Вы находитесь на странице: 1из 21

47th AIAA Aerospace Sciences Meeting Including The New Horizons Forum and Aerospace Exposition

5 - 8 January 2009, Orlando, Florida

AIAA 2009-878

Aerodynamics of Flapping-Wing Micro Air Vehicles


Sergey Shkarayev1 and Dmitro Silin2
The University of Arizona, Tucson, AZ, 85721

The research study outlined in the paper addresses the aerodynamic features of flexible
flapping wings used in micro air vehicles called ornithopters. Aerodynamic force
measurements were conducted for the 25-cm and 74-cm-wing-span models at different
airflow velocities and flapping frequencies. A series of experiments were conducted on the
25-cm flapping-wing model without free stream airflow. In order to study the effect of a
dihedral on generated thrust and normal force, the model was modified to accommodate
three values of dihedral angle. It appeared that the thrust force is higher for higher dihedral
angle. Effects of a wings bending stiffness and of the wing root constraint on the generated
thrust force and required power were investigated. The aerodynamic forces on flapping
wings were studied with the stroke plane angle varied from horizontal to vertical. An
important result was found that flapping wings do not exhibit a typical, abrupt stall seen
with the fixed wings. Experimental results were analyzed in the framework of the
momentum theory. The results of this study were realized in micro ornithopter designs that
was successfully flight tested.

Keywords: flapping, flight, ornithopter, dynamics, wind tunnel, lift, thrust, experiments.

Nomenclature

A
b
C L

CT 0
CT
f
J
N
n
q
S
T
P
X
Y
V
w

SP
1

= area of actuator disk ( A b2 / 4 )


= wing span of flapping-wing model
= lift coefficient
= static thrust coefficient
= thrust coefficient
= flapping frequency
=
=
=
=
=
=
=
=
=
=
=
=

advance ratio
normal component of aerodynamic force
normal vector to the stroke plane
dynamic pressure
wing area
thrust force
electric power input
projection of aerodynamic force in x-direction
projection of aerodynamic force in y-direction
free stream velocity
induced velocity
stroke plane angle

Associate Professor, Department of Aerospace and Mechanical Engineering, University of Arizona, 1130 N.
Mountain Ave, Tucson, AZ 85721. Senior Member of AIAA.
Graduate Research Assistant, Department of Aerospace and Mechanical Engineering, University of Arizona,
1130 N. Mountain Ave, Tucson, AZ 85721.
1

Copyright 2009 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.

= flapping amplitude
= dihedral angle
= air density
= zero free stream (subscript)

I. Introduction

small size of micro air vehicles (MAVs), their reconnaissance capabilities were the driving factor for
Btheecausefirstofgeneration
of vehicles. These developments were concerned MAVs with traditional fixed wings
airplanes and rotorcraft (helicopters).
Flapping-wing micro air vehicles generate lift and thrust for forward motion using their flapping wings,
emulating birds and insects. However, just mimicking the flight of birds and insects may not be sufficient in
designing ornithopters. Here is how this viewpoint was elucidated by Mueller and DeLaurier:1 The primary
motivation for studying animal flight is to explain the physics for a creature that is known to fly. An ornithopter
designer, in contrast, is trying to develop a flying aircraft, and its ability to achieve this is no given fact.
Conversely, a successful micro air vehicle design can provide a verifiable physical model of flight in nature. In this
section, after a brief discussion of the current designs of micro ornithopters, previous studies of the aerodynamics of
flapping wings will be reviewed.
Aerovironment pioneered the designing of radio-controlled micro ornithopters.2 Their most successful vehicle
has a half-ellipse wing planform with a 20-cm wingspan and flapping frequency of 22 Hz. Because of a lack of
understanding of flapping-wing aerodynamics and enabling technologies, the vehicle demonstrated poor
performance.
DeLaurier and his group3 developed a 35-cm ornithopter capable of hovering. A clap and fling mechanism for
enhancing the lift was employed in the kinematics of the 4-wing design. Hovering flights in excess of one minute
were achieved at the flapping frequency of 28 Hz. It was noted that transition to forward flight remains a problem
for this type of vehicle.
The University of Arizona has developed several successful ornithopters.4 The smallest ornithopter (Fig. 1) has a
15-cm wingspan, weighs only 9 grams, and has a flight endurance of 3 min. This ornithopter can fly for more than 3
min at a speed from 1 m/sec to 5 m/sec.
The dynamics of flapping flight was investigated in our previous work5 using the ornithopter equipped with the
autopilot. Essential stability parameters and basic dynamic modes were investigated through extensive laboratory
and flight measurements. Control-fixed flights were performed in order to examine passive aerodynamic stability of
the flapping-wing apparatus. These experiments showed an existence of similar type of low-frequency oscillatory
dynamics in both roll and pitch motions of the ornithopter.

