SELF ASSEMBLY ROBOTS

Seminar Report Submitted by

Aman kukreja

(Roll No. : M140157ME)

In partial fulfillment of the requirements for the award of the degree of

Master of Technology In

Manufacturing Technology

DEPARTMENT OF MECHANICAL ENGINEERING

NATIONAL INSTITUTE OF TECHNOLOGY CALICUT

NIT CAMPUS PO, CALICUT

KERALA, INDIA 673601

May 2015

CERTIFICATE

This is to certify that the report entitledSELF ASSEMBLY ROBOTS is a bonafide record of the Seminar presented by Aman kukreja(Roll No.: M140157ME), in partial fulfillment of the requirements for the award of the degree ofMaster of Technology in

Manufacturing Technology from National Institute of Technology Calicut.

Faculty-in-charge (ME6325 - Seminar)

Dept. of Mechanical Engineering

Professor & Head

Dept. of Mechanical Engineering Place : NIT Calicut Date : 15 may 2015

ABSTRACT

Self assembly robots have number of small modules(or robots) which can stick or bond together to perform various funtions. In this report bonding methods between modules of self assembly robots are analysed . Tests are conducted and the strength of the bonds for each method are presented for different module styles, bonding conditions and breaking conditions in a destructive test, and arecompared with magnetic bonding methods. For 80micrometer modules, bond strengths of up to 500 mN areobserved with thermoplastic bonds, which indicates that theassemblies could be potentially used in high-force structuralapplications of programmable matter, microfluidic channelsor healthcare. And finally simulation was performed using small magnetic modules to show the assembly and disassembly functions.

i

CONTENTS

List of Abbreviations and Tables ii

List of Symbols iii

List of Figures iv

List of Tables

1 Introduction 1

1.1 Introduction 1

1.2 Problem Definition 1

1.3 Outline of the Report 2

2 The Evolution of Manufacturing Industry 3

3 Actuation and Heating Methods 7

3.1 Actuators 7

3.2 Heating methods 10

4 Experimental setup 14

4.1 Module design and fabrication 14

4.2 Bonding using heating 17

4.3 Motion actuation 19

4.4 Module addressing by magnetic disabling 19

5 Bond analysis 22

5.1 Bonding types 23

5.2 Bonding face styles 24

5.3 Bonding temperature 25

5.4 Assembly and bonding demonstration 26

6 Disabling for module addressability 29

6.1 Disabling for magnetic disassembly 30

7 Conclusions 31

References 32

ii

LIST OF ABBREVIATIONS PCM Photo chemical machining process

LIST OF TABLES Table 1 Evolution of self assembly robots Table 2 Comparison between conventional and solid state refrigeration Table 3 comparison of bonding forces

iii

LIST OF SYMBOLS

total magnetic torque;

total magnetic force;

total magnetic field;

magnetic moment;

0 permeability of free surface;

current through the coil;

iv

LIST OF FIGURES

Fig. 1.1 Space molycubes

Fig. 1.2 Kilobots

Fig. 3.1 Magnetic actuation system

Fig. 3.2 Electric actuators

Fig. 3.3 Peltier element

Fig. 3.4 LASER heating

Fig. 3.5 Induction heating

Fig. 4.1 Photolithography technique

Fig. 4.2 Thermoplastic bonding modules

Fig. 4.3 Solder bonding modules

Fig. 4.4 Heating of modules using peltier element

Fig. 4.5 Focussed laser heating

Fig. 4.6 Magnetic coils

Fig. 4.7 Magnetisation curve

Fig. 5.1 Destructive test setup

Fig. 5.2 Change in displacement force with time

Fig. 5.3 Bond face style

Fig. 5.4 Bonding temperature

Fig. 5.5 Assembly demonstration

Fig. 6.1 Sequence of self assembly

Fig. 6.2 Experimental procedure

Fig. 6.3 Magnetic disassembly

1

CHAPTER 1

INTRODUCTION

1.1 INTRODUCTION

Modular self-assembly robotic systems or self-reconfigurable modular robots are

autonomous kinematic machines with variable morphology. Beyond conventional

actuation, sensing and control typically found in fixed-morphology robots, self-

reconfiguring robots are also able to deliberately change their own shape by rearranging

the connectivity of their parts, in order to adapt to new circumstances, perform new tasks,

or recover from damage.

For example, a robot made of such components could assume a worm-like shape to move

through a narrow pipe, reassemble into something with spider-like legs to cross uneven

terrain, then form a third arbitrary object (like a ball or wheel that can spin itself) to move

quickly over a fairly flat terrain; it can also be used for making "fixed" objects, such as

walls, shelters, or buildings.

1.2 PROBLEM DEFINITION AND CHALLENGES

Recent work in reconfigurable robotics has involved the scaling down of individual

robotic modules for increased resolution and access to small spaces. This has brought

with it several challenges in actuation, computation, and module bonding . Many large-

scale modular robotic systems use traditional actuators such as dc motors or shape

memory alloys to power the assembly, reconfiguring, motion, and disassembly processes

Fig. 1.1 Fig. 1.2

2

using algorithms or motion primitives. However, as these systems are scaled down below

the centimeter-scale, compromises must be made which reduce functionality, resulting in

modules with less mobility.

