Chinese Journal of Aeronautics, (2017), 30(3): 988–1003

Chinese Society of Aeronautics and Astronautics

& Beihang University
Chinese Journal of Aeronautics
cja@buaa.edu.cn
www.sciencedirect.com

Aerodynamic optimization and mechanism design of

flexible variable camber trailing-edge flap
Weishuang LU a, Yun TIAN b,*, Peiqing LIU a

a
School of Aeronautic Science and Engineering, Beihang University, Beijing 100083, China
b
National Laboratory for Aeronautics and Astronautics, Beihang University, Beijing 100083, China

Received 28 March 2016; revised 27 October 2016; accepted 14 December 2016

Available online 20 April 2017

KEYWORDS Abstract Trailing-edge flap is traditionally used to improve the takeoff and landing aerodynamic
Aerodynamic optimization; performance of aircraft. In order to improve flight efficiency during takeoff, cruise and landing
GA (W)-2 airfoil; states, the flexible variable camber trailing-edge flap is introduced, capable of changing its shape
Mechanism design; smoothly from 50% flap chord to the rear of the flap. Using a numerical simulation method for
Trailing-edge flap; the case of the GA (W)-2 airfoil, the multi-objective optimization of the overlap, gap, deflection
Variable camber angle, and bending angle of the flap under takeoff and landing configurations is studied. The opti-
mization results show that under takeoff configuration, the variable camber trailing-edge flap can
increase lift coefficient by about 8% and lift-to-drag ratio by about 7% compared with the tradi-
tional flap at a takeoff angle of 8°. Under landing configuration, the flap can improve the lift coef-
ficient at a stall angle of attack about 1.3%. Under cruise state, the flap helps to improve the lift-to-
drag ratio over a wide range of lift coefficients, and the maximum increment is about 30%. Finally,
a corrugated structure–eccentric beam combination bending mechanism is introduced in this paper
to bend the flap by rotating the eccentric beam.
Ó 2017 Chinese Society of Aeronautics and Astronautics. Production and hosting by Elsevier Ltd. This is
an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction ever-changing, but the shape of the wing is almost unchanged.

Aircraft wing design generally takes the efficiency of the cruise
flight and the high-lift performance at takeoff and landing into
consideration. In the actual flight, the flight condition is
* Corresponding author.
E-mail address: aircraft@buaa.edu.cn (Y. TIAN).
Peer review under responsibility of Editorial Committee of CJA.
Production and hosting by Elsevier
In order to improve the efficiency of the mission profile of the
flight, a mission-adaptive wing would be ideal. At present, one
feasible method for improving mission efficiency is to install a
flexible variable camber trailing-edge flap on the wing. Such a
device, combined with the large aircraft high-lift device
concept and structure-deformation technology, has great
application prospects in aircraft wing design.
Trailing-edge high-lift devices have been widely used on
many kinds of aircraft previously. The structure-deformation
technology has also been regarded as promising in the field
of aircraft design. Traditional high-lift devices have a prece-
dent of using the concept of deformation, which is mainly

http://dx.doi.org/10.1016/j.cja.2017.03.003
1000-9361 Ó 2017 Chinese Society of Aeronautics and Astronautics. Production and hosting by Elsevier Ltd.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Aerodynamic optimization and mechanism design 989

applied to the rear of wing, in order to improve the perfor-

mance of takeoff and landing. Among such devices, the
‘‘Smart High-Lift Devices for Next Generation Wings
(SADE)” cooperative research project among Airbus, the
German Aerospace Center (DLR), and 11 other European
institutions, is most representative of studies taking place
outside of China. Smart leading edge (SLE) and smart single
slotted flap (SSSF) are studied respectively in the project,
and stress analyses of flexible variable camber mechanisms
and skins are carried out, but there are less aerodynamic data
available for these two kinds of flexible variable camber
device.1–8 Aeroelastic analyses have been carried out by Li
et al.9 using the SADE concept. The ‘‘Variable Camber Con-
tinuous Trailing Edge Flap (VCCTEF)”10–13project launched
by NASA is another representative study related to deforma-
tion technology and aircraft high-lift devices. Preliminary
results show that the VCCTEF can increase the lift-to-drag
ratio by about 4.85%. Analyses of the mechanism behind Fig. 1 Baseline configurations from Ref.25.
VCCTEF and the rigidity and aeroelasticity of the material
have been carried out by Eric Ting and Sonia Lebofsky respec-
tively.14–18
In addition, Yokozeki et al.19,20 designed the variable cam-
ber morphing airfoil using corrugated structure to change the
trailing-edge camber of a wing. Similarly, the variable-camber
compliant wing is being studied by Joo et al.21,22
In China, Yin23 and Chen et al.24 have carried out similar
research, mainly focusing on the aerodynamic performance
of the trailing edge of variable camber airfoils.
Based on the existing research concerning high-lift devices
and structure-deformation technology, the optimization of flap Fig. 2 Geometry of variable camber trailing-edge flap.
position parameters and the bending angle of the flexible vari-
able camber flap of the GA(W)-2 airfoil at takeoff and landing
is studied in this paper. The aerodynamic characteristics of the
variable camber flap in cruise configuration are investigated.
Finally, a corrugated structure and eccentric beam combina-
tion bending mechanism is designed in this paper, capable of
bending the flap by rotating the eccentric beam.