Fig. 1. University of Arizona ornithopter.

Currently, a small number of successful designs resulted in sustained and controlled flapping flight of
ornithopters. The main reasons for this are in lack of supporting technologies and in poor understanding of the
aerodynamics and controls of flapping flight.
Fortunately, the nature offers its millions-years experience in designing of very effective and robust flying
machines: insects and birds. Some of the notable works on the aerodynamics of flapping flight were completed by
zoologists. Birds and insects have evolved very sophisticated methods of maneuvering capabilities, including
vertical takeoff, landing and hovering. A variety of the kinematic mechanisms of control in insects and birds is
created by differences in the number of wings, their structural arrangement, and features in wing kinematics. Note
that minute features of flapping wing kinematics may affect, in a complex way, unsteady aerodynamics mechanisms
and, therefore, their controls.
The kinematics of flapping wings is a complex phenomenon that is mainly responsible for the flight performance
characteristics. The overall wings motion can be presented as a combination of flapping, rotational, and folding
modes. The flapping motion is restricted to a plane called a stroke plane. The folding mode refers to the wing
motion in the stroke plane. The flapping mode is synchronized with the rotation or pitching of the wing about the
axis connecting the wing base and wing tip. There are two distinct phases of pitching: pronation (from forward to
backward stroke) and supination (from from backward to forward stroke).
This flapping-wing motion includes an articulated motion of the wing as a rigid body and elastic bending and
twisting deformations. It has proven difficult to separate the effects of articulated motion and flexibility of the
flapping wings. Since the center of pressure and shear center do not typically coincide in an actual insect or bird
wing, a resultant torque is applied, causing wing to twist. Thus, pitching can also be achieved by passive
deformations of the wing structure.
In early investigations, the similarities of a flapping wing and propeller were observed, giving argument for the
application of the steady aerodynamic theory to flapping-wing animals. Osborne6 hypothesized that the flapping
wing, at a given time instant, moving with some instantaneous velocity and angle of attack generates the same
amount of instantaneous thrust and lift as a rigid blade moving at the same constant velocity and angle of attack.
Based on this analysis, aerodynamic forces were averaged over the wing stroke. 6 The vertical and horizontal
components of the total force were compared with the weight and thrust, and average lift and drag coefficients were
calculated for flying animals. This method was named the quasi-steady aerodynamic model of flapping flight. The
applicability of the quasi-steady aerodynamic model to fast horizontal flight was reasonably well proven by many
authors. However, when the same question was applied to slow flight or hovering, it remained unanswered or
answered with insufficient confidence for a long time.
It is worth mentioning the difficulties of obtaining direct measurements of cyclic aerodynamic forces in flying
animals and the consequent difficulties of experimental validation of proposed theories. Manmade membrane wings
represent a useful model for the study of the aerodynamics of flapping flight. Two types of wing models have been
investigated: rigid and membrane. Pitching of rigid wings is realized with the help of articulated pitching
mechanism. Flapping membrane wings feature a leading edge spar actuating the flapping motion. A flexible
membrane attached to the spar is twisted during flapping cycle providing wings pitching.
Van den Berg and Ellington7 studied a 10:1 scale model mimicking the flapping motion of the hawkmoth
Manduca. Using flow visualization and analyzing airflow during the downstroke, vortices were found to form along
the leading edge, which moved toward the wing tip, becoming unstable and detatching from the wings surface at 75
percent radial distance of the wing. The vortex diameter was smaller than predicted for linearly translating wings
and a strong axial flow was observed to increase the stability of the vortex, reducing its diameter. The vortex ring
velocity corresponded well with predicted wake impulses, indicating sufficient lift was produced for weight support.
Additionally, the authors confirmed that the dynamic stall and not wing rotation was the mechanism responsible for
leading-edge vortex production. Leading-edge vortices can supply up to two-thirds of the required lift during the
downstroke. It is still unknown what conditions are required for vortex stability, which determines the extent of lift
production.
In the previous studies,8,9 the rotational mechanism of lift augmentation in flapping wings was investigated. The
measurements were conducted on a scaled model of fruit fly wings made of plexiglas that was immersed in mineral
oil. One key element in the production of aerodynamic forces by birds and insect wings lies in the flexibility of the
wing structure, which was not accounted for in the previous works.
Gallivan and DeLaurier10 conducted experimental studies of flapping membrane wings in wind tunnel. During
experiments, lift, thrust, and pitching moment were measured and averaged over a flapping cycle. Significant
scatters in the data were explained by a zero drift from the strain-gage balance. Effects of geometrical, mass and
stiffness parameters of membrane wings were investigated and discussed. It was shown that stiffening the front spar
increases the thrust and has negligible effect on lift. The presence of full chord-length battens was found to produce

