Aerodynamic Efficiency of High Maneuverable Aircraft Applying Adaptive Wing Trailing Edge Section

24TH INTERNATIONAL CONGRESS OF THE AERONAUTICAL SCIENCES

AERODYNAMIC EFFICIENCY OF HIGH MANEUVERABLE AIRCRAFT APPLYING ADAPTIVE WING TRAILING EDGE SECTION

Christian Breitsamter Institute for Fluid Mechanics, Aerodynamics Division, Technische Universität München Boltzmannstrasse 15, 85748 Garching, Germany

Keywords: Adaptive wing, high agility aircraft, high angle of attack, force measurements

Abstract environmental impacts enforce continuous im- The aerodynamic characteristics of a high- provement of the performance of both civil and agility aircraft of canard-delta wing type are military aircraft. A key issue enhancing the air- analyzed in detail comparing the performance craft efficiency is the use of advanced and adap- obtained by smooth, variable camber wing trail- tive wing technologies (Fig. 1) [1-4]. ing-edge sections versus conventional trailing- The aerodynamic characteristics of adap- edge flaps. Wind tunnel tests are conducted on a tive and form-variable wing elements and multi- detailed model fitted with discrete elements rep- functional control surfaces have been inten- resenting the adaptive wing trailing-edge sec- sively studied in several German and European tions. Force and flow field measurements are research programs [5-8]. Results show increased carried out to study the aerodynamic properties. glide number, reduced drag and improved high Lift, drag and pitching moment coefficients are lift performance. A main focus is on structural evaluated to assess aerodynamic efficiency and concepts for adaptive flap systems as well as for maneuver capabilities. Comparing the data for local surface deformations and the integration of adaptive and conventional flap configurations, related sensors, actuators and controllers in the it is shown that smooth, variable camber results aircraft [9]. Local surface thickening known as in an increase in lift at given angle of attack and ‘Shock Control Bump’ (SCB) is used to modify in a decrease in drag at given lift coefficient. the shock development at transonic flight condi- These benefits hold also for trimmed conditions. tions [10]. An appropriate position and shape of They are demonstrated for an angle-of-attack the SCB may lead to a nearly isentropic recom- range of −10 to +30 deg, substantiating a sig- pression along with a reduced pre-shock Mach nificant improvement in flap efficiency for the number. This reduction results in a lower shock adaptive wing concept. Thus, the form-variable strength and, therefore, in a decrease in wave wing demonstrates a remarkable potential to drag. Extensive studies are conducted on the enhance flight and maneuver performance and aerodynamic and structural design of SCBs to to alleviate maneuver loads, respectively. make them applicable for large transport aircraft wings [11, 12]. 1 Introduction Also, advanced structural concepts for variable camber wing trailing-edge sections The forthcoming demands on safe and efficient have been developed. The structural solutions flight operations in context of society needs and were proven by full scale ground demonstrators applying the so called ‘horn concept’ [13], the

Copyright © 2004 by C. Breitsamter. Published by the ‘girdle concept’ [14] and the ‘finger concept’ International Council of the Aeronautical Sciences, with [9]. Further studies concentrate on multifunc- permission. tional control surfaces [15]. For example, small

1 CHRISTIAN BREITSAMTER

2 Experiment and Test Program

Laminar Shock Turbulent 2.1 Model and Facility flow control control flow control Perturbation Experiments are performed on an 1:15 scaled damping detailed steel model of a modern high-agility Boundary turbulent aircraft (Fig. 2). Major parts of the model are layer suction Transition