2. Model design

In this paper, the GA (W)-2 airfoil is selected as an analytical

model; this is an advanced airfoil for general aviation with a
maximum thickness of about 13% of its chord c.25
The baseline configuration of the takeoff and landing con-
figuration25 is shown in Fig. 1, and df is the deflection angle of
the flap, xp is the translation amount in the horizontal direc-
tion and zp is the translation amount in the vertical direction,
which are refer to Ref.25. In this paper, the geometry of the
variable camber trailing-edge flap is shown in Fig. 2. The trail-
ing edge of the flap bends flexibly starting at 50%. Here, dangle
is the bending angle of the flap, cflap is its chord length, and O/
L is the overlap between the main wing and the flap; gap refers
to the width of seam, and d is the deflection angle of the flap.
In order to find the appropriate bending angle and flap
position parameters (overlap, gap, and d), the takeoff and
landing configurations are optimized separately in the iSIGHT
optimization platform, as shown in Fig. 3. The optimization
objective26,27 of the takeoff configuration is the lift coefficient
CL8, and the lift-to-drag ratio, CL8/CD8, is at an attack angle of
8°. The optimization objectives of the landing configuration
are the lift coefficients CL8 and CL12 at attack angles of 8°
Fig. 3 CFD optimization platform architecture based on
iSIGHT.
and 12°, respectively. As shown in Fig. 428, a is the attack
angle, CL is lift coefficient and CL0 is lift coefficient at attack
angle of 0°. For most of the general aircraft, 8° is the normal
attack angle for aircraft takeoff and landing, and 12° is near
the stall angle of attack. The multi-island genetic algorithm
is selected for optimization, and the optimization variables
and objectives are shown in Table 1.
A certain type of general aviation aircraft is chosen to verify
the 2D results, and the main parameters of this aircraft are
shown in Fig. 5 and listed in Table 2. The GA (W)-2 airfoil
is selected as the wing cross-section and maintains equal pro-
portions upon stretching into three dimensions. Here, the flap
chord accounted for 25% of the local wing chord and the flap
was stretched to 70% of the span.
990 W. LU et al.

Table 2 Geometric parameters of

selected general aviation aircraft.
Geometric parameter Value
Fuselage width (m) 2.04
Fuselage length (m) 14.88
Span length L (m) 19.2
Chord length of wing root B0 (m) 2.25
Chord length of wing tip B1 (m) 1.08
Wing reference area S (m2) 32
Aspect ratio 11.5
Ratio of B0/B1 2.08
1/4 chord sweep angle (°) 0
Twist angle (°) 5
28
Fig. 4 Angle of attack limited to fuselage near ground . Dihedral angle (°) 2.6
Incidence angle (°) 3.2

Table 1 Optimization parameters and objectives.

