of attack. Moreover, distribution of spanwise lift coefficient, cross-sectional area, and pressure are also discussed. 3.5.1. Lift coefficient The lift coefficient variation with angle of attack is demonstrated in Fig. 6. As the angle of attack increases, the lift coefficient shows smooth variation. Up to 24◦ , no stall behavior is observed for this airframe in this interval. Among the geometric parameters, low Aspect Ratio and high sweepback angle are responsible for this stall behavior. This issue shows their inappropriateness in the low-Reynolds-number flow regime for this airframe [1]. In general, the BWB configuration introduces high lift-to-drag ratio in a cruise flight condition. The high lift-to-drag ratio increases aerodynamic performance of an airframe. In the first airframe, desired lift coefficient is not satisfied in the cruise flight condition with zero angle of attack. Typically, angle of zero lift is approximately equal to percent camber of an airfoil. In this airframe, the angle of zero lift is 4.5◦ . In addition to this approximation, semi-empirical equation also estimates subsonic lift-curve slope prior to drag-divergent Mach number, which is stated in Eq. (1) [28].

C Lα =

2π AR

 2+

4+

AR2 β 2

η2

(1 +

( tan2 maxle

β2

)

S exposed S ref

)( F )

(1)

Fig. 6. Lift coefficient variation with AoA.

In this equation, maxle indicates sweep angle of a wing in chord location with maximum thickness-to-chord ratio, and η is airfoil efficiency factor, which is ratio of lift curve slope to Mach number parameter for an airfoil. Furthermore, S exposed indicates wing reference area less the parts of the wing covered by the fuselage, and F is lift factor of the fuselage. The lift curve slope of the airframe, according to this equation, is directly affected by the Aspect Ratio, which means the Aspect Ratio of the first airframe is not

P. Dehpanah, A. Nejat / Aerospace Science and Technology 43 (2015) 96–110

101

Fig. 7. Spanwise lift coefficient and thickness-to-chord ratio distribution.

adequate. Thereupon, the second airframe is expected to have a planform with higher Aspect Ratio to satisfy a high lift-to-drag ratio. 3.5.2. Spanwise lift coefficient The spanwise lift distribution is demonstrated in Fig. 7 in cruise flight condition, and also thickness-to-chord ratio distribution across the entire airframe is shown in this figure. The maximum spanwise lift coefficient is obtained in 0.15 m far from the center line, in which the lift coefficient is reduced slightly toward the central locations. The slight reduction of lift coefficient in this portion is because of non-gradual interconnection in trailing edge between swept forward body and swept backward wing. The nongradual interconnection causes high-pressure flow in lower side escapes toward low-pressure flow in upper side. Consequently, it generates vortices, extract energy from flow over the wing, and increases induced drag [10,12]. Therefore, proper spanwise lift distribution and smoother interconnection between the swept forward body and the swept backward wing reduce induced drag. Based on this point, the second airframe is expected to acquire elliptical spanwise lift distribution for minimizing the induced drag. Therefore, the taper ratio is changed, and the thickness-to-chord ratio is expected to vary across the span. 3.5.3. Cross sectional area The second airframe is designed as a commercial transport, in which necessary geometric modification is applied based on study of transonic area rule in the first airframe. The transonic area rule reduces drag at transonic speed. Based on this rule, continuous area distribution diminishes wave drag. Theoretically, the Sears– Haack body has the lowest theoretical wave drag [31]. According to Fig. 8, the area distribution along the central line of the first airframe is shown. The cross-sectional area in each location has been normalized based on the maximum value. In comparison, the cross-sectional area of the Sears–Haack body are provided in this figure as well, and both distributions follow similar trend. However, in the second airframe it is expected varying thickness-tochord ratio across the span makes cross-sectional area distribution closer to the area distribution of the Sear–Haack body. 3.5.4. Drag coefficient The drag coefficient variation with angle of attack is presented in Fig. 9. Although, angle of attack in cruise flight condition is zero, minimum drag coefficient is obtained in a negative angle with positive lift coefficients. One of the major reasons for obtaining minimum drag in a negative angle is high angle of incidence in this airframe. For this purpose, cruise flight condition in zero angle of attack is not economically efficient. The thrust is not minimized in this condition and fuel consumption will raise. Additionally, maximum lift-to-drag ratio is not gained, which affects the aerodynamic performance of the airframe. In the second airframe, it is expected choosing appropriate incidence angle obtains the minimum drag coefficient in between 0◦ and 2◦ angle of attacks.