laminar Load nose section, front fuselage with rotatable can- control ards and a single place canopy, center fuselage with delta wing and a through-flow double air Variable intake underneath, and rear fuselage including camber nozzle section and empennage (fin). The baseline wing configuration is equipped with con- Fig. 1 Adaptive wing technologies. ventional inboard and outboard trailing-edge control surfaces mounted at the trailing-edge of flaps. The model elements designed to represent conventional flaps show an increase in maxi- the smooth, variable camber of the adaptive mum lift of about 7% [16]. Recently, investiga- wing trailing-edge section have the same span- tions on ‘Miniature Trailing Edge Devices’ wise and chordwise dimensions as the conven- (Mini-TEDs) have been performed in wind tun- tional flaps. The form-variable section is com- nel and flight tests within the technology plat- posed by six discrete wing elements which are form AWIATOR (‘Aircraft WIng with Ad- exchanged for the conventional flaps to run the vanced Technology OpeRation’) funded by the adaptive wing tests. The corresponding de- European Commission [8]. flections are defined by the tangent on the Concerning military aircraft, comprehen- trailing-edge including values of ηFT = 0°, 5°, sive flight tests were carried out on the F-111 10°, 15°, 20°, and 25° (Fig. 3). To evaluate the research aircraft to develop and assess the tech- aerodynamic performance obtained by adaptive nology enhancements of the ‘Mission Adaptive wing versus conventional flaps, the comparison Wing’ (MAW), employing smooth contour, is based on a measure of deflection angles with variable camber leading- and trailing-edge sur- respect to the skeleton line. Fig. 4 depicts the faces [17]. Flight tests were conducted in frame corresponding definitions used for conventional of the joint U.S. Air Force/NASA/ Boeing Ad- flap deflection ηK and deflection of the form- vanced Fighter Technology Integration (AFTI/ variable trailing-edge section ηFS. F-111) program. Improvements in glide number, The experiments have been carried out in large reductions in drag at off-design conditions the large low-speed facility A of the Institute for and alleviation of maneuver loads have been Fluid Mechanics (Aerodynamics Division) of successfully demonstrated [18-20]. Technische Universität München. This closed- However, there is a lack of investigations return wind tunnel can be operated with both applying adaptive leading- and trailing-edge open and closed test section at maximum usable surfaces on high-maneuverable aircraft fitted velocities of 75 m/s and 65 m/s, respectively. with slender wing geometry [21]. Research ac- Test section dimensions are 1.8 m in height, 2.4 tivities concentrate mainly on related structural m in width and 4.8 m in length. Because high concepts but there is also a need for systematic angle-of-attack experiments are conducted the aerodynamic investigations. Therefore, the pre- open test section is used. The test section flow sent study uses an aircraft model of canard-delta was carefully inspected and calibrated docu- wing type to assess the aerodynamic perform- menting a turbulence level less than 0.4% and ance of adaptive wing sections versus conven- uncertainties in the spatial and temporal mean tional trailing-edge flaps. velocity distributions of less than 0.067%.

2 AERODYNAMIC EFFICIENCY OF HIGH MANEUVERABLE AIRCRAFT APPLYING ADAPTIVE WING TRAILING EDGE SECTION

2s = 0.740 m ϕW = 50° Index: lµ = 0.360 m ϕW = 45° W: Wing ARW = 2.45 λW = 0.14 C: Canard ARF = 1.38 ϕF = 54° F: Fin Skeleton line

a) Conventional trailing-edge flap.

Skeleton line

b) Form-variable trailing-edge section.

Fig. 4 Definition of trailing-edge deflections.

Fig. 2 Model geometry of delta-canard fighter aircraft.

a) Front view. ηFT = 0°

ηFT = 5°

ηFT = 10°

ηFT = 15° η = 15° FT

η = 20° FT

η = 25° FT

ηFT = 25°

b) Rear views. Fig. 3 Wing trailing-edge elements representing the smooth contour, variable camber trailing-edge section. Fig. 5 Model mounted in test section.

