Design and Experimental Study of a Pneumatic Bionic Stingray

Undulatory Soft Robot

Songzi Guo, Jinhua Zhang, Yuhan Yang, Haiyan Cheng and Jun Hong

Key Laboratory of Education, Ministry for Modern Design and Rotor-Bearing System, Xi’an Jiaotong University,

Xi’an 710049, China

Keywords: Bio-Mimic Design, Flexible Actuation.

Abstract: Underwater organisms have always been providing inspiration for the design and development of novel

underwater propulsion and bionic robots. At present, stingray has been taken as a bionic object due to its

stable motion and robust mobility. In this study, a stingray propelled by flexible pectoral fins was taken as a

bionic object. Based on this, a new idea for the design of high-performance bionic underwater propulsor was

proposed. An analysis was conducted regarding the design, fabrication and experiments of the bionic stingray

wave propulsion soft robot based on pneumatic drive. As revealed by the experiments of propulsion

performance, the influencing factors for average propulsion include varying frequencies, fin stiffness and the

gaps between substrate and the fins. This is expected to provide guidance on our design of a stingray robot in

respect of efficient mobility.

1 INTRODUCTION

With the deepening of research on fish, researchers

have found out that fish is adaptive to different living

environments. By means of distinctive swimming

gaits, fish is capable of high-speed cruise and agile

maneuverability (Webb et al., 1994). Among various

fishes, the stingray shown in Figure.1 is considered to

be an outstanding swimmer especially when they

swim in close proximity of the substrate (B. Liu and

Z. Guo, 2018). Besides, it shows such obvious

advantages high propulsion efficiency, excellent

steering maneuverability and swimming stability,

which makes it attract increasing attention from many

researchers. According to the observation of a

pectoral during swimming, there are multiple

propulsive waves created by their pectoral fin, which

are termed undulatory-swimmer. At present, there

have been plenty of studies conducted on the motion

mechanism of rays (Wu et al., 1961) as well as the

swimming performance (Rosenberger et al., 1999 and

2001). These studies have provided us with a

significant inspiration in the design of bionic robots

(e.g. undulatory soft robotic (Urai, R. Sawada and N.

Hiasa, 2015; Jusufi et al., 2017; S. N. Toda Y et al.,

2006; Moored et al., 2011) and oscillatory soft robotic

(Moored, K. W. and Dewey, P. A, 2011; Chen et al.,

2012; Chew et al., 2015).

Based on the morphological properties of rays

with independent actuators, a mimic soft batoid robot

was developed in the previous work of Urai and

Sawada (Urai, R. Sawada and N. Hiasa, 2015). Such

specialized flexible structure makes the mimic soft

batoid robot adaptive to a wide range of different

applications and complex environmental conditions.

Besides, it was indicated in their work that the fin

stiffness causes swimming performance to vary

significantly. Accordingly, an appropriate design of

the fin stiffness is required to enhance swimming

performance. Using ionic polymer–metal composite

(IPMC), Chen et al (Chen et al., 2012) developed a

manta ray robot to fabricate artificial muscles for the

simulation of swimming behavior exhibited by a

manta ray. According to their experimental results,

this robot is capable to swim at 0.74 cm/s with a

consumption of less than 2.5 W.

In this study, a finalized prototype of a bio-

inspired stingray robot was developed through the

combination of biomimetic functions and

morphological properties. Furthermore, in order to

enhance its swimming performance, the fin-to-wing

extension stiffness for fins based on a jamming

method was designed. However, this jamming

method shows some difference compared with real

rays, it is still a more effective approach to adjusting

the stiffness of a pectoral fin than to replacing fin rays

on it. Then, an experiment platform was constructed

Guo, S., Zhang, J., Yang, Y., Cheng, H. and Hong, J.

Design and Experimental Study of a Pneumatic Bionic Stingray Undulatory Soft Robot.

DOI: 10.5220/0008939703630368

In Proceedings of the 17th International Conference on Informatics in Control, Automation and Robotics (ICINCO 2020), pages 363-368

ISBN: 978-989-758-442-8

363

to study the propulsion performance of the prototype.

