Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Journal of Bionic Engineering 7 (2010) 134–141

Improvement of Artificial Foldable Wing Models by Mimicking the


Unfolding/Folding Mechanism of a Beetle Hind Wing
Azhar Muhammad1,2, Quoc Viet Nguyen1,2, Hoon Cheol Park1,2,4, Do Y. Hwang4,
Doyoung Byun2,3,4, Nam Seo Goo1,2,4
1. Department of Advanced Technology Fusion, Konkuk University, Seoul 143-701, Korea
2. National Research Laboratory for Biomimetics and Microsystems, Konkuk University, Seoul 143-701, Korea
3. Departmenet of Aerospace Engineering, Konkuk University, Seoul 143-701, Korea
4. Artificial Muscle Research Center, Konkuk University, Seoul 143-701, Korea

Abstract
In an attempt to realize a flapping wing micro-air vehicle with morphing wings, we report on improvements to our previous
foldable artificial hind wing. Multiple hinges, which were implemented to mimic the bending zone of a beetle hind wing, were
made of small composite hinge plates and tiny aluminum rivets. The buck-tails of rivets were flared after the hinge plates were
assembled with the rivets so that the folding/unfolding motions could be completed in less time, and the straight shape of the
artificial hind wing could be maintained after fabrication. Folding and unfolding actions were triggered by electrically-activated
Shape Memory Alloy (SMA) wires. For wing folding, the actuation characteristics of the SMA wire actuator were modified
through heat treatment. Through a series of flapping tests, we confirmed that the artificial wings did not fold back and arbitrarily
fluctuate during the flapping motion.
Keywords: hind wing, unfolding, folding, shape memory alloy, folding ratio, artificial wing
Copyright © 2010, Jilin University. Published by Elsevier Limited and Science Press. All rights reserved.
doi: 10.1016/S1672-6529(09)60185-2

of wings: a pair of hardened forewings called elytra, and


1 Introduction a pair of morphable hind wings. The hind wings are
Recent efforts to understand the mechanism of shielded underneath the fore wings during ground lo-
insect flight may help us to realize Nano- and Micro-Air comotion, and extended just before flying. Thus, they can
Vehicles (NAV/MAV)[1]. For an insect-like flyer, fly in the air and crawl on the ground. This unique fea-
aerodynamics in a low Reynolds number regime[2] and ture has encouraged us to study and realize an artificial
flexible wings[3] are important issues. To identify foldable hind wing. In this work, for easier observation
mechanisms of aerodynamic force generation in a low with bare eyes, we selected a beetle called Allomyrina
Reynolds number environment, some features of insect dichotoma (Fig. 1), a species in the order Coleoptera
flight, such as the clap-and-fling mechanism[4], lead- which is relatively larger than others.
ing-edge vortices[5], pitching-up rotation[6], and wake-
capture[7] have been explored.
The wing is an essential member for all flying in-
sects. Insect wings are membranous and fragile. Amaz-
ingly, however, they are still strong enough to endure the
aerodynamic forces produced by flapping wing motion.
Moreover, some winged insects are able to fold and
unfold their wings. Insects equipped with foldable wings
are grouped into several orders including Coleoptera
(beetle) and Dermaptera (earwig). They have two types Fig. 1 Allomyrina dichotoma (male).

