A Miniaturized Tadpole Robot Using an Electromagnetic Oscillatory Actuator
Bu Hyun Shin1, Kyung-Min Lee2, Seung-Yop Lee3
1. Department of Mechanical Engineering, Hanbat National University, Daejeon 305-719, Korea 2. Convergence Technology Examination Division, Korean Intellectual Property Office, Daejeon 302-701, Korea
3. Department of Mechanical Engineering, Sogang University, Seoul 121-742, Korea
Abstract In this paper, we propose a miniaturized tadpole-like robot using an electromagnetic oscillatory actuator. The electro-
magnetic actuator has a simple structure with a moving-magnet type and the body size is 13 mm (length) × 11 mm (height) × 10 mm (width). A tail has the thickness of 100 μm and the length of 20 mm which is twice of the body-length (BL). The tail attached to the oscillatory actuator generates undulatory propulsion for the forward swimming. Moreover, the tadpole robot enables the change of the direction by controlling input signal patterns applied to the oscillatory actuator. Prototypes of the tadpole robot have been manufactured and the thrust force and swimming speed are measured to evaluate the performance of the biomimetic robot in water at various tail-beat frequencies. The maximum thrust force is 42 mN at the tail-beat frequency of 30 Hz with voltage of 3 V, enabling the tadpole robot to swim at the speed of 210 mm·s−1 (6 BL·s−1). The tadpole robot can also change its moving direction with the angular velocity of 21 deg·s−1 at the half pulse pattern of 30 Hz.
Recently, biomimetic underwater microrobots are of great interest for industrial, medical and military ap-plications[1,2]. It is known that a fish tail-like propeller has advantages to underwater microrobots because it makes efficient propulsion, small steering radius and less noise. Biomimetic underwater microrobots mim-icking a tadpole have great advantages in exploring complex, narrow underwater environments, and they even can be used in blood vessels for microsurgery.
Tadpoles are exceptional among vertebrates since they have a globose body with a laterally compressed tail[3]. Compared with most fishes, tadpoles swim uniquely by the waves of relatively high amplitude at both snout and tail tip. It is known that the relative am-plitude of body undulations in tadpoles is significantly larger than those observed in fish. Tail beats of lateral oscillations with large amplitude in tadpoles provide high propulsion[2,3].
Many researchers have developed swimming robots
by adopting various types of actuators. Various swim-ming robots using piezoelectric actuators have been de-veloped[1,3,4]. Piezoelectric actuators have strengths such as fast response, size reduction and high output power for fish robots, but it requires high input voltage and addi-tional structures to amplify small deformations. Shape Memory Alloy (SMA) actuators have the advantages of simple structure and reasonable power, however low speed and slow response are its drawbacks. Some re-searchers have developed swimming robots using SMA actuators[6,7]. Lately, soft polymer actuators have been widely researched and Ionic Polymer Metal Composite (IPMC) is the typical polymer actuator for tadpole ro-bots[2] and fish robots[8–11]. IPMC has the advantages of large bending displacement, low input voltage and mo-bility in underwater, but it has the disadvantages of weak output power and reduced displacement in air. Electro-static actuators are also used in MicroElectroMechanical Systems (MEMS) because it is possible to reduce the size of actuators using microfabrication. A few researchers have applied the electrostatic actuator to fish robots[12].
Journal of Bionic Engineering (2015) Vol.12 No.1 30
Electromagnetic actuators have a lot of advantages such as fast response, simple control law and low cost. A ro-tary electromagnetic motor generally is not suitable for micro swimming robots because a fish robot using the rotary motor is relatively bulky in size[13]. In order to reduce the size of the fish robot, the flagella-like propul-sion methods were proposed to actuate the micro swimming robot[14–17]. The commercial miniaturized fish toys produced by Hexbug and Zuru companies are on the market[18,19]. They have good swimming and turning performances using electromagnetic actuators. The fish toy produced by Zuru company has a moving-coil type electromagnetic actuator and that produced by Hexbug company has a moving-magnet type electromagnetic actuator with two coils.