3 CHRISTIAN BREITSAMTER

gauge bridges. The balance is designed to sus- Wiring to connect strain gauges tain maximum loads of 900 N, 450 N, and 2500 with amplifier modules N for axial, lateral and normal forces, respectively, and maximum moments of 120 Nm, 160 Components for strain gauge balance integration Nm, and 120 Nm for rolling, pitching, and yawing moments, respectively. The accuracy based Strain gauge balance on maximum loads ranges from 0.05% to 0.1% for the force components and from 0.8% to 1.2% for the moment components. The strain Tail sting and Z-adapter to link gauge bridges are connected with six-wire con- balance with support sting ductors to the carrier frequency amplifier modules of a specific measurement system (HBM

Fig. 6 Components for strain gauge balance integration MGC plus). This system is fully computer con- and connection with tail sting adapter. trolled providing also the excitation for the strain gauge bridges. The calculation of forces and moments from the balance signals employs a comprehen- sive balance calibration. The mathematical relationship assigning balance signals to given loads uses a third order approximation of the balance behavior. Pure single loads and combinations of Fig. 7 Six-component strain gauge balance. single loads are stepwise applied to the balance during the calibration. The evaluation of each The model is sting mounted using a spe- loading sequence is based on a least square error cific tail sting with its left cylinder attached to second order polynomial approximation. The the internal strain gauge balance and its right complete calibration coefficient matrix is de- conical part connected with the horizontal sting rived from those evaluations resulting in a sys- of the 3-axis model support (Figs. 5 and 6). The tem of equations. The related solution applies a computer controlled model support provides an least square error method to obtain the aerody- angle-of-attack range of −10° ≤ α ≤ +30°, and namic loads from the balance signals [22]. models may be yawed and rolled 360°. The un- certainty in angle setting is less than 0.05°. 2.3 Test Conditions and Data Processing Force measurements have been made at free 2.2 Force Measurements stream reference velocities of U∞ = 40 m/s and An internal six-component strain gauge balance U∞ = 60 m/s at ambient pressure p∞ and ambient is used to measure aerodynamic forces and mo- temperature T∞ . The corresponding Reynolds ments (Fig. 7). This strain gauge balance was numbers based on the wing mean aerodynamic 6 6 designed and fabricated at the Technical Uni- chord are Relµ = 0.97 x 10 and 1.46 x 10 , re- versity of Darmstadt. It is manufactured from spectively. Thus, turbulent boundary layers are one piece of Maraging steel 250 (1.6359). Side present at wing and control surfaces. The angle and normal forces as well as pitching and yaw- of attack is varied in the range of −10° ≤ α ≤ ing moments are measured at corresponding lat- +30° with steps of ∆α = 2°. Canard and leading- eral/vertical bending positions close to the bal- edge control surfaces are set to 0°. The conven- ance end cylinders. The axial force is measured tional trailing-edge flaps are deployed at ηK = using an axial bending station and the rolling 0°, 5°, 10°, 15°, 20°, and 30°, and adaptive moment is obtained by a parallelogram spring. trailing-edge deflections are set to ηFT = 0°, 5°, All six measurement stations are equipped with 10°, 15°, 20°, and 25°. Prior to the angle-of- advanced, self temperature compensating strain attack tests at wind-on conditions, a polar at

4 AERODYNAMIC EFFICIENCY OF HIGH MANEUVERABLE AIRCRAFT APPLYING ADAPTIVE WING TRAILING EDGE SECTION

wind-off is conducted for each model configura- This agreement is also true for the other aerody- tion. The associated loads caused by mass namic coefficients substantiating correct and forces (model weight) are subtracted from the stiff linkages between tail sting, balance and total measured loads to obtain aerodynamic model. None of the results of repeated meas- forces and moments only. Sting influence is urement runs exceed the balance error levels. considered as well. The comparison of lift coefficients for dif- The voltages of the strain gauge bridges are ferent Reynolds numbers, namely Relµ = 0.97 x 6 6 amplified for optimal signal level and simulta- 10 (U∞ = 40 m/s) and 1.46 x 10 (U∞ = 60 m/s), neously sampled at 15 Hz and digitized with 16 indicates no remarkable differences regarding bit precision. The sampling interval is 33.33 s, the angle-of-attack range tested (Fig. 9). Similar thus providing a sample block of 500 discrete tendencies are found for all other aerodynamic values for each point of the angle-of-attack po- coefficients. The data show there is no signifi- lar. Based on that samples mean values of aero- cant Reynolds number effect for that low-speed dynamic forces (lift L, drag D, side force Y) and region because turbulent boundary layers exist moments (pitching moment m, rolling moment and the flow separates at wing leading-edges l, yawing moment n) are calculated. Moments already at moderate angles of attack. Hence, fur- are related to the balance reference point WBP ther results are presented for U∞ = 40 m/s only. (Fig. 2). The corresponding aerodynamic coefficients CL , CD , CY , Cm , Cl , and Cn are referred Baseline 2 configuration to dynamic free stream pressure q∞ = (ρ∞/2) U∞ (density ρ∞ = f (T∞, p∞)), wing area AW and wing mean aerodynamic chord lµ and wing semi span s, respectively. Here, the coefficients associated down-stroke up-stroke α-motion with the longitudinal motion are discussed: CL α-motion