a positive effect on both lift and thrust forces. However, no specific values of test Reynolds numbers were provided
and measurements were performed only for two flapping axis angles of 0 and 10 degrees.
Designing a flapping-pitching mechanism is not a simple problem and mechanical design drawbacks can
prevent one from obtaining accurate data. Singh and Chopra11 studied the performance of flapping wings with
articulated pitching at no free stream conditions. Static thrust and mechanical power were measured for varied
flapping frequencies, pitching angles, and wing designs. Two flapping-pitching mechanisms were built for pitch
articulation: bistable and passive utilizing a torsion spring. An aluminum frame and composite frame wings were
tested on both pitching setups. The aluminum frame wing showed higher thrust than the composite wing on the
bistable pitch mechanism. However, when the wings were tested on the passive pitching setup, the force magnitudes
were reversed and, therefore, these results are inconsistent.
Studies to date on the aerodynamics of flapping flight, although beneficial to an understanding of the subject,
have not taken into account all the details that are necessary to obtain an accurate representation of the true
aeromechanics of flapping flight. In the present study, wind tunnel measurements of stroke-averaged aerodynamic
forces were conducted on membrane flapping wings. The important aspects of flapping aeromechanics were
investigated: dihedral offset, stiffness of wing structure, constraint of wing base motion, wing size effects, variation
of stroke plane angle from horizontal to vertical, applicability of momentum theory for flapping wings, designing
flapping-wing micro air vehicles.

II. Experimental Setup


A. Wind tunnel facility
Experimental studies were performed in the University of Arizona Wind Tunnel (Fig. 2). This wind tunnel has 3
4 ft test section and velocity range from 2 to 50 m/s. The flow is laminarized in a settling chamber to less than
0.3% turbulence in the axial direction. A flapping-wing model installed in the tunnel is seen in Fig. 2.
The force balance utilized in this study contains six precision flexures with strain gauges for measuring lift, drag,
side force, pitching, rolling and yawing moments. For this sequence of tests, side forces, rolling and yawing
moments were not sought. The force balance is located under the test section of the wind tunnel, as seen in Fig. 3.
Data from strain gages was logged using two National Instruments SCXI-1321 terminal blocks in a low-noise
SCXI-1000 chassis capable of sampling at 330,000 Hz. The resolution of each flexure was about 0.004 N. The zero
drift for flexures did not exceed 1.5% over 15 min of test time with no load applied. When oscillatory load was
applied through the 25-cm flapping wings operating at 20 Hz, the zero drift for the same flexures did not exceed 2%
per 15 min of test run. In order to determine the error in the measured aerodynamic forces, the calibration of the
wind tunnel and the balance was performed before each test series.
Utilizing calibration measurements and the small-sample method,12 the uncertainty intervals in aerodynamic
forces corresponding to a confidence level of 90% were determined. For the range of measured forces 0-5 N, the
uncertainty intervals were: 0.015 N for vertical and 0.007 N for horizontal forces. The value of the lift channel
uncertainty was higher because the lift was measured by two flexures, thus two sources of measurement errors
contributed to the total uncertainty. For the pitching moment of order 0-0.15 Nm the uncertainty interval was of
0.0005 Nm. The standard deviation of measured dynamic pressure was 0.4 N/m2 for the range of velocities studied
in the present work.

Fig. 2. Flapping-wing model in the wind tunnel.

Fig. 3. Force balance.

B. Flapping-wing models
Aerodynamic force measurements were conducted for the 25-cm and 74-cm wing models. Each test model
consists of a mounting rib, flapping wings, and transmission. Generic drawings of a wing are shown in Fig. 4.
Geometrical and mass data for both 25-cm and 74-cm wing are presented in Table 1.

Fig. 4. Geometry of flapping wing model.

The 25-cm wing structure consisted of wing arms, front spar, 4 battens and membrane. Wing arms were made of
0.8 mm music wire. The membrane was 0.015 mm Mylar bonded to the front spar and battens with rubber cement.
In the basic flapping wing model (model A), pultruded carbon rods were used: T315-413 of diameter 0.8 mm for the
front spar and T305-4 of diameter 0.5 mm for battens. Higher bending stiffness in model B was achieved with
additional 0.5 mm diameter carbon rod glued to the root section of the front spar. Higher flexibility in wing model C
is reached by tapering the front spar toward the wing tip. Battens were not rigidly fixed to the front spar, allowing
pitching deformations due to inertia and aerodynamic forces. Original models A-D were of zero dihedral (the angle
in Fig. 4). In order to study the effect of the dihedral on thrust and normal force generation, the model A was
modified by changing the length of connecting rods (Fig. 4).
A wing from commercially available Cybird P2 ornithopter was used as the 74-cm model. Its structure included
front and rear spars, 6 battens, and membrane made of Nylon cloth. Battens were rigidly fixed to both front and rear
spars.

Table 1. Specifications of flapping wing models.