In this paper we propose a miniaturized tadpole robot using an electromagnetic oscillatory actuator with a compact moving-magnet type. The electromagnetic actuator has a simple structure composed of a magnet, a yoke and a coil. A rectangular tail attached to the oscil-latory actuator generates the change of the direction as well as fast swimming in water. We measure the thrust forces and swimming speeds at various tail-beat fre-quencies using prototypes and investigate the optimal swimming conditions.
2 Development of the tadpole robot
A tadpoles consists of a head-body and a tail. Tad-poles vary greatly in size during their development and between species. Fig. 1a shows the tadpole of spring peeper (Pseudacris crucifer) and it usually has the total length of 3.4 cm[20]. The tadpole-like robot in Fig. 1b mimics the tail-induced propulsion of tadpoles by an electromagnetic actuator. An artificial tail is attached to the flapper of the oscillatory actuator. The oscillatory actuator swings the flapper and the tadpole robot swims by the undulatory motion of the tail.
Design and prototype of a compact Electromag-netic Oscillatory Actuator (EOA) are shown in Fig. 2[21]. The EOA has a simple structure consisting of a coil, a magnet and a yoke. The coil surrounds the steel core. A flapper is attached to the moving-magnet. The overall size of EOA is 13 mm in length, 11 mm in height, 10 mm in width and the total mass is 1.5 g. The magnet
Flapper
Oscillatory actuator
Tail
Head-bodyTail
(a)
(b)
1 cm
Fig. 1 (a) Tadpole of spring peeper (Pseudacris crucifer)[20]. It consists of head-body and tail with the 1:2 ratio in length, (b) prototype of the proposed tadpole-like robot with the total length of 3.3 cm.
Fig. 2 Design and prototype of EOA for tadpole robot: (a) De-scription of the electromagnetic field, current and Lorentz force to generate resultant torque of the flapper; (b) the prototype of EOA.
Shin et al.: A Miniaturized Tadpole Robot Using an Electromagnetic Oscillatory Actuator 31
material is ND35, which is a rare earth magnet. The coil has 820 turns and the resister is 58 Ω. The frame and flapper are made of ABS plastics and the flapper is connected to the frame by a steel pin.
If the directions of current and magnetic flux are perpendicular to each other as shown in Fig. 2a, the direction of the Lorentz force becomes perpendicular to both the current and magnetic flux. This electromagnetic force generates the torque in order for the flapper to rotate counterclockwise. When the current flowing in the coil is in the opposite direction, the torque actuates the flapper clockwise. Therefore, the periodic change of the direction of the current applied to the coil causes the oscillatory motion to the flapper. 2.1.2 Dynamic modeling of EOA
There are two types of forces applied to the oscil-latory actuator. One is the magnetic force between the magnet and the steel core, and it holds the flapper on the central position. The magnetic force is called as rein-stating force. The other is the electromagnetic force induced by the current flowing in the coil. The angular direction of the flapper is controlled by the direction of the current of the coil. The dynamic model is similar to that of DC motor because the oscillatory actuator system is theoretically based on DC motor. The equation of the dynamic model is written by
magnetic elecromagnetic ,J F Fθ + = (1)
where J is the mass moment of inertia and θ is the an-gular acceleration. Actually the magnetic force (Fmagnetic) and electromagnetic force (Felectromagnetic) are nonlinear. If the proposed system is assumed to be linear, the two forces can be written as
magnetic ,mF K θ= (2)
elecromagnetic ,tF K I= (3)
where θ is the rotation angle and Km is the reinstating constant which is the same as the spring constant. Kt is the torque constant and I is the current flow in the coil. If the component of inductor in the electric circuit is ig-nored, the circuit equation is written by
,eU R I K θ= + (4)
where U is the voltage, R is the resistor of coil, θ is
angular velocity and Ke is the back-emf constant. Gen-erally the back-emf constant is the same as the torque constant in the case of DC motor. By combining Eqs. (1) and (4), we obtain the second-order linear system. Using Laplace transform, the ratio of the output angle to the input voltage becomes
2
1 .m
tt t
K RJRU s K sK K
Θ=
+ + (5)
The EOA has the magnetic force to hold the flapper between the yoke and the magnet without current. The electromagnetic force should be larger than the magnetic force to generate the oscillatory motion of the tadpole robot. 2.2 Design of the tadpole robot
Tadpoles are young amphibians that usually live in the water. They are exceptional among vertebrates since they have a globose body with a large flattened tail. Tadpoles swim uniquely by lateral undulation of the tail. The undulatory motion of the tail is more efficient to increase the speed of the tadpole-like locomotion[2,3]. The thickness of tail is also one of important design parameters because it directly affects the undulatory motion of the tadpole robot. In general, short tails gen-erate the oscillatory motion rather than the undulatory motion. However, tadpole robots with long tails require more power to swim. A reduction in the surface area of the tail causes a decline in swimming performance. However, it is also known that the tadpoles that exhib-ited predator-induced morphology and naturally possess large tails relative to their body size do not swim better than the tadpoles with smaller tail[22].
