Arama Motoru Optimizasyon Baslangic Rehberi 2

Engineering Materials and Processes
Series Editor
Professor Brian Derby, Professor of Materials Science
Manchester Materials Science Centre, Grosvenor Street, Manchester, M1 7HS, UK

Other titles published in this series
Fusion Bonding of Polymer Composites
C. Ageorges and L. Ye
Composite Materials
D.D.L. Chung
Titanium
G. Ltjering and J.C. Williams
Corrosion of Metals
H. Kaesche
Corrosion and Protection
E. Bardal
Intelligent Macromolecules for Smart Devices
L. Dai
Microstructure of Steels and Cast Irons
M. Durand-Charre
Phase Diagrams and Heterogeneous Equilibria
B. Predel, M. Hoch and M. Pool
Computational Mechanics of Composite
Materials
M. Kamiski
Gallium Nitride Processing for Electronics,
Sensors and Spintronics
S.J. Pearton, C.R. Abernathy and F. Ren

Materials for Information Technology
E. Zschech, C. Whelan and T. Mikolajick
Fuel Cell Technology
N. Sammes
Casting: An Analytical Approach
A. Reikher and M.R. Barkhudarov
Computational Quantum Mechanics for
Materials Engineers
L. Vitos
Modelling of Powder Die Compaction
P.R. Brewin, O. Coube, P. Doremus
and J.H. Tweed
Silver Metallization
D. Adams, T.L. Alford and J.W. Mayer
Microbiologically Influenced Corrosion
R. Javaherdashti
Modeling of Metal Forming and Machining
Processes
P.M. Dixit and U.S. Dixit
Modelling Stochastic Fibrous Materials
with Mathematica

W.W. Sampson

Vitaly Yu. Topolov Christopher R. Bowen
Electromechanical Properties
in Composites Based
on Ferroelectrics
123
Professor Vitaly Yu. Topolov
Department of Physics
Southern Federal University
5 Zorge street
344090 Rostov-on-Don
Russia

Dr Christopher R. Bowen
Department of Mechanical Engineering
Faculty of Engineering and Design
University of Bath
Bath BA2 7AY
United Kingdom

ISBN 978-1-84800-999-8 e-ISBN 978-1-84882-000-5
DOI 10.1007/978-1-84882-000-5
Engineering Materials and Processes ISSN 1619-0181
A catalogue record for this book is available from the British Library
Library of Congress Control Number: 2008934905
2009 Springer-Verlag London Limited
Apart from any fair dealing for the purposes of research or private study, or criticism or review, as
permitted under the Copyright, Designs and Patents Act 1988, this publication may only be repro-
duced, stored or transmitted, in any form or by any means, with the prior permission in writing of the
publishers, or in the case of reprographic reproduction in accordance with the terms of licences issued
by the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms should be
sent to the publishers.
The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of
a specific statement, that such names are exempt from the relevant laws and regulations and therefore
free for general use.
The publisher makes no representation, express or implied, with regard to the accuracy of the infor-
mation contained in this book and cannot accept any legal responsibility or liability for any errors or
omissions that may be made.
Cover design: eStudio Calamar S.L., Girona, Spain
Printed on acid-free paper
9 8 7 6 5 4 3 2 1
springer.com
To our relatives, teachers and pupils
Preface
Composite material, also called Composite, a solid material that
results when two or more different substances, each with its own
characteristics, are combined to create a new substance whose
properties are superior to those of the original components in a
specific application.
Encyclopdia Britannica 2006 Ultimate Reference Suite
... functional composites make use of a number of underlying
ideas including connectivity patterns leading to field and force
concentration; the use of periodicity and scale in resonant structures;
the symmetry of composite structures and its influence on physical
properties; polychromatic percolation and coupled conduction paths;
varistor action and other interfacial effects; sum, combination, and
product properties; coupled phase transformation phenomena; and
the important role that porosity and inner composites play in
composite materials. These ideas provide a basic understanding of
functional composites and have been discussed previously.
R.E. Newnham
After the discoveries of piezoelectricity (1880) and ferroelectricity (1921), many
attempts have been made to obtain new high-effective materials with predictable
physical properties and to develop available technologies. Engineered composites
based on ferroelectrics (i.e., composites with ferro-, piezo- and pyroelectric
properties) have been manufactured and studied since the late 1970s. The 1980s
and 1990s have been concerned with an active development of a previously
untapped piezo-composite landscape in order to study interconnections between
microstructure, composition and properties in novel piezo-active materials,
especially those based on ferroelectrics that belong to so-called smart materials.
The modern piezo-composite landscape is a specific area of research and
innovations, and research developments in this area depend on multi-disciplinary
efforts of scientists and engineers, such as solid state physics, physics of
ferroelectrics, solid-state chemistry, materials science, mechanics of heterogeneous
media, mechanics of electro-elastic media, mechanical engineering, etc. On
studying the area of piezo-composites, it becomes obvious that knowledge of these
materials and prediction of their properties cannot be attained without the good
physico-mathematical basis and understanding of principles of modern materials
science and mechanical engineering.
To the best of our knowledge, there are few books concerned with the piezo-
composites and their effective properties, and these problems were mainly
considered in separate chapters or sections only. Earlier, two books devoted to
methods for prediction of the effective electromechanical properties for some
viii Preface
connectivity patterns of the two-component piezo-composites were written in
Russian and published in the USSR (Khoroshun LP, Maslov BP, Leshchenko PV,
(1989) Prediction of Effective Properties of Piezo-active Composite Materials.
Naukova Dumka, Kiev) and Russia (Sokolkin YuV, Pankov AA, (2003)
Electroelasticity of Piezo-composites with Irregular Structures. Fizmatlit,
Moscow). In such a situation the present book aims to fill a gap in piezoelectric
materials science and be an international edition devoted to different groups of
piezo-composites and their effective properties.
Our book is intended to discuss recent original theoretical and experimental
results on effective electromechanical (piezoelectric, dielectric and elastic)
properties in the piezo-composites based on ferroelectrics. For the last decades,
single-crystal, bulk ceramic and thin-film ferroelectrics have found a number of
applications as a result of their remarkable piezoelectric properties. Piezoelectricity
as a physical phenomenon is concerned with an electromechanical interaction in
acentric single crystals including ferroelectrics. Recent work performed in the
rapidly growing field of smart materials demonstrates that both ferroelectricity and
piezoelectricity represent an important link between solid-state science and
engineering.
The effective electromechanical properties associated with a complicated
distribution of both the electric and elastic fields in piezo-composite samples
depend on a number of factors, such as microstructure, connectivity, chemical
composition and volume fraction of components, poling conditions for ferroelectric
components, etc. Of particular interest are the non-monotonic dependences of the
effective electromechanical properties and related parameters on a series of factors,
such as the shape, spatial distribution and volume fraction of inclusions, porosity of
a matrix, and poling degree of ferroelectric components. Knowledge of the extreme
points of the effective parameters is important for manufacturing the high-
performance piezo-composites for their particular piezotechnical applications. The
characterisation and evaluation of smart materials in general requires a solid basis
for the description of interconnections in the well-known dependency triangle of
composition structure properties. For the piezo-composites based on
ferroelectrics, this triangle becomes intricate as a result of electromechanical
coupling, complex interrelations between the internal mechanical (tensor) and
electric (vector) fields and a dependence of the properties in ferroelectric
components on domain structure, microstructure, heterophase states, and external
fields. Moreover, even in ferroelectric single crystals pertaining to one structural
type, the piezoelectric and dielectric properties measured in weak electric fields
vary significantly.
The present book summarises and generalises a series of the authors works on
the problem of prediction and non-monotonicity of the effective electromechanical
properties in different two- and three-component composites based on the
traditional ferroelectric ceramics and the newly available relaxor-ferroelectric
single crystals. Therein we pay due attention to the analysis of interrelations
between electromechanical constants of the components and to the description of
different analytical schemes of averaging the properties of these materials having
different connectivity and microgeometrical characteristics. Undoubtedly, this
book is concerned with the important topic related to smart materials and
Preface ix
optimisation of their effective properties. The main aim of the book is to show not
only the advantages of different methods for predicting the electromechanical
properties and choosing the optimum components, but to demonstrate the non-
trivial behaviour of specific composite architectures and their parameters that are
valuable for transducer, sensor, actuator, hydroacoustic, and other applications.
This book has been written on the basis of the authors research results obtained
at the Rostov State University (Russia, until December 2006), Southern Federal
University (Russia, since December 2006), University of Bath (UK), and Karlsruhe
Research Centre (Germany). The academic style of presentation of the results and
the discussion about these results indicate that the book would be useful to
engineers, postgraduate students, researchers, and lecturers working in the field of
ferro-, piezoelectric and related materials, dealing with methods of mechanics of
heterogeneous materials and with their different applications. This book will be of
benefit to all specialists looking to understand the modelling and detailed
behaviour of the electromechanical response of the piezo-composites and related
smart materials. Some chapters and sections of the book could serve as a basis for a
university course devoted to piezo-active materials and their properties. In this
connection one can mention the well-known sentence Knowledge itself is power
(F. Bacon). Based on our knowledge and experience, we hope that the 21st
century, being called The Century of New Materials and Technologies, will favour
the further fruitful development of the piezo-composite landscape and new
scientific directions therein and in adjacent areas.
Rostov-on-Don, Russia Prof. Dr. Vitaly Yu. Topolov
Bath, United Kingdom Dr. Christopher R. Bowen
July 2008
Acknowledgements
The authors are grateful to Prof. Dr. A. V. Turik, Prof. Dr. A. E. Panich, Prof. Dr.
V. G. Gavrilyachenko, Prof. Dr. A. A. Grekov, Dr. V. V. Eremkin, and Dr. V. G.
Smotrakov (Southern Federal University, Russia), Prof. Dr. R. Stevens and Prof.
Dr. A. Mileham (University of Bath, UK), Dr. S. A. Wilson (Cranfield University,
UK), Prof. Dr. O. Kraft and Dr. M. Kamlah (Karlsruhe Research Centre,
Germany), Dr. W. S. Kreher (Technical University of Dresden, Germany), Dr. H.
Berger (University of Magdeburg, Germany), Prof. Dr. M. Lethiecq and Dr. F.
Levassort (University of Tours, France), Prof. Dr. P. Bisegna (University of Rome
Tor Vergata, Italy), Prof. Dr. H. L. W. Chan (The Hong Kong Polytechnic
University, P.R. China), Prof. Dr. J. H. Huang (Feng Chia University, Taiwan,
Republic of China), Prof. Dr. W. Cao (The Pennsylvania State University, USA),
Prof. Dr. N. Kikuchi (The University of Michigan, USA), Prof. Dr. A. Safari and
Dr. E. K. Akdogan (Rutgers The State University of New Jersey, USA), Prof. Dr.
E. C. N. Silva (University of So Paulo, Brazil), and Prof. Dr. J. S. Ono Fonseca
(Federal University of Rio Grande do Sul, Brazil) for their interest in the research
problems and for their interesting papers that have been cited in the book.
Furthermore, the authors emphasise the vast geographic area wherein the piezo-
composites are studied or manufactured for various applications. Dr. C. R. Bowen
would also like to thank past supervisors including Prof. Dr. B. Derby (University
of Manchester, UK), Prof. Dr. N. Claussen (Technical University of Hamburg-
Harburg, Germany) and Prof. Dr. R. Stevens (University of Bath, UK).
Special thanks are extended to Prof. Dr. R. E. Newnham and his co-workers
(The Pennsylvania State University, USA) for the pioneering work that stimulated
research activity in the field of piezo-composites and other smart materials in the
last decades. The authors express much thanks to Dr. A. Doyle and Mr. S. Rees
(Springer London Ltd, UK) for their effective and timely co-operation in the field
of editing and producing this book. The authors sincerely thank Mrs. P. Carruthers
and Mr. D. Barker (University of Bath, UK) for the technical help and Mrs. D.
Stern (Karlsruhe Research Centre, Germany) for the bibliographic help. Copyright
permissions obtained from Prof. Dr. P. Bisegna (University of Rome Tor
Vergata, Italy), Springer (www.springer.com), Elsevier (www.elsevier.com), IOP
Publishing (www.iop.org), AIP (www.aip.org), Taylor & Francis
xii Acknowledgements
(www.informaworld.com), and Maney Publishing (www.maney.co.uk) are
acknowledged with due attention and gratitude.
Finally, financial support that promoted the fruitful research collaboration and
writing this book is acknowledged with many thanks. Hereupon gratefully and
proudly the authors mention the timely and effective support from the Royal
Society (London, UK), Karlsruhe Research Centre and Helmholtz Society
(Eggenstein-Leopoldshafen, Germany), EPSRC (Swindon, UK), National Physical
Laboratory, QinetiQ (UK), Great Western Research (GWR, UK), University of
Rome Tor Vergata (Rome, Italy), Rostov State University, and Southern Federal
University (Rostov-on-Don, Russia).
Contents
1 From Smart Materials to Piezo-composites.................................................... 1
1.1 Piezo-composites as an Important Group of Smart Materials.................... 1
1.1.1 Smart Materials .............................................................................. 1
1.1.2 Composites ..................................................................................... 1
1.1.3 Piezo-active Composites ................................................................ 2
1.2 Classification of Composite Materials....................................................... 4
1.2.1 Connectivity ................................................................................... 4
1.2.2 Sizes and Shape .............................................................................. 4
1.2.3 Arrangement ................................................................................... 4
1.2.4 Classification, Connectivity Patterns and Specific Examples ........ 5
References........................................................................................................... 9
2 Effective Electromechanical Properties in Piezo-composites...................... 11
2.1 Piezoelectric Medium and Its Properties ................................................. 11
2.2 Sum, Combination and Product Properties .............................................. 22
2.2.1 Sum Properties.............................................................................. 22
2.2.2 Combination Properties ................................................................ 22
2.2.3 Product Properties ........................................................................ 24
2.3 Methods for Evaluation of Effective Parameters ..................................... 25
2.4 Evolution of DE Connectivity Patterns .................................................. 30
2.5 Modelling of Effective Properties in Piezo-composites
with Planar Interfaces ............................................................................. 32
2.6 Connectivity Links Properties............................................................ 35
References......................................................................................................... 35
3 Non-monotonic Volume-fraction Dependences of Effective Properties in
DE Ceramic / Polymer Piezo-composites..................................................... 43
3.1 Composites with 22 Connectivity, Manufacturing and Applications.... 43
3.1.1 Effective Electromechanical Constants of 22 Composites......... 46
3.1.2 Examples of Volume-fraction Dependences Predicted
for 22 Composites....................................................................... 48
3.2 Composites with 13 Connectivity, Manufacturing and Applications .... 52
xiv Contents
3.2.1 Examples of Volume-fraction Dependences Predicted
for 13 Composites........................................................................ 54
3.2.2 Diagrams of Changes in Piezoelectric Properties Predicted
for 13 Composites........................................................................ 57
3.2.3 Maxima of Squared Figures of Merit of 13 Composites............. 61
3.3 Composites with 03 Connectivity, Manufacturing and Applications.... 62
3.3.1 Modelling of the Effective Properties in 03 Composites
with Spheroidal Inclusions............................................................ 63
3.3.2 Piezoelectric Properties in 03 Composites
Based on PbTiO
3
-type Ceramics................................................... 65
3.3.3 03 Composites with a Hierarchy of Inclusions........................... 75
3.3.4 03 Composites with Planar Microgeometry................................ 82
3.4 Composites with 33 Connectivity, Increasing the Piezoelectric
Sensitivity and Applications.................................................................... 85
3.5 Connectivity Non-monotonic Behaviour Manufacturing................... 92
References......................................................................................................... 93
4 Piezoelectric Response of Porous Ceramic and Composite Materials
Based on (Pb, Zr)TiO
3
.................................................................................. 101
4.1 Porous Materials, General Characteristic and Manufacturing ............... 101
4.2 Volume-fraction Dependences of Parameters of Porous
Piezo-active Materials ........................................................................... 103
4.3 Model of the Modified Layered Composite:
Application to Porous Materials ............................................................ 110
4.4 Comparison of Calculated and Experimental Results............................ 111
4.5 From 33 Connectivity to Porous Materials.......................................... 116
References ....................................................................................................... 118
5 Effective Properties in Novel Piezo-composites Based on
Relaxor-ferroelectric Single Crystals.......................................................... 123
5.1 Relaxor-ferroelectric Solid Solutions and Engineering ......................... 123
5.2 03 Single Crystal / Ceramic Composites.............................................. 125
5.3 013 Single Crystal / Ceramic / Polymer Composites..........................132
5.4 13 Single Crystal / Ceramic Composites.............................................. 142
5.5 103 Single Crystal / Polymer Composites ..........................................150
5.6 High Performance in Wide Ranges........................................................ 154
References ....................................................................................................... 155
6 Comparison of Results on Two-component Piezo-composites.................. 159
6.1 The Pros and Cons of Different Methods .............................................. 159
6.2 03 Versus 22 or 13............................................................................ 160
6.3 Data on 13 Composites.........................................................................165
6.3.1 Composites with Cylindrical Ceramic Rods................................ 165
6.3.2 Composites with Parallelepiped-shaped Ceramic Rods...............168
6.3.3 Variational Bounds for Effective Properties
in 13 Composites with Cylindrical Ceramic Rods.................... 171
Contents xv
6.3.4 Variational Bounds for Effective Properties
in 13 Composites with Parallelepiped-shaped
Ceramic Rods................................................................................173
6.4 Data on 03 Composites .........................................................................175
References ....................................................................................................... 182
7 Conclusions.................................................................................................... 185
7.1 From Components to Effective Electromechanical Properties .............. 185
References ....................................................................................................... 186
Appendix Formulae for Effective Electromechanical Constants
of 22 and 13 Piezo-composites ....................................................................... 187
References .............................................................................................................189
List of Abbreviations.......................................................................................... 191
About the Authors .............................................................................................. 193
Index......................................................................................................................195
1
From Smart Materials to Piezo-composites
1.1 Piezo-composites as an Important Group of Smart Materials
1.1.1 Smart Materials
In recent decades the structure and properties of smart materials (or intelligent
functional materials) has represented a thriving research topic for academia and
industry. This topic is transdisciplinary and attracts the attention of researchers in
the fields of materials science, solid-state physics, mechanics, chemistry,
crystallography, and mechanical engineering. Smart materials are able to adjust
their behaviour to changes in external parameters and show clear analogies to a
variety of biological systems [1, 2]. They can adapt functionality in an intelligent
fashion and find applications as diverse as aeronautics to consumer products [2].
The main directions for the exploitation of smart materials are related to their
performance for transducer, sensor or actuator applications. Advances in materials
science and technology of the 21st century are inconceivable without smart
materials and include ferroelectric, piezoelectric and pyroelectric ceramics and
composites, electrostrictive and magnetostrictive materials, shape memory alloys,
and relaxor-ferroelectrics [1, 2]. The physical properties of smart materials can be
tailored [1, 3, 4] by understanding the composition structure properties
relationships of the materials (Figure 1.1) that ultimately enable the optimisation of
properties for specific applications.
1.1.2 Composites
The majority of smart materials are solid-state heterogeneous or micro-
inhomogeneous materials and include ceramics, polycrystalline films and
composites. Composites can be defined as heterogeneous systems that consist of
two or more components that differ in chemical composition and properties and are
separated by distinct interfaces. Each composite material is characterised by the
two following features [5, 6]. Firstly, the typical size of the separate structural
elements are small in comparison with the whole composite sample (i.e., there is an
element of micro-inhomogeneity). Secondly, the typical sizes of the structural
2 Electromechanical Properties in Composites Based on Ferroelectrics

Properties

Structure Composition
Figure 1.1. Composition structure properties triangle
elements are greater than the size of their individual atoms or molecules so that
each component is regarded as a continuous medium. As a result, to describe the
composite properties it is possible to consider the physical laws and equations
suitable for continuous media.
Each component of the composite represents a set of the structural elements
with equal physical constants. These elements can be distributed regularly
(uniformly) or irregularly and can have a specific, random or variable shape. The
composite as a whole can be described using a set of microstructural
characteristics, for example, connectivity, volume fractions of the components,
spatial distribution of the components, and other parameters [69].
1.1.3 Piezo-active Composites
The piezo-active composites (hereafter referred to as piezo-composites) are an
important group of smart materials. The piezo-composite group is vast due to the
large number of components that may be involved in the design of the composites.
Among these components, ferroelectrics play a leading role [1, 3, 4] and modern
processing technologies enable the synthesis of a variety of ferroelectric single
crystals, ferroelectric ceramics (or polycrystalline materials), ferroelectric films,
and nano-sized ferroelectrics. Ferroelectric properties are generally related to
altering the spontaneous polarisation vector using an external electric field (domain
switching) and such a process takes place in acentric (non-centrosymmetric) crystal
structures. These crystal structures can be divided by symmetry criteria into 21
point groups [3, 4] and piezoelectric properties due to the appearance of positive
and negative charges on the surfaces of a mechanically stressed sample are
particular to the acentric single crystals belonging to the following 20 point groups
[4, 10]: 1 (triclinic system), 2 and m (monoclinic system), 222 and mm2
(orthorhombic system), 3, 32 and 3m (trigonal or rhombohedral system), 4, 422,
4mm, 4 , and 4 2m (tetragonal system), 6, 622, 6 , and 6 m2 (hexagonal system),
and 23 and 4 3m (cubic system). The piezoelectric effect can also take place in
From Smart Materials to Piezo-composites 3
acentric anisotropic media, so-called piezoelectric textures, belonging to one of the
following Curie groups [11]: f, fmm or f/2. Ferroelectric ceramics exhibit
piezoelectricity in the poled state and are piezoelectric textures with fmm
symmetry.
Among the ferroelectric ceramics, the perovskite-type compositions are of
practical interest due to their excellent piezoelectric and dielectric properties, high
electromechanical coupling and large piezoelectric anisotropy [3, 4, 6, 1113]. A
large number of ferroelectric ceramic compositions have been studied over the past
60 years, including BaTiO
3
, Pb(Zr
1x
Ti
x
)O
3
and (1 x)Pb(A
1/3
Nb
2/3
)O
3
xPbTiO
3
,
where A = Mg or Zn. The latter solid solutions exhibit relaxor-ferroelectric
properties and various heterophase states near the morphotropic phase boundary
(MPB) and exhibit very high piezoelectric activity in the single-crystal form with
engineered domain structures [14, 15]. The piezoelectric properties of the
perovskite-type and other ferroelectric ceramics are highly dependent on
manufacturing and poling methods, temperature history, mechanical stress history,
and microstructure [3, 4, 13]. In the perovskite-type ceramics the piezoelectric
charge coefficient (that characterises a charge per unit force) d
33
a 10
2
pC / N [3, 4,
12, 13] and is two orders of magnitude higher than the piezoelectric coefficients d
ij
of quartz (d
11
= 2.25 pC / N and d
14
= 0.853 pC / N [11]), which is an example of
a classic non-ferroelectric piezoelectric single crystal. In single crystals of (1
x)Pb(A
1/3
Nb
2/3
)O
3
xPbTiO
3
with compositions chosen near the MPB, typical
experimental values are even higher with d
33
a 10
3
pC / N [14, 15].

Piezo-composites based on ferroelectrics are the important group of smart
materials due to the ability to vary the microgeometry, effective physical properties
and their anisotropy across a wide range. These significant variations appear upon
combining different components, for example, (i) ceramic and polymer, (ii) single
crystal and ceramic or (iii) single crystal and polymer. The presence of one or more
piezoelectric components enables a set of parameters or properties to be attained
that are not particular to the separate components, and this synergy is important for
many practical transducer applications [1, 4].
Piezoelectricity is an interaction between elastic variables (mechanical stress V
kl

and mechanical strain [
qr
) and electric variables (electric field E and electric
displacement D) [10, 11] and the interaction can be weak or strong in
homogeneous piezoelectric materials, including single-domain ferroelectric single
crystals. The same interaction becomes even more complicated in the piezo-
composites where an electromechanical interaction between the components and a
redistribution of internal electric and mechanical fields strongly depend on
microgeometry and boundary conditions. In general, the piezo-composites based
on ferroelectrics form the final link in hierarchy-of-properties chains of single-
domain single crystal polydomain single crystal ceramic composite and
single-domain single crystal polydomain single crystal heterophase single
crystal composite. As a consequence, the study of piezo-composites and their
effective properties are important in connection with the structureproperties and
compositionproperties sides of the fundamental triangle shown in Figure 1.1.


1.2 Classification of Composite Materials
In order to consider the mechanics and physics of micro-inhomogeneous media it
is first necessary to define the structure of the materials being analysed. In the case
of composites (including piezo-composites), the structure criteria play an important
role and the materials must be classified according to the concepts [69]:
(i) connectivity of the structural elements,
(ii) typical relative sizes and shape of structural elements and
(iii) features of the arrangement of structural elements.

These three concepts will now be described.
1.2.1 Connectivity
Based on knowledge of the connectivity of the structural elements it is possible to
divide the composites into three large groups. The first group are the matrix
composites in which the structural elements of the first component present a
monolithic system (the matrix) and the structural elements of the remaining
components are isolated and do not contact each other at small volume fractions.
These remaining components form a system of isolated inclusions. The second
group comprises composites in which each component presents a monolithic and
interconnected framework. The corresponding composite structures are often
called interpenetrated. In the composites of the third group, the structural elements
of each component are not interpenetrated or isolated. In addition, it is possible to
obtain some structures with the structural elements related to the three different
groups.
1.2.2 Sizes and Shape
With regard to the sizes and shape of the structural elements it is necessary to
consider granular, fibrous and laminated composites. The granular composites
consist of the structural elements having linear dimensions l
i
of one order-of-
magnitude in three co-ordinate directions (i = 1, 2 or 3). If l
1
| l
2
<< l
3
in one of
the components, then the composites are characterised as fibrous. In laminated
composites, a relation l
1
| l
2
>> l
3
holds for one or more components. If the latter
relation is not true for all the components, such a composite can be regarded as a
matrix composite reinforced by plate-like inclusions of one or more components.
Complicated structures that are combinations of granular, fibrous and laminated
structures can also be attained in some composite systems. Typical examples of
such materials are composites reinforced by grains and fibres, porous laminated
and fibrous composites and laminated composites with layers reinforced by fibres.
1.2.3 Arrangement
Taking into consideration the arrangement of the structural elements it is possible
to have many variants of a layout of each component. In this difficult situation,
some degree of regularity of the composite structure becomes a pertinent criterion
for subsequent classification. According to this criterion, composites can be
regarded as materials with a regular (determined) or irregular (stochastic) structure.
The structural elements in the composites with a regular structure are characterised
by a relevant geometric shape (e.g., rectangular parallelepiped or ellipsoid) and a
regular (equidistant and uniform) arrangement in the sample. In the case of the
irregular structure, the arrangement of the structural elements is random, and their
shape can be either appointed or arbitrary. The random character of this
arrangement can change and influence connectivity of the structural elements in
the sample. An additional factor that characterises the arrangement of the
structural elements is related to their spatial orientation. This factor is also to be
taken into account when analysing both microgeometry and connectivity of the
composites, in particular fibrous or laminated composites.
1.2.4 Classification, Connectivity Patterns and Specific Examples
A widespread classification of the two-component (or diphasic
1
) composites with
planar interfaces was first put forward by Newnham et al. [7]. This classification
is based on the concept of connectivity of each component. Connectivity is
regarded as one of the main characteristics of the microstructure and expressed by
the numbers of dimensions (or co-ordinate axes) in which each component is
continuously distributed between limiting surfaces of the composite. The
distribution of a self-connected state of a component can take place along zero,
one, two, or three co-ordinate axes, i.e., connectivity D = 0, 1, 2, or 3 for the
component 1 and connectivity E = 0, 1, 2, or 3 for the component 2. The
connectivity of a two-component composite is written in a general form [1, 7, 9] of
DE where the connectivity of the piezoelectric or most piezo-active component
takes the first position (D). In the case of D d E, the n-component composites are
described by (n + 3)! / (3! n!) connectivities [7, 9], for n = 2 the number of
connectivities is 10. It is also possible to introduce 10 connectivities DE at D t E,
and a scheme of corresponding connectivity patterns is shown in Figure 1.2. Since
D and E can independently vary from 0 to 3, the two-component composite is
described in terms of 4
2
= 16 connectivities as follows: 00, 01, 02, 03, , 3
2, or 33. The scheme shown in Figure 1.2 is also useful for the analysis of
connectivities and microstructural features in the presence of non-planar interfaces.
For example, a composite consisting of a bulk continuous matrix with small
isolated inclusions embedded in the matrix would be described by 30 or 03
connectivity. The term 30 connectivity would be used to describe ferroelectric
ceramics with closed air pores or isolated piezo-passive inclusions, whereas 03
connectivity would be used to describe a ferroelectric ceramic

/

continuous polymer

____________
1
Ferroelectric materials (single crystals and ceramics) can be heterophase, for example,
close to the MPB [3, 1315] in a wide temperature range and the term diphasic would be
substituted by two-component. In this case it is possible to consider a two-component
composite, each component of which could be heterophase, as in the single crystal / ceramic
composite with compositions chosen near the MPB.

Figure 1.2. Scheme of DE connectivity patterns for the two-component composites with
D t E. (XYZ) is the rectangular co-ordinate system, the shaded component is characterised
by D connectivity. Arrows at faces of cubes are used to indicate the connected directions
parallel to the co-ordinate axes (reprinted from Newnham et al. [7], with permission from
Elsevier)
matrix composite with isolated inclusions. If the composite contains three
components, then its connectivity is denoted as DEJ with D, E, J = 0, 1, 2, or 3.
Some examples of the three-component composites based on ferroelectric ceramics
are provided in ref. [1].
The concept of connectivity is fundamental in developing an understanding of
the electromechanical interaction between components within piezo-composites, in
the study of the distribution of internal electric and mechanical fields and in the
interpretation of experimental or calculated data related to composites with specific
connectivity. The connectivity of the piezo-composites is crucial in influencing the
piezoelectric response and electromechanical coupling of these materials.
Moreover, after publishing the pioneering paper [7], connectivities DE, DEJ,
etc., or related connectivity patterns are almost always mentioned when the
research is devoted to the composites based on ferroelectrics or related materials.
Particular examples include piezo-composites with connectivities 03, 13, 22, 2
3, 31, 32, and 33 that have been experimentally studied (see, e.g., refs. [7, 16
21]). Examples of the DE composites based on ferroelectric ceramics are shown
in Figures 1.31.6.

Figure 1.3. Micrograph of the 13 ferroelectric ceramic / polymer composite. eramic
fibres are embedded in epoxy, diameter of each fibre is 8u10
4
m

Figure 1.4. Micrograph of the 03 PbTiO
3
ceramic / epoxy composite with so-called pearl
chains (reprinted from Wilson et al. [18], with permission from IOP Publishing)
The concept of connectivity and the related classification scheme link the
microstructures and effective properties in the piezo-composites and, therefore,
enable us to understand the underlying science of the complex electromechanical
interactions between components connected in various ways. A number of
examples of the effective electromechanical properties of piezo-composites and
their features will be analysed and described in subsequent chapters of this book.