CL = L /( q∞ AW ) (1)

CD = D /( q∞ AW ) (2) up-stroke down-stroke

Cm = m /( q∞ AW lµ) (3)

Flap efficiencies due to changes in lift and Fig. 8 Lift coefficient as function of angle of attack for pitching moment are given by the following up-stroke and down-stroke measurement cycles. derivatives (j = K, FS): Form-variable CLηj = dCLj /dηj (4) trailing-edge

C = dC /dη (5) mηj mj j

CL 3 Results and Discussion

3.1 Preparatory Tests

The accuracy of the complete measurement system has been carefully tested. Results of the lift coefficient as function of angle of attack show a perfect agreement comparing up-stroke and Fig. 9 Lift coefficient as function of angle of attack for down-stroke angle-of-attack motions (Fig. 8). different Reynolds numbers Relµ.

5 CHRISTIAN BREITSAMTER

3.2 Flow Field Characteristics and thus, a sudden expansion of the vortex core takes place. Vortex breakdown is linked to the Three main flow field types develop sequen- transition from stable to unstable core flow evi- tially on the delta wing with increasing angle of dent by the change from jet-type to wake-type attack (Fig. 10; η = 0°). Firstly, the flow is at- FT axial velocity profiles. This transition is accom- tached on the wing upper surface up to moder- panied by maximum turbulence intensities at the ate incidences (α < 4°). Secondly, at α ≈ 4°, the breakdown position and increased turbulence flow separates at the delta wing leading-edges levels in the breakdown wake. Here, vortex with fast transition from bubble- to vortex-type. bursting for the baseline configuration (η = 0°) The roll-up of the separated shear layers starts at FT occurs at the delta wing trailing-edge at α ≈ 12°. the wing tip and progresses to the apex to form At α = 30°, vortex bursting is located close to a fully developed leading-edge vortex. The leading-edge vortices of the delta wing port and the apex. The maximum lift coefficient CLmax is starboard side induce velocities in planes nor- also reached at α ≈ 30°. mal to the free stream direction producing pro- nounced suction peaks in the wing upper surface 3.3 Aerodynamic Characteristics of Adap- pressures. This suction evokes a non-linear in- tive versus Conventional Trailing-Edge Flaps crease in lift. The present wing geometry gener- ate leading-edge vortices of moderate strength 3.3.1 Lift The lift coefficient vs. angle of attack is plotted because the wing sweep is also moderate (φw = 50°). Therefore, the non-linear part, evident in for all form-variable trailing-edge deflections in the lift curve above α > 6°, is small. With fur- Fig. 10. It is shown that positive trailing-edge ther increasing incidence the vortex rotational camber (η > 0) shifts the lift coefficient at zero core grow and vortex trajectories are shifted in- angle of attack, CL0, to markedly higher values. board and upward. Thirdly, an abrupt change in Further, there is a substantial increase in the the vortex core structure well known as vortex maximum lift coefficient CLmax. However, the bursting/breakdown occurs at a certain high an- increase in lift becomes reduced at high angle- gle of attack [23]. This phenomenon is due to of-attack because downstream of vortex burst- the rise in the adverse pressure gradient with ing regions of low velocities and irregular flow increasing angle of attack. The rise in core static arise near the surface. Also, regions of separated pressure causes stagnation of the axial core flow flow occur at the trailing-edge, especially in the