Parameter Maximum objective
order to verify the reliability of the numerical simulation, the
0:02c 6 O=L 6 0:04c; Takeoff: CL8, CL8/CD8; 30P30N airfoil is selected to verify the calculation. The grid
0:01c 6 Gap 6 0:05c; Landing: CL8, CL12 topology of the numerical simulation is shown in Fig. 6, and
10 6 d 6 40 ;
the first-layer grid height of the wall is 1
105 times the ref-
10 6 dangle 6 10
erence chord length.
On the basis of experimental data, the Mach number is set
to 0.2 and the Reynolds number Re is 9
106. The model is
placed at angles of attack in the range of 0°–24°. The pressure
is one atmosphere, and the reference chord length is 1.9.
The FLUENT solution is set to an implicit algorithm,
namely the coupled solver. The S-A model is used for turbu-
lence, which deals with the problem of air flow with a wall
boundary.29,30 The momentum and turbulent kinetic energy
in the equation are two-order upwind schemes. Fig. 7 shows
a comparison of the lift coefficient CL obtained in numerical
simulation with wind tunnel test results.31 At an attack angle
below 19°, the total lift coefficient of the airfoil increases with
the angle of attack, agreeing well with experimental results.
The aerodynamic forces of the leading-edge slat wing and main
body and trailing-edge flap are also in good agreement with
experiment. When the angle of attack is greater than 19°, the
calculated lift coefficient of the leading-edge slat wing is larger,
making the calculated lift coefficient of the multi-airfoil
become relatively larger, and the stall angle of attack is around
23°, which represents an increase of about 2° compared with
the experimental data.
Fig. 831 shows a comparison of the pressure coefficient dis-
tribution, Cp, obtained by numerical simulation and by wind
tunnel test results when the angle of attack is 8°. The
pressure-coefficient distributions of the main body, flap, and
slat are in good agreement.
In order to further verify the reliability of the numerical
simulation, the velocity profiles of the cross-section of the air-
Fig. 5 Geometry of 3D model. foil measured along the direction of a straight line at four rep-
resentative positions (x = 0.45c, x = 0.89c, x = 1.03c,
The geometric parameters of the wing are defined in
x = 1.11c) are compared, as shown in Fig. 931 when the angle
Table 3. The design points of this aircraft are shown in
of attack is 8°. Here, ordinate refers to the distance along the
Table 3.
vertical direction from the surface of the airfoil (n/c) and
abscissa refers to the ratio between measured and farfield
3. CFD verification velocities (V/V1).The wakes of the main body obtained by
both experiment and numerical simulation can be seen from
In this paper, the structure grid is used, and the computational Fig. 9 and are found to be in good agreement. Numerical sim-
fluid dynamics software Fluent is applied as the solver. In ulation of the speed loss of the wake of the slat clearly exceeds
Aerodynamic optimization and mechanism design
Table 3 Design point parameters of selected general aviation aircraft.
Configuration Typical altitude (km) Typical speed (km/h)
Takoff & landing 0 45
Cruise 3 110
Fig. 6 Close-up of grid for 30P30N airfoil.
Fig. 7 Comparison of lift coefficients distribution obtained in
numerical simulation with wind tunnel test 31.
Fig. 8 Comparison of pressure coefficient distributions obtained
in numerical simulation with wind tunnel test 31.
its experimental value. When the angle of attack is 8°, the
boundary layer of the main section (x = 0.45c) is plump and
uniform, whereas that of the flap section is not, which is con-
sistent with the pressure-coefficient curves. The flap section
(x = 0.89c, x = 1.03c, x = 1.11c) shows an inverse-pressure
gradient trend in the flap. The wake of the main body appears
991
Reynolds number Mach number
2.2
106 0.130
11
106 0.313
to experience a slight separation at the flap, with variable
thickness and wake velocity reduction with the growth of
chord length. Overall, the results of the numerical simulation
are in good agreement with experiment.
In summary, the results of numerical simulation are
credible.
4. Results
4.1. Results under takeoff configuration
4.1.1. 2D results
The Pareto optimal solutions of the takeoff configuration are
shown in Fig. 10. According to the optimization objective,
the lift coefficient and lift-to-drag ratio of the optimized con-
figuration are better than those of the baseline configuration.
Three typical results (A, B, and C) on the Pareto Frontier
are selected for further analysis. The geometries of the different
takeoff configurations are shown in Fig. 11, and the geometric
parameters of the selected takeoff configurations are shown in
Table 4. In order to simplify the legend, ‘‘base” is used instead
of ‘‘baseline” in the figure below.
The aerodynamic performances of selected takeoff configu-
rations is shown in Fig. 12, and the lift coefficient and the max-
imum lift-to-drag ratio of the three optimized takeoff
configurations are greater than the baseline configuration
before the stall angle.
In the lift coefficient linear section, the lift coefficients of
configurations A and B are larger than that of the baseline
configuration, and the increment of lift coefficient is about
0.2. The stall angles of attack of configurations A and B are
decreased by 1° compared with the baseline takeoff configura-
tion. Configuration C basically maintains the original stall
angle of attack, and the lift coefficient is 0.1 higher than that
of the baseline configuration.
Compared with the baseline configuration, the lift-to-drag
ratios of the three optimized configurations have been slightly
increased. The maximum lift-to-drag ratio of configurations A
and C is at 8°, whereas that of configuration B is at 4°. At an
angle of 8°, the drag coefficient is obviously increased, leading
to a decrease of the lift-to-drag ratio. When the angle of attack
increases, the drag coefficient is sharply increased to be near
the stall angle of attack (13°), and the lift-to-drag ratio is
decreased.
The pitching moments Cm of the three optimized takeoff
configurations are higher than that of the baseline configura-
tion. Here, the pitching moment of configuration B is the
largest.
The pressure distributions of different takeoff configura-
tions at 8° are shown in Fig. 13. It can be seen that the suction
peaks of the front of main body and the flap of the optimized
takeoff configurations are higher than that of the baseline con-
figuration. Because of the change of the camber and seam
992 W. LU et al.

Fig. 9 Velocity profile of cross-section of airfoil.31

parameters of the flap, the effective camber of the airfoil is

changed, leading to a significant increase in the negative-
pressure value of the flap upper surface, especially for config-
urations B and C. The negative-pressure value of the upper
surface and the positive-pressure value of lower surface of
the main body are higher than those under the baseline config-
uration. The local picture of the streamlines and velocity mag-
nitudes of different takeoff configurations at 14° are shown in
Fig. 14. It can be seen that the main body and flap maintain an
attached flow well without flow separation, except at the cavity
between them. Here, the flow velocity of configuration A is sig-
nificantly higher than those of the other three configurations,
leading to an increase in negative-pressure value of the upper
surface and the positive-pressure value of the lower surface,
so that the lift coefficient of configuration A is higher than
Fig. 10 Pareto optimal solutions of takeoff configuration. the others at a ¼14°.
It can be seen that the effective camber of configuration A is
greater. The lift coefficient and lift-to-drag ratio before the
stall angle are improved because of the small increase of the
effective camber. And the pitching moment of configuration
A is modest. Therefore, configuration A is chosen as the opti-
mal takeoff configuration.

