design, fabrication and thrust drag analysis of improved fish robots actuated by piezoceramic composite actuators

20 Figure 1-23: Fish robot actuated by two LIPCA: a Configuration of the fish robot; b Linkage system.... 74 Figure 4-3: a Configuration of the CLIPCA fish robot, b The CLIPCA fish robot

Design, Fabrication and Thrust/Drag Analysis of Improved Fish Robots Actuated

by Piezoceramic Composite Actuators

Nguyen, Quang Sang

Thesis of the Doctor of Philosophy Department of Advanced Technology Fusion

Graduate School of Konkuk University

Design, Fabrication and Thrust/Drag Analysis of Improved Fish Robots Actuated

by Piezoceramic Composite Actuators

Nguyen, Quang Sang

Thesis of the Doctor of Philosophy Department of Advanced Technology Fusion

Graduate School of Konkuk University

Design, Fabrication and Thrust/Drag Analysis of Improved Fish Robots Actuated

by Piezoceramic Composite Actuators

Nguyen, Quang Sang

A Dissertation Submitted to the Department of Advanced Technology Fusion

and the Graduate School of Konkuk University

in partial fulfillment of the requirements

for the degree of Doctor of Philosophy

Table of Content

List of Figures v

List of Tables ix

ABSTRACT x

1 Introduction 1

1.1 Understanding fish swimming 1

1.1.1 Reynolds number 5

1.1.2 Froude number: 6

1.1.3 Strouhal number 6

1.2 Overview of research on fish 7

1.3 Overview of research on fish robot driven by electromagnetic motor 8

1.4 Overview of fish robot driven by smart material 14

1.5 Objective and approach 21

2 A fish robot driven by piezoceramic actuators 24

2.1 Description of the fish robot 24

2.1.1 Actuator 25

2.1.2 Linkage design 26

2.1.3 External power supply 27

2.1.4 Miniaturized power supply 31

2.2 Swimming test 33

2.2.1 Linkage analysis 33

2.2.2 Thrust measurement 36

2.2.3 Wired-swimming test 38

2.2.4 Free-swimming test 38

2.3 Experimental result 39

2.3.1 Tail-beat angle 39

2.3.2 Thrust 40

2.3.3 Swimming speed 41

2.4 Summary 43

3 An improved fish robot driven by piezoceramic actuators 44

3.1 Design and fabrication of the actuation system 44

3.1.1 Actuator 44

3.1.2 Design and working principle of the linkage system 46

3.1.3 Linkage analysis 48

3.1.4 Fabrication of the actuation system 51

3.2 Fish robot configuration 52

3.2.1 Tail fin of the fish robot 52

3.2.2 Description of the improved fish robot 53

3.3 Evaluation of the improved fish robot 53

3.3.1 Tail-beat angle of the fish robot in water 54

3.3.2 Swimming test of the fish robot 55

3.3.3 Thrust of the fish robot 58

3.3.4 Thrust coefficient 62

3.4 Parameter study 68

3.4.1 Reynolds number 68

3.4.2 Froude number 68

3.4.3 Strouhal number 70

3.5 Summary 70

4 Thrust improvement by using Compressed LIPCAs 71

4.1 Fish robot 71

4.1.1 Actuator 71

4.1.2 Actuation mechanism 72

4.1.3 Fish body 72

4.2 Experiment 73

4.2.1 Swimming experiment 73

4.2.2 Thrust Measurement 77

4.2.3 Drag Estimation 77

4.3 Result 79

4.3.1 Swimming speed 79

4.3.2 Thrust of the Fish Robot 80

4.3.3 Drag of the Fish Robot 80

4.3.4 Drag coefficient 85

4.4 Summary 86

5 CFD simulation 87

5.1 Simulation model 87

5.2 Model validation 90

5.3 Results and discussions 92

6 Concluding remarks and recommendations future work 100

6.1 Concluding remarks 100

6.2 Academic contribution 102

6.3 Recommendation for future work 103

REFERENCE 104

APPENDIX 107

요약문 요약문 108

List of Figures

Figure 1-1: Fish configuration [2] 3

Figure 1-2: Swimming mode associated with (a) BCF propulsion and (b) MPF propulsion [3] 3