Figure 1.5. Micrograph of the 22 volume-fraction gradient composite that comprises PZT
ceramic and Spurr epoxy layers (reprinted from Safari and Akdogan [20], with permission
from Taylor & Francis Group)
Figure 1.6. Micrographs of the 33 porous ferroelectric ceramics (specific case of
composites). In figure (b) two ceramic layers (each 4 mm thick) are the main elements of the
hydrophone

a

b
References
[1] Newnham RE, (1997) Molecular mechanisms in smart materials. Materials Research
Society Bulletin 22(5):2034
[2] Tani J, Takagi J, Qiu J, (1998) Intelligent materials systems: Application of functional
materials. Applied Mechanics Review 51:505521
[3] Xu Y, (1991) Ferroelectric Materials and their Applications. North-Holland,
Amsterdam London New York Toronto
[4] Uchino K, (1997) Piezoelectric Actuators and Ultrasonic Motors. Kluwer, Boston
Dordrecht London
[5] Chistensen RM, (1979) Mechanics of Composite Materials. Wiley, New York
[6] Khoroshun LP, Maslov BP, Leshchenko PV, (1989) Prediction of Effective Properties
of Piezo-active Composite Materials. Naukova Dumka, Kiev (in Russian)
[7] Newnham RE, Skinner DP, Cross LE, (1978) Connectivity and piezoelectric-
pyroelectric composites. Materials Research Bulletin 13:525536
[8] Sokolkin YuV, Pankov AA, (2003) Electroelasticity of Piezo-composites with
Irregular Structures. Fizmatlit, Moscow (in Russian)
[9] Newnham RE, (1994) Nonmechanical properties of composites. In: Kelly A, Cahn
RW, Bever MB (eds.) Concise Encyclopedia of Composite Materials. Elsevier,
Oxford, pp 214220
[10] Ikeda T, (1990) Fundamentals of Piezoelectricity. Oxford University Press, Oxford
New York Toronto
[11] Zheludev IS, (1971) Physics of Crystalline Dielectrics. Vol. 2: Electrical Properties.
Plenum, New York
[12] Dantsiger AYa, Razumovskaya ON, Reznitchenko LA, Grineva LD, Devlikanova
RU, Dudkina SI, Gavrilyatchenko SV, Dergunova NV, Klevtsov AN, (1994) Highly
Effective Piezoceramic Materials (Handbook). Kniga, Rostov-on-Don (in Russian)
[13] Gorish AV, Dudkevich VP, Kupriyanov MF, Panich AE, Turik AV, (1999)
Piezoelectric Device-Making. Vol. 1: Physics of Ferroelectric Ceramics.
Radiotekhnika, Moscow (in Russian)
[14] Zhang R, Jiang B, Cao W, (2001) Elastic, piezoelectric, and dielectric properties of
multidomain 0.67Pb(Mg
1/3
Nb
2/3
)O
3
0.33PbTiO
3
single crystals. Journal of Applied
Physics 90:34713475
[15] Yin J, Cao W, (2002) Effective macroscopic symmetries and materials properties of
multidomain 0.955Pb(Zn
1/3
Nb
2/3
)O
3
0.045PbTiO
3
Physics 92:444448
[16] Ritter T, Geng X, Shung KK, Lopath PD, Park S-E, Shrout TR, (2000) Single crystal
PZN/PT polymer composites for ultrasound transducer applications. IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control 47:792800
[17] Smay JE, Cesarano III J, Tuttle BA, Lewis JA, (2003) Piezoelectric properties of 3X
periodic Pb(Zr
x
Ti
1-x
)O
3
polymer composites. Journal of Applied Physics 92:6119
6127
[18] Wilson SA, Maistros GM, Whatmore RW, (2005) Structure modification of 03
piezoelectric ceramic / polymer composites through dielectrophoresis. Journal of
Physics D: Applied Physics 38:175182
[19] Akdogan EK, Allahverdi M, Safari A, (2005) Piezoelectric composites for sensor and
actuator applications. IEEE Transactions on Ultrasonics, Ferroelectrics, and
Frequency Control 52:746775
[20] Safari A, Akdogan EK, (2006) Rapid prototyping of novel piezoelectric composites.
Ferroelectrics 331:153179

[21] Wang F, He C, Tang Y, Zhao X, Luo H, (2007) Single-crystal 0.7Pb(Mg
1/3
Nb
2/3
)O
3

0.3PbTiO
3
/ epoxy 13 piezoelectric composites prepared by the lamination technique.
Materials Chemistry and Physics 105:273277

2
Effective Electromechanical Properties in
Piezo-composites
2.1 Piezoelectric Medium and Its Properties
The appearance of an electric polarisation in quartz single crystals that were
subjected to a mechanical load was first reported by brothers P. Curie and J. Curie
in experimental work in 1880 [1]. The effect that links a mechanical action (stress
or strain) with an electric response (electric field, displacement or polarisation) is
called the piezoelectric effect or, more exactly, the direct piezoelectric effect. At a
later date, it was experimentally established that the converse piezoelectric effect is
also observed in acentric single crystals in that an external electric field generates a
mechanical response, i.e., a stress or strain of the sample [2, 3], similar to the
electrostriction of dielectrics [3, 4].
It is generally realised that the piezoelectric effect reflects a linear relationship
between electric and mechanical variables and originates from the displacement of
ions of an acentric single crystal under an applied electrical field [14]. This
relationship enables a change in the sign of the effect by switching the external
electric field, unlike electrostriction that always exhibits a quadratic effect [3, 4]
and does not undergo switching. Despite the linear character, the piezoelectric
response of a single crystal is often intricate owing to various interconnections
between the piezoelectric and other properties such as the elastic, dielectric and
thermal properties [3].
A general consideration of the piezoelectric effect in single crystals is
undertaken using thermodynamic functions, such as Helmholtz free energy, Gibbs
free energy, elastic Gibbs energy, and electric Gibbs energy [24]. Each of these
functions has at least three arguments that characterise the mechanical, electric and
thermal states of the single crystal. The first argument can be either mechanical
stress V
kl
or mechanical strain [
jr
, the second argument can be electric field E or
electric displacement (electric flux density) D, and the third argument can be
temperature T or entropy S. In addition, it is possible to develop additional
arguments in terms of magnetic field H or magnetic induction B [5]. Based on
knowledge of the thermodynamic functions and relations between the arguments of
the mechanical, electric and thermal states, it is possible to describe the linked
response actions in terms of [
jr
, E
i
and T as follows:
V
kl
= [
E
kljr
c
jr
e
ikl
E
i
+ E
kl
'T (2.1)
D
i
= e
ikl
[
kl
+ E
[
H
if
f
+ p
i
'T (2.2)
'S = E
jr
[
jr
+ p
i
E
i
+ (U / T
0
) c 'T (2.3)
In Equations 2.12.3 the constants written in the tensor form are elastic moduli at
constant electric field ( ), piezoelectric coefficients (e
E
kljr
c
ikl
), coefficients of thermal
stress (E
kl
), dielectric permittivities at constant mechanical strain ( ), pyroelectric
coefficients (p
[
H
if
i
), density (U), and specific heat (c). The increments 'T and 'S
denote differences T T
0
and S S
0
, respectively, where T
0
is the initial
temperature and S
0
is the initial entropy of the single crystal. It is assumed that the
linear relations in Equations 2.1 2.3 hold in the case of relatively weak external
fields. For example, the dependence V
kl
([
jr
) from Equation 2.1 obeys Hookes law
in an anisotropic medium at small strains [
jr
. The dependence D
i
(E
f
) from
Equation 2.2 is linear at relatively low levels of the electric field E
f
applied to the
single crystal. With regard to ferroelectric single crystals, similar requirements
concerning the linear dependences V
kl
([
jr
) and D
i
(E
f
) are possible, but the E
f
range
becomes narrower than that in acentric linear dielectric single crystals. It is well
known that in the presence of a low electric field, the polarisation of ferroelectric
single crystals linearly depends on field [3, 4] so that domain-wall displacements
are reversible and no nuclei of reoriented domains will form. In this connection
the electric field E
f
is usually regarded to be a few times lower than the coercive
field E
c
which is determined from a ferroelectric hysteresis loop [3, 4, 6].
Equations 2.1 2.3 enable the determination of a set of isothermal constants for
a single crystal. An analogous set of adiabatic constants can also be derived using
the thermodynamic functions and a combination of three arguments including 'S
(for example, [
jr
, E
i
and 'S). According to experimental data, the difference
between the related isothermic and adiabatic constants of the piezoelectric single
crystals is approximately 1% or less [3, 5].
The effect of thermal fields on the elastic and electric responses of
piezoelectrics is often neglected, and the description of the piezoelectric effect is
carried out on the basis of Equations 2.1 and 2.2 at 'T = 0. These equations
written in the matrix form then reduce to [2, 3]
V
p
= [
E
pq
c
q
e
fp
E
f
(2.4)
D
k
= e
kl
[
l
+ E
[
H
kr
r
(2.5)
and represents the first pair of piezoelectric equations linking the variables of strain
Effective Electromechanical Properties in Piezo-composites 13
and field ([, E). The first term in the right part of Equation 2.5, P
k
= e
kl
[
l
, describes
the piezoelectric polarisation produced by an external strain as a result of the direct
piezoelectric effect. In addition, the converse piezoelectric effect is described by
the e
fp
E
f
term in Equation 2.4. From the thermodynamic treatment of the
interrelations between the electric and elastic fields [2, 3], the converse and direct
piezoelectric effects can also be described by the three following pairs of
equations:
[
p
= V
E
pq
s
q
+ d
fp
E
f
(2.6)
D
k
= d
kl
V
l
+ E
V
H
kr
r
(2.7)
in variables (V, E),
[
p
= V
D
pq
s
q
+ g
fp
D
f
(2.8)
E
k
= g
kl
V
l
+ D
V
E
kr
r
(2.9)
in variables (V, D) and
V
p
= [
D
pq
c
q
h
fp
D
f
(2.10)
E
k
= h
kl
[
l
+ D
[
E
kr
r
(2.11)
in variables ([, D). Superscripts V and D denote measurement conditions at V =
const and D = const, respectively, constants and are determined using
equalities = G
V
E
kr
[
E
kr
V
E
kr
V
H
rv
kv
and = G
[
E
kr
[
H
rv
kv
, and elastic moduli or and elastic
compliances or are related by conditions = G
E
pq
c
D
pq
c
E
pq
s
D
pq
s
E
pq
c
E
qr
s
pr
and = G
D
pq
c
D
qr
s
pr
,
where G
kv
is the Kronecker symbol. It should be noted that Equations 2.42.11
contain four types of piezoelectric coefficients, namely e
kl
, d
kl
, g
kl
, and h
kl
.
The piezoelectric coefficient e
kl
(stress coefficient) relates the electric field to
the mechanical stress (Equation 2.4) and the mechanical strain to the electric
displacement (Equation 2.5). The piezoelectric coefficient d
kl
(charge coefficient)
characterises the relation between the electric field and the mechanical strain
(Equation 2.6) and the relation between the mechanical stress and the electric
displacement (Equation 2.7). The piezoelectric coefficient g
kl
(voltage coefficient)
is introduced to relate the electric displacement to the mechanical strain (Equation
2.8) and the mechanical stress to the electric field (Equation 2.9). Finally, the
piezoelectric coefficient h
kl
(strain coefficient) relates the electric displacement to
the mechanical stress (Equation 2.10) and the mechanical strain to the electric field
(Equation 2.11). In general, each piezoelectric coefficient characterises the
relationship between components of two fields that are described by first- and
second-rank tensors, and all the piezoelectric coefficients introduced in Equations
2.42.11 are defined to be components of the third-rank tensors, but written in the
matrix form (with two subscripts) [26] only. These individual piezoelectric
coefficients can be used to aid in the selection of piezoelectric materials for
specific applications. For example, for an actuator application [4] it may be
necessary to select materials with a high strain per unit applied electric field (i.e., a
high d
kl
). For a pressure sensor, the electric field generated per unit mechanical
stress is likely to be important (i.e., a high g
kl
).
The interrelationships of the piezoelectric coefficients follow from the
thermodynamic description [2, 3] and are expressed by a set of equalities
d
fp
= g
V
H
fk
kp
= e
fq
E
qp
s (2.12)
e
fp
= h
[
H
fk
kp
= d
fq
E
qp
c (2.13)
g
fp
= d
V
E
fk
kp
= h
fq
D
qp
s (2.14)
h
fp
= e
[
E
fk
kp
= g
fq
D
qp
c (2.15)
All four types of piezoelectric coefficients from Equations 2.122.15 are also
involved in relations [2] between dielectric or elastic constants measured on
different conditions:
= d
V
H
kr
[
H
kr
kf
e
rf
(2.16)
= g
[
E
kr
V
E
kr
kf
e
rf
(2.17)
= e
D
pq
c
E
pq
c
fp
h
fq
(2.18)
= d
D
pq
s
E
pq
s
fp
g
fq
(2.19)
The close connections between the elastic, piezoelectric and dielectric constants of
a single crystal (see Equations 2.122.19) or a piezoelectric medium enable them
to be collectively described as electromechanical constants. These constants
characterise electromechanical properties in a variety of piezo-active materials,
such as piezoelectric or ferroelectric single crystals, poled ferroelectric ceramics,
thin ferroelectric films, piezo-composites, etc.
It follows from Equations 2.1, 2.2 and 2.42.11 that the piezoelectric properties
are described by third-rank tensors and are represented in the matrix form. Tables
of matrices of the piezoelectric coefficients of the different symmetry classes have
been collected in a series of monographs [25]. It should be noted that the matrices
of the piezoelectric coefficients || d || and || g || differ from the matrices of the
piezoelectric coefficients || e || and || h || for 3, 32, 3m, 6 , and 6 m2 classes [2, 3].
The conditions for measuring the piezoelectric coefficients are concerned with the
electric and mechanical variables that are fixed for the direct or converse
piezoelectric effect. For example, if the direct piezoelectric effect is considered,
then the corresponding piezoelectric coefficients are measured under the following
conditions [3]: [ = 0 and E = 0 (for d
kl
), [ = 0 and D = 0 (for g
kl
), V = 0 and D = 0
(for h
kl
), and V = 0 and E = 0 (for e
kl
). In the case of the converse piezoelectric
effect, the following conditions should be used: V = 0 and E = const (for d
kl
), V = 0
and D = const (for g
kl
), [ = 0 and D = const (for h
kl
), and [ = 0 and E = const (for
e
kl
). The above-given conditions are determined on the basis of Equations 2.42.11
for the four types of piezoelectric coefficients.
It has been shown that there are close connections between the variety of
electromechanical constants, and two groups of these constants (i.e., piezoelectric /
dielectric and elastic / piezoelectric) are described by Equations 2.122.19. The
elastic, piezoelectric and dielectric constants are factors that ultimately measure the
effectiveness of the conversion of electric energy into mechanical energy and vice
versa. From the analysis of energy conversion [6], the effectiveness strongly
depends on differences between dielectric permittivities from Equation 2.16 or
differences between elastic compliances from Equation 2.19, and these differences
stem from the piezoelectric properties in the medium. In a general form, the
effectiveness of the energy conversion is described by an electromechanical
coupling factor [3, 5, 6]
k = w
piezo
/
mech el
w w
(2.20)
and is expressed in terms of volume densities of piezoelectric (or mutual) energy
w
piezo
= V
i
d
mi
E
m
/ 2, electric energy w
el
= E
k
V
H
kr
E
r
/ 2, and mechanical energy w
mech

= V
p
E
pq
s V
q
/ 2. The k
2
value determined using Equation 2.20 characterises [26] a
ratio of stored mechanical energy to electric energy input or a ratio of stored
electric energy to mechanical energy input. The absolute value of the
electromechanical coupling factor depends [2, 4, 6] on the vibration mode,
measurement methodology, sample shape, and the electromechanical constants of
the piezoelectric material.
To date, we have considered the constitutive equations and the
electromechanical properties in linear piezo-active media. Equations 2.42.11 are
often used to describe the linear behaviour of ferroelectric single crystals (single-
domain, polydomain or unipolar) [3, 4] poled ferroelectric ceramics (piezo-
ceramics) [6, 7] and piezo-active composites [5] under weak electric and
mechanical fields. As discussed by Hall [7], the linear description provides a
reasonable approximation of the functional characteristics of piezo-ceramics at low
levels of the applied electric field and stress, but becomes increasingly inaccurate
as these levels increase. Non-linear behaviour of the electromechanical properties
in various perovskite-type ferroelectric ceramics is observed above the threshold
values [7] of the electric field E
thr
and mechanical stress V
thr
. According to data
[7], the typical values equal E
thr
| 10 kV / m for soft PZT ceramics (i.e., Pb(Zr,
Ti)O
3
-type ceramics with low coercive field E
c
) and E
thr
| 300 kV / m for well-
aged hard PZT ceramics (i.e., having high E
c
values). Similarly, non-linearity of
mechanical properties in both the soft and hard PZT ceramics becomes evident
as mechanical stress exceeds V
thr
| 20 MPa [7].
The hysteresis and non-linear behaviour of electromechanical properties in
ferroelectric single crystals and ceramics are associated with the reorientation of
the spontaneous polarisation vectors P
s
of domains under the external electric field
[3, 4, 6, 7]. The domain structure of ferroelectrics is of interest due to the
significant influence of both the domain arrangement and reorientations on the
electromechanical properties in these materials. Ferroelectric single crystals with
the perovskite-type structure represent the vast group of multiaxial ferroelectrics
and have various 180q and non-180q domain structures in ferroelectric phases [3,
4]. Many examples of complicated domain structures have also been studied [4] in
related perovskite-type ferroelectric ceramics. An effect of the domain structure on
the electromechanical properties in the well-known perovskite-type single crystals
has been studied for cases of the 180q (BaTiO
3
[8]) and 90q (BaTiO
3
[9], PbTiO
3

[10]) domains in the ferroelectric 4mm phase. The domain structures are
characterised [810] as regular laminar structures (Figure 2.1) with planar domain
walls that are motionless under weak external fields.
An orientation contribution caused by the elastic (or reversible) 90q domain-
wall displacements under applied electric fields has been determined [11] for
BaTiO
3
single crystals separated into 90q domains. The contribution of domain
orientation to the electromechanical constants of a single crystal strongly depends
on the mobility of the 90q domain walls within the material. The important role of
the 60q, 90q and 120q domain-wall displacements on the piezoelectric coefficient
d
33
of the polydomain KNbO
3
single crystal (mm2 phase) has been analysed in ref.
[12]. The contributions of domain-wall displacements to the d
33
under an electric
field correlate with some crystallographic characteristics of the non-180q domain
structures.
Full sets of the electromechanical constants of polydomain single crystals are
used to study the relationships between electromechanical properties in a single
grain and a poled ceramic [10, 1315] consisting of a large number of grains with
different orientations of main crystallographic axes. This study is based on solving
a problem [13, 14] of the electromechanical interaction between a spherical piezo-
active grain and a piezo-ceramic medium surrounding this grain. Such an
interaction is considered within the framework of the effective medium method
[13, 14] that takes into account not only the piezoelectric effect but also the
anisotropy of the piezoelectric properties in the grain and the ceramic medium.
This anisotropy can be considerable, particularly in the presence of non-180q
domain reorientations in the grains [16]. Modelling of the 180q and 90q
reorientations under an electric field results in the full set of the electromechanical
constants, calculated for the BaTiO
3
ceramic [16] at various values of P
r
/ P
s
, where
P
r
is the remanent polarisation of the ceramic and P
s
is the spontaneous polarisation
of the separate domain. In addition, the effect of the 90q domain-wall
displacements within the grains on the electromechanical properties in poled
BaTiO
3
ceramic was studied in [11, 17, 18] where contributions from these
domain-wall displacements to a variety of constants for the ceramic were
evaluated. Of particular interest is the significant contribution of domain-wall
displacements (3772%) to the piezoelectric coefficients d
3j
[17, 18]. The
analogous contribution to the elastic compliances is evaluated to be in a range
from 11 to 43 % [18]. Differences between the experimental [6] and averaged [13]
constants of the BaTiO
E
pq
s
3
ceramic were accounted for by the evaluated contributions
[11, 17, 18] from the 90q domain-wall displacements.

a b
Figure 2.1. Schematic representation of the polydomain single crystal with the 180q (a) or
90q (b) domain structure. Spontaneous polarisation vectors of the domains are shown by
arrows
In contrast to BaTiO
3
ceramics, the electromechanical properties in the PbTiO
3
ceramic are determined [10, 15] in the absence of 90q domain-wall displacements
within the grains. According to results [15, 19], the 90q domain structure of the
grains has a strong influence on the piezoelectric coefficients d
fp
and their
anisotropy d
33
/ d
31
in the PbTiO
3
-type ceramics. In particular, it is possible to
attain large piezoelectric anisotropy (i.e., d
33
/ | d
31
| >> 1) by doping the ceramic
[20], varying temperature [15] and by changing the volume fractions of the 90q
domains [15] within the ceramic grains.
A comparative analysis of the linear and non-linear effective properties in
BaTiO
3
ceramics was carried out within the framework of the model of a
polydomain grain in an effective medium [21, 22]. Results on the modelling of the
piezoelectric behaviour of the ceramic and on the electromechanical properties
determined using the effective medium method demonstrates that the non-linear
behaviour is a result of 90q domain-wall displacements in the grains. These
displacements in the presence of an externally applied field become an important
extrinsic factor that influences the piezoelectric coefficients d
fp
of the ceramic at
different poling states. For the fully poled BaTiO
3
ceramic, the aforementioned
contributions from the 90q domain-wall displacements to d
fp
attain 68 % and 64 %
[22] for fp = 31 and 33, respectively.
Therefore, subsequent averaging procedures facilitate the consideration of the
hierarchy-of-properties chain of single-domain single crystal polydomain single
crystal ceramic in the presence of the 90q domain-wall displacements (BaTiO
3
)
or on condition of the motionless 90q domain walls (PbTiO
3
) [10, 13, 15, 19, 20].
The main aims are to analyse factors that influence the piezoelectric properties in
the related ferroelectric materials and to study the correlation between the
microstructure, domain-wall displacements and electromechanical properties in the
ferroelectric ceramic. Knowledge of the full sets of the electromechanical constants
of ferroelectric ceramics enables appropriate material selection for various
piezotechnical applications and offers a route to using the ceramics as piezo-active
components for composite materials.
At present, there is a limited number of full sets of electromechanical constants
available in the literature for poled ferroelectric ceramics, primarily due to the
effort involved in manufacturing and poling a variety of test samples.
Electromechanical constants are often measured using standard methods [23, 24] at
room temperature and with weak electric fields. The majority of the ferroelectric
ceramics are manufactured as perovskite-type solid solutions [35, 2527] with
compositions varied and tailored to improve specific properties. Attempts to
classify the ferroelectric ceramics have been made in different papers [2527].
From the classification scheme [25] and by considering the conventional
perovskite-type ceramics based on Pb(Zr
1x
Ti
x
)O
3
, it is possible to identify
materials that are stable with regard to electric and mechanical actions, materials
with high dielectric permittivity , materials having high sensitivity with respect
to mechanical stress fields (high piezoelectric coefficients g
V
H
33
fp
), materials with the
large piezoelectric anisotropy d
33
/ d
31
, materials with high stability to the
resonance frequency, materials with low dielectric permittivity , and high-
temperature materials (high Curie temperature T
V
H
33
C
). From the classification method
proposed in monograph [26], the Pb(Zr
1x
Ti
x
)O
3
-based ceramics are divided into
four groups with specific performances. These groups are related to molar
concentrations of Ti from ranges 0 d x d 0.1, 0.1 d x d 0.4, 0.4 d x d 0.6, and 0.6 d
x d 1, and some advantages of each group are mentioned [26] in connection with a
variety of piezotechnical applications.

Examples of the ferroelectric ceramics and full sets of their room-temperature
electromechanical constants used in Equations 2.4 and 2.5 are listed in Table 2.1.
Analogous sets related to the advanced relaxor-ferroelectric single crystals with
engineered domain structures are also shown in Table 2.2 for comparison.
Differences between the individual constants of the materials with the same
composition (e.g., BaTiO
3
and (Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
ceramics from Table
2.1 or the 0.70Pb(Mg
1/3
Nb
2/3
)O
3
0.30PbTiO
3
single crystals from Table 2.2) are
the result of technological, microstructural and poling factors, the selection of a
composition in the proximity of the MPB, fluctuations of the composition over
samples and a variable mobility of the domain walls. These factors can influence
the domain structures [44], heterophase states [45, 46] and electromechanical
properties [47]. For example, changes in volume fractions of the 90q domains or in
the mobility of the 90q domain walls in the grains of the PbTiO
3
-type ceramics
affect their piezoelectric and other properties [15, 19, 26, 47]. Furthermore, in the
presence of an external electric field E, displacements of interfaces separating
morphotropic phases in the 0.92Pb(Zn
1/3
Nb
2/3
)O
3
0.08PbTiO
3
single crystal
promote an appreciable contribution to the piezoelectric coefficient d
33
[46]. This
contribution can exceed 50 % in a wide E range. The full sets of experimental
electromechanical constants from Tables 2.1 and 2.2 have been used in our study
on the piezo-composites, and we will analyse and discuss the results in Chapters
36.
In the next section we consider some of the physical factors that influence the
effective properties of two-component composites based on ferroelectrics.
Table 2.1. Experimental room-temperature values of elastic moduli (in 10
E
pq
c
10
Pa),
a

piezoelectric coefficients e
fp
(in C / m
2
) and dielectric permittivities of ferroelectric
ceramics
[
H
ff

BaTiO
3
[28]
b
BaTiO
3
[6]
b
(Ba
0.917
Ca
0.083
)
.
.
TiO
3
[6, 29]
(Pb
0.9625
La
0.025
)
.
.
(Ti
0.99
Mn
0.01
)O
3
(I) [30]
b
(Pb
0.855
Nd
0.11
)
.
.
(Ti
0.94
Mn
0.02
.
.
In
0.04
)O
3
[31]
E
c
11

16.6 15.0 15.8 14.33 17.3
E
c
12

7.66 6.6 6.91 3.220 5.31
E
c
13
7.75 6.6 6.75 2.413 5.68
E
c
33
16.2 14.6 15.0 13.16 16.6
E
c
44

4.29 4.4 4.51 5.587 6.09
e
31
4.34 4.35 3.14 0.4584 2.00
e
33
18.6 17.5 13.5 6.499 8.81
e
15
11.6 11.4 10.9 5.923 4.84
[
H
11
/H
0

1270 1115 1000 210 270
[
H
33
/H
0
1420 1260 910 140 198
(Pb
0.9625
La
0.025
)
.
.
(Ti
0.99
Mn
0.01
)O
3
(II) [31]
b
(Pb
0.85
Nd
0.10
)
.
.
(Ti
0.99
Mn
0.01
)O
3
[31]
(Pb, Ca)TiO
3
[32]
TBK-3
[5]
c
TBKS
[5]
d
E
c
11

15.8 17.5 15.05 15.7 14.6
E
c
12

4.20 5.49 3.68 6.3 5.18
E
c
13
4.55 5.65 3.09 6.2 5.06
E
c
33
15.2 16.5 12.78 15.3 14.2
E
c
44

6.32 6.26 1.20 4.4 4.9
e
31
1.32 1.80 1.71 3.2 0.68
e
33
6.80 8.83 8.8 12.5 7.71
e
15
3.35 5.41 0.33 9.1 4.56
[
H
11
/H
0

203 265 170 950 410
[
H
33
/H
0
133 193 140 950 336
NBS-1
[5]
e
ZTBS-3
[5]
f
ZTS-19
[5]
g
PZT-4
[6]
g
PZT-5
[6]
PZT-5A
[32]
E
c
11

11.0 15.1 10.9 13.9 12.1 12.04
E
c
12

4.22 7.9 6.1 7.78 7.54 7.52
E
c
13
4.12 8.0 5.4 7.43 7.52 7.51
E
c
33
10.0 13.6 9.3 11.5 11.1 11.09
E
c
44

3.45 2.9 2.4 2.56 2.11 2.3
e
31
3.3 7.9 4.9 5.2 5.4 5.4
e
33
14.8 17.7 14.9 15.1 15.8 15.8
e
15
7.7 15.4 10.6 12.7 12.3 12.3
[
H
11
/H
0

900 1610 820 730 916 540
[
H
33
/H
0
890 1280 840 635 830 830
Table 2.1 (continued)
PZ 27
[33]
PZ 34
[33]
PCR-1
[25]
h
PCR-7
[25]
PCR-7M
[25]
PCR-8
[25]
E
c
11

14.28 18.84 16.0 12.5 13.3 14.4
E
c
12

9.98 8.02 10.1 8.4 9.2 8.5
E
c
13
9.22 8.23 9.6 8.1 9.1 7.7
E
c
33
11.45 12.72 13.3 12.1 12.5 11.6
E
c
44

2.28 5.89 2.58 2.36 2.28 2.83
e
31
3.71 3.81 3.7 9.0 9.5 7.5
e
33
15.77 6.87 11.1 28.3 31.1 13.6
e
15
11.74 2.31 10.8 17.9 20.0 11.6
[
H
11
/H
0
1121 181 570 1430 1980 830
[
H
33
/H
0
878 154 300 1350 1810 740
PCR-13
[25]
PCR-21
[25]
PCR-62
[25]
PCR-63
[25]
PCR-73
[25]
E
c
11

11.9 13.4 12.7 14.1 17.3
E
c
12

4.8 7.3 4.5 6.6 13.3
E
c
13
3.1 6.2 0.8 5.6 12.5
E
c
33
10.0 11.8 11.3 11.2 14.4
E
c
44

3.50 2.45 4.70 4.15 2.29
e
31
6.0 7.3 6.6 4.6 9.1
e
33
10.4 16.2 8.9 11.8 29.2
e
15
7.1 12.0 6.0 6.9 22.4
[
H
11
/H
0
710 890 560 830 2220
[
H
33
/H
0
710 890 590 1000 2380
Notes to Table 2.1
a. Ferroelectric ceramics poled along the OX
3
axis are characterised [26] by fmm
symmetry and a relation c
66
= ( ) / 2
E
c
11
E
c
12
b. Different processing conditions lead to different sets of the electromechanical constants as
shown for the BaTiO
3
and (Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
ceramics
c. Based on (Ba
x
Ca
1x
)TiO
3

d. Based on (Ba
x
Ca
1xy
Pb
y
)TiO
3

e. Based on (Ba
1x
Pb
x
)Nb
2
O
6

f. Based on (Ba
1x
Pb
x
)(Zr
1y
Ti
y
)O
3

g. Based on Pb(Zr
1x
Ti
x
)O
3

h. PCR is the abbreviation for the group piezoelectric ceramics from Rostov-on-Don.
These materials based mainly on Pb(Zr
1x
Ti
x
)O
3
are manufactured [25] by the conventional
method (PRC-13 and PCR-62) or by means of hot pressing (PCR-1, PCR-7, PCR-7M, PCR-
8, PCR-21, PCR-63, and PCR-73)

Table 2.2. Experimental room-temperature values of elastic moduli
E
pq
c
(in 10
10
Pa),
piezoelectric coefficients e (in C / m
fp
2
) and dielectric permittivities
[
H
ff of (1
x)Pb(A
1/3
Nb
2/3
)O
3
xPbTiO
3
single crystals (4mm symmetry)
A= Mg,
x= 0.30
[34]
A= Mg,
x= 0.30
[35]
A= Mg,
x= 0.33
[36]
A= Mg,
x= 0.38
[37]
A= Mg,
x= 0.42
[38]
A= Zn,
x= 0.045
[39]
A= Zn,
x= 0.07
[40]
E
c
11
11.7 16.04 11.5 21.25 17.51 11.1 11.3
E
c
12
10.3 14.96 10.3 14.33 8.51 10.2 10.3
E
c
13

10.1 7.51 10.2 13.51 8.3 10.1 10.5
E
c
33

10.8 12.0 10.3 9.92
a
10.5 10.5 10.91
E
c
44
7.1 5.38 6.9 5.56 2.85 6.4 6.30
E
c
66

6.6 2.87 6.6 6.95 8.0 6.3 7.10
e
31
2.4 5.22 3.9 3.26 2.1 3.7 2.3
e
33
27.1 30.4 20.3 11.46 12.2 15.0 15.1
e
15
13.6 31.84 10.1 21.12 37.50
c
8.9 11.1
[
H
11
/H
0

3307 4963 1434 2171
b
3054 3000 2779
[
H
33
/H
0
1242 1386 680 255 259 1000 823
A = Zn,
x = 0.08 [41]
A = Zn,
x = 0.08 [42]
A = Zn,
x = 0.08 [43]
E
c
11
11.5 11.1 11.47
E
c
12
10.5 10.2 10.30
E
c
13

10.9 9.97 10.37
E
c
33

11.51 10.2 10.71
E
c
44
6.34 6.1 6.70
E
c
66

6.50 6.56 3.05
e
31
5.1 9.15 5.83
e
33
15.4 12.2 12.60
e
15
10.1 9.39 6.22
[
H
11
/H
0

2720 2848 1415
[
H
33
/H
0
984 620 560
Notes to Table 2.2
a. According to our calculations based on formula [2] = G
E
pq
c
E
qr
s
pr
and values = 11.73,
= 2.71, = 8.17, and = 21.53 (in 10
E
s
11
E
s
12
E
s
13
E
s
33
12
Pa
1
) [37], it would be = 14.89u10
E
c
33
10
Pa
instead of 9.92u10
10
Pa
b. According to our calculations based on formula [2] = (1 k
[
H
kr
V
H
kr
15
2
) and experimental
values / H
V
H
11
0
= 4301 and k
15
= 0.459 [37], it would be / H
[
H
11
0
= 3395 instead of 2171. The
value / H
[
H
11
0
= 3395 would also be attained using Equation 2.16, / H
V
H
11
0
= 4301, d
15
= 380
pC / N, and e = 21.12 C / m
2
from ref. [37]
15
c.

According to our calculations based on Equation 2.13 and experimental values d
15
= 131
pC / N and
E
c
44

= 2.85u10
10
Pa [38], it would be e
15
= 3.73 C / m
2
instead of 37.50 C / m
2
2.2 Sum, Combination and Product Properties
Ferroelectrics and related piezo-active materials are characterised by the full sets of
electromechanical constants that are listed in Tables 2.1 and 2.2. Ferroelectric
ceramics and single crystals are used as components of piezo-composites and
actively influence the effective properties of the composites. Each physical
property of a piezo-composite is analytically determined using specific formulae or
procedures of averaging (see, e.g., [5, 32, 33]). The physical property can also be
studied by experimental methods and characterisation. In the case of piezo-
composites, results on averaging depend on the volume fractions of components
and their properties, microstructure and connectivity [5, 33, 48, 49]. Numerous
theoretical and experimental results point out that there are three main effects [48,
49] related to forming the effective properties in the piezo-composites, namely
(i) the effect of the addition of the properties (forming the sum properties),
(ii) the effect of the combination of the properties (forming the combination
properties) and
(iii) the effect of the generation of the properties (forming the product
properties).
2.2.1 Sum Properties
For example, the effective property Y* of a two-component piezo-composite
1

obeys a condition Y
(1)
< Y*< Y
(2)
, where Y
(n)
is the analogous property of the nth
component, n = 1 and 2. Y* belongs to the sum properties when it is characterised
by a monotonic dependence (Figure 2.2a) on the volume fraction m of one of the
components. It is the simplest example of the volume-fraction dependence of the
effective property, that is determined by a general formula Y*(m) = Y
(1)
m + Y
(2)
(1
m) + 'Y, where 'Y depends on Y
(n)
and m and characterises deviations of Y*(m)
from the linear dependence (compare solid and dotted lines in Figure 2.2a). Such
behaviour is typical of elastic moduli , piezoelectric coefficients and
dielectric permittivities of 03 or 13 composites [5, 5054] comprising a
PZT-type ceramic and a piezo-passive polymer.
E
ab
c
* *
ij
d
V
H
*
pr
2.2.2 Combination Properties
In this case, the properties Y
(n)
and Z
(n)
of the components (n = 1 and 2) form the
effective properties Y*(m) and Z*(m) that are monotonic dependences (Figures
2.2a and b). It is assumed that )
(n)
is expressed in terms of Y
(n)
and Z
(n)
and also

____________
1
Hereafter, we denote all of the effective constants and parameters of composites by an
asterisk. Following Equations 2.6 and 2.7, the full set of effective constants is written as ,
and . The similar set of the constants , and is derived from Equations 2.8
and 2.9. Finally, the full set of the effective constants , and can be determined
using Equations 2.10 and 2.11.

