See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/281836834

Smart Materials and Structures

Research · September 2015

CITATIONS READS

5 33,258

1 author:

Bo-Hua Sun

277 PUBLICATIONS 1,051 CITATIONS

SEE PROFILE

All content following this page was uploaded by Bo-Hua Sun on 14 January 2021.

The user has requested enhancement of the downloaded file.

SMART MATERIALS
AND STRUCTURES
Lecture at Swiss Federal Institute of Technology Zürich (ETH)
Zürich, 14 Sept.,2015

Bohua Sun
Member of Academy of Science of South Africa(ASSAf)
Professor in Engineering
Cape Peninsula University of Technology
Cape Town
South Africa
[email protected]

CAPE PENINSULA UNIVERSITY OF TECHNOLOGY (2015)

Copyright ⃝2015
c by Bohua Sun. All rights reserved.

Published by Cape Peninsula Univerisity of Technology

Cape Town, South Afica.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form
or by any means, electronic, mechanical, photocopying, recording, scanning.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in
preparing this book, they make no representations or warranties with respect to the accuracy or
completeness of the contents of this book and specifically disclaim any implied warranties of
merchantability or fitness for a particular purpose. No warranty may be created or extended by sales
representatives or written sales materials. The advice and strategies contained herin may not be
suitable for your situation. You should consult with a professional where appropriate. Neither the
publisher nor author shall be liable for any loss of profit or any other commercial damages, including
but not limited to special, incidental, consequential, or other damages.

Smart Materials and Structures / Bohua Sun . . . [et al.].

Printed in South Africa.

 To My Dear Father, Mother
and Family

[I hope that I may succeed in deserving and obtaining your confidence. But in the first place, I can
ask nothing of you but to bring with you, above all, a trust in science and a trust in yourselves.
The love of truth, faith in the power of mind, is the first condition in Philosophy. Man, because
he is Mind, should and must deem himself worthy of the highest; he cannot think too highly of the
greatness and the power of his mind, and, with this belief, nothing will be so difficult and hard
that it will not reveal itself to him. ]
—Georg Wilhelm Friedrich Hegel, Oct. 28, 1816 at Heidelberg University
PREFACE

Smart structures or smart materials systems are those, which incorporate actuators and
sensors that integrate into the structures and have structural functionality, as well as highly
integrated control logic, signal conditioning, and signal power amplification electronics.
Such actuating, sensing and controlling are incorporated into a structure for the purpose
of influencing its states or characteristics, be they mechanical, thermal, optical, chemical,
electrical, or magnetic. For example, a mechanically smart structure is capable of altering
both its mechanical states (its position or velocity) or its mechanical characteristics (its
stiffness or damping). Optically smart structures could, for example, change colour to
match its background.
In future, it is expected that there will be widespread application of the technology that
is under development, in its current and evolutionary forms. The breadth of application of
this technology is expected not only towards high-tech but also towards civilian fields.
These lecture notes have especially been prepared for the Seminar at Institute of Struc-
tural Engineering at ETH. I would like to take this opportunity to address some important
issues regarding smart materials and structures and to introduce some work from my re-
search group.
I would like to express my deep gratitude to Prof. Dr. Eleni Chatzi1 for her warm
hospitality and to the South African National Research Foundation for its financial support.

B OHUA S UN
Cape Town, South Africa and Zürich, Switzerland

1 Prof. Dr. Eleni Chatzi is the Chair of Structural Mechanics at ETH

v
CONTENTS

Preface v
List of Figures xi
List of Tables xv

1 INTRODUCTION 1
1.1 The Needs 1
1.2 Smart Materials Age 2
1.3 Smart Structures and Development Background 3
1.4 Smart Materials 3
1.5 Smart Structures 4
1.6 Critical Component Technologies of Smart Structures 5
1.7 Smart Materials and MEMS 7

2 Piezoelectric Materials 9
2.1 General Concepts 9
2.2 Piezo Transducers 11
2.3 PZT Application 12

3 Shape Memory Materials 17

3.1 General Concepts 17
3.1.1 Ferromagnetic Shape Memory Alloys 19
vii
viii CONTENTS

3.2 SMA Application 20

4 Magnetostrictive Materials 23
4.1 General Concepts 23
4.2 Basics of Magnetostriction 25
4.3 Magnetosrtictive Materials Applications 26

5 Electrorheological Fluids 29
5.1 General Concepts 29
5.2 Electrorheological Materials Applications 29

6 Magnetorheological fluids 31
6.1 Basic Concepts 31
6.2 MR Actuators 33
6.3 Design of MR Systems 34

7 Fibre Optic Sensors 37

7.1 The Structure of Optical Fibres 37
7.2 Characteristics, Advantages and Capabilities of Fibre Optic Sensors 38
7.3 Capabilities and Significance 39
7.4 Intensiometric Fibre Optic Sensors 39
7.5 Interferometric fibre optic sensors 40
7.6 Bragg (Grating) Fibre Optic Sensor 40
7.7 Polarimetric Fibre Optic Sensors 42
7.8 Modalmetric Fibre Optic Sensors 42

8 DISTRIBUTED PIEZOELECTRIC ACTUATING SMART STRUCTURES 43

9 SOME GENERAL ASPECTS OF PIEZOELECTRIC SENSOR MECHANICS 51

9.1 Introduction 51
9.2 Smart structure and active control 51
9.3 Sensory elements 52
9.4 Working Principle of Piezoelectric Smart Structures 53
9.5 Sensor analysis 55
9.5.1 Piezo Composite Beam 55
9.5.2 Modal sensor 56

10 SMART PIEZO COMPOSITE MINDLIN BEAMS 57

10.1 Introduction 57
10.2 Formulations 58
10.3 Numerical Example 61
 CONTENTS ix

11 RECTANGULAR AND CIRCULAR SHAPE DISTRIBUTED PIEZOELECTRIC

ACTUATOR 65
11.1 Introduction 65
11.2 Rectangular DPA element 66
11.2.1 Polarization of Rectangular DPA Element 66
11.2.2 Distributed Charge Density Analysis 67
11.2.3 Capacitance Analysis of the DPA Element 67
11.2.4 The Mechanical Performance Analysis of the Rectangular
DPA Element 68
11.3 The Electrical Field Distribution in the Part-circular Shape DPA 68

12 ELECTRO-MECHANICAL PERFORMANCE OF C-SHAPE PIEZOELECTRIC

ACTUATOR 71
12.1 Introduction 71
12.2 Displacement and Force Equations 72
12.3 Displacement Analysis and Discussion 74
12.4 Force Analysis and Discussion 74
12.5 Conclusions and Recommendations 75

13 MONITORING STRUCTURAL INTEGRITY USING FIBRE OPTIC SENSORS 77

13.1 Formulations 77
13.2 Conclusions 79
References 81
LIST OF FIGURES

1.1 Approach for classification of active materials 5

1.2 Active material devices used as actuator (top) and sensor (bottom) 6
2.1 Piezoelectric materials 10
2.2 Representation of domain rotation and switching during poling of a
polycrystalline ceramic 11
2.3 Polarization and electrostrictive 11
2.4 Piezo transducers (a) 13
2.5 Piezo transducers (b) 14
2.6 Active Fibre Composite (Piezoelectric fibre) from MIT 15
3.1 Shape memory materials performance 17
3.2 Illustration of the Shape Memory Effect: Original single crystal lattice
structure in (a); upon cooling down to the martensite finish temperature
(Mf) self-accommodated martensite forms without significant change
in the external dimensions (b); upon mechanical deformation (c and d)
the system minimizes energy through twinning (lattice invariant shear),
while maintaining atomic bonds; and heating above the austenite
finish temperature reverts variants to the parent phase in the original
orientation (e) 18
3.3 Stress-strain curves of SMA at phase of austenitic and martensite 19
xi
xii LIST OF FIGURES

3.4 Ferromagnetic Shape Memory Alloys. 19

3.5 Pipe coupling using SMA 21
3.6 Mechanism of pipe coupling using SMA 21
4.1 Strain versus magnetic field. 24
4.2 Thermal expansion of a magnetic material 25
4.3 A prototype magnetostrictive actuator 26
5.1 Effect of applying electrical field to ER materials. 30
5.2 Utilization of ER fluids in power transmission 30
6.1 The Basic Structure of MR Fluid 32
6.2 Strain-stress curve of MR 32
6.3 Shock absorber using MR fluid. 34
6.4 Force-velocity curve 35
7.1 Cross-sectional view of typical optical fibre 38
7.2 Fibre optic evanescent wave absorption technique. 40
7.3 Response characteristics of fibre optic Bragg grating 41
8.1 Schematic representation of a discrete actuator a) and a distributed
actuator b). Distributed actuator systems are often based on multiple
individual strain actuators (modularity); note that the actual contact
area is very thin and can be typically a mechanical connection in a) and
an adhesive layer in b) 44
8.2 Development of distributed actuator system in early 1980s 45
8.3 Development of distributed actuator system in mid 1980s. 46
8.4 Development of distributed actuator system in late 1980s. 47
8.5 Development of distributed actuator system in early 1990s. 48
8.6 Development of distributed actuator system in mid 1990s. 49
8.7 Development of distributed actuator system in late 1990s. 49
10.1 Laminated beam with integrated piezoelectric sensor and actuator 58
10.2 Cross section geometry of an n-layered laminated beam. 60
10.3 Cross section geometry of an n-layered laminated beam. 62
10.4 Effect of negative velocity feedback gain on the tip transient response
[0/90/90/03]. 63
10.5 The tip defection of beam versus feedback control gain. 63
 LIST OF FIGURES xiii

11.1 The principal type of DPA element with interdigital electrode (IDE). 66

11.2 The DPA element. 69

11.3 The DPA element. 69

11.4 The strain result comparison between the analytical model and the FEM. 70

11.5 The DPA element with part-circular shape. 70

12.1 C-shape piezoceramic actuator. 73

12.2 Unimorph C-shape actuator. 73

12.3 Results for Displacement for substrate/PZT thickness ratios (for 1mm
PZT thickness, at 50 V). 75
12.4 Results for the force for substrate/PZT thickness ratios (for 1mm PZT
thickness, at 50 V). 75
12.5 Numerical simulations of deflection and forces. 76

13.1 Embedded optical fibres into the test piece. 77

13.2 Comparison of experimental and theoretical prediction for optical
power through a fibre subjected to elongation owning to an axial force. 79
LIST OF TABLES

10.1 The material properties of the main structure and piezoelectric 62

11.1 The constants in the charge density with different h/a and b/a. 68

xv
CHAPTER 1

INTRODUCTION

[study hard, improve every day.]

— BY C HAIRMAN Z EDONG M AO

1.1 The Needs

The demand for a new generation of industrial, military, commercial, medical, automotive
and aerospace products has fuelled research and development activities, which focus on
advanced materials and smart structures. This situation has been further stimulated by the
intellectual curiosity of humankind in synthesising new classes of bio-mimetic materials.
And, of course, global competition among the principal industrial nations has also been a
parameter in the equation that governs the rate of technological progress. A fundamental
axiom of this field of advanced materials is that the ultimate materials are biological mate-
rials, which replicate such characteristics and properties in synthetic materials and, which
can be employed in diverse scientific and technological applications.
Thus, by integrating knowledge bases that are associated with the mega-technologies
of advanced materials, information technology and biotechnology, the creation of a new
generation of biomimetic materials and structures can be facilitated with inherent brain-
s, nervous systems and actuation systems –this is at present a mere skeleton compared
with the anatomy perceived in the not-too-distant future. This quantum jump in materials
technology will revolutionise the future in ways that are far more dramatic than ways in
Smart Materials and Structures. 1
By Bohua Sun Copyright ⃝ c 2017 Bohua Sun
2 INTRODUCTION

which the electronic chip has impacted on our lifestyles. These new materials are termed
Smart Materials or Intelligent Materials and they will typically feature fibrous polymeric
composite materials that are embedded with powerful computer chips of gallium arsenic,
which will be interfaced with both embedded sensors and embedded actuators by networks
of embedded optical-fibre wave-guides, through which large volumes of data will be trans-
mitted at high speeds.
Today’s material revolution is the cornerstone of the triumvirate of mega-technologies,
which comprise the essential integrates of this embryonic field. These technologies will
have a mutually symbiotic relationship and will significantly impact on one another, re-
sulting in synergistic technological advances which cannot be foreseen today. However,
a natural consequence of advancing these technological disciplines will be the impending
revolution in smart materials and structures.
The classes of smart materials and intelligent structures are diverse and their applica-
tions are largely unknown. However, what is known is that this new generation of materials
will certainly revolutionise our quality of life as dramatically as the state-of-art material-
s did in the past, with stone implements triggering the Stone Age, alloys of copper and
tin triggering the Bronze Age, while the smelting of iron ore triggers the Iron Age. The
time-line of humankind is located at the dawn of a new age,namely The Smart Materials
Age.

1.2 Smart Materials Age

Human civilisation has been so profoundly influenced by materials technologies that histo-
rians have defined time periods by materials that dominated these eras. Thus, as humankind
embarked on a continuous quest for superior products and weaponry that were fabricated
from superior materials, terms such as the Stone Age, the Bronze Age, and the Iron Age
entered the vocabulary. The current Synthetic Materials Age, featuring plastics and fibrous
composites is provides a viable precursor to the dawn of a new era, namely the Smart ma-
terials Age, which will capitalise on these synthetic materials in order to exploit several
eclectic emerging technologies for the synthesis of smart materials that exhibit nervous
systems, brains, and muscular capabilities. The degree of sophistication displayed by this
new generation of materials will depend mostly on the individual applications; however,
it is anticipated that various innovations in diverse science fields will emerge such as nan-
otechnology, biomimetics, neural networking, artificial intelligence, materials science, and
molecular electronics, for example.
This new generation of smart materials will significantly impact on civilisation. For
example, some classes of materials will be able to select and execute specific functions
autonomously in response to changing environmental stimuli and others will only feature
embedded sensory capabilities in order that a structural member is manufactured to comply
with the quality control specifications. Self-repair, self-diagnosis, self-multiplication and
self-degradation are also some of the characteristics that are anticipated to be a feature
of the supreme classes of smart or intelligent materials in an engineering context. All
aspects of civilisation will be influenced by these new generations of innovative materials,
as designers capitalise on their unique capabilities in industries as diverse as aerospace,
manufacturing, automotive, sporting goods, medicine, semi-conductive technology and
civil engineering.
 SMART STRUCTURES AND DEVELOPMENT BACKGROUND 3

1.3 Smart Structures and Development Background

Smart structures or smart materials systems are those, which incorporate actuators and
sensors that integrate into structures and have structural functionality, as well as highly
integrated control logic, signal conditioning, and signal power amplification electronics.
Such actuating, sensing and controlling are incorporated into a structure for the purpose
of influencing its states or characteristics, be they mechanical, thermal, optical, chemical,
electrical, or magnetic. For example, a mechanically smart structure is capable of altering
both its mechanical states (its position or velocity) or its mechanical characteristics (its
stiffness or damping). Optically smart structures, for example, could change colour to
match its background.
Three historical trends have combined to establish the potential feasibility of smart
structures. The first is a transition to laminated materials. In the past structures were
manufactured from large pieces of monolithic materials, which were machined, forged, or
formed to a final structural shape, making it difficult to imagine the incorporation of active
elements. However, in the past 40 years a transition to laminated materials technology
has occurred. Laminated materials, which are built up from smaller constitutive elements
allow for the easy incorporation of active elements within the structural form. One can
now envision the incorporation of a smart ply carrying actuators, sensors, processors, and
inter-connections within the laminated materials.
The second trend has been the exploitation of the off-diagonal terms in the material con-
stitutive relations, which currently enable smart structures. The full constitutive relations
of materials include characterisation of its mechanical, optical, electromagnetic, chemi-
cal, physical, and thermal properties. For the most part, researchers have focused only on
block diagonal terms. Those who are interested in exploiting materials for their structural
benefits have focused only on mechanical characterisation. However, much can be gained
by exploiting the off-diagonal terms in constitutive relations, which, for example, couple
the mechanical and electrical properties. The characterisation and exploitation of these
off-diagonal material constitutive relations have led to progress in the creation of smart
structures.
The third and perhaps most obvious advancements come in the electrical engineering
and computer science disciplines. These include the development of microelectronics, bus
architectures, switching circuitry, and fibre-optic technology. Also, central to the emer-
gence of smart structures is the development of information processing, artificial intelli-
gence, and control disciplines.
The sum of these three evolving technologies (the transition to laminated materials, the
exploitation of the off-diagonal terms in material constitutive relations, and the advance-
ments in microelectronics) have created enabling infrastructure in which smart structures
can develop.

1.4 Smart Materials

The technological field of “smart materials”is not transparent or clearly structured. It has
evolved over the past few decades with increasing pace during the 1990s to become what
it is today, and for transition into the future. Generally speaking, these materials respond
with a change in shape upon application of externally applied driving forces. Typically,
this shape change is reflected in an elongation of the sample, thus allowing the use of, for
example, a small linear motor.
4 INTRODUCTION

The term ”smart materials”, which is sometimes also called intelligent materials or ac-
tive materials, describes a group of material systems with unique properties. At this stage,
the following materials are the active ones:

Piezoelectric materials;

Shape memory alloys;

Magnetostrictive materials;

Electrorheological materials; and

Magnetorheological fluids.