Since the early demonstrations of early modular self-reconfiguring systems, the size,

robustness and performance has been continuously improving. In parallel, planning and

control algorithms have been progressing to handle thousands of units. There are,

however, several key steps that are necessary for these systems to realize their promise

of adaptability, robustness and low cost. These steps can be broken down into challenges

in the hardware design, in planning and control algorithms and in application. These

challenges are often intertwined.

1.3 OUTLINE OF THE REPORT

Depending on the application area, desirable capabilities for micro-scale modular

robotassemblies could include:-

Small module size

Physical presence (large assembly size)

Addressable modules

Creation of arbitrary 2D/3D shapes

Mechanical strength

Electrical conductivity/continuity

Reconfiguration

Disassembly

Here in this report, small size robotic modules are made using photo chemical

machining(PCM) process. When assembled together using actuation mechnism it can

form large assembly. Each module is addresed separately i.e. we should be able to control

motion of each module separately. Here it is controlled using magnetic field in different

samples having different composition of magnetic particles. Assembly creates various

2D/3D shapes. Bonding of these small modules can be done using thermal bonding or

magnetic attraction. Strength of the bonds are analysed using destructive test. And then

assembly and disassembly is shown by performing simulation.

3

CHAPTER 2

THE EVOLUTION OF SELF ASSEMBLY ROBOTS

The roots of the concept of modular self-reconfigurable robots can be traced back to the

quick change end effector and automatic tool changers in computer numerical

controlled machining centers in the 1970s. Here, special modules each with a common

connection mechanism could be automatically swapped out on the end of a robotic arm.

However, taking the basic concept of the common connection mechanism and applying it

to the whole robot was introduced by Toshio Fukuda with the CEBOT (short for cellular

robot) in the late 1980s.

The early 1990s saw further development from Greg Chirikjian, Mark Yim, Joseph

Michael, and Satoshi Murata. Chirikjian, Michael, and Murata developed lattice

reconfiguration systems and Yim developed a chain based system. While these

researchers started with from a mechanical engineering emphasis, designing and building

modules then developing code to program them, the work of Daniela Rus and Wei-min

Shen developed hardware but had a greater impact on the programming aspects. They

started a trend towards provable or verifiable distributed algorithms for the control of

large numbers of modules.

One of the more interesting hardware platforms recently has been the MTRAN II and III

systems developed by Satoshi Murata et al. This system is a hybrid chain and lattice

system. It has the advantage of being able to achieve tasks more easily like chain

systems, yet reconfigure like a lattice system.

More recently new efforts in stochastic self-assembly have been pursued by Hod

Lipson and Eric Klavins. A large effort atCarnegie Mellon University headed by Seth

Goldstein and Todd Mowry has started looking at issues in developing millions of

modules.

Many tasks have been shown to be achievable, especially with chain reconfiguration

modules. This demonstrates the versatility of these systems however, the other two

advantages, robustness and low cost have not been demonstrated. In general the prototype

systems developed in the labs have been fragile and expensive as would be expected

during any initial development.

4

There is a growing number of research groups actively involved in modular robotics

research. To date, about 30 systems have been designed and constructed, some of which

are shown below.

Table 1: Physical systems created

System Class, DOF

Author Year

CEBOT Mobile Fukuda et al. (Tsukuba) 1988

Polypod chain, 2, 3D

Yim (Stanford) 1993

Metamorphic lattice, 6, 2D

Chirikjian (Caltech) 1993

Fracta lattice, 3 2D

Murata (MEL) 1994

Fractal Robots lattice, 3D Michael(UK) 1995

Tetrobot chain, 1 3D

Hamline et al. (RPI) 1996

3D Fracta lattice, 6 3D

Murata et al. (MEL) 1998

Molecule lattice, 4 3D

Kotay & Rus (Dartmouth) 1998

CONRO chain, 2 3D

Will & Shen (USC/ISI) 1998

PolyBot chain, 1 3D

Yim et al. (PARC) 1998

TeleCube lattice, 6 3D

Suh et al., (PARC) 1998

Vertical lattice, 2D Hosakawa et al., (Riken) 1998

Crystalline lattice, 4 2D

Vona & Rus, (Dartmouth) 1999

I-Cube lattice, 3D Unsal, (CMU) 1999

Micro Unit lattice, 2 2D

Murata et al.(AIST) 1999

M-TRAN I hybrid, 2 3D

Murata et al.(AIST) 1999

5

Pneumatic lattice, 2D Inou et al., (TiTech) 2002

Uni Rover mobile, 2 2D

Hirose et al., (TiTech) 2002

M-TRAN II hybrid, 2 3D

Murata et al., (AIST) 2002

Atron lattice, 1 3D

Stoy et al., (U.S Denmark) 2003

S-bot mobile, 3 2D

Mondada et al., (EPFL) 2003

Stochastic lattice, 0 3D

White, Kopanski, Lipson (Cornell) 2004

Superbot hybrid, 3 3D

Shen et al., (USC/ISI) 2004

Y1 Modules chain, 1 3D

Gonzalez-Gomez et al., (UAM) 2004

M-TRAN III hybrid, 2 3D

Kurokawa et al., (AIST) 2005

AMOEBA-I Mobile, 7 3D

Liu JG et al., (SIA) 2005

Catom lattice, 0 2D

Goldstein et al., (CMU) 2005

Stochastic-3D lattice, 0 3D

White, Zykov, Lipson (Cornell) 2005

Molecubes hybrid, 1 3D

Zykov, Mytilinaios, Lipson (Cornell) 2005

Prog. parts lattice, 0 2D

Klavins, (U. Washington) 2005

Miche lattice, 0 3D

Rus et al., (MIT) 2006

GZ-I Modules chain, 1 3D

Zhang & Gonzalez-Gomez (U. Hamburg, UAM)