E
pq
s
*
*
fp
d
V
H
*
kr
D
pq
s
* *
fp
g
V
E
*
kr
D
pq
c
* *
fp
h
[
E
*
kr

Figure 2.2. Sum properties Y*(m) and Z*(m) (a and b) and combinations properties )*(m)
(ce) in two-component composites. m is the volume fraction of the first component. In
graphs a and b, possible versions of the volume-fraction dependence are shown with dotted
lines
characterises some property of the nth component. The monotonic character of
Y*(m) and Z*(m) does not imply a monotonic volume-fraction behaviour of the
effective property )*. In this case a non-monotonic )*(m) dependence (often
with values |)*(m) | > |)
(n)
|, Figures 2.2c and d) is concerned with the effect of
combination of the properties. This effect is important, for example, when
considering the piezoelectric coefficients = / and = / (j = 1
and 3) of various piezo-composites [50, 52, 54] (see also Sections 3.13.3 and 5.4).
*
3 j
g
*
3 j
d
V
H
*
33
*
3 j
h
*
3 j
e
[
H
33
Furthermore, the combination property )*(m) can be described by a monotonic
volume-fraction dependence (Figure 2.2e), but a configuration of the )*(m) curve
differs from both the Y*(m) and Z*(m) curves. The effect of combination of three
properties (i.e., elastic, piezoelectric and dielectric properties) is concerned with
the volume-fraction dependence of electromechanical coupling factors of the 13
ceramic / polymer composites [32, 51].
2.2.3 Product Properties
The effect of generation of properties in a two-component composite is described
as follows. It is assumed that a response V
1
of the first component is the result of an
external action X. Simultaneously, the second component does not show a response
on the action X, but generates only a response V
2
in response to V
1
. This implies
that the internal action V
1
from the first component due to the external action X is
the reason for the new response V
2
. As a result, the formation of product properties
is associated with two different responses, V
1
and V
2
. Such an effect takes place, for
example, in composites [55, 56] comprising one magnetostrictive (CoFe
2
O
4
or
other ferrites) and one piezoelectric (BaTiO
3
, PbTiO
3
or PZT-type) ceramics. In
the presence of an external magnetic field, a mechanical strain is generated in the
magnetostrictive ceramic. This strain is transferred into the piezoelectric
component of the composite that generates an electric field in the piezoelectric
ceramic. Thus, the magnetic field applied to the composite sample generates an
electric field due to the combination of the magnetostrictive (but piezo-passive)
and piezo-active (but showing no magnetostriction) components. In this composite
the role of both the electromechanical and electromagnetic interactions between the
components is obviously of importance.
In general, the product property appears due to the interaction of two
components that have different properties being mutually complementary. O ther
examples of the product properties [49] in the composites based on ferroelectrics
are concerned with pyroelectricity (taking into account piezoelectricity of the first
component and thermal expansion of the second component) and photostriction
(taking into account the photovoltaic effect in the first component and
electrostriction in the second component).
Recent work [5759] has been devoted to the study of more complicated effects
in the piezo-composites and disordered piezo-active heterogeneous materials.
Taking into account electric conductivities of the components, it is possible to
achieve giant dielectric permittivity [57], giant piezoelectric and dielectric
enhancement [58].
Our further consideration of the effective electromechanical properties is
closely connected with examples of non-monotonic dependences of these
properties that are often caused by the effect of the combination of properties. A
transition to an analysis of the properties would be impossible without a
characterisation of methods for determination of the effective properties in the
piezo-composites with different connectivity patterns and components. These will
now be described.
2.3 Methods for Evaluation of Effective Parameters
It is well known [26, 26, 29] that Equations 2.42.19 are valid not only for
piezoelectric single crystals, but also for poled ferroelectric polydomain single
crystals, ferroelectric ceramics and composites containing at least one piezo-active
component. The applicability of Equations 2.42.19 to describe the
electromechanical properties of piezo-composites is reasonable in cases when these
properties can be regarded as effective (or overall), i.e., characterising the response
of the composite sample as a whole. The effective electromechanical properties
and related parameters enable us to move away from the properties of the
individual components, the specifics of microgeometry and other features of
heterogeneous materials and to estimate the potential advantages and applications
of these materials.
The problem of prediction of the effective physical properties in advanced
composites is of interest to many specialists undertaking theoretical and
experimental studies of these heterogeneous materials. A formulation of the related
problem on the effective electromechanical properties in the piezo-composite is
given in ref. [5]. Equations 2.4 and 2.5 are examples of constitutive equations that
describe the behaviour of the piezo-active material under electric and mechanical
fields. If averaged components of the electric field <E
f
> and the mechanical strain
<[
q
> in a macroscopic region of the piezo-composite are independent of co-
ordinates x
j
(j = 1, 2 and 3), then Equations 2.4 and 2.5 can be rewritten as follows:
<V
p
> = <[
E
pq
c
*
q
> <E
*
fp
e f
> (2.21)
<D
k
> = <[
*
kl
e l
> + <E
[
H
*
kr
r
> (2.22)
In Equations 2.21 and 2.22 elastic moduli , piezoelectric coefficients and
dielectric permittivities
E
pq
c
* *
fp
e
[
H
*
kr

constitute the full set of effective electromechanical
constants of the piezo-composite. These constants are determined [5] taking into
account equations of electric and mechanical equilibrium and boundary conditions
at the surface of the macroscopic region in the heterogeneous medium.
Calculations of the effective electromechanical constants of the piezo-composite
are closely connected with averaging a series of vector and tensor components of
the electric and mechanical fields, for example, [
q
, E
f
, V
p
, and D
k
(see Equations
2.42.11). This averaging is performed on volume fractions of the composite
components
2
(n = 1, 2, ) for which the full sets of electromechanical constants,
such as , and or other constants involved in Equations 2.62.11, are
known. The formulation of the problem allows the development of a variety of
methodologies for the determination of the full sets of electromechanical constants
and various methods of numerical realisation for specific connectivity patterns and
arrangement, etc. [5, 32, 33, 5154, 6063].
E n
pq
c
), ( ) (n
fp
e
[
H
), (n
kr

____________
2
In order to avoid confusion owing to vector (tensor) components and composite
components, the latter are also called constituents.
The central aspect of this problem is a method of homogenisation of the
electromechanical properties [33, 63]. It is clear that the effective properties are to
be regarded as properties in a material at distances exceeding the linear dimensions
of an individual inclusion or the thickness of a separate layer in a composite
sample. However, the same inclusions can be inhomogeneous (for example,
ferroelectric ceramic particles that are split into domains or are heterophase). It is
therefore necessary to attain an additional homogenisation of the properties in
these inclusions. Such a situation needs a compromise. For example, the effective
electromechanical properties in the 22 piezo-composite are determined [64, 65]
by assuming that their components are characterised by an equivalent homogeneity
[66]. The authors of refs. [67, 68] have considered a variety of applications of the
approach developed with attention to piezo-composite and biological materials. It
is also noted that homogenisation methods are applicable [67] to media with
periodic and quasi-periodic structures.
Based on the detailed analysis of results concerning composite materials and
their effective properties, Khoroshun et al. [5] placed emphasis on two main
scientific directions in modern mechanics of composites. The first direction
represents the mechanics of the composites having regular (periodic) structures.
The second direction is related to the mechanics of stochastically heterogeneous
materials, including the composites with arbitrary connectivity, shape and
distribution of structural elements.
The first direction deals with periodic solutions of equations of mechanics of
microinhomogeneous media in which the regular distribution of the structural
elements is present. As a rule, such solutions are applicable to the matrix piezo-
composites with inclusions having the appropriate shape, e.g., spheroids. The
accompanying method, called the method of regularisation of structure [5], is
applied to the composites [32, 33, 60, 63, 69] with a certain disorder in the shape
and distribution of the structural elements. This method makes it possible to
substitute the real composite by a model material with regularly distributed
structural elements having the appropriate shape only. In such a model material, a
periodic representative volume element (similar to unit cells in single crystals, as
shown, for example, in the Banno cube model [69]) is chosen and analysed using
micromechanical methods including the finite-element method (FEM). A
representative volume element used for a prediction of effective parameters of the
porous piezo-ceramic [70] with 33 connectivity is shown in Figure 2.3.
The second direction implies using random functions to describe the properties
in the composite with a stochastic structure. These properties are regarded as
randomly varied ones over the composite sample, and the homogenisation can be
performed with regard to mathematical features of the random functions. This
approach is suitable for the prediction of the effective properties in the piezo-
composites without restrictions on the shape and distribution of inclusions and
connectivity.
A considerable amount of research is devoted to the 03 piezo-composites [5,
33, 50, 53, 6062] that contain a system of isolated inclusions embedded in a
matrix. The inclusion is often made from a ferroelectric ceramic, and the matrix is
regarded as a polymer medium being either piezoelectric or piezo-passive. Such
piezo-composites are manufactured using a variety of techniques (see, e.g., reviews

a b
Figure 2.3. Micrograph of the porous PZT ceramic (a) and example of the representative
FEM image (b) that is used for calculations of the effective electromechanical properties. x,
y and z are co-ordinate axes (reprinted from Ramesh et al. [70], with permission from
Elsevier)
[48, 71, 72]) and studied using a series of experimental methods [7379].
Theoretical studies of 03 piezo-composites are focused on the effective
electromechanical properties that are strongly dependent on the properties and
volume fractions of the components [5, 33, 50, 61], on the shape of the inclusions
and features of their distribution in the matrix [5, 60, 80] as well as on poling
conditions [75, 77] for the composite sample as a whole. In these studies, it is often
assumed that the inclusions have the shape of a spheroid [53, 81] or a rectangular
parallelepiped [33] and aligned regularly in the matrix. An electromechanical
interaction between each inclusion and the surrounding matrix is taken into
account [5, 60, 61] in connection with internal electric and mechanical fields and
boundary conditions for the specific shape of the inclusion.
An approach developed by Banno [69, 82, 83] stems from concepts [84]
initially applied to 22 composites. Based on these concepts, the 03 composite is
represented [83] by the Banno unit cell (or the representative volume element) that
comprises the parallelepiped-like inclusion and the surrounding matrix. The
effective electromechanical properties in this composite are determined by direct
averaging of the properties in the components. In this averaging procedure, the
shape of the parallelepiped (i.e., a ratio of their edges), but not the interaction
between the components plays an important role.
In an alternative matrix method [33] proposed for the prediction of the
electromechanical properties in the 03 composite, the theoretical description is
based on the matrix approach [64, 85] previously applied to the 22 composite.
This approach enabled the authors of ref. [33] to perform the averaging procedure
for the 03 composite with a regular distribution of the parallelepiped-like
inclusions. The method developed by Levassort et al. [33] has been extended to 3
3 and 13 ceramic / polymer piezo-composites [86, 87] with planar interfaces. In
refs. [33, 86, 87], the electromechanical constants of the components ,
and are involved in matrices of the properties (9 u 9) and boundary
conditions. To avoid mathematical complications concerned with the presence of
the matrix elements with strongly differing degrees of values (e.g., a (10
E n
pq
c
), ( ) (n
fp
e
[
H
), (n
kr
E n
pq
c
), (
10

10
11
) Pa and a (110) C / m
) (n
fp
e
2
for the nth component), the matrix approach has
been modified and successfully applied to the 1E piezo-composites with E = 1, 2
and 3 [88, 89]. The matrices describing the electromechanical properties in the
components comprise elastic compliances a (10
E n
pq
s
), (
11
10
10
) Pa
-1
, piezoelectric
coefficients a (10
) (n
fp
d
11
10
9
) pC / N and dielectric permittivities a (10
V
H
), (n
kr
11

10
8
) F / m, and use of such a set of electromechanical constants [88, 89] avoids the
aforementioned significant differences in the degrees of the values of the
electromechanical constants.

Many publications on 03 composites are concerned with a piezoelectric
ellipsoidal inclusion problem [5, 53, 60, 9095]. This problem stems from the
classical Eshelby analysis [96, 97] of the stress and strain fields in and around an
elastic ellipsoidal inclusion. Results [96, 97] on the uniform stress and strain fields
in such inclusions and the equivalent inclusion method stimulated intensive
investigations into heterogeneous media and their effective physical properties
[98]. Work on the coupled electric and mechanical fields in heterogeneous
materials implied the application of the equivalent inclusion method [96, 97] for
the description of the piezo-composites, poled ferroelectric ceramics and piezo-
active materials with pores or cracks [94, 95, 99101]. The inclusion can be either
piezo-active or piezo-passive, and no restrictions on its symmetry are introduced.
The matrix can be either piezo-active (transversely isotropic) or piezo-passive
(isotropic) [9195, 99, 101].
The analysis of the coupled electromechanical behaviour of an inclusions
matrix system is based on knowledge of the Eshelby tensors. These tensors
represent the strain and electric fields that are induced in the inclusion due to a
constraint of the matrix in the external fields. The Eshelby tensors are expressed in
the explicit form for the following inclusion shapes [101, 102]: spheroid, elliptic
cylinder, rod, penny, and ribbon-like. All of the inclusions are assumed to be
aligned along the axis of anisotropy, i.e., along the poling axis in the transversely
isotropic medium. As in the work on elastic inclusions in an elastic matrix [96, 97],
the Eshelby tensors depend on the shape of the inclusion and on the properties in
the matrix. Further averaging of the electromechanical properties is often carried
out according to a self-consistent scheme. The main variants of this scheme are
concerned with the effective medium method [94, 103, 104] and the effective field
method (or the mean field method) [53, 94, 95, 104106].
In the effective medium method, the inclusion is regarded as a region in a
continuous medium (i.e., in a matrix containing an ensemble of similar inclusions).
The effective properties in this medium are found using the Eshelby tensors for the
specified shape of the inclusions, and calculations are performed within the
framework of iterative (self-consistent) procedures.
The effective field method (EFM) stems from the Mori Tanaka method [107]
based on the concept of an average stress in the matrix containing the inclusions
with so-called eigenstrains. The eigenstrains represent the strains appearing in a
stress-free region, similar to the spontaneous strains in ferroelectric or ferroelastic
phases. In the EFM, the inclusion interacts with the matrix subjected to an action
of an external electric and mechanical field. Due to the presence of the ensemble
of similar inclusions in the matrix, the field can be regarded as the effective field

that acts on each inclusion therein. This means that the effective (or average)
electric and mechanical fields in the composite sample (see Equations 2.21 and
2.22) play a significant role in forming the overall effective electromechanical
properties. The EFM is one of the most powerful and most suitable methods [105]
for studying the overall behaviour of piezo-composites and for predicting their
effective electromechanical properties. Moreover, the EFM is known to produce
physically realistic results with regard to the effective properties in the piezo-
composites [53, 104, 106] in a wide range of inclusion volume fractions. This
method has been applied to determine the effective electromechanical properties in
the 03 ferroelectric ceramic / polymer [53, 80], ceramic / ceramic [61, 105] and
single crystal / ceramic [108] composites.
In addition to the above approaches, the effective electromechanical properties
in a piezo-composite with spheroidal inclusions have also been determined using
the dilute concentration limit (or dilute solution) [94, 106, 109], the differential
scheme [94] and the conditional moments method [5]. These methods are based on
different averaging procedures and take into account features of the arrangement of
the inclusions and their electromechanical interaction.
In addition to the piezoelectric effect in the composites studied, the
piezomagnetic and magnetoelectric effects have also been examined in 03
composite systems [110112]. These coupled effects are of interest due to the
involvement of three subsystems (e.g., electric, magnetic and mechanical) in
materials whereby an interaction between piezoelectric and piezomagnetic
components [110, 111] occur. The study of these effects is based on knowing the
magneto-electro-elastic Eshelby tensors for spheroidal inclusions [112]. The
authors of refs. [110112] generalised earlier results [101] for the case of a 03
piezoelectricpiezomagnetic composite.
The determination of the effective properties in piezo-composites with arbitrary
connectivity is often carried out by means of the FEM [63, 113118]. This method
employs numerical modelling techniques that take into account boundary
conditions for electric and mechanical fields in a composite system. The model of
the composite is based on the unit cell being the periodic representative volume
element, and this element is assumed to be characteristic of the whole composite
structure. An important advantage of the FEM is that there are no restrictions on
connectivity of the composite, sizes of the inclusions and their spatial distribution,
the number of the components and their properties. Interesting examples of the
numerical realisation of the FEM for evaluating the effective electromechanical
constants are given in refs. [116, 117] (03 connectivity), [63, 68, 113, 118] (13
connectivity) and [114, 115] (30, 33 and mixed connectivity patterns). Trends in
the study of the piezo-composites and other piezo-active materials show that the
FEM and related finite-element software becomes increasingly more widespread in
mechanical-engineering, smart-materials and piezotechnical applications
(transducers, sensors, actuators, hydrophones, acoustic antennae, etc.). However,
advances in numerical techniques do not exclude the development of analytical
methods for connectivity patterns such as 03, 13 and 22.

2.4 Evolution of DE Connectivity Patterns
An example of the evolution of the connectivity patterns in the two-component
composites becomes clear with reference to the problem of a piezo-composite with
ellipsoidal inclusions (see Section 2.3). In a general case, a composite containing
isolated ellipsoidal inclusion is characterised by 03 or 30 connectivity, where the
index 3 refers to the number of co-ordinate axes along which the matrix of the
composite is distributed continuously. In the case of 03 connectivity, this
corresponds to a ferroelectric ceramic / polymer composite with ceramic
inclusions. Infinite elongation of one of the axes of the ellipsoid enables us to
consider the inclusions having the shape of the elliptic cylinder or a lengthy rod,
and this microgeometry results in 13 or 31 connectivity of the composite.
However, the structure of the 03 composite containing penny-shaped inclusions
approaches a structure of the 22 composite. These changes in connectivity are
accompanied by changes in the Eshelby tensors [101] so that they can be utilised to
predict the effective properties in the 13, 31 or 22 composites with one or two
piezoelectric components.
Another example of the evolution of the connectivity patterns is associated with
various modifications of the 22 composite structure. According to Levassort et
al. [33, 86, 87], the modified patterns are described by 03, 13 or 33
connectivity. The effective properties in the corresponding piezo-composites are
determined by means of the matrix method [33, 64].
It can be seen from the examples above that some connectivities of the two-
component composite can be associated with slight modifications of the initial
connectivity pattern. In this section, we consider the entire complex of the DE
connectivity patterns (Figure 1.1), their evolution and interconnections between
them [89] in the two-component composites with planar interfaces.
It is assumed that there is a group of junction connectivity patterns and that
their combination can form new connectivity patterns. Each composite with a
regular arrangement of components is represented by the Banno unit cell (or
representative volume element) [82, 83]. The presence of planar interfaces and
their certain hierarchy inherent in the DE composite suggest that the 22
connectivity pattern (i.e., a system of series or parallel layers) may be regarded as
one of the junction connectivity patterns. For example, a simple multiplication of
the 22 connectivity pattern results in the 11 connectivity pattern [88], see Figure
2.4 and the connection between the 22 and 11 structure. A further simplification
of the 11 connectivity pattern enables us to obtain the 13 or 31 connectivity
pattern.
3
It should be noted that this simplification has similarities to the formation
or rearrangement of ferroelectric domain structures, in which not all domain types
are present even though these types are predicted [119] in accordance with
symmetry changes at phase transitions.

____________
3
It is seen from Figure 2.4 that the above-mentioned 22, 11, 13, and 31 connectivity
patterns make up the full set of the junction connectivity patterns. These patterns show
numerous interconnections between the structures of the two-component composites from
Figure 1.1.
Figure 2.4. Links between the DE connectivity patterns. C1 and C2 denote components 1
and 2, respectively. Double lines connecting the 22 and 11 connectivity patterns as well as
the 11 and 00 connectivity patterns denote a multiplication of structural elements at
transformations 22 o 11 and 11 o 00. Dashed lines connecting the 11 and 13
connectivity patterns as well as the 11 and 31 connectivity patterns denote a
simplification of the 11 structure at transformations 11 o 31 and 11 o 13 (reprinted
from Topolov and Glushanin [89], with permission from IOP Publishing)
The main interconnections between the DE connectivity patterns are shown in
the schematic in Figure 2.4 that contains links between the junction connectivity
patterns and other connectivity patterns in the DE composites. All the DE
composite structures can be generated by systems of the layers. These layers are
either homogeneous (see components C1 and C2 at the base of Figure 2.4) or
heterogeneous. The heterogeneous layers are characterised by the junction
connectivity pattern, by the combination of the junction connectivity patterns or by
the combination of the junction connectivity pattern and one of the components.
The evolution of the connectivity patterns demonstrates the important role of
the junction connectivity patterns. The main features of this evolution are described
as follows.
1. From the symmetrical arrangement of the junction connectivity patterns and
other elements in the scheme (Figure 2.4), it can be concluded that the evolution of
the connectivity patterns is primarily caused by a motion along the symmetry axis
(i.e., the line that mentally connects the connectivity patterns with D = E) and
adjacent branches.
2. The structures of the interfaces are rather complicated for 01 and 12
connectivity patterns (Figure 1.1) that are generated by three or two junction
connectivity patterns. The 00 connectivity pattern is situated at a dead-end
between the 11 connectivity pattern and the component Ci (i = 1 or 2). As a
consequence, the 00 connectivity pattern does not influence the formation of the
0E structures with E = 1, 2 and 3.
3. The maximum distance between the 33 connectivity pattern and the
components Ci, along with the lack of direct links between them are also typical of
the evolution of the connectivity patterns. It is shown in Figure 2.4 that the
remaining connectivity patterns are located in between 33 and Ci. The continuous
distribution of the C1 or C2 component or each component in the 33 connectivity
pattern along three co-ordinate axes represents two extreme cases of the
arrangement in the scheme (see upper and bottom parts of Figure 2.4).
4. The number of systems of the interfaces decreases at D = 3 or E = 3. Such a
decrease is a result of a continuous distribution of one (at D = 3 z E or E = 3 z D)
or two (at D = E = 3) components along three co-ordinate axes. In fact, any
introduction of an additional type of interface would destroy this continuity.
Knowledge of the evolution of the connectivity patterns enables the analysis of
the DE piezo-composites with planar interfaces as well as understanding the
resulting effective electromechanical properties in these heterogeneous materials.
The concepts developed in this section can ease both the modelling and
optimisation of the effective electromechanical properties, especially in the piezo-
composites with complicated connectivity patterns. An example of the 12
composite and a series of parameters introduced for determination of its effective
electromechanical properties are considered below.
2.5 Modelling of Effective Properties in Piezo-composites with
Planar Interfaces
An important step following the analysis of the evolution of the connectivity
patterns is the synthesis, i.e., a combination of parts to form a connected composite.
An approach developed in ref. [89] is based on the assumption that different DE
composite structures with planar interfaces can be generated using a system of
layers (Figure 2.5). It is assumed that the layers with specified connectivity
regularly alternate along the OX
3
axis that represents the poling direction. The C1
component is shown as the dotted area in all cross-sections (Figure 2.5) of the DE
connectivity patterns (D E) and in the schematic arrangement of the 12
composite (Figure 2.5, see bottom right). In order to determine the effective
electromechanical properties of the 12 composite, one can introduce a series of
concentration parameters [89], for example, t = |AC| / |AE|, n = |FH| / |FL|, tc = |AB|
/ |AC|, nc = |FG| / |FH|, and v = |MN| / |MP|. These parameters
4
characterise volume

____________
4
The concentration parameters t, n and tc, which are related to the 00 and 01 connectivity
patterns (Figure 2.5), are introduced by analogy with the concentration parameters related to
the 12 composite. The DE connectivity patterns obeying the condition D E are depicted
by the similar schemes wherein the dotted areas shown in Figure 2.5 become white and the
white areas become dotted.

Figure 2.5. Systems of layers forming the DE connectivity patterns (reprinted from
Topolov and Glushanin [89], with permission from IOP Publishing)
fractions of the C1 and C2 components in the 12 composite and are assumed to be
constant over the composite sample. Any variations of the concentration
parameters mean that relative widths (or lengths) of several parallelepipeds are
simultaneously varied in all the layers of the composite. The layers forming the D
E composites are parallel to the (X
1
OX
2
) plane as shown for the 12 composite
(Figure 2.5, see bottom right).
The calculation of the effective electromechanical constants of the DE
composites can be performed by means of the matrix method [64] (as shown for
the 11, 12 and 13 ferroelectric ceramic / polymer composites [88, 89]) or the
FEM. In these methods the electromechanical interaction between the adjacent
regions is taken into account in averaging procedures. For this calculation, full sets
of electromechanical constants of both components are required. The effective
properties in the aforementioned 1E composites with E = 1, 2 and 3 are
determined [88, 89], for example, using elastic compliances , piezoelectric
coefficients and dielectric permittivities of the ferroelectric ceramic (n =
1) and polymer (n = 2).
E n
pq
s
), (
) (n
fp
d
V
H
), (n
kk
In a general case, the matrix method [64] is applied to predict the effective
electromechanical properties in the periodically layered 22 composite or the
ferroelectric single crystal with a laminar domain structure [120, 121]. This
approach can be applied, providing the wavelength of an incident acoustic field is
much longer that the thickness of the separate layer. In a range of low frequencies
Z of the AC electric field, the condition ZW << 1 holds (where W is the Maxwell
Wagner relaxation time) and the electric conductivities of the layers are to be taken
into account [57] for the correct prediction of the effective electromechanical
properties in the 22-type piezo-composite. The matrix method [64], which does
not take into account the electric conductivities, can be applied at frequencies that
are considerably more than those where the MaxwellWagner relaxation is in
progress. At the same time, the frequency of the electric field should be well below
a piezoelectric resonant frequency of the composite sample for specific oscillation
modes.
The modelling of the effective properties in the piezo-composites is closely
connected with their potential applications. Clearly, for successful and optimum
piezotechnical applications, the piezo-composites with different connectivity
patterns, volume fractions of the components and other parameters are required.
The optimisation of the effective properties and related parameters as well as
manufacturing the appropriate piezo-composites are of independent interest. It
should be noted that structure processing property relations in the ferroelectric
ceramic / polymer composites are described using connectivity as the unifying
central concept [71]. Computer-aided design-based fabrication methods have
advanced in recent years [72]. In particular, novel processing technologies make it
possible to attain unprecedented structural complexity [71] as compared to the
traditional methods [48, 69] for manufacturing the ceramic / polymer composites.
2.6 Connectivity Links Properties
Chapter 2 has been devoted to the description of the electromechanical properties
in piezo-active materials. An emphasis has been placed on the piezo-composites,
their connectivity and trends in forming their effective electromechanical
properties. Three types of effective properties (i.e., sum, combination and product
properties) have been considered and applied to the two-component composites
based on ferroelectrics. Results on the evolution of the DE connectivity patterns
and the role of systems of planar interfaces inherent in the DE piezo-composites
have been discussed. A scheme demonstrating the links between the DE
connectivity patterns has been put forward for the modelling of the effective
properties in the piezo-composites with planar interfaces. In general, Chapter 2
can be regarded as a transition from the connectivity concept to determination of
the volume-fraction dependences of the effective properties in different groups of
piezo-composites, such as ferroelectric ceramic / polymer, ferroelectric single
crystal / polymer, ferroelectric single crystal / ferroelectric ceramic, with two or
more piezoelectric components, etc. Further discussion will examine the
determination and interpretation of calculated data relating to specific connectivity
patterns and the aforementioned groups of the piezo-composites. In Chapter 3 we
will consider the most important DE connectivity patterns as well as features of
the effective electromechanical properties and applications of the two-component
composites based on ferroelectric ceramics.
References
[1] Levanyuk AP, Sannikov DG, (1994) Piezoelectrics. In: Prokhorov AM (ed.) Physics
Encyclopaedia. Vol. 4. Bolshaya Rossiyskaya Entsiklopedia, Moscow (in Russian),
pp 188189
[2] Ikeda T, (1990) Fundamentals of Piezoelectricity. Oxford University Press, Oxford
New York Toronto
[3] Zheludev IS, (1971) Physics of Crystalline Dielectrics. Vol. 2: Electrical Properties.
Plenum, New York
[4] Uchino K, (1997) Piezoelectric Actuators and Ultrasonic Motors. Kluwer, Boston
Dordrecht London
[6] Berlincourt DA, Cerran DR, Jaffe H, (1964) Piezoelectric and piezomagnetic
materials and their function in transducers. In: Mason W (ed.) Physical Acoustics.
Principles and Methods. Vol. 1: Methods and Devices. Pt A. Academic Press, New
York London, pp 169270
[7] Hall DA, (2001) Review. Nonlinearity in piezoelectric ceramics. Journal of Materials
Science 36:45754601
[8] Turik AV, Bondarenko EI, (1974) Effect of domain structure on physical properties of
ferroelectrics. Ferroelectrics 7:303305
[9] Turik AV, (1970) Elastic, piezoelectric, and dielectric properties of single crystals of
BaTiO
3
with a laminar domain structure. Soviet Physics Solid State 12:688693
[10] Turik AV, Topolov VYu, Aleshin VI (2000) On a correlation between remanent
polarization and piezoelectric coefficients of perovskite-type ferroelectric ceramics.
Journal of Physics D: Applied Physics 33:738743
[11] Aleshin VI, (1990) Domain-orientation contribution into constants of ferroelectric
polydomain single crystal. Zhurnal Tekhnicheskoi Fiziki 60:179183 (in Russian)
[12] Topolov VYu, (2003) Domain wall displacements and piezoelectric activity of
KNbO
3
single crystals. Journal of Physics: Condensed Matter 15:561565
[13] Turik AV, Chernobabov AI, (1977) On an orientation contribution in dielectric,
piezoelectric and elastic constants of ferroelectric ceramics. Zhurnal Tekhnicheskoi
Fiziki 47:19441948 (in Russian)
[14] Aleshin VI, (1991) Spherical inclusion in an anisotropic piezo-active medium.
Kristallografiya 36:13521357 (in Russian)
[15] Topolov VYu, Bondarenko EI, Turik AV, Chernobabov AI (1993) The effect of
domain structure on electromechanical properties of PbTiO
3
-based ferroelectrics.
[16] Aleshin VI, (1987) Properties of textures being formed on the basis of non-180q
reorientations. Kristallografiya 32:422426 (in Russian)
[17] Bondarenko EI, Topolov VYu, Turik AV, (1990) The effect of 90 domain wall
displacements on piezoelectric and dielectric constants of perovskite ceramics.
[18] Bondarenko EI, Topolov VYu, Turik AV, (1991) The role of 90 domain wall
displacements in forming physical properties of perovskite ferroelectric ceramics.
Ferroelectrics. Letters Section 13:1319
[19] Turik AV, Topolov VYu, (1997) Ferroelectric ceramics with a large piezoelectric
anisotropy. Journal of Physics D: Applied Physics 30:15411549
[20] Topolov VYu, Turik AV, Chernobabov AI, (1994) On the mechanism of high
piezoelectric anisotropy in lead titanate-based ferroelectrics. Crystallography Reports
39:805809
[21] Rdel J, Kreher WS, (2000) Self-consistent modeling of non-linear effective
properties of polycrystalline ferroelectric ceramics. Computational Materials Science
19:123132
[22] Rdel J, Kreher WS, (2003) Modeling of linear and nonlinear behavior of
polycrystalline ferroelectric ceramics. Journal of the European Ceramic Society
23:22972306
[23] Ruschmeyer K, Helke G, Koch J, Lubitz K, Mckl T, Petersen A, Riedel M,
Schnecker A, (1995) Piezokeramik: Grundlagen, Werkstoffe, Applikationen. Expert-
Verlag, Renningen-Malmsheim
[24] Alguer M, Alemany C, Pardo L, Gonzlez AM, (2004) Method for obtaining the full
set of linear electric, mechanical and electromechanical coefficients and all related
losses of a piezoelectric ceramic. Journal of the American Ceramic Society 87:209
215
[25] Dantsiger AYa, Razumovskaya ON, Reznitchenko LA, Grineva LD, Devlikanova
RU, Dudkina SI, Gavrilyatchenko SV, Dergunova NV, Klevtsov AN, (1994) Highly
Effective Piezoceramic Materials (Handbook). Kniga, Rostov-on-Don (in Russian)
Piezoelectric Device-making. Vol. 1: Physics of Ferroelectric Ceramics.
[27] Haertling G, (1999) Ferroelectric ceramics: history and technology. Journal of the
American Ceramic Society 82:797818
[28] Bechmann R, (1956) Elastic, piezoelectric, and dielectric constants of polarized
barium titanate ceramics and some applications of the piezoelectric equations. Journal
of the Acoustical Society of America 28:347350
[29] Jaffe B, Cook WR, Jaffe H, (1971) Piezoelectric Ceramics. Academic Press, London
New York
[30] Ikegami S, Ueda I, Nagata T, (1971)

