The Influence of Bat Wings For Producing Efficient Net Body Forces in
Bio-inspired Flapping Robots
Julian Colorado
1
, Claudio Rossi
2
, Antonio Barrientos
2
and Diego Patino
1
1
School of Engineering, Pontificia Universidad Javeriana, Cr. 7 No. 40-62, Bogot
´
a, Colombia
2
Centre for Automation and Robotics, Universidad Polit
´
ecnica de Madrid, C/ Jos
´
e Abascal 2, 28006, Madrid, Spain
Keywords:
Bio-inspired MAV, Shape Memory Alloy Actuators, Bats.
Abstract:
Bat wings contain dozens of joints that enable the animal to perform aggressive maneuvers by means of
changing the wing shape during flight. There is evidence that the inertial forces produced by their wings during
flapping have a key role in the attitude movements of the animal, i.e. aerial rotations. In fact, bats efficiently
generate net body forces to manoeuvre by taking advantage of their large wing-to-body mass ratio. In this
paper, the following question is formulated: Could a Micro Aerial Vehicle (MAV) inspired by the biomechanics
of bats take advantage of the morphing-wings aimed at increasing net body forces? Using BaTboT, a novel
bat-like MAV with highly articulated wings actuated by shape memory alloy actuators, our goal is to quantify
the effects of different wing modulation patterns on the generation of net body forces. Experiments are carried
out to confirm the important physical role that changing the wing shape enables: the contraction time of the
wings (upstroke) should be faster than the extension time (downstroke), taking about 37.5% of the wingbeat
period. This modulation pattern has enabled a lift force increment of 22% (from L = 0.92N to L = 1.12N),
abrupt drag reduction (from D = 0.22N to D = 0.11N) and also an increase of net body forces (F
net
) about
28% compared to those wing modulation patterns defined with equal periods for contraction/extension. These
findings can be useful for accurate dynamics modelling and efficient design of flight controllers applied to
morphing-wing micro aerial vehicles.
1 INTRODUCTION
Morphing-wing aircrafts have emerged as a direction
to enhance the efficiency of flight by changing the
wing profile (Gomez and Garcia, 2011). There is
a growing interest in the energy cost of flight and
learning from nature is the key to optimise efficiency
(Lentink and Biewener, 2010), (Hedenstrom et al.,
2009). Unlike insects or birds, bats have heavy mus-
cular wings with joints that enable higher degree of
dexterity (Swartz et al., 2012).
This allows bats to save energy during flight
than any other flying creature (Winter and Helversen,
1998), (Riskin et al., 2012). Also, their massive wings
undergo large accelerations that are caused by inertial
forces with a significant contribution for maneuvering
(Iriarte-Diaz et al., 2011), (Riskin et al., 2010). This
opens the path to a different way of attitude control for
Micro Aerial Vehicles (MAV) that currently manoeu-
vre either by means of multiple rotating-propellers
(like helicopters) or moving appendices such as flaps
and rudders as in the case of fixed-wing MAV. In-
elbow
joint
digits
wrist joint
silicone
membrane
shoulder
joint
SMA artificial
muscles
elbow joint
contraction
extension
θ
elbow
SMA stroke: 4mm
Figure 1: BaTbot, a bat-like morphing-wing MAV based on
Bahlman et al., robotic wing design (Bahlman et al., 2013)
but driven by Shape Memory Alloy (SMA) artificial mus-
cles.
ertial forces are significant in bats because the mass
of their wings comprise a significant portion of total
body mass (large wing-to-body mass ratio) (Tholles-
son and Norberg, 1991).
Inspired by the morphology of the Cynopterus
brachiotus, (Bahlman et al., 2013) proposed the de-
528
Colorado J., Rossi C., Barrientos A. and Patino D..
The Influence of Bat Wings For Producing Efficient Net Body Forces in Bio-inspired Flapping Robots.
DOI: 10.5220/0005085805280532
In Proceedings of the 11th International Conference on Informatics in Control, Automation and Robotics (ICINCO-2014), pages 528-532
ISBN: 978-989-758-040-6
Copyright
c
2014 SCITEPRESS (Science and Technology Publications, Lda.)
(lift)
(weight)
(thrust)
airflow
α
6D force
sensor
y
0
z
0
F
net
Downtroke
(drag)
f
D
f
L
f
Z
f
T
(a) (b)
Figure 2: Lift, drag and thrust are measured as a function of the airflow speed (V
air
) and angle of attack (α); a) end of
downstroke shown from the simulation dynamic-model, b) end of upstroke shown from the real setup.
sign of a multi-articulated wing structure from which
the robotic model in Figure 1 was based. Results
from (Bahlman et al., 2013) reported experimental
measurements that detail the inertial and aerodynamic
power involved in the cost of flapping and the contri-
bution of the wing inertia in the overall cost of flight.