Fig. 8. Cross-sectional area distribution.

Fig. 9. Drag coefficient variation with AoA.

3.5.5. Drag polar As already mentioned, induced drag minimization is important for improving the aerodynamic performance. Furthermore, for an airframe, drag coefficient is defined according to Eq. (2), which includes minimum drag and induced drag coefficients. Meanwhile, when total drag is minimum, parasite drag is equal to induced drag. The total drag is sum of the parasite drag, induced drag, and wave drag. The parasite drag is consist of form drag, skin-friction drag, and interference drag. The form drag results because of shape of an object. The skin-friction drag depends on wetted area, and interference drag is the increase in the drag of a component due to change of the airflow caused by other component [28]. The profile drag is sum of the form and skin-friction drag.

C D = C D0 + K C L 2

(2)

The induced drag results because high-pressure flow in lower side of the airframe escapes toward upper side in the wingtips, which indicated by second term of the aforementioned equation. For this reason, lift vector rotates backward, and a component of the lift lies in drag direction, which affects aerodynamic performance of an airframe.

102

P. Dehpanah, A. Nejat / Aerospace Science and Technology 43 (2015) 96–110

Fig. 10. Drag polar diagram.

Fig. 11. Pitching moment coefficient variation with AoA.

Meanwhile, strength of tip vortices decreases by increasing the Aspect Ratio, which are produced due to escape of the highpressure flow toward upper side. Therefore, increasing the Aspect Ratio weakens the component of the lift which lies in drag direction, and consequently diminishes the induced drag [32]. For the first airframe, Drag Polar diagram is demonstrated in Fig. 10. The drag coefficient increases dramatically in high lift coefficients, which inhibits lift-to-drag ratio to meet its maximum value. The maximum lift-to-drag ratio is obtained by drawing a tangent line on the curve from the origin. By increasing the Aspect Ratio, the parabolic curve in this diagram is widen, and higher lift with lower drag coefficient is obtained. Therefore, it increases the lift-to-drag ratio. The increase of Aspect Ratio is restricted by structural constraints. Moreover, the parabolic curve in this diagram offsets vertically. The vertical offset is small which implies approximately the minimum drag coefficient equals to zero-lift drag coefficient. 3.5.6. Pitching moment coefficient The pitching moment coefficient variation with angle of attack is demonstrated in Fig. 11. As angle of attack increases, the pitching moment decreases. According to the right-hand rule, positive direction of pitching moment around lateral axis is associated with the counterclockwise rotational direction, indicating downward ro-

Fig. 12. Lift-to-drag ratio variation with lift coefficient.

Fig. 13. Lift-to-drag ratio variation with AoA.

tation of nose. The pitching moment coefficient is negative after −4◦ , and also the curve slope is negative before 5◦ and after of 11◦ . Therefore, as angle of attack increases, pitching moment coefficient decreases, implying the airframe’s nose is turning upward. In between 5◦ and 11◦ , pitching moment coefficient is almost constant because its curve slope is almost zero. For this reason, the airframe needs to be considered for pitch up avoidance in climb condition. In addition, the static margin is negative ratio of pitching moment curve slope to lift curve slope. From the aforementioned points, the second airframe is expected to obtain at least 1% positive static margin with appropriate distribution of airfoil stack across the span. 3.5.7. Lift-to-drag ratio The lift-to-drag ratio variation with lift coefficient is illustrated in Fig. 12. The maximum lift-to-drag ratio is obtained in 0.175 lift coefficient. Accordingly, higher lift coefficient and lower drag coefficient increases the ratio, which is desirable for decreasing the fuel consumption and increasing the range. The lift-to-drag ratio variation with angle of attack is illustrated in Fig. 13. Its maximum value is 8.5 obtained in zero angle of attack. Setting the maximum lift-to-drag ratio in cruise flight

P. Dehpanah, A. Nejat / Aerospace Science and Technology 43 (2015) 96–110

103

in negative angle of attacks. The longitudinal location varies approximately around 23.5 cm far from the nose. Comparing with aerodynamic center and center of gravity, this location is important in order to acquire static margin in cruise flight condition.