Figure 1-3: Diagram showing the relation between swimming propulsors and swimming functions [4] 4

Figure 1-4: (a) The forces acting on a swimming fish; (b) Pitch, yaw, and roll definitions [5] 4

Figure 1-5: Typical velocity of the largest possible variety of swimmers as a function of the Reynolds number [1] 6

Figure 1-6: Flow visualization of velocity field in the x-y planes [13] 10

Figure 1-7: Viscous flow around a swimming fish: (a) mesh around the fish model; (b) detail of a cut through the mesh [14] 10

Figure 1-8: Schematic view of the eight internal links of the MIT’s Robot Tuna mechanism [18] 11

Figure 1-9: Draper Lab’s hydraulic-actuated Vorticity Control Unmanned Undersea Vehicle’s configuration [19] 11

Figure 1-10: RoboPike [20] 12

Figure 1-11: Design of Boxybot fish robot [21] 12

Figure 1-12: Prototype of the experimental robot fish [22] 13

Figure 1-13: Apparatus of pectoral fin motion [23] 13

Figure 1-14: Pressure sensor and its location on the fish body [24] 15

Figure 1-15: Drag measurement apparatus [25] 15

Figure 1-16: Geometry and position of neutral axis of LIPCA [28] 17

Figure 1-17: An IPMC strip under a low voltage [29] 17

Figure 1-18: Large actuation displacement and actuation force of SMA with low responsiveness [31] 18

Figure 1-19: Lamprey robot actuated by SMA [32] 18

Figure 1-20: Two type robots driven by IPMC [33] 19

Figure 1-21: The tadpole robot: (a) configuration of the microrobot; (b) miniaturized of the battery, electrode and embedded controller located inside the body [34] 19

Figure 1-22: The floating boat driven by THUNDER [35] 20

Figure 1-23: Fish robot actuated by two LIPCA: (a) Configuration of the fish robot; (b) Linkage system 20

Figure 2-1: Assembly of the fish 25

Figure 2-2: Geometry and position of layers in a LIPCA 25

Figure 2-3: Linkage system of the fish robot 27

Figure 2-4: Function generator (Agilent 33220A) 29

Figure 2-5: Voltage amplifier (MATSUSADA model AML-1.5B40-LC) 29

Figure 2-6: Oscilloscope (Tektroniks TDS 2024) 29

Figure 2-7: The schematic of the MIPAD 30

Figure 2-8: Hardware implementation of the MIPAD 32

Figure 2-9: MIPAD response to square wave command 32

Figure 2-10: Tail-beat angle analysis by vector calculus 34

Figure 2-11: Tail-beat angle measurement apparatus 35

Figure 2-12: Load cell (Nano 17 Transducer) 36

Figure 2-13: The apparatus of the thrust measurement 37

Figure 2-14: Schematic diagram of the wired-swimming test 37

Figure 2-15: Fish robot in the free-swimming test 38

Figure 2-16: Tail-beat angle by vector calculus 40

Figure 2-17: Tail-beat angle of the fish robot in water 40

Figure 2-18: The average thrust of the fish robot 42

Figure 3-1: Fabricated LIPCA 45

Figure 3-2: Design of the actuation mechanism 45

Figure 3-3: Working principle of the actuation mechanism 47

Figure 3-4: Configuration of linkage system 49

Figure 3-5: The tail-beat angle by vector calculus 50

Figure 3-6: Fabricated actuation mechanism 50

Figure 3-7: The limited tail-beat angle 51

Figure 3-8: Comparison of the linkage systems 52

Figure 3-9: The four scaled tail fins 54

Figure 3-10: The fish robot 54

Figure 3-11: Schematic diagram of the tail-beat angle measurement 56

Figure 3-12: The tail-beat angle in water 56

Figure 3-13: Schematic diagram of the swimming test 57

Figure 3-14: The swimming speed of the fish robot 59

Figure 3-15: Measured thrust of a rigid connection 59

Figure 3-16: Effect of LIPCA’s inertial force 60

Figure 3-17: Schematic diagram of the thrust measurement 61

Figure 3-18: The average thrust of the fish robot 62

Figure 3-19: The thrust data of the fish robot at a tail beat frequency of 3.7 Hz.62 Figure 3-20: Thrust coefficient of the fish robot (see Appendix for more detail) 63