2006

The Distributed Flight Array lattice, 6 3D

Oung & D'Andrea (ETH Zurich) 2008

Evolve chain, 2 3D

Chang Fanxi, Francis (NUS) 2008

6

Odin Hybrid, 3 3D

Lyder et al., Modular Robotics Research Lab, (USD)

2008

EM-Cube Lattice, 2 2D

An, (Dran Computer Science Lab) 2008

Roombots Hybrid, 3 3D

Sproewitz, Moeckel, Ijspeert, Biorobotics Laboratory, (EPFL)

2009

Programmable Matter by Folding

Sheet, 3D Wood, Rus, Demaine et al., (Harvard & MIT) 2010

Sambot Hybrid, 3D HY Li, HX Wei, TM Wang et al., (Beihang University)

2010

Moteins Chain, 1 3D

Center for Bits and Atoms, (MIT) 2011

ModRED Chain, 4 3D

C-MANTIC Lab, (UNO/UNL) 2011

Programmable Smart Sheet Sheet, 3D An & Rus, (MIT) 2011

SMORES Hybrid, 4, 3D

Davey, Kwok, Yim (UNSW, UPenn) 2012

Symbrion Hybrid, 3D EU Projects Symbrion and Replicator 2013

ReBiS - Re-configurable Bipedal Snake

Chain, 1, 3D

Rohan, Ajinkya, Sachin, S. Chiddarwar, K. Bhurchandi (VNIT, Nagpur)

2014

7

CHAPTER 3

ACTUATION AND HEATING METHODS

3.1 Actuators

An actuator is a type of motor that is responsible for moving or controlling a mechanism

or system. There are several actuation methods available, but in the case of self assembly

robots it can mainly be divided into two categories as:

i) Actuation through external magnetic field

ii) Actuation through electric signals

3.1.1 Actuation using magnetic field

A MEMS magnetic actuator is a device that uses the microelectromechanical

systems (MEMS) to convert an electrical current into a mechanical output by employing

the well-known Lorentz Force Equation or the theory of Magnetism .

Motion is accomplished by rolling, direct pulling or by vibration based stick slip

crawling.

.

Fig. 3.1

8

3.1.1.1 Mathematic analysis of magnetic actuator

Case Magnetic modules can be controlled by the magnetic coilssurrounding the

workspace, and also interact with eachother. The total magnetic torque and force

thatgovern these interactions are:

(1)

(2)

The magnetic field and its spatial gradients depend linearlyon the currents

through the coils, and so the fieldand gradient terms can be expressed as:

(3)

(4)

where each element ofI is current through each of the ccoils, H is a 3 c

matrix mapping these coil currents to themagnetic field vectorand Hx , Hy ,

Hzare the 3 cmatrices mapping the coil currents to the magnetic field

spatialgradients in the x, y and z directions, respectively. Thesemapping

matrices are calculated for a given coil arrangementby treating the coils as

magnetic dipoles in space andare calibrated through workspace

measurements.

9

Thus, for a desired field and force on a single magneticmicro-robot we arrive

atas:

(5)

In this way, we can achieve 5 DOF control of magnetic modules,enabling

motion on 2D surfaces for assembly tasks as will be shown in the

experimental demonstrations. Motion is accomplished by rolling,

directpulling or by vibration-based stick-slip crawling, methodswhich have

been demonstrated previously. Feedbackcontrol of single or teams of micro-

robots moving in3D are not fully exploited in this work, but the feasibility

ofsuch control has been shown in our previous studies.

3.1.2 Actuation using electricity

An electric actuator is powered by a motor that converts electrical energy

into mechanical torque. The electrical energy is used to actuate equipment

such as multi-turn valves. It is one of the cleanest and most readily available

forms of actuator because it does not involve oil.

Fig. 3.2

10

3.2 Heating methods

A heating system is a mechanism for maintaining temperatures at an acceptable level; by

using thermal energy. As you will see in the coming chapters, how are these method s are

being used to form strong bonds with the neighboring modules.

We can heat the small modules of assembly robots using three methods, which have their

own advantages and disadvantages. These methods are:

i) Heating by Peltier element

ii) Heating by focused laser

iii) Inductive heating by high power AC fields

3.2.1 Heating by Peltier element

The Peltier effect thermoelectric heat pump is a semiconductor based electronic

component that functions as a heat pump. Just by applying a low DC voltage to this

module, one surface gets cold and the other surface gets hot. And just by reversing the

applied DC voltage, the heat moves to the other direction. Thus this thermoelectric device

works as a heater or a cooler.

The Peltier thermoelectric heat pumps have been used for medical devices sensor

technology, cooling integrated circuits, automotive applications and military applications.

3.2.1.1 Principle and operation

The thermoelectric heat pump was discovered by a French watchmaker during the 19th

century. It is described as a solid state method of heat transfer generated primarily

through the use of dissimilar semiconductor material (P-type and N-type).

.