As shown in Fig. 1b, the tadpole-like robot consists of two parts. The oscillatory actuator is like the head-body of real tadpoles and a tail is attached to the flapper of the oscillatory actuator. The oscillatory ac-tuator swings the flapper and generates the undulatory motion of the tail. The undulatory propulsion by the tail moves the tadpole robot forward. The tail is made of PET film and has the length of 20 mm and the thickness of 100 μm. The total length and height of the tadpole robot are 33 mm and 11 mm.
The length of tail is chosen as 20 mm that is about twice of the body of the tadpole robot, and the ratio between the body and tail is similar to that of a real
Journal of Bionic Engineering (2015) Vol.12 No.1 32
tadpole. The shape of the tail is simply rectangular be-cause the tail is similar to that of a real tadpole and it is easy of production. Based on previous researches by the authors using the EOA with the tail of trapezoidal shape like fish[20], we use the thickness of the tail with 100 μm. The tadpole robot with the thickness of 100 μm is faster than that of 200 μm because the undulation motion eas-ily occurs with the thinner tail. In this study, we focus on the optimal value of the driving frequency in order to increase the thrust force and swimming speed of the tadpole robot with the rectangular tail.
The driving circuit is simply configured using a power op-amp L272 as shown in Fig. 3. In order to switch the direction of the current, two op-amps are used for the bipolar drive because it is easier to get a single power in general. The output of the driving circuit gen-erates square waves. Figs. 3b and 3c show the periodic pulse patterns of input signal for the forward and turning motions of the tadpole robot. In order to change the direction of the tadpole robot, the input signal of the one op-amp is ground while the other has a periodic square wave.
(a)
(b)
(c)
L272 L272
3
−3
3
3 V
R
R
0
Time (s)
Time (s)
Time (s)
Vol
tage
(V)
Vol
tage
(V)
Fig. 3 (a) Driving circuit using power op-amp; (b) input signal for the forward motion; (c) input signal for the turning motion.
3 Experimental results
3.1 Evaluation of the EOA with tail The experimental setup is shown in Fig. 4. The
force is measured using load cell (CAS PW4M-300g). The displacement of the tail is measured using a laser displacement sensor (Keyence IL-065), and the meas-ured displacements are used to calculate the rotating angle and the dynamic response.
To evaluate the performance of the EOA with tail, the physical parameters of Eq. (5) are experimentally measured. In order to obtain the restoring and torque constants, we measure the torque of the flapper as a function of the rotational angle and input current. Here, the value of torque is the measured force multiplied by the distance from the axis to the measured point of the flapper. Firstly, we measure torque induced by the EOA for different input currents when the flapper locates in the central position, as shown in Fig. 5a. The torque constant is also calculated from the electromagnetic forces meas-ured using the load cell for different input currents, while the flapper is hold without moving. The experimental torque constant of the EOA is Kt = 5.46 mN·m·A−1, which is the slope of the linearized current-torque curve.