Wing model

Wingspan, b, cm
Flapping amplitude, , deg
Wing area S, cm2
Batten 1, ll1, cmdeg
Batten 2, l22, cmdeg
Batten 3, l33, cmdeg
Intersection point, l4, cm
Root chord, cr, cm
Wingtip deflection, cm/g
Mass, g

25
72
137
3.870
877
7
0.18
1.15

25
72
137
3.870
877
7
0.127
1.28

25
72
137
3.870
877
7
0.233
1.18

74
55
991
7.580
16.575
25.565
27
16.8
0.95
15.8

For the purposes of this study, the advance ratio parameter was defined as J V / bf , which was varied in the
range from 0 (static conditions) to 1.2. The variation of advance ratio was achieved by varying the free stream
velocity, V , for flapping frequencies, f , held constant at 22 Hz and 9.2 Hz, for models A and D, respectively.
Correspondently, the free stream velocities used in the experiments are presented in Table 2.
For the flapping frequency measurements, the experimental setup was equipped with a built-in optical
tachometer. The tachometer consists of a CP-36 photodiode and ECG-3038 phototransistor connected to the data
acquisition board.
Table 2. Free stream velocities, V , m/sec.
Wing
Model

J
0.4

0.6

0.8

1.2

2.77

4.15

5.53

6.94

8.30

2.61

3.92

5.23

6.54

7.84

The aerodynamic forces X , Y and moment, M , were determined in the wind tunnel frame of reference, as
shown in Fig. 5. The orientation of a flapping wing model with respect to a free stream is described by a stroke
plane angle, SP , which is defined by the normal vector, n , to the stroke plane.

Fig. 5. Conventions used for positive forces and angle.

III. Experimental Results


A. Effect of dihedral
First, the effect of dihedral angle, , on aerodynamic forces was studied. Wind tunnel measurements on
flapping wings were conducted without airflow, i.e. V = 0. The 25-cm flapping-wing model A was tested for three
values of the dihedral: 0 , 8 , and 19 . The model was mounted on top of the pylon fixture of the
balance and tested at constant flapping frequencies varying from 5 to 23 Hz in increment of 2 Hz. In addition to
force measurements, an electric power input, P, was recorded in all tests.
Measured static thrust, T0 , normal force, N 0 , and total force-to-power variations with flapping frequency are
presented in Figs. 6, 7, and 8, respectively. It appeared from Figure 6 that as the flapping frequency increased, the
thrust slope increased and the thrust force was higher for higher dihedral angle: at f = 23 Hz, the thrust for 19
was 20% greater than for 0 .
Flapping wings without dihedral are moving symmetrically and, therefore, should not produce any normal
forces at 0 . The results presented in Fig. 7 proved it. The data points were scattered within 0.02 N about
zero. However, the normal force increased dramatically at 19 to the maximum of 0.11 N at 23 Hz. It resulted
in the increase of the total force,

T02 N 02 by 8% , and change in the direction of the total force from 0 to 24 .

Obviously, with the dihedral increase the interference between wings increases changing aerodynamic forces.
This phenomenon is analogous to the near fling mechanism for lift enhancing employed by some insects and birds. 16
In this mechanism, the wings move close to each other (near fling) or even clap at the end of the upstroke. Then, the
wings fling rotating about a trailing edge. The previous study17 showed that the effectiveness of the near fling was
increased significantly with distance between wings and opening angle decrease. Large aerodynamic forces resulted
from these kinematic patterns were explained by added masses and increased circulation. Thus, increasing
aerodynamic forces with increased dihedral, observed in the present study, can be explained by the near fling
mechanism.
The plots of the total force-to-power illustrate a power efficiency of the flapping propulsion. What is noticeable
from Figure 8 is a decrease of

T02 N 02 / P0 with the flapping frequency increases. Thus, the power efficiency of

flapping wings decreased on average by 20% with the dihedral increase from 0 to 19 .

One conclusion from these observations was that the near fling mechanism can be very useful for changing the
total aerodynamic force. It is theorized that a significant change in the direction of the total aerodynamic force could
be utilized in controls of a flapping-wing air vehicle through pointing the resultant force in the desired direction of
flight.

0.3
0.25

T 0, N

0.2
0.15
0.1
=19
=8
=0

0.05
0
4

10

12

14 16
f , Hz

18

20

22

24

Fig. 6. Effect of dihedral angle on the thrust.

0.12
=19
=8
=0

0.1
0.08

N 0, N

0.06
0.04
0.02
0
-0.02
-0.04
4

10

12

14
f , Hz

16

18

20

22

Fig. 7. Effect of dihedral angle on the normal force.

24

Fig. 8. Propulsion effectiveness of flapping wings.