Fig. 3.3

11

Figure 3.3 shows a Peltier effect heat pump typical mechanical and electrical installation.

Like conventional refrigeration, Peltier modules obey the laws of thermodynamics.

Basically the refrigerant in both liquid and vapor form is replaced by two dissimilar

conductors. The solid junction (evaporator surface) becomes cold through absorption of

energy by the electrons as they

pass from the low energy level to the high energy level. The compressor is replaced by a

DC power source that pumps the electrons from one semiconductor to another one. A

heat sink replaces the conventional condenser fins, discharging the accumulated heat

from the system. The following table outlines the differences and similarities between the

thermoelectric module and the conventional refrigerator.

Table 2 : Comparison between conventional and solid state refrigeration.

The evaporator Allows the pressurized

refrigerant to expand, boil and

evaporate (The heat is absorbed

during the change of state from

liquid to gas).

At the cold junction, heat is

absorbed by the electron as

they pass from a low energy

level (P-type) to a high energy

level (N-type)

The compressor Acts on the refrigerant and

recompresses the gas to liquid.

The refrigerant leaves the

compressor as a vapor.

The power supply provides the

energy to move the electrons.

The condenser Expels the heat absorbed at the

evaporator to the environment

plus the heat produced during

compression into the ambient.

Also, the refrigerant returns to

theliquid phase.

At the hot junction, heat is

expelled to the heat sink as the

electrons move from high

energy level to the low energy

level.

12

3.2.2 Focused LASER:

n this process the laser is used to heat the surface of materials. Any subsurface heating is

accomplished by conduction. For intensity values up to about 1x104 W/cm2, the absorbed

power depends on the wavelength ( ), the material (and its surface condition). Generally,

as the material temperature increases, so does the absorption of laser light. The most

relevant processing parameters are the laser power and the beam/material interaction

time. In metals, local surface heating is very rapid and produces a thin hot layer on a

relatively cool bulk material. This conducts heat away from the surface very quickly.

Cooling rates of the order of several thousand degrees per second are possible, which can

be used to advantage in producing microstructural changes, for example, in

transformation hardening which uses IR lasers.

3.2.3 Induction heating:

Induction heating is the process of heating an electrically conducting object (usually a

metal) by electromagnetic induction, through heat generated in the object by eddy

currents (also called Foucault currents).An induction heater consists of an electromagnet,

and an electronic oscillator which passes a high-frequency alternating current (AC)

through the electromagnet. The rapidly alternating magnetic fieldpenetrates the object,

generating electric currents inside the conductor called eddy currents. The eddy currents

flowing through the resistance of the material heat it by Joule heating.

In ferromagnetic (andferrimagnetic) materials like iron, heat may also be generated by

Fig. 3.4

13

magnetic hysteresis losses. Thefrequency of current used depends on the object size,

material type, coupling (between the work coil and the object to be heated) and the

penetration depth.An important feature of the induction heating process is that the heat is

generated inside the object itself, instead of by an external heat source via heat

conduction. Thus objects can be very rapidly heated. In addition there need not be any

external contact, which can be important where contamination is an issue.

Fig. 3.5

14

CHAPTER 4

EXPERIMENTAL SETUP

4.1Module design and fabrication: Modules were fabricated by using PCM technique, Photolithography, also

termed optical lithography or UV lithography, is a process used in microfabrication to

pattern parts of a thin film or the bulk of a substrate. It uses light to transfer a geometric

pattern from a photomask to a light-sensitivechemical "photoresist", or simply "resist," on

the substrate. A series of chemical treatments then either engraves the exposure pattern

into, or enables deposition of a new material in the desired pattern upon, the material

underneath the photo resist.

This process is generally used for manufacturing of electronic circuits, about 0.0125 mm

tolerance can be achieved with this process.

Fig. 4.1

15

4.1.1Thermoplastic bonding modules:

Modules with bonding sites are fabricated in a multi-step molding process,First, a

polyurethanemodule with embedded magnetic composite particles isfabricated in a batch

process using a micro-molding techniqueusing SU-8 photolithography.

They are composed of a mixture of neodymium-iron-boron(NdFeB) particles

(Magnequench MQP-15-7) suspended ina polyurethane matrix (TC-892, BJB

enterprises). The shapewas chosen as roughly hexagonal to allow tessellation ofmany

modules in 2D. Curved edges are used to aid in alignmentof adjacent faces prior to

bonding. The module hasvoids along its edges of approximate size several hundredm,

which will ultimately becomethe binding sites. As a low-temperature thermoplastic

forbonding, commercially available Ethylene-vinyl acetatehot-melt adhesive is used.

These adhesives are availablewith a variety of properties, including different

meltingtemperatures. To fill the binding sites with thermoplasticmaterial in a controlled

manner, the module is placed in asoft rubber jig designed to match the shape of the

module without voids. The thermoplastic is thendeposited under high temperature (130

C) to melt into thevoids, bubbles are removed by placing in a vacuum chamber,and the

remaining material is scraped clean using a sharp edge. After cooling, the moduleis

removed from the jig and is ready for bonding withother modules under low heat

(approximately 70 C). the whole process is shown in the figures below:

Fig. 4.2

16

4.1.2Solder bonding modules:

Modules with solder bonding sites are fabricated in a similarmanner to the thermoplastic

bonding modules, with thethermoplastic replaced by low-melting point metal.