∆CLmax

Form-variable Cross-flow velocities U∞ trailing-edge over starboard wing at planes x/cr = 0.4, and 1.0, at α = 24° Vortex ∆CL0 bursting (ηFT = 0° - 25°) CL Non-linear increase in lift CLα ≈ 2.922

x/cr = 0.4 Leading-edge vortices of moderate strength

x/cr = 1.0

Fig. 10 Lift coefficient as function of angle of attack for the configuration with form-variable trailing-edge section at deflections of ηFT = 0°, 5°, 10°, 15°, 20°, and 25°.

6 AERODYNAMIC EFFICIENCY OF HIGH MANEUVERABLE AIRCRAFT APPLYING ADAPTIVE WING TRAILING EDGE SECTION

11). Even a remarkable increase in maximum lift can be detected for all investigated trailing-

∆CL (ηFS;ηK) edge deflections. The differences in lift coefficient between the configurations with form- ηFS = ηK = 20° ∆CL (ηFS;ηK) variable and conventional trailing-edge flaps increase with flap deflection. The potential of the ηFS = ηK = 10° CL adaptive wing trailing-edge section in generat- ing more lift is attributed to the smooth, variable camber based on continuously rising contour gradient and curvature. The differences in lift become markedly lower at α > 22° because vortex bursting dominates the overall flow behavior on the wing upper side. Fig. 12 depicts the differences in lift coefficient between the configu- Fig. 11 Lift coefficient as function of angle of attack rations applying form-variable and conventional comparing the configurations with form-variable trailing- trailing-edge flaps normalizing the results by the edge section vs. conventional flaps deflected at ηFS ,ηK = 5°, 10°, 15°, and 20°. values of the conventional case. For example, the form-variable trailing-edge deflected at ηFS = 15° provides higher lift of 13.4% to 2.6% in (CL(ηFS) − CL(ηK))/ CL(ηK) x 100 [%] the angle-of-attack range of 4° ≤ α ≤ 30°. 3.3.2 Drag The lift-to-drag (Lilienthal) polars demonstrate a shift in minimum drag coefficient to higher lift coefficients if the trailing-edge flap is deflected + 13.4 % (Fig. 13). It reflects the rise in glide number with increasing trailing-edge camber. The results of the form-variable cases indicate a strong reduction in drag at given lift coefficients com- + 2.6 % pared to the data of the conventional flap configuration. The lower drag is mainly caused by an improved drag coefficient at zero lift due to better flow alignment at the smooth, variable camber contour alleviating the separation ten- Fig. 12 Relative differences in lift coefficient between dency at the trailing-edge. Relative differences configurations with form-variable trailing-edge section and conventional flaps as function of angle of attack for in the drag coefficients between adaptive and ηFS , ηK = 5°, 10°, 15°, and 20°. conventional flap configurations are plotted in Fig. 14 for ηK = ηFS = 5°, and 15°. The relative outboard (wing tip) area. Therefore, flap effi- difference (CD(ηFS) − CD(ηK))/ CD(ηK) for ηK = ciency of conventional outboard flaps is typi- ηFS = 15° is −26.5% at CL = 0.4 and still −9.5% cally lower than that of inboard flaps. To at CL = 1.2. The drag reduction decreases at counter his effect an adaptive trailing-edge sec- higher angles of attack as induced drag grows tion will provide a continuous decrease of out- and wing upper flow is largely separated. board and increase of inboard deflections at Lilienthal polars related to form-variable high incidences. and conventional cases of trimmed conditions The comparison of data for form-variable are shown in Fig. 15. A reference center of vs. conventional flaps shows that smooth, vari- gravity (c.g., Fig. 2) with respect to an original able camber results in a higher lift coefficient aircraft is chosen to re-calculate the aerody- over the whole angle-of-attack range tested (Fig. namic coefficients. The graphs exhibit again a

7 CHRISTIAN BREITSAMTER

∆CD (ηFS;ηK)

(CL = const.) ∆CD (ηFS;ηK) CL (CL = const.)