Electromechanical properties of PbTiO
3
ceramics
containing La and Mn. Journal of the Acoustical Society of America 50:10601066
[31] Nagatsuma K, Ito Y, Jyomura S, Takeuchi H, Ashida S, (1985) Elastic properties of
modified PbTiO
3
ceramics with zero temperature coefficients. In: Taylor GW (ed.)
Ferroelectricity and Related Phenomena. Vol. 4: Piezoelectricity. Gordon and Breach
Science Publishers, New York London Paris Montreux Tokyo, pp 167176
[32] Nelli Silva EC, Ono Fonseca JS, Kikuchi N, (1997) Optimal design of piezoelectric
microstructures. Computational Mechanics 19:397410
[33] Levassort F, Lethiecq M, Certon D, Patat F, (1997)

A matrix method for modeling
electroelastic moduli of 03 piezo-composites. IEEE Transactions on Ultrasonics,
Ferroelectrics, and Frequency Control 44:445452
[34] Zhang R, Jiang W, Jiang B, Cao W, (2002) Elastic, dielectric and piezoelectric
coefficients of domain engineered 0.70Pb(Mg
1/3
Nb
2/3
)O
3
0.30PbTiO
3
single crystal.
In: Cohen RE (ed.) Fundamental Physics of Ferroelectrics. American Institute of
Physics, Melville, pp 188197
[35] Peng J, Luo H, He T, Xu H, Lin D, (2005) Elastic, dielectric, and piezoelectric
characterization of 0.70Pb(Mg
1/3
Nb
2/3
)O
3
0.30PbTiO
3
single crystal. Materials
Letters 59:640643
1/3
Nb
2/3
)O
3
0.33PbTiO
3
Physics 90:34713475
[37] Cao H, Luo H, (2002) Elastic, piezoelectric and dielectric properties of
Pb(Mg
1/3
Nb
2/3
)O
3
38% PbTiO
3
single crystal. Ferroelectrics 274:309315
[38] Cao H, Schmidt VH, Zhang R, Cao W, Luo H, (2004) Elastic, piezoelectric, and
dielectric properties of 0.58Pb(Mg
1/3
Nb
2/3
)O
3
0.42PbTiO
3
single crystal. Journal of
Applied Physics 96:549554
[39] Yin J, Jiang B, Cao W, (2000) Elastic, piezoelectric, and dielectric properties of
0.955Pb(Zn
1/3
Nb
2/3
)O
3
0.045PbTiO
3
single crystals. IEEE Transactions on
Ultrasonics, Ferroelectrics, and Frequency Control 47:285291
[40] Zhang R, Jiang B, Cao W, Amin A, (2002) Complete set of material constants of
0.93Pb(Zn
1/3
Nb
2/3
)O
3
0.07PbTiO
3
domain engineered single crystal. Journal of
Materials Science Letters 21:18771879
[41] Jiang W, Zhang R, Jiang B, Cao W, (2003) Characterization of piezoelectric materials
with large piezoelectric and electromechanical coupling coefficients. Ultrasonics
41:5563
[42] Yin J, Jiang B, Cao W, (1999) Determination of elastic, piezoelectric and dielectric
properties of Pb(Zn
1/3
Nb
2/3
)O
3
PbTiO
3
single crystals. SPIE Conference on
Ultrasonic Transducer Engineering. San Diego, California, February 1999. SPIE
3664:239246
[44] Ye Z-G, Topolov VYu, (2001) Complex domain and heterophase structures in (1
x)Pb(Mg
1/3
Nb
2/3
)O
3
xPbTiO
3
single crystals. Ferroelectrics 253:7986
[45] Topolov VYu, (2002) Intermediate monoclinic phase and elastic matching in
perovskite-type solid solutions. Physical Review B 65:094207 6 p
[46] Topolov VYu, Turik AV, (2002) An intermediate monoclinic phase and
electromechanical interactions in xPbTiO
3
(1 x)Pb(Zn
1/3
Nb
2/3
)O
3
crystals. Solid
State Physics 44:13551362
[47] Aleshin VI, Luchaninov AG, (2001) Influence of mobility of the 90q domain walls on
the effective properties of PbTiO
3
ceramics. Journal of Physics D: Applied Physics
34:23532358
[48] Gururaja TR, Safari A, Newnham RE, Cross LE, (1988) Piezoelectric ceramic-
polymer composites for transducer applications. In: Levinson LM (ed.) Electronic
Ceramics: Properties, Devices, and Applications. Marcel Dekker, New York Basel,
pp 92128
[49] Newnham RE, (1994) Nonmechanical properties of composites. In: Kelly A, Cahn
RW, Bever MB (eds.) Concise Encyclopedia of Composite Materials. Elsevier,
Oxford, pp 214220
[50] Pardo L, Mendiola J, Alemany C, (1988) Theoretical treatment of ferroelectric
composites using Monte Carlo calculations. Journal of Applied Physics 64:50925097
[51] Chan HLW, Unsworth J, (1989) Simple model for piezoelectric ceramic / polymer 1
3 composites used in ultrasonic transducer applications. IEEE Transactions on
[52] Grekov AA, Kramarov SO, Kuprienko AA, (1989) Effective properties of a
transversely isotropic piezoelectric composite with cylindrical inclusions. Mechanics
of Composite Materials, 25:5461
[53] Levin VM, Rakovskaja MI, Kreher WS, (1999) The effective thermoelectroelastic
properties of microinhomogeneous materials. International Journal of Solids and
Structures 36:26832705
[54] Jensen H, (1991) Determination of macroscopic electro-mechanical characteristics of
13 piezoceramic / polymer composites by a concentric tube model. IEEE
[55] Uchino K, (2000) Ferroelectric Devices. Marcell Dekker, New York
[56] Nan C-W, Lin Y, (2002) Microstructure property linkages in multi-phase
electroceramics. Key Engineering Materials 228229:3742
[57] Turik AV, Radchenko GS, (2002) MaxwellWagner relaxation in piezoactive media.
[58] Radchenko GS, Turik AV, (2003) Giant piezoelectric effect in layered ferroelectric -
polymer composites. Physics of the Solid State 45:17591762
[59] Turik AV, Chernobabov AI, Radchenko GS, Turik SA, (2004) Giant piezoelectric and
dielectric enhancement in disordered heterogeneous systems. Physics of the Solid
State 46:22132216
[61] Luchaninov AG, (2002) Piezoelectric Effect in Non-polar Heterogeneous
Ferroelectric Materials. Volgograd State Academy of Architecture and Construction,
Volgograd (in Russian)
[62] Levin VM, (1995) The overall properties of piezoactive matrix composite materials.
In: Markov KZ (ed.) Continuum Models and Discrete Systems: Proceedings of the 8th
International Symposium, June 1116, 1995, Varna, Bulgaria. World Scientific,
Singapoure, pp 225232
[63] Berger H, Kari S, Gabbert U, Rodrguez-Ramos R, Bravo-Castillero J, Guinovart-
Daz R, (2005) A comprehensive numerical homogenization technique for calculating
effective coefficients of uniaxial piezoelectric fibre composites. Materials Science and
Engineering A 412:5360
[64] Akcakaya E, Farnell GW, (1988) Effective elastic and piezoelectric constants of
superlattices. Journal of Applied Physics 64:4469

4473
[65] Grekov AA,

Kramarov SO,

Kuprienko AA, (1987) Anomalous behavior

of

the

two-
phase lamellar piezoelectric texture. Ferroelectrics 76:4348
[66] Christensen RM, (1979) Mechanics of Composite Materials. Wiley, New York
[67] Telega JJ, (1990) Piezoelectricity and homogenization. Application to biomechanics.
In: Continuum Models and Discrete Systems. Vol. 2. Longman, London, pp 220230
[68] Agbossou A, Viet HN, Pastor J, (1999) Homogenization techniques and application to
piezoelectric composite materials. International Journal of Applied Electromagnetics
and Mechanics 10:391403
[69] Banno H, (1983) Recent development of piezoelectric ceramic products and
composite of synthetic rubber and piezoelectric ceramic particle. Ferroelectrics 50:3
12
[70] Ramesh R, Cara H, Bowen C R, (2004) Characteristics of piezoceramic and 33
piezocomposite hydrophones evaluated by finite element modelling. Computational
Materials Science 30:397403
[73] Furukawa T, Ishida K, Fukada E, (1979) Piezoelectric properties in the composite
systems of polymers and PZT ceramics. Journal of Applied Physics 50:49044912
[74] Yamamoto T, Urabe K, Banno H, (1993) BaTiO
3
particle-size dependence of
ferroelectricity in BaTiO
3
/ polymer composites. Japanese Journal of Applied Physics.
Pt. 1 32:42724276
[75] Chan HLW, Ng PKL, Choy CL, (1999) Effect of poling procedure on the properties
of lead zirconate titanate / vinylidene fluoride-trifluoroethylene composites. Applied
Physics Letters 74:30293031
[76] Chan HLW, Cheung MC, Choy CL, (1999) Study on BaTiO
3
/ P(VDF-TrFE) 03
composites. Ferroelectrics 224:113120
[77] Ng KL, Chan HLW, Choy CL, (2000) Piezoelectric and pyroelectric properties of
PZT / P(VDF-TrFE) composites with constituent phases poled in parallel or
antiparallel directions. IEEE Transactions on Ultrasonics, Ferroelectrics, and
[78] Fang D-N, Soh AK, Li C-Q, Jiang B, (2001) Nonlinear behavior of 03 type
ferroelectric composites with polymer matrices. Journal of Materials Science
36:52815288
[79] Lam KH, Chan HLW, (2005) Piezoelectric and pyroelectric properties of 65PMN
35PT / P(VDF-TrFE) 03 composites. Composites Science and Technology 65:1107
1111
[80] Glushanin SV, Topolov VYu, (2005) A hierarchy of inclusions and electromechanical
properties of 03 ceramic / polymer composites. Journal of Physics D: Applied
Physics 38:24602467
[81] Glushanin SV, Topolov VYu, Krivoruchko AV, (2006) Features of piezoelectric
properties of 03-type ceramic / polymer composites. Materials Chemistry and
Physics 97:357364
[82] Banno H, Saito S, (1983) Piezoelectric and dielectric properties of composites of
synthetic rubber and PbTiO
3
or PZT. Japanese Journal of Applied Physics 22 (Suppl.
2):6769
[83] Banno H, (1995)

Theoretical equations for dielectric, elastic and piezoelectric
constants of diphasic composite changing its connectivity from 30 to 03 via 33. In:

Pandey RK, Liu M, Safari A (eds.) ISAF94:

Proceedings of the Ninth IEEE
International Symposium on Applications of Ferroelectrics, University Park,

PA,

USA, August 710, 1994.

IEEE, Piscataway, pp

186189
[85] Hashimoto KY, Yamaguchi M, (1986) Elastic, piezoelectric and dielectric properties
of composite materials. In: Proceedings of IEEE Ultrasonic Symposium,
Williamsburg, Va, November 1719, 1986. Vol. 2. New York, pp 697702.
[86] Levassort F, Lethiecq M, Millar C, Pourcelot L, (1998) Modeling of highly loaded 0
3 piezoelectric composites using a matrix method. IEEE Transactions on Ultrasonics,
[87] Levassort F, Topolov VYu, Lethiecq M, (2000) A comparative study of different
methods of evaluating effective electromechanical properties of 03 and 13 ceramic /
polymer composites. Journal of Physics D: Applied Physics 33:20642068
[88] Glushanin SV, Topolov VYu, (2001) Features of electromechanical properties of
piezoelectric composites with elements of connectivity 11. Journal of Physics D:
[89] Topolov VYu, Glushanin SV, (2002) Evolution of connectivity patterns and links
between interfaces and piezoelectric properties of two-component composites. Journal
of Physics D: Applied Physics 35:20082014
[90] Benveniste Y, (1992) The determination of the elastic and electric fields in a
piezoelectric inhomogeneity. Journal of Applied Physics 72:1086-1095
[91] Wang B, (1992)

Three-dimensional analysis of an ellipsoidal inclusion in a
piezoelectric material. International Journal of

Solids and Structures 29:293308
[92] Dunn ML, Taya M, (1993) An analysis of piezoelectric composite materials
containing ellipsoidal inhomogeneities. Proceedings of the

Royal Society (London),
Pt A 443:265287
[93] Dunn ML, Taya M, (1993) Micromechanics predictions of the effective electroelastic
moduli of piezoelectric composites. International Journal of Solids and Structures
30:161175
[94] Dunn ML, (1993) Micromechanics of coupled electroelastic composites: Effective
thermal expansion and pyroelectric coefficients. Journal of Applied Physics 73:5131
5140
[95] Dunn ML, Wienecke HA, (1997) Inclusions and inhomogeneities in transversely
isotropic piezoelectric solids. International Journal of Solids and Structures 34:3571
3582
[96] Eshelby J, (1957) The determination of the elastic field of an ellipsoidal inclusion, and
related problems. Proceedings of the

Royal Society (London), Pt A 241:376396
[97] Eshelby J, (1959) The elastic field outside an ellipsoidal inclusion. Proceedings of the

Royal Society (London), Pt A 252:561569
[98] Mura T, (1987) Micromechanics of Defects in Solids. 2nd edn. Martins Nijhoff,
Dordrecht
[99] Dunn ML, Wienecke HA, Li JY, (1997). Multiple-scale micromechanics of
heterogeneous piezoelectric media: defects, ceramics, and composites. In: Inoue K,
Shen SIY, Taya M (eds.) Proceedings of the First US Japan Workshop on Smart
Materials and Structures, Warrendale, 1996. The Minerals, Metals & Materials
Society, Warrendale, pp 203215
[100] Wu T-L, (2000) Micromechanics determination of electroelastic properties of
piezoelectric materials containing voids. Materials Science and Engineering
A280:320327
[101] Huang JH, Yu S, (1994) Electroelastic Eshelby tensors for an ellipsoidal piezoelectric
inclusion. Composites Engineering 4:11691182
[102] Mikata Y, (2001) Explicit determination of piezoelectric Eshelby tensors for a
spheroidal inclusion. International Journal of Solids and Structures 38:70457063
[103] Nan C-W, (1994) Effective-medium theory of piezoelectric composites. Journal of
[104] Fakri N, Azrar L, El Bakkali L, (2003) Electroelastic behavior modeling of
piezoelectric composite materials containing spatially oriented reinforcements.
International Journal of Solids and Structures 40:361384
[105] Huang JH, Kuo W-S, (1996) Micromechanics determination of the effective
properties of piezoelectric composites containing spatially oriented short fibers. Acta
Materialia 44:48894898
[106] Jiang B, Fang D-N, Hwang K-C, (1999) A unified model for piezocomposites with
non-piezoelectric matrix and piezoelectric ellipsoidal inclusions. International Journal
of Solids and Structures 36:27072733
[107] Mori T, Tanaka K, (1973) Average stress in matrix and average elastic energy of
materials with misfitting inclusions. Acta Metallurgica 21:571574
[108] Topolov VYu, Kamlah M, (2004) Piezoelectric properties of PbTiO
3
-based 03 and
013 composites. Journal of Physics D: Applied Physics 37:15761585
[109] Dunn ML, Taya M, (1993) Electromechanical properties of porous piezoelectric
ceramics. Journal of the American Ceramic Society 76:16971706
[110] Huang JH, Chiu Y-H, Liu H-K, (1988) Magneto-electro-elastic Eshelby tensors for a
piezoelectric-piezomagnetic composite reinforced by ellipsoidal inclusions.

Journal of
[111] Huang JH, (1998) Analytical predictions for the magnetoelectric coupling in
piezoelectric materials reinforced by piezoelectric ellipsoidal inclusions. Physical
Review B 58:1215
[112] Wu T-L, Huang JH, (2000) Closed-form solutions for the magnetoelectric coupling
coefficients in fibrous composites with piezoelectric and piezomagnetic phases.
[113] Pettermann HE, Suresh S, (2000) A comprehensive unit cell model: a study of
coupled effects in piezoelectric 13 composites. International Journal of Solids and
[114] Bowen CR, Perry A, Stevens R, Mahon S, (2001) Analytical and numerical modelling
of 33 piezoelectric composites. Integrated Ferroelectrics 32:333342
[115] Kara H, Perry A, Stevens R, Bowen CR, (2002) Interpenetrating PZT / polymer
composites for hydrophones: Models and experiments. Ferroelectrics 265:317332
[116] Jayasundere N, Smith BV, (1993) Dielectric constant for binary piezoelectric 03
composite. Journal of Applied Physics 73:24622466
[117] Poizat C, Sester M, (2001) Homognisation priodique de composites
pizolectriques 03: influence de la distribution. Revue des Composites et des
Matriaux Avancs 11:6574
[118] Kar-Gupta R, Venkatesh TA, (2005) Electromechanical response of 13 piezoelectric
composites: effect of poling characteristics. Journal of Applied Physics 98:054102
14 p
[119] Shuvalov LA, Ourousovskaya AA, Zheludev IS, Zalessky AV, Semiletov SA,
Grechushnikov BN, Chistyakov IG, Pikin SA, (1981) Modern Crystallography. Vol.
4: Physical Properties of Crystals. Nauka, Moscow (in Russian)
[120] Topolov VYu, (1995) Anisotropy of electromechanical properties in KNbO
3
crystals
with S-type domain boundaries. Journal of Physics: Condensed Matter 7:74057408
[121] Topolov VYu, Turik AV, (1998) Electromechanical constants and their anisotropy in
LiNbO
3
-type crystals having 180q inclined domain walls. Journal of Physics:
Condensed Matter 10:451459

3
Non-monotonic Volume-fraction Dependences of
Effective Properties in o| Ceramic / Polymer
Piezo-composites
3.1 Composites with 22 Connectivity, Manufacturing and
Applications
Since the 1970s, there has been significant progress in the development and
application of piezo-composites consisting of ferroelectric ceramic and polymer
components [16]. The o| ferroelectric ceramic / polymer composites combine
high piezoelectric activity of a ceramic with the beneficial mechanical properties of
a polymer. A polymer is often chosen as the piezo-passive material and is
characterised by low dielectric permittivity and low elastic modulus. By combining
the physical properties of the ceramic and polymer components and tailoring the
microgeometrical features (see, e.g., Figure 1.2) significant advantages can be
achieved over the performance of the individual piezo-active component.
Some advantages of the o| piezo-composites are concerned with a variety of
non-monotonic dependences of the effective electromechanical properties on the
volume fractions of the ceramic components. Many examples of such dependences
are known from work on the 22 ceramic / polymer composites with the series
connection (Figure 3.1a) and the parallel connection (Figure 3.1b) of layers. The
22 series-connected composites based on the Pb(Zr, Ti)O
3
-type ceramic are
characterised by several types of volume-fraction dependence of the thickness
electromechanical coupling factor
*
t
k
=
*
33
e / (
D
c
*
33

c
*
33
)
1/2
[7, 8] and elastic
modulus
E
c
*
11
measured at an electric field E = const [9], where
*
33
e is the
piezoelectric coefficient,
D
c
*
33
is the elastic modulus (at electric displacement D =
const) and
c
*
33
is the dielectric permittivity (at mechanical strain = const). The
calculated volume-fraction dependences of the piezoelectric coefficient
*
31
e
and
dielectric permittivity
o
c
*
11
show maxima, and their values are highly dependent on
the ratio between the elastic moduli of the individual components [9]. The

Figure 3.1. Schematic drawing of the 22 series-connected (a) and parallel-connected (b)
composites. m and 1 m are the volume fractions of components C1 (n = 1) and C2 (n = 2).
(X
1
X
2
X
3
) is the rectangular co-ordinate system
piezoelectric coefficient
*
31
e of the 22 composite [10] can pass through the zero
value and reach max
*
31
e > 0, although both the components (n = 1 and 2) have
negative values of the piezoelectric coefficient
) (
31
n
e . A squared strainvoltage
Non-monotonic Volume-fraction Dependences of Effective Properties 45
figure of merit
2 *
33
) (Q =
*
33
d
*
33
g
also exhibits a maxima [11] in the 22 composites
based on the Pb(Zr, Ti)O
3
-type ceramic.
In comparison to the series-connected composite in Figure 3.1a, the parallel-
connected composite (Figure 3.1b) contains layers alternating in the OX
1
direction.
This structure is characterised by an anisotropy of the electromechanical properties
in the (X
1
OX
2
) plane. The corresponding 22 composite based on the Pb(Zr, Ti)O
3
-
type ceramic also exhibits the non-monotonic behaviour of a number of
parameters. For example, the thickness electromechanical coupling factor
*
t
k
and
the hydrostatic (or hydrophone) piezoelectric coefficients
*
h
d
=
*
33
d
+
*
32
d
+
*
31
d
and
*
h
g
=
*
33
g
+
*
32
g
+
*
31
g
pass through maxima [12, 13] on varying the volume fraction
of the ceramic component.
The simplicity of the 22 structure (Figure 1.2) and the possibility of varying
and tailoring the volume fractions of the components (Figure 3.1) in a wide range,
enable the manufacture and design of composites with extreme values of
parameters. Methods suitable for manufacturing the 22 ferroelectric ceramic /
polymer composites are discussed in a series of review papers (see, e.g., [46]).
The simplest method is the dice-and-fill process [4, 5] whereby a series of parallel
cuts is made in a ceramic bulk sample (block). The cuts attached to the ceramic
base are backfilled with polymer, and the resulting filled composite consisting of a
system of alternating ceramic and polymer layers is finally removed from the
ceramic base. In the injection-moulding method [4, 5, 14] preforms containing a
number of thin ceramic sheets are initially prepared and then heat treated. The heat
treatment enables the removal of the binder and subsequent sintering. Final filling
of the structure with a polymer results in a fine-scale 22 composite. The tape-
lamination method [4, 5, 15] utilises thin ceramic sheets that are either laminated
with a passive material or stacked with a spacer, and then filled with polymer. Jet
machining [4, 5, 16] is applied to a green ceramic tape that contains the powder
and binder. The main stages of this method are lamination with a photoresist,
exposing through a photomask, the development of the photoresist, and jet
machining. This method is often used to manufacture the 22 ceramic / polymer
composite with a system of annular rings [4, 5]. The co-extrusion method is also
suitable for manufacturing the 22 composites [4, 5]. The initial stage of co-
extrusion involves mixing ceramic powder with a thermoplastic polymer and then
extruded the mixture to form a sheet. This sheet is then laminated with a carbon-
filled thermoplastic sheet. Finally, the co-extrusion of these laminates is
implemented and results in the formation of the 22 composite.
Recently, rapid prototyping methods (or solid freeform fabrication) [6] have
become attractive and enable the manufacture of fine-scale 22 and other piezo-
composites using computer-aided design. Among these methods, those suitable for
the fabrication of separate polymer and ceramic parts [5, 6] with specified
microgeometry are of interest. The main stages of manufacturing are the fused
deposition of the ceramic, filament fabrication, the build process (using computer-
aided design), and the post-processing stage. As a result, a number of 22 piezo-
composites [6, 7, 17, 18] have been manufactured, namely, composites with curved
ceramic skeletons, composites with volume-fraction gradient, radial composites
and oriented composites.
The 22 piezo-composites find various applications, mainly as active elements
of sensors, actuators and transducers for medical-imaging applications [5, 6]. In
addition, the transducer can perform various functions via conversion of acoustic
energy into the electric energy or vice versa. The transducer can also combine
passive and active functions in, for example, the pulseecho mode. The 22 piezo-
composites based on the Pb(Zr, Ti)O
3
-type ceramic are often used as cantilevers or
multilayers [19]. The 22 ceramic / polymer composites that exploit the transverse
piezoelectric mode for hydroacoustic applications are also of interest [19, 20]. In
the 1990s, novel 22-type structures became widespread due to the development of
moonies and cymbals [19, 20] that exhibit high actuator performance owing to
the combination of ferroelectric ceramic layers and metal elements as strain
transformers. These novel transducers provide both a considerable displacement
and a large generative force for hydrophone, actuator and transceiver applications.
3.1.1 Effective Electromechanical Constants of 22 Composites
In the following section we will consider the effective electromechanical properties
in simple series- and parallel-connected composites (Figure 3.1) that are termed 2
2 ceramic / polymer composites. The orientation of the interfaces between the
layers with respect to the poling direction influences these properties. Alternating
the ceramic (C1) and polymer (C2) layers along the poling direction OX
3
(Figure
3.1a) or along the OX
1
direction, perpendicular to the poling direction (Figure
3.1b), results in different boundary conditions [13, 21] for the internal electric and
mechanical fields and, as a consequence, influences electromechanical coupling in
the 22 structures.
It is assumed that the layers of the 22 composite have a large length in the OX
1
and OX
2
directions (Figure 3.1a) or in the OX
2
and OX
3
directions (Figure 3.1b).
The interfaces separating these layers represent planes, and perfect bonding
between the adjacent layers is assumed. To determine the full set of effective
electromechanical constants of the 22 composite, a matrix method is used [21,
22]. The electromechanical constants of the components are given by the 9 9
matrices
|| C
(n)
||

=

|
|
.
|
\
|
|| ||
|| ||
|| ||
|| ||
), (
) (
) (
), (
o
c
n
t n
n
E n
d
d
s
(3.1)
where || s
(n),E
|| is the 6 6 matrix of elastic compliances, || d
(n)
|| is the 6 3 matrix
of piezoelectric coefficients and || c
(n),o
|| is the 3 3 matrix of dielectric
permittivities of the ceramic (n = 1) and polymer (n = 2) components, and
superscript t denotes the transposition. A matrix || C* || of the effective
electromechanical properties in the 22 composite is written [21, 22] in the form
|| C* || = [|| C
(1)
||
.
|| M || m + || C
(2)
|| (1 m)] [|| M || m + || I || (1 m)]
1
(3.2)

The structure of the matrix || C* || is similar to that shown in Equation 3.1. The
effective properties of the composite are determined using Equation 3.2 in a long-
wave approximation [21], i.e., in a case when the wavelength of an external
acoustic field is much longer than the thickness of the separate layer (C1 or C2, see
Figure 3.1) in the composite. In Equation 3.2 || M || is the matrix related to the
boundary conditions at interfaces x
i
= const (i = 1 or 3), and || I || is the identity
matrix. The boundary conditions, for example, at x
1
= const (Figure 3.1b) imply a
continuity of components of mechanical stress o
11
, o
12
and o
13
, strain
22
,
23
and
33
, electric displacement D
1
, and electric field E
2
and E
3
. By taking into account
the transversal isotropy of the poled ceramic component and that the polymer
component can be isotropic (piezo-passive) or transversely isotropic (piezo-active)
[10, 12, 13], the matrix || M || for interfaces x
1
= const is || M || = || m
1
||
1
|| m
2
||
where

|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
.
|
\
|
=
1 0 0 0 0 0 0 0 0
0 1 0 0 0 0 0 0 0
0 0 0 0 0 0 0
0 0 0 1 0 0 0 0 0
0 0 0 0 1 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0 0 0 0 1
|| ||
), (
11
) (
15
) (
15
), (
44
) (
33
), (
33
), (
13
), (
13
) (
31
), (
13
), (
11
), (
12
o
c
n u
n E n
n E n E n E n
n E n E n E n
n
d
d s
d s s s
d s s s
m

is written in terms of the electromechanical constants of the components (n = 1 or
2).
Taking the matrix elements
*
ab
C from Equation 3.2 and using Equations 2.12
2.19, it is possible to determine the effective piezoelectric coefficients
*
kl
d
(m),
*
kl
e
(m),
*
kl
g
(m), and
*
kl
h
(m) for the 22 composites (Figure 3.1) where the volume
fraction m is varied from 0 to 1.

3.1.2 Examples of Volume-fraction Dependences Predicted for 22 Composites
In this section we consider the piezoelectric properties of 22 composites based on
the Pb(Zr, Ti)O
3
- or PbTiO
3
-type ceramics. As can be seen in Table 2.1, these
ferroelectric ceramics are distinguished by a piezoelectric anisotropy, dielectric
properties and signs of the piezoelectric coefficients
) 1 (
31
e
. Calculations are
performed using the room-temperature experimental data on typical ferroelectric
ceramics (see Table 2.1) and polymer (araldite from ref. [23]).
Of particular interest are the volume-fraction dependences of the four types of
piezoelectric coefficients
*
3l
( = d, e, g, or h) that characterise an electric
response of the composites in the poling direction OX
3
when an external
mechanical field is applied along one of the co-ordinate axes OX
l
(Figure 3.1). The
22 composites examined are characterised by mm (series connection) or mm2
(parallel connection) symmetry. A relation
*
31
=
*
32
=
*
33
holds for the ceramic /
polymer series-connected composites with interfaces x
3
= const (Figure 3.1a). In
this case there is no distinction between the response along the OX
1
and OX
2
axes.
The orientation of interfaces x
1
= const (Figure 3.1b) results in a reduction of the
macroscopic symmetry of the composite so that a relation
*
31
=
*
32
=
*
33
is held.
The series-connected composites are characterised by relatively small values of
*
3l
e
and
*
3l
d
at volume fractions 0 < m < 0.9 due to the poor orientation of the layers
(Figure 3.1a) with respect to the poling direction. In the external electric field E ||
OX
3
a considerable depolarisation field appears in the ceramic layers C1, and these
layers provide a small contribution to the piezoelectric response of the composite
as a whole. For example, at m = 0.5 the PZT-5 ceramic / araldite composite is
characterised by the following values:
*
31
e
= 1.0710
2
) 1 (
31
e
,
*
33
e
= 5.2710
4
) 1 (
31
e
,
*
31
d
= 4.6910
3
) 1 (
31
d
, and
*
33
d
= 3.5910
3
) 1 (
33
d
. In addition, the studied composite
attains values
*
31
g
= 0.956
) 1 (
31
g
,
*
33
g
= 0.729
) 1 (
33
g
and
*
31
h
= 1.07
) 1 (
31
h
. For the
(Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
(I) ceramic / araldite composite at m = 0.5 the
values
*
33
e
= 2.6910
3
) 1 (
33
e
,
*
33
d
= 2.5310
2
) 1 (
33
d
,
*
31
g
= 0.957
) 1 (
31
g
, and
*
33
g
=
0.543
) 1 (
33
g
are predicted. A relatively slow increase in the dielectric permittivities
o
c
*
33
(m) and
c
*
33
(m) of the composites leads to relatively large values of the

*
3l
g
(m) =
*
3l
d
(m) /
o
c
*
33
(m) (3.3)
*
3l
h
(m) =
*
3l
e
(m) /
c
*
33
(m) (3.4)
in comparison with
*
3l
d
(m) and
*
3l
e
(m) (see Equations 2.122.15).
The volume-fraction dependence of the piezoelectric coefficients of the
parallel-connected composites (Tables 3.1 and 3.2) indicates that such an
arrangement of the layers C1 and C2 (see Figure 3.1b) leads to

a significant
Table 3.1. Volume-fraction dependences of the piezoelectric coefficients
*
3l
e
(in C / m
2
),
*
3l
d
(in pC / N),
*
3l
g
(in mV

m / N), and
*
3l
h
(in 10
8
V / m), which have been calculated for the
22 PZT-5 ceramic / araldite parallel-connected composite
m
*
31
e
*
32
e
*
33
e
*
31
d
*
32
d
*
33
d
0 0 0 0 0 0 0
0.05 0.0187 0.112 0.944 43.7 61.2 136
0.1 0.0394 0.225 1.89 67.0 92.4 205
0.2 0.0878 0.456 3.78 93.1 124 273
0.3 0.149 0.695 5.66 109 140 307
0.4 0.228 0.945 7.53 121 149 328
0.5 0.335 1.21 9.38 131 156 342
0.6 0.487 1.51 11.2 140 160 352
0.7 0.721 1.86 13.0 148 163 359
0.8 1.13 2.31 14.6 155 167 365
0.9 2.01 3.07 16.0 163 169 369
1 5.40 5.40 15.8 170 170 373
m
*
31
g
a *
32
g
b *
33
g
c *
31
h
*
32
h
*
33
h
d
0 0 0 0 0 0 0
0.05 79.4 111 247 0.452 2.70 22.9
0.1 55.7 76.8 170 0.497 2.85 23.9
0.2 35.2 46.9 103 0.568 2.95 24.5
0.3 26.3 33.7 74.0 0.647 3.02 24.6
0.4 21.3 26.2 57.6 0.747 3.10 24.7
0.5 18.1 21.5 47.2 0.880 3.19 24.7
0.6 15.9 18.2 40.0 1.07 3.32 24.6
0.7 14.3 15.8 34.6 1.36 3.51 24.5
0.8 13.1 13.9 30.6 1.87 3.83 24.3
0.9 12.1 12.5 27.3 2.98 4.55 23.7
1 11.3 11.3 24.7 7.34 7.34 21.5
Notes to Table 3.1
a.