Figure 3-21: Changes in the duty ratio 64

Figure 3-22: Schematic diagram of the turning experiment 65

Figure 3-23: Fish robot in the turning experiment 65

Figure 3-24: The turning radii of the fish robot 66

Figure 3-25: The Reynolds number of the fish robot 66

Figure 3-26: The Froude number of the fish robot 69

Figure 3-27: The Strouhal number of the fish robot 69

Figure 4-1: Deformation of LIPCA under compressive mechanical load 74

Figure 4-2: The actuation mechanism using four CLIPCAs 74

Figure 4-3: (a) Configuration of the CLIPCA fish robot, (b) The CLIPCA fish robot can be excited by LIPCA and compressed LIPCA 75

Figure 4-4: Schematic diagram of drag measurement 76

Figure 4-5: Schematic diagram of bearing friction measurement 76

Figure 4-6: The computation domain of the CLIPCA fish robot: (a) a view of meshed domain; (b) detail of this mesh near the fish robot 81

Figure 4-7: The boundary condition of the computation model 81

Figure 4-8: The swimming speed of the CLIPCA fish robot actuated by LIPCAs 82

Figure 4-9: The swimming speed of the CLIPCA fish robot actuated by CLIPCAs 83

Figure 4-10: The average thrust of the CLIPCA fish robot for 0.6A tail fin 83

Figure 4-11: The drag of the LIPCA fish robot 84

Figure 4-12: Comparison of drags of the CLIPCA fish robot and the LIPCA fish robot (without electric wires) 84

Figure 4-13: Drag coefficient of fish robots 85

Figure 5-1: Tail fin kinematic 91

Figure 5-2: Tail-beat angle 91

Figure 5-3: Simulation model 94

Figure 5-4: Thrust comparison between analysis and measurement at 3.7 Hz tail-beat frequency 94

Figure 5-5: Position of the cross section 95

Figure 5-6: Thrust analysis of the fish robot at 3.7 Hz tail-beat frequency 95

Figure 5-7: Velocity, vortex, and pressure at the cross-section of tail fin 97

Figure 5-8: Average thrust comparison between analysis and measurement 98

Figure 5-9: Thrust analysis comparison 98

Figure 5-10: Vortex comparison 99

List of Tables

Table 1-1: Reynolds numbers of fish 2

Table 2-1: Material properties of LIPCA’s layers 26

Table 2-2: Dimension of linkages of the fish robot 35

Table 3-1: Dimension of linkages of the improved fish robot 46

ABSTRACT

Design, Fabrication and Thrust/Drag Analysis of Improved Fish

Robots actuated by Piezoceramic Composite Actuators

Quang-Sang Nguyen

Department of Advanced Technology Fusion

Graduate School of Konkuk University

This dissertation focuses on the design, fabrication and thrust/drag analysis of fish robots actuated by Lightweight Piezoelectric Composite Actuator (LIPCA)

In the first model, four LIPCAs were utilized in the actuation system The linkage system was composed of links and two gears to amplify actuation displacement produced by LIPCAs The tail-beat angle by means of vector calculus to evaluate the actuation mechanism was estimated The calculated tail-beat angle was compared with the measured tail-beat angle in air at a low frequency (about 0.5 Hz) Tests of the first model in water were carried out to measure the tail-beat angle, thrust, and swimming speed for various tail-beat frequencies from 0.5 Hz to 2 Hz They were compared with those of the fish robot actuated by two LIPCAs in the previous study The best tail-beat frequency

of the fish robot was 1.4 Hz and the maximum average thrust force was 0.0048

N A miniaturized power supply, which was developed to excite the LIPCAs, was installed inside the fish robot body for free swimming The maximum free-swimming speed was 3.2 cm/s due to the increased size