Thisprocedure is shown in figure below. To permit the solder to wetthe the module,

copper is sputtered onto the top and bottom

polyurethane surface. Some copper is deposited into thebinding sites on the side of the

modules during this sputtering.The solder is then applied manually to each bonding

site using tweezers in an optical microscope at 80 C. Anindium alloy (Fields metal) is

used as low melting temperaturesolder, with a sharp melting point of about 62 C.

Thismetal fuses well with itself at high temperature, when in air,water or oil.

4.1.3Magnetic bonding modules:

Bonding by magnetic attraction alone is a simple methodwhich follows naturally from the

use of magneticallyactuatedmodules. This method is widely used in large-scalerobotics,

was investigated for micro-scale robotic elementsin detail in our previous work, and has

the major advantageof being an easily reversible bond. However, the majorlimitation is

that the bonding strength is low, and it requiresspecial out-of-plane module geometry to

resist contact slidingand rotation, as the magnetic force provides no interfaceshear

strength. In addition, modules can only be bondedin a limited number of magnetically

stable configurations.The most stable shapes are long straight chains, althoughsome other

geometries can be assembled, albeit with alower bonding strength.

Fig. 4.3

17

4.2Bonding using heating:

Module bonding occurs by heating to temperatures around5090 C, depending on the

bonding material. As manypotential applications of micro-robots occur in liquid

environments,all experiments are conducted in silicone oil orwater. One heating approach

involves raising the temperatureof the entire liquid volume. This method has

theadvantage of simplicity, and does not require precise knowledgeof the module

locations. However, it melts all bondingsites in the workspace, and takes a longer time to

heat andcool than other methods. It also required mechanical accessto the workspace

area. An experimental setup has been createdto heat the workspace liquid by

electrothermal heating.This setup, shown in Fig. 4.4 , is based around a well ofwater or

oil on a glass slide, which forms the experimentalworkspace. Underneath the glass slide

is glued a Peltierheating element, heated by a 1.0A power supply. In theexperiments in

this paper, the water temperature is monitoredby a thermocouple, and is assumed constant

overthe water volume due to the thin water layer and the highthermal conductivity of

water when compared with the air above it.

Once the desired bonding temperature is reached,the heating current is removed or

reversed and the systemcools to room temperature, at which point adjacent modules

Fig. 4.4

18

are bonded together. For a thin 2mm water layer, heatingfrom 30 C to 70 C with 1A

takes approximately 12 secwhile cooling takes approximately 30 sec with no

currentapplied. Larger volumes of liquid take a longer duration toheat and cool. Heating

by electrical means is used as the primarymethod in this paper due to its simplicity and

precisetemperature control capability.

Laser heating is a second possible method which wouldachieve fast, localized bonding.

This method could be integratedinto an experimental setup for targeted bonding ofonly a

single glue site pair. As a proof-of-concept demonstrationof this method, a commercial

CO2 laser (PinnacleV-Series Laser Engraver, 35W) is used to bond two modules, as

shown in Fig. 4.5. Here the laser power is approximately 3W with a spot size of

approximately 100m, andis traced across the module bond at a speed of 0.55 m/sfour

times with a spacing of 0.5 s. This power providesenough heat to bond the modules

without damaging thepolyurethane module base.

As a third potential method for heating, inductive heatingby high-power AC fields could

be used remotely withouta line-of-sight to the modules. While inductive heating

ofmagnetic particles and particle suspensions have been studied, achieving the desired

temperature rise of over 15 Ccould be a practical challenge with this method, especially

as the module size is reduced below hundreds of micronsize. Magnetic modules are

fabricatedusing the same methods used to make the thermoplastic andsolder modules, but

without the bonding sites. As the bondingstrength is dependent on the magnetic moment

of themodules, a strong magnetic moment is necessary.

Fig. 4.5

19

4.3 Motion actuation:

Magnetic modules are actuated by a set of independentelectromagnetic coils, aligned

pointing towards a commoncenter point, with an open space of approximately 10.4 cm.

The coils are operated with an air or iron core, dependingon the desired magnetic fields

and gradients. The maximumfields produced by the system driven at maximum current

(19A each) are 6.6 kA/m using air cores, and 19.4 kA/m using iron cores. Similarly,

maximum field spatial gradientsare 271 A/m2 using air cores, and 812 A/m2 usingiron

cores. Fields are measured using a Hall effect sensor(Allegro A1321) with an error of

about 80 A/m. Control ofthe currents driving the electromagnetic coils are performedby a

PC with data acquisition system at a control bandwidthof 20 kHz, and the coils are

powered by linear electronicamplifiers (SyRen 25). A photograph of the experimental

coil system is shown in Fig. 4.6.

Magnetic influences have already been explained in 3.1.1.1, using the same effect

magnetic modules are moved to form different assemblies.

4.4 Module addressing by magnetic disabling:

Remotely and selectively turning on and off the magnetizationof individual modules

could allow for addressablecontrol of each module. We have developed a

Fig. 4.6

20

compositematerial whose net magnetic moment can be selectivelyturned on or off by

application of a large magnetic fieldpulse. The material is made from a mixture ofmicron-

scale neodymium-iron-boron and ferrite particles,and can be formed into arbitrary

actuator shapes using thesimple molding procedure discussed before. Themagnetic

coercivity and remanence (retained magnetizationvalue when the applied field His

reduced to zero)are distinct, which allows for moderately-large fields tore-magnetize the

ferrite while maintaining the NdFeB magnetization.By applying a pulse in the desired

directiongreater than the coercivity field (Hc) of the ferrite, its magnetizationdirection can

be switched instantly. In addition,the magnetic states of both NdFeB and ferrite can be

preservedwhen driving an actuator using small fields of lessthan 12 kA/m.