Fig. 14. Center of pressure longitudinal location.

Fig. 15. Total pressure line at AoA = 2◦ and Re = 200K over the upper surface. (For interpretation of the colors in this figure, the reader is referred to the web version of this article.)

condition improves the aerodynamic performance; however, the maximum value is not desirable. The maximum lift-to-drag ratio is relatively lower than the desirable value for the airframe, resulted by small lift-curve slope with large drag coefficient. 3.5.8. Center of pressure location The longitudinal location of the center of pressure varies slightly in term of angle of attack according to Fig. 14. Along the central chord of the airframe, this displacement reaches to 8%

3.5.9. Pressure distribution The pressure line distribution is demonstrated in Fig. 15. Based on the pressure line pattern in the interconnection between swept forward body and swept backward wing, high-pressure flow escaped toward upper surface in this region. Therefore, smoother interconnection diminishes induced drag and viscous separation drag, which is desirable in the second planform. The pressure distribution over upper and lower surface of the airframe is provided in Fig. 16. Over the lower surface, pressure coefficient fairly remains uniform, except there is a slight variation narrowly in the leading edge. Over upper surface, there is a little gap between low and high pressure coefficients in trailing-edge interconnection between the swept forward body and the swept backward wing. The gap between low and high pressure coefficients develops toward wingtips, which makes airflow vulnerable to separate. The gap produces vortices which extract energy from the flow and develop induced drag. The spanwise pressure coefficient distribution is demonstrated in Fig. 17 for seven locations across the span. Respectively, Reynolds number and angle of attack are 2 × 105 and 2◦ . Among the spanwise locations, two are situated on the body, one is situated before and after the tailing edge interconnection, another is situated on the pressure gap after the interconnection, and two are situated near the wingtip. Their chordwise profiles are also shown below their pressure coefficient distribution. They demonstrates similar profile twisted across the span. Their locations are indicated based on the semi-span percentage. The pressure coefficient in 10% semi-span location remains almost constant in lower surface, and increases sharply till 20% of chord in upper surface, then decreases gradually. In 20% semi-span location, the pressure coefficient follows similar trend in the upper surface, except its peak is increased slightly in 20% of chord location. In 40% and 50% semispan locations, which are located before and after the trailing-edge interconnection, the pressure coefficient increases sharply until the 10% of the chord in the upper surface, decreases gradually until 40% of the chord, remains almost constant until 80% of the chord, and then decrease to reach to similar pressure coefficient to the lower surface. The peak of the upper-surface pressure coefficient is increased in these sections in 10% of chord location in comparison with the previous section. In these semi-span locations also, the pressure coefficient in the lower surface is almost constant. In 60% semi-span location, the pressure coefficient increases sharply until 10% of chord in the upper surface, decreases gradually until 40%

Fig. 16. Total pressure distribution at AoA = 2◦ and Re = 200K over upper and lower surfaces. (For interpretation of the colors in this figure, the reader is referred to the web version of this article.)

104

P. Dehpanah, A. Nejat / Aerospace Science and Technology 43 (2015) 96–110

Fig. 17. Comparison of chordwise pressure distribution at Re = 200K and AoA = 2◦ . (For interpretation of the colors in this figure, the reader is referred to the web version of this article.)

of the chord, remains constant until 70% of the chord, increases again slightly until 80% of chord, and then decreases to reach to a value lower than pressure coefficient in lower surface. The pressure coefficient also in this section remains almost constant in lower surface. In 80% and 90% semi-span locations, the pressure coefficient increases sharply until 10% of chord in the upper surface, decrease gradually until 40% of chord, remains constant until 80% of chord, increases again slightly until almost 90% of chord, and then decrease to reach to a value similar to the lowersurface pressure coefficient. The lower-surface pressure coefficient in 80% semi-span location remains almost constant; however, in 90% semi-span location, it decreases slightly in 90% of chord location. From the aforementioned points, in the second airframe, it is expected the trailing-edge interconnection with relatively smoother curvature and lower swept forward angle to be considered in the planform. Moreover, for improving the spanwise pressure distribution, which consequently leads toward effective spanwise lift distribution, different spanwise profiles are expected to be exploited. Because of transonic flow regime in the second airframe, supercritical airfoils are needed to be situated in the outboard wing. For this purpose, camber line of airfoil stack is transformed gradually