As revealed by the experimental results of propulsion

performance, there is a significant improvement of

average propulsion with the rising frequencies under

1Hz. Besides, it is determined by fin stiffness and

shows an increase of negative pressure in jamming

chamber. Then, an investigation was conducted into

the impact of the bionic stingray robot on the

propulsion performance when swimming took place

at different heights from a substrate. The

experimental results demonstrated that swimming

performance can be enhanced by swimming close to

the substrate at proper heights. The contributions of

this study are threefold. Firstly, the novel pneumatic

stingray robot was developed. Secondly, a jamming

method was proposed for active variable stiffness to

enhance propulsion force. Thirdly and lastly, the

ground effect of the stingray robot on propulsion

performance was verified.

Figure 1: Biological stingrays.

2 FROM BIOLOGICAL FISH TO

ROBOTIC FISH

Figure 2: The skeletal structure of a stingray.

A rajiform swimmer, bluespotted stingray, is taken as

an biomimetic subject, which is because its

swimming model embodies a fine balance reached

between speed and maneuverability, thus facilitating

the design of a next-generation underwater vehicle.

On this basis, a scheme was proposed to develop a

stingray robot by combining both morphological and

kinematic properties of the stingray. In order to

illustrate their excellent swimming performance, a

series of vivisection experiments were conducted on

a bluespotted stingray. The musculoskeletal structure

of the subject is shown in Figure. 2, which reveals that

the radial cartilage skeleton is embedded in a circular

pectoral fin, which is covered with planate muscles.

This sort of radially musculoskeletal structure is

considered favorable for the agile and flexible motion

of the pectoral fin by the successive stimulation of

planate muscles. Moreover, an observation was made

of the interosseous cartilage structure along the

chordwise direction, and such a conjunction structure

between two radial cartilages increases the chordwise

stiffness in the pectoral fin required for the actual

transmission of traveling wave in the chordwise

direction (Fiazza et al., 2010). On the other hand, the

compliant tissue, particularly on the pectoral fin,

allows for bending to a greater extent, which prompts

the shift of our focus to the enhancement of material

properties that make the stingray robot closer to the

object of simulation, for example, flexibility and

malleability.

Based on the biological observations of

anatomical structure, it is deemed necessary to

simulate the existence of a radially musculoskeletal

structure and the compliance of the pectoral fin,

which is achieved through the symmetrical and radial

deployment of twelve soft pneumatic actuators on the

pectoral fin. In addition, it is essential to choose an

appropriate material to approach the simulation

object. Therefore, a variety of hyperelastic silicone

rubber (ELASTOSIL, M4601) was adopted to

fabricate the bionic actuators due to its excellent

flexibility and malleability. The actuator is designed

to combine an actuating chamber with two antagonist

chambers and a jamming chamber as shown in

Figure.3.a, in which the pressures are controlled

separately. Constrained by winding fiber, this

actuating chamber is covered with an external jacket

made of the same material. A jamming chamber

capable to adjust stiffness distribution on the pectoral

fin is attached at the tip of an actuating chamber by

wacker E41 as adhesive. As packed with plastic

granules 2mm in diameter (Figure.3.d), While being

depressurized, the cavity makes uncompacted granule

compressed tightly and presents different stiffness.

This method is effective in enhancing swimming

performance, which will be verified in the following

experiments. The total length of the actuator is

105mm, while the actuating part and jamming part are

70mm and 35mm in length, respectively, as shown in

Figure.3.b. The fabrication process of the actuating

chamber is illustrated in Figure.3 c. With the M4601

ICINCO 2020 - 17th International Conference on Informatics in Control, Automation and Robotics

364

polymer poured into a 3D printed bladder mold and

cured, the preliminary bladder is achieved after the

first-time demolding. Then, it is winded with fiber to

withstand the deformation caused in radial direction,

as a result of which curvature can be generated when

the pressure is increased in one chamber. Afterwards,

the bladder is covered with jacket layer. After the

second-time demolding and attachment of the

jamming chamber at the tip of the fin ray, the

fabrication of the entire actuator is completed.

Figure 3: The design of the actuating fin ray with two

degrees of freedom, (a) an isometric view, (b) a side view,

consisting of actuating chamber and jamming chamber. (c)

The fabrication process of the actuating fin rays involves

bladder molding and jacket molding. (d) The fabrication

process of the the jamming chamber. When depressurized,

it makes uncompacted granules compressed tightly and

exhibits varying stiffness.