Bahlman’s original wing design was modified in pre-
vious work (Colorado et al., 2012) so the robot could
be actuated by Shape Memory Alloy (SMA) artificial
muscles. Table 1 shows morphological parameters of
BaTboT. Each wing of the robot has six degrees of
freedom (dof): two-dof at the shoulder driven by a
standard DC-motor, one-dof at the elbow driven di-
rectly by the SMA actuators, and three under-actuated
dof at the wrist joint. The body is a six-dof floating
base with rotations designated to roll, pitch and yaw
following aerodynamic conventions. In this paper,
we aim at identifying how to properly modulate the
wing geometry for increasing net body forces (F
net
)
that generate propulsion (Colorado et al., 2013) (cf.
Figure 2b).
2 METHODS
Force measurements have been used for analysing the
effects of applying different wing modulation patterns
on lift generation, drag reduction and net body force
production. By using data acquired with a wind-
tunnel setup (Colorado et al., 2012; Colorado et al.,
2013), our goal is to quantify how net body forces
can be increased by changing the wing’s geometry in
a proper way, using the morphing-wing mechanism
presented in prior work (Colorado et al., 2012). To
change the wing shape, we have used Nickel Titanium
(NiTi) shape memory alloy actuators manufactured
by Migamotors (cf. Figure 1). The very light struc-
Table 1: Morphological parameters.
Parameter [unit] BaTboT
Total mass m
t
[g] 125
Extended wing length B [m] 0.245
Body width l
m
[m] 0.04
Body mass m
b
[g] 18
Extended wingspan S = l
m
+ 2B [m] 0.53
Extended wing area A
b
[m
2
] 0.05
Humerus length l
h
[m] 0.055
Humerus average diameter 2r
h
[m] 0.006
Radius length l
r
[m] 0.070
Radius average diameter 2r
r
[m] 0.005
Wing mass
1
m
w
[g] 8.2
Membrane thickness [m] 0.0001
1
Wing mass is composed by: humerus, radius, digits-III,IV,V
and SMA actuators.
ture of the SMA actuators make them suitable for the
construction of light wings with muscle-like actuation
similar to the one found in bats. The morphing-wing
mechanism is composed by an antagonistic pair of
SMA actuators acting as artificial triceps and biceps
muscles, which enable BaTboT to track bio-inspired
wing motion trajectory patterns. More details regard-
ing the SMA actuation mechanism can be found in
(Colorado et al., 2012). Basically, the SMA act di-
rectly on the elbow joint (θ
elbow
), thus enabling BaT-
boT to change its wingspan from 40.8cm (wings re-
tracted) to 53cm (wings fully extended). As shown in
the inset of Figure 1, the SMA actuators are connected
to the elbow and wrist joints, allowing the wings to
contract and extend at each wingbeat cycle. During
the downstroke, wings are fully extended in order to
maximise the area and increase lift, whereas during
the upstroke, wings are folded in order to reduce drag.
These modulation patterns for wing contraction and
TheInfluenceofBatWingsForProducingEfficientNetBodyForcesinBio-inspiredFlappingRobots
529
extension are generated by applying a heating signal
(electrical current) that actives the contraction stage
of the SMA actuators.
To accelerate during forward flight, BaTboT must
produce a net force to counteract gravity and over-
come drag. As shown in Figure 2b, this force can
be decomposed into a net force component in the di-
rection of flight that corresponds to the difference be-
tween thrust ( f
T
) and drag (D), and a perpendicular
component that corresponds to the difference between
lift (L) and weight ( f
z
). Thus, net body forces can be
calculated as:
F
net
= ( f
T
− D) + (L − f
z
),
(1)
In (1), both lift (L) and drag (D) force components
depend on the angle of attack α as shown in Figure
2b. Equation (2) shows these force components.
L = f
L
cos(α) − f
D
sin(α)
D = f
D
cos(α) + f
L
sin(α)
(2)
3 RESULTS
Could a Micro Aerial Vehicle (MAV) inspired by the
biomechanics of bats take advantage of the morphing-
wings aimed at increasing net body forces?. To test
such hypothesis, we controlled the SMA actuators to
apply three different morphing patterns for the wings,
as shown in Fig. 3a. The plots show a close-up to a
wingbeat cycle described by the elbow joint (θ
elbow
).
Such profiles differ in the proportion that the wing
takes for contraction and extension during a wingbeat.