Fig. 18. Schematic view of pitching moments around the airframe.

along the span from the S-shaped curvature to the supercritical curvature. 3.6. Longitudinal stability and controllability The pitching moments in cruise flight condition around the lateral axis are identified in Fig. 18. Accordingly for obtaining the

P. Dehpanah, A. Nejat / Aerospace Science and Technology 43 (2015) 96–110

105

Fig. 19. Location of aerodynamic center and types of winglets.

Fig. 20. Schematic view of the c.g. movement for balancing the airframe.

pitch trim, pitching moment due to external flow identifies pitching moment produced by pressure distribution over the airframe. The pitching moment due to thrust vector identifies pitching moment because of its vertical distance in aft body. The pitching moment due to c.g. location identifies pitching moment because of its location in front of aerodynamic center. As stated earlier, the pitching moment curve slope is negative for the airframe, and it is also negative in zero angle of attack. Therefore, aerodynamic pitching moment turns the nose in clockwise rotational direction. Besides, because vertical distance of the thrust vector is located above the c.g., it turns the nose in counterclockwise rotational direction [36]. Moreover, the aerodynamic center of the first airframe is located behind the c.g. according to Fig. 19a. Therefore, it turns the nose in counterclockwise rotational direction. The aerodynamic center is calculated based on subsonic semi-empirical equations for similar geometries [6]. In general in a BWB configuration, because of small moment arm, advanced control systems and flight computers is being used for turning around the lateral axis in addition to obtaining longitudinal static margin in a cruise flight condition. In fact, several control surfaces simultaneously contribute in stable and controllable turn [24]. The first AEROPP airframe is a one-percent research airframe in which electric engines are considered for free flight tests. For this reason, the aircraft weight does not vary during a test because of fuel consumption. In this case, it is possible to obtain the pitch trim and longitudinal static margin in the cruise flight condition by adding an external mass moving along the central line. The external mass is indicated in Fig. 20. It can exactly match the c.g. location with the aerodynamic center and remove the pitching moment due to c.g. location. Furthermore, because

aerodynamic pitching moment turns the nose upward by increasing the angle of attack and thrust vector turns the nose downward, pitch trim is achievable. By moving the external mass backward, which moves the c.g. location backward, we can balance the airframe in the cruise flight condition. As stated earlier, this approach is appropriate for obtaining the pitch trim in this airframe because of using electric engines. In the first airframe, similar to the Boeing BWB concept, yaw motion is considered with a Split Drag-Rudder and a DoubleHinged Rudder, which are situated on the winglets as control surfaces. They are schematically shown in Fig. 19b. Their functionalities enable the airframe to spin around the c.g. without sliding in transversal direction. Meanwhile, based on aerodynamic forces of the winglets, they produce appropriate vector which acts together with the forward c.g. motion in a bank maneuver. 3.7. Usable space The blended-wing-body configuration provides adequate space for accommodating passengers and cargo. The space is not usable unless an airfoil with appropriate thickness-to-chord ratio to be used in the central body. The first airframe does not provide adequate space for accommodating passengers and cargo, which is shown in Fig. 21, and the Selig S5010 airfoil with 10% thickness-to-chord ratio is not proper. It is expected the second airframe to be designed with appropriate airfoil, which provides sufficient thickness-to-chord ratio for accommodating 800 passengers and cargo. Meanwhile, because the central airfoil-liked body has small compressibility drag, it can be thicken enough to accommodate the passengers in a double-deck interior and