4.1.2. 3D results
The locations of the spanwise stations are shown in Fig. 15. By
stretching the wing sections given by configuration A and the
baseline configuration into three dimensions while maintaining
equal proportions, the wing and fuselage are determined.
Because the twist angle is set to 5°, the geometrical shapes
Fig. 11 Geometries of different takeoff configurations.
Aerodynamic optimization and mechanism design
Table 4 Geometrical parameters of different takeoff configurations.
Configuration O/L (c)
Base 0.025
A 0.010
B 0.004
C 0.005
Fig. 12 Aerodynamic performance of selected takeoff configurations.
Fig. 13 Pressure distributions of different takeoff configurations
at a ¼8°.
of different spanwise positions are shown in Fig. 16. The half
models and surface meshes are shown in Fig. 17.
The 3D results of the optimal and baseline takeoff configu-
rations are shown in Fig. 18. In the 3D case, the growth trend
of the aerodynamic performance is similar to that of the 2D
results. Because of the presence of the incidence angle of the
wing and the washing effect of the fuselage, the wing and fuse-
lage stall angle of attack is smaller by about 4° in the 3D case.
The pressure coefficient distributions of various takeoff
configurations are shown in Fig. 19.The negative pressure val-
ues of the main upper surface decrease gradually from root to
tip, and the pressure distribution of the flap remains basically
unchanged. Compared with the baseline configuration, the
suction peak, the upper-surface negative-pressure value, and
lower-surface positive-pressure value of the main body and
flap are larger, as was the case for the 2D calculation results;
993
Gap (c) d (°) dangle (°)
0.040 10 0
0.028 12 4
0.024 21 3
0.036 14 1
this is especially true in case of the suction peak at the wing
tip of the optimized takeoff configuration.
There are slight differences between the baseline configura-
tion and the optimized takeoff configurations in terms of
streamlines and pressure, as shown in Fig. 20. Only in the indi-
vidual position, the upper-surface pressure of the optimized
configuration is slightly lower.
4.2. Results under landing configuration
4.2.1. 2D results
The Pareto optimal solutions of the landing configurations are
shown in Fig. 21. According to the optimization objective for
the landing configuration, three better landing configurations
A, B, and C are selected from the optimization result. A com-
parative analysis of the aerodynamic characteristics of the
three optimized landing configurations and the baseline land-
ing configuration is studied. The geometries of the landing
configuration are shown in Fig. 22, and the geometric param-
eters of the different takeoff configurations are shown in
Table 5.
The aerodynamic performances of different landing config-
urations is shown in Fig. 23. It can be seen that the lift coeffi-
cient of configuration C in the linear section is larger than the
others by 0.15, but the stall characteristics of configuration C
are clearly inferior to the others in terms of the stall angle of
attack.
When the angle of attack is less than 8°, the drag coefficient
of configuration A is slightly less than that of the baseline land-
ing configuration. Therefore, the advantages of configuration
A in terms of the lift-to-drag ratio are obvious, and this config-
uration, having the maximum lift-to-drag ratio, maintains this
advantage up to the stall angle. However, the lift-to-drag ratios
of configurations B and C are lower than that of the baseline
994
Fig. 14 Local pictures of streamlines and velocities for different takeoff configurations at a ¼14°.
Fig. 15 Locations of spanwise stations.
configuration. When the angle of attack is about 12°, these lift-
to-drag ratios decrease sharply.
The pitching moments of the three optimized landing con-
figurations are higher than that of the baseline configuration.
Here, the growth of the pitching moment of configuration A
is lower.
The pressure distributions of different landing configura-
tions at a ¼8° are shown in Fig. 24. It can be seen that the suc-
tion peak, negative-pressure value of the upper surface, and
positive-pressure value of the lower surface of the main bodies
Fig. 16 Geometries of takeoff configuration at different spanwise stations.
W. LU et al.
of configurations A and C are higher than those of the baseline
configuration. The negative-pressure values of the flap fronts
of configurations A and C are lower than that of the baseline
configuration.
The local streamlines and velocity contours for different
landing configurations at a = 12° are shown in Fig. 25. It
can be seen that the upper surface of configuration C produces
serious flow separation. Moreover, the flow velocities of other
two optimized landing configurations increase slightly com-
pared with the flow velocity of the baseline configuration.
It can be seen that the effective camber of configuration A is
slightly increased. The small increase in the effective camber
not only improves both the separation at the trailing edge of
the flap at a small angle of attack and the lift coefficient and
lift-to-drag ratio below the stall angle but also maintains the
stall angle of attack of the baseline landing configuration.
Therefore, configuration A is chosen as optimal for landing
conditions.
4.2.2. 3D results
By stretching the wing sections given by configuration A and
the baseline configuration into three dimensions while main-
taining equal proportions, the wing and fuselage are deter-
mined The geometrical shapes of the cross-sections at
different spanwise positions are shown in Fig. 26. The half
models and surface mesh are shown in Fig. 27.
The 3D calculation results for the optimal and baseline
landing configurations are shown in Fig. 28, and they appear
Aerodynamic optimization and mechanism design 995