Firstly, we can note that the aerodynamic response
is clearly affected by changing the wing modulation
pattern. Figure 3b shows the lift (C
L
) and drag (C
D
)
coefficients that correspond to each wing modulation
pattern of column (a). Both coefficients have been
calculated as:
C
L
= 2L
ρV
2
air
A
b
−1
,
C
D
= 2D
ρV
2
air
A
b
−1
,
(3)
where ρ = 1.2(Kgm
3
) is the air density, A
b
=
0.05(m
2
) is the wing area, L and D are the measured
lift and drag forces at an angle of attack of α = 10
o
.
We measured aerodynamic data using this configu-
ration since the lift-to-drag ratio (L/D) is maximum
at that angle of attack (see the inset at the top plot
of Fig. 3b). On average, the corresponding lift L
and drag forces D are 1.12N and 0.112N respectively.
These measurements correspond to the wing modu-
lation pattern described by the top plot in Fig. 3a.
Table 2 details the numerical values. Despite keep-
ing the same flapping frequency and airspeed within
the wind-tunnel, note how lift and drag are affected
after applying the wing profile described in both mid-
dle and bottom plots in Fig. 3a. The lift coefficient
decreases whereas the drag coefficient increases de-
pending on the angle of attack.
Similarly, net body force production is affected by
changing the wing modulation patterns. We observed
from the Fig. 3c that the upstroke portion of the wing-
beat generates less drag due to the fact that the wing
contracts sufficiently to reduce the area at minimum
span. The experiments have confirmed that the gen-
eration of net body forces (F
net
) decreases when the
wing modulation pattern is defined with equal pro-
portions of contraction/extension during a wingbeat.
The readings are even worst when the contraction dur-
ing upstroke takes longer, as described by the bottom
plots in Fig. 3.
To respond the question formulated at the begin-
ning of this section, experimental results show that
the contraction time of the wing (upstroke) should be
faster than the extension time (downstroke), taking on
average 37.5% of the wingbeat period. We performed
extensive experiments with different wing modulation
patterns; the three presented in Fig. 3a are the most
representative of the studied cases. By comparing the
bias of the net body forces between the middle and
top plots of Fig. 3c, net body force production is in-
creased by 28%. Table 2 summarises the numerical
data corresponding to the experiments carried out in
Fig. 3.
4 CONCLUSIONS
Our findings support the idea that by properly con-
trolling the modulation of the wing geometry, more
efficient flight can be achieved, just as bats efficiently
generate net body forces taking advantage of their
large wing-to-body mass ratio. In fact, we have
demonstrated that with the proper wing kinematics,
lift and net body forces can be increased whereas drag
can be reduced. This leads to a significant increase
in flight efficiency at the benefit of energy consump-
tion. This approach opens the path to the development
of novel attitude controllers (flight control) applied to
flapping MAV, where the inertial forces produced by
the wings are vital on the production of net body ac-
celerations. The possibility of controlling the shape
of the wings has great potential to improve the ma-
neuverability of micro aerial vehicles for operating in
confined spaces.
ICINCO2014-11thInternationalConferenceonInformaticsinControl,AutomationandRobotics
530
Table 2: Influence of wing modulation on flight performance.
Experiment α(deg) V
air
(ms
−1
) f (Hz) C
L
C
D
¯
L(N)
¯
D(N) f
z
(N)
¯
F
net
(N)
Top 10 5 2 − 2.5 1.5 0.152
¯
1.12
¯
0.11 0.77 0.115
Middle 10 5 2 − 2.5 1.23 0.3
¯
0.92
¯
0.22 0.77 0.09
Bottom 10 5 2 − 2.5 0.48 0.17
¯
0.36
¯
0.12 0.77 0.021
0 5 10 15
0
1
2
3
4
(deg)
C
L
,C
D
0 5 10 15
5
10
(deg)
F
L
/F
D
0 5 10 15
0
0.6
1.2
1.8
(deg)
C
L
,C
D
0 5 10 15
0
0.5
1
(deg)
C
L
,C
D
0 2 4 6
0.02
0.021
0.022
0.02
0.021
t (s)
F
net
(N)
0 2 4 6
0.05
0.1
0.15
t (s)
F
net
(N)
0 2 4 6
0.08
0.09
0.1
0.11
0.12
t (s)
F
net
(N)
0 0.1 0.2 0.3 0.4
0
50
100
t (s)
q
3
(deg)
0 0.1 0.2 0.3 0.4
0
50
100
60
t(s)
q
3
(deg)
0.1 0.2 0.3 0.4
0
25
50
75
t (s)
q
3
(deg)
0
C
L
C
D
C
L
C
D
C
L
C
D
extension
(downstroke)
extension
(downstroke)
extension
(downstroke)
contraction
(upstroke)
contraction
(upstroke)
contraction
(upstroke)
(a) (b)
q
elbow
(deg)
q
elbow
(deg)
q
elbow
(deg)
0
(c)
20
40
0.021
0.022
θ
elbow
deg
( )
θ
elbow
deg
( )
L
D
θ
elbow
deg
( )
Figure 3: (Experimental) influence of wing modulation patterns (θ
elbow
) on lift, drag and net body force production (airspeed
of V
air
= 5(ms
−1
) and wingbeat frequency of f = 2.5(Hz)): a) close-up to a wingbeat with changes on wing modulation
patterns, b) impact of changing wing modulation patterns from (a) on the lift coefficient C
L
and the drag coefficient C
D
, c)
impact of changing wing modulation patterns from (a) on net body force production F
net
.