In addition, some material systems that do not exhibit a shape change, but rather have
other significant properties are sometimes also called smart materials (while other, for
some strange reason, are not). Examples of ”other” smart materials include electro- and
magneto-theological fluids. These fluids can change viscosity over many orders of magni-
tude upon application of an external magnetic or electric field.
Consequently, the term “smart materials”is not too very well defined and is frequent-
ly used to describe different systems and systems’s behaviors. Although there have been
approaches to quantify and classify different levels of smartness or intelligence in systems,
from a practical standpoint, it is most important to understand that none of the classifi-
cations are established and used as a standard in the academic, scientific, or industrial
community. Furthermore, one should note that the terms：

Smart materials;

Intelligent Materials;

Active Materials;

Adaptive Materials; and

and to some extent “actuators”and “sensors”.

are almost always used interchangeably. This can sometimes lead to confusion, as different
terms can really describe the same effect or property of a material.
To add to the confusion, the terms “smart devices”, “smart systems”or “smart
structure”are often carelessly used. Here one should note that, in general, the system com-
plexity increases from the unit “material”to “device”to “systems”to “structures”.
Any permutation of the adjective (smart, active) with the subject (material, device,…) is
more or less meaningful and seems to have been used already in one way or another in
published reports and papers. Much more important than the actual word definition, is the
general understanding of the field.

1.5 Smart Structures

The term ”smart structure” is more commonly applied to a super-system where intrinsically
adaptive materials are employed. Smart structures are structures that are made of smart
materials; in other words, they are those, which incorporate actuators and sensors that are
integrated into the structure and have structural functionality, as well as integrated control
 CRITICAL COMPONENT TECHNOLOGIES OF SMART STRUCTURES 5

logic, signal conditioning and power amplification electronics. Such actuating, sensing and
signal processing elements are incorporated into a structure for the purpose of influencing
its states or characteristics, be they mechanical, thermal, optical, chemical, electrical or
magnetic.
For example, a mechanically intelligent structure is capable of altering both mechani-
cal states, namely. its position or velocity, or its mechanical characteristic, in terms of its
stiffness or damping. An optically intelligent structure could, for example, change colour
to match its background. The truly intelligent structural system learns and adapts its be-
haviour in response to the external stimulation, which is provided by the environment in
which it operates. However, there is a wide variety of less sophisticated smart materials
and structures, which exploit the basic sub-disciplines, which define three classes of smart
materials. These include materials with only sensing capabilities, those with only actuation
capabilities, and those with both sensing and actuation capabilities, at a primitive level that
is relative to notions of intelligence.

1.6 Critical Component Technologies of Smart Structures

In the context of intelligent materials there is considerable focus on sensors and actuators,
as well as control capabilities. The current generation of smart materials and structures
incorporate one or more of the following features which are outlined below.

Figure 1.1 Approach for classification of active materials

Sensors, which are either embedded within structural materials or else bonded to the
surface of that material. Alternatively, the sensing function can be performed by a func-
tional material, which, for example, measures the intensity of the stimulus that is associated
with a stress, strain, and electrical, thermal, radioactive, or chemical phenomenon. This
functional material may in some circumstances also serve as a structural material.
Actuators, which are embedded within a structural material or else bonded to the sur-
face of the material. These actuators are typically excited by an external stimulus such as
electricity in order to either changes their geometrical configuration or else change their
stiffness and energy-dissipation properties in a controlled manner. Alternatively, the ac-
6 INTRODUCTION

tuator function can be performed directly by a hybrid material, which serves as both a
structural material as well as a functional material.
Control capabilities, which permit the behaviour of the material to respond to an exter-
nal stimulus, according to a prescribed functional relationship or control algorithm. These
capabilities typically involve one or more microprocessors and data transmission links,
which are based on the utilisation of an automatic control theory.
To obtain a better understanding of the active materials field, it is appropriate to intro-
duce an approach to classify different smart materials. Ideally, the classification should be
collectively exhaustive and mutually exclusive. The most common way of structuring is
by considering the input and output of a material system, as illustrated in Figure 1.1.
The input or stimulus can be a change in temperature or in the magnetic field, for exam-
ple. The material then intrinsically responds with an output, which in turn can be a change
in the length of the material, a change in viscosity, or a change in electrical conductivity,
for example.
Active materials can be divided into two groups. One group comprises the classical
active materials, as viewed by the academic community and is characterized by the type
of response that these materials generate. Upon application of a stimulus, the materials
respond with a change in shape and/or in the length of the material (Figure 1.2).

Figure 1.2 Active material devices used as actuator (top) and sensor (bottom)

Thus, input is always transformed into strain, which can be used to introduce motion or
dynamics into a system. These materials are the most widely used group for the design of
smart structures, where active materials are integrated into a mechanical host structure (for
example, a building or a helicopter rotor blade) with the goal to change the geometrical
dimensions of the structures.
The desired change in geometrical dimensions is mostly time dependant and is often
the steady state of the structure in a dynamic system, where integrated active materials
or devices are constantly agitated to change the characteristics of the host in real time.
 SMART MATERIALS AND MEMS 7

Devices based on materials that respond with a change in length are often referred to as
actuators or solid state actuators, to be more specific.
Conversely, active materials can also be used as sensors where a strain that is applied
on the material is transformed into a signal that allows computation of the strain levels in
the system. The figure below illustrates the basic principles of an actuator/sensor smart
material system. Depending on the stimulus-response-direction, an active materials device
can be used as both actuator and sensor.
The second group consists of materials that respond to stimuli with a change in a key
material property, for example, electrical conductivity or viscosity. While they are equally
important from a scientific point of view, they are less frequently integrated into mechanical
structures, but are rather used to design complex modules, for example, clutches, fasteners,
valves or various switches. Frequently, these materials are used as sensors.
Although materials in this group do not produce strain upon application of an external
stimulus, they are sometimes also referred to as actuator systems. Examples include the
electro- and magnetorheological fluids, which respond with an increase in viscosity upon
application of an external electrical or magnetic field.

1.7 Smart Materials and MEMS

Ultra-small machines are everywhere these days. Tiny mechanical devices which are so
minute that a hundred thousand could sit on a pencil eraser, are responsible for triggering
one’s airbags during an accident, spitting colours out in precise detail on your inkjet printer,
and projecting light in the newest digital theatres.
MEMS123 : Micromachining methods, based on techniques, that are utilized in the man-
ufacture of microelectronic devices, are being used to produce a growing array of microme-
chanical structures, including membranes, beams, valves, gears and so on. The marriage
of electrical and mechanical functions on a single chip has led to the development of ”mi-
croelectromechanical systems” (MEMS), which have over the last decade become well
embedded in the high-tech landscape. The MEMS is a small machine, which operates on
a small circuit.
NEMS: Now engineers and physicists are taking the next step in machine miniaturiza-
tion, building mechanical devices on the nanometer scale (a billionth of a meter). If the
researchers succeed, their work could lead to ultra sensitive sensors that can detect even
the most subtle genetic alterations, which are responsible for a disease, or to ultra strong
artificial muscles that might replace damaged human tissue or power tiny robots.
This next frontier in mechanization is called nano-electromechanical systems (or NEM-
S). With MEMS, one could make a mirror, and though it is still a mirror, but smaller. With
NEMS the whole interaction of matter with light is different. You get completely new
physical properties, which is a big opportunity for new devices.

1 Bohua Sun, MEMS Technology, CPUT, Cape Town, 2010

2 Bohua Sun and Bo Zhang, Analysis of MOEMES gyroscope, Journal of Astronautics, Vol30, No.5, 2009,
1925-1928.
3 Bohua Sun and Lin Wang, A novel MEMS single mass 3-axis accelerator and single mass 3-axis gyro, Journal

of Mechanics and MEMS, 3(1) 2011. pp.11-23.

CHAPTER 2

PIEZOELECTRIC MATERIALS

2.1 General Concepts

There are several material systems that exhibit an electromechanical coupling that results
in a displacement of the material upon application of an electric field. The two main classes
are piezoelectric materials (PZT). Both are typically ceramics, with the exception of some
polymers (like PVDF) that exhibit piezoelectric behavior. The conceptual difference be-
tween piezoceramics and electrostrictors is their response upon reversing the electric field.
Piezoceramics can be elongated and compressed, while electrostrictors only exhibit an e-
longation, independent of the direction of the applied electric field. Different underlying
physical principles must thus govern these behaviors. The general idea of electromechani-
cal coupling in a piezoceramic is illustrated Figure 2.1 below.
Piezoelectricity, discovered in Rochelle salt in 1880 by brothers Jaques and Pierre Curie,
is defined as a change in electric (not to be confused with electronic) polarization; with a
change in applied stress, which is usually referred to as the direct piezoelectric effect.
The converse piezoelectric effect is an analogy of the change in strain for a free crystal
(or stress for a clamped system) with a change in the applied field. Thus, the converse
piezoelectric effect is typically used when a material system is used an an actuator: input
is a voltage gradient, while output is a strain. For low fields there is a linear relationship
between the strain and electrical fields. Reversing the field also reverses the direction of
the strain.

Smart Materials and Structures. 9

By Bohua Sun Copyright ⃝ c 2017 Bohua Sun
10 PIEZOELECTRIC MATERIALS

Figure 2.1 Piezoelectric materials

Historically, Rochelle salt and quartz are the most frequently used piezoelectric materi-
als. However, only relatively new material systems such as PZT offer properties that enable
the development of active structural devices.
The microscopic origin of the piezoelectric effect is the displacement of ionic charges
within a crystal structure. In the absence of the external strain, the charge distribution
within the crystal is symmetric and the net electric dipole moment is zero. However, when
an external stress is applied, the charges are displaced and the charge distribution is no
longer symmetric.
A net polarization develops and results in an internal electric field. A material can be
only piezoelectric if the unit cell has no center of inversion. Virtually all piezoelectric
materials crystallize in the perovskite structure.
It should also be furthermore noted that all piezoceramics (like all materials) are also
electrostrictive. However, the piezoeffect usually dominates and the electrostrictive contri-
bution is virtually always ignored in piezoceramics.
Each unit cell within a material has a net polarization vector. A region of equally ori-
ented polarization vectors within a material is called a domain (in analogy to magnetic
domains in ferromagnetic materials. Piezoelectrics are always ferroelectric, which in turn
owes much of its terminology to magnetism. Ferroelectrics have, for example, a Curie
point, above which a material looses both its ferroelectric and piezoelectric material. Typ-
ical Curie temperatures for piezoceramics are between 2000 C and 3000 C).
Ideally, polarization vectors of all unit cells would add up and result in the total po-
larization, forming one large domain within the crystal. However, commercially available
piezoceramics are always polycrystalline. The sum of all polarization vectors of all u-
nit cells (and of all polarization vectors of all domains) results in an electrically neutral
sample, as vectors cancel each other out owning to a random statistical distribution of di-
rections. The trick is then to pole the ceramic at high electric fields to force the domains
to rotate and switch into the desired direction. This is represented in Figure 2.2 and 2.3
below, respectively.
The result is never a full orientation off all domains, nevertheless, the polycrystalline ce-
ramic exhibits significant net polarization that can be used for actuation or sensing purpos-
es. The simple reason that singly crystal piezoceramics are not used is that so far nobody
 PIEZO TRANSDUCERS 11

Figure 2.2 Representation of domain rotation and switching during poling of a polycrystalline
ceramic

Figure 2.3 Polarization and electrostrictive

was able to grow crystals of sufficient size at reasonable costs (Single crystal piezoce-
ramics, as well as lead-free systems - for environmental reasons - are currently an area of
significant research activities).
After poling, linear constitutive relationships describe material behaviour as follows:
( ) [ ]( ) ( ) [ ]( )
D ϵs e E D eT d E
= , = , (2.1)
σ −et c E
ε ε dt s E
σ

where D is electric displacement, T is stress, E is electric field and S is strain, while dij
is piezoelectric constant.

2.2 Piezo Transducers

Transducers convert one form of energy into another. Piezo motors (actuators) convert
electrical energy into mechanical energy, and piezo generators (sensors) convert mechan-
ical energy into electrical energy. In most cases the same element can be used to perform
either task.
12 PIEZOELECTRIC MATERIALS

Single sheets can be energized to produce motion in the thickness, length, and width
directions. They may be stretched or compressed to generate electrical output.
Thin 2-layer elements are the most versatile configuration of all. They may be used
like single sheets (made up of 2 layers), and can be used to bend or to extend. ”Benders”
achieve large deflections, which are relative to other piezo transducers.
Multilayered piezo stacks can deliver and support high force loads with minimal com-
pliance, but they deliver small motions.

2.3 PZT Application

Piezoelectric actuators have been used for active shape, vibration and acoustic control of
structures because of their adaptability and light weight. Their ability to be easily integrat-
ed into structures makes them very attractive in structural control since all the moving parts
encountered with conventional actuators are eliminated. Structural control is achieved by
simply embedding PZT actuators in the structure or bonding them on the surface of the
structure (Figures 2.4 and 2.5).
Active Fiber Composites (AFCs) (Figure 2.6)), which are developed by MIT, are an
innovative combination of active and passive materials that create a new hybrid material,
capable of meeting the increasingly high demands of current and upcoming applications
in the aerospace industry. Its potential applications include the following: Vibration Sup-
pression - Rotorcraft and Airplanes increased structure life, increased passenger comfort,
improved flight performance (better fuel efficiency, higher payloads, increased cruising
speeds), Acoustic Control- Helicopter/Aircraft Fuselages and Submarine’s decreased cab-
in noise, minimizing acoustic signatures in aircrafts, submarines and torpedos, and airplane
wing de-icing micro-air vehicles.
Another proposed application for piezoelectrics is in flow-induced vibration reductions.
An application for the vibration control of a cantilever plate by using the digital PID. The
plate composed of two PVDF (Polyvinylidene Fluoride) piezoelectric films and a plastic
base plate that were placed parallel to the flow in a low turbulent wind tunnel.
An integrated distributed actuator design methodology is presented based on the con-
verse piezoelectric effect, aimed at actively controlling the in-vacuo flapping and lagging
eigen-vibration characteristics of rotating blades carrying a tip mass. The helicopter blade
is modelled as a thin/thick walled closed cross-section untwisted cantilevered beam that
rotates with constant angular velocity.
State-of-the-art research in the application of PZT materials indicates that numerous
issues remain unanswered about the application of piezoelectrics and their limitations. Re-
search on modelling is required to predict the behavior. On the materials engineering
side investigation is required to enhance the mechanical properties. Previous research that
was conducted on PZT-actuated beam and plate structures led to models describing their
response. However, much less research has been performed on structures, which have
curvatures, while further research in this area is required.
 PZT APPLICATION 13

Figure 2.4 Piezo transducers (a)

14 PIEZOELECTRIC MATERIALS

Figure 2.5 Piezo transducers (b)

 PZT APPLICATION 15

Figure 2.6 Active Fibre Composite (Piezoelectric fibre) from MIT

CHAPTER 3

SHAPE MEMORY MATERIALS

3.1 General Concepts

The most common group of materials that respond to a temperature change with a shape
change or elongation is the Shape Memory Alloys (SMAs)(Figure 3.1).

Figure 3.1 Shape memory materials performance

Smart Materials and Structures. 17

By Bohua Sun Copyright ⃝ c 2017 Bohua Sun
18 SHAPE MEMORY MATERIALS

These alloys, which are mainly used for commercial purpose, mostly use Titanium-
Nickel alloys that undergo a phase transformation upon temperature change. This austenitic
to martensite transformation can directly lead to a volume and change the shape of the sam-
ple. Furthermore, if properly designed, the material (for example, in wire or spring form)
can be trained to transform fully reversibly in an exact reproducible manner, thus repre-
senting a micro-motor that can be driven by temperature increase and decrease.