When all of the twelve flexible fin rays are

obtained, they are adhered to a disc-like transparent

silicon rubber plate with a 0.5mm thickness.

3 EXPERIMENTS AND RESULTS

3.1 The Establishment of the

Experimental Platform

Figure 4: The developed prototype of the ray-like soft robot.

The bio-inspired stingray robot and its specifications

are shown in Figure.4. The pectoral fin consists of

dozen of the bionic fin rays deployed radially around

the flexible disc-like plate. Besides, it is fixed on a

skeleton made out of photosensitive resin. In order to

provide and regulate the driving pressure, an air tube

(SMC, 2mm) is supposed to be connected to each of

Figure 5: An experiment system to test the swimming

performance.

the antagonist chambers actuated separately by a

pneumatic system, as shown in Figure.5. The

pneumatic system primarily consists of an air

compressor module capable of supplying compressed

air to the actuators and a digital output module

achieved by a PLC module (OMRON, CJ1W-

DA08V, CJ1W-OC21), which is capable to regulate

the states of solenoid valves and proportional valves,

respectively. The pneumatic system is placed above

the water. Air tubes are applied to connect the

pneumatic system to the bionic stingray robot.

Considering that one of the most representative

properties of bio-inspired robots, flexibility, can be

affected when they are combined with rigid electronic

components, a simplified prototype with open loop

control was developed without any type of electrical

components fitted on the bionic stingray robot. This

open loop control is capable of regulating the

undulatory attitudes, frequencies and the stiffness of

the fin rays. More specifically, it can regulate the

compressed air generated by the air compressor

module and make adjustment to the pressure in the

two antagonist chambers when the solenoid valves

and proportional valves are successively exerted

sinusoidal signals generated by the PLC module to

control its states. Then, the bionic stingray robot can

perform distinct undulatory gaits. The deformation

caused to the pectoral fins of the bionic stingray robot

is illustrated in Figure. 6.a, which shows that it is

driven by a pressure regulated by the PLC module.

The maximum pressure in the actuating chambers is

0.08Mpa, and the actuating frequency is 0.5Hz. As

shown in Figure. 6.b, the bionic stingray robot is

snapped while performing an undulatory gait.

Design and Experimental Study of a Pneumatic Bionic Stingray Undulatory Soft Robot

365

Figure 6: The deformation on the pectoral fins of the

stingray robot.

3.2 Effects of Frequencies on

Propulsion Performance

In order to explore the relationship between the

frequency and propulsion, the stingray robot was

placed in a water tank which is 1.5m long, 1m wide

and 0.5m deep, the locomotion of which was

restricted to one degree of freedom for a forward

translation with a steel bar. The swimming behavior

on propulsion was observed by periodically driving

the fin rays in phase at varying frequencies. The

average propulsion was obtained by measuring the

propulsion within one period. In the experiments, a 6-

axis force torque sensor (ATI-Nano17) was employed

to record the forces generated. More specifically, the

undulatory amplitude was significantly down-

regulated to actuate frequencies over 2.5Hz due to the

limitation on the response time of the pneumatic

system. Thus, the maximum driving frequency was

restricted to under 2Hz. It was found out that a decent

undulatory amplitude appeared on the fin under the

driving pressure P =0.05Mpa. Therefore, the

experiments were conducted under the actuating

pressure of 0.05MPa.

Figure 7: The propulsion performance for four different

frequencies f= 0.33, 0.50, 1.00, 2.00 Hz.

The experimental results indicate the relationship

between average propulsion and the actuating

frequencies, as shown in Figure.7. The average

propulsion is affected by the changes in frequency.

On the one thing, the average propulsion is improved

with the rise of frequency and one maximum average

propulsion can reach 1.1N. In the meantime, the

driving frequency and pressure are 2.00Hz and

0.05MPa, respectively. When the actuating frequency

exceeds 1Hz, nevertheless, this upward trend begins

to slow down. On the other hand, a higher level of

frequency requires the consumption of more energy.