ACKNOWLEDGEMENTS
This work was supported by the Robotics and Cy-
bernetics Group at Technical University of Madrid
(Spain), and by the project entitled Dise
˜
no y con-
strucci
´
on de una plataforma rob
´
otica de exploraci
´
on
y reparaci
´
on de tuber
´
ıas hidrosanitarias, operada
remotamente, identifier 120350227348, funded by
COLCIENCIAS y H&U. It was also funded by
The Spanish Ministry of Education and Science
(DPI2010-17998). The authors are thankful to pro-
fessors Kenny Breuer and Sharon Swartz from Brown
University for their knowledge about bat flight and for
providing the wind-tunnel facility for the initials ex-
periments carried out in (Colorado et al., 2012; Col-
orado et al., 2013). Furthermore, to Joe Bahlman
from Brown University for the original design of
the multi-articulated robotic wing used by BaTboT
(Bahlman et al., 2013).
REFERENCES
Bahlman, J., Swartz, S., and Breuer, K. (2013). Design
and characterization of a multi-articulated robotic bat
wing. Bioinspir. Biomim., (8):016009.
Colorado, J., Barrientos, A., Rossi, C., Bahlman, J., and
Breuer, K. (2012). Biomechanics of smart wings in a
bat robot: morphing wings using sma actuators. Bioin-
spir. Biomim., (7):036006.
Colorado, J., Barrientos, A., Rossi, C., and Parra, C. (2013).
Inertial attitude control of a bat-like morphing-wing
air vehicle. Bioinspir. Biomim., (8):016001.
Gomez, J. and Garcia, E. (2011). Morphing unmanned
aerial vehicles. Smart Mater Struct, (20):1–16.
Hedenstrom, A., Johansson, L., and Spedding, G. (2009).
Bird or bat: comparing airframe design and flight per-
formance. Bioinsp. Biomim, vol. 4, no.1, pp. 5001.
Iriarte-Diaz, J., Riskin, D., Willis, D., Breuer, K., and
Swartz, S. (2011). Whole-body kinematics of a fruit
bat reveal the influence of wing inertia on body accel-
TheInfluenceofBatWingsForProducingEfficientNetBodyForcesinBio-inspiredFlappingRobots
531
erations. Journal of Experimental Biology, vol. 214,
no. 9, pp. 1546-1553.
Lentink, D. and Biewener, A. (2010). Nature inspired
flight-beyond the leap. Bioinsp. Biomim, vol. 5, no.
4, p040201.
Riskin, D., Bergou, A., Breuer, K., and Swartz, S. (2012).
Upstroke wing flexion and the inertial cost of bat
flight. Proceedings of the Royal Society B: Biologi-
cal Sciences, vol. 279, p2945-2950.
Riskin, D., Iriarte-Diaz, J., Middleton, K., Breuer, K., and
Swartz, S. (2010). The effect of body size on the
wing movements of pteropodid bats, with insights into
thrust and lift production. The Journal of Experimen-
tal Biology, vol. 213 (23) pp. 4110-4122.
Swartz, S., Iriarte-Diaz, J., Riskin, D., and Breuer, K.
(2012). A bird? a plane? no, its a bat: an introduc-
tion to the biomechanics of bat flight. Evolutionary
History of Bats Fossils, Molecules and Morphology,
p317-352, Cambridge.
Thollesson, M. and Norberg, U. (1991). Moments of iner-
tia of bat wings and body. Journal of Experimental
Biology, vol. 158, 1991, pp. 19-35.
Winter, Y. and Helversen, O. V. (1998). The energy cost of
flight: do small bats fly more cheaply than birds? J
Comp Physiol B 168: 105-111.
ICINCO2014-11thInternationalConferenceonInformaticsinControl,AutomationandRobotics
532