Figure 3.2 Illustration of the Shape Memory Effect: Original single crystal lattice structure in (a);
upon cooling down to the martensite finish temperature (Mf) self-accommodated martensite forms
without significant change in the external dimensions (b); upon mechanical deformation (c and d) the
system minimizes energy through twinning (lattice invariant shear), while maintaining atomic bonds;
and heating above the austenite finish temperature reverts variants to the parent phase in the original
orientation (e)

Shape Memory Alloys can be operated in a one-way mechanism, for example, upon
heating above a critical temperature (the austenite finish temperature) the shape memory
system adapts to the desired shape that it had originally.
However, it is also possible to ”train” the material to operate fully reversible, which
is known as the two-way shape memory effect. Here the twinning (see ”c” and ”d” in
Figure 3.2) during each cycle takes place in exactly the same manner, resulting in a nearly
identical microstructure after each deformation. Generally, the one-way shape memory
effect leads to higher strains than the two-way effect.
Many metals exhibit a phase change as they are heated and cooled. We can illustrate
this by using a crude ‘stick and ball’model of the metallic lattice.
The higher temperature austenitic structure has the characteristic stress-strain curve of
most metals. The lower temperature martensitic structure has a stress-strain curve more
like that of an elastomer in which there is a ‘plateau’stress. All the deformation up
to about 8/100 is ’elastic’or, in other words, it can be recovered –but not by simply
relaxing the stress whilst it in the martensitic condition (Figure 3.3)).
 GENERAL CONCEPTS 19

Figure 3.3 Stress-strain curves of SMA at phase of austenitic and martensite

There are numerous alloys, which exhibit this memory. The most useful for fluid fit-
ting applications is the Titanium-Nickel family, known by the Raychem Corporation Tinel
name and particularly the Titanium-Nickel-Iron Alloy A and higher strength A-HS, which
exert very high forces on recovery and have transformation temperatures in the region of
−100C to −150C .

3.1.1 Ferromagnetic Shape Memory Alloys

Ferromagnetic Shape Memory Alloys (FSMA) (Figure 3.4)) are a recently discovered class
of actuator material, whose salient features are magnetically driven actuation (field inten-
sity varies, about 3KG and larger) and large strains (around 6/100). FSMA are still in the
development phase; so far only alloys in the Ni-Mn-Ga ternary have been proven to show
satisfactory performance, while other systems have shown interesting potential.

Figure 3.4 Ferromagnetic Shape Memory Alloys.

20 SHAPE MEMORY MATERIALS

As the name suggests, FSMAs are ferromagnetic alloys, which also support the shape
memory effect. This means that they undergo the characteristic martensitic transformation
upon cooling, and show all features of conventional shape memory alloys. However, the
actuation mechanism in these materials is radically different to ordinary shape memory
alloys. Rather than being heat driven, they are magnetically driven. This difference allows
for increased frequency response (fast actuation).
FSMAs operate solely in the martensitic phase and the martensitic phase transforma-
tion is immaterial for anything other than setting an upper operation temperature (if the
temperature rises above the austenite start temperature. Then the material transforms to
austenite (the high temperature phase) and cannot be magnetically driven.) The deforma-
tion mechanism relies on two facts, namely, that the martensitic phase of an FSMA has
twinning as its preferential deformation mechanism, and that the martensitic phase has a
high magneto-crystalline anisotropy, that is, the magnetization is strongly pinned to spe-
cific crystallographic orientations. In other words, there are axes in the crystal where the
material is easily magnetized (easy directions), and increasingly higher field-intensities are
required to align the magnetization with the field on directions away from these easy axes.
Unlike slip (dislocations), twinning is a planar defect. The twinning process is a pure
shear of the crystal which transforms its crystallographic structure into a new one. For the
materials of interest, the twinned structure is mirror symmetric to the original one (type
I twins, mixed type twins). The structures that are related by a twinning shear are called
twin variants. Because each is a sheared version of the other, twin variants have different
magnetic easy axes.
In a situation where two twin variants exist in a single crystal, a magnetic field will tend
to align the magnetic moments. But because of the high magneto-crystalline anisotropy,
the magnetic moments will not readily away from their easy axes. Instead, it is possible
that one crystal twin shears into the other variant so that there the magnetic easy axes is
better aligned with the field. This is a magnetically driven twin boundary motion, which is
the mechanism of operation of the FSMA.

3.2 SMA Application

The coupling of using shape memory alloys has been the best application for single mate-
rials so far (Figures 3.5, and 3.6).
Couplings are machined from Tinel bar-stock to a size where their inside diameter is
slightly smaller than the outside diameter of the tubes that they are intended to join. The
couplings are then immersed in liquid nitrogen so that the metal becomes martensitic.
They are then mechanically expanded by driving a tapered steel mandrel through the
bore, using a hydraulic press. As long as they are kept cold they maintain this expanded
shape in which the bore is now large enough to accommodate the tube outside of the
diameter.
It is in this condition that they are delivered to the user. If they are allowed to warm up
they will undergo ’free recovery’, that is, they will shrink back to the as-machined size and
again will be too small to slip over the tube.
Conversely, constrained recovery is what happens when we place something in the way
of the recovery back to the original austenitic shape.
In the case of the couplings, it is the tube, which prevents complete recovery.
 SMA APPLICATION 21

Figure 3.5 Pipe coupling using SMA

Figure 3.6 Mechanism of pipe coupling using SMA

The Tinel alloy from which the couplings are machined recovers with tremendous force,
and if a tube is interposed, it will be swaged as the coupling tries to get back to its original
shape.
The tubing resists this force and the two come into equilibrium with about 2/100 strain
remaining in the coupling, but this is a dynamic equilibrium; if the tube grows or shrinks
the coupling follows as, it is always trying to get back to its as-machined dimensions.
22 SHAPE MEMORY MATERIALS

Unlike mechanically swaged couplings, whose pressure on the tube relaxes as soon the
swage head is removed, this constrained recovery or ‘live crimp’provides a dramatic
benefit in terms of the reliability of the coupling. Even in high vibration conditions, or
with the coupling slipped or rotated about the tube, the seal is still good.
CHAPTER 4

MAGNETOSTRICTIVE MATERIALS

4.1 General Concepts

Magnetostriction is the changing of a material’s physical dimensions in response to chang-

ing its magnetization. In other words, a magnetostrictive material will change shape when
it is subjected to a magnetic field. Most ferromagnetic materials exhibit some measurable
magnetostriction.
The highest room temperature magnetostriction of a pure element is that of Co which
saturates at 60 microstrain. Fortunately, by alloying elements one can achieve ”giant”
magnetostriction under relatively small fields. The highest known magnetostriction are
those cubic laves phase iron alloys that contains the rare earth elements Dysprosium, Dy,
or Terbium, Tb; DyFe2 , and TbFe2 . However, these materials have tremendous magnetic
anisotropy, which necessitates a large magnetic field to drive the magnetostriction. Not-
ing that these materials have anisotropies in opposite directions, Clark and his co-workers
at NSWC-Carderock prepared alloys that contains Fe, Dy, and Tb. These alloys are gen-
erally stochiometric, of the form Tbx Dy1−x Fe2 and have been coined Terfenol-D. When
Terfenol-D is operated under a mechanical-bias, it strains to about 2000 microstrain at
room temperature. For typical transducer and actuator applications, Terfenol-D is the most
commonly used engineering magnetostrictive material.
The mechanism of magnetostriction at an atomic level is relatively complex subject
matter, but on a macroscopic level, may be segregated into two distinct processes. The
first process is dominated by the migration of domain walls within the material in response
to external magnetic fields, while the second involves rotation of the domains. These two

Smart Materials and Structures. 23

By Bohua Sun Copyright ⃝ c 2017 Bohua Sun
24 MAGNETOSTRICTIVE MATERIALS

Figure 4.1 Strain versus magnetic field.

mechanisms allow the material to change the domain orientation, which in turn causes a
dimensional change. Since the deformation is isochoric there is an opposite dimensional
change in the orthogonal direction. Although there may be many mechanisms involved
in the reorientation of the domains, the basic idea, represented in the figure, remains that
the rotation and movement of magnetic domains cause a physical length change in the
material.
Magnetostrictive materials (MS) are typically mechanically biased in normal opera-
tions. A compressive load is applied to the material, which owning to the magneto-elastic
coupling, forces the domain structure to orient perpendicular to the applied force. Then, as
a magnetic field is introduced, the domain structure rotates, producing the maximum pos-
sible strain in the material. A tensile preload should orient the domain structure parallel
to the applied force, though this has not yet been observed owning to the brittleness of the
material in tension.
The first positive identification of a magnetostrictive effect was in 1842, when James
Joule observed that a sample of nickel changed in length when it was magnetised. Subse-
quently, cobalt, iron and alloys of these materials were found to show a significant mag-
netostrictive effect with strains of about 50ppm. The reciprocal effect, where a stress is
applied to a body, is known as the Villari effect. An additional phenomenon, which is
known as the Wiedemann effect is a twisting in a material when a helical magnetic field
is applied. The inverse of this, called the Matteuci effect, is the creation of a helical field
when a magnetostrictive material is subject to a torque.
One of the first practical applications of magnetostriction was its use in SONAR devices
in echo location during the Second World War. Another early application included torque
sensing and these applications are as important today as they were then. The nickel based
materials that are used in these devices had saturation magnetostriction values of 50 parts
per million (ppm). These strains are quite low, and have limited applications.
 BASICS OF MAGNETOSTRICTION 25

In the 1960s the rare-earth elements terbium (Tb) and dysprosium (Dy) were found to
have between 100 and 10,000 times the magnetostrictive strains that are found in nickel
alloys. However, because this property only occurs at low temperatures, applications that
operate at ambient temperature and above were not possible.
What researchers were looking for was a material, which would operate at high temper-
atures, have a large magnetostrictive strain, but would only require a low magnetic field.
They found that the addition of iron to Tb and Dy to form the compounds TbFe2 and
DyFe2 brought the magnetostrictive properties to room temperature. These materials re-
quired large magnetic fields to generate large strains. By alloying the two compounds it was
found that the magnetic field that are required to produce saturated strains were consider-
ably reduced. The resulting alloy Tb.27Dy.73Fe1.95 (commercially known as Terfenol-D)
is presently the most widely used magnetostrictive material. Terfenol is capable of strains
as high as 1500ppm and, since the 1980’s, has been a commercially available material for
application in many different fields.

4.2 Basics of Magnetostriction

In ferromagnetic materials, the existence of local magnetic moments is inextricably related

to the electronic configurations. The wave functions of the electrons in that a multi-atom
system are related to the interatomic distance in a self-consistent manner. They differ
significantly from atomic orbital for short interatomic distances such as those are found in
solids.

Figure 4.2 Thermal expansion of a magnetic material

Compared to the atomic orbitals that electrons would occupy, the energy of the elec-
tronic states when there is a local magnetic moment is higher (lower). This results in an
increased (diminished) interatomic distance compared to non magnetic species, which is
an effect that can generally be described as a strain. This strain is termed magnetostriction.
Magnetostriction is further affected by long-range magnetic ordering, such as that takes
place below the Curie temperature TC in ferromagnetic materials. In as much as the mag-
netic ordering can be modified by applied magnetic fields, the latter will also influence the
observed magnetostriction.
26 MAGNETOSTRICTIVE MATERIALS

A better picture of the different types of magnetostriction can be gained from the various
contributions to the thermal expansion of a magnetic material, as depicted in Figure4.2.
The dotted line (marked ”no local moment”) indicates normal linear thermal expansion for
a solid. The dot-dashed line labeled ”Local moment” is displaced from the first to indicate
the volume expansion that accompanies the formation of a local magnetic moment.
Because the local magnetic moment does not vanish immediately above TC, but merely
loses its long-range ordering, the internal pressure, which is associated with it does not
vanish completely above TC (Figure 4.2, solid line). The solid lines show the form of the
thermal expansion for a ferromagnet above and below TC. Below TC, additional magneto-
volume effects owning to long-range magnetic ordering are turned on; and they may add
to or subtract from the volume expansion owning to the presence of a local moment.
The slope of these solid lines (the thermal expansion coefficient, alpha) can be of ei-
ther sign just below the Curie temperature (a property that is used in invar alloys such as
Fe70Ni30). All of these effects are isotropic, involving the bulk modulus. These isotropic
effects are called volume magnetostriction or, when the magnetic ordering is produced by
an applied field, they are called forced magnetostriction.
Figure 4.2, which is a schematic representation of the thermal expansion of a mag-
netic material as a function of temperature, illustrates the increased volume owning to
the presence of a local magnetic moment and the onset of magnetic anomalies below the
Curie temperature. A small anisotropic strain, depending on the direction of magnetization
(circled inset), is also observed below TC. The latter is usually referred to as anisotropic
magnetostriction.
On a smaller scale the volume expansion can show an anisotropy for T < TC , namely
the linear strain is shown in different directions relative to the direction of magnetization
(Figure 4.2, circled inset). That is, the magnetization vector has associated with it a stress
that is directed along M, which causes a uniaxial mechanical deformation.

4.3 Magnetosrtictive Materials Applications

Figure 4.3 A prototype magnetostrictive actuator

 MAGNETOSRTICTIVE MATERIALS APPLICATIONS 27

Magnetosrtictive devices may be incorporated in multifunctional composites to tailor

mechanical deformations, as well as to sense of these deformations/forces. When dis-
tributed as microscale devices in a host, magnetostrictives can act as distributed sensors in
multifunctional composites. Alternative applications include distributed actuation for com-
posites, which are capable of vibration suppression, micropositioning, damage mitigation
and shape control.
Preliminary results show that simple analog control systems that consist of Terfenol-D
actuators can significantly reduce flexural vibrations in a rotating shaft with application in
machine tools. Passive damping and velocity sensing. which use magnetostrictive trans-
duction, has been discussed.
A prototype magnetostrictive actuator is shown in Figure 4.3.
An integrated model, which analyzes the embedded magnetostrictive mini actuators
(MMA) and the vibration suppression capability of a cantilever beam embedded with M-
MA, was developed. The simulation results indicated that by embedding MMA in the
host structure, the vibration suppression of the host structure can be achieved by using the
magneto-thermoelastic.
CHAPTER 5

ELECTRORHEOLOGICAL FLUIDS

5.1 General Concepts

Electrorheological(ER) materials are suspensions, which experience reversible changes in

rheological properties such as viscosity, plasticity and elasticity when subjected to elec-
tric fields. These reversible changes occur because of the controllable interaction between
micron-sized dielectric particles within the ER suspension. The polarization of these par-
ticles leads to configuration changes, which in turn result in significant changes in rheo-
logical properties. In essence, ER material behaviour transforms from that of a liquid to
characteristics of a solid-like gel when an electric field is applied. ER adaptive structures
are based on controlling the pre-yield rheology of the ER material, which is achieved by
applying different electrical fields to the ER material (Figure 5.1).

5.2 Electrorheological Materials Applications

ER material applications can be categorized into two classes, namely controllable devices
and adaptive structures. ER based controllable devices that have been studied include,
valves, mounts, clutches, brakes and dampers. Adaptive structures are structures that have
tunable properties owning to the incorporation of a controllable component such as an ER
material.
ER fluids have been studied for active control of flexible rotor blades. Results of testing
the increased damping that is supplied by an ER fluid was reported for a cantilevered com-
Smart Materials and Structures. 29
By Bohua Sun Copyright ⃝ c 2017 Bohua Sun
30 ELECTRORHEOLOGICAL FLUIDS

Figure 5.1 Effect of applying electrical field to ER materials.

posite beam with approximately 40 percent volume fraction of ER fluid. The experimental
results tend to suggest that much stronger ER materials need to be developed for structural
applications such as a helicopter fluid-filled flexible beam. Utilization of ER fluids in car
suspension systems were also studied. The suspension system comprises three elements:
Electro-Rheological fluid; Rotary shock absorber; and A control algorithm.
The unique design and operating characteristics of the computer controlled Rotary Ac-
tive Control Damper (RACD), assembled with ER fluid and associated sensors, allow for
continuous adjustment of the damping force applied to the vehicle The required power is
supplied by the existing vehicle is electric systems. Utilization of ER fluids in power trans-
mission systems such as clutches, speed reducer, and brakes is shown in Figure 5.2 below.

Figure 5.2 Utilization of ER fluids in power transmission

CHAPTER 6

MAGNETORHEOLOGICAL FLUIDS

6.1 Basic Concepts

Magnetorheological Fluids(MR) materials are suspensions, which suddenly becomes s-

mart and actuating as soon as a magnetic field passes through it. Magnetic fluids can be
classified into two types: colloidal ferro-fluids and magnetorheological (MR) fluids. Each
of these types of fluid consists of finely divided magnetizable particles that are suspended
in a carrier liquid such as a mineral oil.
Ferro-fluids are composed of small (3 - 15 nm) particles of solid, magnetic, single-
domain particles that are coated with a molecular layer of dispersant. True ferro-fluids
do not settle out even though a slight concentration gradient may be established in the
container.
By comparison, MR fluid comprises of larger (3 - 8 um) particles, usually carbonyl
iron, which is in suspension in the carrier fluid with the aid of a chemical dispersant. The
two types of fluid are substantially different in magnetic properties. MR fluids essentially
solidify in the presence of a magnetic field, while ferro-fluids simply change viscosity.
Ferro-fluids are also about 1000 times as expensive to produce, and for this reason alone,
MR technology is the choice for practical applications.
The basic structure of MR fluids was first described in the late 1940’s and is shown
in Figure 6.1. In the absence of a magnetic field, the particles are randomly distributed
throughout the fluid, and the fluid behaves as a Newtonian liquid with a linear relationship
between shear stress and strain rate. However, in the presence of a magnetic field, the
particles tend to align themselves. In order for relative motion of the fluid to occur, these

Smart Materials and Structures. 31

By Bohua Sun Copyright ⃝ c 2017 Bohua Sun
32 MAGNETORHEOLOGICAL FLUIDS

Figure 6.1 The Basic Structure of MR Fluid

chains of iron particles must be continuously broken and re-linked. Until sufficient force
is applied to break the chains, no motion will occur. This provides the solid feel of the MR
fluid.