B. Effect of wings stiffness


Since the center of pressure and the shear center do not typically coincide in the actual wings of insects and
birds, a resultant torque causes their wings to twist. Hence, the key element in the production of the thrust by
membrane flapping wings lies in the flexibility or stiffness of the wing frame.
In order to investigate the effects of wing stiffness on the developed thrust and required power, two more wing
models were built models B and C, with increased and decreased stiffness, respectively (Table 1). All three wings
A, B, and C were built with the dihedral of 19 .
The models were tested under the static conditions (no airflow). The thrust force was measured for flapping
frequencies up to 23 Hz, and the results are presented in Fig. 9. The stiffer wing of model B shows the highest
thrust, followed by the most flexible model C. At f 23 Hz, the model B generates almost 45% higher thrust than
model A.
Also, the electric power input was recorded and the total force-to-power ratio is presented in Fig. 10. Models B
and C appeared to be 1.5-2 times less power effective than the model A.
Based on presented results, stiffer wings produce higher thrust; however, they suffer a significant decrease in
their power efficiency. The optimal design of the flapping-wing structure should depend on specifics of a flight
mission. If a long range flight is desirable, then more flexible wings would be preferable. Contrary, if a high
maneuverability or hovering capabilities are required, then stiffer wings would generate larger forces providing
additional aerodynamic controls for performance.

10

0.45
0.4
0.35
0.3

T 0, N

0.25
0.2

0.15
Model A
Model B
Model C

0.1
0.05
0
4

10

12

14
f , Hz

16

18

20

22

24

Fig. 9. Thrust variation with flapping frequency.

Fig. 10. Variation of total force-to-power ratio with flapping frequency.

C. Constraint on wing base motion


Kinematics of flapping wing motion depends on the constraints on the wing base motion. In order to study this
effect, experiments were conducted with the wings structure released along the wing base and only a small portion at
the wings apex remained fixed (Fig. 11). The results of measurements showed almost no changes in thrust force
indicating that the near root part of the wing does not produce any significant thrust.

11

Fig. 11. Schematics of flapping-wing models with fixed and released root rib.

D. Wing size effect and variation of stroke plane angle


During the transitioning between maneuvers and from horizontal flight to vertical and back, high stroke plane
angles, SP , often occur. Wind gusts can also generate high stroke plane angles. The flapping-wing models A and
D were tested at the stroke plane angle varying from the horizontal ( SP 0 ) to the vertical position ( SP 90 ).
For each experiment, the frequency of flapping was held constant while the model was tilted from 0 to 90 in
increments of 1.22 per data point. Note that the wing D is almost three times larger than wing model A.
The lift curves are typically one of the first things to look at when designing an aircraft or analyzing its
performance. The results for the stroke-averaged vertical component of the aerodynamic force are presented in Figs.
12 and 13 for models D and A, respectively. Overall, the plots exhibit similar behaviors. The data are less scattered
for model D, which can be explained by higher measured force values for this model. These non-zero vertical forces
at SP 0 may result from non-symmetry in flapping kinematics, geometrical or stiffness properties of a
flapping-wing model.
The values of slopes can be determined keeping the data corresponding to the stroke plane angle up to 15. In
terms of trends, it appeared that as the velocity increased, the curve slopes of the vertical force increased.
The vertical force plots demonstrate nonlinear behavior at higher angles. The values of maximum forces and the
corresponding stroke plane angles are presented in plots in Figs. 12 and 13. The maximum forces for the model D
are 8.2-9.5 times greater than for the model A, which is close to the ratio of these wings areas of 7.2.
After reaching maxima, force magnitudes slowly decreased, when the stroke plane angle approached SP 90 .
It is seen from the Figs. 12 and 13 that the flapping wings do not exhibit a typical, abrupt stall seen with the fixed
wings. This important result necessitates further studies of the physics of flow in flapping wings at a high SP .
7
L MAX=5.2 @ " SP=76

6
L MAX=4.4 @ " SP=65

5
L MAX=3.7 @ " SP=55

Y, N

4
3
2

J=1.2
J=1
J=0.8
J=0.6
J=0.4

1
0

L MAX=2.9 @ " SP=44

L MAX=2.3 @ " SP=65

-1
0

15

30

45

SP , deg

60

75

90

Fig. 12. Variation of vertical force with stroke plane angle and advance ratio (Model D).

12

0.7

J=1.2
J=1
J=0.8
J=0.6
J=0.4

0.6

Y,N

0.5

L MAX=0.6 @ " SP=51

L MAX=0.54 @ " SP=51

0.4
0.3
0.2
L MAX=0.27 @ " SP=56

0.1
L MAX=0.31 @ " SP=48

L MAX=0.4 @ " SP=48

0
0

15

30

45

SP , deg

60

75

90

Fig. 13. Variation of vertical force component with stroke plane angle and advance ratio (model A).
Similar to fixed wings, the horizontal force plots for flapping wings in Figs. 14 and 15 are concave. In these
plots, the negative force actually means a pointing-forward thrust force, and a zero value means that thrust and drag
, corresponding to the balance are shown in Figs. 14 and 15.
are balanced. The stroke plane angles, SP

J=1.2
J=0.8
J=0.4

J=1
J=0.6

X, N

' SP=31
2

' SP=27
1

' SP=22

0
-1

' SP=36.7

' SP=57.5

-2
0

15

30

45

SP , deg

60

75

90

Fig. 14. Variation of horizontal force with stroke plane angle and advance ratio (Model D).