Corresponding author: Hoon Cheol Park


E-mail: hcpark@konkuk.ac.kr
Muhammad et al.: Improvement of Artificial Foldable Wing Models by Mimicking the
Unfolding/Folding Mechanism of a Beetle Hind Wing 135
[8]
In our previous work , we demonstrated two types 2 Unfolding and folding of beetle hind wing
of artificial hind wings which can be unfolded by ap-
We investigated the hind wing folding motions of a
plying an electric field to the installed Shape Memory
beetle, Allomyrina dichotoma, after tethering the beetle.
Alloy (SMA) actuators[9]. One of the artificial wings was
A carbon rod was attached to a place at its dorsal and the
designed based on a so-called basic model[10] and could
other end was attached to a fixture. The unfolding
be unfolded in the out-of-plane mode. After observing
process was captured by using a high speed camera
the wing unfolding behavior of Allomyrina dichotoma,
recorder (Photron, Fastcam Ultima APX, model 120K)
we found that it actually unfolds its hind wing nearly in a
at 4000 frames per second, shown sequentially in Fig. 2.
plane. Therefore, we made another artificial wing that
The folding images were captured by a digital camera
can be unfolded in a plane by activating the installed
(Canon G9) at 30 frames per second. Because a full set
SMA wires. Hass and Beutel explained the details of
muscle actuation and principles for wing unfolding of of sequential pictures for wing unfolding and time spent
Pachnoda marginata[11]. The wing unfolding is very for the wing unfolding (~0.6 s) were reported in Ref. [8],
similar to that of Allomyrina dichotoma, except Pach- we include only a few images in Fig. 2a. The figure
noda marginata does not open the elytra during the hind shows that the beetle consecutively opens its forewings
wing unfolding. As Hass and Beutel suggested, we and folded hind wings, sweeps the hind wings forward,
examined the change in angles between two major veins starts to flap the hind wings, and fully unfolds the hind
– the Radius Anterior (RA) and Medial Posterior (MP) – wings nearly in a plane, just before flight. Thus, this
and tried to reproduce the so-called “scissor-like mo- species fully unfolds the outer part of its hind wings
tion” of the two major veins[11] in an artificial wing. during the initial flapping motion within a relatively
In this work, we first observed wing folding and short time.
unfolding of Allomyrina dichotoma, and then confirmed In contrast to unfolding, the folding process of the
that wing folding is also occurring in a plane. The fold- beetle’s hind wing is more random and requires more
ing motion of an artificial wing was demonstrated by time. This may be because a beetle can still crawl even
reversing the actuation deformation of the SMA wire with partially-folded wing configuration, while the hind
actuators, which were properly heat-treated before being wing must be unfolded immediately for flight. Fig. 2b
installed in the artificial wing. We also modified the shows that the hind wings are completely folded after the
hinges such that the folding and unfolding motions be- elytra are closed. Thus, the folding also takes place in a
come faster and wing veins remained straight after un- plane.
folding and during the flapping motion. All the charac- The wing folding and unfolding motions are trig-
terization parameters were evaluated to confirm the gered by contracting various flight muscles in the thorax,
similarity between the artificial wing and the real beetle as Hass and Beutel pointed out[11]. Similarly, a proper
wing. Afterward, a series of flapping tests was con- actuation of muscles may cause morphological change
ducted to examine whether or not the artificial hind wing in the hind wings of Allomyrina dichotoma. Structural
folds back during wing flapping, and to inspect its members of a hind wing, as shown in Fig. 3, create
flexibility. three main mechanisms with which to complete the wing

(a) Unfolding

(b) Folding
Fig. 2 Sequence of free unfolding and folding motions.
136 Journal of Bionic Engineering (2010) Vol.7 No.2

shape change: (1) scissor-like motion between the RA 3 Artificial foldable hind wing
and MP with the Medial Bridge (MB) as the pivot; (2)
3.1 Design principle
bending motion of the Bending Zone (BZ); and (3) ro-
As previously described, there are many muscles
tation motion of the Marginal Joint (MJ). These are
and wing members involved in wing folding and un-
nearly the same mechanisms found in another beetle,
folding motions. Hass and Beutel suggested other pas-
Pachnoda marginata[11], even though it does not open its
sive mechanisms such as elastic energy storage during
elytra during wing unfolding. Thus, the MP, RA, RA3,
wing folding at specific locations in a hind wing, where
RP3+4, BZ, MJ, MB, and the movable joint work like an
resilin, a highly elastic material, is present[10]. Large
in-plane hinge structure, resulting in stable in-plane
elastic bending deformations of the BZ, in Fig. 3, is
folding and unfolding motions of the hind wing.
another important element in making in-plane wing
Because details of morphological change during
folding and unfolding possible. However, exactly
wing unfolding were explained in Ref. [8], we included
copying a beetle hind wing is almost impossible, be-
only the final angle changes among vein structures in
cause no available material can reproduce real flight
Fig. 4. As shown in Fig. 4, a hind wing of Allomyrina
muscle and resilin. Therefore, we need to properly ap-
dichotoma is folded at two locations: the anal fold (fold
proximate the wing folding and unfolding, and similarly
of anal field to medial field in Fig. 3) and the apical fold
replace the vein structures by using existing and avail-
(fold of apical field to medial field). A large elastic
able engineering materials and elements.
bending deformation in BZ is also noticeable in Fig. 4b.
In this study, we neglected the anal fold because of
its relatively small contribution to the folding ratio. The
vein structures were made from composite materials and
the BZ was mimicked by using multiple hinges. Unlike
our previous work[8], tiny aluminum rivets were used for
the hinges to improve folding and unfolding perform-
ance. Because two-way SMA wires were not available in
(a) Regions and names of veins
the market, we separately demonstrated wing folding
and unfolding by using two artificial wings equipped
with SMA wires which have different one-way shape
memory effects; i.e., one for folding and the other for
unfolding. We did not attempt to exactly mimic the
geometry and stiffness of a real hind wing, because the
main purpose of the current work was to explore feasible
(b) Joints methods of mimicking the folding and unfolding of a
beetle hind wing.