Figure 6.2 Strain-stress curve of MR

The characteristic shear stress - strain rate characteristics of a typical MR fluid is shown
in Figure Figure 6.2 as a function of the applied magnetic field. Note that in the absence
of a field, the stress-strain rate relationship is linear, which suggests Newtonian behaviour.
 MR ACTUATORS 33

Many different carrier fluids have been used for formulation of MR fluid, which provides
a range of off-state viscosities from 10 to 10,000 Pa-sec.
When a magnetic field is applied the fluid develops a yield stress, which must be over-
come before any strain rate will occur. This appears as an upward displacement of the stress
level, depending on the strength of the magnetic field. However, once the yield stress level
at a given magnetic field strength is achieved, the fluid exhibits a viscosity, which is nearly
equal to the off-state viscosity.
MR fluids have been developed, which generate yield strengths of up to 100 kPa under
magnetic fields of 3,000 Oersted. Like all magnetically active materials, MR fluids exhibit
a magnetic saturation property in which additional magnetic field strength will generate
little or no further yield stress. For high performance MR fluid this occurs at about 4,000
to 5,000 Oersted.
Some limited measurements were made on the speed of the MR fluid by measuring the
dynamic change in shear stress as a function of magnetic field. These measurements are
significantly complicated by the difficulty in directly measuring magnetic field strength.
From these early experiments, the response of the yield strength to the applied field was
found to be approximately exponentially asymptotic, in nature with a 90/100 rise time of
less than 10 ms. For the purpose of real-time suspension control, this was determined to
be sufficiently fast.

6.2 MR Actuators

MR technology and several classes of MR actuators (axial and rotary dampers and brakes)
were originally described in the late 1940s. The original fluid formulations suffered from
instability and the actuators were bulky and had poor durability. This resulted in a hiatus
in the interest of MR technology that lasted about 40 years.
During this hiatus, significant advances were made in a number of collateral technolo-
gies that had the potential to relieve the technical shortfalls of the original work on MR. In
addition, the late 1980s brought a significant interest in electronically controlled suspen-
sions, particularly these that focused on passenger car primary suspensions. MR technolo-
gy is ideally suited for these electronically controlled suspension systems.
The development of MR based actuators follows the principle of operation of the fluid
itself: imposition of a magnetic field on the fluid allows control of the force, which is
required to move the fluid. A schematic diagram of an axial shock absorber that utilize one
manifestation of this principle, as shown in Figure 6.3. Note that this is one of a number
of effective configurations for an MR actuator.
The actuator shown in Figure 6.3 is a shock absorber of the pressurised monotube type.
This design is probably the simplest shock absorber type, and incorporates an accumulator
of high pressure nitrogen, separated from the hydraulic (MR) fluid by a floating piston. The
purpose of the accumulator is to accommodate the change in the internal volume, which is
caused by excursions of the piston rod. The electromagnet in the actuator is contained in
the piston itself and controls the magnetic field of the MR fluid only as it passes through
the orifice. This innovation dramatically reduces the power requirement of the device.
Figure 6.4 shows the typical force-velocity characteristics of an MR actuator. The re-
semblance of these characteristics to those of the fluid should be obvious. An important
characteristic about MR actuators (and all other dampers) is that they can only produce
force in a direction opposite to the direction of the velocity. This means that they can only
absorb mechanical energy from a suspension, but cannot add energy to the system.
34 MAGNETORHEOLOGICAL FLUIDS

Figure 6.3 Shock absorber using MR fluid.

MR actuators can be designed to produce virtually any force characteristic within the
design constraints of physical size. MR Technologies, Inc. has conducted preliminary
design exercises on actuators with force levels ranging from a few pounds to over 20,000
pounds.

6.3 Design of MR Systems

MR axial dampers offer advantages in vibration and shock isolation, which are not avail-
able through the use of passive shock absorbers or with fully active systems. With passive
dampers, vibration and shock isolation capabilities are dependent on the damping charac-
teristics, spring rates, and stroke length that are selected during the system design process,
and are not adjustable in use. Fully active systems, conversely, require large power sup-
plies to be effective over similar force ranges and often have limited frequency response,
stroke length, and excessive computational and power requirements.
Passive suspension system design involves strict engineering trade-offs. The designer is
often faced with mutually exclusive performance goals. For example, outstanding vibration
isolation can be achieved in a passive suspension system, which has a low natural frequency
and little damping. This system behaves like a soft, lightly-damped spring. However,
good shock attenuation is achieved by a system that has a relative stiff, highly-damped
system. The soft, lightly-damped system will perform poorly in the presence of large
shock loads by offering almost no resistance to motion and large load excursions. The
characteristics that make it good at vibration isolation (soft, little damping) is precisely
what makes it bad at shock attenuation. Similarly, a stiff, highly damped system will
perform poorly, as a vibration isolator offers almost no significant vibration attenuation.
Again, the characteristics that make the system good at shock attenuation (stiff, highly-
damped), is precisely what makes it bad at vibration isolation.
 DESIGN OF MR SYSTEMS 35

Figure 6.4 Force-velocity curve

Digitally controlled MR actuators have the ability to combine previously contradictory

shock and vibration attenuation design goals into a single, functioning shock and vibra-
tion isolation system. The key characteristics that allow use of MR in combined-mode
applications, and are as follows:
1) Low off-state force levels, which allow the actuator to exert little resistance to motion
in the presence of steady state or random vibration about an equilibrium point;
2) High on-state force levels. This allow the actuator to exert the necessary resistive
forces in the presence of high shock loads. Referring to force/velocity curves in Figure 6.4
these forces can be generated at low velocities; and
3) Fast response time. MR actuators can switch between any two conditions within its
performance envelope in milliseconds. How the actuator responds and to what external
stimuli, is are a function of how the digital controller is programmed and the nature of
related sensors and signal conditioning equipment.
Major components of a complete magneto-rheological SVI system will consist of a soft
spring to support the payload (low natural frequency), an MR actuator with low off-state
force level (lightly damped) and high enough force generation capability to control the
payload without excessive excursions under the severest anticipated shock loads, together
with an electronic controller, algorithm, and related electronics. More advanced MR ac-
tuator designs incorporate a gas spring into the actuator itself, minimizing parts count and
allowing dynamic ”tuning” of the spring rate.
CHAPTER 7

FIBRE OPTIC SENSORS

Smart materials and structures, at the most sophisticated level in the foreseeable future, will
feature a network of embedded sensors and datelines. These sensory and data-transmission
systems must typically be small, lightweight, possess geometrical flexibility, operate with
low power consumption, feature a wide bandwidth, and be able to withstand the manufac-
turing environment imposed on the host structural material without suffering performance
degradation. Although previous materials can also be used as sensors, fibre-optic sensors
have more features.

7.1 The Structure of Optical Fibres

Optical fibres are dielectric wave-guiding devices that are used to confine and guide light.
The majority of optical fibres that are used in sensing applications have silica glass cores,
but some special materials such as sapphire, fluoride glasses and Neodynium doped silica
are utilised for specialised applications.
The core is surrounded by a dielectric material called the cladding. The cladding’s
refractive index must be lower than the core to satisfy Snell’s Law for total internal
reflection, and thus propagation of the light along the fibre core. The cladding is generally
made from silica glass, although for some applications a plastic or doped silica cladding
is used. A barrier layer of plastic (the “buffer”) is used to jacket the fibre; this provides
the fibre with mechanical strength and protects it from damage or moisture absorption. In
Smart Materials and Structures. 37
By Bohua Sun Copyright ⃝ c 2017 Bohua Sun
38 FIBRE OPTIC SENSORS

Figure 7.1 Cross-sectional view of typical optical fibre

some sensing application a specialised buffer is used to enhance the fibre’s measurement
sensitivity. Figure 7.1 illustrates a cross-sectional view of a standard optical fibre.
A simplistic definition of the modes that are guided by an optical fibre can be explained
as the locus of all the light rays, which are launched at different entry angles into the core
of the fibre. For any optical fibre the number of modes that will be guided by the fibre is
dependent on core size, the ratio of core/cladding refractive indices, and the wavelength of
operation. Depending on the optical property that is utilised for the sensing application,
the number of modes that are propagated by an optical fibre is an important parameter.

7.2 Characteristics, Advantages and Capabilities of Fibre Optic Sensors

The high expectations of optical fibres as information carriers in communication systems

have been justified by their performance over the past two decades. Due to their high band-
width, low attenuation and mechanical properties, each fibre is capable of replacing 1000
copper wires in telecommunication systems. With these characteristics, it is no surprise
that optical fibres have become the most affordable and efficient medium available in the
field of telecommunications.
But optical fibres can be more than mere signal carriers. Light that is launched into
and confined to the fibre core propagates along the length of the fibre, unperturbed, unless
acted upon by an external influence. Any disturbance of the fibre alters the characteristics
of the guided light; such alterations can be monitored, and related to the magnitude of the
disturbing influence. The characteristics of the light which may be monitored in sensing
applications include; amplitude; polarisation; phase; modal distribution; wavelength and
time-of-flight. Such modulation of the light makes possible the measurement of a wide
range of events and conditions.
Fibre optic sensors may be intrinsic or extrinsic, depending on whether the fibre is
the sensing element or the information carrier, respectively. They are designated point
sensors when the sensing gauge length is localised to discrete regions. If the sensor
 CAPABILITIES AND SIGNIFICANCE 39

is capable of sensing a mean field continuously over its entire length, it is known as a
“distributed”sensor; “quasi-distributed”sensors utilise point sensors at various loca-
tions along the fibre length. Fibre optic sensors can be transmissive or can be used in a
reflective configuration by mirroring the fibre end-face.

7.3 Capabilities and Significance

Fibre optic sensors and systems have proven their advantages and capabilities in various
applications and environments. Now, fibre optic sensors offer great potential for the con-
dition monitoring applications. Their significance to condition monitoring applications
stems from the fact that: they have many operational advantages over existing convention-
al sensors; they are made from a durable material (silica) that is corrosion resistant and can
withstand high tensile loading (can withstand up to 5/100 elongation - that is 50,000 μe);
they can measure temperature from −2000 C to 8000 C for silica, 20000 C for sapphire
fibres, with better than 0.10 C resolution; they are capable of having an extremely wide
dynamic range (DC to MHz) with a uniform response characteristic; they can be applied
to complex surfaces and difficult to reach areas (around the circumference of a round ob-
ject, around sharp corners or across welds); point to long sensing lengths are possible (¡1
mm to kms); they offer immunity to EMI (can operate in electrically noisy environments
and have no EMI noise pick-up over very long leads) and electrical isolation (they are non-
conductive); they provide high spatial resolution (0.1µe easily achieved, 1 pe demonstrated
- that is one millionth of a µe); they can be used at any pressure, deep under the sea or in
space, without adjustment; they can monitor the composition of gases or liquids with an
accuracy approaching parts per million; they may be incorporated within many materials,
structures, machines or components without significant modification of the environment
or the fibre’s own dielectric nature; they can perform the sensing in-situ at the location
where a problem is expected or occurring (they can be located much closer to the problem
source); they are capable of monitoring different parameters simultaneously, thus reducing
system cost and complexity; they provide the ability for on-line analysis and the opportu-
nity for feedback control; and they are already cost-competitive against many conventional
sensors.

7.4 Intensiometric Fibre Optic Sensors

Intensiometric fibre optic sensors are based on the modulation of light intensity in the fibre
and are generally configured as a distributed sensor. The variation in light intensity can oc-
cur in the form of fracture loss, time-of-flight, refractive index, amplitude and wavelength.
Fracture-loss techniques offer simple and effective threshold sensors, but they are not
useful for quantitative analysis. Absorption-based techniques, conversely, have great po-
tential for chemical analysis applications. Absorption techniques involve the monitoring
of absorption of broadband light propagating through an absorbing medium. A spectrom-
eter can be utilised to determine the absorption spectra of the received light, and hence
the chemical species or concentrations present (they essentially represent chemical finger-
prints).
Coupling optical fibres with a spectrometer permits remote, real-time monitoring. The
optical air-path and sensing cell in the spectrometer may be replaced by the optical fibres,
provided that the fibres have sufficient optical range in the infrared. The fibre optic sen-
sor may then be configured extrinsically, delivering the light to/from a test chamber, or
40 FIBRE OPTIC SENSORS

Figure 7.2 Fibre optic evanescent wave absorption technique.

intrinsically by configuring the optical fibre in such a way that the light, which passes a-
long the fibre is not wholly constrained by the fibre, but part of it (the evanescent wave)
propagates a short distance beyond the bounding surface of the glass core. This evanescent
wave is modified by absorption in the contiguous medium, and can thus be used to convey
analytical information. Figure 7.2 illustrates the fibre optic evanescent wave absorption
technique.

7.5 Interferometric fibre optic sensors

Interferometric fibre optic sensors are a large class of extremely sensitive fibre optic sen-
sors. Fibre optic interferometers are analogous to their respective classic bulk optic inter-
ferometers. They are typically used when ultra-high sensitivities are required, and/or in
applications of localised measurements, although sensor lengths longer than 1 meter are
sometimes possible. This sensing technique is based primarily on detecting the optical
phase change that is induced in the light, as it propagates along the optical fibre. Fibre
optic interferometers are generally intrinsic sensors in which light from a source is equally
divided to follow two (or more) fibre-guided paths. The beams are then recombined to
mix coherently and form a “fringe pattern”, which is directly related to the optical phase
difference experienced between the different optical beams. Single mode fibre and associ-
ated components are used because they maintain the spatial coherence of the light beam,
whereas multi-mode fibres do not. The most common configurations of the interferometers
are the Mach-Zehnder, Michelson and Fabry-Perot fibre optic sensors.

7.6 Bragg (Grating) Fibre Optic Sensor

The Bragg grating fibre optic sensor is a relatively new type of fibre optic sensor. It is
generally classed as an interferometer. With this technique a stable Bragg grating is per-
 BRAGG (GRATING) FIBRE OPTIC SENSOR 41

Figure 7.3 Response characteristics of fibre optic Bragg grating

manently impressed (photo-etched) into the core of a germanosilicate fibre by exposure to

a two-beam ultraviolet interference pattern. The region of periodic variation in the index
of refraction of the fibre core then acts as a narrow-band reflection filter. The reflection
signal is, therefore, a narrow spike with a centre wave-length that is linearly dependent on
the back reflected Bragg wave-length and the mean index of refraction of the core. Figure
7.3 illustrates the working principles of a fibre optic Bragg grating.
Consequently, any external parameters which act to alter the grating characteristics re-
sults in a shift in the reflected Bragg wavelength and constitutes the mode of measurement.
A strain resolution of 0.1µe and temperature resolution of 0.1o C is readily achievable.
The Bragg sensor has several advantages over all other fibre optic sensors: it is con-
structed from a single, unweakened fibre; its response is linear to the measurand, and
temperature compensation is possible by overlaying two gratings. All other interferome-
ters have a sinusiodal response characteristic, which requires recalibration on initialisation
(following power interruption). The Bragg sensor, conversely, is initially calibrated (de-
termined from grating characteristics) and any deviations from the Bragg wave-length are
proportionately related to an exact parameter (strain).
Bragg fibre optic sensors could be particularly useful when the Bragg gratings are ar-
ranged along the fibre length so that the gratings are written into the core of the optical
fibre at various Bragg wave-lengths. Each of the reflected wave-length signals from the
corresponding gratings could be monitored by the use of a coupler, detector and tuneable
optical filter, thus achieving single-fibre multiplexing of the sensors.
42 FIBRE OPTIC SENSORS

7.7 Polarimetric Fibre Optic Sensors

Polarimetric fibre optic sensors are an attractive alternative to interferometric sensors when
ultra-high sensitivity is not required, and longer sensor gauge lengths are desired. The prin-
ciple of polarisation sensing is related to birefringence within the fibre. A polarised light
beam is launched at to the principle axis of a single-mode fibre, thus the two orthogonal
polarisation eigenmodes are equally excited and degenerate. When an external force acts
on the fibre, the degeneracy is lifted and a phase difference is induced in the two eigen-
modes. Consequently, a measurement of a change of the state of polarisation in the fibre is
directly related to the measurand. The polarimetric fibre optic sensor is sometimes referred
to as a differential interferometer.
The polarimetric fibre optic sensor is capable of detecting most parameters previously
listed with the advantaged of being configured as a point sensor or distributed. Drawbacks
of this sensor include the high cost of components and the complexity of the system.