Fig. 17 Half models and surface mesh of takeoff configuration.

Fig. 18 Aerodynamic performances of different takeoff configurations in 3D.

Fig. 19 Pressure coefficient distributions of takeoff configurations at a = 4°.

Fig. 20 Streamlines and pressure of takeoff configurations at a = 8°.

996 W. LU et al.

Fig. 21 Pareto optimal solutions for landing configuration.

Fig. 24 Pressure distributions of different landing configurations

at a = 8°.

It can be seen from Fig. 29 that the negative-pressure peak

Fig. 22 Geometries of different landing configurations.
similar to the results for the takeoff configuration. In the 3D
case, the optimized configurations still retain their advantages
in terms of lift coefficient and lift-to-drag ratio. The 3D stall
angle of attack is smaller by about 3°, and the lift coefficient
of the 3D wing is significantly smaller than that of the 2D air-
foil. The lift-to-drag ratio is decreased and the pitching
moment is reduced.
decreases gradually from root to tip, and the negative-pressure
value on the flap’s upper surface gradually increases. For an
optimized landing configuration, the pressure distribution on
the main body is similar to that of the baseline configuration,
but the negative pressure on the flap value is decreased at the
middle of the wing, thereby reducing the inverse-pressure gra-
dient on the flap’s upper surface and delaying the flow separa-
tion (Fig. 30).
From the picture of the pressure contours and streamlines
at each individual position at a ¼7° (Fig. 30), it can be seen
that the upper-surface pressure of the optimized configuration
is slightly lower. Flow separation appears at the flap. For the
baseline configuration, the flow separation position is at the
front of the flap, and for optimized configurations this separa-
tion appears at the rear of flap.

Table 5 Geometric parameters of different landing configurations.

Configuration O/L (c) Gap (c) d (°) dangle (°)
Base 0.005 0.035 30 0
A 0.002 0.028 23 7
B 0.02 0.036 32 5
C 0.006 0.0265 30 10

Fig. 23 Aerodynamic performances of different landing configurations.

Aerodynamic optimization and mechanism design 997

Fig. 25 Local pictures of streamlines and velocities for different landing configurations at a = 12°.

Fig. 26 Geometries of landing configurations at different spanwise stations.

Fig. 27 Half models and surface mesh of landing configuration.

Fig. 28 Aerodynamic performances of various landing configurations in 3D.

998 W. LU et al.

Fig. 29 Pressure coefficient distributions of landing configurations at a = 4°.

Fig. 30 Streamlines and pressures of landing configurations at a = 7°.

4.3. Results under cruise configuration

4.3.1. 2D results
Under cruise configuration, the different conditions for the
bending angle of the flexible variable-camber flap are calcu-
lated separately using the method of computational fluid
dynamics. The values of the bending angle are 0°, 6°, 10°,
6°, and 10°, and the geometries of the cruise configurations
with different bending angles are shown in Fig. 31.
Fig. 31 Geometries of variable-camber trailing edge flap in
The effects of the variable-camber trailing-edge flap upon
cruise configuration.
the aerodynamic force in the cruise configuration are shown
in Fig. 32, and as the lift coefficient and drag coefficient