In the second model, a new actuation mechanism was developed which was simple to fabricate because it works without gears With the new mechanism, the

series of tests of the second model in water were carried out to measure the beat angle, thrust, swimming speed for various tail-beat frequencies from 1 Hz to

tail-5 Hz, and turning radius The maximum swimming speed of the fish robot was 7.7 cm/s at a best tail-beat frequency of 3.9 Hz A turning experiment at the best frequency showed that the swimming direction of the second model could be controlled by changing the duty ratio of the driving voltage; it showed a turning radius of 0.41 m for a left turn and 0.68m for a right turn

Actuation force of the LIPCA can be improved by applying a compressive force

on the LIPCA Thrust measurement was conducted for two fish robots with the same tail fin areas to investigate the thrust force improvement by using the compressed LIPCA One fish robot was actuated by the compressed LIPCAs, and the other one was actuated by the original LIPCAs The results showed that the use of compressed LIPCAs could increase the best tail-beat frequency of fish robot from 3.7 Hz to 4.2 Hz, resulting in 10% increase in hydrodynamic force However, when springs were utilized to apply the compressive force on LIPCAs, the cross section became slightly larger than that of the fish body actuated by original LIPCAs To evaluate effect of the larger cross section, drag of the fish robots was estimated by measurement and computational fluid dynamics (CFD) analysis with the ANSYS-CFX software package The CFD analysis and measurement showed about 4.9% and 8.7% increment of drag, respectively, due

to about 8.4% larger cross section

To identify the mechanism to produce a highest thrust, a three-dimensional CFD simulation with the ADINA software was carried out by implementing the tail-beat motion captured by using a high speed camera (Photron Ultima APX 120K) The thrust simulation was done for the 0.6A tail fin at 3.7 Hz tail-beat frequency which can create the fastest swimming speed The calculated time history of the thrust agreed well with that acquired from experimental with about 14% difference This means the force measurement apparatus worked properly

providing fairly reliable data Besides, based on velocity, vortex, and pressure images at a cross section of the tail fin, peak thrust generation progress was explained To confirm the best frequency, which was found out from thrust experiment, thrust was calculated for the 0.6 A tail fin at three tail-beat frequencies: 3.2 Hz, 3.7 Hz and 4 Hz The simulation result showed that highest average thrust was produced at 3.7 Hz tail-beat frequency

This work may provide a guideline for the design, analysis, and experimentation

of fish robots propelled by artificial muscle-type actuators

1 Introduction

1.1 Understanding fish swimming

After million years of evolution, fish became master of art of propulsive interaction with water with more than 22,000 fish species [1] Due to high density of water, force of gravity on fish body is mostly close to its buoyancy Therefore, some fish pieces are very good in maneuvering or hovering On the other hand, fish easily transfer its momentum into water surrounding it due to incompressibility characteristic of water Hence, some others pieces are fanta1stic cruise swimmers [2]

Figure 1-1 shows configuration of fish Fishes swim by transferring momentum from the fish to surrounding water To create thrust, fish bends their bodies into backward-moving propulsive wave from their caudal fin which is called body and/or caudal fin (BCF) locomotion mode On the other hand, fish uses median and pectoral fins for maneuvering and stabilization which is called median and/or paired fin (MPF) locomotion mode as shown in Figure 1-2 Figure 1-3 clearly shows about the relationship between propulsors and swimming functions

During swimming, as shown in Figure 1-4 (a), there are four main forces acting

on the fish body such as via buoyancy plus hydrodynamic lift, weight, thrust, and resistance For balancing the vertical forces, fish can create supplement lift force by using their pectoral fins extended For forward swimming, thrust created by fish can be larger than resistance which was called sometime drag Drag has three main components: viscous (or friction drag), form drag, vortex (or induced drag) Viscous drag is skin friction between the fish and the boundary layer of water It depends on the wetted area and swimming speed of

the fish and property of layer water around the fish Form drag is generated by the distortion of flow around solid bodies and depends on their shape Induced drag is energy lost in the vortices formed by the caudal and pectoral fins when they generate lift or thrust On the other hand, pitch, roll, and yaw motions as shown in Figure 1-4 (b) are also important terms when the hydrodynamic stability and direction of fish movement are considered

Table 1-1: Reynolds numbers of fish

Species Velocity

(ms-1)