7.8 Modalmetric Fibre Optic Sensors

Optical fibres were found to be microphonic quite early in the development of fibre optic
sensors. Research reveals this sensitivity to be based on the modulation in the distribution
of modal energy in the optical fibre. This finding resulted in the first truly intrinsic fibre
optic sensor; the modalmetric fibre optic sensor. Although this type of sensor is sensitive,
the modulation of the modal pattern is generally nonlinearly related to all disturbances,
which resultes in deep fading and drifting of the output signal. This behaviour limits the
use of the sensor for quantitative strain measurements, but nonetheless it can be used as a
threshold-type sensor.
More complex techniques overcome the random behaviour by involving modal interfer-
ometry in a single, circular or elliptic core, two-mode optical fibre. Modalmetric sensors
are capable of sensing most parameters, however, their sensitivities are generally far low-
er than interferometric sensors and localisation of the sensing region is difficult. Usually,
these sensors are employed in conjunction with other techniques, such as micro-bending,
polarimetry, and optical time domain reflectometry (OTDR).
CHAPTER 8

DISTRIBUTED PIEZOELECTRIC
ACTUATING SMART STRUCTURES

The following presentation provides an overview of smart structures by using distributed

actuator systems based on piezo-ceramics.
Discussion of the historical developments is limited to piezoelectric actuators, as these
have mostly been used for distributed actuator systems during the past decades.
Initial developments of the materials are provided next. While several specific applica-
tions were realized in the 1930s-1950s using piezoelectric materials (for example, phono-
graph pickups) it was not until the 1980s that piezoelectric actuators were used as structural
strain actuators to enable the development of what is presently known as smart or actively
controlled structures.
Traditionally, piezoelectric actuators were designed as discrete actuators. Like a small
linear motor, they were used to transduce strain or to force to a very small, geometrically
well defined area. Mechanical energy was exchanged with the system only in this small
contact area.
Distributed actuators (often called strain actuators) , conversely, are characterized by
a relatively large contact area. The difference is schematically illustrated in the figure
below. Distributed actuators are mostly thin wafers or plies, and therefore, allow smooth
integration onto or beneath the surfaces of mechanical host structures(Figure 8.1).
The following presentation illustrates some of the developments and challenges with
regards to the structures level. The vast majority of research on the materials and discrete
actuator level occurred during the 20th century with a peak level of research activity during
and in the post World War II area (Figure 8.2).

Smart Materials and Structures. 43

By Bohua Sun Copyright ⃝ c 2017 Bohua Sun
44 DISTRIBUTED PIEZOELECTRIC ACTUATING SMART STRUCTURES

Figure 8.1 Schematic representation of a discrete actuator a) and a distributed actuator b).
Distributed actuator systems are often based on multiple individual strain actuators (modularity);
note that the actual contact area is very thin and can be typically a mechanical connection in a) and
an adhesive layer in b)

A lot of the research activities were driven by the need for underwater transducers and
sonar for marine and navy applications. Research that deal with distributed actuator sys-
tems began in the early 1980s.
Early1980s (Figure 8.2):
First Developments of Distributed Actuator Systems in the early to mid-1980s: develop-
ment of the theory and first small beam experiments. Theoretical considerations combined
with first lab-scale experiments of actively controlled structures date back to early-to-mid
1980s. Most of the work was structured on both a modelling and an experimental part
of simple beams, which resulted in different levels of agreement between measured and
predicted vibration reduction results.
Since control experiments of vibration reduction require high band-width active ma-
terials, most of the studies that were performed focused on piezoelectric strain actuators.
Both ceramic (Lead-Zirconate-Titanate, PZT) and polymeric piezoelectric (Polyvinylidene
Fluoride, PVDF) materials were used.
One of the most critical issues that were identified in these studies is the correct spatial
placement of sensors and actuators. In general, distributed actuators are placed in posi-
tions, which are characterized by high modal strain energy.The importance of a structured
systems approach for the design of controlled structures is crucial in order to avoid unnec-
essarily complicated and inappropriate choices of measurement variables.
In the middle 1980s (Figure 8.3): depicts a prototype one-dimensional structure using
piezoceramic wafers for active control of a beam (after de Luis). The plot shows sample
open and closed loop experiments for vibration control.
 45

Figure 8.2 Development of distributed actuator system in early 1980s

One of the most critical issues that was identified in these studies is the correct spatial
placement of sensors and actuators.
The late 1980s (Figure 8.4): shows a rototype of two-dimensional active structure using
piezoceramic wafers for active control of an plate-like airfoil (after Lazarus):
Based on the early work on actively controlled beams, the research in the late 1980s
shifted towards the development of more complex structures that were a transition step
towards the integration of active elements into intelligent structures.
Crawley and Lazarus 1 developed a two-dimensional plate-like lifting surface (span of
29.2 cm, chord of 15 cm) covered on both sides to approximately 70/100 with surface
bonded piezoelectric wafers. The design as a thin, low-aspect-ratio lifting surface was
chosen to account for aeroelastic, as well as and transverse plate effects.
Both an aluminum based and an graphite/epoxy based airfoil were developed. Closed-
loop bench-top and wind tunnel experiments were used to validate the models, and to
control law development procedures.
The results demonstrated the ability of integrated strain actuators to effect high-authority
multi-variable control on flexible plate-like structures. Furthermore, the airfoil wind tunnel
testing demonstrated the ability to successfully address control objectives such as allevia-
tion, command following, and flutter suppression, which directly relate to aircraft design
criteria such as ride quality, maximum wing loading, maneuverability, flight envelope and
overall performance.

1 E.F. Crawley and K.B. Lazarus, Induced strain actuation of isotropic and anisotropic plates, American Institute
of Aeronautics and Astronautics, Journal 29, 944-951, 1991.
46 DISTRIBUTED PIEZOELECTRIC ACTUATING SMART STRUCTURES

Figure 8.3 Development of distributed actuator system in mid 1980s.

The early 1990s (Figure 8.5): depicts an active wing segment developed by MIT/NASA
Langley to investigate flutter control and gust alleviation; uncovered (left) and covered
(right) wing segment; (courtesy of Active Control eXperts, Cambridge, MA, USA):
Based on the experiences gained in the 1980s large scale pilot projects were initiated
soon thereafter, and these were mainly fueled by several U.S. government organizations
(DARPA, NASA, Office of Naval Research ONR, Army Research Office ARO, Air Force)
that began to support developments in the smart materials and structures field. An example
of a pilot was the joint development of an active wing by MIT and NASA Langley.
The goal of this study was to enhance flutter control and gust alleviation in aircraft
wings. A total of 72 piezoelectric actuators (each containing four piezo elements) were
mounted onto a 1,83m long by 0,61 chord wing segment, and produced span wise bending
and changes in the wing’s camber and twist. While vibration reduction was demonstrat-
ed, a remaining challenge was the modelling of optimal placement of the actuators on the
wing segment.
Unnecessarily many actuators lead to added weight and added to the hardware wiring
and control loop complexity. Another example of a pilot project was the reduction of buffet
vibrations in a high performance twin tail aircraft by active control using surface mounted
piezoelectric wafers on the aircraft tails.
Most of the research projects on distributed actuators focused on potential applications
for the aerospace industry in the early 1990s. Research focused mostly on vibration sup-
pression, interior and exterior noise mitigation, actively adjusted precision structures, and
active helicopter rotor blade control. In addition, underwater transducer and sonar applica-
tions were further developed. The requirements for active material systems for aerospace
and underwater applications began to converge. In addition to performance data character-
 47

Figure 8.4 Development of distributed actuator system in late 1980s.

istics like robustness, reliability, and fatigue, resistance became more and more important
as real applications were being developed and implemented.
The first application, which used distributed piezoelectric actuators and, was geared
towards the civilian mass market, was probably the introduction of the “smart ski”in the
mid-1990s based on a collaboration between K2 (Vashon, WA, USA) and Active Control
eXperts Inc., (ACX, Cambridge, MA, USA).
Piezoelectric wafers, which were wired to an analog-logic electrical circuit, were em-
bedded in the laminate of the ski in order to detect and dampen vibration. Based on this
technology, soon similar sports equipment was developed where vibrations could be effec-
tively addressed. Examples include water-skis, snowboards or baseball bats.
Mid 1990s(Figure 8.6) showed development of flexible actuators based on piezoelectric
fibers embedded in a polymer matrix (Active Fiber Composites) and sandwiched between
interdigitated electrodes.
Some of the greatest barriers to implementation for monolithic piezoceramic actuators
are brittleness and catastrophic failure once a crack reaches a critical length in the ceram-
ic. One approach to mitigate these limitations is to embed piezoelectric components for
protection in a polymer matrix.
Piezoelectric composites have, therefore, been commonly used during the past decades
to enable certain applications. However, owing to their limited performance they have not
been utilized on a larger scale until the mid-to-end 1990s when piezofiber composites with
interdigitated electrodes were developed and commercialized.
Active Fiber Composites (AFCs) have been successfully developed with the goal to
offer structural robustness, conformability, and directionality, while still offering sufficient
48 DISTRIBUTED PIEZOELECTRIC ACTUATING SMART STRUCTURES

Figure 8.5 Development of distributed actuator system in early 1990s.

actuation authority. Therefore, they directly helped to overcome some of the problems
mentioned above.
The concept was to unidirectionally align piezoceramic fibers in a monolayer, and to
embedded in a polymer matrix that protects the fibers and provides a path for load transfer.
Driving voltages are delivered by means of two interdigitated electrodes that are attached
to both sides of the active ply.
The late 1990s noted (Figure 8.7) shows application of piezo wafers to alleviate the
bugged problem in a twin tail aircraft.
The latest development in smart structures will be incorporated into the semiconductor
industry and MEMS/NEMS for high-performance sensors, and even in wearable format.
 49

Figure 8.6 Development of distributed actuator system in mid 1990s.

Figure 8.7 Development of distributed actuator system in late 1990s.

CHAPTER 9

SOME GENERAL ASPECTS OF

PIEZOELECTRIC SENSOR MECHANICS

9.1 Introduction

Some general problems of piezoelectric sensor mechanics are discussed below. It shows
that modal sensors cannot be constructed for general cases of beam, plate and shell, except
in the case of Lee and Moon (1990)1 . This means that distributed sensors generally detect
a number of vibration modes simultaneously, and can reduce but not completely remove
the control spillover. Random sensors, noise effects, finite element interpolation of sensor
equation and smart finite element formulation of smart structures are also presented. Al-
though with emphasis on plates and beam, all the results that presented here are also valid
for shells.

9.2 Smart structure and active control

In the context of intelligent materials and structures, there is considerable focus on sensors,
actuators and control capabilities.
The goal of creating a design that limits vibration amplitudes and duration in the pres-
ence of both shock and vibration disturbances faces limitations in the form of constraints
regarding choice of mass, damping, and stiffness (static deflection) values. For instance, in
the design of an isolation system, it often occurs that the desired calls for a value of stiff-

1 Lee, C.K. and Moon, F C, Modal sensors/actuators, ASME J. Appl. Mech., 434/Vol.57, June 1990.

Smart Materials and Structures. 51

By Bohua Sun Copyright ⃝ c 2017 Bohua Sun
52 SOME GENERAL ASPECTS OF PIEZOELECTRIC SENSOR MECHANICS

ness that results in a static deflection is too large for the intended application. Sometimes,
a given isolation design might be required to operate over a load range that is impossible
to meet with a single choice of mass and stiffness. In addition, once the materials are fixed
for a given system, it is difficult to change the mass and stiffness of the system more than
a few percent. Basically, the choice of the physical parameter m, c, and k determines the
response of the system. The choice of these parameters to obtain a desired response is the
design problem. This design procedure can be thought of as passive control (for example,
adding mass to a machine base to lower its frequency). If the constraints on mass (M),
damping (C), and stiffness (K) are such that the desired response cannot be obtained by
changing M, C, and K, active control may provide an attractive alternative.
Active control uses an external adjustable (or active) device, which is called an actua-
tor to provide a force to the device, or structure whose vibration properties should to be
changed. The force that is applied to the structure by the actuator is dependent on a mea-
surement of the response of the system. This is called feedback control. If the goal of the
active control system is to remove unwanted vibration, then the control system is called
active vibration suppression, which consists of measuring the output or response of the
structures to determine the force that should be applied to the mass to obtain the desired
response. The device that is used to apply the force, together with the sensor used to mea-
sure the response of the mass and the electronic circuit is called the control system. The
mathematical rule that is used to apply the force from the sensor measurement is called the
control law.
Active control systems provide increased versatility and better performance in the de-
sign of vibration suppression systems. However, they do so with substantial increase in
cost and potential decrease in reliability. Even in the face of increased costs and complex-
ity, active control methods for vibration suppression are often the only alternatives.

9.3 Sensory elements

It is clearly evident, therefore, to understand sensor’s mechanism is essential to obtain the

full picture of the smart structure (Crawley, 1994). Sensory elements of smart structures
must be sensitive to the mechanical states of the structure and capable of being highly
distributed. The ideal sensor for a smart structure converts strain or displacement (or their
temporal derivatives) directly into electrical outputs. The primary function requirements
for such sensors are their sensitivity to the strain or displacement (or their time derivatives),
spatial resolution, and band-width. Secondary requirements include the transverse and
temperature sensitivity, linearity and hysteresis, electromagnetic compatibility, and size of
sensor packaging. Although actuators are so large they must be explicitly accommodated
in the built-in laminates, hence it is desirable to make sensors small enough to be placed in
inter-laminar or otherwise unobtrusive positions.
Lee and Moon (1990) proposed various modal sensors; and Callahan and Baruh 2 de-
veloped the segmented piezoelectric film sensor. Tzou, et al3 discovered membrane modal
sensitivity and transverse modal sensitivity for piezoelectric shell sensor. Nevertheless,

2 Callahan, J and Baruh, H., Modal analysis using segmented piezoelectric sensors, AIAA Journal, Vol.33, No.12,

December 1995.
3 18. Tzou, H S et al, Sensor mechanics of distributed shell convolving sensors applied to flexible rings, ASME J

Appl Mech, 40/Vol.115, January 1993.

 WORKING PRINCIPLE OF PIEZOELECTRIC SMART STRUCTURES 53

some problems should still be reconsidered, for instance, nonlinear effects, random effects
and the possibility of a general modal sensor. This comes to our motivation.
Some general problems of piezoelectric sensor mechanics are discussed below. It has
shown that modal sensors cannot be constructed for general case of beam, plate and shell,
except in the case of Lee and Moon (1990). This means that distributed sensors generally
detect a number of vibration modes simultaneously, and can reduce but not completely
remove the control spillover. Random sensors, noise effects, finite element interpolation of
sensor equation and smart finite element formulation of smart structures are also presented.
Although with emphasis on plates and beam, all the results that are presented here are also
valid for shells.

9.4 Working Principle of Piezoelectric Smart Structures

The deformation of piezoelectric structures will allow sensor generate charge; the total
charge that is developed on the surface of the sensor layer is q(t). When piezoelectric
sensors are used as strain sensors, the output charge can be transformed to the sensor cur-
rent i(t) = dq/dt, and current is converted into the open circuit sensor voltage output
of controller V (t) = −Gc i(t), which is actually the applied load on the actuators of s-
mart structures, which are by the controller. The actuator layers will deform and result as
dampers of the smart structures.
According the working principle, the simulation of smart structures consists of follow-
ing steps:
Sensor analysis: this step is to formulate sensor equation and to obtain currents I(t):.
Since no external electric field is applied to the sensor layer, the electric displacement
that is developed on the sensor surface is directly proportional to the strain that acts on the
sensor. If the poling direction is z for sensors with the electrodes on the upper and lower
surfaces, the electric displacement is

Dz = e31 εx + e32 εy + e36 γxy . (9.1)

The total charge on the sensor layer will be

∫ ∫
q(t) = Dx dxdy, (9.2)
A

where A is the domain of the sensor. The current can be obtained by

∫ ∫
dq(t)
i(t) = , i(t) = (e31 ε̇x + e32 ε̇y + e36 γ̇xy )dxdy (9.3)
dt A

which can be further expressed in terms of velocity.