Thus, it is necessary for both propulsion and energy

consumption to be considered carefully for meeting

different requirements.

3.3 Effects of Fin Stiffness on

Propulsion Performance

Given that fishes can create a greater propulsion force

to change the stiffness of their fins and have

interaction with the surrounding ﬂuid in a flexible

way (Fiazza et al., 2010), our stingray robot with a

jamming chamber at the tip of the fin ray is optimized

to enhance the propulsion performance while

allowing it to get adaptive to various hydrodynamic

circumstances. The experiments were performed to

better understand how the propulsion force varies for

the fins to achieve different stiffness after being

depressurized to different pressures. The propulsion

forces were measured during the one period T=3s

with jamming chamber for three different pressures

(Pj= 0kPa, -30kPa, -60kPa), and the twelve fin rays

were driven under pressure P =0.05MPa.

Figure 8: Propulsion of the stingray robot with three

different fin stiffness regulated by negative pressure in the

jamming chamber (Pj stands for the negative pressure in

jamming chamber).

The difference of propulsion forces at varying

negative pressure is shown in Figure.8, in which the

propulsion forces of three different negative pressures

are plotted. According to the experimental results, the

propulsion force is maximized when the jamming

chambers are depressurized to -60kPa. By contrast,

the propulsion force is minimized under the pressure

ICINCO 2020 - 17th International Conference on Informatics in Control, Automation and Robotics

366

P=0kPa. Additionally, each side of the fin rays is

divided into three pairs driven by the actuating signal

in different phases. Since the motion of fins is based

on three pairs of fin rays, three different propulsion

peaks can be observed in Figure.9 and the largest

propulsion force is indicated by the middle peak,

which suggests that the middle pair plays a crucial

role in this stipulated pattern of motion. Moreover, it

is worth noting that the maximum undulatory

amplitude of the stingray robot is reduced as negative

pressure rises (0kPa through -60kPa) in jamming

chamber. In comparison, the driving pressure and

frequency remain unchanged. Therefore, the stiffness

of the fins is required to meet the various

requirements.

3.4 The Ground Effect on Swimming

Performance

It is noteworthy that when fishes swim near the

substrate, its generation of propulsion is subject to a

significant impact from the distance between the fins

and the substrate (B. Liu and Z. Guo, 2018). In the

experiments conducted in this study, the propulsion

performance of the fins was investigated with three

different gaps (d=10cm, 20cm, 30cm) between the

fins and ground. Moreover, each of the experiment

was conducted under the same driving pressure

P=0.05MPa and frequency f=0.5Hz.

Figure 9: Propulsion of the stingray robot with different

distances between the substrate and the bionic pectoral fin.

The result is shown in Figure.9, which reveals that

the maximum value can reach 1.6 N when the gap is

10cm. In this case, its trajectory of propulsion force

shows a slight difference with gap d= 20cm and gap

d=30cm, the instantaneous maximum of which can

reach 1.28 N and 0.92 N, respectively. In addition,

near-ground swimming can generate three different

propulsion peaks in one period, which is associated

with the motion of fins. The propulsion peaks will be

observed when the fin rays close to the ground,

suggesting that the stingray robot can achieve a

superior propulsion performance when it swims at a

proper height (10cm), which is conducive to

enhancing the efficiency of underwater propulsor.

4 CONCLUSION

In this study, a prototype of stingray robot based on

the combination of biological functions and

morphological properties was proposed to make the

stingray robot capable of achieving ray-like flexible

locomotion through a pneumatic system. The

biological functions including rapid responding and

agile motion are estimated by propulsion forces and

the average propulsion can reach 1.1N at maximum.

In comparison, the driving frequency and pressure are

2.00Hz and 0.05MPa, respectively. Moreover, the

propulsion performance of the stingray robot can be

enhanced by adjusting the stiffness of their fins and

having interaction with the surrounding ﬂuid in a

flexible way as achieved by jamming method and

flexible biomimetic material. The inherent

advantages of the pneumatic actuator, especially in

weight and deformation, provide a new idea for the

design of underwater vehicles fit for complex

working environments. Moreover, the constant

change made to the stiffness on bionic fin rays using

the depressurized jamming method allows bionic

underwater vehicles to adjust stiffness distribution.