13

0.7

0.2

J=1.2
J=1
J=0.8
J=0.6
J=0.4
' SP=29.4

' SP=24.5

0.1

' SP=22

0.6
0.5

X, N

0.4
0.3

' SP=50.2

' SP=34.3

-0.1
-0.2
0

15

30

45

60

75

90

SP , deg

Fig. 15. Variation of horizontal force component with stroke plane angle and advance ratio (model A).

IV. Discussions
Static thrust coefficient is usually defined as,
CT 0

(1)

f 2b 4

Figure 16 shows the plots of thrust coefficients for data obtained by Singh and Chopra11 (soft spring in Fig. 21a) and
data presented in Figure 6 for 0 .

0.18
0.16
0.14

CT 0

0.12
0.1
0.08
0.06
0.04

Singh and Chopra


Model A

0.02
0
4

10

12

14

16

18

20

22

24

f , Hz
Fig. 16. Variation of static thrust coefficient with flapping frequency.
In Ref. [15] data were given on a model of wing-propeller combination at angles of attack up to 90 . The results
of the previous work showed that the increase in propeller thrust increased the maximum lift and the corresponding

14

angle of attack. The propeller slip stream also significantly diminished the reduction of the lift force in the range
from the maximum angle of attack to 90 . This behavior in aerodynamic force is analogous to the one observed for
the lift in the present study (Figs. 12 and 13). The definitions of the lift and thrust coefficients given in Ref. [15]
were used in the present analysis. Following the notations in Fig. 5, the lift and thrust coefficients were defined as

C L

Y
q S

(2)

CT

X cos SP Y sin SP
T

q S
q S

(3)

Here, the dynamic pressure, q , is a sum of free stream and flapping wings slip stream components

q 0.5V2

T0
A

(4)

Using the data presented in Figs. 12-15, the aerodynamic coefficients were calculated using Equations (2-4) and
plotted in Figs. 17-20.
1.4
1.2
1

C"L

0.8
0.6

J=1.2
J=1
J=0.8
J=0.6
J=0.4

0.4
0.2
0
0

15

30

45

SP , deg

60

75

90

Fig. 17. Variation of lift coefficient with stroke plane angle and advance ratio (Model D).

15

1.2

J=1.2
J=1
J=0.8
J=0.6
J=0.4

C"L

0.8
0.6
0.4
0.2
0
0

15

30

45

SP , deg

60

75

90

Fig. 18. Variation of lift coefficient with stroke plane angle and advance ratio (Model A).
1.2

J=1.2
J=1
J=0.8
J=0.6
J=0.4

C"T

0.8
0.6
0.4
0.2
0
0

15

30

45

SP , deg

60

75

90

Fig. 19. Variation of thrust coefficient with stroke plane angle and advance ratio (Model D).

16

1.2

J=1.2
J=1
J=0.8
J=0.6
J=0.4

C"T

0.8
0.6
0.4
0.2
0
0

15

30

45

60

75

90

SP , deg

Fig. 20. Variation of thrust coefficient with stroke plane angle and advance ratio (Model A).
A classical theory of rotating propellers combines the blade element theory and the momentum principle15
providing propellers thrust as well as induced velocity behind the propeller disk. Ellington18 used this theory in
studies of the aerodynamics of insects. A partial actuator disk was proposed in,18 which is a better model of actual
flapping wings.
In order to verify the applicability of this theory to the performance of flapping wings in the current study,
additional tests were conducted on the wing model A. In these tests, the thrust force generated by flapping wings
was measured for the input power held constant at 1, 2, and 3 W. The values of the static thrust were determined of
0.086, 0.153, and 0.211 N for 1, 2 and 3 W, respectively. Then, the induced velocity was found using the formula

w0 T0 / 2A . The stroke plane angle was varied as SP 0, 28, 57, 85.


Theoretical values of the thrust at a given forward velocity and stroke plane angle were determined using the
following equations:

w
0

w V
w
2

w w cos SP w
0

0
0

T V
w

cos
T0 w0
w0

w 1
0

(5)

(6)

Theoretically, the thrust generated by the actuator disk does not depend on the input power. The experimental
results presented in Figure 21 are consistent with the theory within the margin of error of 27% and follow the
theoretical trend. Deviations of the experimental data from the theory increased dramatically with the stroke plane
angle increased (Fig. 22). Overall, the momentum theory underpredicts the performance of flapping wings in terms
of available thrust.