Controller analysis: this step is to design a control mechanism, which is also called
feedback analysis; usually this step is simulated by introducing a series of gain Gc and
negative velocity feedback control.
The input for controller is i(t), and the output of the controller will be in voltage:

V (t) = −Gc i(t). (9.4)

The gain Gc plays a role as amplifier and − presents a negative feedback control, that is
−Gc is like a controller in reality.
54 SOME GENERAL ASPECTS OF PIEZOELECTRIC SENSOR MECHANICS

Actuator analysis: This step formulates the actuator equation, that is, the reaction forces
and the moment that is supplied by the actuator layers.
The actuator layers are thin so that the electric field intensity E will be uniform for each
layer. The following applies for the k-th layer:

EZk = V k /hk , (9.5)

where V k is the applied voltage across the k-th layer and hk is the thickness of the k-th
layer.
According to the definition of resultant forces and moment of laminated structures
(beams, plates and shells), the contribution of actuator layers on the resultant forces and
moments are
∑
n
Np = Ezk ek31 (zk − zk−1 ),
k=1
∑n (9.6)
Mp = 1
2 Ezk ek32 (zk2 − zk−1
2
),
k=1

where zk is the distance of k-th layer to the reference surface of the smart structures, and
number n denotes the total number of layers.
Smart structure analysis: this last step is to combine all the above results to formulate
a governing equation in terms of the Hamilton principle, which is called smart structures
equations (SSE).
By using the Hamilton principle, the smart structure equations can be formulated. In
finite element method form, the SSE can be written as:

M ü + Gc u̇ + Ku = Fext . (9.7)

This equation is quite general for linear cases and it clearly shows that on actuator is like an
active damper. The stiffness and mass matrix are the same as ordinary laminated structures.
This remarkable conclusion makes the numerical calculation easily, and all elements that
are developed for non-smart structures can also be used for the analysis of smart structures.
The only different is that each element has one more degree of freedom, which is the
electric voltage or to be precisely the gain Gc . According to modal analysis, the SSE can be
transformed into a standard principle coordinate form üi (t)+2ζi ωi ṙi (t)+ωi2 ui (t) = fi (t),
whiose solution is

∫ t
1
ui (t) = di exp(−ζi ωi t) sin(ωdi +ϕi )+ fi (τ ) exp(ζi ωi τ ) sin ωdi (t− τ )dτ, (9.8)
ωdi 0

√ (u̇ +ζ ω u )2 +(u̇2i0 )
where ωdi = ωi 1 − ζi2 , ζi = 2m Gc Ci
i ωi
, ϕi = tan−1 u̇i0u̇+ζ
i0 ωdi
i ωi ui0
, di = i0 i iω2i0
di
The above equation is merely to show how the gain Gc influences the damping coefficient
ζi , frequency ωi , amplitude di and phase ϕi .
The analysis steps show that we can easily see that sensor analysis is essential for pick
up information of the motion of structure. Without the correct sensor information, control
of smart structures would be a dream.
 SENSOR ANALYSIS 55

9.5 Sensor analysis

A linear piezoelectric classical plates theory that incorporates the piezoelectric effect into
the elastic laminate constitutive equation has been studied by Lee4 , and Lee and Moon
(1990), Callahan and Baruh(1995). All of them assumed the in-plane displacement u and v
to be negligible compared to the transverse deflection w and the output signal generated by
w only. Here we will not omit the in-plane displacement, because it should be considered
in cases that have laminated plates, unless the sensors were in the mid-surface. Besides
this, the effects of transverse deformation are also considered because the inter-laminar
shear Modula are usually much smaller than the in-plane Young’s Modula.
If one uses the Mindlin plates theory, that is the charge generated by each piezoelectric
sensor is given by:
∫ ∫
qr (t) = Fr P0 [e]T [L][u] + zr [θ])dxdy, r = 1, 2, ·, N. (9.9)
A
which is generally called a sensor equation. The signal output qr (t) is an average quantity
over the sensor coverage. It proved that frequency of charge qr (t) is equal to the frequency
of sensors and their host structure, where z = (zk + zk−1 )/2 is the average distance from
the geometric centre of the rth sensor to the mid-plane of the laminate and N is the number
of piezosensors, A is the sensor coverage and is the polarization profile (Lee and Moon,
1990), and where F (x, y) is Heavyside type function
{
1 , if (x,y) is within the r-th sensor coverage A
F (x, y) = (9.10)
0 , otherwise
and
   
e31 ∂
0 ( ) ( )
∂x
   ∂  ux θx
[e] = e32  , [L] =  0 ∂y 
, [u] = , [θ] = . (9.11)
∂ ∂
uy θy
e36 ∂y ∂x

As Tzou et al (1993) point out that signal output of sensors can be divided into two parts,
called sensitivities: (1) transverse modal sensitivity; and (2) membrane modal sensitivity.
In general, the transverse modal sensitivity is defined for transverse natural modes and
the membrane modal sensitivity for in-plane natural modes. However, in equation 9.1 the
transverse modal sensitivity is defined for rotation or for the precise change of curvature, so
there is no explicit contribution from transverse deflection w; the reason for this is because
the sensor’s constitutive equation for both e33 and e34 are null. In order to capture the
contribution of w, a piezoelectric material that has no-zero e33 and e34 , should be chosen.

9.5.1 Piezo Composite Beam

In this case, uy = 0 displacement is the only function of x, hence the sensor equation 9.9
simplifies to the one-dimensional sensor equation
∫ a ∫ b/2
qr (t) = e31 Γr (x)(u, x + zr θx,x )dx, Γr = Fr P0 dy. (9.12)
0 −b/2

4 LeeC. K., Theory of Laminated Piezoelectric Plates for the Design of Distributed Sensors/Actuators. Part I:
Governing Equations and Reciprocal Relationships, Journal of Acoustical Society of America, 87(3), pp.1144-
1158, (1990).
56 SOME GENERAL ASPECTS OF PIEZOELECTRIC SENSOR MECHANICS

According to the linear vibration theory, the vibration solution can be expressed in mode
function um , θm as:
∞
∑ ∞
∑
u θ
u(x, t) = um (x)Tm (t), θx (x, t) = θm (x)Tm (t). (9.13)
m=1 m=1

when one substitutes this solution into the sensor equation, one finds:
∞
∑
u θ
qr (t) = [Am Tm (t) + zr Bm Tm (t)], (9.14)
m=1
∫a ∫a
where Am = e31 0
Γr (x)um , xdx and Bm = e31 0
Γr (x)θm , xdx

9.5.2 Modal sensor

Am and Bm are constant coefficients. By using the classical plate theory and omitting
in-plane displacement u and v, Lee and Moon 5 have shown that if the spatial electrode
pattern of the rth lamina is proportional to the modal strain distribution along the length of
the plate, one can obtain a true modal sensor based on the orthogonality of the eigenvalue-
solution of the beams with respect to their stiffness. In other words, observer spillover
will not be present in systems that adopts these types of sensors. Further more, Lee and
Moon also propose modal actuator, which excites each particular mode independently so
that actuator spillover will not be an issue in one-dimensional plates that are actuated by
this type of actuator.
However, for both linear and nonlinear Mindlin plate theory, since there is no general
orthogonality for both um and θm , it is impossible to have the mode sensor like beam
whether or not we choose Γr (x). In other words, observer spillover will always be present
in the system that adopts the Mindlin type of sensor. As a consequence, the modal actuator
will not exist in either of the plates.In only one exceptional case can the sensor be used as
a membrane sensor if it is placed on the mid-surface, zr = 0. In this case, one can obtain
a membrane modal sensor and membrane modal actuator for the beam type plates.

5 Lee, C.K. and Moon, F C, Modal sensors/actuators, ASME J. Appl. Mech., 434/Vol.57, June 1990.
CHAPTER 10

SMART PIEZO COMPOSITE MINDLIN

BEAMS

10.1 Introduction

Due to the increasing demands for high structural performance requirements, the studies
of embedded or surface-mounted piezoelectric materials in structures have received con-
siderable attention in recent years. Smart structures technology, which feature a network
of sensors and actuators, real-time control capabilities, computational capabilities and host
materials will have a tremendous impact on the design, development and manufacture of
the next generation of products in diverse industries. The idea of applying smart materials
to mechanical and structural systems has been studied by researchers in various disci-
plines. Polyvinylidene fluoride (PVDF) was initially discovered by Kawai1 in 1969. Raw
polymetric PVDF (α phase) is an electrical insulator and it does not have any intrinsic
piezoelectric properties. If the raw material is polarized during the manufacturing process,
PVDF transforms to β-phase - a tough and flexible semi-crystalline material and it can be
made to strain in either one or two directions in the film plane. Since β-phase PVDF pos-
sesses a strong direct piezoelectric effect, it has been used in many transducer applications:
for example, sonar, medical ultrasonic equipment, robot tactile sensors, acoustic pick-ups,
forces and strains gauges and so on. Due to its distinct characteristics such as flexibili-
ty, durability, manufacturability, and so on., PVDF is an ideal material for the distributed
sensing and vibration suppression/control of distributed parameter systems (for example,
beams, plates, shells, and so on.).

1 H. Kawai, 1969, The piezoelectricity of ply(vinylidene fluoride, Japan Journal of Applied Physics 8, 975.)

Smart Materials and Structures. 57

By Bohua Sun Copyright ⃝ c 2017 Bohua Sun
58 SMART PIEZO COMPOSITE MINDLIN BEAMS

In order to utilize the strain-sensing and actuating properties of piezoelectric materials,

the interaction between the structure and SSA (strain sensing and actuating) material must
be well understood.

10.2 Formulations

Figure 10.1 Laminated beam with integrated piezoelectric sensor and actuator

Constitutive equations are developed for laminated beams that have piezoelectric sen-
sors and actuator layers. The shear deformation effect is incorporated in the formulation
by using Mindlin plate theory. Consider a smart laminated beam, which has length L,
width b, and thickness h Figure(10.1). The electric field is applied through the thickness of
the piezoelectric material. The constitutive equations, including piezoelectric effects with
respect to the plane (laminate) co-ordinates is (x, y, z), namely:

[σ]k = [Q̄]k [ε]k − [ē]Tk [E]k , (10.1)

[D]k = [ē]k [ε]k + [ḡ]k [E]k , (10.2)

where [D], [E], [ε], [σ]are the electric displacements, electric field, strain and stress vec-
tors, and [Q̄], [ē], [ḡ] are the elasticity, piezoelectric and permittivity constant matrices,
respectively.
In this example the PVDF is chosen as the material of both the sensors and the ac-
tuators. Piezoelectric material layers are polarized in the thickness direction and exhibit
transversely isotropic properties in the xy-plane. Considering piezoelectric materials while
retaining the anisotropic behaviour of the master structure, the constitutive equations can
 FORMULATIONS 59

be rewritten as:
       
σx Q̄11 Q̄12 0 0Q̄16 εx 0 0 ē32
         
 σy  Q̄12 Q̄22 0 0Q̄26   εy   0 0 ē31  0
         
τyz  =  0 0  γyz  − ē25 0 
   0 Q̄44 Q̄45     ē15  0 ,
       
τxz   0 0 Q̄45 Q̄55 0  γxz  ē15 ē25 0  Ez k
τxy Q̄16 ¯
Q̄ 0 0 Q̄66 k γxy k 0 0 0 k
26
k
(10.3)
where ē31 = e13 ,ē15 = e15 (cos2 θ − sin2 θ) and ē25 = −2e15 sin θ cos θ.
Thus, because of the above statement, for a beam problem one can use σy = τyz =
τxy = 0, while assuming that εy ̸= γyz ̸= γxy ̸= 0, in order to obtain the following
reduced constitutive equations of smart composite beams from the equation (Eqn. 10.3).
( ) [ ] ( ) ( )
σx Q̃11 0 εx ẽ31
= − Ezk (10.4)
τxy k 0 Q̃55 k γxz k 0

where the constants Qij

Q̄16 Q̄26 − Q̄12 Q̄66 Q̄12 Q̄26 − Q̄16 Q̄22

Q̃11 = Q̄11 + Q̄12 + Q̄16 (10.5)
Q̄22 Q̄66 − Q̄26
2 Q̄22 Q̄66 − Q̄226
Q̄2
Q̃55 = Q̄55 − 45 (10.6)
Q̄44
( )
Q̄12 Q̄26 − Q̄16 Q̄22
e31 = 1− ē 31 . (10.7)
Q̄22 Q̄66 − Q̄226
The displacement field for the present composite beams based on the Mindlin plate theory,
can be written as:
u1 (x, t) = u0 (x, t) + zψ0 (x, t),
u2 (x, t) = 0, (10.8)
u3 (x, t) = w0 (x, t).
The strain displacement relations of a laminated beam, based on a first order shear defor-
mation theory, which is associated with the displacement field are given by:

εx = ε0x + zκ1x , 0
γxz = γxz , (10.9)

where
∂u0 ∂ψ0 ∂w0
ε0x = , κ1x = 0
, γxz = ψ0 + (10.10)
∂x ∂x ∂x
In the present beam model, it is assumed that the bottom and top layers are sensor and
actuator layers, respectively. For sensor laminae, no external electric field is applied to this
layer. Then, the electric field intensity for the sensor is zero.
Substituting equation (Eqn.10.8) into equation (Eqn.10.4), setting the sensor layer as
Ezk = 0, and integrating through the thickness, the stress resultants can be obtained as
    0   p
Nx Ā11 B̄11 0 εx Nx
     1   p
 Mx  = B̄11 D̄11 0   κx  − Mx  , (10.11)
0
Qxz 0 0 Ā55 γxz 0
60 SMART PIEZO COMPOSITE MINDLIN BEAMS

Figure 10.2 Cross section geometry of an n-layered laminated beam.

where the matrix constants are

n ∫
∑ zk
(Ā11 , B̄11 , D̄11 ) = Q̃11 (1, z, z 2 )dz, (10.12)
k=1 zk −1

∑n ∫ zk
Ā55 = Q̄55 dz, (10.13)
k=1 zk −1
∫ zk
(Nxp , Mxp ) = ẽ31 Ezk (1, z)dz, (10.14)
zk −1

in which zk and z0 = −h/2, zn = h/2 are shown in Figure(10.2).

Regarding the beam problem, the varying of y-axis is not considered because any phys-
ical variables are uniformly distributed through the y direction. The electric potential Φ of
smart composite beams can be written as:

Φ(x, z, t) = f (z)ϕ0 (x, t) (10.15)

Since the thickness of the piezoelectric layers is thin, it is also assumed that the voltage is
uniformly distributed through the thickness (Z direction) of the piezoelectric layers. That
is, f (z) = 1. Hence, the electric field intensity Ezk : can be expressed as

ϕ0 (x, t)
Ezk = . (10.16)
hp

where hp is the thickness of the piezoelectric layers.

 NUMERICAL EXAMPLE 61

Piezoelectric materials are polarized in the thickness direction and exhibit transversely
isotropic properties in the XY-plane. So, for equation (Eqn.10.2), only Dz is of interest
here. Considering piezoelectric materials while retaining the anisotropic behaviour of the
master structure, and the beam ϵy = 0, the constitutive equation (10.2) can be written as:

Dzk = ē31 εx + ē31 εy + ḡ33 Ezk = ē31 εx + ḡ33 Ezk , (10.17)

Q̄ ē2
where g̃33 = ḡ33 + Q̄22 Q̄6666 −31Q̄2 .
26
From the Hamilton principle, the governing equations can be obtained as follows:
∂Nx
δu0 : I1 ü0 + I2 ψ¨0 − = 0, (10.18)
∂x
∂Qxz
δw0 : I1 ẅ0 − = p, (10.19)
∂x
∂Mx
δψ : I2 ü0 + I3 ψ¨0 − + Qxz = 0 (10.20)
∂x
δϕ0 : Gkz + q0 = 0, (10.21)
n ∫
∑ zk
where (I1 , I2 , I3 ) = zk −1
ρk (1, z, z 2 )dz and
k=1

n ∫
∑ zk
Dzk
Gkz = dz. (10.22)
zk −1 hp
k=1

When one substitutes equation (10.17) into equation (10.22), one finds that the electric
potential should be expressed as:
∂u0 ∂ψ0
ϕ0 (x, t) = α1 + α2 + ϕA (x, t), (10.23)
∂x ∂x
n ∫
∑ zk
where α1 = −ẽ31 Gk1 /(g̃33 Gk3 ), α2 = −ẽ31 Gk2 /(g̃33 Gk3 ), and Gk1 = zk −1
(1/hp )dz,
k=1
n ∫
∑ n ∫
∑
zk zk
Gk2 = zk −1
(1/hp )zdz, and Gk3 = zk −1
(1/hp )2 dz.
k=1 k=1
For piezoelectric sensor layers, ϕA = 0, hence the electric potential within the sensor
can be approximately proposed in the form
∂u0 ∂ψ0
ϕ0 |sensor (x, t) = α1 + α2 . (10.24)
∂x ∂x
In the active vibration control application, the electric force term can be regarded as the
feedback control force. The electrical potential input of the actuator layer can be written
in terms of the output signal of the piezoelectric sensor layer multiplied with a gain Gi
∂u0 ∂ψ0
ϕ0 |actuator (x, t) = −Gi (α1 + α2 ). (10.25)
∂x ∂x

10.3 Numerical Example

An intelligent beam structure which contains a distributed piezoelectric sensor/actuator on

both the top and bottom surface is shown in Figure (10.3) above.
62 SMART PIEZO COMPOSITE MINDLIN BEAMS

Figure 10.3 Cross section geometry of an n-layered laminated beam.