Magnetic forces are measured using the load cell for the cases of various rotational angles with zero cur-rent at each coil. The magnetic force at each angular position of the flapper is measured to calculate the
Laser SensorKeyence IL-065
Tadpole robot
Load cellPW4M-300g
Function Generator
Oscilloscope Fig. 4 Experimental setup of the EOA with tail.
Shin et al.: A Miniaturized Tadpole Robot Using an Electromagnetic Oscillatory Actuator 33
(a)Current (A)
0.20
0.16
0.12
0.08
0.04
0.000.00 0.01 0.02 0.03 0.04
y = 5.4627x
(b)
Rotating angle (rad)
0.16
0.12
0.08
0.04
0.000.00 0.01 0.02 0.03 0.04
y = 0.418x
Fig. 5 (a) The torques measured experimentally for various input currents. The slope is the torque constant. (b) The measured torques depending on rotational angles. The slope is the reinstat-ing constant. reinstating constant. In Fig. 5b, the reinstating constant of the EOA is Km = 0.42 mN·m·rad−1, which is the slope of the linearized angle-torque curve. The experimental values of the parameters are summarized in Table 1.
To find the static response, the rotational angle is measured for various input voltages and the experi-mental results are shown in Fig. 6. The rotational angle increases with the input voltage in the early stage, but it is saturated over the region of 2.5 V. The tail is rotated by 24 deg at the input voltage of 3V. We have also per-formed experiments to measure the dynamic response of the actuator system using sinusoidal inputs. Using a function generator, we make the sinusoidal input from 1 Hz to 100 Hz. The displacement signals of the tail measured by the laser displacement sensor are recorded in the oscilloscope. After Fast Fourier Transform (FFT) analysis, the experimental frequency response of the EOA with tail is plotted in Fig. 7, and it is also compared with the theoretical frequency response by Eq. (5). The oscillatory actuator with tail has the resonance frequency at 23 Hz when it moves in air.
Table 1 Physical parameters measured in experiments
Parameters Values
Moment of Inertia (J) 23 × 10−9 kg·m−2
Resistor (R) 58 Ω
Torque constant (Kt) 5.46 mN·m·A−1
Reinstating constant (Km) 0.42 mN·m·rad−1
Rot
atin
g an
gle
(deg
.)
Fig. 6 Rotating angles measured experimentally for various input voltages.
Fig. 7 Frequency response of the EOA with tail in air.
3.2 Experiment of the tadpole robot
We carried out experiments in water to measure the thrust force and the speed of the tadpole robot. The input voltage is set to 3 V commonly used for inte-grated circuits. The experimental setup is similar to the one used in Fig. 4. The power op-amp transfers the input signals of various tail-beat frequencies by the function generator to the tadpole robot. Two wires with the length of 500 mm and the diameter of 200 μm from the power op-amp hang the tadpole robot as shown in Fig. 8a. A high-speed camera is used to measure the swimming speed of the tadpole robot (Fig. 8b). The frame rate of the captured video is 240 fps at low resolution and 30 fps at high resolution. The camera takes the video of the robot from the top of the water
Journal of Bionic Engineering (2015) Vol.12 No.1 34
(a) (b)
(c)
Load cell
Fig. 8 (a) Experimental setup in water; (b) the measurement of swimming speed; (c) the measurement of thrust force using a load cell.
tank. Power is supplied to the tadpole robot through enamel-coated copper coils with the diameter of 20 μm. The thrust force by the tadpole actuator is measured using a load cell (CAS PW4M 300g) as shown in Fig. 8c.
We also conducted experiments in order to deter-mine the maximum swimming speed and thrust force of the tadpole robot. The measurement results of the thrust force for various tail-beat frequencies are shown in Fig. 9a. The maximum thrust force is 42 mN at 30 Hz and 3 V. We plot the swimming speed curve of the tadpole robot in Fig. 9b. The maximum speed is 210 mm·s−1, which is corresponding to 6.36 BL·s−1. The maximum swimming speed occurs at the tail-beat frequency of
50
40
30
20
10
0
Frequency (Hz)
250
200
150
0.0 10 20 30 40 50 60
100
50
0
(a)
(b)
Fig. 9 Experimental results for various input frequencies at 3 V. (a) Thrust force; (b) swimming speed.