Fig. 32 Effects of variable-camber trailing-edge flap upon aerodynamic force in cruise configuration.
Aerodynamic optimization and mechanism design
increase, the stall angle decreases with the increase of flap cam-
ber. Below the stall angle, the lift coefficient undergoes a clear
increase when bending angle is greater than 0°, and the greater
the value of dangle is, the larger the increment of the lift coeffi-
cient will be. But this increment is close to 0 when the attack
angle is near the stall angle of attack, and even afterward,
the lift coefficient presents negative growth. At the same time,
the increase of the flap camber leads to flow separation in
advance, reducing the stall angle of attack by 2° when the
bending angle is 10°. The pitching moment increases along
with the effective airfoil camber.
From the lift-to-drag ratio curve, the increase of effective
airfoil camber reduces the angle of attack corresponding to
the maximum lift-to-drag ratio. When the bending angle is
greater than 6°, the increase of the camber is conducive to
enhancing the maximum lift-to-drag ratio, but doing so will
be harmful. From Fig. 32(c), which show the envelope of the
lift-to-drag ratio, it can be seen that this ratio can be improved
with the lift coefficient ranging from 0 to 2 by changing the flap
bending angle during the cruise.
The effect of the variable camber trailing-edge flap upon the
pressure coefficient distribution in the cruise configuration at
8° is shown in Fig. 33. It can be seen that negative-pressure
value of the upper surface and the positive-pressure value of
the lower surface are higher for the optimized configurations
than those for the baseline configuration when the bending
angle is greater than 0°, leading to an increased inverse-
pressure gradient between the upper and lower surfaces. More-
over, there is no sudden pressure change at the bending posi-
tion. Compared with the traditional flap and aileron, a
flexible variable camber flap reduces the loss of lift at the
deflection position.
4.3.2. 3D results
Using the GA(W)-2 airfoil as a wing section and stretching it
into three dimensions whilst maintaining the same proportions,
Fig. 33 Effect of variable camber trailing-edge flap on pressure-
coefficient distribution in cruise configuration.
Fig. 34 Geometry of cruise configuration at different spanwise stations.
999
the wing and fuselage are determined. The incidence angle is set
to 3.2° and the twist angle is set to 5°. The geometrical shapes
at different spanwise positions are shown in Fig. 34. The half
model and surface meshes are shown in Fig. 35.
The 3D calculation results for the cruise configuration are
shown in Fig. 36, and appear to be similar to those for the
takeoff and landing configurations. The 3D stall angle of
attack is smaller by about 3° than the 2D case, and the lift
coefficient of the 3D wing is significantly smaller than that of
the 2D airfoil. The lift-to-drag ratio is also decreased.
It can be seen from Fig. 37 that negative-pressure value on
upper-wing surface decreases gradually from root to tip. As
with the 2D calculation results, the larger the bending angle
is, the larger the negative-pressure value on upper surface
and the positive-pressure value on lower surface are; this is
especially true at the wing tip because of spanwise flow.
The pictures of the pressure contours and streamlines, as
shown in Fig. 38, indicate that larger bending angles corre-
spond to lower upper-surface pressure values on the wing.
The flow separation is most serious when the bending angle
is 10°.
5. Mechanism design
Throughout the above analysis of the effect of a flexible vari-
able camber flap upon airfoil aerodynamic performance and
the optimization results, the bending angle was not assumed
to be too large. At the same time, stress and strain analyses
of skin materials in the literature4 have indicated that skin is
prone to wrinkling with bending angles beyond 8°. Therefore,
in the design of the bending mechanism, the bending angle is
set from 8° to 8°.
In the design of the bending mechanism for the flexible
variable-camber trailing-edge flap, as shown in Fig. 39, a cor-
rugated structure is used as the stringer in the skin (Fig. 39(b))
and an eccentric beam is used as the deformation mechanism
from the 50% flap chord to the rear of the flap (Fig. 39(c)).
Here, the corrugated structure is not in the stress state, and
the flap is not distorted.
The corrugated structure is strongly deformable along the
chordwise direction but strongly resistant to deformation
along the spanwise direction. This mechanism can decrease
the drive force required for the realization of flap deformation.
It can also better maintain the flap shape along the spanwise
direction and decrease the number of eccentric beams, thereby
decreasing its necessary weight.
In order to realize the flexible deformation of the flap, it is
necessary to use the deflection curve of the skin material. The
deflection is the centroid of the cross-section along the vertical
axis, and the direction of the line displacement refers to the
bending deformation. The axis of the beam will be changed
into a plane curve in the longitudinal plane of the beam, which
1000 W. LU et al.

Fig. 35 Half models and surface mesh of cruise configuration.

Fig. 36 Aerodynamic performances of cruise configurations.

Fig. 37 Pressure coefficient distribution of cruise configurations at 4°.

Fig. 38 Streamlines and pressure distributions of cruise configurations at a = 10°.

Aerodynamic optimization and mechanism design 1001

Fig. 39 Bending mechanism.

32
Fig. 40 Deflection curve .

Fig. 41 Eccentric beam.