106

P. Dehpanah, A. Nejat / Aerospace Science and Technology 43 (2015) 96–110

Table 5 Geometric parameters of the second AEROPP airframe. Parameters

Unit

Value

(t /c )central

– m m2 – – m m

0.18 0.8018 0.0659 10 4.3416 0.0970 0.1465 37.49 35.24 32.7 4 .5 4 3 0.1345 0.22

b S ref AR AR w MAC Y

LE 0.25C 0.5C

◦ ◦ ◦

i

◦

θ 

◦

CR

λ

◦

m –

Table 6 Design constraints of the second AEROPP airframe. Constraints

Unit

Value

Passengers Crew Flight attendants Range Diversion Diversion altitude Loiter endurance Cruise altitude Cruise speed Field length

– – – km km m min m Mach m

800 2 20 121,964 277.8 6100 35 10,000 0.85 3690

Fig. 21. Schematic view of usable space for the first AEROPP airframe.

participate in the lift generation. The thick central body hold major load over the airframe allowing outboard wing to be thinned adequately and preventing wave drag development over the wings. 4. The second airframe The second airframe is designed and modified based on evaluation of the first airframe. In this section, design procedure of the second airframe is described which includes the conceptual design approach, baseline geometry, interior arrangement, and control surfaces. 4.1. Conceptual design approach In the second airframe, required space for 800 passengers is acquired by choosing a very thick airfoil in the central body. Further, thickness-to-chord ratio of the airfoil stack is changed from 18% to 10% across the span. Other modifications are applied to this airframe based on the assessment of the first airframe. Among them, Aspect Ratio of the wings are increased to improve the maximum lift-to-drag ratio, in particular wetted Aspect Ratio. Moreover, five different airfoils with different thickness-to-chord ratio are situated across the span. With respect to the increase of Aspect Ratio, the span is decreased in comparison with the first airframe. However, the wing area is increased just 33% comparatively. The geometric parameters of the one-percent second airframe is stated in Table 5. In addition, design constraints for this airframe is demonstrated in Table 6. They comply with FAR Part 25 regulations for general aviation aircraft. In this airframe, the reserved mission profile is also added which is indicated by diversion parameters in this table. The second airframe performance is illustrated in Table 7. The weight is estimated using a buildup method together with a refined method. In the refined method, weight fractions are corrected based on wing loading and thrust-to-weight estimations.

Table 7 Design performance of the second AEROPP airframe. Parameters

Unit

Value

MTOGW OEW Fuel burned L / D (cruise) Wing Span Wing Area(trap) Wetted Area T /W Thrust(total)

kg kg kg – m m2 m2 – N

358,838.99 178,603.86 103,463.38 21.65 80.18 670.076 2756.348 0.2009 2 × 353,484

Meanwhile, operational empty weight and fuel burned weight are calculated in the weight estimation. Besides, the c.g. location is calculated using a group weights method, and DATCOM method’s semi-empirical approach is implemented for aerodynamic performance estimation. 4.2. Baseline geometry The full-scaled baseline geometry of the second airframe is demonstrated by its technical drawing in Fig. 22. It shows dimensions of the second configuration in top, front, and side views. The control surfaces are also sized in this airframe, and they are indicated in this figure. 4.3. Interior arrangement The lower and upper passenger cabins are demonstrated in Fig. 23 in the second airframe. The upper deck includes first, business, and economic seats class located in 7 cabin bays. In every bay, galley and lavatory are located at the rear. The lower deck is accessible through two stairs in front of the upper deck. Additional stairs toward the lower deck are also situated in rear of the each bay.

P. Dehpanah, A. Nejat / Aerospace Science and Technology 43 (2015) 96–110

Fig. 22. Technical drawing of the second AEROPP airframe.

Fig. 23. Schematic views of the interior arrangement at (a) upper deck and (b) lower deck.

107

108

P. Dehpanah, A. Nejat / Aerospace Science and Technology 43 (2015) 96–110

Fig. 24. Sectional view of the interior arrangment.