30 Hz, where the thrust force is also maximized. As shown in Fig. 7, the maximum displacement in the fre-quency response of the tadpole robot occurs at 23 Hz in air. However, the resonance occurs around 30 Hz when it moves in water. Since the undulatory motion of the tail is maximized at the resonance frequency, both the thrust force and swimming speed are also maximized. It is noted that the optimal tail-beat frequency to maximize the swimming speed is experimentally determined.
The sequential images of the tadpole robot during one stroke are shown in Fig. 10. It is noted that the tail has the undulatory motion over one cycle. Kim et al.[2] developed a tadpole-like robot using IPMC where the lengths of the head-body and tail are 42 mm and 54 mm, respectively. The maximum swimming speed of the IPMC based tadpole robot was 23.6 mm·s−1 (0.24 BL·s−1) at the tail-beat frequency of 4 Hz. It is remarkable that the proposed tadpole robot using EOA swims over 6 BL·s−1.
In order to change the direction of the tadpole robot, the asymmetric square wave as shown in Fig. 3c, is ap-plied to the oscillatory actuator to cause the half oscil-latory motion of the flapper and to generate the thrust
Fig. 10 Sequential images of the captured video during one stroke at 30 Hz and 3 V. (a) t = 0 s; (b) t = 0.0042 s; (c) t = 0.0083 s; (d) t = 0.0125 s; (e) t = 0.0167 s; (f) t = 0.0208 s; (g) t = 0.025 s; (h) t = 0.0292 s.
Shin et al.: A Miniaturized Tadpole Robot Using an Electromagnetic Oscillatory Actuator 35
Fig. 11 Right turn motion of the tadpole robot at 30 Hz and 3 V: (a) t = 0 s; (b) t = 0.167 s (3.5 deg); (c) t = 0.333 s (7 deg); (d) t = 0.5 s (10.5 deg); (e) sequential images of the right turn motion. force in the right-hand direction. The actuator has the magnetic force to hold the flapper between the yoke and the magnet without current. The asymmetric thrust force is larger than the magnetic force, and the difference of these two forces change the direction of the tadpole robot as shown in Fig. 11. The tadpole robot enables the turning motion with the angular speed of 21 deg·s−1 at 30 Hz and 3 V. The undulatory motion of the tail at-tached to the oscillatory actuator supplies highly effi-cient propulsion, enabling fast swimming as well as easy turning. The biomimetic tadpole robot has the capability of steering and it can be applied to the underwater tad-pole robots.
4 Conclusion
We have developed a miniaturized tadpole-like robot with both high swimming speed and easy turning. An EOA is designed to have a simple structure com-posed of a magnet, a yoke and a coil. The proposed biomimetic tadpole robot consists of the actuator body and a tail which is attached to the flapper of the oscil-latory actuator. The maximum thrust force and swim-ming speed are over 40 mN and 200 mm·s−1 at the os-cillation of 30 Hz. It is noted that the maximum speed of the proposed tadpole robot is over 6 BL·s−1, which is the
remarkable advantage compared to previous researches by others. The direction of the tadpole robot is also eas-ily changed by asymmetric thrust forces applied to the tail. The tadpole robot shows the right or left turn with the angle of 15 deg at the half pulse pattern of 30 Hz and 3 V. Experimental results show that the undulatory mo-tion of the tail generates fast swimming as well as easy turning. The advantages of the proposed tadpole robot are cost-effectiveness for mass production and high-efficiency propulsion in the water.
Acknowledgment
This research was supported by both National Nu-clear R&D Program (2012M2A8A4055325) and Basic Science Research Program (2013R1A2A2A01068159) through the National Research Foundation of Korea funded by Ministry of Science ICT & Future Planning.
References
[1] Kósa G, Shoham M, Zaaroor M. Propulsion method for swimming microrobots. IEEE Transactions on Robotics, 2007, 23, 137–150.