CD Fig. 13 Lift coefficient as function of drag coefficient Fig. 15 Lift coefficient as function of drag coefficient for comparing the configurations with form-variable trailing- trimmed conditions comparing configurations with form- edge section vs. conventional flaps deflected at ηFS ,ηK = variable trailing-edge section vs. conventional flaps. 5°, and 15°. WBP (C (η ) − C (η ))/ C (η ) x 100 [%] Cm D FS D K D K

Vortex bursting

Form-variable trailing-edge α = 18°

Vortex bursting CL at trailing-edge − Cm0 dCm/dCL ≈ 0.185

Fig. 14 Relative differences in drag coefficient between configurations with form-variable trailing-edge section and conventional flaps as function of lift coefficient for Cm ηFS , ηK = 5°, and 15°. Fig. 16 Lift coefficient as function of pitching moment significantly improved lift-to-drag performance coefficient for the configuration with form-variable trailing-edge section at deflections of ηFT = 0°, 5°, 10°, 15°, applying variable camber trailing-edge section. 20°, and 25°. 3.3.3 Pitching Moment The nose-up pitching moment increases with reference point (WBP) causing a positive pitch- angle of attack and lift coefficient, respectively ing moment becomes reduced. Approaching (Fig. 16). The curve of the baseline configura- maximum lift, the gradient of pitching moment coefficient related to lift coefficient dCm/dCL tion (ηFT = 0°) exhibits a kink at α ≈ 18° indicating a local drop in the nose-up behavior. This changes its sign indicating the loss of lift on the kink results from the upstream shift of vortex wing upper (suction) side. bursting with increasing angle of attack. Conse- A positive trailing-edge deflection leads to quently, the amount of lift ahead of the moment the typical behavior of an increase in nose-down

8 AERODYNAMIC EFFICIENCY OF HIGH MANEUVERABLE AIRCRAFT APPLYING ADAPTIVE WING TRAILING EDGE SECTION

CLηFS > CLηK ∆Cm (ηFS;ηK)

(CL = const.)

CL CL

Fig. 17 Lift coefficient as function of pitching moment Fig. 18 Lift coefficient as function of deflection angles coefficient comparing the configurations with form- ηK and ηFS, respectively, at several angles of attack. variable trailing-edge section vs. conventional flaps deflected at ηFS ,ηK = 5°, 10°, 15°, and 20°. pitching moment, particularly in Cm0 (Figs. 16 and 17). Compared to the conventional case, the trailing-edge variable camber produces a larger nose-down pitching moment because of higher lift generated in the trailing-edge region. The Cm form-variable configuration is characterized by a quasi linear trend for the rise in the nose-down CmηFS  >  CmηK  pitching moment at zero lift, Cm0, with deflection angle ηFT. This trend can be approximated by the following relationship based on linear potential theory, with xK denoting the relative position of the variable camber section: 3 Cm0 = −1.715 √(x k (1-xk)) ηFT (6) Fig. 19 Pitching moment coefficient as function of deflection angles ηK and ηFS, respectively, at several angles of attack.

3.3.4 Flap Efficiency are found analyzing the pitching moment coef- The values of lift coefficient as function of the ficients as function of the deflection angles (Fig. deflection of conventional trailing-edge flap η K 19). and form-variable trailing-edge section η , res- FS For further evaluation, the required form- pectively, are plotted for several angles of attack variable trailing-edge deflections η and in Fig. 18. It is shown that the gradient of the FS-CL-eq η to achieve equivalent increase in lift curves, C , is larger for the form-variable cases FS-Cm-eq Lη and pitching moment, respectively, are com- compared to the associated conventional flap puted: deflections. In particular, increased gradients of -1 CLη are evident for both large deflection angles ηFS-CL-eq =(CLηFS) (CLηK) ·1[°] (7) and at high lift coefficients, substantiating an -1 enhanced flap efficiency obtained by adaptive ηFS-Cm-eq=(CmηFS) (CmηK) ·1[°] (8) wing geometry. Corresponding characteristics