The first seats class in the upper deck includes two cabin bays. One bay is located in front of the deck and another one is located in one of the central bays. The lower deck includes first and economics seats class included in 5 bays. The first seats class in the lower deck is located in front with wide view toward the sky, and other bays in this deck are dedicated to the economic seats class. The entire interior of the second airframe, including passenger cabins and cargo bays, are illustrated by the middle sectional view in Fig. 24. The passenger bays in the center are surrounded by the cargo bays. As it is shown, the double deck interior arrangements appropriately fits in the provided space in this airframe. 4.4. Airframe sections and control surfaces The airframe layout of the second airframe is demonstrated in Fig. 25a. Accordingly, passengers and cargo are placed in the central body, and fuel tanks are situated in the outboard wings. Moreover, tip-to-tip control surfaces are sized for this airframe for providing 10% extra controllability. Among the tip-to-tip control surfaces, a single slotted fowler flap is placed between the two engines. In the trailing interconnection between swept back wing and swept forward body, a double slotted flap is situated. In this region, horizontal line of the trailing edge improves its performance. Between the fowler and double slotted flaps, elevators are sized. In addition, ailerons are located toward the tips at the end of the double slotted flap. In the outboard wing, a leading edge slot is sized. Similar to the first airframe, a Split Drag-Rudder and a Double-Hinged Rudder are situated on the winglets demonstrated in Fig. 25b. 5. Airframe comparison

Fig. 25. Second airframe layout and types of winglets.

A comparison between the second and the first one-percent airframes is presented in this section. Comparatively in the second airframe, wing span is doubled, and also wing reference area increased by 33%. Moreover, wing Aspect Ratio is almost doubled, and MAC increased by 72%. The leading edge sweep angle of the wing increased by 4 degrees, and the angle of incidence increased by 2 degrees. Furthermore, the wing twist angle raised by 2 degrees, and also the dihedral angle raised by 1 degrees. The central chord length is doubled. The wing taper ratio relatively remains constant. The thickness-to-chord ratio in central section is doubled. As earlier stated, appropriate spanwise lift distribution across the span play crucial role in improving the aerodynamic performance [26,27]. This distribution can be adjusted by changing the taper ratio and the twist angle of the airfoil stack across the span. Meanwhile, trend studies states the taper ratio depends on Mach number, which defines the wing sweep angle. Therefore, the wing

Fig. 26. Baseline geometry comparison between the first and the second airframes.

P. Dehpanah, A. Nejat / Aerospace Science and Technology 43 (2015) 96–110

Fig. 27. Planform comparison between the fist and the second airframes.

109

lence model. The aerodynamic performance investigation includes variation of lift, drag, and pitching moment coefficients. In addition, spanwise lift and cross-sectional area distribution are studied for improving the performance of the second airframe. Meanwhile, Drag Polar diagram, lift-to-drag ratio, location of center of pressure, pressure distribution, longitudinal stability and controllability, and usable space are also studied in this airframe. Through this investigation, advantages and disadvantages of the geometric parameters are identified in the first airframe. Then, based on assessing the first airframe, the second airframe is designed and modified. In design procedure of the second airframe, conceptual design approach is implemented, in which a buildup method together with a refined method for weight estimation, semi-empirical method for aerodynamic performance estimation, and group weights method for calculating the c.g. location are being used. Meanwhile, design constraints are applied to geometric parameters of the second airframe based on main and reserved missions. In this airframe, interior arrangement and control surfaces are also investigated. In summary, baseline geometries and planform of the first and the second airframes are compared, and the second airframe is proposed as a high-capacity long-range blended-wing-body commercial transport. As the future works, aerodynamic shape optimization of the spanwise airfoils and multiobjective planform optimization improve performance of the airframe further, which are currently under investigation by the authors. Conflict of interest statement

Fig. 28. Artist’s rendering of the second AEROPP configuration.

sweep angle also affects the spanwise lift distribution. From these points, shape of the wings is modified in the second airframe with aim of improving the spanwise lift distribution in cruise condition. The comparison between the first and the second baseline geometries are demonstrated in Fig. 26. The planform comparison between the first and the second airframes are provided in Fig. 27. Accordingly, Aspect Ratio is increased notably. Moreover, for decreasing the wetted area, leading edge curvature is curved inward in forepart of the body. Altogether, in the second airframe, two turbofan engines are connected by their pylons at the rear, two winglets are sized and placed in the wingtips, and a bullet-liked nose is added in front of the airframe. An artist’s rendering of the second AEROPP configuration is shown in Fig. 28. 6. Conclusions Combining wing and body in the blended-wing-body configuration is an innovative idea which benefits from its inherent aerodynamic potential. However, it needs creative and careful revision at its stage of evolution as a proper candidate for the future generation of commercial transport. In this work, aerodynamic performance of a blended-wing-body airframe is studied with aim of improvement in early stage of the conceptual design. For this purpose, key design parameters are identified for two sequential airframes using computational fluid dynamics. In this procedure, the first airframe is designed as a simple research model. Its aerodynamic performance for a mission profile is investigated based on RANS simulation with SA turbu-