17

Fig. 21. Variations of thrust with free stream velocity and input power at zero stroke plane angle.

Fig. 22. Variations of thrust with free stream velocity and stroke plane angle for the input power of 3 W.

V. Ornithopter design
Designs of successful flapping-wing micro air vehicles are presented here. A computer modeled rendering of a
typical University of Arizona ornithopter design is presented in Fig. 23, and the mass breakdown of two vehicles are
presented in Table 3.
The design goal for the 15-cm ornithopter was to demonstrate that current progress in miniature radio-control
components and electric power trains allows the creation of a flapping- wing MAV within the sub-15-cm linear
scale. The uniqueness of the 20-cm ornithopter is in its high agility and, specifically, hovering capability.

18

Fig. 23. SolidworksTM rendering of a typical ornithopter design.


Table 3. Mass breakdown of ornithopters
Components

Mass, g
15 cm

20 cm

Wing

0.3

0.4

Fuselage & Tail

1.8

4.5

Motor

2.9

12

Battery

2.4

10

Speed Controller

0.4

RC Receiver

0.4

Servos

0.8

3.5

Total

32.4

The battery and motor account for 55% of the total weight of the ornithopters, making the powertrain the
heaviest system in the aircraftit remains the biggest design issue. The transmission is powered by an electric
motor. The following motors are used: Super Slick 3.3 Ohm and Micro DC5-2.4 coreless electric motors for the 15cm and 20-cm ornithopters, respectively. The motor is regulated by a speed controller and radio receiver. Energy is
provided by lithium polymer batteries.
The motor drives a crankshaft mechanism through a reduction gearbox. The crankshaft initiates a flapping
motion of the front spar by the action of cranks connected to push-rods. The 15-cm ornithopter had its motor aligned
with the longitudinal axis of the vehicle. Designs with a streamlined longitudinal motor arrangement proved difficult
to trim for level flight due to the powerful rolling moment produced by the motor torque. The motor for the 20-cm
vehicle was mounted in the transverse direction. For this design, the torque reaction produces a positive (nose-up)
pitching moment, improving stability and controls. The cranks are counter-rotating in the 15-cm and rotating in the
same direction in the 20-cm vehicle.
The fuselage frame was made from balsa wood and carbon rods. The frame and components are protected from
the impact of landing by a EPP foam nose. The wing structure consists of the carbon rod frame and a membrane.
The wing of 15-cm vehicle has a front spar made of 0.5 mm diameter carbon rod. The 20-cm ornithopter utilizes 0.8

19

mm main spar and 0.5 mm root section reinforcement for high output thrust needed for hovering. The following
materials were used for the membrane: Mylar of 16.8 g/m2 density, and Mylar of 7.0 g/m2 density for the 20-cm, and
15-cm ornithopters, respectively. The 15-cm and 20-cm aircraft use conventional tail with rudder and elevator.

Summary
Aerodynamic force measurements were conducted at the University of Arizona wind tunnel using the 25-cm and
74-cm flapping wing models. The effects of geometric, elastic, and kinematic parameters of flapping wings on
stroke-averaged aerodynamic forces were analyzed.
In a series of static tests of the 25-cm model, the flapping frequency was varied from 5 to 23 Hz in increment of
2 Hz. In addition to force measurements, an electric power input, P, was recorded in these tests. The results showed
that with the flapping frequency increased, the thrust slope increased and the thrust force was higher for higher
dihedral. The normal force increased dramatically at a high value of the dihedral. It was theorized that a significant
change in the direction of the total aerodynamic force could be utilized in controls of a flapping-wing air vehicle
through pointing the resultant force in the desired direction of a flight.
In order to study the effects of wing frame stiffness on the developed thrust and required power, two more wing
models were built. The stiffer wing model showed the highest thrust, followed by the most flexible model, however,
it suffered a significant decrease in power efficiency. The optimal design of the flapping-wing structure should
depend on the specifics of a flight mission. If a long range flight is desirable, then more flexible wings would be
preferable. Contrary, if a high maneuverability or hovering capabilities are required, then stiffer wings would
generate larger forces providing additional aerodynamic controls for performance.
During the transitioning between maneuvers and from horizontal flight to vertical and back, high stroke plane
angles often occur. Wind gusts can also generate high stroke plane angles. The 25-cm and 74-cm flapping wing
models were tested at the stroke plane angle varying from the horizontal to the vertical position. In terms of trends, it
appeared that as the velocity increased, the curve slopes of the vertical force increased. After reaching maximum, of
force magnitudes slowly decreased with the stroke plane angle approaching 90 . It was noted that the flapping
wings do not exhibit a typical, abrupt stall seen with the fixed wings. This important result necessitates further
studies of the physics of flow in flapping wings at high stroke plane angles.
In order to verify the applicability of the momentum theory to the performance of flapping wings, additional
tests were conducted on the 25-cm wing model. The stroke-averaged thrust was measured for the input power held
constant. Theoretically, the thrust generated by the actuator disk does not depend on the input power and
experiments showed consistency with the theory. In terms of the thrust variations with free stream velocity and
stroke plane angle, the momentum theory underpredicts the performance of flapping wings.
Designs of successful flapping-wing micro air vehicles were presented and discussed.