9 CHRISTIAN BREITSAMTER

periments the form-variable trailing-edge sec- C Lη tion is exchanged for the conventional inboard and outboard flaps. The configurations tested form-variable include six discrete elements representing smooth, variable camber contours with trailing- edge deflections in the range of 0° to 25°. Con- ventional flaps are deflected from 0° to 30°. The comparison of data is based on equivalent de- conventional flection angles with respect to the trailing-edge skeleton line. The comparison of the measured aerodynamic forces and moments between the configu- (CLηK/CLηFS) = 0.761 rations of adaptive wing and conventional trailing-edge flaps demonstrates that the adaptive wing configuration exhibits both significantly higher lift coefficients at given angles of attack Fig. 20 Flap efficiency CLη as function of angle of attack and substantially lower drag coefficients at comparing configurations with form-variable trailing- given lift coefficients. This improvement in edge section vs. conventional flaps. aerodynamic performance is further reflected by

increased efficiencies in altering lift and pitch- Improved flap efficiencies of the adaptive wing ing moment over the regarded angle-of-attack result in lower trailing-edge deflections η FS-CL-eq range of −10° to +30°. The adaptive wing aero- and η with respect to the conventional FS-Cm-eq dynamic potential enables also effective maneu- flap deflections η for comparative alteration of K ver load alleviation. lift or pitching moment. For example, the form- variable trailing-edge deflection is only about 75% of the conventional one to obtain an Acknowledgment equivalent increase in lift at angle of attack of α = 24° (Fig. 20). The decrease in the gradients This work was supported by the EADS GmbH (Military Aircraft). The author would like to ap- CLη and Cmη, typically observed for conventional trailing-edge flaps at high angle-of-attack, is preciate the long period of fruitful cooperations markedly lower for the adaptive wing. with Dr.-Ing. J. Becker (EADS MT2).

3 Conclusions References The aerodynamic characteristics of a high ma- [1] Hilbig H, and Wagner H. Variable wing camber control for civil transport aircraft. Proceedings of the 15th neuverable aircraft of canard-delta wing type Congress of the International Council of the Aero- have been intensively investigated focusing on nautical Sciences, Toulouse, France, ICAS-84-5.2.1, the effectiveness of a form-variable trailing- pp 107-112, 1984. edge section. Aerodynamic efficiencies related [2] Renken J H. Mission adaptive wing camber control to smooth, variable camber trailing-edge ge- systems for transport aircraft. AIAA Paper 85-5006, ometry are compared with those achieved with Oct. 1985. conventional trailing-edge flaps. The compari- [3] Greff E. The development and design integration of a variable camber wing for long/medium range aircraft. son of flight and maneuver performance is The Aeronautical Journal, Vol. 94, No. 939, pp 301- based on force and flow field measurements. 312, 1990. Aerodynamic forces and moments are measured [4] Spillmann J J. The use of variable camber to reduce applying an advanced internal six-component drag, weight and costs for transport aircraft. The strain gauge balance connected to a high accu- Aeronautical Journal, Vol. 96, No. 951, pp 1-9, racy measurement system. To perform the ex- 1992.