Length (m)

Sea urchin sperm 0.0002 0.00015 3 × 10-2

Figure 1-1: Fish configuration [2]

Figure 1-2: Swimming mode associated with (a) BCF propulsion and (b) MPF

propulsion [3]

dorsal fin

caudal fin (tail)

median

anal fin pectoral fin

paired

pelvic fin

caudal peduncle fin-base

main axis

Figure 1-3: Diagram showing the relation between swimming propulsors and

MPF propulsion

Oscillations Undulations Fin Oscillations Fin Undulations

Transient movements

Periodic swimming ACCELERATING CRUISING MENOEUVRING

resistance boyancy plus hydrodynamic lift

yaw

To evaluate locomotion of fish such as cruise, and maneuvering, researchers considered fish in terms of some dimensionless numbers: Reynolds number, Strouhal number, and Froude number

1.1.1 Reynolds number

The Reynolds number (Re) is a dimensionless number which shows ratio of inertial forces to viscous forces [1] The Reynolds number is used to characterize different flow regimes, such as laminar or turbulent flow When Re is low which means the viscous forces were dominant, laminar flow occurs and flow is characterized by smooth, constant fluid motion On the contrary, turbulent flow occurred at high Reynolds numbers The inertial force is domination at this time and produces random eddies, vortices and other flow instabilities The Reynolds number for a fish is calculated as follows:

Re  U L ρ

where U is velocity of fish, L is length of the fish, ρ is the density or weight of fresh water (which is about 1000 kgm-3 at 1 atm and 15oC), and µ is viscosity of water

In fish research field, the Reynolds number is used to evaluate cruise performance of the fish The Reynolds number of aquatic animals is from less than 1 for bacteria to 108 for the blue whale as shown Table 1-1 [1] Depending

on value of the Reynolds number, there are two groups: plankton and nekton as shown in Figure 1-5 In plankton group, animals are too small and have slow speed Therefore, their main concern is to avoid sinking Their weight and buoyancy are the important forces While, aquatic animals in nekton group are larger than plankton, they have powerful enough for fast swimming for example whales, dolphins and squid

Figure 1-5: Typical velocity of the largest possible variety of swimmers as a

function of the Reynolds number [1]

1.1.3 Strouhal number

Plankton

Blue whale

Nekton

Fishes Copepods

Sperm Bacteria

In fluid analysis, the Strouhal number (St) is a dimensionless number describing oscillating flow mechanisms and is essentially ratio of unsteady force to inertial force The Strouhal number is related how fast vortices are being generated and the space between them Based on experimental studies, Triantafyllou et al concluded that maximum spatial amplification and optimum creation of thrust lie

in a narrow range of St which was from 0.25 to 0.35 with peaking at 3.0 [6] The Strouhal number of a fish is defined as follows:

where A is width of the wake, which is approximated the amplitude of tail oscillation; and f is the frequency of oscillation which is tail-beat frequency

1.2 Overview of research on fish

To understand how fish swims, many theories have been developed to explain the swimming mechanism of fish Lighthill was highly successful with the best-known theory for the hydrodynamics of undulatory fish swimming, in which the curvature of the fish was assumed small and the effect on the fish of the vortex wake was neglected [7, 8] Treating the fish as an elastic plate, Wu developed a two-dimensional waving plate theory [9] which formed the basis for Lighthill’s elongated-body theory [7, 10] To be able to consider carangiform swimming, Lighthill developed the original theory to cater for fish motions of arbitrary amplitude, leading to the large-amplitude elongated-body theory [11] While,

Cheng et al made assumption that fish was infinitely thin and the undulations of

small amplitude [12] Base on those theories, researchers have used numerical models to simulate fish swimming Studies on simulation help researchers more

understand about the flow field around fish during its motion Adkins et al

proposed a CFD simulation to investigate fluid around the fish during motion

[13] The flow around the fish can be examined by changing of amplitude and frequency of the swimming motion (Figure 1-6) To estimate vorticity and thrust

produced by swimming motion, Z Zhang et al built a model (Figure 1-7) based

on the solution of the Navier-Stokes equations on unstructured meshes [14]