This research is carried out without any kinds of connections and involvements from any organizations and industrial projects. We confirm there are no conflict of interest associated with this publication and there has been no financial support for this work that influenced its outcome. The manuscript is read and approved by all authors and there are no persons who satisfied the criteria for authorship but are not listed. Appendix A. Supplementary material Supplementary material related to this article can be found online at http://dx.doi.org/10.1016/j.ast.2015.02.015. References [1] J.D. Anderson Jr, Fundamentals of Aerodynamics, Tata McGraw–Hill Education, 1985. [2] K.R. Bradley, A sizing methodology for the conceptual design of blended-wingbody transports, NASA/CR-2004-213016, 2004. [3] D.L. Brown, Miles Aircraft Since 1925, Putnam Limited, 1970. [4] A. Diedrich, J. Hileman, D. Tan, K. Willcox, Z. Spakovszky, Multidisciplinary design and optimization of the silent aircraft, AIAA Paper 1223, 2006. [5] R. Nangia, M. Palmer, R. Doe, Aerodynamic design studies of conventional and unconventional wings with winglets, in: 24th AIAA Applied Aerodynamics Conference, American Institute of Aeronautics and Astronautics, 2006. [6] R. Fink, D. Hoak, USAF Stability and Control DATCOM, Air Force Flight Dynamics Laboratory, Wright-Patterson AFB, Ohio, 1996. [7] J. Hileman, Z. Spakovszky, M. Drela, M. Sargeant, Aerodynamic and aeroacoustic three-dimensional design for a silent aircraft, in: 44th AIAA Aerospace Sciences Meeting and Exhibit, 2006, 2006-241. [8] J. Hileman, Z. Spakovszky, M. Drela, M. Sargeant, A. Jones, Airframe design for silent fuel-efficient aircraft, J. Aircr. 47 (3) (2010) 956–969. [9] A. Jameson, A perspective on computational algorithms for aerodynamic analysis and design, Prog. Aerosp. Sci. 37 (2) (2001) 197–243. [10] R.T. Jones, The Spanwise Distribution of Lift For Minimum Induced Drag of Wings Having a Given Lift and a Given Bending Moment, National Advisory Committee for Aeronautics, 1950. [11] A. Klemin, Junkers G-38, progress of the year and more, Technology 142 (1930) 150. [12] I. Kroo, Drag due to lift: concepts for prediction and reduction, Annu. Rev. Fluid Mech. 33 (1) (2001) 587–617.