10 AERODYNAMIC EFFICIENCY OF HIGH MANEUVERABLE AIRCRAFT APPLYING ADAPTIVE WING TRAILING EDGE SECTION

[5] Mertens J. Some important results of the technology [17] Bonnema K, and Smith S. AFTI/F-111 Mission program RaWid. New Results in Numerical and Ex- Adaptive Wing Flight Research Program. AIAA Pa- perimental Fluid Mechanics II, Notes on Numerical per 88-2118, May 1988. Fluid Mechanics, Vol. 72, vieweg, pp. 315-322, [18] Smith S B, and Nelson D W. Determination of the 1999. aerodynamic characteristics of the mission adaptive [6] Stanewsky E. Aerodynamic benefits of adaptive wing wing. AIAA Journal of Aircraft, Vol. 27, No. 11, pp. technology. Aerospace Science and Technology, Vol. 950-958, 1990. 4, No. 5, pp 439-452, 2000. [19] Goecke Powers S, Webb D L, Friend E L, and Lokos [7] Henke R. ProHMS – Prozesskette Hochauftrieb mit W A. Flight test results from a supercritical mission multifunktionalen Steuerflächen. Proceedings of the adaptive wing with smooth variable camber. NASA- German Aerospace Congress 2002, Stuttgart, Ger- TM-4415, Nov. 1992. many, DGLR-2002-094, 2002. [20] Thornton S V. Reduction of structural loads using [8] Henke R. Forschung in Europa: Ergänzung der natio- maneuver load control on the Advanced Fighter nalen Arbeiten und Wegbereitung ins Produkt. 13th Technology Integration (AFTI)/F-111 Mission Adap- DGLR Symposium of STAB, Technische Universität tive Wing. NASA-TM-4526, Sept. 1993. München, Munich, Germany, 2002. [21] Becker J, and Luber W. The role of buffeting on the [9] Monner H P, Breitbach E, Bein Th, and Hanselka H. structural design of Eurofighter. AIAA Paper 2003- Design aspects of the adaptive wing – the elastic 1889, April 2003. trailing edge and the local spoiler bump. The Aero- [22] Hufnagel K, and Ewald B. Force testing with internal nautical Journal, Vol. 104, No. 1032, pp 89-95, strain gage balances, DGLR Symposium on Cryo- 2000. genic Wind Tunnel Testing, Technische Universität [10] Birkemeyer J. Widerstandsminimierung für den München, Munich, Germany, 2001. transsonischen Flügel durch Ventilation und adaptive [23] Breitsamter C. Turbulente Strömungstrukturen an Konturbeule. Deutsches Zentrum für Luft- und Flugzeugkonfigurationen mit Vorderkantenwirbeln. Raumfahrt, Technical Report DLR-FB-99-28, 1999. Dissertation, ISBN 3-89675-201-4, Herbert Utz Ver- [11] Sommerer A, Lutz Th, and Wagner S. Numerical lag Wissenschaft, 1997. optimisation of adaptive transonic airfoils with vari- nd able camber. Proceedings of the 22 Congress of the International Council of the Aeronautical Sciences, Harrogate, UK, ICAS-2000-2.11.1, 2000. [12] Pritschow G, and Wadehn W. Design of an adaptive wind tunnel mock-up with shape memory alloys. Proceedings of the 8th International Conference on New Actuators, Actuator 2002, Bremen, Germany, 2002. [13] Müller D, and Reiners D. Adaptiver Tragflügel, In- tegration des Hornkonzepts in ein Landeklappensys- tem. Proceedings of the German Aerospace Congress 1999, Berlin, Germany, DGLR-JT99-048, 1999. [14] Campanile L F, Sachau D, and Seelecke S. Ein adap- tives Windkanalmodell für einen adaptiven Tragflü- gel. Proceedings of the German Aerospace Congress 1999, Berlin, Germany, DGLR-JT99-050, 1999. [15] Recksiek M, Hoffmann M, Schmitt N, Lembcke T, and Schievelbusch B. Entwurf eines Hochauftriebs- systems zur Ansteuerung multifunktionaler Steuer- flächen. Proceedings of the German Aerospace Con- gress 2002, Stuttgart, Germany, DGLR-2002-098, 2002. [16] Dargel G, Hansen H, Wild J, Streit T, Rosemann H, and Richter K. Aerodynamische Flügelauslegung mit multifunktionalen Steuerflächen. Proceedings of the German Aerospace Congress 2002, Stuttgart, Ger- many, DGLR-2002-096, 2002.