Professional Documents
Culture Documents
a r t i c l e
i n f o
a b s t r a c t
Article history:
Received 2 October 2007
Received in revised form 6 July 2009
Accepted 9 July 2009
Available online 20 August 2009
Keywords:
Planar membrane wing
Flapping wing
MAV
This paper presents the design and aerodynamic performance of a planar membrane wing as shape airfoil
for the micro aerial vehicle. This simulation calculates the average lift force, L as the criteria weight of
the apping wing (weight must be lower than 8.78 g), to make one ultra-light, small size apping wing
MAV. In here two phases are discussed. First, the 3D aerodynamic calculation and ow eld simulation
of a planar membrane wing as shape airfoil for a MAV were studied. Analyzing the apping wing under
different frequencies and angles of attack, investigates the pressure distribution, the airfoil-tip vortex and
the up-wash situation of the air ow. Second is to average lift force, L 8.78 g for designing weight limit
of the MAV. The specications of apping wing MAV are 8 g gross weight, the 15 cm wingspan, and
5 cm chord length. In this vehicle, we employed the concept of four-bar linkage to design a apping
mechanism which simulates the apping motion of a bird. The angles of upstroke and downstroke can
be varied in the design. The total apping angle is 73 . The apping frequency of wing is 25.58 Hz.
The power source comes from motor with a LiH battery. A simple ight test was carried out and the
result of the ight is going well. The actual ight distance is approximately 8 m, and the primary goal is
achieved. By the way, we found the rigidity of tail wing is crucial and should be enhanced to prevent the
apping-wing MAV will be unable to revise if the MAV in a crooked condition and it will cause a crash.
2009 Elsevier Masson SAS. All rights reserved.
1. Introduction
The micro aerial vehicle, in English is abbreviated as a MAV,
according to the Defense Advanced Research Projects Agency
(DARPA) of USA, the size of various aspects of micro aerial vehicle (MAV) is limited to 15 cm, the ying speed is 1020 m/s, the
Reynolds number must be below 106 . Regarding a apping wing
for a MAV, the most important issue at present is the aerodynamic performance. The Reynolds number of a MAV is about 105 ,
this range of Reynolds number will cause laminar separation phenomenon occurred on the surfaces of the body. Moreover, since
the denition of a MAV includes size limit, and the challenge of
this work is to design an ultra-light and small size of a apping
wing MAV comparing all literatures [1,7,9,10,13], therefore by using very low aspect ratio of MAV to obtain enough lifting force,
L. However, small aspect ratio will increase the three-dimensional
effects on ow eld. The MAV is small and the speed is low, the
ight stability of a MAV is affected easily by the external wind
shear or other disturbances.
This research applied dynamic moving grid technology and analyzed a planar membrane wing under the low Reynolds number.
Each pattern of the ap movement initiates a complex and un-
*
1
384
Nomenclature
L
P
L
T
T
CL
CL
CD
CD
K
AOA
the ying speed. The ying speed and angle of attack combine
to generate the lift force, L for ying. This paper is the most important reference to us. In recent, the design of precision balance
and aerodynamic characteristic for micro aerial vehicle to measure
lift, drag, rolling-moment, and pitching-moment of a MAV was reported by Suhariyono et al. [12], but measurement is for the xed
wing MAV only, not for apping wing, the measurement of apping wing is critical. Only Singh et al. [11] studied an experimental
apparatus that incorporates apping wings and measures the small
amount of thrust generated by these wing motions is described.
This methodology is used to measure the thrust generated by two
wings at different wing pitch settings. Also, the effect of change
in pitch phase during a apping cycle is examined experimentally.
Regarding the simulation, Larijani [6] proposed a nonlinear aeroelastic model for the study of apping wing ight in the 2001, this
paper conducted the Huangs [5] numerical analysis for the apping wing MAV later.