3.2 Fabrication
For easier fabrication, we made artificial wings with
a 10.5 cm span, which is about two times longer than a
typical hind wing of Allomyrina dichotoma. The pattern
(c) Folding structures
of veins was taken from a real beetle hind wing as shown
Fig. 3 Real hind wing of Allomyrina dichotoma.
in Fig. 5. The vein structures were made of 0.27 mm
thick glass/epoxy composites (GEP, SK Chemicals,
Korea), which were attached to the upper and lower
sides of an artificial membrane. The artificial membrane
was made of 15 ȝm thick polypropylene film, even
(a) Unfolding (b) Folding though a real beetle membrane has non-uniform thick-
Fig. 4 Unfolding and folding configurations of Allomyrina ness[12]. Because the size of each part was small, we
dichotoma hind wing[8].
Muhammad et al.: Improvement of Artificial Foldable Wing Models by Mimicking the
Unfolding/Folding Mechanism of a Beetle Hind Wing 137
constructed CAD models of all parts by using commer- carbon rods used to mimic the MB, MJ, and the move-
cial CAD software (CATIA, Version 7, Dassault able joint were also replaced by the same rivets as shown
Systemes, France) prior to the cutting process. For high in Fig. 5b. In the current artificial wing, the diameter of
precision cutting, we used a CNC machine (Manix, each hole was slightly larger than 1.15 mm and the shaft
MM300-S, Woosung E&I Ltd, South Korea). diameter of an aluminum rivet was 1.15 mm. The rivet
In our previous work[8], multiple in-plane hinges buck-tails were flared to obtain tight clearance in each
were used to mimic the BZ of a real beetle hind wing, as in-plane hinge. Thus, a sliding fit was available. In this
shown in Fig. 3a. A piece of composite hinge plate and way, we could reduce the friction in each hinge and
two short carbon rods, which are marked with white maintain a straight wing shape after assembly. Also, the
dotted circles in Fig. 5a, were used to form each in-plane artificial wing did not arbitrarily fluctuate during flap-
hinge. As shown in Fig. 6a, each carbon rod was tightly ping motion.
fit (interference fit) to a hole in its own hinge plate be-
cause the free ends of the rods could not be flared. Be- 3.3 Actuation system
cause of this, friction in the in-plane hinges was rela- For the actuation system to fold and unfold the
tively large, and fast wing folding and unfolding motions proposed artificial hind wing, we used a set of SMA
were impossible. In the current work, we replaced the wires (Kantoc, Ltd.) which were also used for unfolding
carbon rod with a tiny rivet to reduce friction in the the previous artificial wing[8]. SMA wire, which is also
in-plane hinges, as shown in Fig. 5b and Fig. 6b. Short known as “muscle wire,” is a metal alloy that can
memorize its initial shape. SMA wire is typically weak
and plastically deformable when its temperature is lower
than its transformation range. A typical SMA wire
available in the market is specially treated so that it has a
one-way shape memory effect. Therefore, if we arbi-
trarily bend a SMA wire and then increase its tempera-
(a) Previous model ture within or higher than the transformation range, the
wire becomes stiffer and more elastic, and then returns to
its original shape[9], as shown in Fig. 7a. In Fig. 7a, the
original shape of the SMA wire is straight. The SMA was
arbitrarily deformed and then heated up to recover its
original shape. If we treat it differently, we can reverse
(b) Current model
the deformation, but the SMA then loses its original
Fig. 5 Artificial foldable hind wings. shape memory ability. Fig. 7b shows a heat-treated SMA
wire whose original shape is arbitrary bent. The wire
then goes back to its original shape when it is heated up
after being straightened.

(a) Previous model

(a) Original SMA wire

(b) Current mode (b) Heat-treated SMA wire


Fig. 6 Configurations of in-plane hinges. Fig. 7 SMA wire behaviors.
138 Journal of Bionic Engineering (2010) Vol.7 No.2