min
*
31
g
=
*
31
g
(0.016) = 102 mV

m / N

b. min
*
32
g
=
*
32
g
(0.016) = 143 mV

m / N
c. max
*
33
g
=
*
33
g
(0.015) = 320 mV

m / N
d. A diffuse maximum of
*
33
h
(
*
33
h
(m) ~ 2.4710
9
V / m) is attained at 0.334 s m s 0.546
increase in
*
33
e
(m),
*
33
d
(m) and
*
3l
h
(m) on increasing the volume fraction m of the
ceramic layers. In both composites the piezoelectric coefficients
*
3l
g
(m) show non-
monotonic behaviour at fairly small m values. This behaviour is due to the
relatively rapid increase in the piezoelectric coefficients
*
3l
d
(m) at small dielectric
permittivity
o
c
*
33
(m) in a range of 0 < m << 1 (see Equation 3.3). In contrast to the
peaks of
*
3l
g
(m), a diffuse maximum is peculiar to
*
33
h
(m) in the composite based
on the PZT-5 ceramic (see Table 3.1): in this case a slow increase in
c
*
33
(m) cannot
Table 3.2. Volume-fraction dependences of the piezoelectric coefficients
*
3l
e
(in C / m
2
),
*
3l
d
(in pC / N),
*
3l
g
(in mV

m / N), and
*
3l
h
(in 10
8
V / m), which have been calculated for
the 22 (Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
(I) ceramic / araldite parallel-connected composite
m
*
31
e
*
32
e
*
33
e
*
31
d
a *
32
d
b *
33
d
0 0 0 0 0 0 0
0.05 0.00134 0.0185 0.322 12.3 5.22 28.8
0.1 0.00283 0.0370 0.644 15.9 5.39 37.2
0.2 0.00631 0.0741 1.29 17.6 5.11 43.8
0.3 0.0107 0.111 1.93 17.1 4.89 46.5
0.4 0.0164 0.149 2.58 15.8 4.75 48.0
0.5 0.0242 0.187 3.22 14.2 4.64 49.0
0.6 0.0354 0.226 3.86 12.4 4.57 49.6
0.7 0.0528 0.266 4.51 10.5 4.51 50.1
0.8 0.0837 0.308 5.16 8.50 4.47 50.5
0.9 0.153 0.359 5.81 6.47 4.43 50.8
1 0.458 0.458 6.50 4.40 4.40 51.0
m
*
31
g
c *
32
g
d *
33
g
e *
31
h
*
32
h
*
33
h
0 0 0 0 0 0 0
0.05 116 49.1 270 0.138 1.90 33.1
0.1 88.0 29.8 205 0.179 2.35 40.9
0.2 52.8 15.3 131 0.227 2.67 46.3
0.3 35.1 10.0 95.4 0.269 2.80 48.5
0.4 24.7 7.40 74.9 0.317 2.87 49.7
0.5 17.8 5.84 61.6 0.379 2.93 50.4
0.6 13.1 4.81 52.3 0.467 2.97 50.9
0.7 9.50 4.09 45.4 0.601 3.02 51.3
0.8 6.76 3.55 40.1 0.837 3.09 51.6
0.9 4.58 3.14 35.9 1.37 3.21 51.9
1 2.81 2.81 32.6 3.70 3.70 52.5
Notes to Table 3.2
a.

min
*
31
d
=
*
31
d
(0.207) = 17.6 pC / N

b. min
*
32
d
=
*
32
d
(0.085) = 5.41 pC / N
c. min
*
31
g
=
*
31
g
(0.034) = 121 mV

m / N
d. min
*
32
g
=
*
32
g
(0.021) = 62.8 mV

m / N
e. max
*
33
g
=
*
33
g
(0.031) = 285 mV

m / N
compete with a rapid increase in
*
33
e
(m) in the wide m range (see Equation 3.4). It
is seen that both the composites show the large anisotropy of
*
3l
e
(m) and
*
3l
h
(m) at
various m values, i.e., conditions
*
33
e
/ |
*
3 f
e
| >> 1 and
*
33
h
/ |
*
3 f
h
| >> 1 (3.5)
hold, where f = 1 and 2.
The composite based on the PbTiO
3
-type ceramic is also of interest due to
positive values of the piezoelectric coefficients
*
3l
e
(m) and
*
3l
h
(m) in the complete
m range (l = 1, 2 and 3, see Table 3.2). At the same time, the piezoelectric
coefficients
*
3l
d
(m) have different signs. According to Equation 2.12, these
piezoelectric coefficients in the parallel-connected composite (mm2 symmetry) are
interrelated by formulae

*
31
d
=
*
31
e
E
s
*
11
+
*
32
e
E
s
*
12
+
*
33
e
E
s
*
13
(3.6)

*
32
d
=
*
31
e
E
s
*
12
+
*
32
e
E
s
*
22
+
*
33
e
E
s
*
23
(3.7)

*
33
d
=
*
31
e
E
s
*
13
+
*
32
e
E
s
*
23
+
*
33
e
E
s
*
33
(3.8)
An analysis of Equations 3.63.8 in connection with data from Table 3.2 and
conditions 3.5 suggests that the volume-fraction dependence
*
31
d
(m) is primarily
due to the item
*
33
e
E
s
*
13
< 0 and closely connected with a competition between
items
*
32
e
E
s
*
22
> 0 and
*
33
e
E
s
*
23
< 0. In addition, an item
*
33
e
E
s
*
33
> 0 plays a
significant role in forming the dependence
*
33
d
(m). It is obvious that the large
piezoelectric anisotropy, peculiar to the PbTiO
3
-type ceramic layers (wherein
) 1 (
33
d /
|
) 1 (
31
d | > 10), cannot affect both the
*
33
d
/
*
31
d
and
*
33
d
/
*
32
d
ratios in the wide m
range (Table 3.2) due to changes in elastic compliances
E
ab
s
*
from Equations 3.6
3.8. Undoubtedly, these changes are associated with the parallel-connected layers
(Figure 3.1b) that promote the appreciable elastic anisotropy at various volume
fractions m.
It should be added that analytical formulae are known for the effective
electromechanical constants of both the series-connected [13] and parallel-
connected [12, 13] piezo-composites (Figure 3.1) with a regular arrangement of the
layers C1 and C2. A comparison of data from Tables 3.1 and 3.2 with the results
obtained using formulae [12, 13] shows that a difference between the similar
constants does not exceed 1% in the range of 0 < m < 1. A further comparison of
the volume-fraction dependences of the electromechanical properties in the series-
connected composites can be made in order to account for [2426] the electric
conductivity of the layers. The volume-fraction dependences of the effective elastic
properties determined using Equation 3.2 coincide with those predicted [24] taking
into account the electric conductivity of the layers. A discrepancy between the
effective properties (for example, volume-fraction dependences
*
3l
d
(m) or
o
c
*
33
(m))
predicted using the matrix approach [21, 22] and the same properties determined
using formulae [24, 26] is accounted for by the distribution of internal electric and
mechanical fields in the layered structure with an element of conductivity of the
components. A possible giant increase in the dielectric and piezoelectric
properties in the series-connected composites is expected [24, 26] at low electric
field frequencies, i.e., at the condition et << 1 where e is the circular frequency of
the electric field E and t is the MaxwellWagner relaxation time. In this case, the
role of the electric conductivity of the layers of the composite and the
redistribution of electric charges near the interfaces separating these layers and
oriented perpendicular to the electric field E is of importance. Such behaviour of
the effective properties in the 22 series-connected piezo-composites could lead to
potentially new applications for these composite materials in the infralow-
frequency range.
3.2 Composites with 13 Connectivity, Manufacturing
and Applications
In the 13 ceramic / polymer composite system, the ceramic component is self-
connected in one dimension (often along the poling axis) and the polymer
component is self-connected in three dimensions. The 13 composites (Figures 1.2
and 3.2) are widespread due to the ease of poling and a variety of advantages over
the poled monolithic ferroelectric ceramics [36, 2731]. The ceramic component
can be represented by a continuous row of particles, relatively long rods or fibres,
or a series of discs. The combination of the relatively low dielectric permittivity
and high piezoelectric activity at small volume fractions of the ceramic component
(less than 10%) results in significant piezoelectric sensitivity of the 13
ferroelectric ceramic / polymer composite [32, 33]. These particular composite
properties stimulated intensive research into the development and manufacture of
13 ceramic / polymer composites in the last decades (see, e.g., [46, 2729, 34
36]).
Simultaneously, much research has been devoted to the analysis of the non-
monotonic dependence of the 13 composites parameters on the volume fraction
of the ceramic component. For example, a number of papers report on extreme
*
3 j
d
[27, 3739],
*
33
e
[23, 38, 40, 41],
*
33
g
[33, 42], and
*
33
h

[33], hydrostatic piezoelectric coefficients
*
h
d
[27, 29, 32, 34] and
*
h
g
[27, 32, 34,
43], the thickness electromechanical coupling factor
*
t
k
[23, 28, 29], and squared
hydrostatic figure of merit
2 *
) (
h
Q
=
*
h
d
*
h
g
[30, 34, 43, 44]. Due to the rod shape of
the highly piezo-active inclusions (Figure 3.2) oriented along the poling field the
piezoelectric anisotropy values
,
e
=
*
33
e
/
*
31
e
(3.9)
,
d
=
*
33
d
/
*
31
d
(3.10)
are varied over wide ranges [45].


Figure 3.2. Schematic drawing of the 13 composite with cylindrical rods. m and 1 m are
volume fractions of the rods (component 1) and the matrix (component 2). (X
1
X
2
X
3
) is the
rectangular co-ordinate system
Various methods for manufacturing 13 ferroelectric ceramic / polymer
composites have been developed. The first attempts to create 13 composites were
associated with using a coarse PZT powder (i.e., a composition of the Pb(Zr, Ti)O
3

type) and polyurethane [3]. The average particle size approached the thickness of
the composite sample and promoted saturation poling under an external electric
field. The 13 composites were also made of PZT spheres or discs embedded in
polymers [3]. Klicker et al. [30] proposed embedding PZT rods, aligned in a
fixture, in a polymer matrix. A further development of this processing resulted in a
rod placement (or pick-and-place) method [4, 5]. In this method, which is
appropriate for a wide ceramic volume-fraction range, different microgeometrical
features such as arrangement of rods, rod size and shape of the base of the rod, etc.
are considered.
The methods for manufacturing the 22 ceramic / polymer composites (see
Section 3.1) also became popular for the 13 composite system. These methods
include dice-and-fill, injection moulding, co-extrusion, tape-lamination, and rapid
prototyping methods [36, 31]. The lost-mould method for manufacturing the 13
composites [4] uses a plastic mould that has been pre-prepared. The mould
contains a negative structure relative to an intended composite structure and the
negative mould is filled with ceramic / polymer slurry and subjected to a heat
treatment. During the heat treatment, the binder is burned out and the ceramic
structure is sintered. After sintering, this structure is filled with the polymer
component, i.e., the 13 composite is formed. Dielectrophoresis [4, 46] is also
suitable for manufacturing the 13 ceramic / polymer composites. This method
provides an ordering of a colloidal suspension of ceramic particles in a liquid
polymer under an external electric field. Due to the non-uniform electric field in
the medium, the ceramic particles attract each other and form a chain-like structure
that can be continuously distributed between electrodes.
The 13 ferroelectric ceramic / polymer composites that exhibit high
performance find various applications, for example, as active elements of
hydrophones, actuators, ultrasonic transducers operating in the pulse-echo mode,
medical-imaging devices [35, 28, 31]. In a high-frequency range (~ 10
6
10
7
Hz),
the 13 composites are exploited for non-destructive testing and biomedical
diagnostics [4, 28, 31].
The 13 piezo-composites represent interesting examples of volume fraction
dependencies of the effective piezoelectric properties. In the following section we
consider two typical cases of volume-fraction behaviour of these properties in the
composites based on the Pb(Zr, Ti)O
3
or PbTiO
3
ceramics.
3.2.1 Examples of Volume-fraction Dependences Predicted for 13
Composites
During modelling of 13 composite system it is assumed that the composite
consists of very long ceramic rods (component 1) that are regularly distributed in
the continuous polymer matrix (component 2). The rods are considered to be in the
form of circular cylinders aligned parallel to the poling OX
3
axis and continuous in
the same direction. In the (X
1
OX
2
) plane a square arrangement of the cylinders is
assumed. Both the components can be piezo-active in a general case. The 13
piezo-composite as a whole is regarded as a transversely isotropic material.
Electromechanical properties of its components are described by sets of the elastic
moduli
E n
ij
c
), (
, piezoelectric coefficients
) (n
kl
e
and dielectric permittivities
c
), (n
pp

where n = 1 and 2. The volume fractions of the ceramic and polymer components
are m and 1 m, respectively.
Examples of calculated volume-fraction dependences of the four types of
effective piezoelectric coefficients of the 13 ceramic / polymer composites are
shown in Figures 3.3 and 3.4. The components are identical to those chosen for the
analysis of the piezoelectric properties of the 22 composites (see Section 3.1.2).
Calculations were performed on the basis of the EFM that enables analytical
expressions to be obtained [32] for
*
kl
e
,
*
kl
d
,
E
ij
c
*
, and
o
c
*
pp
.
A simple comparison of results from Figures 3.3 and 3.4 suggests that,
independently of sgn
) (n
kl
e
, conditions 3.3 hold over the range 0 < m s 0.6. It is
obvious that the vertical arrangement of the piezo-active rods in the 13 structure
promotes the large anisotropy values of ,
e
from Equation 3.9. The piezoelectric
coefficients
*
3l
d
are determined using Equations 3.6 3.8 for the composite with
mm symmetry and it follows that:

*
31
d
=
*
32
d
=
*
31
e
(
E
s
*
11
+
E
s
*
12
) +
*
33
e
E
s
*
13
(3.11)

*
33
d
= 2
*
31
e
E
s
*
13
+
*
33
e
E
s
*
33
(3.12)
In Equation 3.11 both items connected with
*
3l
e
are negative at 0 < m s 0.6 for the
PZT-5-based composite. In the composite based on the PbTiO
3
-type ceramic a
competition between the items from Equation 3.11 (i.e., positive and negative
contributions into
*
31
d
) results in a non-monotonic dependence
*
31
d
(m), as shown in
Figure 3.4. In both of the composites studied here the dependence
*
33
d
(m) is
primarily caused by the second item in Equation 3.12. As follows from the data in
Figures 3.3 and 3.4, the piezoelectric anisotropy ,
d
from Equation 3.10 varies in a
more restricted range compared with ,
e
.
Figure 3.3. Volume-fraction dependences of the piezoelectric coefficients
*
3l
e
(in C / m
2
),
*
3l
d
(in pC / N),
*
3l
g
(in mV

m / N), and
*
3l
h
(in 10
8
the 13 PZT-5 ceramic / araldite composite: (a) l = 3 and (b) l = 1

a

b
Figure 3.4. Volume-fraction dependences of the piezoelectric coefficients
*
3l
e
(in C / m
2
),
*
3l
d
(in pC / N),
*
3l
g
(in mV

m / N), and
*
3l
h
(in 10
8
the 13 (Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
(I) ceramic / araldite composite: (a) l = 3 and (b) l = 1

a

b
As with the piezoelectric coefficients
*
3l
g
(m) of the 22 parallel-connected
composites (see Tables 3.1 and 3.2), non-monotonic behaviour of
*
31
g
(m) and
*
33
g
(m) in the 13 composites occurs at relatively small m values (see Figures 3.3
and 3.4), in accordance with Equation 3.3. In this volume-fraction range the
*
3l
d
(m) change relatively quickly in comparison with
o
c
*
33
(m), and this results in the considerable piezoelectric
sensitivity at 0 < m < 0.1.
3.2.2 Diagrams of Changes in Piezoelectric Properties Predicted for 13
Composites
The determination of the effective electromechanical properties of the 13
composite is usually carried out [10, 27, 29, 32, 33, 42, 47] in terms of two sets of
constants of components. An approach proposed in ref. [35] is intended to analyse
the effective properties using specific ratios of the component constants. This
approach enables the derivation of all the volume-fraction dependences of the
effective properties via a limited number of the basic constants of one of the
components. Hereafter, the elastic modulus
E
c
), 2 (
33
, piezoelectric coefficient
) 2 (
33
e
and
o
c
), 2 (
33
of the matrix (component 2) are chosen as the basic
constants. The volume-fraction dependences of the electromechanical properties
calculated using the EFM
1
[32] are given by
E
fg
c
*
(m) =
E
fg
c
, 0
(m)
E
c
), 2 (
33
,
*
ij
e
(m) =
0
ij
e
(m)
) 2 (
33
e
and
o
c
*
pp
(m) =
o
c
, 0
pp
(m)
o
c
), 2 (
33
(3.13)
In Equations 3.13
E
fg
c
, 0
(m),
0
ij
e
(m) and
o
c
, 0
pp
(m) are dimensionless coefficients that
depend on the volume fraction m of the rods. To diversify the volume-fraction
dependences that are calculated on the basis of formulae in ref. [32], we consider
the possibility of varying the properties of the components. This variation is
implemented on changing the ratios [35] between the electromechanical constants
of the ceramic rod and the polymer matrix,
r
c,fg
=
E
fg
c
), 1 (
/
E
fg
c
), 2 (
, r
e,ij
=
) 1 (
ij
e
/
) 2 (
ij
e
and r
c,pp
=
o
c
), 1 (
33
/
o
c
), 2 (
33
(3.14)
An effect of varying r
c,fg
and r
e,3j
from Equations 3.14 on the piezoelectric
properties and their anisotropy in the 13 composite with the piezo-active
polyvinylidene fluoride (PVDF) matrix is shown in Figures 3.5 and 3.6. The

____________
1
It should be noted that the electromechanical constants calculated using the formulae in
ref. [32] are almost equal to those calculated using the Eshelby tensors for the circular
cylinder [48] or to those calculated according to formulae in ref. [29]: the difference
between the related constants at m = const does not exceed 1% over the wide volume-
fraction range.

Figure 3.5. Graphic presentation of changes in volume-fraction dependences of the
normalised piezoelectric coefficients
0
3 j
e
that have been calculated for the 13 composite
with the PVDF matrix (reprinted from Topolov and Turik [35], with permission from the
American Institute of Physics)

Figure 3.6. Graphic presentation of changes in volume-fraction dependences of the
normalised piezoelectric coefficients
0
3 j
d
that have been calculated for the 13 composite
with the PVDF matrix (reprinted from Topolov and Turik [35], with permission from the
American Institute of Physics)

electromechanical properties of PVDF are taken from experimental data [10]. It is
assumed that the components have almost equal anisotropy of the
electromechanical properties and piezoelectric coefficients
) (
31
n
e
< 0 and
) (
33
n
e
> 0.
Figures 3.5 and 3.6 contain a system of graphs in cells related to fixed values of
ratios from Equations 3.14. In Figure 3.5 these ratios are r
e,3j
= 0.01, r
c,fg
= 0.01,
0.1, 1, 10, and 100 for cells 1, 2, 3, 4, and 5, respectively; r
e,3j
= 0.1, k
c,fg
= 0.01,
0.1, 1, 10, and 100 for cells 6, 7, 8, 9, and 10, respectively; r
e,3j
= 1, k
c,fg
> 0 for cell
11; r
e,3j
= 10, r
c,fg
= 0.01, 0.1, 1, 10, and 100 for cells 12, 13, 14, 15, and 16,
respectively. The piezoelectric coefficients
*
3 j
e
(m) of the 13 composite are
expressed in terms of the basic constants from Equations 3.13 as
*
3 j
e
(m)
=
0
3 j
e
(m)
) 2 (
33
e
.
In Figure 3.6 for cells 1 10, the same r
e,3j
and r
c,fg
values are assumed as those
related to cells 1 10 of Figure 3.5; r
e,3j
= 1, r
c,fg
= 0.01, 0.1, 1, 10, and 100 for
cells 11, 12, 13, 14, and 15, respectively; r
e,3j
= 10, r
c,fg
= 0.01, 0.1, 1, 10, and 100
for cells 16, 17, 18, 19, and 20, respectively. The piezoelectric coefficients
*
3 j
d
(m)
of the 13 composite are expressed in terms of the basic constants from Equations
3.13 as
*
3 j
d
(m) =
0
3 j
d
(m)
) 2 (
33
e
/
E
c
), 2 (
33
.
The graphs shown in cells of Figures 3.5 and 3.6 have been built for a series of
hypothetical 13 composites in which the electromechanical properties are varied
within a few orders of magnitude. Such large variations enable different
combinations of the components to be considered, for example, almost piezo-
passive or with high piezoelectric activity, elastically soft or hard, etc.
Changes in the piezoelectric coefficients
*
3 j
e
(m) show that weak min
0
33
e
(m) < 0
or max
0
33
e
(m) > 0 can be observed at r
c,fg
> 1 (see cells 4, 5, 9, 10, 15, and 16 in
Figure 3.5). The volume-fraction behaviour of the anisotropy value ,
e
(m) from
Equation 3.9 is concerned with a convexity of the
0
3 j
e
(m) curves. This behaviour
undergoes changes at r
c,fg
= 1 or r
e,3j
= 1 (see cells 2 4, 7 9, 13 15, or a
sequence of cells 7, 11, 15, etc., in Figure 3.5).
As follows from the graphs shown in Figure 3.6, the
0
3 j
d
(m) curves do not
undergo significant changes on increasing r
c,fg
or r
e,3j
. Contrary to ,
e
(m), changes in
the anisotropy value ,
d
(m) from Equation 3.10 are observed in a fairly narrow
range (Figure 3.6). Such behaviour is mainly associated with the relatively small
,
e
(0) and ,
d
(0) values of the piezo-active PVDF [10].
It is important to note that the changes in volume-fraction dependences
0
3 j
e
(m),
0
3 j
d
(m), ,
e
(m), and ,
d
(m) (Figures 3.5 and 3.6) do not depend on any r
c,pp
variations
related to the dielectric properties of the individual components. This independence
is caused by the 13 structure (see Figure 3.2) and boundary conditions for the
electric-field vector components in the composite poled along the OX
3
axis.
Points of weak max
0
33
e
(m) or min
0
33
e
(m) appear at r
c,fg
> 1, r
e,3j
= const and
almost equal values of the volume fraction m (Figure 3.5). Such extreme points are
closely connected with the elastic moduli
E n
fg
c
), (
and piezoelectric coefficients
) (
3
n
j
e
:
these electromechanical constants are characterised by the equal anisotropy in both
components (see Equations 3.12). Any changes in the elastic anisotropy of one of
the components at r
e,3j
= const result in a shifting of the extreme point of
0
33
e
(m) or
can even result in the lack of an extreme point in the range of 0 < m < 1 [35].
Changes in the volume-fraction behaviour of
0
33
e
(m) are associated with changes in
ratios from Equations 3.14. These latter changes result in a redistribution of the
mechanical field (see vertical directions in Figure 3.5) or the electric field (see
horizontal directions in Figure 3.5) in the 13 composite along the poling axis OX
3
.
The same redistribution has a small influence on the
0
31
e
(m) dependence: the
corresponding curves shown in Figure 3.5 do not undergo considerable changes at
r
c,fg
> 1. As follows from results in ref. [35], inequalities r
e,3j
< 1 and r
c,fg
> 1 are
necessary for reaching min
0
33
e
(m), and inequalities r
e,3j
> 1 and r
c,fg
> 1 are
necessary for reaching max
0
33
e
(m). An effect of the elastically hard rods on the
0
33
e
(m) dependence demonstrates the important role of the internal stress field in
forming the non-monotonic piezoelectric properties in the 13 composite. It is
noteworthy that max
*
33
e
(m) also takes place in the presence of the piezo-passive
polymer matrix. A similar maximum was analysed in studies on the 1| ceramic /
polymer composites with planar interfaces [38, 41] where | = 1, 2 or 3.
Thus, the approach developed in ref. [35] can be applied for prediction of the
effective electromechanical properties in 13 piezo-composites and the
interpretation of the non-monotonic volume-fraction dependence of different
effective parameters, especially when the properties in each component can be
varied in certain ranges. Particular cases for the application of this approach to the
two-component piezo-composites are considered in ref. [45].
3.2.3 Maxima of Squared Figures of Merit of 13 Composites
An interesting example of the volume-fraction behaviour of the effective
parameters is concerned with maxima of squared figures of merit
2 *
33
) (Q
and
2 *
) (
h
Q

in the 13 composites based on the Pb(Zr, Ti)O
3
-type ceramics. These parameters
are used [3, 20] to describe the sensor signal-to-noise ratio of the piezoelectric
element and to characterise its piezoelectric sensitivity. For the 13 composite
(Figure 3.2) poled along the OX
3
axis and being transversely isotropic, the squared
figures of merit are given by

2 *
33
) (Q
=
*
33
d
*
33
g
= (
*
33
d
)
2
/
o
c
*
33
(3.15)

2 *
) (
h
Q
=
*
h
d
*
h
g
= (
*
h
d
)
2
/
o
c
*
33
(3.16)
where
*
h
d
=
*
33
d
+ 2
*
31
d
and
*
h
g
=
*
33
g
+ 2
*
31
g
=
*
h
d
/
o
c
*
33
are the hydrostatic
piezoelectric coefficients [32, 34, 44]. It is well known [34, 43, 44] that values of
max [
*
33
Q
(m)]
2
and max[
*
h
Q
(m)]
2
in the 13 composite, based on the Pb(Zr, Ti)O
3
-
type ceramic, are several tens of times larger than the [
*
33
Q
(1)]
2
and [
*
h
Q
(1)]
2

values related to the monolithic ceramic.
As a rule, for various 13 ceramic / polymer composites, max [
*
33
Q
(m)]
2
and
max [
*
h
Q
(m)]
2
are attained at relatively small volume fractions of the Pb(Zr, Ti)O
3
type ceramic, i.e., at 0 < m < 0.1 [32, 34, 43]. In the same m range, the 13
composite is characterised by large anisotropy of the piezoelectric coefficients
*
3 j
e
,
i.e., | ,
e
| >> 1. Taking into account these facts, one can rewrite Equations (3.15)
and (3.16) as follows [44]:

2 *
) (
h
Q
~
*
,h elas
q
(
*
33
e
/
E
c
*
33
)
2
/
o
c
*
33
~
*
,h elas
q
[m
) 1 (
33
e
/ (m
E
c
), 1 (
33
+
E
c
), 2 (
33
)]
2
/ (m
o
c
), 1 (
33
)
and
2 *
33
) (Q
~ (
*
33 , elas
q
/
*
,h elas
q
)
2 *
) (
h
Q
(3.17)
where
*
,h elas
q
= {[1 + (
*
12
| )
1
2(
*
13
|
)
1
] / [1 + (
*
12
| )
1
2(
*
13
|
)
1
(
*
33
|
/
*
13
|
)]}
2
,
*
33 , elas
q
= {[1 + (
*
12
| )
1
] / [1 + (
*
12
| )
1
2(
*
13
|
)
1
(
*
33
|
/
*
13
|
)]}
2
and
*
ab
|
=
E
c
*
11
/
E
ab
c
*

depend on the elastic properties in the composite only, and an inequality
) 1 (
33
e
>>
) 2 (
33
e
holds. Both the squared figures of merit from Equations 3.17 depend on a ratio
= m
) 1 (
33
e
/ (m
E
c
), 1 (
33
+
E
c
), 2 (
33
) so that a higher
) 1 (
33
e
/
E
c
), 1 (
33
ratio leads to higher
2 *
) (
h
Q
and
2 *
33
) (Q
values for 0 < m < 0.1. It is worth noting that among the perovskite-type
ferroelectric ceramics listed in Table 2.1, PCR-7M has the highest
) 1 (
33
e
/
E
c
), 1 (
33
ratio
at room temperature. For example, max[
*
h
Q
(m)]
2
= 6240 (PPa)
1
and max[
*
33
Q
(m)]
2

= 6.2410
5
(PPa)
1
are attained in the 13 PCR-7M / elastomer composite [44]. The
max [
*
h
Q
(m)]
2
value is about 3 times larger than that of the typical 13 piezo-
composites [5]. Moreover, the same max[
*
h
Q
(m)]
2
value is approximately 1.5 times
larger than the (
*
h
Q
)
2
value typical of the 130 Pb(Zr, Ti)O
3
-based composite [5]
with a porous polymer matrix. These findings could promote new hydroacoustic,
medical and other piezotechnical applications of the studied 13 and related
composites.

3.3 Composites with 03 Connectivity, Manufacturing and
Applications
The 03 ferroelectric ceramic / polymer composite system is one of the most
common composite types used for sensor applications. The 03 composite (Figure
1.4) consists of the three-dimensionally connected polymer matrix reinforced by a
system of discrete (isolated) ceramic inclusions. The most common ceramic
components are perovskite-type compositions based on solid solutions of Pb(Zr,
Ti)O
3
(e.g., PZT, PZ, Navy, ZTS, PKI, APC, and PCR) [2, 3, 10, 20, 40, 49] or
Pb(Ca, Ti)O
3
[5052]. The polymer components can be either piezo-active or
piezo-passive. As the effective parameters of the 03 piezo-composites are highly
dependent [2, 3, 5053] on their microstructure, properties and volume fractions of
the components and poling conditions a large research effort has studied the
interrelations between microgeometry and the effective electromechanical
properties.
The 03 ceramic / polymer composite can be made in a variety of forms [3]
with a range of mechanical flexibility (e.g., piezorubber [2, 19, 20]). These
performances differ from those of the 22, 13 and 33 composites. Technological
aspects of processing and manufacture of 03 piezo-composites have been studied
[5358]. As with 13 composites, dielectrophoresis can also be used to
manufacture 03 composites. A variety of chain-like structures with 03
connectivity are formed in a colloidal suspension of the ceramic particles under the
electric field [58]. These structures contain prolate ceramic inclusions formed due
to the attraction of the particles and mutual dielectrophoresis. Some aspects related
to the poling procedure and choosing the filler materials for the 03 composites
have been discussed in ref. [3].
Various 03 composite films and cables are manufactured by hot-rolling [3, 54]
the ceramic particles into the rubber matrix. These highly effective materials find
applications as active elements for flexible hydrophones, keyboards, blood-
pressure cuffs and sensor applications [20].
3.3.1 Modelling of the Effective Properties in 03 Composites with Spheroidal
Inclusions
The electromechanical properties of the 03 piezo-composites can be predicted by
a variety of methods and averaging procedures [10, 5968]. The most popular
models for predicting the properties consider regularly shaped inclusions (for
example, sphere [63, 66, 67], spheroid [60, 67, 68] or parallelepiped [38, 59, 64])
or the hybrid inclusions [51, 52] that form the Banno unit cell [2, 54]. An approach
developed in monographs [10, 40] also enables the consideration of 03 piezo-
composites with an irregular arrangement of piezoelectric inclusions in the matrix.
However, most research deals with a regular distribution of inclusions with a
specific geometric shape, such as spheroid or parallelepiped.
Of practical interest is the averaging procedure [60, 62, 6871] that allows for
an electromechanical interaction between the piezo-active inclusions in the matrix
of the 03 composite. This procedure is based on the EFM, i.e., the known Mori
Tanaka method [72] generalised for heterogeneous piezoelectric media [69]. In this
method, an individual inclusion within the matrix, which is under an external
electroelastic field, is subjected to the action of an unknown average electroelastic
field. This average (or effective) field is a result of the external field applied to the
matrix (and, therefore, to the composite sample as a whole) and by the interaction
between the inclusions therein. Using the equivalent inclusion approach, the
coupled electroelastic behaviour of a 03 piezo-composite was modelled by Dunn
and Taya [73]. This modelling can be performed for the composite with aligned
ellipsoidal (spheroidal) piezo-active inclusions [73] that are isolated (not touching)
and characterised by a volume fraction 0 < m s 0.5. The EFM can also be applied
|
|
.
|
\
|
-
-
-
-
|| ||
|| e ||
|| ||e
|| ||c
t E
in the range of 0.5 s m < 1 [74], but model predictions are approximate only since
an error due to near-touching or touching of inclusions would increase as m 1.
In the model [68] of the 03 ceramic / polymer composite with spheroidal
inclusions (Figure 3.7), their shape is described by the equation
(x
1
/ a
1
)
2
+ (x
2
/ a
1
)
2
+ (x
3
/ a
3
)
2
= 1 (3.18)
relative to the axes of the rectangular co-ordinate system (X
1
X
2
X
3
). In Equation
3.18 the semi-axes of the spheroid are a
1
= a
2
and a
3
. The aspect ratio of the
spheroid is defined as = a
1
/ a
3
. The spheroidal ceramic inclusions are uniformly
aligned in the continuous polymer matrix, so that the remanent polarisation vector
of each inclusion is oriented along the OX
3
axis, which is the poling axis for the
sample as a whole. It is assumed that centres of symmetry of the inclusions occupy
sites of a simple tetragonal lattice with unit-cell vectors parallel to the OX
j
axes.
For improved poling the composite sample, the ceramic (n = 1) and polymer (n =
2) components obey an inequality
(2)

(1)
for electrical conductivities
(n)
. It is
also assumed that surface charges appearing owing to the ferroelectric polarisation
of the inclusions are to be fully screened by free charge carriers that would flow to
interfaces separating the components.
The effective electromechanical constants of the 03 composite are written in a
matrix form as

|| C*

||

=

(3.19)

where the superscript t denotes the transposition and the matrices || c*
E
||, || e* || and
|| *

|| contain the constants from Equations 2.21 and 2.22. Taking into account the
electromechanical interaction between the piezo-active inclusions,

the effective
properties in the 03 ceramic / polymer composite are determined by means of the
EFM [62, 6870]. The corresponding 9 9 matrix of the effective constants is
given by

Figure 3.7. Schematic of the 03 composite with spheroidal inclusions. The polarisation
vector P of the inclusion is denoted by an arrow. (X
1
X
2
X
3
) is the rectangular co-ordinate
system, a
i
are semi-axes of the spheroid, m is the inclusion volume fraction, and 1 m is the
matrix volume fraction
||

C* ||

=

||

C
(2)
||

+

m

(||

C
(1)
||

||

C
(2)
||) [||

I

||

+

(1

m) ||

S

||

||

C
(2)
||
1
(||

C
(1)
||

||

C
(2)
||)]
1
(3.20)