In general it is difficult to demagnetize a single magnetby applying a single

demagnetizing field because theslope of the hysteresis loop (i.e. the magnetic

permeability)near the demagnetized state is very steep. Thus, such ademagnetization

process must be very precise to accuratelydemagnetize a magnet. While steadily

decreasing AC fieldscan be used to demagnetize amagnetic material, thismethoddoes not

allow for addressable demagnetization becauseit will disable all magnets in the

workspace. This motivatesthe use of a magnetic composite to enable

untetheredaddressable magnetic disabling.

We employ a demagnetization procedure to achieve amore precise demagnetization by

employing two materials,both operating near saturation where the permeabilityis

relatively low. In this method, an applied switching fieldHpulsecan be applied to switch

only one materials (ferrite)magnetization without affecting the second material(NdFeB).

Fig. 4.7

21

As the two materials are mixed in one magneticmodule, this switching allows the device

to be switchedbetween on and off states as the magnetic moments add inthe on state or

cancel in the off state. While the internal fieldof the magnet at any point will not be zero,

the net field outsidethe magnet will be nearly zero, resulting in negligiblenet magnetic

actuation forces and torques.

When fields are applied below the NdFeB coercivity, theNdFeB acts as a permanent

magnet, biasing the device magnetization,as shown in Fig. 4.9 for Hpulseup to about 240

kA/m. Traversing the hysteresis loop, the device begins inthe off state at point A, where

motion actuation fields, indicatedby the 12 kA/m range, only magnetize the device

toabout 0.08 Am2, resulting in minimal motion actuation.To turn the device on, a 240

kA/m pulse is applied in theforward direction, bringing the device to point B. After

thepulse, the device returns to point C, in the on state. Here,motion actuation fields vary

the device moment betweenabout 1.7 and 1.8 Am2. To turn the device off, a pulse inthe

backward direction is applied, traversing point D, andreturning to the off state at point A

at the conclusion of thepulse. For small motion actuation fields in the lateral direction,the

device is expected to show even lower permeabilityin the on or off state due to the shape

anisotropy inducedduring the molding process.Thus, modules can be magnetically

disabled by a shortfield pulse of high strength. This will allow for multiplemodules to be

added to an assembly at a time and sequentiallydisabled. As the process is reversible, the

modules canbe re-enabled magnetically once the assembly process iscomplete, allowing

for the entire assembly to be actuated.

22

CHAPTER 5

BOND ANALYSIS

Module bonding strength is measured in a destructive testwhich pulls two modules

directly apart while measuring theloading force using a load cell, as shown in Fig. 5.1.

Themodule pair is glued to a 3-axis motion stage with manuallinear motion. The loading

beam is placed on the loadcell and a fulcrum, and serves to reduce the force

transmittedto the load cell, as shown schematically in the figureinset. The modules are

first glued on a removable plastictip by instant adhesive. Then the tip is fixed on the

motionstage and aligned manually using the camera such that themodules come into

contact with the loading beam, which iscoated in a thin layer of adhesive. When the

adhesive hascured, the motion stage is lowered at a constant rate untilthe bond between

the modules breaks. The load cell datais acquired by amplifier/conditioner (TMO-2) and

DAQ ata rate of 10 kHz. To get the breaking force, the differencebetween the maximum

value during the breaking and thevalue of zero load after the break is calculated, as

shown inFig. 5.2. Calibration is made by a 20 g proof mass placed atthe same position on

the loading beam in a separate measurement.

Fig. 5.2

Fig. 5.1

23

The bond breaking phase is not instantaneous dueto the viscoelastic nature of the

thermoplastic at elevatedtemperatures.

5.1 Bonding types

Representative thermoplastic, solder, and magnetic bondingmodule pairs are compared

with respect to bonding force,with results shown in Table 3. Module style refers to

thesize, number and shape of the bonding sites, and can be referencedin Fig. 5.3. This

table shows the mean and standarddeviation in force for five module pairs for each type

oftest. Modules assembled by hand are pushed together usingtweezers in a microscope

and then heated. Those assembledby coils are pushed together in the magnetic coil

systemusing magnetic force on one module while the secondmodule remains in place.

Assembly by hand using tweezersis also directly compared with assembly using

magneticforces in the magnetic coil system, showing that while handassembly can result

in larger bond force, the coil-assembledpairs maintain an adequate force. Bonding by

laser wasonly performed for one module pair as a proof-of-conceptdemonstration due to

the difficulty in aligning the laser inthe current setup. Results for solder bonding show

moderatebond force, where it is noted that the failure mode is the delamination of the

solder with the copper seed layerrather than failure of the bulk solder. As expected,

magneticbonding is by far the weakest bonding style, but has theadvantage of easily

reversible bonding.