Kowalczyk et al used a commercial software ANSYS (CFX-10.0) to model

undulatory and oscillatory fin movements [15] The authors studied the influences of different fin-like movements of the surrounding fluid such as vortex and velocity

1.3 Overview of research on fish robot driven by electromagnetic motor

Propeller-driven designs have been developed and widely used for underwater vehicles for a long time but it was not very efficient for small underwater vehicle Their efficiencies are no more than 40 percent and they show power maneuverability [16] Swimming speed and maneuverability of propeller driven designs are inversely proportional to each other; a high-speed vehicle shows low maneuverability In addition, power efficiency of propeller is only optimal in a range of operation swimming speeds To overcome these restrictions, researchers have been studying locomotion of aquatic animals In addition, autonomous underwater vehicles (AUV’s) missions require a variety of capabilities can be mutually exclusive: high transit speed, long range and duration, maneuverability, and station-keeping ability Whereas, fish can cruise great distances at significant, maneuver in tight space, and accelerate and decelerate quickly from rest or low speed, especially highly efficient-fast swimming and silent propulsion For example, dolphins follow ships at 20 nautical miles per hour (knots) with impressive grace and apparent ease, playfully bursting Tuna, which

used by fish can potentially provide inspiration for a design of propulsors that will outperform the thrusters currently in use

There were a lot of models of fish robots presented by now The most famous one is MIT’s motor-actuated RoboTuna, which was used to study fish propulsion mechanism [18] Robot tuna is 1.25 m long, covered with a flexible skin and equipped with a tail fin The outer shape of the robot is the same with that of a bluefin tuna The caudal fin has maximum (vertical) span 0.32 m Figure 1-8 shows the mechanism of the robot, which is assembled of eight links made of anodized aluminium

Another famous fish robot is Draper Lab’s hydraulic-actuated Vorticity Control Unmanned Undersea Vehicle (VCUUV) [19] This vehicle is the first mission-scale, autonomous underwater vehicle that used vorticity control propulsion and maneuvering To study the energetic and maneuvering performance of fish-swimming propulsion, the VCUUV mimicked the morphology and kinematics of

a yellowfin tuna Figure 1-9 shows configuration of the VCUUV The VCUUV can swim at a stable speed of 1.2 m/sec and has maneuvering with turning rates

up to 75 degrees per second In addition, there are two others famous fish robot: RoboPike [20], Boxybot [21] as shown in Figure 1-10 and 1-11, respectively

Wang presented a novel unmanned underwater vehicle biomimetic robot fish which was able to carry out underwater mission [22] The fish robot was mimicked as carangiform swimmer The fish robot has four links, which were connected with four servomotors, the control system and battery, which were equipped in the plastic fish head (Figure 1-12)

Figure 1-6: Flow visualization of velocity field in the x-y planes [13]

Figure 1-7: Viscous flow around a swimming fish: (a) mesh around the fish

model; (b) detail of a cut through the mesh [14]

Figure 1-8: Schematic view of the eight internal links of the MIT’s Robot Tuna

mechanism [18]

Figure 1-9: Draper Lab’s hydraulic-actuated Vorticity Control Unmanned

Undersea Vehicle’s configuration [19]

The fish robots research mentioned above are based on the BCF propulsion mode More recent researches turn their interest to the MBF mode Kato tried to control a fish robot with pectoral fins in horizontal plane The fish robot was guided and controlled with a pair of apparatus of pectoral fin motion on both sides in horizontal plane (Figure 1-13) The fish robot with a pair of apparatus of

pectoral fin motion has high performance of maneuverability such as turning at the same position [23]

Figure 1-10: RoboPike [20]

Figure 1-11: Design of Boxybot fish robot [21]

Figure 1-12: Prototype of the experimental robot fish [22]

Figure 1-13: Apparatus of pectoral fin motion [23]

Other researchers studied on real fish Arthur measured forces on the tail surface

of bluefish, such as thrust, drag and acceleration [24] Two pressure transducers were used to measure forces on the right and left sides of the tail, as shown in