In Equation 3.20 the matrices of the electromechanical constants || C
(1)
||
(ceramic) and || C
(2)
|| (polymer) have a form similar to that shown in Equation
3.19, m is the volume fraction of the ceramic, || I || is the identity matrix, and || S ||
is the matrix that contains the Eshelby tensor components [48] depending on the
elements of || C
(2)
|| and the aspect ratio . The polymer medium surrounding the
ceramic inclusions can be either piezo-passive (isotropic) or piezo-active
(transversely isotropic). The 03 composite as a whole (see Figure 3.7) is described
by mm symmetry. Based on the effective electromechanical constants from
Equation 3.20 and taking into account Equations 2.122.15,

it is possible to
represent the piezoelectric coefficients in the matrix form as follows:
|| d* || = || e* |||| c*
E
||
1
, || g* || = || d* |||| *

||
1
and || h* || = || e* |||| *

||
1

(3.21)
In Equations 3.21 the piezoelectric coefficients || d* ||, || e* ||, || g* ||, and || h* || are
written as 3 6 matrices and determined as functions of m and .
The electromechanical constants of the 03 Pb(Zr, Ti)O
3
-type ceramic /
polymer composite that are calculated using Equations 3.20 and 3.21 are in full
agreement with the similar constants evaluated in ref. [60]. Despite the different
formulae for the Eshelby tensor components used in refs. [48, 68] and [60], the
difference between the constants remains less than 1% over a wide range of m and
. A small difference is observed in the limiting case of = 0, as the composite
shown in Figure 3.7 approaches that of a 13 geometry and the piezo-active
inclusion is elongated in the poling axis (see Figure 3.2) due to a
3
. The
corresponding electromechanical constants, evaluated for the 13 Pb(Zr, Ti)O
3
-
type ceramic / polymer composite by means of Equations 3.20 and 3.21 and the
Eshelby tensor [48] for the rod-shaped inclusion are more or less equal to constants
calculated using formulae from refs. [29], [32] or [40] over a wide range of m: the
difference between the related constants is less than 0.5%. Another limiting case of
(i.e., the piezo-active inclusion is elongated normal to the poling axis)
enables one to attain a good correlation between the electromechanical constants,
calculated using Equations 3.20 and 3.21 and the Eshelby tensor [48] for the
penny-shaped inclusions, and the constants evaluated using formulae from ref. [13]
for the series-connected 22 composite. For > 10
2
, a difference between the
values of the related constants becomes less than 1% over the range of 0 < m < 1.
Further examples of comparison of the volume-fraction dependences of the
effective parameters of the 03 composites are considered in Chapter 6.
3.3.2 Piezoelectric Properties in 03 Composites Based on PbTiO
3
-type
Ceramics
As is seen from Table 2.1, the PbTiO
3
-type ceramics have the piezoelectric
coefficients [10, 75, 76]

) 1 (
31
e
> 0,
) 1 (
33
e
> 0 and
) 1 (
15
e
> 0 (3.22)
and exhibit large anisotropy of the piezoelectric coefficients
) 1 (
3 j
d
and
electromechanical coupling factors
) 1 (
3 j
k
: across a wide temperature range,
inequalities
) 1 (
33
d
/ |
) 1 (
31
d
| >> 1 and
) 1 (
33
k
/ |
) 1 (
31
k
| >> 1 hold [77]. Important examples of
the large piezoelectric anisotropy were considered in work on the (Pb
1x
Me
x
)TiO
3

ceramics (Me = Ba, Ca or Sr) [7880]. In this section we analyse how the highly
anisotropic piezoelectric component influences the piezoelectric properties in the
03 ceramic / polymer composite with spheroidal inclusions. We will focus our
attention on two examples of the PbTiO
3
-based composites. In the first example,
the ceramics of the (Pb, A)(Ti, B)O
3
type are involved, where both the Pb and Ti
ions are substituted by specified ions A and B (or even B' and B) in order to vary
the electromechanical properties and, in particular, the
) 1 (
33
e
/
) 1 (
31
e
ratio (Table 2.1).
The second example is related to the (Pb
1x
Ca
x
)TiO
3
solid solutions (Table 3.3) that
are characterised by the non-monotonic behaviour of both the piezoelectric
anisotropy values,
) 1 (
33
e
/
) 1 (
31
e
and
) 1 (
33
d
/
) 1 (
31
d
, at 0.10 s x s 0.30. The effective
piezoelectric coefficients in the 03 composites are calculated using Equations
3.20 and 3.21 in specific ranges of m and . In the case of 0 < s 1, the volume
fraction m of the inclusions is varied from 0 to 0.5 at the upper bound m
u
= t / 6 for
the spherical inclusions ( = 1) in a primitive cubic cell.
The calculated volume-fraction dependences of four types of effective
piezoelectric coefficients are shown in Figure 3.8. The piezoelectric response of
the composite studied is characterised by the presence of min
*
31
d
(m, )|
= const
(see,
for example, curve 3 in Figure 3.8a) and the corresponding appearance of
min
*
31
g
(m, )|
= const
and max
*
33
g
(m, )|
= const
(curves 5 and 6 in Figures 3.8a and
b) where is taken from a range of aspect ratios 0 < <
*
. It can be seen that for
various combinations of the PbTiO
3
-type ceramics from Table 2.1 and piezo-
passive polymers, the
*
value belongs to a range [0.11; 0.16] and increases
slightly with an increase in the elastic moduli
) 2 (
ab
c
of polymers. The m
f
values that
satisfy conditions
*
31
g
(m
2
, ) =
*
, 31 m
g
and
*
33
g
(m
3
, ) =
*
, 33 m
g
are almost equal, and
an inequality | m
2
m
3
| 0.01 holds. The calculated
*
33
e
(m, ),
*
33
d
(m, ) and
*
3 j
g
(m, ) dependences (curves 2, 46 in Figure 3.8) are similar to those
determined [59, 60, 63, 67] for various 03 composites based on the perovskite
ceramics with sgn
) 1 (
31
e
= sgn
) 1 (
33
e
< 0.
The
*
, 33 m
g
values calculated for << 1 (i.e., elongated inclusions along the
poling axis) are highly dependent on the elastic moduli of the polymer component.
For example, at = 0.01 in the (Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
(II) / araldite
composite, a ratio
*
, 33 m
g
/
) 1 (
33
g
= 9.13 is observed, and at = 0.01 in the
(Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
(II) / polyethylene composite, the same ratio
increases to 24.6. Moreover, in these composites at 0 < <
*
, the piezoelectric
Table 3.3. Calculated values of elastic moduli
E
ab
c
), 1 (
(in 10
10
Pa),
a
) 1 (
ij
e
(in C / m
2
), relative dielectric permittivities
c
), 1 (
pp

/ c
0
, piezoelectric coefficient
) 1 (
33
g
(in
mV m / N), and piezoelectric anisotropy value
) 1 (
33
d

/
) 1 (
31
d
of the (Pb
1x
Ca
x
)TiO
3
ceramic
b

(reprinted from Glushanin et al. [68], with permission from Elsevier)
x
E
c
), 1 (
11
E
c
), 1 (
12
E
c
), 1 (
13
E
c
), 1 (
33
E
c
), 1 (
44
) 1 (
31
e
0.10 19.1 5.22 5.15 18.6 6.85 0.676
0.15 18.8 5.09 5.01 18.3 6.76 0.751
0.20 18.5 4.76 4.68 18.0 6.79 0.844
0.23 18.7 5.03 4.93 18.2 6.76 1.06
0.24 18.4 4.72 4.61 17.9 6.76 1.07
0.25 18.9 5.19 5.06 18.3 6.76 1.33
0.26 18.9 5.14 5.02 18.3 6.77 1.36
0.30 18.6 4.89 4.80 18.1 6.77 1.08
x
) 1 (
33
e
) 1 (
15
e

c
), 1 (
11
/ c
0

c
), 1 (
33
/ c
0
) 1 (
33
g
) 1 (
33
d
/
) 1 (
31
d
0.10 3.90 1.62 111 109 21.1 11.7
0.15 4.15 1.73 118 115 21.4 13.0
0.20 4.31 1.77 126 123 20.6 18.7
0.23 4.55 1.80 137 133 19.7 30.3
0.24 4.69 1.87 145 141 19.4 39.8
0.25 5.09 1.95 152 147 19.5 71.9
0.26 5.30 2.03 161 156 19.1 65.2
0.30 5.48 2.27 202 197 20.2 17.6
Notes to Table 3.3
a. For the poled ceramic with mm symmetry,
) 1 (
66
c
= (
E
c
), 1 (
11

E
c
), 1 (
12
) / 2 according to
formulae in refs. [10, 40]
b. The electromechanical constants have been calculated by the self-consistent (effective
medium) method [79, 80]. Spherical grains of the ceramic sample are assumed to be split
into 90 lamellar domains of two types with equal volume fractions and these domains are
separated by planar 90 domain walls. These domain walls are assumed to be practically
motionless so that a contribution from their displacements under external electric or
mechanical fields into the electromechanical constants of the ceramic approaches zero [79,
80]. The electromechanical constants of the single-domain crystal sample have been
evaluated by the method from ref. [79] on the basis of experimental room-temperature data
on the (Pb
1x
Ca
x
)TiO
3
solid solutions

anisotropy is characterised by values ,
e
>> 1 and ,
d
35. The distinctions
between the effective piezoelectric coefficients
*
3 j
e
and
*
3 j
d
demonstrate the
important role of the elastic properties of the polymer component in forming the
piezoelectric response of the 03 composite with highly prolate ceramic inclusions
( << 1) with the piezoelectric coefficients
) 1 (
3 j
e
that obey conditions 3.22.
In contrast to the above-considered PbTiO
3
-based composites, min
*
31
d
(m, ) at
<< 1 is not observed in the Pb(Zr, Ti)O
3
- or BaTiO
3
-based 03 composites [59,
60, 70]. The piezoelectric coefficient
*
31
d
of the 03 composites is expressed in
terms of the piezoelectric coefficients
*
3 j
e
and elastic compliances
E
ab
s
*
in
accordance with Equation 3.11. It is known that
*
31
d
< 0 is observed within a wide
range of m and values (see, for example, curve 3 in Figure 3.8), and as a
consequence, an inequality ,
e
> (
E
s
*
11
+
E
s
*
12
) /
E
s
*
13
is true for the composites with
*
3 j
e
> 0.
The analogous inequality is written using the elastic moduli
E
ab
c
*
of the same
composite in a form
,
e
>
E
c
*
33
/
E
c
*
13
(3.23)
Figure 3.8. Effective piezoelectric coefficients
*
3 j
e
(m, ) (in C / m
2
),
*
3 j
d
(m, ) (in pC / N),
*
3 j
g
(m, ) (in mV m / N), and
*
3 j
h
(m, ) (in 10
8
V / m) that have been calculated for the 03
(Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
(II) ceramic / araldite composite containing spheroidal
inclusions with = 0.01 (a), = 0.10 (b) or = 1 (c). Curve 1 corresponds to 10
2
*
31
e
, curve 2
corresponds to
*
33
e
(graphs a and b) or 10
2
*
33
e
(graph c), curve 3 corresponds to
*
31
d
(graphs
a and b) or 10
*
31
d
(graph c), curve 4 corresponds to 10
1
*
33
d
(graphs a and b) or
*
33
d
(graph
c), curve 5 corresponds to 10
1
*
31
g
(graphs a and b) or
*
31
g
10
2
*
33
g
(graphs a and b) or 10
1
*
33
g
*
31
h
, and curve 8
corresponds to 10
1
*
33
h
(graphs a and b) or
*
33
h
(graph c) (reprinted from Glushanin et al.
[68], with permission from Elsevier)


Figure 3.8 (continued)
Both the increase in the volume fraction m of the ceramic inclusions and an
increase in their aspect ratio result in the reduction in the ,
e
value for small
changes in the
E
c
*
33
/
E
c
*
13
ratio. A violation of condition 3.23 leads to a change in
the shape of the
*
31
d
curve on increasing (compare curves 3 in graphs ac of
Figure 3.8). A fulfilment of condition 3.23 promotes the non-monotonic volume-
fraction behaviour of
*
31
d
at << 1.
The transition from prolate spheroidal inclusions ( << 1) to spherical
inclusions ( = 1) leads to a significant redistribution of the internal electric and
mechanical fields in the composite sample. As a consequence, a significant
decrease in |
*
3 j
e
|, |
*
3 j
d
|, |
*
3 j
g
|, and |
*
3 j
h
| at m = const (compare graphs in Figures
3.8a and c) and ,
e
is observed in the 03 composites studied on increasing the
aspect ratio in the range 0 < < 1. In contrast, some increase in |,
d
| is observed
in the same range. This trend is the result of a relatively slow decrease in the
*
31
d
(m) dependence at = 1 (see curve 3 in Figure 3.8c).
The range 0 < < * is of practical interest due to the large ,
e
and |
*
3 j
g
| values,
where the
*
is taken from the aforementioned range [0.11; 0.16]. The studied 03
composites also have relatively large hydrostatic piezoelectric coefficient
*
h
g
. Its
local maximum can be represented as
*
,m h
g
=
*
h
g
(m
h
, ) at = const. The
*
,m h
g

values, calculated for the (Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
(II) / polymer composite
at 0.01 0.10, changes from 90.3 mV m / N to 52.6 mV m / N (the composite
with the araldite matrix) or from 114 mV m / N to 54.7 mV m / N (the composite
with the polyurethane matrix). The calculated
*
,m h
g
values are comparable with the
experimental
*
exp , h
g
values [3, 52] known for a variety of the 03 PbTiO
3
-type
ceramic / polymer composites (
*
exp , h
g
= (47100) mV m / N) and significantly
exceed an experimental value
*
exp , h
g
= 8 mV m / N [3] related to the 03 PZT /
polyurethane composite.
The next interesting example of the high piezoelectric sensitivity is related to
the 03 (Pb
1x
Ca
x
)TiO
3
ceramic / polyurethane composites. The (Pb
1x
Ca
x
)TiO
3

ceramic in the molar concentration range of 0.10 x 0.30 exhibits non-
monotonic dependences of both the piezoelectric anisotropy values
) 1 (
33
e
/
) 1 (
31
e
and
) 1 (
33
d
/
) 1 (
31
d
and an almost constant piezoelectric coefficient
) 1 (
33
g
(Table 3.3). The
slight dependence
) 1 (
33
g
on x (see Table 3.3) and the large
) 1 (
33
d
/ |
) 1 (
31
d
| values result
in a narrow range of values of the hydrostatic piezoelectric coefficient, i.e., 17.5
mV m / N
) 1 (
h
g
(x) 18.9 mV m / N. As a consequence, the local maxima
*
, 33 m
g
=
max
*
33
g
(m, , x)|
; x= const
=
*
33
g
(m
3
, , x) and
*
,m h
g
= max
*
h
g
(m, , x)|
; x= const
=
*
h
g
(m
h
, , x) undergo small changes at 0.10 x 0.30 and = const. It is obvious
that the presence of prolate ceramic inclusions in the 03 composite favours an
increase in the
*
33
g
and
*
h
g
values in comparison with the monolithic ceramic
component and both the
*
, 33 m
g
and
*
,m h
g
values remain large for fixed values of the
molar concentration x. The volume fraction m
3
that corresponds to the
*
, 33 m
g
values
is varied within a narrow range: the m
3
values are almost constant for = const and
0.02 m
3
0.12 for 0.01 0.20. The volume fraction m
h
related to the
*
,m h
g

values exhibits similar behaviour at the variation of the molar concentration x: m
h

const for = const and 0.02 m
h
0.19 for 0.01 0.20. The relatively low
volume fractions m
3
and m
h
of the ceramic inclusions lead to a small effective
o
c
*
33
and the relatively large effective piezoelectric
coefficient
*
33
d
. Both these factors, along with the small dielectric anisotropy of the
PbTiO
3
-type ceramics (i.e.,
o
c
), 1 (
11
/
o
c
), 1 (
33
s 1.3) [77, 80] favour the high piezoelectric
sensitivity of the 03 composite with the prolate ceramic inclusions. The lesser
*
,m h
g
/
) 1 (
h
g
values in comparison with the
*
, 33 m
g
/
) 1 (
33
g
values are a result of moderate
,
d
values.

As is known from the data on the (Pb
1x
Ca
x
)TiO
3
ceramic (Table 3.3), max(
) 1 (
33
d

/ |
) 1 (
31
d
|) is attained in the vicinity of x = 0.25. The volume-fraction dependences of
the normalised piezoelectric coefficient
*
33
g
/
) 1 (
33
g
of the 03 (Pb
0.75
Ca
0.25
)TiO
3

ceramic / polyurethane composite show that a significant increase in the
piezoelectric sensitivity is attained at << 1 and 0 < m s 0.20 (curves 13 in
Figure 3.9). Increasing and/or m results in a reduction of
*
33
g
values (see, for
example, curve 4 in Figure 3.9) due to a decrease of the
*
33
d
value (compare curves
4 in Figures 3.8b and c) or an increase of the
o
c
*
33
value as m increases.

The calculated
*
33
g
/
) 1 (
33
g
ratios (curves 14 in Figure 3.9) are comparable with
experimental values (sets 5 and 6 in Figure 3.9) for the related 03 PbTiO
3
-based
composites [51, 81]. The calculated curves 13 in Figure 3.9 and values of local
maximum
*
, 33 m
g
demonstrate the promising performance of the 03 composite with
highly prolate spheroidal inclusions made of the highly anisotropic ferroelectric
ceramic. In this case large
*
33
g
/
) 1 (
33
g
values are attained, even in the presence of
the non-piezo-active polymers, whereas the experimental
*
33
g
/
) 1 (
33
g
values (see sets
5 and 6 in Figure 3.9) correspond to the 03 ceramic / polymer composites with
both components being piezo-active. Thus, the large piezoelectric anisotropy
(
) 1 (
33
d
/|
) 1 (
31
d
| >> 1) of the ceramic component and the system of the aligned prolate
spheroidal ceramic inclusions favour the high piezoelectric sensitivity of the 03
composites studied.
Now we consider how the piezoelectric anisotropy
) 1 (
33
d
/
) 1 (
31
d
of the ceramic
component influences the normalised piezoelectric coefficient
*
33
d
/
) 1 (
33
d
in the 03
composites studied. Despite the large anisotropy of the piezoelectric coefficients
) 1 (
3 j
d
and positive signs of both the piezoelectric coefficients
) 1 (
3 j
e
(see inequalities
3.22) in the PbTiO
3
-type ceramics, the
*
33
d
(m) dependence calculated for the
Figure 3.9. Calculated (curves 1 4) and experimental (sets 5 and 6) volume-fraction
dependences of the
*
33
g
/
) 1 (
33
g
ratio of the 03 ceramic / polymer composites. Curves 1, 2, 3,
and 4 are plotted for the (Pb
0.75
Ca
0.25
)TiO
3
ceramic / polyurethane composite with = 0.01,
0.05, 0.10, and 0.15, respectively, experimental dependences 5 and 6 are related to the
PbTiO
3
ceramic / 70 / 30 mol.% copolymer of vinylidene fluoride and trifluoroethylene
composite [81] and (Pb,Ca)TiO
3
ceramic / 75 / 25 mol.% copolymer of vinylidene fluoride
and trifluoroethylene composite [51], respectively (reprinted from Glushanin et al. [68],
with permission from Elsevier)
related 03 composites is analogous to that determined [60] for other 03
composites based, for example, on the BaTiO
3
or Pb(Zr,Ti)O
3
ceramics. The
*
33
d
is written for transversely isotropic materials (mm
symmetry) in accordance with Equations 2.12 and 3.12 in the form [68]

*
33
d
=
*
31
e
E
s
*
33
[2(
E
s
*
13
/
E
s
*
33
) + ,
e
] (3.24)
In the 03 PbTiO
3
-based composites examined here,
*
31
e
(m, ) > 0 (see, for
example, curves 1 in Figures 3.8a c),
E
s
*
13
(m, ) < 0 and
E
s
*
33
(m, ) > 0 for 0.01 s
m s 0.50 and 0.01 s s 1. The piezoelectric anisotropy value ,
e
(m, )(
= const
> 0
changes from a few tens (at = 0.01) to a few units (at = 1). The
E
s
*
13
/
E
s
*
33
ratio
involved in Equation 3.24 reduces to
E
s
*
13
/
E
s
*
33
=
E
c
*
13
/ (
E
c
*
11
+
E
c
*
12
), i.e., we have
|
E
s
*
13
| /
E
s
*
33
< 1 at various m and values. The piezoelectric coefficient
*
33
d
(m, )
exhibits a monotonic increase [68] at = const due to the increase in
*
31
e
(m, ) > 0
and the large ,
e
(m, ) values.
Some examples of the volume-fraction dependence of the normalised
*
33
d
/
) 1 (
33
d
of the 03 ceramic / polymer composites are
shown in Figure 3.10. The largest
*
33
d
/
) 1 (
33
d
values (curves 1 and 2 in Figures
3.10a and c, and curve 5 in Figure 3.10b) are related to the PbTiO
3
-based
composites with highly prolate inclusions. The simple comparison of values of the
normalised piezoelectric coefficient
*
33
d
/
) 1 (
33
d
calculated for the PbTiO
3
- and
Pb(Zr, Ti)O
3
-based composites at fixed aspect ratios << 1 (see curves 1, 2, 5,
and 6 in Figure 3.10b) suggests that the lower
*
33
d
/
) 1 (
33
d
values are attained in the
03 composites based on the Pb(Zr, Ti)O
3
-type ceramics. Such a response is due to
the relatively small anisotropy of the piezoelectric coefficients
) 1 (
33
d
/
) 1 (
31
d
and the
Figure 3.10. Volume-fraction dependences of the normalised piezoelectric coefficient
*
33
d
/
) 1 (
33
d
of the 03 ceramic / polymer composites:
(a) curves 1, 2, 3, and 4 are calculated for the (Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
(II) / araldite
composite with = 0.01, 0.05, 0.10, and 0.15, respectively; experimental dependences 5 and
6 are related to the PbTiO
3
ceramic / 70/30 mol.% copolymer of vinylidene fluoride and
trifluoroethylene composite [81] and (Pb, Ca)TiO
3
ceramic / 75/25 mol.% copolymer of
vinylidene fluoride and trifluoroethylene composite [51], respectively,
(b) curves 1, 2, 3, and 4 are calculated for the (Pb
0.75
Ca
0.25
)TiO
3
ceramic / araldite composite
with = 0.01, 0.05, 0.10, and 0.15, respectively; curves 5 and 6 are calculated for the PZT
ceramic / polymer composite with = 0.01 and 0.10 [60], respectively, and
(c) curves 1, 2, 3, and 4 are calculated for the (Pb
0.75
Ca
0.25
)TiO
3
ceramic / polyurethane
composite with = 0.01, 0.05, 0.10, and 0.15, respectively (reprinted from Glushanin et al.
[68], with permission from Elsevier)
equality sgn
) 1 (
33
e
= sgn
) 1 (
31
e
in the Pb(Zr, Ti)O
3
-type ceramics [10, 40]. In addition,
the large piezoelectric anisotropy
) 1 (
33
d
/
) 1 (
31
d
of the PbTiO
3
-type ceramic inclusions
and the relatively soft polymer matrix promote the normalised piezoelectric
coefficient
*
33
d
/
) 1 (
33
d
~ 1 over a wide volume-fraction range (see curve 1 in Figure
3.10c), and this tendency is similar to that in the 13 ceramic / polymer composites
[32] independently of the signs of the piezoelectric coefficients
) 1 (
ij
e
of the ceramic
component.
In general, the presence of the highly prolate ceramic inclusions with the large
piezoelectric anisotropy
) 1 (
33
d
/
) 1 (
31
d
and the piezoelectric coefficients
) 1 (
ij
e
obeying
inequalities 3.22 lead to a simple combination of electromechanical constants that
would form the piezoelectric response of the 03 composite along the poling axis.
It is seen that advantages of the 03 PbTiO
3
-based composites are associated with
the significantly large
*
h
g
,
*
33
d
/
) 1 (
33
d
and
*
33
g
/
) 1 (
33
g
values and the possibility of
tailored piezoelectric anisotropy.
3.3.3 03 Composites with a Hierarchy of Inclusions
In this section we will consider a relationship between the microstructure and
electromechanical properties in the 03 composite based on the Pb(Zr, Ti)O
3
-type
ceramic. It is important to note that microgeometrical features of a distribution of
the ceramic inclusions within a polymer matrix can result in either pure 03 or
mixed connectivity [82, 83]. The microstructure of the composite with mixed
connectivity is often represented by a system of parallelepipeds (or cubes) [2, 50
52, 54, 82] in the matrix. This composite is often described in terms of 03, 13 or
33 connectivity patterns [38, 59] that influence the effective electromechanical
properties in a variety of aspects. Jayasundere et al. [83] proposed a finite-element
model suitable for the ceramic / polymer composite with 03 / 13 mixed
connectivity. This form of mixed connectivity is thought to be a representative
microstructure where the average size of the inclusion is comparable to the
thickness of the composite sample. On increasing the volume fraction of the
ceramic component, the 13 connectivity elements become the main component
responsible for the piezoelectric properties in the composite studied [83]. In this
case, predicted properties based on the 03 model of the composite with isolated
spherical inclusions are lower [56] than the experimental values of the
*
33
d
, dielectric permittivity
o
c
*
33
, etc.
A new approach to the study of the electromechanical properties in the ceramic
/ polymer composites is concerned with a hierarchy of the inclusions [84, 85] that
provide 03 connectivity over a wide volume-fraction range. To compare the
behaviour of the effective parameters of different composites, one can introduce
the normalised dielectric permittivity
|
c
=
o
c
*
33
(m
FC
) /
o
c
*
33
(1) =
o
c
*
33
(m
FC
) /
o
c
), 1 (
33
(3.25)
and the normalised piezoelectric coefficient
|
d
=
*
33
d
(m
FC
) /
*
33
d
(1) =
*
33
d
(m
FC
) /
) 1 (
33
d
(3.26)
where m
FC
is the volume fraction of the ceramic component. The experimental data
on |
c
and |
d
as a function of m
FC
for 03 ceramic / polymer composites based on
Pb(Zr, Ti)O
3
[55, 56] or Pb(Ca, Ti)O
3
[51, 52] ceramics show that changes in the
|
c
(m
FC
) dependence (Equation 3.25) are different, even in composites with a
volume fraction m
FC
that is varied over a wide range. There are also distinctions
between the experimental |
c
(m
FC
) dependences of these 03 composites [51, 52,
55, 56].
The main reasons for the differences between the |
c
(m
FC
) dependences are
associated [85] with different
o
c
*
33
(0) /
o
c
*
33
(1) ratios, where
o
c
*
33
(0) and
o
c
*
33
(1)
characterise dielectric properties of the polymer and ceramic component,
respectively. Features of microstructure, poling conditions and differences between
the electromechanical properties of the components are important in forming the
|
d
(m
FC
) dependence from Equation 3.26. In addition, evaluations made for |
c
and
|
d
within the framework of the model of the 03 composite [68] with spherical
ceramic inclusions demonstrate that good agreement between the calculated and
experimental [51, 52, 55, 56] values are attained only when m
FC
s 0.1. It seems
probable that microstructural elements of the composite are having a considerable
influence on the electromechanical properties at m
FC
> 0.1. Our further analysis of
this influence is carried out within the framework of the following model [84, 85].
A heterogeneous medium is represented by two types of layers that randomly
alternate along the poling axis OX
3
(Figure 3.11). The first layer, with the volume
fraction 1 m, comprises of spherical ceramic inclusions and a polymer matrix.
These inclusions have a relatively small volume fraction n
0
and are uniformly
distributed in the aforementioned layer. In the second layer, the inclusions in the
layer with the volume fraction m form a structure similar to a cylindrical rod
(Figure 3.11, insert). It should be indicated for comparison that, according to
scanning electron micrographs [86] of the fracture surfaces of the 03 PZT-based
composite with m
FC
= 0.49 and 0.67, the ceramic inclusions are distributed quasi-
uniformly. This implies that the ceramic component can be represented as a series
of chains, rod-like elements or other quasi-continuous structures and remain
distributed relatively uniformly in separate regions. In the same composite with
m
FC
= 0.67 [86], the volume fraction of the regions with a quasi-continuous
arrangement of the ceramic inclusions becomes dominant.
The presence of the rod-like elements oriented along the OX
3
axis favours
poling of the composite, particularly at high volume fractions m
FC
. It is assumed
[85] that the electromechanical constants of the rod-like elements are
approximately equal to the constants of the ceramic. This approximation may be
valid at low volume fractions of the polymer component within these elements and
with an almost continuous distribution of the ceramic inclusions along the OX
3
axis
in separate regions of the composite. In the model of the 03 composite (Figure
3.11), the rod-like elements are approximated by circular cylinders, and their
volume fraction n
1
in the above-described layers is varied from 0 to t / 4. The

Figure 3.11. Schematic diagram illustrating the 03 ceramic / polymer composite structure
with the hierarchy of inclusions. m is the volume fraction of layers comprising the rod-like
elements (see the insert on the left side) that are characterised by the volume fraction n
1
in
the separate layer. 1 m is the volume fraction of layers comprising spherical inclusions that
are characterised by the volume fraction n
0
in the separate layer. (X
1
X
2
X
3
) is the rectangular
co-ordinate system. The ceramic component is shown as a shaded region (reprinted from
Glushanin and Topolov [85], with permission from IOP Publishing)
upper bound of n
1
equals the ratio of the area of a circle, inscribed into a square
lying in the (X
1
OX
2
) plane, to the area of this square.
The volume fraction of the ceramic component in the 03 composite shown in
Figure 3.11 equals
m
FC
= n
1
m + n
0
(1 m) (3.27)
The presence of the three parameters m, n
0
and n
1
in Equation 3.27 suggests that
the effective electromechanical properties of the 03 composite can be determined
in three stages associated with procedures of averaging of these parameters. In the
first stage of averaging, the electromechanical properties in the layer containing the
spherical ceramic inclusions in the polymer matrix (03 connectivity) are
determined as a function of n
0
by means of the EFM (see Section 3.3.1).
In the second stage of averaging, the electromechanical properties in the layer
containing the rod-like ceramic inclusions in the polymer matrix (13 connectivity)
are determined as a function of n
1
by means of the EFM [32]. There is a small
difference between the effective constants of the 13 ceramic / polymer composite
with the finite elongated rods [87, 88] and the effective constants evaluated for
similar structures with the infinite rods [32]. The difference depends on the aspect
ratio o
r
, i.e., the height-to-width ratio of the separate ceramic rod. On increasing
the o
r
value, this difference decreases and becomes less than 10% for o
r
> 5, as
shown for effective electromechanical coupling factors k
t,eff
, k
31,eff
, k
33,eff
,
piezoelectric coefficients d
31,eff
, d
33,eff
and other parameters of the 13 PZT /
polymer composite [87] with a periodic structure. Elastic moduli
E
eff ab
c
,
,
eff ij
e
,
c
eff pp,
are calculated as
functions of the volume fraction of the ceramic rods [88] for the 13 PZT-5 /
polymer composites with the finite (o
r
= 5) or infinite (o
r
) cylindrical rods.
According to Berger et al. [88], the effective constants
E
eff ab
c
,
(ab = 13 and 33),
eff ij
e
,
(ij = 15, 31 and 33) and
c
eff pp,
(p = 1 and 3) of the 13 composite with o
r
= 5
show very good agreement with those evaluated for a similar composite with o
r

: in the range of n
1
= 0.10.6, differences between the similar parameters do not
exceed 5%. For the effective elastic moduli
E
eff ab
c
,
with ab = 11, 12 and 44, the
aforementioned differences attain about 10 15% at n
1
> 0.5 and less than 10% at
n
1
< 0.5.
The effective constants of the individual 03 and 13 layers are used in the
third stage of averaging [85] on the volume fraction m (22 series-connected
composite, see Figure 3.1a) according to formulae in ref. [10] that take into
account the electromechanical interaction between the piezo-active layers
distributed randomly along the poling direction.
2
As a result, the effective
electromechanical constants of the 03 composite studied are calculated on
dependence of n
0
, n
1
and m, and finally, the volume-fraction dependences from
Equations 3.25 and 3.26 are calculated over the wide range of m
FC
.
For these calculations, the full sets of electromechanical constants of the
ceramic and polymer components are required. To the best of our knowledge, the
full sets of such constants have not been published for the components of the 03
composites studied in refs. [55, 56]. To minimise possible discrepancies, it is
possible to use experimental constants of PZ 27 ceramic [89] and 75 / 25 mol.%
copolymer of vinylidene fluoride and trifluoroethylene [29] for calculations and
comparison with the experimental data [55]. As with PKI 502 ceramic [55], PZ 27
belongs to the Navy II ceramic group [89], so that
) 1 (
33
d
/
) 2 (
33
d
,
) 1 (
33
d
/
o
c
), 1 (
33
and
o
c
), 1 (
33
/
o
c
), 2 (
33
ratios calculated using data on the PZ 27 ceramic and the 75 / 25 mol.%
copolymer of vinylidene fluoride and trifluoroethylene are in good agreement with
the analogous ratios evaluated for the components of the 03 composite from ref.
[55].
The calculated volume-fraction dependences and experimental values of |
c
and
|
d
from Equations 3.25 and 3.26 are shown in Figure 3.12. The |
c
(m
FC
)
dependence undergoes small changes after replacing the piezo-active matrix by the
piezo-passive one (compare Figure 3.12a and Figure 3.12c). The presence of two
piezo-active components poled in the same direction results in a variation of the |
d

value in a wider range (Figure 3.12b) mainly due to the different signs of the
piezoelectric coefficients (
) 1 (
33
d
> 0 and
) 2 (
33
d
< 0 [29, 55]). In comparison with