Table 10.1 The material properties of the main structure and piezoelectric
P roperty PVDF Graphite/epoxy
E1 0.2E + 10N/m2 0.98E + 11N/m2
E2 0.2E + 10N/m2 0.79E + 10N/m2
2
G12 0.775E + 9N/m 0.56E + 10N/m2
G23 − 0.38E + 10N/m2
ν12 0.29 0.28
ρ 1800kg/m3 1520kg/m3
e31 0.046C/m2 −
e32 0.46C/m2 −
e33 0.1062E − 9F/m −
g11 0.2E + 10N/m 2
−
g22 0.2E + 10N/m2 −
g33 0.2E + 10N/m 2
−
t 0.1E − 3m 0.12E − 3m

In this structure, the piezoelectric of the bottom layer is considered as a sensor in order
to sense the strain and generate the electrical potential and the piezoelectric of the top layer
as an actuator to control the vibration of the structure. All material properties that are used
are shown in Table(10.1).
A cantilever laminated beam with a distributed piezoelectric PVDF layer that serves
as a distributed actuator on the top surface, and another PVDF on the bottom surface
as a distributed sensor, will be used for the case study. The beam dimensions that were
considered are length l=100mm and width b=5mm. The thickness of the beam can be
generally written as h = 0.125 × 10−3 n, and the piezoelectric PVDF layer is taken as
0.1 × 10−3 . The applied transverse load is uniformly distributed and has a magnitude of
p1 (x) = p2 (x) = 2.5 × 103 N/m2 .
 NUMERICAL EXAMPLE 63

Figure 10.4 Effect of negative velocity feedback gain on the tip transient response [0/90/90/03].

Figure 10.5 The tip defection of beam versus feedback control gain.

Here, all the graphical outputs are obtained by using the Software Package MATHE-
MATICA. The transverse displacement of four-layer laminated composite beams with an
actuator and sensor layer on the top and bottom surfaces, respectively, for feedback gains
of 0, 40, 100 and 140, and are shown in the following figures. In the following graphical
results, the frequency of the external applied force is taken as 10 Hz.
Figure (10.4) shows the effect of negative velocity feedback control gain on the tip
transient response.
Figure10.5 shows that the tip deflection of the beam versus feedback control gain for
the different ply orientations. From this figure, the tip defection (amplitude) of the beam
64 SMART PIEZO COMPOSITE MINDLIN BEAMS

decreases quickly, while the feedback control gain increases. When the control gain Gi
is less than 100 C/A, the control purpose is effective. However, when the control gain
Gi > 100C/A, the tip defection decreases slowly. All these phenomena indicate that the
optimal feedback control gain of the present beam model is about 100C/A.
By observing Figure(10.5), the mathematical relation of the amplitude of the beam vi-
bration and feedback control gain can be approximately expressed in the form of:

f = A(1 − eαGi ) + B, (10.26)

where f is the amplitute of vibration and Gi is the feedback control gain. The coefficient
A, B and α can be determined by choosing three points from the above graphical results.
The numerical results of the dynamic response and analysis of the state governing equa-
tions for smart laminated beams, show that the displacement decays amplitude, while the
feedback gain increases. Hence, the present method is correct and effective.
CHAPTER 11

RECTANGULAR AND CIRCULAR SHAPE

DISTRIBUTED PIEZOELECTRIC
ACTUATOR

11.1 Introduction

Actuators and sensors have been identified as two of the critical component technologies
for the advancement of intelligent structures. Although intelligent, smart, active, con-
trolled and adaptive are often used interchangeably, a useful paradigm is the one, which
is presented by Wada et al 1 . This identifies intelligent structures as a subset of all other
types of structure, capturing the highest level of complexity and functionality of system
components, which can sense and respond to stimuli within their environment.
In the intelligent (or smart) technology field, the distributed piezoelectric actuator (DPA)
is one of the most useful types of actuator. Firstly, DPAs are of a solid actuator and can
be easily embedded into any composite structures. Secondly, they can be controlled by an
electrical signal with high bandwidth and high precision. So they can be applied to many
different fields such as high-resolution positioning, noise and vibration detection and shape
control.

1 Wada B K, Fanson J L and Crawley E F, 1990, Adaptive structures, J. Intell. Mater. Syst. Struct., 1, 157–74.

Smart Materials and Structures. 65

By Bohua Sun Copyright ⃝ c 2017 Bohua Sun
66 RECTANGULAR AND CIRCULAR SHAPE DISTRIBUTED PIEZOELECTRIC ACTUATOR

The DPA can be tailored into any shape to fit the host structure, while the rectangular and
circular shapes are the most popular shapes, which have been investigated comprehensively
by Sun and Qiu, Below are some of their studies 2 3 .

11.2 Rectangular DPA element

11.2.1 Polarization of Rectangular DPA Element

In Figure (11.1) the rectangular DPA element, the poles of the actuator, have been cancelled
between the two closest IDE branches on one side. And on the other side, the poles have
been cancelled with the same situation. The direction of the majority of the electrical field
and polarization tensors in the actuator element follows the piezo-electric material length
direction,which is the Y axis.

Figure 11.1 The principal type of DPA element with interdigital electrode (IDE).

For each element of the rectangular DPA, considering the structure of the element, the
minimum assumptions have been made, which are outlined below:
1. The properties of the piezoelectric material and the cover material (epoxy) in the
actuator are uniform.
2. Due to the general dimensions of the actuator in the X axis always being much bigger
than the IDE branch centre distance 2b, the electrical and mechanical effects on the X
axis have been ignored.

2 Bohua Sun and Yan Qiu, Analysis of circular shape distributed piezoelectric actuators, Composite Structures,

62, 177-191, 2003.

3 Bohua Sun and Yan Qiu, Rectangular shape distributed piezoelectric actuator: analytical analysis, Smart Mater.

Struct. 13,337-349, 2003.

 RECTANGULAR DPA ELEMENT 67

3. The following analysis processes have supposed that the dimensions of the actuator
in the Y axis are much bigger than the IDE branch centre distance 2b. The electrical
edge effects for each element have been ignored.

11.2.2 Distributed Charge Density Analysis

For the analytical analysis of the electrical field in each DPA element, the charge distributed
density on each of the poles should first be understood and described, which this has been
ignored in previous work.
According to the theory of electro-magnetics, each point on the poles will have the same
electric potential, and the interelectrical field tensor Eint = 0 equals zero. Consider the
structure of an IDE branch and the distribution of poles, the Eint for each point on the pole
in one element can be obtained by the following formula:
∑n ∫ ∫
ρ(s)
Eint = 2
rds, (11.1)
i pole 4πεR

where ρ(s) is the is the surface charge distributed density on the electrodes in each element.
It can be resolved by the product of two line charge densities:

ρ(s) = ρX (l)ρY (l), (11.2)

where ρX (l) is the line charge density on the X axis, and ρY (l) is the line charge density
on the Y axis. The function ρX (l) is equal to a constant that depends on the applied voltage
for the actuator. The function ρY (l) has been contained by three boundary conditions:
dρY (l) dρY (l)
|y=0 = 0, ρY (l)|y=±a = ∞, |y=±a = ∞. (11.3)
dy dy
Using the above boundary conditions, the charge density ρY (l) can be expressed as

ρY (l) = f (y)g(y)h(y), (11.4)

where the functions are

f (y) = 1
−|y|+1 , g(y) = [A(− | Dy | +DC)−B ]−1 ,
∑
n (11.5)
h(y) = Ai (− | y | +C)i + C1 .
i=1

The constants can be found in the table (11.1) below.

For the case of n = 2, the rectangular DPA element is illustrated in Figure (11.2).

11.2.3 Capacitance Analysis of the DPA Element

A capacitor is a device which stores electric charge and hence electric energy. It consists
of two conductors that are separated by an insulating medium. Its capacitance is defined as
the ratio of the charge on one of the conductors to the applied voltage (potential difference).
Thus, the capacitance C of a capacitor is C = Q/V , where C is the capacitance in farads,
F, Q is the charge in coulombs, C, and V is potential difference in volts, V, hence:
∫
ρ(s)ds
C = ∫sa . (11.6)
Es dl˙
b
68 RECTANGULAR AND CIRCULAR SHAPE DISTRIBUTED PIEZOELECTRIC ACTUATOR

Table 11.1 The constants in the charge density with different h/a and b/a.

The components of the electrical field intensity Es can be found in Sun and Qiu 4 . The
numerical simulation is shown in Figure 11.3 below.

11.2.4 The Mechanical Performance Analysis of the Rectangular DPA Ele-

ment

After building the analytical analysis models of the electrical field in the DPA element,
the next step, which is required develop the DPA technology here is the analysis of the
mechanical performance of the DPA element, applying the models that was developed
above.

11.3 The Electrical Field Distribution in the Part-circular Shape DPA

For the part-circular shape DPA shown in Figure 11.5 below, considering the structure of
element, the minimum hypotheses are as follows:
1. The material properties of piezoelectric material and cover material (epoxy) in the
actuator are uniform.

4 Bohua Sun and Y Qiu, Rectangular shape distributed piezoelectric actuator: analytical analysis, Smart Mater.

Struct. 13 (2004) 337-349.

 THE ELECTRICAL FIELD DISTRIBUTION IN THE PART-CIRCULAR SHAPE DPA 69

Figure 11.2 The DPA element.

Figure 11.3 The DPA element.

2. The ratio of the IDE branch center radian 2b and the IDE width radius 2a have been
kept the same in the part-circular shape DPA.

3. The following analysis processes have supposed that the dimension of the actuator in
the radius direction is much bigger than the IDE branch center distance. The electrical
edge effects for each element are ignored.

A similar analysis and numerical results of the circular distributed actuator can be found
in Sun and Qiu 5 .

5 Bohua Sun and Yan Qiu, Analysis of circular shape distributed piezoelectric actuators, Composite Structures,

62, 177-191, 2003.

70 RECTANGULAR AND CIRCULAR SHAPE DISTRIBUTED PIEZOELECTRIC ACTUATOR

Figure 11.4 The strain result comparison between the analytical model and the FEM.

Figure 11.5 The DPA element with part-circular shape.

CHAPTER 12

ELECTRO-MECHANICAL PERFORMANCE
OF C-SHAPE PIEZOELECTRIC ACTUATOR

12.1 Introduction

Widespread application of curved shaped piezoelectric actuators in certain smart structures

and devices demand a comprehensive study on this particular shape. Active controls of
satellite dishes or mirrors are examples of applications of curved piezoelectric actuators 1 2
the smart flap drive for individual blade control [3], circular multilayered diaphragm-type
actuators for MEMS, micropumps and acoustic devices, Deshpande and Saggere 3 .
Performance characteristics of a piezoelectric device are determined by material, ge-
ometry and electromechanical coupling properties. Woo et al. 4 , reportthat dome height
and stored elastic energy, induced by thermal deformation during curing, have an influence
on the flexural displacement and load capability of a plate-type piezocomposite actuator.
In order to obtain a wide motion range and ultra-precision simultaneously, Dong et al 5 ,

1 D.S. Stampleman, A.H. von Flotow, Microgravity isolation mounts based on piezoelectric films in active and
vibration control, ASMEWAM (1990) 57–67.
2 G.W. Bahaman, V.H. Schmidt, R.J. Conant, Piezoelectric Polymer Actuators in a vibration isolation application,

Proc. SPIE 3987 (2000) 331–342.

3 M. Deshpande, L. Saggere, An analytical model and working equations for static deflections of a circular multi-

layered diaphragm-type piezoelectric actuator, Sens. Actuators A: Phys. 136 (2) (2007) 673–689.
4 S.-C. Woo, K.H. Park, N.S. Goo, Influences of dome height and stored elastic energy on the actuating perfor-

mance of a plate-type piezoelectric composite actuator, Sens. Actuators A: Phys. 137 (1) (2007) 110–119.
5 W. Dong, L.N. Sun, Z.J. Du, Design of a precision compliant parallel positioner driven by dual piezoelectric

actuators, Sens. Actuators A: Phys. 135 (1) (2007) 250–256.

Smart Materials and Structures. 71

By Bohua Sun Copyright ⃝ c 2017 Bohua Sun
72 ELECTRO-MECHANICAL PERFORMANCE OF C-SHAPE PIEZOELECTRIC ACTUATOR

developed a macro/micro dual compliant parallel positioner system in which piezoelectric

motors and piezoelectric ceramics are integrated into one system.
The influence of substrate/PZT thickness ratio on the optimum performance of the C-
shape actuator is the main focus of this work.
The C-shape piezoelectric actuator, which is a semicircular shell,was invented by Moska-
lik and Brei in 1996 6 . When individual C-shape actuators are combined in a series and/or
parallel, it is possible to generate displacement and force larger than a comparable straight
bender. The force that is produced by an array of C-shape actuators is proportional to the
number of individual C-shape actuators in a parallel arrangement, while the resulting dis-
placement equals the sum of displacements of individual actuators in a series arrangement
7
.
A unimorph individual C-shape piezoelectric actuator (Figure 12.1) consists of three
layers, which are laminated together to form a semicircular shell, namely, one active layer
(piezoceramic) and passive layers (bonding and substrate). The piezoceramic (PZT) layer
is preplated with electrode layers on its inner and outer surfaces. The piezoceramic lay-
er, together with its electrode, is bonded on the outer surface of the substrate. Epoxy is
used as the bonding material and a strong bond is created between the piezolayer and the
substrate. This ensures that all loads that are applied by the active layer are transmitted
fully to the passive layer. With the unimorph actuator, when the piezoelectric layer ex-
pands/contracts in the radial direction the strain in the plane normal to the poling direction
( in the circumferential direction) undergoes a contraction/expansion.
The “perfect bond”between the piezoceramic layer and the substrate leads to a con-
centrated couple or moment M pe at the edges of the piezoelectric actuator layer once a
voltage is applied 8 . The concentrated moment will eventually cause the whole structure to
flex. The assumption of a “perfect bond”also implies continuity of displacement/strain
between the interfaces of the laminate. For the PZT material, the unimorph configuration
is preferred, particularly when the actuating device is subjected to both tensile and com-
pressive loads. PZT materials are weak in tension; therefore, for an actuator to perform
well and in safety, it is important that when determining the stacking sequence of the lay-
ers in the laminate, the piezoceramic layer must be positioned on one side of the neutral
axis. This will ensure that the PZT layer experiences only compressive stresses during its
operation. In addition, by locating the active layer as far as possible from the neutral axis,
a larger moment arm will be obtained and hence a larger bending moment. The location
of the neutral axis is normally achieved by the introduction of passive layers (substrate) of
appropriate geometry (thickness) and elastic modulus.

12.2 Displacement and Force Equations

Since the C-shape actuator is considered to have a thin cross section, it is reasonable to
consider a unidirectional state of loading. The radial displacement and force at the tip of
the free end can be defined by the force-displacement equation which is obtained by using
the Castigliano method for thin piezoelectric curved beams.

6 A.J. Moskalik, Diann Brei, Quasi-static behavior of individual C-block piezoelectric actuator, J. Intell. Mater.
Syst. Struct. 3 (1997) 577–587.
7 A.J. Moskalik, Diann Brei, Force-displacement behavior of piezoelectric C-block actuator arrays, Smart Struct.

Mater. 8 (1999) 531–543.

8 E.F. Crawley, J. de Luis, Use of piezoelectric actuators as elements of intelligent structures, AIAA J. 25 (10)

(1987) 1373–1385.
 DISPLACEMENT AND FORCE EQUATIONS 73

Figure 12.1 C-shape piezoceramic actuator.

Figure 12.2 Unimorph C-shape actuator.

The internal moment M at any angular position Figure 12.2 is the sum of moment owing
to externally applied forces and the piezoelectric moment, hence:
M = −Px Rna sin θ + M pe , (12.1)
where Px is the available force at the free end tip, M pe is the piezoelectric moment, and
Rna is the radius from the origin of the semi-circle to the neutral axis.
∂M
= −Rna sin θ. (12.2)
∂Px
From Figure 12.2 above whows that the complementary energy is obtained by the equation:
∫ π
1
Uc = M 2 Rna dθ, (12.3)
2QIz 0
where Q is the Young’s modulus of elasticity of the material and Iz is the moment of
inertia of the cross-section of an actuator about the neutral axis.
The displacement of the free end can be obtained by using complementary energy as
follows:
∂Uc 2M pe Rna2 3
πPx Rna
δx = = − , (12.4)
∂Px C 2C
where δx is the radial displacement at the free end, and C = QIz is the composite bending
stiffness.
Expressing the available force in terms of the piezoelectric moment and radial displace-
ment, a net force is obtained at at the tip of the free end:
4M pe 2Cδx
Px = − 3
. (12.5)
πRna πRna
74 ELECTRO-MECHANICAL PERFORMANCE OF C-SHAPE PIEZOELECTRIC ACTUATOR

The blocking force (the force obtained when the displacement of the tip of the free end is
equal to zero in the equation 12.4) is given by:

4M pe
Pblock = . (12.6)
πRna
The free-displacement at the free end ( when the piezoelectric forcing term is not resisted
exremally in any way) becomes:

2M pe Rna
2
δf ree = . (12.7)
C
The piezoelectric moment M pe , the radiu of neutral axis Rna and the bending composite
stiffness C can be obtained by
∑
n ∑
n
Mpe = 1
2 bj Qj (h2j − h2j−1 )(d31 E3 )j , C= 1
3 bj Qj (h3j − h3j−1 ),
j=1 ∑n j=1 (12.8)
j=1 0.5Qj bj (rj −rj−1
2 2
Rna = ∑ n
j=1 Q j b j (r j −r j−1 ) ,

where bj is the width of the jth layer, hj is the coordinate of the outer surface of the jth
layer, hj−1 is the coordinate of the inner surface of the jth layer, hj = rj − Rna , and d31
is the piezoelectric strain constant.
For a thin piezoelectric actuator it is sufficient to assume that an electric field is equal
to the change of voltage between the electrodes V that are divided by the thickness t of the
piezoceramic layer concerned, i.e. t = hj − hj−1 ; hence

V
E3 = . (12.9)
hj − jj−1

12.3 Displacement Analysis and Discussion

The results that were obtained indicate that actuators with smaller substrate/PZT thickness
ratio give larger displacement than the ones with greater ratios. The experimental result-
s concur with the theoretical, as well as with those from FEM analysis that is, there is
increased displacement when the substrate/PZT is increased, and subsequently decreased
displacement when the ratio of approximately 0.28 is reached.
Figure 12.3 shows typical experimental, theoretical and finite element analysis results
for displacement of the free end tip of the C-shape actuator.