[2] Kim B, Kim D, Jung J, Park J. A biomimetic undulatory tadpole robot using ionic polymer-metal composite actuators. Smart Material and Structure, 2005, 14, 1–7.
[3] Liu H, Wassersug R, Kawachi K. The three-dimensional hydrodynamics of tadpole locomotion. The Journal of Ex-perimental Biology, 1997, 200, 2807–2819.
[4] Nguyen Q S, Park H C, Byun D. Thrust analysis of a fish robot actuated by piezoceramic composite actuators. Jour-nal of Bionic Engineering, 2011, 8, 158–164.
[5] Heo S, Wiguna T, Park H C, Seo N S. Effect of an artificial caudal fin on the performance of a biomimetic fish robot propelled by piezoelectric actuators. Journal of Bionic En-gineering, 2007, 4, 151–158.
[6] Shinjo N, Swain G W. Use of a shape memory alloy for the design of an oscillatory propulsion system. IEEE Journal of Oceanic Engineering, 2004, 29, 750–755.
[7] Wang Z, Hang G, Li J, Wang Y, Xiao K. A micro-robot fish with embedded SMA wire actuated flexible biomimetic fin, Sensors and Actuators A, 2008, 144, 354–360.
[8] Paquette, J W, Kim, K J. Ionomeric electroactive polymer artificial muscle for naval applications. IEEE Journal of Oceanic Engineering, 2004, 29, 729–737.
[9] Guo S, Fukuda T, Asaka K. A new type of fish-like under-water microrobot. IEEE/ASME Transactions of Mechatron-ics, 2003, 8, 136–141.
[10] Ye X, Su Y, Guo S. A centimeter-scale autonomous robotic
Journal of Bionic Engineering (2015) Vol.12 No.1 36
fish actuated by IPMC actuator. Proceedings of IEEE In-ternational Conference of Robotics and Biomimetics, Sanya, China, 2007, 262–267.
[11] Chen Z, Shatara S, Tan X. Modeling of biomimetic robotic fish propelled by an lonic polymer-metal Composite caudal fin. IEEE/ASME Transactions of Mechatronics, 2010, 15, 448–459.
[12] Zhang Z G, Yamashita N, Gondo M, Yamamoto A, Higuchi T. Electrostatically actuated robotic fish: Design and control for high-mobility open-loop swimming. IEEE Transactions of Robotics, 2008, 24, 118–129.
[13] Yu J, Wang S, Tan M. A simplified propulsive model of bio-mimetic robot fish. Robotica, 2005, 23, 101–107.
[14] Behkam B, Sitti M. Modeling and testing of a biomimetic flagellar propulsion method for microscale biomedical swimming robots. Proceedings of the IEEE/ASME Interna-tional Conference on Advanced Intelligent Mechatronics, Monterey, California, USA, 2005, 37–42.
[15] Behkam B, Sitti M. Design methodology for biomimetic propulsion of miniature swimming robots. ASME Journal of Dynamic Systems, Measurement and Control, 2006, 128,
36–43. [16] Chen B, Liu Y D, Chen S, Jiang S R, Wu H T. A biomimetic
spermatozoa propulsion method for interventional micro robot. Journal of Bionic Engineering, 2008, 5, 106–112.
[17] Ha N S, Goo N S. Propulsion modeling and analysis of a biomimetic swimmer. Journal of Bionic Engineering, 2010, 7, 259–266.
[19] ZURU company, Robofish. http://www.zururobofish.com/ [20] A Field Guide to Amphibian Larvae and Eggs of Minnesota,
U.S. Geological Survey, Wisconsin, and Iowa. http://www.umesc.usgs.gov/terrestrial/amphibians/mknutson_5003869_field_guide.html
[21] Shin B H, Lee S Y. Micro mobile robots using electromag-netic oscillatory actuator. IEEE RAS & EMBS International Conference on Biomedical Robotics and Biomechatronics, Rome, Italy, 2012, 575–580.
[22] Van Buskirk J, McCollum S A. Influence of tail shape on tadpole swimming performance. Journal of Experimental Biology, 2000, 203, 2149–2158.