____________
2
Formulae for the full set of the electromechanical constants of the 22 series-composite
with the random distribution of the piezo-active layers are collected in Appendix.
Figure 3.12. Volume-fraction dependences 10
2
|
c
(m
FC
) (graphs a and b) and 10
2
|
d
(m
FC
)
graphs c and d) that have been calculated for the 03 PZ 27 / 75/25 mol.% copolymer of
vinylidene fluoride and trifluoroethylene composite with both the components poled in the
OX
3
direction (graphs a and b) or with the piezo-passive polymer component (graphs c and
d). In all the graphs, n
0
= 0.10. Curves 1, 2, 3, 4, 5, and 6 are related to n
1
= 0.20, 0.30, 0.40,
0.50, 0.60, and 0.70, respectively. Dotted lines 7, 8, 9, 10, 11, and 12 are related to m

= 0.45,
0.55, 0.65, 0.75, 0.85, and 0.95, respectively. Experimental values [55] of the parameters of
the related composite are indicated by triangles (reprinted from Glushanin and Topolov [85],
with permission from IOP Publishing)
Figure 3.12b, the |
d
(m
FC
) values in Figure 3.12d vary within a narrower range.
A comparison of the calculated and experimental dependences shown in
Figures 3.12a and b demonstrates the emphasis of the role of the rod-like
inclusions (Figure 3.11) in forming the dielectric and piezoelectric responses of the
03 composite along the poling axis. Even at relatively low volume fractions of the
ceramic (m
FC
= 0.20.3), the volume fraction of these inclusions can be n
1
~ 0.3
0.4 (see curves 2 and 3 in Figures 3.12a and b) in the layers with the volume
fraction m

~ 0.5 0.7 (see dotted lines 810 in Figures 3.12a and b). It is obvious
that these inclusions provide the largest contribution to the volume fraction m
FC

(see Equation 3.27) owing to conditions n
1
m > n
0
(1 m) and n
0
= 0.1. Figures
3.12a and b show that small differences in the parameters (n
1
, m) related to the
experimental |
c
and |
d
values are attained at constant m
FC
values from a wide
volume-fraction range. This demonstrates a significant correlation between the
|
c
(m
FC
) and |
d
(m
FC
) dependences in the composite with the components poled in
the same direction. In the case of the composite with a piezo-passive matrix, this
correlation becomes less considerable (see, for example, a mutual arrangement of
curves 16 and experimental points in Figures 3.12c and d). This volume fraction
behaviour is concerned [85] with poling features and a reduction in
electromechanical coupling in the 03 composite.
Yamada et al. [90] attempted to explain the volume-fraction behaviour of some
of the electromechanical constants of a 03 PZT / PVDF composite within the
framework of a model with a spheroidal ceramic inclusion in the polymer matrix,
but without taking into account the electromechanical interaction among the
components. Despite this approximation, it was shown [90] that the ceramic
inclusion suitable for the prediction of the effective properties would be a prolate
spheroid with a
3
/ a
1
= 2.8 (see Figure 3.7 and Equation 3.18). In our model,
geometrical characteristics of the prolate ceramic inclusions in the layers (Figure
3.11) change as a result of variations in the parameters n
1
and m from Equation
3.27. Approaching m
FC
0.6 corresponds to m ~ 0.90 0.95 (see Figure 3.12),
i.e., the 03 structure shown in Figure 3.11 becomes very much like a 13 structure
shown in Figure 3.2. This is in good agreement with the experimental results [55].
Even at m
FC
~ 0.6, a continuous distribution of the ceramic component was not
observed [55, 86] in the thickness direction. The composite based on the Pb(Zr,
Ti)O
3
-type ceramic is described by 03 connectivity at 0 < m
FC
s 0.6. It should be
added that the model of the 03 composite (Figure 3.11) was also applied [85] to
interpret the experimental data [51, 52] on the composites based on the PbTiO
3
-
type ceramics.
Good agreement between the calculated [84, 85] and experimental [51, 52, 55]
parameters over the wide m
FC
range testifies to the correct consideration of the
hierarchy of the inclusions in the studied 03 ceramic / polymer composites and,
therefore, to the plausibility of the developed model. According to this model, in
the Pb(Zr, Ti)O
3
-based composite, the system of rod-like ceramic inclusions plays
an important role in forming the dielectric and piezoelectric properties at 0.20 s
m
FC
s 0.60 independently of the poled state of the polymer matrix. In the (Pb,
Ca)TiO
3
-based composite, the similar rod-like inclusions strongly influence the
piezoelectric properties at 0.20 s m
FC
s 0.65 [85]. It seems likely that a further
increase in the m
FC
value would result in a continuous distribution of the ceramic
component on the OX
3
direction (13 connectivity) and smooth changes in the
effective electromechanical properties.
3.3.4 03 Composites with Planar Microgeometry
Examples of a 03 connectivity pattern with planar interfaces are shown in Figures
1.2 and 3.13. A system of ceramic inclusions is regularly distributed in a
continuous polymer matrix and represents a 03 piezo-composite that can be
manufactured by means of rapid prototyping from the computer-aided design files
[6]. Such a composite is completely described by the cubic Banno unit cell (Figure
3.13) with a single inclusion. This inclusion is in the form of a rectangular
parallelepiped so that its length, width and height make up the tth, nth and hth parts
of the Banno unit cell edge, respectively (Figure 3.13).
The electromechanical constants of the ceramic and polymer components are
averaged by the matrix method that has also been applied to 22, 03, 11, and 13
connectivity patterns [22, 38, 59]. The averaging procedure for the 03 composite
implies [64] the determination of the effective electromechanical constants of a
laminar ceramicpolymer structure by taking into account the electrical and
mechanical boundary conditions at interfaces
(i) x
1
= const,
(ii) x
2
= const
and (iii) x
3
= const
separately. The ceramic (p = 1) and polymer (p = 2) components are characterised
by the elastic compliances
E p
ab
s
), (
, piezoelectric coefficients
) ( p
kl
d
and dielectric
permittivities
o
c
), ( p
ff
. The full set of effective elastic compliances
E
ab
s
*
, piezoelectric
coefficients
*
kl
d
o
c
*
ff
of the 03 composite is calculated
as a function of the parameters t, n and h that characterise microgeometry and the
volume fraction of the ceramic component m = tnh (Figure 3.13).
An interesting example of an influence of the geometric factor on the effective
parameters of the 03 composite is concerned with the variation of length and
width of the ceramic inclusion (i.e., 0 < t < 1 and 0 < n < 1, see Figure 3.13) at its
constant height (h
0
= const, 0 < h
0
< 1). This composite poled along the OX
3
axis is
described by mm2 symmetry, and its effective electromechanical properties obey
conditions
E
ab
s
*
(t, n, h
0
) =
E
ab
s
*
(t', n', h
0
),
*
kl
d
(t, n, h
0
) =
*
kl
d
(t', n', h
0
) and
o
c
*
ff
(t, n, h
0
)
Figure 3.13. Schematic representation of the Banno unit cell for the 03 composite with
planar interfaces. (X
1
X
2
X
3
) is the rectangular co-ordinate system
=
o
c
*
ff
(t', n', h
0
) where t' = n and n' = t. For example,
*
33
d
(0.1, 0.7, 0.9) =
*
33
d
(0.7,
0.1, 0.9). Our further consideration in this section is related to the 03 PbTiO
3
-type
ceramic / polymer composite.
By variation of n and t it is possible to establish maxima of the effective
parameters
*
h
g
=
*
33
g
+
*
32
g
+
*
31
g
,
2 *
33
) (Q =
*
33
d
*
33
g
and
2 *
) (
h
Q =
*
h
d
*
h
g
that
characterise piezoelectric sensitivity of the composite with the 03 particulate
microgeometry. The values of these maxima increase with increasing h
0
(i.e., the
height of the inclusion shown in Figure 3.13), and in the limiting case of h
0
= 1 the
composite has 13 connectivity, where the absolute maxima of
*
h
g
,
2 *
33
) (Q and
2 *
) (
h
Q are achieved. As follows from results in ref. [64], a difference between
values of the local max
*
h
g
(t, n, 0.9) and the absolute max
*
h
g
does not exceed 5%,
and any trend is therefore typical of the volume-fraction dependences of
2 *
33
) (Q
and
2 *
) (
h
Q . Examples of the dependences of the effective parameters on t and n at
h
0
= 0.95 are shown in Figure 3.14. The corresponding 03 composite displays
relatively high values of
*
h
g
~ 4
) 1 (
h
g
and
2 *
33
) (Q ~ 15(
) 1 (
33
Q )
2
in comparison with the
parameters of the ceramic component. The locations of max
*
h
g
and max
2 *
33
) (Q are
realised on conditions that
t << 1 and n 1 or n << 1 and t 1 (3.28)
Conditions 3.28 correspond to the use of plate-like or fibre-like ceramic inclusions
having the least edge lying along one of the non-polar (OX
1
or OX
2
) axes (see
Figures 3.13 and 3.14a). In contrast to the
*
h
g
and
2 *
33
) (Q values, the squared
hydrostatic figure of merit
2 *
) (
h
Q exhibits a monotonic dependence on both t and n
from ranges of 0 < (t, n) < 0.99 (see, e.g., Figure 3.14c). However, the largest
2 *
) (
h
Q value remains two orders of magnitude lower than
2 ) 1 (
) (
h
Q of the bulk
(Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
ceramic due to the presence of isolated ceramic
inclusions and a discontinuous distribution of the ceramic component in the
polymer matrix along the poling axis OX
3
(Figure 3.13). At such a discontinuous
distribution, the piezoelectric coefficients
*
33
d
and
*
h
d
remain relatively small in
comparison with the constants of the monolithic ceramic material and do not
provide large
2 *
) (
h
Q values. A considerable difference between the
2 *
33
) (Q and
2 *
) (
h
Q values satisfying conditions 3.28 is accounted for by the small volume
fraction m of the ceramic component. At low m values the piezoelectric anisotropy
) 1 (
33
d
/
) 1 (
31
d
of the ceramic is unable to influence a redistribution of internal electric
and mechanical fields and possibly promote an increase in the
2 *
) (
h
Q value of the
composite.
Thus, our study on the 03 piezo-composites enables us to determine volume-
fraction dependences of the effective electromechanical properties and related
Figure 3.14. Effective parameters
*
h
g
(a), (b) and
2 *
) (
h
Q (c) that have been calculated for
the 03 (Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
(II) / elastomer composite with inclusions shown in
Figure 3.13, where h = 0.95, 0.01 s t s 0.05 and 0.01 s n s 0.99 (reprinted from Bowen and
Topolov [64], with permission from Elsevier)
parameters and establish important extreme points thereof. Due to the lack of
connectivity of the piezo-active ceramic inclusions, the piezoelectric coefficients
*
33
d
and
*
h
d
are fairly low. At the same time, the reduced dielectric permittivity
o
c
*
33

promotes the enhanced piezoelectric coefficients
*
33
g
and
*
h
g
. Therefore, such 03
composites are suitable for sensing applications (e.g., hydrophones or sensor
elements), rather than active transducers. It is important to note that high
piezoelectric sensitivity depends not only on the microgeometry of the composite
(for example, prolate inclusions oriented along the poling axis), but also on the
volume fraction of the ceramic. The common feature of the 03 ceramic / polymer
composites studied in Sections 3.3.2 and 3.3.4 is that the relatively small volume
fraction of the ceramic component and the presence of the ceramic inclusions being
prolate with respect to the poling axis OX
3
can be regarded as a necessary
condition for high piezoelectric sensitivity.
3.4 Composites with 33 Connectivity, Increasing
the Piezoelectric Sensitivity and Applications
In the 33 composite both the components being continuously distributed in three
dimensions (Figure 1.2) form complicated interpenetrated structures. The presence
of the piezo-active framework offers possibilities to vary both the microgeometry
and the effective electromechanical properties in the composite over wide ranges.
In contrast to the 03 composite system, the continuous distribution of the piezo-
active component between electrodes provides favourable conditions for poling a
33 composite sample. The 33 ceramic / polymer composites are of interest due
to a series of effective parameters that exhibit a non-monotonic volume-fraction
behaviour. For example, max
*
h
d
is typical of the 33 composites based on the
Pb(Zr, Ti)O
3
-type ceramics [91], max
*
33
d
and max
*
t
k
were experimentally
established in the 33 (Pb, Ca)(Ti, Co, W)O
3
ceramic / epoxy composite [92],
max
*
33
g
and max [
2 *
) (
h
Q ] were found in the 33 PZT / epoxy composites [93] with
a certain arrangement of ceramic elements. In the 33 PZT / polymer composite
[94] with a three-dimensional lattice of ceramic rods distributed regularly in a
polymer matrix, max
*
33
d
and max [
2 *
) (
h
Q ] are also attainable.
Recently, a variety of methods and technologies have been applied to fabricate
the 33 ceramic / polymer composites. Skinner et al. [95] first fabricated these
composites by means of the lost-wax method and replamine process. A simpler
method developed in refs. [91, 96] is based on the so-called burned-out plastic
spheres (BURPS) technique. In this method, a ceramic powder, for example, of the
Pb(Zr, Ti)O
3
type, and plastic spheres are mixed in an organic binder. The sintering
procedure makes it possible to form a ceramic skeleton with a well-developed
structure of pores. The skeleton is later back-filled with a polymer, and as a result,
33 connectivity is observed. The corresponding ceramic / polymer composites are
manufactured to provide the volume fraction of the ceramic component up to 80%
[92] and performance can be better than that of the related 03 and 13
composites. Other methods for manufacturing the 33 composites include relic
processing [97, 98], distorted reticulated ceramics [99], rapid prototyping [6], and
robotic deposition technique (robocasting) [94]. The rapid prototyping and
robocasting methods enable one to attain a fine-scale periodic structure with 33
connectivity in a wide volume-fraction range.
The 33 ceramic / polymer composites are traditionally characterised by
relatively high piezoelectric sensitivity, low density and relatively low dielectric
permittivity [9194]. A next stage in the development of the 33 structure can be
concerned with creation of a piezo-composite that would provide extremely high
piezoelectric sensitivity at a relatively large volume fraction m of the ceramic
component, i.e., a piezo-active material for which conditions [64]

*
h
g
(m) /
*
h
g
(1) >> 1, [
*
33
Q
(m)]
2
/ [
*
33
Q
(1)]
2
>> 1, [
*
h
Q
(m)]
2
/ [
*
h
Q
(1)]
2
>> 1,
and m ~ 1 m (3.29)
would be fulfilled simultaneously.
It is obvious that an increase in piezoelectric sensitivity of the composite is
closely connected with increasing the connectivity of its piezo-active component:
this tendency is clearly observed on comparing the effective parameters of the 13
and 03 composites (see Sections 3.2.1 and 3.3.2). In the 33 composite system the
connectivity of both the components is maximum. This high degree of connectivity
favours high piezoelectric activity at relatively low dielectric permittivity, and as a
consequence, such a combination of the dielectric and piezoelectric properties can
result in a fulfilment of conditions 3.29.
A typical Banno unit cell for the 33 composite is shown in Figure 3.15a,
where the interpenetrating components are indicated. It is assumed that the
components are distributed regularly and the composite is poled along the OX
3
axis
of the rectangular co-ordinate system (X
1
X
2
X
3
). As the dielectric permittivity
o
c
), 1 (
33
of the piezo-ceramic is significantly greater than that of the polymer, it will
dominate the effective permittivity
o
c
*
33
of the composite. As follows from analysis
of Figure 3.15b, the piezoelectric column parallel to the poling direction OX
3

contributes to
o
c
*
33
of the composite (volume 1). The other volumes shown in
Figure 3.15b contribute little to
o
c
*
33
since they consist of either the low-permittivity
polymer (volume 2) or the piezo-active component being in series with a polymer
in the OX
3
direction (volumes 3 and 4). The effective dielectric permittivity of the
studied composite can therefore be approximated [100] by a formula

o
c
*
33
=
o
c
), 1 (
33
m
(c)
(3.30)
where m
(c)
= L
2
/ (L + l)
2
is the volume fraction of the piezo-active component
contributing to
o
c
*
33
.
To calculate the effective piezoelectric coefficient
*
33
d
, the fraction of a force
experienced by the piezo-active component must be determined. If a force F
3
is
applied on the OX
3
direction, the total force is distributed through four volumes
indicated in Figure 3.15b. The polymer component will experience less force,
being of significantly higher compliance than the ceramic and has led to the
Figure 3.15. Banno unit cell and distribution of forces in the 33 ceramic / polymer
composite: (a) model of the interpenetrating structures of two components (piezoelectric and
polymer phases) with volume fractions depending on the piezoelectric column width L and
length l; (b) force F
3
in the OX
3
direction is distributed through volumes 1 4; (c) force F
1

in the OX
1
direction is distributed through volumes 5 8 (reprinted from Bowen and
Topolov [64], with permission from Elsevier)
assumption in some models of complete stress transfer into the piezo-ceramic
[100]. However, the actual distribution of the force between the four volumes
depends on the respective compliances of the volumes and the cross-sectional area
of each volume. For volume 1, the elastic compliance is
) 1 (
33
V
s
=
E
s
), 1 (
33
and the area
of an intersection by the (X
1
OX
2
) plane is a
(V1)
= L
2
. For volume 2, the compliance
is
) 2 (
33
V
s
=
) 2 (
33
s
and the area of the similar intersection is a
(V2)
= l
2
. Volumes 3 and 4
are identical in geometry and consist of the polymer and piezo-active components
connected in series in the OX
3
direction. For both these volumes, the area of the
intersection by the (X
1
OX
2
) plane is a
(V3)
= a
(V4)
= Ll and the elastic compliance
connected with the OX
3
direction can be calculated using a series model by means
of a formula [64]

) 3 (
33
V
s

=
) 4 (
33
V
s
= m
(V3)
E
s
), 1 (
33
+ (1 m
(V3)
)
) 2 (
33
s
= (L
E
s
), 1 (
33
+ l
) 2 (
33
s
) / (L + l) (3.31)
where m
(V3)
is the volume fraction of the piezo-active component in volume 3. At
high or intermediate volume fractions of the ceramic component, only volume 1
can be considered to contribute to the effective piezoelectric coefficient
*
33
d
of the
33 composite. Volume 2 consists only of the piezo-passive polymer while the
piezo-active ceramics in volumes 3 and 4 are in series with the polymer in the OX
3

direction, which results in a negligible contribution to the
*
33
d
value.
The effective piezoelectric coefficient
*
33
d
of the composite shown in Figure
3.15 can therefore be calculated [100, 101]

from the following formula:

*
33
d
=
) 1 (
33
d
fraction of load experienced by volume 1
=
) 1 (
33
d
(L
2
/
E
s
), 1 (
33
)[(L
2
/
E
s
), 1 (
33
) + (l
2
/
) 2 (
33
s
) + 2 L l (L + l) / (L
E
s
), 1 (
33
+ l
) 2 (
33
s
)]
1
(3.32)

By varying the L and l values, the piezoelectric coefficient
*
33
d
can be
calculated as a function of the volume fraction of the piezo-ceramic and the aspect
ratio of the unit cell. The same procedure is undertaken to calculate the
*
31
d
of the composite [100]. For this calculation, one can
consider a force F
1
acting in the OX
1
direction: as seen from Figure 3.15c, this
force is distributed through volumes 58. The volume of the piezo-ceramic that
contributes to the
*
31
d
value is the cube in volume 6, as indicated in Figure 3.15c
(the remainder of volume 6 is neglected since it is in series in the OX
3
direction
with the polymer component), and the piezo-ceramic in volume 7. Based on these
statements,
*
31
d
can be calculated using a formula [64]

*
31
d
=
) 1 (
31
d
[L /(L + l)]
-1
fraction of load experienced by volume 6 and 7
=
) 1 (
31
d
[L / (L + l)]{[L l (L + l) / (L
E
s
), 1 (
11
+ l
) 2 (
11
s
) ] + (L
2
/
E
s
), 1 (
11
)]{(l
2
/
E
s
), 1 (
11
) + [2L l (L
+ l) / (L
E
s
), 1 (
11
+ l
) 2 (
11
s
)] + (L
2
/
E
s
), 1 (
11
)}
1
(3.33)

Thus, the volume-fraction dependences of the effective hydrostatic piezoelectric
coefficients
*
h
d
and
*
h
g
of the 33 ceramic / polymer composite can be determined
using Equations 3.303.33.
Calculations were made for 33 composites consisting of a piezo-ceramic and a
polymer being ten times more compliant than the ceramic (i.e.,
) 2 (
ab
s
= 10
E
ab
s
), 1 (
) [64].
The most important results on the dielectric and piezoelectric responses of the
studied 33 composites are graphically represented in Figure 3.16. As would be
expected, there is a gradual reduction in
o
c
*
33
as the low-permittivity polymer
replaces the ceramic (Figure 3.16a). Figure 3.16b shows that the hydrostatic
*
h
d
reaches a maximum at polymer volume fractions about
0.40.5 for the PCR-63, PCR-62 and Pb(Zr
1x
Ti
x
)O
3
ceramics. The highest
*
h
d
values are achieved in the composites based on ceramics with the highest
) 1 (
33
d

values (e.g., PCR-63 among the ceramic compositions chosen). It should be
mentioned for comparison that in the 13 composites, the maximum in
*
h
d
is
generally achieved at higher polymer volume fractions of 0.70.8 [43]. The
mechanism by which
*
h
d
increases for the 33 composites is similar to that for 1
3s [29, 32]: in both cases the composite structure leads to decoupling of
mechanical stress and a large reduction in the |
*
31
d
| value relative to
*
33
d
. It is for
this reason that the ceramic components that experience the greatest enhancement
of
*
h
d
relative to
) 1 (
h
d
are those with a small anisotropy value
) 1 (
33
d
/|
) 1 (
31
d
| (i.e., with a
relatively high |
) 1 (
31
d
| value), such as PCR-62 and PCR-63 (see Figure 3.16b and
Table 2.1). For the composite based on the modified PbTiO
3
ceramic,
3
no increase
in
*
h
d
is observed (Figure 3.16b) as the ceramic component inherently exhibits a
high
) 1 (
33
d
/ |
) 1 (
31
d
| ratio due to its relatively high
) 1 (
33
d
(51 pC / N), low
) 1 (
31
d

(4.4
pC / N)

and initial high
) 1 (
h
d
values. Incorporating polymer into this modified
PbTiO
3
ceramic merely serves to gradually decrease
) 1 (
33
d
and the associated the
) 1 (
h
d
value (Figure 3.16b). For a particular ceramic / polymer composite system,
the reduction in
*
31
d
relative to
*
33
d
(and as a consequence, the resulting increase in
*
h
d
) is greater for 33 composites than 13 composites. This is due to the relatively
high stiffness of the ceramic along the OX
1
and OX
2
axes of the 33 composite
(Figure 3.15a), which acts to stiffen the structure in these directions and reduce
|
*
31
d
|. It is for this reason that top and bottom metallic cover plates are used in 13
composites, where stiffening the structure in the OX
1
and OX
2
directions really
results in the increase in
*
h
d
[43].
The hydrostatic piezoelectric coefficient
*
h
g
increases for all the composites
____________
3
In this section the modified PbTiO
3
ceramic means the composition of
(Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
(II). The full set of electromechanical constants of this
ceramic is given in Table 2.1.
0
200
400
600
800
1000
1200
0 0.2 0.4 0.6 0.8 1
Polymer volume fraction
R
e
l
a
t
i
v
e

p
e
r
m
i
t
t
i
v
t
y

c
3
3
T
PCR-63
PCR-62
x=0.4
x=0.3
x=0.2
x=0.1
Mod PbTiO3

a
0
10
20
30
40
50
60
70
80
90
100
0 0.2 0.4 0.6 0.8 1
d
h

(
i
n

1
0
-
1
2

m

V
-
1
)
PCR-63
PCR-62
x=0.4
x=0.3
x=0.2
x=0.1
Mod PbTiO3

b
Figure 3.16. Effective relative dielectric permittivity
o
c
*
33
/
0
c
(a), hydrostatic piezoelectric
coefficient
*
h
d
(b), hydrostatic piezoelectric coefficient
*
h
g
(c), and squared hydrostatic
figure of merit (
*
h
Q
)
2
(d) that have been calculated for 33 ceramic / polymer composites
with elements of microgeometry shown in Figure 3.15a. Ferroelectric ceramics used as
piezo-active components include modified PbTiO
3
((Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
(II)),
PCR-62, PCR-63, and Pb(Zr
1x
Ti
x
)O
3
ceramics where x = 0.1, 0.2, 0.3, and 0.4 (reprinted
from Bowen and Topolov [64], with permission from Elsevier)

0
50
100
150
200
250
300
350
0 0.2 0.4 0.6 0.8 1
g
h

(
i
n

1
0
-
3
V

m
-
1

P
a
-
1
)
PCR-63
PCR-62
x=0.4
x=0.3
x=0.2
x=0.1
Mod. PbTiO3

c

d
(Figure 3.16c) as polymer is incorporated into the 33-type ceramic skeleton and
the ceramic volume fraction decreases. This is generally a consequence of the
increase in
*
h
d
(Figure 3.16b) and decrease in
o
c
*
33
(Figure 3.16a) of the 33
composite. The highest hydrostatic piezoelectric coefficient
*
h
g
is observed for the
composite based on the modified PbTiO
3
ceramic, due to its relatively low
dielectric permittivity (
o
c
), 1 (
33
/
0
c
= 170) and relatively high piezoelectric coefficient
(
) 1 (
33
d
= 51 pC / N). The squared hydrostatic figure of merit
2 *
) (
h
Q reaches a
maximum at polymer volume fractions of approximately 0.8 and, unlike 03
ceramic / polymer composites, large ratios of
2 *
) (
h
Q /
2 ) 1 (
) (
h
Q are attained (Figure
3.16d). As an example, a 33 PCR-63 / polymer composite exhibits max[
2 *
) (
h
Q ] =
3.410
12
Pa
1
(Figure 3.16d), while the ceramic component is characterised by a
value of
2 ) 1 (
) (
h
Q = 0.0410
12
Pa
1
. It should be added that a similar max[
2 *
) (
h
Q ] is
observed for the 13 ceramic / polymer composites, where the maximum is located
at slightly higher polymer volume fractions of 0.9 (see Section 3.2.3 and refs. [32,
43, 44]).

The 33 connectivity leads to relatively large
*
h
d
,
*
h
g
and
2 *
) (
h
Q values and
represents various examples of fulfilment of conditions 3.29. Such performance of
the 33 composites studied in this section promotes both signal-generation and
signal-sensing applications assuming that the composite has an optimum
architecture. The optimum architecture in terms of choice of the ceramic volume
fraction and choice of the ceramic composition depends on the application and the
appropriate figure of merit that needs to be maximised.
3.5 Connectivity Non-monotonic Behaviour Manufacturing
In this chapter we have analysed and summarised results on non-monotonic
volume-fraction dependences of effective electromechanical properties and related
parameters in a variety of o| ceramic / polymer composites. As follows from a
series of review papers [26, 20, 31, 47, 51, 102] written in the last decades, these
composites are of interest for both research and applications. We have applied
different methods to evaluate the effective parameters of these composites and to
find extreme points of the effective parameters. Along with extreme points,
appropriate methods for manufacturing the o| composites have been mentioned
in Table 3.4. Undoubtedly, the data listed in Table 3.4 cannot be restricted by the
connectivity patterns considered in Chapter 3. It is believed that the available list of
connectivity indices and examples of the non-monotonic behaviour of the effective
parameters will be renewed in the future due to advances in manufacturing,
experimental studies and promising applications of the o| composites.

Table 3.4. Non-monotonic effective parameters and methods for manufacturing the o|
ceramic / polymer composites
o| Examples of extreme points of effective
parameters
Methods for
manufacturing the
composites
22
min
E
c
*
11
[9], max
*
31
e
[9, 10, 40], max
o
c
*
11
[9],
max
*
t
k
[7,8,13], max[(
*
33
Q
)
2
] [11] (series
connection) or max
*
t
k
[13], max
*
h
d
[12],
max
*
h
g
[12] (parallel connection)
Dice and fill [4,5],
injection moulding [4,5,
14], tape lamination [4,5,
14], jet machining [4,5,
16], co-extrusion [4,5],
rapid prototyping [6]
13
min
*
31
d
[37,39], max
*
33
d
[38], max
*
33
e [35,
38,40,41], max
*
33
g
[33,42], max
*
33
h
[33],
max
*
h
d
[29,32,34,40], max
*
h
g
[32,34,43],
max
*
t
k
[29,40], max[(
*
33
Q
)
2
] [44],
max[(
*
h
Q
)
2
] [30, 34, 43, 44]
Rod placement (pick and
place) [4,5], dice and fill
[4,5], injection moulding
[4,5,14], co-extrusion [4,
5], tape lamination [4,5,
14], rapid prototyping [6],
lost mould [4],
dielectrophoresis [4, 46],
robocasting [94]
03
min
*
31
d
[68], max
*
33
g
[68], min
*
31
g
[68],
max
*
h
g
[64,68], max[(
*
33
Q
)
2
] [64]
Dielectrophoresis [58],
pressing mixture of
ceramic and polymer
[102], colloid processing
method [102], rapid
prototyping [6], mixing
and hot rolling [3,54]
33
max
*
33
d
[92], max
*
33
g
[93], max
*
h
d
[64,91],
max
*
t
k
[92], max[(
*
h
Q
)
2
] [64,93]
Lost-wax method and
replamine process [95],
BURPS [91,96], relic
procession [97,98],
distorted reticulated
ceramics [99], rapid
prototyping [6],
robocasting [94]
11

max
*
31
d
or min
*
31
d
[22], min
*
32
d
or max
*
32
d

[22], max
*
31
e
or min
*
31
e
[22], min
*
32
e
or
max
*
32
e
[22], min
*
33
e
[22] or max
*
33
e
[41],
min
*
31
g
[22], min
*
32
g
[22], max
*
33
g
[22],
max
*
31
k
[22], min
*
32
k
[22], max[(
*
33
Q
)
2
]
[22], max[(
*
h
Q
)
2
] [22]
Laminating the layers
[103]
References
[2] Banno H, (1983) Recent development of piezoelectric ceramic products and
composite of synthetic rubber and piezoelectric ceramic particles. Ferroelectrics 50:3
12
[3] Gururaja TR, Safari A, Newnham RE, Cross LE, (1988) Piezoelectric ceramic-
polymer composites for transducer applications. In: Levinson M (ed.) Electronic
Ceramics: Properties, Devices, and Applications. Marcel Dekker, New York Basel,
pp 92128
[4] Janas VF, Safari A, (1995) Overview of fine-scale piezoelectric ceramic/polymer
composite processing. Journal of the American Ceramic Society 78:29452955
[7] Huebner W, Reidmeyer MR, Stevenson JW, Busse L, (1995) Fabrication of 22
connectivity PZT / thermoplastic composites for high efficiency linear arrays. In:
Pandey RK, Liu M, Safari A (eds.) ISAF94: Proceedings of the Ninth IEEE
International Symposium on Applications of Ferroelectrics. University Park,
Pennsylvania, USA, August 710, 1994. IEEE, Piscataway, pp 206209
[8] Zhang QM, Geng X, (1994) Dynamic modeling of piezoceramic polymer composite
with 22 connectivity. Journal of Applied Physics 76:60146016
[9] Getman IP, Ryabov AP, Ustinov YuA, (1987) On possibilities of the averaging
method in a task on propagation of waves in an electroelastic layer with a periodic
thickness inhomogeneity. Izvestiya Akademii Nauk SSSR. Mekhanika Tverdogo Tela
(3):118124 (in Russian)
[11] Cao W, Zhang QM, Cross LE, (1993) Theoretical study on the static performance of
piezoelectric ceramicpolymer composites with 22 connectivity. IEEE Transactions
on Ultrasonics, Ferroelectrics, and Frequency Control 40:103109
[12] Grekov AA,

Kramarov SO,

Kuprienko AA, (1987) Anomalous behavior

of

the

two-
phase lamellar piezoelectric texture. Ferroelectrics 76:4348
[13] Hashimoto KY, Yamaguchi M, (1986) Elastic, piezoelectric and dielectric properties
of composite materials. In: Proceedings of IEEE Ultrasonic Symposium,
Williamsburg, Va, November 1719, 1986. V. 2. New York, pp 697702
[14] Bowen LJ, French KW, (1992) Fabrication of piezoelectric ceramic / polymer
composites by injection molding. In: Liu M, Safari A, Kingon A, Haertling G (eds.)
ISAF92: Proceedings of Eighth IEEE International Symposium on the Applications
of Ferroelectrics. Greenville, SC, USA, August 30 September 2, 1992. IEEE,
Piscataway, pp 160163
[15] Stevenson JW, Reidmeyer MR, Huebner W, (1994) Fabrication and characterization
of PZT / thermoplastic polymer composites for high-frequency phased linear arrays.
Journal of the American Ceramic Society 77:24812484
[16] Lubitz K, Wolff A, Preu G, (1992) New piezoelectric composites for ultrasonic
transducers. Ferroelectrics 133:2126
[17] Turcu S, Jadidian B, Danfort SC, Safari A, (2002) Piezoelectric properties of novel
oriented ceramic polymer composites with 22 and 33 connectivity. Journal of
Electroceramics 9:165171
[18] McNultry T, Stanefield DJ, Danfort SC, Safari A, (1999). Dispersion of lead zirconate
titanate for fused deposition of ceramics. Journal of the American Ceramic Society
82:1757 1760
[19] Fernandez JF, Dogan A, Zhang QM, Tressler JF, Newnham RE, (1996). Hollow
piezoelectric composites. Sensors and Actuators A 51:183192
[20] Newnham RE, (1997) Molecular mechanisms in smart materials. Materials Research
Society Bulletin 22(5):2034