We installed the original SMA wires (used in the We raised the temperature of the SMA wires by
previous work[8]) in the current modified artificial wing applying 5 V and 1.5 A from an external power supply
to demonstrate the unfolding motion. Meanwhile, wing (Agilent, E3646A). The sequential unfolding and fold-
folding motion could be achieved by installing another ing motions of the artificial wing are shown in Fig. 9.
set of SMA wires with a reversed shape memory effect. The angle between RA and MP increased during un-
We properly treated the original SMA wire by following folding and decreased during folding due to scissor-like
a heat treatment process described in Ref. [8], where a motion triggered by the installed SMA wires at the wing
SMA wire was plastically deformed such that it had the base. During folding, the artificial BZ created a slight
desired shape and then baked at 250 ÛC for 30 min. This curvature so that the artificial RA could bend, and the
process was selected to make the wire’s shape trans- artificial RA3 generated a large rotation about the mar-
formation temperature the same as that of the original ginal joint. On average, wing unfolding was completed
SMA wire. The process was carried out in an environ- within about 3 s and the wing folded in about 4 s.
mental chamber (Model EC1x, Sun Electronic Systems
Inc.). Because a heat-treated wire loses its original shape
memory effect, we had to fabricate another artificial
wing to demonstrate wing folding using heat-treated
SMA wire. Thus, the artificial wing folding and un-
folding were separately demonstrated with two artificial
wings equipped with differently heat-treated SMA
wires.
Fig. 8 shows the two sets of SMA wires used in this
work: the original SMA wires for wing unfolding, and
the heat-treated wire used for wing folding. The shapes Fig. 8 Shapes of the SMA wires installed on the in-plane artificial
of the original SMA wires used for unfolding are shown hind wing when unfolding and folding.
at the bottom of Fig. 8. The bent angle of the SMA wire
installed at the wing base is about 20Û, and the other
SMA wire has a slight curvature so that it can stay along
the leading edge vein (RA through RA3) when they are
installed in wing vein structures. Then, the two pieces of
SMA wire were properly bent to construct the folded
configuration. The shapes of the two deformed SMA
wires look like the shapes shown at the top of Fig. 8. In
this case, the bent angle of the SMA wire installed at the
folded wing base is about 10Û. When heated, this SMA
wire created a scissor-like motion of the RA and MP,
and the other SMA wire installed along the leading edge
vein generated rotation as well as translation of the
apical field (outer wing).
To demonstrate wing folding, we prepared the
heat-treated SMA wires as previously described, and
another artificial wing. The memorized shapes of the
two heat-treated SMA wires looked like the shapes at the
top of Fig. 8. They were installed in the unfolded wing
after being properly deformed, just like the shapes at the
bottom of Fig. 8. When heated, the two SMA wires re-
covered their own shapes and thus wing folding was (a) Unfolding

completed. Fig. 9 Demonstration of artificial foldable hind wing.


Muhammad et al.: Improvement of Artificial Foldable Wing Models by Mimicking the
Unfolding/Folding Mechanism of a Beetle Hind Wing 139
and folding ratio. These data were acquired by examin-
ing images taken by a high-speed camera and are sum-
marized in Table 1. The current artificial wing is both
larger and heavier than a real beetle hind wing. A typical
unfolding time was about 3 s, which was less than half of
the time spent by the previous wing[8], but is larger than
that required by a real beetle wing (0.05 s). We believe
that 3 s to 4 s for artificial wing folding and/or unfolding
is still fast enough for a real application. We made other
improvements in the current artificial wing model. After
complete wing unfolding, the angle between RA and MP
was 19Û and another angle between RA and MJ through
the tip of RA3 was 152Û became closer to those of a real
beetle wing, which are 20Û and 154Û, respectively. In
wing folding, the angle between RA and MP was 12Û, is
close enough to that of a real beetle hind wing of 10Û, yet
the angle between RA and MJ through the tip of RA3
was 50Û which was still larger than the real one of 39Û.
The discrepancy mainly came from a lack of actuation
force produced by the SMA wires, which might not be
enough for compact folding around marginal joint and
(b) Folding
Fig.9 Continued.
moving joint areas. The folding ratio, which is the ratio
between the areas of unfolded and folded configurations,
4 Comparison was improved in the two current wings. Note that the
We compared a real beetle hind wing, a previous folding ratio of a real beetle hind wing changes from
artificial wing (only for unfolding), and two current 2.32 to 1.70 if the anal fold of the real wing is neglected
modified artificial wings (for folding and unfolding) in for a fair comparison with the folding ratios of the arti-
terms of geometry, time spent for folding and unfolding, ficial wings, where the anal folds were ignored.

Table 1 Comparison between real and artificial hind wings

Artificial hind wing


Parameter Real hind wing Previous Present
(unfolding only) Folding wing Unfolding wing
Length (cm) 5.2 10.5 10.5
Weight (g) 0.16 1.06 1.1
Thickness of membrane (Pm) 3~6 15 15

Typical time Folding Random Not available 4 Not available


spent (s) Unfolding 0.05 7~8 Not available 3

Projected wing Folding 3.29 21.88 21.22 20.317


area (cm2) Unfolding 7.62 32.79 35.76 34.93

Angle between Folding 10 12 12 10


RA and MP (Û) Unfolding 20 16 19 19

Angle between RA and Folding 39 51 50 41


MJ – tip of RA3 (Û) Unfolding 154 148 154 152
Folding ratio 2.32 (1.70*) 1.50* 1.70* 1.72*

*Without anal fold.