From Refs. [9] and [7] we know the three-dimensional movement of many birds apping is used the standard NACA shape
airfoil as wings, but actual apping wing of MAV to be restricted
in the volume and the weight. Its unlikely to use the NACA series
of wing section. On the contrary, the most of the apping wing
for MAV, a planar membrane wing are used primarily. In order to
imitating the insect utter and the ight pattern, therefore, this
investigation does take the planar membrane wing as a study target vehicle, discussing its aerodynamic characteristic and to predict
average lift force, L as the criteria weight to manufacture a future
MAV. The actual MAV was made by the wingspan is 15 cm, the
mean chord is 5 cm, the weight is 8 g, the wing area is 75 cm2 ,
the apping frequency is 25.58 Hz of a apping wing MAV.
2. Numerical analysis
Wm
U
Re
J
AR
different weight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . g
ying speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . m/s
Reynolds number
advance ratio
aspect ratio
density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . kg/m3
length of chord . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . m
the frontal area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . m2
time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . s
slope angle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . rad
dynamic viscosity
c
S
t
C k =
(2)
2L
CL =
(3)
U 2 S
Thrust coecient:
CT =
2T
(4)
U 2 S
Reduced frequency:
fc
K=
(5)
Advance ratio:
J=
U
2 f R
(6)
Reynolds number:
Re =
B n n B 0 0 < (given value)
C Ut
4 f R 2
AR
(7)
1
g t
r grad ) = S
g + div( u
(1)
385
Fig. 1. The grid disposition of the rigid wing section for a three-dimensional planar membrane wing.
Fig. 2. The grid system of the rigid wing section for the three-dimensional NACA
2412 under the upstroke and downstroke.
386
Fig. 3. The comparison of lift coecient, C L under different frequencies for the NACA 2412.
Fig. 4. The comparison of the lift and drag coecient, C L , C D under different K (AOA = 0 ).
fore it may bring the upward lifting force, L for the plate wing.
While the t / T = 3/6 in the perigee position of downstroke, instantaneously, the turbulent ow can be absorbed because of the big
reacting force. While the t / T = 4/6 starting upstroke, then curls
up the clockwise rotation of the wing tip vortex, the strong turbulent ow causes of the low pressure region for lower wing surface,
therefore the negative lifting force, L is not favor for the MAV
ight.
While K = 0.1, no matter how the downstroke or upstroke
is, the wing tip vortex appears smooth. While K = 0.3, the wing
387
(a) t / T = 0/6
(b) t / T = 2/6
(c) t / T = 3/6
(d) t / T = 5/6
tip vortex can be seen obviously and the average vortex velocity
is 8.02 m/s for the wing tip. As a result of the K increase can
cause the maximum vortex velocity increasing quickly for the wing
tip, wing tip vortex became obvious, it affects the pressure between upper and lower surfaces of airfoil, and inuences on lifting
force, L and thrust force, T as well. Regardless of the increasing
of K , the upstroke and downstroke have the same clockwise and
counterclockwise strength of the vortex, therefore the positive and
the negative of lifting force, L is mutually offset. This causes the
average lifting force equal to zero. This result veries that C L and
C D of different K at AOA = 0 as our expectation.
3.4. A three-dimensional case of planar membrane wing in different
angle of attack K = 0.3, AOA = 0 , 5 and 10
3.4.1. Lifting force and thrust force
K = 0.3, AOA = 0 , 5 and 10 , investigates the increasing of K
to inuence on the lift coecient, C L and the drag coecient, C D .
Fig. 8 is the comparison of the lift coecient, C L and the drag
coecient, C D under the different angle of attack, so the increasing angle of attack conducive to favor the lifting force, L and the
thrust force, T generation, while in downstroke the positive lift coecient, C L becomes the proportion to increase. While AOA = 0 ,
the maximum lift coecient, C L is 0.3. While AOA = 5 , the maximum lift coecient, C L is 0.5. While AOA = 10 , the maximum
lift coecient, C L is 0.7. While in upstroke, the negative lift coefcient, C L becomes the proportional reducing actually. While the
AOA = 0 , the smallest lift coecient, C L is 0.3. While AOA = 5 ,
the smallest lift coecient, C L is 0.15. While AOA = 10 , the
smallest lift coecient, C L is 0. According to this, while AOA = 10 ,
the lifting force, L is no longer negative. Thus, the angle of attack
388
(a) t / T = 0/6
(b) t / T = 2/6
(c) t / T = 3/6
(d) t / T = 5/6
Fig. 7. The velocity vector diagram of a apping cycle (K = 0.3, AOA = 0 ).