Table 3 Comparison of bonding forces

24

5.2 Bonding face style

For thermoplastic bonds, the bonding faces of each modulecontain voids for

thermoplastic to reside. By varyingthe number, shape and size of these voids, different

bondingstrengths are achieved, with results shown in Fig. 5.3for five samples of each

face style. Six styles of modulewith the same outer diameter size of 800m but different

bonding faces are designed and fabricated for direct comparison.Below the bar graph, the

polyurethane module bodyis shown in white while the thermoplastic is shown in grey.To

obtain a thin layer of thermoplastic in style A with novoids, the 800m module is

placed into a jig with diameterof810m. All other module styles are prepared in a jigA

three-dimensional system approach is

presented in this paper as a decision-

support framework for environmentally

sustainable manufacturing. This system

approach considers the three components

of manufacturing: technology,of size

800m. Results in the bar graph show the

advantagesof using dedicated

thermoplastic voids over the A

style. This indicates that the design of

bonding faces withvoids has improved the

bonding force between modulesfrom

approximately 100mN to several hundred mN. It isseen that all void styles show adequate

bond force, but thesingle-void of wide size but shallow depth (style C) showsthe highest

bond force. It is expected that the wider bondingsite increases the bond strength as it

should depend primarilyon the cross-sectional bond area, but the depth of style

Dprevents it from having high adhesion as the glue needs topenetrate deep into the

recess to adhere fully.

Thus, whileit is determined that any module style with voids resultsin acceptable bond

force, further study could improve thebond force even further through optimized void

style. It isthus desired to increase the cross-sectional bond area whilemaintaining full glue

penetration.

Fig. 5.3

25

5.3 Bonding temperature:

To determine the required temperature to form a securebond, the breaking force of

thermoplastic module pairsbonded under different temperatures is investigated,

withresults shown in Fig. 5.4a. Here

heat is applied carefully toeach C-

style module pair using a heat gun in

an air environment.The temperature

at the bonding site is monitoredusing

a thermocouple. After cooling to

room temperature,each module pair

is mounted in the destructive test, for

breaking at 22 C. Results show that

a critical temperatureof about 55 C

is necessary to form a bond.

Modulepairs indicated with 0

bonding force achievement no

binding atall, as indicated by them

not being able to support their

ownweight when lifted with tweezers. At temperatures above55 C, high bonding force is

relatively consistent (with onefailed bond at 65 C). However, above about 80 C

thethermoplastic melts to such a degree that it flows from thebond sites. While this can

allow for high bonding force, thegeometry of the module pair is distorted by the pool of

thermoplasticwhich forms around the base of the modules. Theduration of heating was

not observed to have an effect onbond strength.

The strength of a thermoplastic bond broken at high temperature is investigated in Fig.

5.4 b, where identical modulesbonded at 70 C are broken at varying temperatures.

Hightemperature during the break test is achieved using a Peltierthermo-electric element,

driven by 1A current, and monitoredwith a thermocouple glued to the Peltier elementface

near the module pair. The plot shows that for temperatureslower than about 40 C, the

bond force is highand relatively insensitive to temperature. However, at

Fig. 5.4 a, b

26

highertemperatures the bond force is lower. The bond force neverreduces to zero due to

the strong capillary and visco-elasticnature of the melted thermoplastic. The results from

bothof these temperature tests are specific to the thermoplasticused.

5.4 Assembling and bonding demonstration:

A demonstration to show the bonding of 2D tessellating shapes is shown in Fig. 8. Here,

six F thermoplastic modulesare moved in the magnetic coil system for assemblyin Fig.

5.5. Modules are bonded by applying heat

using theintegrated resistive heating setup.

The insetin Fig. 5.5b shows the capillary

drawing action which pullsadjacent modules

into intimate contact during the reflowprocess.

In this experiment, only one module is

magnetic,allowing for the step-by-step

addition of the non-magnetic modules.When

the assembly is moved adjacent to a

newmodule, heat is applied to bond it to the

assembly. Thisproves that subsequent bonding

cycles can be performed inthe presence of existing bonds at other sites without

negativeeffect. The single magnetic module is able to carrythe non-magnetic modules,

although when the assemblyreaches a size of five modules, as in Fig. 5.5e, the mobilityis

somewhat reduced, making it more difficult to preciselyadd more modules to the

assembly in the desired configuration. However, the large assembly is still very mobile

usingthe methods of stick-slip crawling or tumbling cartwheellikerolling motion. The

bond force is high enough that theassembly is not broken by moving in the coil system

undermagnetic torques and forces, even when the temperature iselevated to 70 C.

Fig. 5.5

27

CHAPTER 6

DISABLING FOR MODULE ADDRESSABILITY

To add more magnetic modules to an assembly, each moduleis magnetically disabled

after assembly, as shown schematicallyin Fig. 6.1. This process allows individual

modulesto be added to the assembly one at a time into arbitrarylocations on the assembly

without regard for the magneticattraction or repulsion associated with that assembly

location.Compared with bonding by magnetic attraction, thisallows for a much wider

range of assembly morphologiesto

be made. New modules can be

added to the assemblyin any

position or orientation. However, it

is desirable tobond all modules

with the same orientation so that

they canbe disabled and enabled as

a group using a single globalpulse.

Heat bonding of modules is

accomplished by individualheating

(i.e. by laser or inductive heating),

or by globalheating (i.e. heat

conduction through the entire

medium).Once the assembly is

completed, it is re-enabled magnetically and is free to move through the workspace as a

singleunit.