140 Journal of Bionic Engineering (2010) Vol.7 No.2

5 Flapping test 6 Conclusion


A beetle hind wing does not fold back during flight In this work, our previous artificial foldable hind
unless the beetle is exhausted. This is another curiosity wing was successfully improved by modifying the
because the leading edge of a hind wing is almost hinges. Through the modification, the current artificial
separated near the MJ. We conducted a series of flapping wing demonstrated swifter folding/unfolding motions,
tests for the artificial wings after installing them in a closer angles among veins before and after folding, and a
motor-driven flapper to see if they would fold back higher folding ratio. In addition, folding motion was
during the flapping motion. The flapping system could demonstrated by reversing the actuation deformation of
flap at 30 Hz with 120Û flapping angle without wings. the SMA wires through a heat treatment process. Flap-
When we installed the artificial wings, however, the ping test showed that the artificial hind wing did not fold
flapping frequency happened to drop to 9 Hz for the back during the flapping motion, and possesses an
same 120Û flapping angle. Fig. 10 shows the flapping adequate flexibility when flapped at a 9 Hz wing beat
motions during upstrokes and downstrokes. From these frequency with a 120Û flapping angle. Information ac-
figures, we note that the artificial hind wings do not quired from this research should be helpful in develop-
experience folding back motion or arbitrary fluctuation ing portable flapping-wing MAVs with morphing wings.
during flapping. Because the wing vein structures re-
Acknowledgement
main almost straight, we conclude that the artificial
wings are stiff enough to sustain aerodynamic and iner- This research was financially supported by the
tial forces when the wings are under flapping motion at Korea Science and Engineering Foundation Grant (Na-
9 Hz with a 120Û flapping angle. tional Research Laboratory Program, R0A-2007-000-
200012-0) and the Korea Research Foundation (KRF-
006-005-J03301). This work was partially supported by
the 2009 KU Brain Pool of Konkuk University. The
authors gratefully acknowledge this support.

References
[1] Pines D J, Bohorquez F. Challenges facing future mi-
cro-air-vehicle development. Journal of Aircraft, 2006, 43,
290–305.
[2] Shyy W, Lian Y, Tang J, Viieru D, Liu H. Aerodynamics of
Low Reynold Numbers Flyers, Cambridge University Press,
New York, USA, 2008.
(a) Downstroke [3] Shyy W, Berg M, Ljungqvist D. Flapping and flexible wings
for biological and micro air vehicles. Progress in Aerospace
Sciences, 1999, 35, 405–505.
[4] Fogh T W. Quick estimates of flight fitness in hovering
animals, including novel mechanisms for lift production.
Journal of Experimental Biology, 1973, 59, 169–230.
[5] Ellington C P, Berg C V D, Willmott A P, Thomas A L R.
Leading-edge vortices in insect flight. Nature, 1996, 384,
626–630.
[6] Sun M, Tang J. Unsteady aerodynamic force generation by a
model fruit fly wing in flapping motion. Journal of Ex-
perimental Biology, 2002, 205, 55–70.
[7] Dickinson M H, Lehmann F O, Sane S P. Wing rotation and
(b) Upstroke
the aerodynamic basis of insect flight. Science, 1999, 284,
Fig. 10 Flapping test. 1954–1960.
Muhammad et al.: Improvement of Artificial Foldable Wing Models by Mimicking the
Unfolding/Folding Mechanism of a Beetle Hind Wing 141

[8] Muhammad A, Park H C, Hwang D Y, Byun D, Goo N S. Science, 1996, 263, 1651–1658.
Mimicking unfolding motion of a beetle hind wing. Chinese [11] Haas F, Beutel R G. Wing folding and the functional mor-
Science Bulletin, 2009, 54, 2416–2424. phology of the wing base in Coleoptera. Zoology, 2001, 104,
[9] Duerig T W, Melton K N, Stockel D, Waymanm C M. En- 123–141.
gineering Aspects of Shape Memory Alloys, Butterworth- [12] Jin T, Goo N S, Woo S C, Park H C. Use of a digital image
Heinmann Ltd, London, 1990. correlation technique for measuring the material properties
[10] Haas F, Wotton R J. Two basic mechanisms in insect wing of beetle wing. Journal of Bionic Engineering, 2009, 6,
folding. Proceedings of the Royal Society B, Biological 224–231.

You might also like