Fig. 8. The comparison of the lift and drag coecient, C L , C D under different AOA (K = 0.3).
(a) t / T = 0/6
389
(b) t / T = 2/6
(c) t / T = 3/6
(d) t / T = 5/6
Fig. 9. The velocity vector diagram of a ap cycle (K = 0.3, AOA = 10 ).
Table 1
The average lifting force and thrust force of the different angle of attack (K = 0.3).
AOA = 0
AOA = 5
AOA = 10
CL
CD
L (g)
T (g)
0
0.1875
0.3625
0.018
0.0325
0.0775
0
4.39
8.4874
0.4214
0.7609
1.8146
Fig. 10. Design drawing and three views of the apping wing MAV.
390
Table 2
The various components weight of the MAV.
Components
Weight (grams)
2.3
1.5
1.5
0. 7
2
8
and the wing thickness is 0.03 cm. The overall design conguration
is shown in Fig. 10.
4.1. Overall design
4.1.1. Estimation of total weight to take off
The prediction of numerical calculation tells us, the gross
weight has to be lower than 8.78 g to be able to y. Therefore
aspects of weight to take off is estimated, weights of various components are listed in Table 2. From Table 2 we can see that the
total weight estimate to take off is 8 g, lower than 8.78 g. Eliminating crosswind shear and projection angle problems, this apping
wing micro aerial vehicle should be able to y ideally.
391
(a)
(b)
(c)
(d)
Fig. 12. The MAV ying test.
392
(3)
(4)
(5)
(6)
Acknowledgements
The authors acknowledge the support of the MRL of the ITRI,
ROC 2007 and the funding of the National Science Council in Taiwan under the contracts of NSC 94-2212-E-216-004.
References
[1] S.K. Banala, S.K. Agrawal, Design and optimization of a mechanism for out-ofplane insect wing-like motion with twist, Journal of Mechanical Design (2005)
841844.
[2] Electric remote control airplane design, 2006, http://home.kimo.com.tw/
epowery/diypage.htm.
[3] Fluent 6.1, Users Guide, 2003.
[4] Gambit 2.0, Users Guide, 2003.
[5] Shi-Ming Huang, Numerical simulation of ow over a apping wing, PhD Dissertation, National Cheng Kung University, 2004.
[6] R.F. Larijani, J.D. DeLaurier, A nonlinear aeroelastic model for the study of apping wing ight, in: T.J. Mueller (Ed.), Fixed and Flapping Wing Aerodynamics
for Micro Air Vehicle Applications, in: AIAA Progress in Aeronautics and Astronautics, vol. 195, AIAA, 2001, pp. 399428 (Chapter 18).
[7] C.S. Lin, C. Hwu, W.B. Young, The thrust and lift of an ornithopters membrane
wings with simple apping motion, Aerospace Science and Technology 10 (2)
(March 2006) 111119.
[8] Micro aerial vehicle, 2006, http://bicoreeos.solarbotics.net/2nd.gen-mavCG.mav/.
[9] M.F. Neff, D. Hummel, Euler solutions for a nite-span apping wing, in: Conference on Fixed, Flapping and Rotary Wing Vehicles at Very Low Reynolds
Numbers, 2000, pp. 57.
[10] T. Nick Pornsin-sirirak, Y.C. Tai, H. Nassef, C.M. Ho, Titanium-alloy MEMS wing
technology for a micro aerial vehicle application, Sensors and Actuators A:
Physical 89 (2001) 95103.
[11] B. Singh, M. Ramasamy, I. Chopra, J.G. Leighman, Insect-based apping wings
for micro hovering air vehicles: Experimental investigations, in: Proceedings of
the 60th Annual Forum of the American Helicopter Society, June 2004.
[12] A. Suhariyono, J.H. Kim, N.S. Goo, H.C. Park, K.J. Yoon, Design of precision
balance and aerodynamic characteristic measurement system for micro aerial
vehicles, Aerospace Science and Technology 10 (2) (March 2006) 9299.
[13] I.H. Tuncer, M. Kaya, Thrust generation caused by apping airfoils in a biplane
conguration, Journal of Aircraft 40 (2003) 509515.