Acknowledgments
This project was sponsored by the College of Engineering at the University of Arizona. The authors also would
like to thank the other members of the Micro Air Vehicle Project at the University of Arizona for their contributions
to this work: Roman Krashanitsa and Malladi Bharani.

References
1

Mueller, T.J., and DeLaurier, J.D., An Overview of Micro Air Vehicle Aerodynamics, In: Fixed and Flapping
Wing Aerodynamics for Micro Air Vehicle Applications, edited by T. J. Mueller, Vol. 195, AIAA, Reston, VA,
2001, pp. 1-12.
2
Keennon, M., and Grasmeyer, J., Development of Two MAVs and a Brief Vision for the Future of MAVs
Design, AIAA-2003-2901, AIAA International Air and Space Symposium and Exposition: The Next 100 Years,
Dayton, Ohio, July 14-17, 2003.
3
DeLaurier, J., Ornithopter Research, 2003 Bioflight Workshop, NASA Langley Research Center, August 7-8,
2003.
4
Olson, D. H., Silin, D., Aki, M., Murrieta, C., Tyler, J., Kochevar, A., Jehle, A., and Shkarayev, S., Wind
Tunnel Testing and Design of Fixed and Flapping Wing Micro Air Vehicles at the University of Arizona, Micro
Air Vehicle Design Papers, Konkuk University, South Korea, 2005, 9 p.

20

Krashanitsa, R., Silin, D., Shkarayev, S., and Abate, G., Flight Dynamics of Flapping-Wing Air Vehicle,
AIAA AFM Conference and Exhibit, August 18-21, 2008, Honolulu, Hawaii, AIAA-2008-6698.
6
Osborne, M. F. M., Aerodynamics of Flapping Flight with Application to Insects, Journal of Experimental
Biology, Vol. 28, 1951, pp. 221245.
7
Coen van den Berg, and Ellington, C. P., The vortex wake of hovering model hawkmoth, Philosophical
Transactions: Biological Sciences, Vol. 352, Issue 1351, pp. 317-328, 1997.
8
Lehmann, F.-O., and Dickinson, M., The control of wing kinematics and flight forces in fruit flies
(Drosophila SPP.), The Journal of Experimental Biology, 201, pp. 385-401, 1998.
9
Dickinson, M., Lehmann, F.-O., and Sane, S., Wing rotation and the aerodynamic basis of insect flight,
Science, Vol. 284, No. 5422, pp. 1881-2044, June, 1999.
10
Gallivan, P. and DeLaurier, J., An Experimental Study of Flapping Membrane Wings, Canadian Aeronautics
and Space Journal, Vol. 53, No. 2, June 2007, pp. 35-46.
11
Singh, B., and Chopra, I., Insect-Based Hover-Capable Flapping Wings for Micro Air Vehicles: Experiments
and Analysis, AIAA Journal, Vol. 46, No. 9, September 2008, pp. 2115-2135.
12
Kline, S. J., and McClintock, F. A., Describing Uncertainties in Single-Sample Experiments, Mechanical
Engineering, Vol. 75, No.1, 1953, pp. 3-8.
13
Material Data Sheets, CST - The Composites Store, Inc., Tehachapi, California, 2006.
14
Kuhn, R. E., and Draper, J. W., Investigation of the Aerodynamic Characteristics of a Model Wing-propeller
Combination and of the Wing and Propeller Separately, NACA Technical Report 1263, Jan 1956, pp. 247-286.
15
McCormick, B. W., Jr., Aerodynamics of V/STOL Flight, Dover Publications, Inc., Mineola, New York, 1999,
328 p.
16
Weis-Fogh, T., Quick Estimates of Flight Fitness in Hovering Animals, Including Novel Mechanisms for Lift
Production, Journal of Experimental Biology, Vol. 59, 1973, pp. 169-230.
17
Sunada, S., Kawachi, K., Watanabe, I., and Azuma, A., Fundamental Analysis of Three-dimensional Nearfling, Journal of Experimental Biology, Vol. 183, 1993, pp. 217-248.
18
Ellimgton, C. P., The Aerodynamics of Hovering Insect Flight: V. A vortex theory. Philos. Trans. R. Soc.,
London, Ser. B, 1984; 305, pp. 11544.

21

Вам также может понравиться