Figure 1-14 As important consideration of fish robots, thrust and drag were studied in literatures The study can be done by careful experimentation and

computational fluid dynamics (CFD) Barrett et al built an apparatus for drag

measurement to estimate the drag of a robotic fish [25] In this apparatus, the fish robot was towed by a streamlined strut (Figure 1-15) With the same method,

the drag of the autonomous fish robot was measured by Listak et al [26]

1.4 Overview of fish robot driven by smart material

The fish robots actuated by electromagnetic motor have large size due to a complex mechanism to transfer torque from motor to tail fin jawing motion Therefore, power efficiency of the fish robot is mostly low Besides, the electromagnetic motor generates electric noise during operation

To develop a new class of actuator, called artificial muscles, researchers have proposed various types of actuators Piezoelectric materials have been widely used for sensors and actuators, because they respond reasonably fast and have a relatively large actuation force The thin layer composite unimorph piezoelectric driver THUNDER is a representative PZT-based unimorph actuator with a thin PZT layer as an actuating element [27] Light-weight piezocomposite actuator (LIPCA) is a variation of THUNDER in which metals layers are replaced with multiple composite layers as shown in Figure 1-16 The actuation performance

of the LIPCA can be better than that of THUNDER due to a design technique that utilizes the “coefficient of a unimorph actuator” [28] However, the LIPCA can produce an actuation displacement of only a few millimeters under an input

of hundreds of volts

Figure 1-14: Pressure sensor and its location on the fish body [24]

Figure 1-15: Drag measurement apparatus [25]

Polymer-based actuators provide a large actuation displacement or actuation force A representative polymer-based actuator is the ionic polymer-metal composite (IPMC), which can create a large deformation with the application of

just a few volts (Figure 1-17) [29] However, because its actuation force is very limited, the actuation displacement is considerably reduced when the IPMC is under a preload in an application device Dielectric materials also have potential

as a polymer actuator [30] The electro-active material can produce a large volume change under a kilo-volt level input to a membrane with a thickness of a few millimeters This behavior is rather inconvenient in terms of the actuator packaging and the power supply to an application device

A metal-based actuator, such as a shape memory alloy (SMA), can simultaneously produce a large shape change and a large actuation force as shown in Figure 1-18 [31] Unfortunately, its actuation response is too slow The SMA actuator needs another mechanism to recover its shape as quick as possible for most applications Power consumption is another issue in SMA actuators

Thus, each actuator has pros and cons due to its inherent properties Despite the present limitation of these actuators, they still have merits over electromagnetic motors, particularly the silent propulsion, which minimizes noise Many researchers have demonstrated the application of these actuators by implementing them in robotic systems; they have also proposed ways of overcoming the defects of actuators

Figure 1-16: Geometry and position of neutral axis of LIPCA [28]

Figure 1-17: An IPMC strip under a low voltage [29]

Figure 1-18: Large actuation displacement and actuation force of SMA with low

responsiveness [31]

Figure 1-19: Lamprey robot actuated by SMA [32]

Figure 1-20: Two type robots driven by IPMC [33]

Figure 1-21: The tadpole robot: (a) configuration of the microrobot; (b) miniaturized of the battery, electrode and embedded controller located inside the

body [34]

Figure 1-22: The floating boat driven by THUNDER [35]

Figure 1-23: Fish robot actuated by two LIPCA: (a) Configuration of the fish

robot; (b) Linkage system

(b) (a)

Ayers et al used SMA actuators to create the undulating motion of their lamprey fish robot (Figure 1-19) [32] For a faster actuation response, multiple SMA actuators were installed in such a way that they could be cooled by contact with water Ionic conducting polymer film is another type of IPMC Because it cannot generate a large actuation force, it is used for a miniaturized vehicle A 0.2 mm x

3 mm x 15 mm fin made of ionic conducting polymer film as shown in Figure

1-20 was used for an underwater robot that could swim at a speed of 0.52 cm/s with a 2.5 V input [33] A biomimetic undulatory tadpole robot using ionic polymer-metal composite actuators was presented by Kim [34] This robot was a biomimetic undulatory motion of the fin tail Size of the robot was 96 mm in length, 24 mm in width, and 25 mm in thickness (Figure 1-21)