12.4 Force Analysis and Discussion

Larger force is obtained from an actuator that has larger substrate/PZT thickness ratios
(stiffness dependence). It is observed that the actuator output force is increased when the
substrate/PZT thickness ratio is increased, at least up to a ratio of 1:1 after which a decrease
is also noted. This is obviously owning to the fact that the actuator becomes too thick to
be deformed by the generated piezoelectric actuator force.
Experimental, theoretical finite element results for the actuator are shown in Figure 12.4
above.
 CONCLUSIONS AND RECOMMENDATIONS 75

Figure 12.3 Results for Displacement for substrate/PZT thickness ratios (for 1mm PZT thickness,
at 50 V).

Figure 12.4 Results for the force for substrate/PZT thickness ratios (for 1mm PZT thickness, at
50 V).

12.5 Conclusions and Recommendations

Numerical simulations have been animated in Figure 12.5 above, which can be concluded
that, for a given PZT actuator material and applied voltage, free displacement and blocked
forces are influenced by the actuator stiffness, which is partly contributed by the substrate
material’s elastic properties and the geometry (in this case, the thickness). The thickness
of the substrate material, apart from altering the actuator stiffness, also has the role of
76 ELECTRO-MECHANICAL PERFORMANCE OF C-SHAPE PIEZOELECTRIC ACTUATOR

Figure 12.5 Numerical simulations of deflection and forces.

determining the location of the neutral axis (N.A.), and thus the moment arm, which is
defined as the distance between the midline of the PZT layer ( where the piezoelectric
force is assumed to act) to the neutral axis. On the other hand, the thickness ratio also
determines the portion/part of the cross-section of the PZT layer, which will be subjected
to the compression/tension load during the forward and backward strokes of the device,
especially when subjected to alternating voltage.
For a unimorph actuator, where a PZT layer is on top (outside of the substrate layer, at
lower thickness ratios), a bigger portion will suffer tensile load during the inward stroke
and compressive load during the outward stroke. At higher ratios, the situation will be the
other way round. The farther the neutral axis goes away from the midline of the structure,
the larger the moment arm, and hence the larger the piezoelectric moment. At the ratio of
1:1, the whole PZT layer will either be fully loaded by a tensile load or a compressive load
during the inward stroke and vice versa. In general, the location of the neutral axis will
alter the distribution of stress along the thickness and thus the amount of displacement and
force.
Piezoceramic materials are reported weak in tension, therefore, the location of the neu-
tral axis of the composite is important with regard to its operating life expectancy.
For the thickness ratio, which gives the peak displacement, low voltages can be used to
produce the same amount of displacement, which would have demanded higher voltages if
a much thicker actuator was used. The study, indicates that thickness ratios between 0.25
and 0.30 produce maximum displacement and a relatively smaller force. For ratios beyond
this range, the displacement decreases, while the force increases.
The study concludes that total actuator thickness alone cannot be used to determine the
actuator performance. The results should be of interest to designers who wish to establish
how much force will be sacrificed by choosing to have a certain amount of displacement
and vice versa. The results also help to determine the appropriate geometry ( thickness
ratio) if one aims for large displacement and/or a large force.
CHAPTER 13

MONITORING STRUCTURAL INTEGRITY

USING FIBRE OPTIC SENSORS

13.1 Formulations

In the past few decades, we have witnessed drastic changes in materials technology. Latest
developments such as the use of materials with embedded devices have changed ways of
thinking in modern design and manufacturing.

Figure 13.1 Embedded optical fibres into the test piece.

Smart Materials and Structures. 77

By Bohua Sun Copyright ⃝ c 2017 Bohua Sun
78 MONITORING STRUCTURAL INTEGRITY USING FIBRE OPTIC SENSORS

Crack propagation in materials is considered to be a mechanical fault that needs atten-

tion because it may develop as a result of mechanical deformation. In aerospace, aviation,
construction, mining and many other industries there are certain structures where, owning
to dynamic and/or static loading in specific points, mechanical deformation may lead to
crack initiation, propagation and subsequent catastrophic failure. Mechanical deformation
from fatigue loading has been taken as a major influencing factor in crack creation and
propagation. Conversely, some researchers consider a crack to serve a positive role as a
stress relief mechanism (residual stress).
Due to the fact that cracks sometimes develop from the inner side (core) of the material,
it is difficult to detect or visualize their occurrence, and in most cases to locate their prop-
agation path. Several existing non-destructive testing (NDT) techniques such as magnetic
particles detection, acoustic emission, ultrasonic, electronic speckle pattern interferometry
(ESPI), shearography, as well as physical visualization, are employed in order to detect
crack occurrence and propagation. Recently, researchers have attempted to detect crack
orientation by using strain measured with optical fibres, The optical fibre has been proven
to have a reasonably high enough elastic yield when subjected to a tensile force. Optical
fibres are also considered to have the mechanical properties, which permit them to with-
stand deformation under a reasonable extension, and hence when experiencing geometrical
changes, it will change the optical radiation characteristics of the light that they may trans-
mit through them. Since light is characterized by amplitude (intensity), phase, frequency
and polarization, any one of these parameters may be singled out to be monitored as it
might undergo changes. The paper reports on research which aimed to develop a system
that can detect and warn of a crack initiating and propagating within a critical component.
When the optical fibre that transmits a light beam is stretched, the light beam encounters
the necked section of the fiber, whereby an analogous effect of a reduced aperture occurs.
Utou 1 shows, that the input/output optical power transmission ratio, with respect to the
change in optical fiber diameter, can be predicted by:
P0 △d α
=1−( ) , (13.1)
Pr d
where Pr and P0 are the optical power through the fiber before and after deformation,
respectively, △d and d are the reduced and original diameters of the optical fiber, and a is
the attenuation coefficient per unit length of the optical fibre material. Recalling that the
elongation of the optical fibre equals the crack opening of the host specimen (expressed in
terms of elastic and plastic behaviour of the host material), at the position where the fibre
is fixed, one can obtain a relationship of the optical fibre diameter to the applied load, fibre
properties (subscripted with f), and physical parameters of the host material10 as follows:
√
4 K2 0.4(W − a)vp
df = [ I + ]. (13.2)
πFf lf Ef 2σE 0.4W + 0.6a

The change in optical fiber diameter is obtained by differentiating the above equation with
respect to the applied force with all other quantities forming a constant C:
√
C 1 3/2
df = , △df = − C 1/2 Ff . (13.3)
Ff 2

1 F.
Utou, Fibre sensors ensuring structural integrity, Doctorate Thesis, Cape P. Uni. of Tech, Cape Town, South
Africa, 2005.
 CONCLUSIONS 79

√
4 KI2 0.4(W −a)Vp
where C = πlf Ef [ 2σE + 0.4W +0.6a , KI is fist fracture mode intensity factor, Vp is
plastic components of the measured displacement, and Ff is the axial applied force on
fibre.
Henw conbine the above equations, we obtain an expression, which characterizes the
optical power output through the fibre that has deformed as a result of the host specimen
that experiences a propagating crack.
[ √ ]α
P0 △Ff C
= 1− . (13.4)
Pr 2d Ff

Figure 13.2 Comparison of experimental and theoretical prediction for optical power through a
fibre subjected to elongation owning to an axial force.

Figure 13.2 above depicts the predicted and experimental values of the optical power
ratio through the optical fibre, as it elongates by being fixed across the path of a crack that
propagates and widens in the host specimen.

13.2 Conclusions

The sensor, based on the principles of fiber optics, may be embedded in a critical section of
a component monitoring structural health. In particular, the sensor, by being capable of de-
tecting unwanted excessive distortion, displacement, initiation and propagation of a crack,
presents us with a promising method that warns of impeding catastrophic events. Research
on the subject accomplished the development of a complete system, including the sensor,
and the light emitting and detecting modules. While there is a fair degree of confidence in
the experimental work that was conducted, the theoretical analysis aimed at predicting the
optical output power through the optical fibre that was subjected to dimensional changes,
requires further attention towards increasing the accuracy of the predicted results.
Further work is also envisaged in the process of not only detecting and monitoring crack
propagation with optical fibres, but also with regard to incorporation of shape memory
alloy wires that are embedded (as part of the host material) to form a smart structure,
which heals or retards the crack mechanism.
REFERENCES

1. Crawley E F and Anderson E H, Detailed models of piezoelectric actuation of beams, J. Intell.

Mater. Syst. Struct., 1 4-25, 1990.
2. Wada B K, Fanson J L, and E F Crawley, Adqptive Structures, Journal of Intelligent Material
Systems and Structures,Vol 1 No.2 pp 157-174, 1990.
3. Gandhi, M.V. and Thompson, B S., Smart Materials and Structures, Chapman and Hall,
1992.
4. Crawley E F, Intelligent Structures for Aerospace: A Technology Overview and Assessment,
AIAA Journal, Vol.32 No.8, pp1689-1699, August, 1994.
5. Vijay K. Varadan, et al, Smart Material Systems and MEMS, John Wiley and Sons, 2006.
6. Mel Schwartz, The Encyclopedia of Smart Materials, John Wiley and Sons, 2002.
7. Udd E, Fiber Optic Smart Structure, John Wiley and Sons Ltd, 1995.
8. http://www.piezo.com/，http://www.mide.com/
9. Marino Kekana and Bohua Sun, Smart materials and structures, The South African Mechani-
cal Engineer, Vol 48, 13-17, July 1998
10. Bohua Sun, Some problems of piezoelectric sensor mechanics, Journal of Wave-Material In-
teraction, Vol.14, No.1-2, January/April 1999.
11. B Sun and D Huang, On the feedback control gain of smart composite beams based on Mindlin-
Reissner lamination theory, R and D Journal, 16(2), 41-45, 2000.
12. Bohua Sun and D Huang, Analytical vibration suppression analysis of composite beams with
piezoelectric laminae, Smart Materials and Structures, 9, 751-760, 2000.
13. D Huang and B Sun, Approximate Analytical Solutions of Smart Composite Mindlin Beams,
Journal of Sounds and Vibration, 244(3), 379-394, 2001.

Smart Materials and Structures. 81

By Bohua Sun Copyright ⃝ c 2017 Bohua Sun
82 REFERENCES

14. Bohua Sun and Yan Qiu, Analysis of circular shape distributed piezoelectric actuators, Com-
posite Structures, 62,177-191, 2003.
15. Bohua Sun and Yan Qiu, Rectangular shape distributed piezoelectric actuator: analytical analy-
sis, Smart Mater. Struct, 13, 337-349, 2004.
16. Bohua Sun, Overview of smart materials and structures, World Engineers Convention 2004,
Shanghai, China, November 2-6, 2004.
17. A.N. Mtawa, Bohua Sun and J Gryzagoridis, An investigation of the influence of substrate
geometry and material properties on the performance of the C-shape piezoelectric actuator,
Smart Mater. Struct., 16, 1036–1042, 2007.
18. A.N. Mtawa, Bohua Sun and J Gryzagoridis., Effect of substrate to piezoceramic layer thick-
ness ratio on the performance of a C-shape piezoelectric actuator, Sensors and Actuators: A.
Physical, Vol 141/1, 173-181, 2007.
19. F E Utou, J Gryzagoridis and B Sun Parameters affecting the performance of fiber optic dis-
placement sensors，Smart Mater. Struct, 15, S154-S157, 2006.
20. J Gryzagoridis, F Utou and Bohua Sun, Monitoring structural integrity using fibre optic sensors,
Insight, The British Institute of Non-Destructive Testing, Vol49, No5, 264-266, May 2007.
21. F.E. Utou, J.Gryzagoridis and B.Sun, Detection of Crack Travel and Opening of a Standard
Notched Specimen Mode I Using Embedded FOS Device, Applied Mechanics and Materials,
Vol. 442, pp 360-366, 2007.
22. Bohua Sun, Smart Materials and Smart Systems, Inaugural Professorial Lecture, Cape Penin-
sula University of Technology, August 31, 2001.
23. Da Huang, Approximate analytical solution for vibration control of smart composite beams,
Master Thesis, Cape Peninsula University of Technology, 1999.
24. Yan Qiu, Distributed piezoelectric actuator with complex shape, Master’s Thesis. Cape Penin-
sula University of Technology, 2002.
25. O Philander, The development of a computational design tool for use in the design of SMA
actuator systems, Doctorate Thesis, Cape Peninsula University of Technology, 2004.
26. F E Utou, Fiber optical sensors ensuring structural integrity, Doctorate Thesis. Cape Peninsula
University of Technology, 2005.
27. A N Mtawa, Influence of geometry and material properties on the optimum performance of the
C-shape piezo-composite actuator, Doctorate Thesis. Cape Peninsula University of Technolo-
gy, 2007.
28. Fei Guo, Micromachined capacitive accelerometer with crab-shape, Master’s Thesis. Cape
Peninsula University of Technology, 2005
29. Rui Zhang, mechanics of micromachined bridge-type accelerometer, Master’s Thesis. Cape
Peninsula University of Technology, 2005
30. in Wang, Mechanics of micro capacitive accelerometer with U-shape cantilever beam, Master’s
Thesis. Cape Peninsula University of Technology, 2005.
31. Jose Olivera, Deformation and damage analysis of composite beams equipped with polyvinyli-
dene fluoride film sensors, Master’s Thesis. Cape Peninsula University of Technology, 2008.
32. Lifeng Han, The Mechanical Analysis of Micromachined Accelerometer with Piezoelectric
Thin Films Read-out, Master’s Thesis. Jinan University, 2006
33. Yuyang Jiang, Mechanics analysis of capacitive gyro with electrostatic actuation, Master The-
sis. Jinan University, 2008
34. Fangyin Diao, Mechanics analysis of gyroscopic decelerator, Master’s Thesis. Jinan University,
2009.
 REFERENCES 83

35. Fei Chen, Mechanics analysis of uncooled nonlocal infrared sensor, Master’s Thesis. Jinan
University, 2010.
36. Zhaojun Qin, Studies based on one-plane system of electro-wetting on dielectric (EWOD), Mas-
ter’s Thesis. Jinan University, 2011.
37. Ding Wang, Interaction of protein-protein interaction of Hemagglutinin(HA) and Neu-
raminidase(NA) on surface of influenza a virus, Master’s Thesis. Jinan University, 2010.
 84 REFERENCES

ETH Zürich (Swiss Federal Institute of Technology in Zürich) is an engineering, sci-

ence, technology, mathematics and management university in the city of Zürich, Switzer-
land. Like its sister institution EPFL, it is an integral part of the Swiss Federal Institutes of
Technology Domain (ETH Domain) which is directly subordinate to Switzerland’s Federal
Department of Economic Affairs, Education and Research.
ETH Zürich is consistently ranked among the top universities in the world. It is currently
ranked as the 3rd best university in the world in engineering, science and technology, just
behind the Massachusetts Institute of Technology and Stanford University in the QS World
University Rankings.
Twenty-one Nobel Prizes have been awarded to students or professors of the Institute
in the past; the most famous of whom is Albert Einstein in 1921, and the most recent is
Richard F. Heck in 2010. It is a founding member of the IDEA League and the International
Alliance of Research Universities (IARU) and a member of the CESAER network.
The school was founded by the Swiss Federal Government in 1854 with the stated mis-
sion to educate engineers and scientists, serve as a national center of excellence in science
and technology, and provide a hub for interaction between the scientific community and
industry.

The Cape Peninsula University of Technology, a university in Cape Town, South Africa,
is the only university of Technology in the Western Cape province with over 32,000 stu-
dents.
It was formed in January 2005, resulting from a merger between the Cape Technikon
and the Peninsula Technikon, following years of change in the higher education landscape
of South Africa. In 1993 the Technikons Act was promulgated, which allowed Technikons
to offer Bachelor’s degrees, Master’s and Doctoral degrees in Technology. In March 2001
Kader Asmal (then Minister of Education) announced the National Plan on Higher Educa-
tion, and in May 2002 he announced the possible merger of the two institutions, with the
national working committee also recommending that the University of the Western Cape
should be included in the merger. Towards the end of 2002 the final merger was announced,
and in October 2003 the new name was approved.

View publication stats