As further demonstration of the assembly plus heat activated bonding, a 2D ship-in-bottle

morphology is createdin a microfluidic chip environment, as shown in Fig. 6.2.Creating a

ship in a bottle requires individual modules topass through a small opening (the

bottleneck) and assembleinto a ship shape one at a time. The simple 2D ship shapemade

here consists of nine modules, and consists of a hull,mast and sail. Such a demonstration

shows the capability ofthe presented addressability, bonding and control method

toachieve the creation of arbitrary shapes in remote inaccessibleareas. The assembled

Fig. 6.1

28

ship is enabled magnetically bya magnetic pulse for actuation as a rigid body. The

strengthof the assembly is demonstrated by fast actuation after it is assembled, and is

promising for future physical interactionswith the environment.

6.1 Disabling for magnetic disassembly:

A final study is conducted using magnetic disabling to show that magnetic module bonds

can be broken directly. Whilewe have demonstrated that magnetic bonds aremuch

weakerthan the new thermally-activated bonds, they could be usefulfor reversible

bonding capability. Our previous workshowed such magnetic bonds can be broken by

anchoringand pulling modules apart. However, such a methodrequired electrostatic

anchoring to apply pulling forces. Wenow introduce the concept of magnetic disabling to

such magnetic bonds for reversible control without the needfor electrostatic anchoring. In

this method, a magneticallystableassembly is created, with the restriction that allmagnetic

modules must be assembled oriented in the samedirection. To disassemble the group, all

modules are magneticallydisabled using a magnetic field pulse, in themanner of the

previous section. Thus, the magnetic bondingforce reduces to near-zero and the assembly

Fig. 6.2

29

is brokenapart by small magnetic actuation forces and torques. Atthis point, individual

modules can be re-enabled for re-usein further assemblies. The process is thus completely

repeatable.A demonstration of this type of disassembly is shownin Fig. 6.3. Assembly in

this case was performed by handusing tweezers. Magnetic modules in this study are

identicalto those used in the heat-assisted bonding study in thischapter, but without heat

activation. This demonstrates thatthe principle could be used to create easily

disassembledassemblies for reconfigurable magnetic micro-robotic tasks.Further

investigations of the limits and usefulness of thistechnique could lead to a robust and

simple disassemblymethod for magnetically-bonded assemblies.

Fig. 6.3

30

CHAPTER 7

CONCLUSIONS

In this report, a comparison of severalbonding methods for micro-scale modular robotics

is shown. As areversible bonding method, we investigated magnetic attractionbetween

modules, and as a stronger but non-reversiblebonding method,heat-activated metal solder

and thermoplastic bonding methods were shown.We showed solder and

thermoplasticbonds activated by increasing the temperaturethrough heating of the entire

workspace, or as a proof-ofconceptdemonstration, through focused laser heating forfaster

heating/cooling in environments which may not beconducive to large heating elements.

As a potential methodfor use in applications which lack line-of-sight view, weproposed

the use of inductive heating to perform bondingwith additional study required. The

temperature increaserequired for bonding could be reduced through the use ofdifferent

temperature sensitive materials with lower meltingpoints. Bond strength was tested for a

variety of differentmodule styles for the thermoplastic bonds, and thetemperatures

required for solid bonding and cooling wereinvestigated. As a demonstration, the

magnetic moduleswere moved in a magnetic coil system as addressable

microroboticagents to remotely form a target structure from upto nine independent

modules in a fluid environment accessiblefrom only a small opening. Such an assembly

methodcould potentially be used to form complex desired shapes ininaccessible small

spaces for microfluidic manipulation orhealthcare applications. As an additional bonding

technique,we also demonstrated reversible magnetic bonding using amagnetic disabling

technique for easy bond reversal. Futurework will involve developing reversible adhesive

bonds,and creating larger complex assemblies for tasks insidefluid channels. In addition,

the formation of out-of-plane3D assemblies will be investigated for applications suchas

in-situ heterogeneous tissue scaffold construction. Suchapplications may require reduced

scale using fabricationmethods such as flip-chip assembly and biocompatibilitythrough

proper choice of materials and coatings.

31

REFERENCES

[1] Diller E, Zhang N, SittiM , 2013, Bonding methods for modular micro-robotic

assemblies, International conference on robotics and automation, pp 25732578.

manufacturing, 61, pp. 39-42.

[2] http://users.wfu.edu/ucerkb/Nan242/L15-Photolithography.pdf

[3] Alaoui C, 2011, Peltier Thermoelectric Modules Modeling and Evaluation,

International Journal of Engineering., 5, pp. 115-121.

[4] Diller E, Pawashe C, Floyd S, Sitti M, 2011, Assembly and disassembly of

magnetic mobile micro-robots towards deterministic 2-D reconfigurable micro-

systems, International Journal of Robot Res., 14, pp 16671680.

[5] Christopher J. Morris, Brian Isaacson, Michael D. Grapes, and Madan Dubey, 2011,

Self-Assembly of Microscale Parts through Magnetic and Capillary Interactions ,

Micromachines., 2, pp 69-81.

[6] Mastrangeli M, Abbasi S, Varel C, Van Hoof C, Celis JP, Bohringer KF (2009)

Self-assembly from milli- to nanoscales:methods and applications. Journal of

MicromechMicroeng 19(8):083,001