110

P. Dehpanah, A. Nejat / Aerospace Science and Technology 43 (2015) 96–110

[13] N.B. Kuntawala, J.E. Hicken, D.W. Zingg, Preliminary aerodynamic shape optimization of a blended-wing-body aircraft configuration, in: 49th AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition, Orlando, Florida, 2011. [14] R. Liebeck, Blended wing body design challenges, in: AIAA International Air and Space Symposium and Exposition: The Next 100 Years, American Institute of Aeronautics and Astronautics, 2003, pp. 14–17. [15] R.H. Liebeck, Design of the blended wing body subsonic transport, J. Aircr. 41 (1) (2004) 10–25. [16] R.H. Liebeck, M.A. Page, B.K. Rawdon, Evolution of the Revolutionary BlendedWing-Body, McDonnell–Douglas Aerospace, Transportation Beyond 2000: Technologies Needed for Engineering Design, 1996, pp. 431–459 (SEE N 96-26302 09-01). [17] R.H. Liebeck, M.A. Page, B.K. Rawdon, Blended-wing-body subsonic commercial transport, AIAA Paper 438, 1998. [18] Z. Lyu, J. Martins, RANS-based aerodynamic shape optimization of a blendedwing-body aircraft, in: 21st AIAA Computational Fluid Dynamics Conference, American Institute of Aeronautics and Astronautics, 2013. [19] Z. Lyu, J. Martins, Aerodynamic shape optimization of a blended-wing-body aircraft, in: 51st AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition, American Institute of Aeronautics and Astronautics, 2013, 2013-0283. [20] C.A. Mader, J.R. Martins, Stability-constrained aerodynamic shape optimization of flying wings, J. Aircr. 50 (5) (2013) 1431–1449. [21] V. Mukhopadhyay, Blended wing body (BWB) fuselage structural design for weight reduction, in: 46th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference, American Institute of Aeronautics and Astronautics, 2005. [22] V. Mukhopadhyay, J. Sobieszczanski-Sobieski, I. Kosaka, G. Quinn, G. Vanderpaats, Analysis, design, and optimization of noncylindrical fuselage for blended-wing-body vehicle, J. Aircr. 41 (4) (2004) 925–930. [23] D. Myhra, The Horten Brothers and Their All-Wing Aircraft, Schiffer Publishing, 1998. [24] K. Nickel, M. Wohlfahrt, E.M. Brown, Tailless Aircraft in Theory and Practice, American Institute of Aeronautics and Astronautics, 1994. [25] M. Potsdam, M. Page, R. Liebeck, Blended wing body analysis and design, in: 15th Applied Aerodynamics Conference, American Institute of Aeronautics and Astronautics, 1997.

[26] N. Qin, A. Vavalle, A. Le Moigne, M. Laban, K. Huckett, P. Weinerfelt, Aerodynamic studies of blended wing body aircraft, AIAA Paper 5448, 2002, p. 2002. [27] N. Qin, A. Vavalle, A.L. Moigne, Spanwise lift distribution for blended wing body aircraft, J. Aircr. 42 (2) (2005) 356–365. [28] D.P. Raymer, et al., Aircraft Design: A Conceptual Approach, American Institute of Aeronautics and Astronautics Reston, 2006. [29] T.A. Reist, D.W. Zingg, Aerodynamic shape optimization of a blended-wingbody regional transport for a short range mission, in: 31st AIAA Applied Aerodynamics Conference, American Institute of Aeronautics and Astronautics, 2013. [30] David R. Townend, Clipped Wings – The History of Aborted Aircraft Projects, AeroFile Publications, Markham, Ontario, ISBN 978-0-9732020-4-5, 2007. [31] D. Roman, R. Gilmore, S. Wakayama, Aerodynamics of high-subsonic blendedwing-body configurations, AIAA Paper 554, 2003, p. 2003. [32] J. Roskam, Airplane Design: Part 5-Component Weight Estimation, DAR corporation, 1985. [33] M. Sargeant, T. Hynes, W. Graham, J. Hileman, M. Drela, Z. Spakovszky, Stability of hybrid-wing-body-type aircraft with centerbody leading-edge carving, J. Aircr. 47 (3) (2010) 970–974. [34] W. Sears, Flying-wing airplanes – The XB-35/YB-49 program. [35] W.B. Stout, A low-A/R, all-wing light plane with pilot and engine within wing, British Patent # 149,708, 1919. [36] M. Voskuijl, G. La Rocca, F. Dircken, Controllability of blended wing body aircraft, in: Proceedings of the 26th International Congress of the Aronautical Sciences, ICAS 2008, including the 8th AIAA Aviation Technology, Integration and Operations (AIO) Conference, Anchorage, Alaska, September 14–19, 2008, Optimage Ltd., ISBN 0-9533991-9-2, 2008. [37] S. Wakayama, Blended-wing-body optimization problem setup, AIAA Paper 4740, 2000. [38] S. Wakayama, Multidisciplinary design optimization of the blended-wing-body, AIAA Paper 4938, 1998. [39] S. Wakayama, I. Kroo, Subsonic wing planform design using multidisciplinary optimization, J. Aircr. 32 (4) (1995) 746–753. [40] S. Wakayama, I. Kroo, The challenge and promise of blended-wing-body optimization, AIAA Paper 98-4736, 1998. [41] S. Wakayama, M. Page, R. Liebeck, Multidisciplinary optimization on an advanced composite wing, AIAA Paper 4003, 1996.