Because piezoceramic actuators can produce a displacement of only a few millimeters, they require another mechanism to amplify or augment the actuation displacement and force Borgen et al installed two sets of THUNDER actuators side by side to form a hinge linkage system (Figure 1-22); this configuration amplified the bending motion of the two actuators and generated a large tail-beat motion in a floating boat [35]

A fish robot driving by LIPCA was proposed by Teddy [36] In that actuation system, two LIPCAs, a rack, two pinions, and coupling links were used to amplify the limited bending actuation displacement of the LIPCAs as shown in Figure 1-23 The fish robot could swim at 2.5 cm/s when the LIPCAs were excited by 300 V peak-to-peak (Vpp) at 0.9 Hz tail-beat frequency

1.5 Objective and approach

Fish robots using the artificial muscle actuators such as polymer, metal and based actuator have been developing to overcome limitations of fish robots driving by electromagnetic motor, for example, complex mechanism The

PZT-LIPCA is a representative of PZT-based actuator It has some advantages, fast responsiveness and high actuation force and has been proven in other work [28]

In the previous study, the LIPCA was available used to drive a fish robot [36] However, its swimming speed was slow due to actuation force limitation This research focuses on design, fabrication and thrust/drag analysis of improved fish robots driven by piezoceramic composite actuators In the fish robot designs, LIPCA and Compressed LIPCA (CLIPCA) are used as artificial muscle for actuating the fish robots

A linkage system is design to transform the bending motion from LIPCAs into a large tail-beat motion The linkage system is integrated with the fish robot body The fish robot can be driven by a miniaturized power supply that fitted in the fish body The fish robot can mimic some characteristics of fish such as swimming mode (BCF), and shape of tail fin Tail-beat angle, thrust, and swimming speed of the fish robot have been estimated

An improved model of fish robot actuated by four LIPCAs is suggested The fish robot has a new propulsion system equipped with a simplified linkage; which result in a smaller cross section Therefore, the new fish robot can swim at a higher speed than that of the previous one Turning motion is demonstrated only

by changing tail-beat motion which is created by modifying duty ratio of the input voltage supply Thrust improvement is also attempted by applying a compressive force on the four LIPCAs

In addition, a force measurement apparatus is developed to investigate the thrust and drag of a fish robot at various tail-beat frequencies A best tail-beat frequency can be determined by conducting swimming speed Besides, a more detail of effects of tail fin shape and tail-beat frequency has been experimentally figured out for the best configuration of the fish robot A drag measurement test

confirm the measured thrust and check if the thrust measurement apparatus works properly

2 A fish robot driven by piezoceramic actuators

In the previous study, the fish robot actuated by two LIPCAs was presented [36] However, the swimming speed of the fish robot was slow due to limited actuation force of two LIPCAs In addition, size of the fish robot was large because its linkage system was placed in the horizontal direction In this chapter,

a new fish robot actuated by LIPCAs that could swim at a higher speed than that

of the previous one was proposed To increase actuation force, four LIPCAs were used in the actuation mechanism of the fish robot The fish robot consists

of a new linkage system which was placed in the vertical direction In this way, the cross section of the fish robot was smaller than that of the previous fish robot which was actuated by two LIPCAs [36] For a free-swimming fish robot, a miniaturized power supply was developed which can install inside and drive the fish robot Experimental test showed that swimming speed of the fish robot actuated by four LIPCAs was about 46% higher than that of the previous one; even though the fish robot was driven by a lower voltage apply (250Vpp) Besides, the fish robot can swim freely by using the miniaturized power supply

at a speed of 3.2 cm/s

2.1 Description of the fish robot

The fish robot, which has dimensions (L × H × W) of 40 cm × 15 cm × 4 cm, has three main parts: a miniaturized power supply, actuators, and a linkage system Figure 2-1 shows the assembly of a fish robot actuated by light-weight piezocomposite actuators (LIPCAs) The body of the fish robot was made of acrylic material The body and/or caudal fin movements were mimicked to achieve a large thrust [37] The shape of the tail fin was modeled on a thunniform fin, and the tail was driven by the LIPCAs through the linkage