is known as the beam’s deflection curve.31 This curve is shown

Fig. 42 Deformation of flap.
in Fig. 40.32 The l is the length of curve in the horizontal direc-
tion and y is the deformation in the vertical direction, which
can be seen from the Ref.32. The functional expression of
deflection curve, mðxÞ, is usually expressed up to third order,
as in Eq. (1):
mðxÞ ¼ Ax3 þ Bx2 þ Cx þ D ð1Þ
The constants A and B are related to the material properties
and stress conditions whereas C and D are arbitrary.
The eccentric beam parameters are shown in Fig. 41. This Fig. 43 Geometry shape of deformation of flap.
design guarantees that the curve of flap shape does not mani-
fest second-order discontinuous points over the whole defor-
mation process, and also ensures a smooth transition in the 6. Conclusions
pressure distribution under flap deformation. The eccentric
beam is rotated 90° upward and downward, which respectively In this paper, the optimization of the takeoff and landing con-
raises and presses the trailing edge of the flap to realize vari- figurations of the GA (W)-2 airfoil with a 25%c flexible vari-
able camber from 50% of the flap chord, as shown in Fig. 42. able camber trailing-edge flap was carried out. This trailing-
The use of the bending mechanism combining a corrugated edge flap can smoothly change its shape from 50% of its chord
structure and eccentric beam can achieve the requirements of to its rear. The optimization variables included the overlap,
variable camber from 50% of the flap chord, and the maxi- gap, deflection angle of the flap, and bending angle. Under
mum bending angle is 8°. Fig. 43 shows the geometric shapes the takeoff configuration, the optimization objective was to
of the flap bent 8° upward and downward. maximize the lift coefficient, CL8, and the lift-to-drag ratio,
1002
CL8/CD8, at a takeoff angle of 8°. Under the landing configu-
ration, the optimization objective was to maximize the lift
coefficient CL8 at an attack angle of 8° and the lift coefficient
CL12 at an attack angle of 12°. The influence of the bending
angle upon the aerodynamic performance in the cruise state
has been analyzed. A bending mechanism has been designed
according to the deformation characteristics of the flexible
variable camber trailing-edge flap.
(1) For the optimized takeoff configuration, the effective
camber of the airfoil was increased slightly, which
improved the lift coefficient at the takeoff angle (8°)
by 8% and the lift-to-drag ratio by 7%, although the
stall angle of attack decreased by 1°.
(2) For the optimized landing configuration, the deflection
angle of the flap was decreased. The flexible variable
camber trailing-edge flap not only made up for the defi-
ciency of flap deflection but also improved the lift coef-
ficient at the stall angle by 1.5% while maintaining the
original stall characteristics.
(3) For the cruise configuration, the flexible variable camber
trailing-edge flap not only improved the lift characteris-
tics by 0.4 at a cruise angle of 3° in the 2D case but also
improved the lift-to-drag ratio in the lift-coefficient
range from 0 to 2. Here, the maximum lift-to-drag ratio
was increased by 1.2%.
(4) The corrugated structure was used as the stringer in the
skin, and the eccentric beam was used as the deforma-
tion mechanism. The bending mechanism satisfied the
requirements of deformation from 8° to 8°.
References
1. Sinapius M, Monner HP, Kintscher M, Riemenschneider J.
DLR’s morphing wing activities within the European network.
Procedia IUTAM 2014;10(1):416–26.
2. Ahmed S, Guo S. Optimal design and analysis of a wing with
morphing high lift devices54th AIAA/ASME/ASCE/AHS/ASC
structures, structural dynamics, and materials confer-
ence. Reston: AIAA; 2013.
3. Matteo ND, Guo S, Morishima R. Optimization of leading edge
and flap with actuation system for a variable camber wing53rd
AIAA/ASME/ASCE/AHS/ASC structures structural dynamics and
materials conference. Reston: AIAA; 2012.
4. Matteo ND, Guo S, Li D. Morphing trailing edge flap for high lift
wing52nd AIAA/ASME/ASCE/AHS/ASC structures, structural
dynamics and materials conference. Reston: AIAA; 2011.
5. Matteo ND, Guo S, Ahmed S, Li D. Design and analysis of a
morphing flap structure for high lift wing51st AIAA/ASME/
ASCE/AHS/ASC structures, structural dynamics, and materials
conference. Reston: AIAA; 2010.
6. Perera M, He Y, Guo S. Structural and dynamic analysis of
seamless aeroelastic wing51st AIAA/ASME/ASCE/AHS/ASC
Structures, structural dynamics, and materials confer-
ence. Reston: AIAA; 2010.
7. Perera M, Guo S. Optimal design of a seamless aeroelastic wing
structure50th AIAA/ASME/ASCE/AHS/ASC structures, struc-
tural dynamics, and materials conference. Reston: AIAA; 2009.
8. Monner HP, Kintscher M, Lorkowski T, Storm S. Design of a
smart droop nose as leading edge high liftsystem for transporta-