According to the results of ground effect on

swimming performance, a proper operational depth is

also beneficial to improve the propulsion

performance of ray-like underwater vehicles.

In the future, a further study in this regard should

focus on two aspects as follows. Firstly, judging on a

serial of frames recorded by the high-speed camera,

the undulatory amplitude of the pectoral fin is

partially limited by the disc-like base that radially

carries dozen of the fin rays. In addition, this

constraint placed on the fin rays results in a slightly

lateral distortion and vibration when these fin rays are

actuated over a certain pressure to perform undulatory

gaits. It is considered as a disadvantage, especially

when loop-controls are introduced into the robotic

system. Therefore, one of the improvement that can

be made in the further research is to chose a more

resilient material to minimize this potential

interference. Secondly, the introduction of loop-

controls to the robotic system is effective in

improving the swimming performance of the bionic

stingray robot. As a support for gathering the

information about motion, a control method based on

Design and Experimental Study of a Pneumatic Bionic Stingray Undulatory Soft Robot

367

nonlinear error feedback controller will be introduced

and a flexible sensor will be fitted on our next-

generation prototype.

ACKNOWLEDGEMENTS

This research was financially supported by the

National Nature Science Foundation of China (No.

91748123).

REFERENCES

Webb, P. (1994). The biology of fish swimming. In L.

Maddock, Q. Bone, & J. Rayner (Eds.), The Mechanics

and Physiology of Animal Swimming (pp. 45-62).

Cambridge: Cambridge University Press.

Liu and Z. Guo, (2018). Ground effect on the hydrodynamic

performance of a flexible hinge-connected fin.

IEEE/ASME Int. Conf. Adv. Intell. Mechatronics,

AIM, vol. 2018-July, pp. 881–886, 2018.

Wu, T. . (1961). Swimming of waving plate. journal of fluid

mechanics, 100.

Rosenberger, L. J., & Westneat, M. W. (1999). Functional

morphology of undulatory pectoral fin locomotion in

the stingray taeniura lymma (chondrichthyes:

dasyatidae).

Rosenberger, L. J., (2001). Pectoral fin locomotion in

batoid fishes: undulation versus oscillation. Journal of

Experimental Biology, 204(Pt 2), 379-394.K.

Urai, R. Sawada, N. Hiasa, M. Yokota, and F. DallaLibera.

.(2015). Design and control of a ray-mimicking soft

robot based on morphological features for adaptive

deformation. Artificial Life & Robotics, 20(3), 237-

243.

Jusufi, A., Vogt, D. M., Wood, R. J., & Lauder, G. V..

(2017). Undulatory swimming performance and body

stiffness modulation in a soft robotic fish-inspired

physical model. soft robot, 202-210.

S. N. Toda Y, Sanada Y, and Danno M. .(2006). The motion

of a fish-like under-water vehicle with two undulating

side fins. Proceedings of the 3rd International

Symposium on Aero Aqua Bio-mechanisms.

Moored, K. W. , Fish, F. E. , & Kemp, T. H. . (2011). Batoid

fishes: inspiration for the next generation of underwater

robots. Marine Technology Society Journal, 45(4).

Moored, K. W. , Dewey, P. A. , Leftwich, M. C. , Bart-

Smith, H. , & Smits, A. J. . (2011). Bioinspired

propulsion mechanisms based on manta ray

locomotion. Marine Technology Society journal, 45(4),

p.110-118.

Chen, Z. , & Bart-Smith, H. . (2012). Bio-inspired robotic

manta ray powered by ionic polymer-metal composite

artificial muscles. International Journal of Smart &

Nano Materials, 3(4), p.296-308.

Chew, C. M. , Lim, Q. Y. , & Yeo, K. S. . (2015).

Development of propulsion mechanism for Robot

Manta Ray. IEEE International Conference on Robotics

& Biomimetics. IEEE.

Fiazza, C. ,Salumae, T. , Listak, M. , Kulikovskis, G. , &

Kruusmaa, M. (2010). Biomimetic mechanical design

for soft-bodied underwater vehicles. Oceans IEEE-

SYDNEY. IEEE.

ICINCO 2020 - 17th International Conference on Informatics in Control, Automation and Robotics

368