tion aircrafts50th AIAA/ASME/ASCE/AHS/ASC structures,
structural dynamics, and materials conference. Reston: AIAA;
2009.
W. LU et al.
9. Li DC, Guo SJ, He YY, Xiang JW. Nonlinear aeroelastic analysis
of morphing flap. Inter J Bifurcation and Chaos 2012;22(5):125009.
10. Nguyen N, Precup N, Urrnes J, Nelson C, Lebofsky S, Ting E,
et al. Experimental investigation of a flexible wing with a variable
camber continuous trailing edge flap design32nd AIAA applied
aerodynamics conference. Reston: AIAA; 2014.
11. Lebofsky S, Ting E, Nguyen N. Aeroelasticmodeling and drag
optimization of aircraft wing with variable camber continuous
trailing edge flap32nd AIAA applied aerodynamics confer-
ence. Reston: AIAA; 2014.
12. Urnes JM, Morris C, Sheahan J, Dykman J, Clingman D. Control
system design for a variable camber continuous trailing edge flap
system on an elastic wing55th AIAA/ASMe/ASCE/AHS/SC
structures, structural dynamics, and materials confer-
ence. Reston: AIAA; 2014.
13. Urnes JM, Nguyen N, Ippolito C, Totah J, Trinh K, Ting E. A
mission-adaptive variable camber flap control system to optimize
high lift and cruise lift-to-drag ratios of future N+3 transport
aircraft51st AIAA aerospace sciences meeting including the new
horizons forum and aerospace exposition. Reston: AIAA; 2013.
14. Ting E, Dao T, Nguyen N. Aerodynamic load analysis of a
variable camber continuous trailing edge flap system on a flexible
wing aircraft56th AIAA/ASCE/AHS/ASC structures, structural
dynamics, and materials conference. Reston: AIAA; 2015.
15. Lebofsky S, Ting E, Nguyen N. Multidisciplinary drag optimiza-
tion of reduced stiffness flexible wing aircraft with variable camber
continuous trailingedge flap56th AIAA/ASCE/AHS/ASC struc-
tures, structural dynamics, and materials conference.. Reston: AIAA;
2015.
16. Nguyen N, Ting E, Lebofsky S. Aeroelasticanalysis of a flexible
wing wind tunnel model with variable camber continuous trailing
edge flap design56th AIAA/ASCE/AHS/ASC structures, structural
dynamics, and materials conference. Reston: AIAA; 2015.
17. Nguyen N, Tal E. A multi-objective flight control approach for
performance adaptive aeroelasticwing56th AIAA/ASCE/AHS/
ASC structures, structural dynamics, and materials confer-
ence. Reston: AIAA; 2015.
18. Rodriguez DL, Aftosmis MJ, Nemec M, Anderson GR. Opti-
mized off-design performance of flexible wings with continuous
trailing-edge flaps56th AIAA/ASCE/AHS/ASC structures, struc-
tural dynamics, and materials conference. Reston: AIAA; 2015.
19. Yokozeki T, Sugiura A, Hirano Y. Development of variable
camber morphing airfoil using corrugated structure. J Aircraft
2014;51(3):1023–9.
20. Yokozeki T, Sugiura A, Hirano Y. Development and wind tunnel
test of variable camber morphing wing22nd AIAA/ASME/AHS
adaptive structures conference. Reston: AIAA; 2014.
21. Joo JJ, Mark CR, Zientarski L, Culler AJ. Variable camber
compliant wing - design23rd AIAA/AHS adaptive structures
conference. Reston: AIAA; 2015.
22. Marks CR, Zientarski L, Culler A, Hagen B, Smyers B, Joo JJ.
Variable camber compliant wing - wind tunnel testing23rd AIAA/
AHS adaptive structures conference. Reston: AIAA; 2015.
23. Yin WL. Aerodynamic characteristics of multi-sectional variable
trailing-edge airfoil. Journal of Harbin Institute of Technology
2010;42(11):1758–61 [Chinese].
24. Chen J, Bai P, Yin WL, Leng JS. Analysis on the aerodynamic
characteristics of variable camber airfoils with continuous smooth
morphing trailing edge. Acta aerodynamics sinica 2010;28(1):46–53
[Chinese].
25. Wentz WH, Seetbaram HC, Fiscko KA. Pressure distribution for
the GA (W)-2 airfoil with 20% aileron 25% slotted flap and 30%
Fowler flap. Washington, D.C.: National Aeronautics and Space
Administration; 1978. Report No.: NASA- CR-2948.
26. Wang Q, Zhao QJ, Wu Q. Aerodynamic shape optimization for
alleviating dynamic stall characteristics of helicopter rotor airfoil.
Chin J Aeronaut 2015;28(2):346–56.
Aerodynamic optimization and mechanism design
27. Zhao K, Gao ZH, Huang JT, Li Q. Aerodynamic optimization of
rotor airfoil based on multi-layer hierarchical constraint method.
Chin J Aeronaut 2016;29(6):1541–52.
28. Reckzeh D. Aerodynamic design of Airbus high-lift wings.
Braunschweig: DLR; 2005. Germany.
29. Han ZZ, Wang J, Lan XP. The case and application of FLUENT
fluid engineering simulation. Beijing: Beijing Institute of Technol-
ogy Press; 2004.
1003
30. Phan MK, Shin J. Numerical investigation of aerodynamic flow
actuation produced by surface plasma actuator on 2D oscillating
airfoil. Chin J Aeronaut 2016;29(4):882–92.
31. Chin VD, Peters DW, Spaid FW, Mcghee RJ. Flow field measure-
ments about a multi-element airfoil at high Reynolds numbers24th
AIAA fluid dynamics conference. Reston: AIAA; 1993.
32. Hu QM. Encyclopedia of China. Beijing: China Encyclopedia
Press; 1987 [Chinese].