4473
[24] Turik AV, Radchenko GS, (2002) MaxwellWagner relaxation in piezoactive media.
[25] Turik AV, Radchenko GS, (2003) MaxwellWagner relaxation of elastic constants of
layered polar dielectrics. Physics of the Solid State 45:10601064
[26] Radchenko GS, Turik AV, (2003) Giant piezoelectric effect in layered ferroelectric -
polymer composites. Physics of the Solid State 45:17591762
[27] Smith WA, (1993) Modeling 13 composite piezoelectrics: hydrostatic response.
IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control 40:4149
[28] Hossack JA, Bedi RL, (1994) Design of composite piezoelectric transducers. Key
Engineering Materials 9293:301322
[29] Taunaumang H, Guy IL, Chan HLW, (1994) Electromechanical properties of 13
piezoelectric ceramic / piezoelectric polymer composites. Journal of Applied Physics
76:484489
[30] Klicker KA, Biggers JV, Newnham RE, (1981) Composites of PZT and epoxy for
hydrostatic transducer applications. Journal of the American Ceramic Society 64:59
[31] Smith WA, Shaulov AA, (1988) Composite piezoelectrics: basic research to a
practical device. Ferroelectrics 87:309320
[34] Li L, Sottos NR, (1995) Improving hydrostatic performance of 13 piezocomposites.
Journal of Applied Physics 77:45954603
[35] Topolov VYu, Turik AV, (1999) Nonmonotonic concentration dependence of
piezoelectric coefficients of 13 composites. Journal of Applied Physics 85:372379
[36] Kar-Gupta R, Venkatesh TA, (2005) Electromechanical response of 13 piezoelectric
composites: effect of poling characteristics. Journal of Applied Physics 98:054102
14 p
[37] Shaulov AA, Smith WA, Ting RY, (1989) Modified-lead-titanate / polymer
composites for hydrophone applications. Ferroelectrics 93:177182
[41] Topolov VYu, Glushanin SV (2002) Evolution of connectivity patterns and links
[42] Pardo L, Mendiola J, Alemany C, (1988) Theoretical treatment of ferroelectric
composites using Monte Carlo calculations. Journal of Applied Physics 64:50925097
[43] Bennett J, Hayward G, (1997) Design of 13 piezocomposite hydrophones using
finite element analysis. IEEE Transactions on Ultrasonics, Ferroelectrics, and
[44] Topolov VYu, Turik AV, (2001) Porous piezoelectric composites with extremely high
reception parameters. Technical Physics 46:10931100
[45] Topolov VYu, Turik AV, (1999) A comparative analysis of electromechanical
properties and their anisotropy in two-component composites with different
connectivities. Ferroelectric 222:131136
[46] Randall CA, Miller DV, Adair JH, Bhalla AS, (1993) Processing of electroceramic
polymer composites using the electrorheological effect. Journal of Materials Research
8:899904
[47] Wersing W, (1986) Composite piezoelectrics for ultrasonic transducers. In: ISAF86:
Proceedings of the Sixth International Symposium on Applications of Ferroelectrics.
Lehigh University, Bethlehem, PA, USA, 811 June, 1986. IEEE, Piscataway, pp
212223
[49] Haertling G, (1999) Ferroelectric ceramics: history and technology. Journal of the
[50] Garner GM, Shorrocks NM, Whatmore RW, Goosey MT, Seth P, Ainger FW, (1989)
03 piezoelectric composites for large area hydrophones. Ferroelectrics 93:169176
[51] Dias CJ, Das-Gupta DK, (1994) Electroactive polymer-ceramic composites. In:
Proceedings of the 4th International Conference on Properties and Applications of
Dielectric Materials. July 38, 1994, Brisbane, Australia. IEEE, Piscataway, pp 175
178
[52] Wenger MP, Das-Gupta DK, (1999) Mixed connectivity composite material
characterization for electroactive sensors. Polymer Engineering and Science 39:1176
1188
[53] Chiang CK, Popielarz R, (2002) Polymer composites with high dielectric constant.
[54] Banno H, Saito S, (1983) Piezoelectric and dielectric properties of composites of
synthetic rubber and PbTiO
3
or PZT. Japanese Journal of Applied Physics 22 (Suppl.
2):6769
[55] Ng KL, Chan HLW, Choy CL, (2000) Piezoelectric and pyroelectric properties of
PZT/P(VDF-TrFE) composites with constituent phases poled in parallel or antiparallel
directions. IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
47:13081315
[56] Venkatragavaraj E, Satish B, Vinod PR, Vijaya MS, (2001) Piezoelectric properties of
ferroelectric PZTpolymer composites. Journal of Physics D: Applied Physics
34:487492
[57] Chan HLW, Ng PKL, Choy CL, (1999) Effect of poling procedure on the properties
of lead zirconate titanate / vinylidene fluoride-trifluoroethylene composites. Applied
Physics Letters 74:30293031
[58] Wilson SA, Maistros GM, Whatmore RW, (2005) Structure modification of 03
piezoelectric ceramic / polymer composites through dielectrophoresis. Journal of
[59] Levassort F, Lethiecq M, Millar C, Pourcelot L, (1998) Modeling of highly loaded 0
3 piezoelectric composites using a matrix method. IEEE Transactions on Ultrasonics,
[61] Nan C-W, (1994) Effective-medium theory of piezoelectric composites. Journal of
[62] Huang JH, Kuo W-S, (1996) Micromechanics determination of the effective
properties of piezoelectric composites containing spatially oriented short fibers. Acta
Materialia 44:48894898
[64] Bowen CR, Topolov VYu, (2003) Piezoelectric sensitivity of PbTiO
3
-based ceramic /
polymer composites with 03 and 33 connectivity. Acta Materialia 51:49654976
[65] Nan CW, Weng GJ, (2000) Influence of polarization orientation on the effective
properties of piezoelectric composites. Journal of Applied Physics 88:416423
[66] Wong CK, Poon YM, Shin FG, (2003) Explicit formulas for effective piezoelectric
coefficients of ferroelectric 03 composites based on effective medium theory.
[67] Poizat C, Sester M, (2001) Homognisation priodique de composites
pizolectriques 03: influence de la distribution. Revue des Composites et des
Matriaux Avancs 11:6574
[68] Glushanin SV, Topolov VYu, Krivoruchko AV, (2006) Features of piezoelectric
properties of 03 PbTiO
3
-type ceramic / polymer composites. Materials Chemistry
and Physics 97:357364
30:161175
[71] Dunn ML, Wienecke HA, (1997) Inclusions and inhomogeneities in transversely
isotropic piezoelectric solids. International Journal of Solids and Structures 34:3571
3582
[72] Mori T, Tanaka K, (1973) Average stress in matrix and average elastic energy of
materials with misfitting inclusions. Acta Metallurgica 21:571574

Royal Society (London), Pt
A 443:265287

3
ceramics
modified PbTiO
3
[77] Turik AV, Topolov VYu, (1997) Ferroelectric ceramics with a large piezoelectric
anisotropy. Journal of Physics D: Applied Physics 30:15411549
[78] Ichinose N,

Fuse Y,

Yamada Y, Sato R, (1989)

Piezoelectric anisotropy in the
modified PbTiO
3
ceramics. Japanese Journal of Applied Physics 28 (Suppl. 2): 87 90
[79] Topolov VYu, Turik AV, Chernobabov AI, (1994) On the mechanism of high
piezoelectric anisotropy in lead titanate-based ferroelectrics. Crystallography Reports
39:805809
[80] Topolov VYu, Turik AV, Chernobabov AI, (1994) On the piezoelectric anisotropy in
modified PbTiO
3
ceramics. Ferroelectrics 154:271276
[81] Ngoma JB, Cavaille JY, Paletto J, Perez J, (1990) Dielectric and piezoelectric
properties of copolymer-ferroelectric composite. Ferroelectrics 109:205210
[82] Banno H, (1995)

Theoretical equations for dielectric, elastic and piezoelectric
constants of diphasic composite changing its connectivity from 30 to 03 via 33. In:

Pandey RK, Liu M, Safari A (eds.) ISAF94:

Proceedings of the Ninth IEEE
International Symposium on Applications of Ferroelectrics, University Park,

PA,

USA, August 710, 1994.

IEEE, Piscataway, pp

186189
[83] Jayasundere N, Smith BV, Dunn JR, (1994) Piezoelectric constant for binary
piezoelectric 03 connectivity composites and the effect of mixed connectivity.
[84] Glushanin SV, Topolov VYu, (2005) Features of the electromechanical properties of
03 composites of the Pb(Zr, Ti)O
3
-based ferroelectric ceramics polymer type.
Technical Physics Letters 31:346348
[85] Glushanin SV, Topolov VYu, (2005) A hierarchy of inclusions and electromechanical
properties of 03 ceramic / polymer composites. Journal of Physics D: Applied
Physics 38:24602467
[86] Chan HLW, Chen Y, Choy CL, (1996) Thermal hysteresis in the permittivity and
polarization of lead zirconate titanate / vinylidenefloride-trifluoroethylene 03
composites. IEEE Transactions on Dielectrics and Electrical Insulation 3:800805
[87] Hirata Y, Numazawa T, Takada H, (1997) Effects of aspect ratio of lead zirconate
titanate on 13 piezoelectric composite properties. Japanese Journal of Applied
Physics. Pt 1 36:60626064
[89] Alguer M, Alemany C, Pardo L, Gonzlez AM, (2004) Method for obtaining the full
set of linear electric, mechanical and electromechanical coefficients and all related
losses of a piezoelectric ceramic. Journal of the American Ceramic Society 87:209
215
[90] Yamada T, Ueda T, Kitayama T, (1982) Piezoelectricity of a high-content lead
zirconate titanate / polymer composite. Journal of Applied Physics 53:43284332
[91] Rittenmyer K, Shrout T, Schulze WA, Newnham RE, (1982) Piezoelectric 33
[92] Yoon KH, Yoo JH, Kim WS, Kim ES, (1996) Piezoelectric properties of (Pb, Ca)(Ti,
Co, W)O
3
/ epoxy composites. Ferroelectrics 186:169172
[93] Jadidian B, Allahverdi M, Mohammadi F, Safari A, (2002) Processing of piezoelectric
fiber / polymer composites with 33 connectivity. Journal of Electroceramics 8:209
214
[94] Smay JE, Cesarano III J, Tuttle AB, Lewis JA, (2002) Piezoelectric properties of 3X
periodic Pb(Zr
x
Ti
1-x
)O
3
polymer composites. Journal of Applied Physics 92:6119
6127
[95] Skinner DP, Newnham RE, Cross LE, (1978) Flexible composite transducers.
Materials Research Bulletin 13:599607
[96] Shrout TR, Schluze WA, Biggers JV, (1979) Simplified fabrication of PZT / polymer
composites. Materials Research Bulletin 14:15531559
[97] Waller DJ, Safari A, (1992) Piezoelectric lead zirconate titanate ceramic fiber /
polymer composites. Journal of the American Ceramic Society 75:16481655
[98] Ting SM, Janas VF, Safari A, (1996) Relic processing of fine-scale, large-area
piezoelectric ceramic fiber / polymer composites. Journal of the American Ceramic
Society 79:16891692
[99] Creedon MJ, Schulze WA, (1994) Axially distorted 33 piezoelectric composites for
hydrophone applications. Ferroelectrics 153:333339
[100] Bowen CR, Perry A, Kara H, Mahon SW, (2001) Analytical modelling of 33
piezoelectric composites. Journal of the European Ceramic Society 21:14631467
[101] Bowen C R, Perry A, Stevens R, Mahon S, (2001) Analytical and numerical
modelling of 33 piezoelectric composites. Integrated Ferroelectrics 32:333342
[102] Safari A, (1994), Development of piezoelectric composites for transducers. Journal de
Physique III (France) 4:11291149
[103] Nagahara H, Sato T, Adachi A, Amemiya K, Shioya K, Igaki E, Tanahashi M, (2002)
Method of manufacturing a piezoelectric element. Patent 6,538,363 (USA), publ.
25.03.03

4
Piezoelectric Response of Porous Ceramic and
Composite Materials Based on Pb(Zr, Ti)O
3
4.1 Porous Materials, General Characteristic and Manufacturing
This chapter will describe the study of porous materials based on ferroelectric
ceramics of the Pb(Zr, Ti)O
3
type. The ceramic compositions are chosen, as a rule,
near the MPB to optimise the electromechanical properties, and many examples of
the electromechanical constants of the appropriate Pb(Zr, Ti)O
3
-type ceramics are
provided in Table 2.1. Porous materials, whether they are a composite or a ceramic
(Figures 1.6 and 2.3a), or contains closed pores (air inclusions, voids) or open
pores (air channels), exhibit electromechanical properties that depend on the
volume fraction and microgeometry of pores [15]. As shown by Banno [1],
Wersing et al. [2, 3], Khoroshun et al. [6], and other authors, the electromechanical
properties of porous ferroelectric ceramics can be described on the basis of
averaging procedures similar to those suitable for the piezo-active composites.
Following this principle, we will consider piezoelectric performance and features
of the porous ceramics and composites as two groups of related heterogeneous
materials.
The piezoelectric and other physical properties of a porous materials depend on
the properties of the piezo-active component, porosity m
p
(i.e., the volume fraction
of the pores), the shape and size of the pores and their orientation. Results of the
large body of work [14, 615] on porous Pb(Zr, Ti)O
3
-type ceramics show that
these materials are characterised by reduced values of the piezoelectric coefficients
, and | |, elastic moduli and , dielectric permittivities and
, electromechanical coupling factors , | | and | |. Although there is a
reduction in the above-mentioned properties, there is also a simultaneous increase
in the piezoelectric coefficients , | |, , and , figures of merit and
in porous ceramics [1012, 14, 15], as compared to the monolithic ceramic
material. It is also known that the piezoelectric coefficient and
electromechanical coupling factor remain almost constant [1012] over wide
*
33
e
*
33
h
*
31
d
E
ab
c
* D
ab
c
*
V
H
*
ff
[
H
*
ff
*
33
k
*
31
k
*
p
k
*
33
g
*
31
g
*
h
g
*
h
d
*
33
Q
*
h
Q
*
33
d
*
t
k
m
p
ranges. It should be mentioned that the Pb(Zr, Ti)O
3
-type ceramics with the
perovskite-type structure represent a vast group of ferroelectric ceramic materials
that have the piezoelectric coefficients [16]
< 0, > 0 and > 0 (4.1)
) (
31
FC
e
) (
33
FC
e
) (
15
FC
e
(see Table 2.1).
1
Inequalities 4.1 differ from similar conditions 3.22 written for the
PbTiO
3
-type ceramics. As a result, in contrast to the PbTiO
3
-type ceramics, the
Pb(Zr, Ti)O
3
-type materials do not exhibit the large anisotropy of the piezoelectric
coefficients [16].
) (
3
FC
j
d
A variety of methods for manufacturing porous materials based on ferroelectric
ceramics have been proposed and developed in the last decades. These methods are
similar to those concerned with the 33 composites (Section 3.4). For example, to
manufacture porous ceramics, Skinner et al. [17] used a coral replamine technique,
Creedon et al. [18] used polymer foams that were coated with ceramic particles. In
refs. [3, 10, 14, 1926] a ceramic powder and plastic pellets (spheres) were mixed
with an organic binder and compacted. The compacted elements were subjected to
heating and a burnout cycle (BURPS technique). A polymeric sponge technique
with a subsequent burnout cycle was also applied [14, 22] to manufacture porous
ceramics and composites with various microstructures. As noted by Kara et al.
[14], the BURPS and polymeric sponge techniques are relatively easy to
implement for a large-scale production. After the burnout cycle Hayashi et al. [23]
carried out hot-isostatic pressing of the sintered samples without encapsulating.
The porous PZT ceramic manufactured in this way was characterised by 33
connectivity and a homogeneous microstructure with a bimodal distribution of
small and large pores in the ceramic matrix [23]. The authors of refs. [2426]
modified the BURPS method and studied the role of the pore-forming agent.
Various patterns of porosity in the PZT-based samples were attained using agents
such as polyvinyl chloride, polymethyl methacrylate, polyethylene oxide [24],
polyethylene glycol, saccharose, and ammonium oxalate [26]. The performance of
porous ceramics was studied [2326] in connection with their microgeometry, pore
size and other factors.
The aerogel process [27] and tape-casting methods [28] have also been
employed to manufacture porous ceramics. Tape-casting technology combined
with a fugitive ink was applied by Kahn et al. [9] to form arrays of ordered plate-
like air pores in a ceramic matrix. A pattern with a regular arrangement of pores
was formed on an ink-deposition screen that was used for subsequent deposition of
the ink onto a green ceramic tape. The tapes were subsequently stacked and
laminated, and organic components were finally burned out to form the porous
medium. A porous ceramic with 31 connectivity was manufactured by Bast and
Wersing [29] using the multilayer method, i.e., a combination of photolithography
with stacking, pressing and firing the ceramic sheets. Lamellar pores oriented
____________
1
In Chapter 4 we use superscript (FC) to denote the electromechanical constants and other
parameters of the monolithic ferroelectric ceramic.
Piezoelectric Response of Porous Ceramic and Composite Materials 103
parallel or perpendicular to the poling axis were formed by Lopatin et al. [30] in
ceramic samples.
It is obvious that each method enables a distinctive microstructure, size and
arrangement of pores to be obtained. Their volume fraction m
p
can be varied in
specific ranges. For example, in the samples prepared using the BURPS technique,
m
p
is usually varied between 0.30 and 0.70, the foam or sponge technique enables
m
p
= 0.850.90 to be obtained [14, 21, 22]. Various porous ceramics and
composites with m
p
= 0.090.92 were manufactured using the BURPS and sponge
techniques and studied in refs. [14, 21]. It should be noted that an increase in the
m
p
value results in various fractions of closed and open porosity [21] and results in
changes in the connectivity of porous samples. However, experimental works
contain restricted data on features of microstructure of the porous ceramics and
composites (see, e.g., [14, 18, 29, 31]). These factors make it impossible to
experimentally analyse the behaviour of the effective electromechanical properties
of the porous ceramics and / or composites in the complete m
p
range for a fixed
pore size and morphology.
4.2 Volume-fraction Dependences of Parameters of Porous
Piezo-active Materials
Experimental dependences of a series of the effective parameters )
*
()
*
= ,
, , , , , etc.) of the porous ceramic on its porosity m
p
in a range 0
< m
p
d 0.66 were first generalised in ref. [10]. These dependences were measured
on porous samples made of the Pb(Zr, Ti)O
3
-type ceramics such as PCR-1 (with
the large value), ZTSSt-2 (with large value) and ZTStBS-2 (with the
moderate and values). The porous ceramic materials were
characterised by 33 connectivity. Based on numerous experimental data, Lopatin
and Lupeiko [10] approximated the )
*
(m
p
) dependence by means of a formula
*
3 j
d
*
3 j
g
*
33
e
*
33
h
V
H
*
33
*
3 j
k
) (
33
FC
g
) (
33
FC
d
) (
33
FC
g
) (
33
FC
d
)
*
(m
p
) / )
(FC)
= A (m
p
2
+ B)
-1
+ Cm
p
+ G r H (4.2)
In Equation 4.2 coefficients A, B, C, G, and H were calculated for different
parameters of the porous ceramics by the least-square method and )
(FC)
= )
*
(0).
Examples of the )
*
(m
p
) / )
(FC)
ratios are given below for the cases of d()
*
/ )
(FC)
)
/ dm
p
< 0 ()
*
= , , , and ), d()
*
/ )
(FC)
) / dm
p
| 0 ()
*
= ) and
d()
*
/ )
(FC)
) / dm
p
> 0 ()
*
= and ) [10]:
*
31
d
*
33
e
*
33
h
V
H
*
33
*
33
d
*
31
g
*
33
g
/ = 10.239 (m
p
2
+ 2.0)
1
2.429 m
p
+ 6.128 r 0.045,
*
31
d
) (
31
FC
d
/ = 0.206 (m
p
2
+ 1.0)
1
+ 0.047 m
p
+ 1.202 r 0.044,
*
33
d
) (
33
FC
d
/ = 2.319 (m
p
2
1.0)
1
1.260 m
p
1.270 r 0.011,
*
31
g
) (
31
FC
g
/ = 5.316 (m
p
2
0.95)
1
+ 0.625 m
p
4.531 r 0.010,
*
33
g
) (
33
FC
g
/ = 15.373 (m
p
2
+ 2.5)
1
2.737 m
p
+ 7.129 r 0.020,
*
33
e
) (
33
FC
e
/ = 0.421 (m
p
2
+ 0.4)
1
1.746 m
p
+ 2.055 r 0.039, and
*
33
h
) (
33
FC
h
/ = 81.195 (m
p
2
+ 10)
1
1.731 m
p
+ 9.108 r 0.023.
V
H
*
33
V
H
), (
33
FC
In refs. [25, 1113] devoted to porous piezo-ceramics, a variety of approaches
were proposed to predict the dependences of the effective electromechanical
properties and related parameters on porosity m
p
. The study by Wersing et al. [3] is
based on a comparative analysis of models and formulae for calculating these
dependences in the wide m
p
range. The models involved with the analysis and
porous ferroelectrics include Wagners model, Marutakes model and the unit-cell
models related to 30 or 33 connectivity [2, 3]. It is shown that Bruggemans
formulae (first applied to isotropic heterogeneous media) [32] can be used to
describe the isotropic and quasi-isotropic properties of a porous ceramic. Such a
description is correct in a case when there are only small anisotropies of the
dielectric and elastic properties, and the good agreement between calculated and
experimental results [3] demonstrates the correctness of the approach proposed by
Bruggeman [32].
Based on the modified cube model, Banno collected formulae [1] for the
effective electromechanical constants of a ceramic with closed pores (30
connectivity). The calculated data obtained using these formulae are in agreement
with various experimental dependences of the effective parameters of porous
ceramics on m
p
in a range of 0 < m
p
< 0.2 [1]. Simultaneously, the theoretical
treatment within the framework of the Banno unit cell with planar interfaces [1, 12]
cannot examine the dependence of the electromechanical properties on the size and
distribution of pores within the ceramic medium.
In contrast to the work of refs. [1, 12], Dunn and Taya [4, 11] and Wu [13]
developed an approach to determine the full set of effective electromechanical
constants of a transversely isotropic piezo-active material containing spherical
pores. The procedure is similar to that described in Section 3.3.1 where the
effective properties of the 03 ceramic / polymer composites are considered. The
methodology is based on the EFM, i.e., the Mori Tanaka method applied to a
heterogeneous material with a piezo-active matrix. As is shown in ref. [11], this
method accounts for the electromechanical interaction at small and moderate m
p
values (0 < m
p
d 0.5). It should be added for comparison that the touching of
identical sized spheres would be at volume fractions m
p
> S / 6, i.e., the upper
value 0.5 is near the lower bound related to sphere touching. Numerical results
illustrate an effect of the shape of air pores on the effective electromechanical
properties in the porous ceramics. An agreement between the experimental and
calculated values of a series of parameters is attained [11] for the porous ceramics
of the Pb(Zr, Ti)O
3
type with various microstructures. Problems of modelling of
the electromechanical properties in the porous ceramics were also discussed in refs.
[15, 26, 3339].
Levassort et al. [26] considered representative unit cells for 33 and 30
connectivity patterns as well as a 33 unit cell with a 30 matrix. Based on the
unit-cell concepts, Levassort et al. analysed different cases of the open- and closed-
ceramics. Comparing the calculated and experimental data, it was shown that
elements of both 30 and 33 connectivities would be important at m
p
t 0.25.
Results on the finite-element modelling of the effective electromechanical
properties in the porous PZT ceramic are discussed in refs. [3739]. Kar-Gupta and
Venkatesh [37] put forward a classification of porous ceramics. According to this
classification, these ceramics are characterised by 30, 33 or 31 connectivity.
The porous ceramics with 31 connectivity can be regarded as either
longitudinally porous or transversely porous materials, depending on a mutual
arrangement of the porosity axis and the poling direction. A further systematisation
of data on the effective properties is discussed in ref. [37]. Features of
microgeometry and symmetry of the porous 31 material are considered in ref.
[38]. It is shown that the full set of electromechanical constants in such a material
can comprise as many as 45 independent values, and the number of values depends
on the symmetry of the monolithic material, the porosity orientation and
distribution therein [38]. Asai et al. [39] presented and discussed results on the
finite-element modelling of the properties in the porous PZT ceramics with
complex three-dimensional microstructures. An approach developed in ref. [39] is
concerned with an iterative solver that is suitable for the prediction of effective
electromechanical properties in a wide m
p
range.
Important examples of the influence microgeometry on porous PZT ceramics
and related composites are studied in refs. [4042]. It is of interest to note that the
authors of refs. [4042] used optimal design methods and then constructed and
manufactured novel porous materials intended for hydrophone and other
piezotechnical applications. As follows from a comparison, good agreement
between the predicted and measured parameters [40] is attained. However, a
discrepancy between the predicted and measured parameters [42] is accounted for
by an insufficient polarisation caused by a field distribution (or non-uniform
electric field) during poling.
Examples of experimental dependences of the normalised parameters )
*
/
)
(PZT-5H)
on the volume fraction of the PZT-5H ceramic m
PZT
= 1 m
p
in porous
materials are shown in Figures 4.14.5. These dependences were determined for
both porous ceramics (PZT-5Hair) and polymer-impregnated ceramic foams
(PZT-5Hpolymer) over wide m
PZT
ranges. Some technological aspects of
manufacturing these materials and features of their microstructure were considered
and discussed in refs. [14, 20, 21]. Low-density foams (m
PZT
< 0.3) were prepared
by coating the polymeric foams with a ceramic suspension and by subsequent
sintering. The porous materials with m
PZT
> 0.3 were processed using a variety of
volatile additives added to the ceramic powder before uniaxial pressing. The
fugitive additives include polymethyl methacrylate (PMMA), polyethylene oxide
(PEO), self-raising flour (SR), and corn starch (CS), which all produce porosity
within the ceramic medium on volatilisation during heat treatment. The scatter in
the data in graphs (Figures 4.14.5) is thought to originate from the different
microstructures formed by the different inclusion types and sizes. For example,
PMMA-based materials exhibit cracking parallel to the pressing (and poling)
direction [14, 21]: this cracking leads to fairly large / ratios (Figure
*
33
d
) 5 (
33
H PZT
d

porosity patterns and calculated a series of the effective parameters for porous PZT
Figure 4.1. Experimental dependence of the normalised piezoelectric coefficient /
on the relative density (i.e., volume fraction of the ceramic component m
PZT
) for
different porous materials based on the PZT-5H ceramic (reprinted from Topolov et al. [15],
with permission from Maney Publishing)
*
33
d
) 5 (
33
H PZT
d

Figure 4.2. Experimental dependence of the normalised piezoelectric coefficient /
PZT
) for
*
31
d
) 5 (
31
H PZT
d

Figure 4.3. Experimental dependence of the normalised hydrostatic piezoelectric coefficient
/ on the relative density (i.e., volume fraction of the ceramic component m
PZT
)
for different porous materials based on the PZT-5H ceramic (reprinted from Topolov et al.
[15], with permission from Maney Publishing)
*
h
d
) 5 ( H PZT
h
d

Figure 4.4. Experimental dependence of the normalised dielectric permittivity /
PZT
) for
V
H
*
33
V
H
), 5 (
33
H PZT
Figure 4.5. Experimental dependence of the normalised hydrostatic piezoelectric coefficient
/ on the relative density (i.e., volume fraction of the ceramic component m
PZT
)
for different porous materials based on the PZT-5H ceramic (reprinted from Topolov et al.
[15], with permission from Maney Publishing)
*
h
g
) 5 ( H PZT
h
g

4.1) along with small / and / ratios (Figures 4.2 and 4.4)
in a wide m
PZT
range.
*
31
d
) 5 (
31
H PZT
d
V
H
*
33
V
H
), 5 (
33
H PZT
Figure 4.1 shows the dependence of the piezoelectric coefficient on m
PZT
for foams and porous materials manufactured from the different inclusion
materials. The general trend observed is a decrease in the value with
decreasing m
PZT
. This trend shows a distinct change in at m
PZT
< 0.75 and a
slighter change at m
PZT
> 0.75. It is found that different processing methods result
in different values for the same relative density or porosity m
p
, and, moreover,
the condition | const established in a wide m
p
range [1012] is violated. This
implies that the piezoelectric coefficient also depends on features of
microstructure, manufacturing and poling conditions and other factors that are to
be taken into consideration in future studies. In other words, many of the models
assume that the piezo-active material is fully poled along a common axis without
regard to regions surrounding the air inclusions or pores, and features of
inhomogeneous poling should be taken into consideration for future modelling
activities.
*
33
d
*
33
d
*
33
d
*
33
d
*
33
d
*
33
d
Variations among additives are related to different microstructures formed in
the porous materials. The variation of the piezoelectric coefficient with m
PZT
is
shown in Figure 4.2. The general trend observed here consists in a decrease in the
*
31
d
*
31
d
/ ratio with decreasing m
PZT
, but at a more rapid rate than /
(see Figure 4.1), particularly in the high density range, at m
PZT
> 0.5.
Such a decrease in / or | | suggests that an electromechanical
interaction between some piezo-active elements aligned in the poling direction OX
3
becomes weaker as the porosity of the material surrounding these elements
increases. The rapid decrease in | | and the subsequent increase in the
piezoelectric anisotropy / | | results in high values of the hydrostatic
piezoelectric coefficient (Figure 4.3). A maximum of is present at volume
fractions m
PZT
from a range 0.5 0.75. The highest values are attained using
PMMA as the inclusion material. The foam-based materials, while having a
modest value (100 200 pC / N), still show a higher value than that for the
monolithic PZT-5H ceramic (around 50 pC / N). These foam materials have many
other effective property advantages in their favour, including low density and
acoustic impedance, although the relatively low strength of these materials should
also be considered.
) 5 (
31
H PZT
d
*
33
d
) 5 (
33
H PZT
d

*
31
d
) 5 (
31
H PZT
d
*
31
d
*
31
d
*
33
d
*
31
d
*
h
d
*
h
d
*
h
d
*
h
d
Figure 4.4 shows how the normalised dielectric permittivity /
depends on the volume fraction m
PZT
. The results for samples made with
PEO and SR have similar values over the wide m
PZT
range. However, the samples
made with PMMA have lower values than expected; and is a result of the
presence of cracks [14, 21] that are oriented with their length mainly perpendicular
to the poling direction OX
3
. In contrast, the samples made with CS have a higher
permittivity , possibly due to the relatively small pore size. For a further
comparison, it should be added that the porous PZT ceramic manufactured using
capsule-free hot-isostatic pressing [23] shows high dielectric permittivity
along with good connectivity on the electric field direction.
V
H
*
33
V
H
), 5 (
33
H PZT
V
H
*
33
V
H
*
33
V
H
*
33
The hydrostatic piezoelectric coefficient reaches fairly large values (i.e.,
50 and more) at volume fractions m
PZT
< 0.5 (Figure 4.5) due to relatively
high hydrostatic piezoelectric activity (Figure 4.3) and low dielectric permittivity
(Figure 4.4). The large values are important for hydroacoustic applications. It
should be noted that the samples made with PMMA have the large values
arising from their suppressed and values caused by cracking of these
samples [14] during manufacture. It can also be seen that the samples made with
CS have a lower value than the samples made with PEO and SR, owing to the
large of the samples made with CS (see Figures 4.4 and 4.5). The PZT-5H
air-foam-based materials exhibit the large values but are difficult to use due to
their relatively poor mechanical strength. However, the PZT-5Hpolymer-foam-
based materials, while having a lower value, are more robust and still well
suited to hydrophone applications.
*
h
g
) 5 ( H PZT
h
g

*
h
g
*
h
g
V
H
*
33
*
31
d
*
h
g
V
H
*
33
*
h
g
*
h
g
In the next section we consider an example of prediction of the effective
electromechanical properties in porous ceramics and composites and compare
results of modelling and known experimental data. This consideration is concerned
with porous materials based on the Pb(Zr, Ti)O
3
-type ceramics.
4.3 Model of the Modified Layered Composite: Application to
Porous Materials
The possibility of creating a three-component piezo-composite based on
ferroelectric ceramics that exhibit considerable piezoelectric anisotropy was first
discussed by Topolov and Turik [43]. Such a model can be adapted to describe a
porous material with different arrangements of porosity and second phase. As
follows from results [44] on the 22 series-connected composite shown in Figure
3.1a, absolute values of the piezoelectric anisotropy ]
d
can be higher than the /
| | ratios related to the piezo-active components (n = 1 or 2). An increase in |]
d
|
is possible even in the presence of two piezo-active components with small /
| | ratios (for example, less than 3) [44]. The influence of the 22-type structure
on ]
d
can be enhanced in a modified layered composite [43]. Such a structure
contains two types of series-connected layers (see Figure 3.1a). The layer of the
first type is made of a polymer matrix that is reinforced by ceramic rods distributed
continuously along the poling axis OX
3
, i.e., 13 connectivity takes place in this
layer. The layer of the second type comprises of ceramic only. The aforementioned
layers alternate in the OX
3
direction and promote a redistribution of internal electric
and mechanical fields. This redistribution results in large |]
d
| and |]
e
| values [43] in
specific ranges of volume fractions of the ceramic components.
) (
33
n
d
) (
31
n
d
) (
33
n
d
) (
31
n
d
Based on results [43], a similar model containing two types of piezo-active
layers (Figure 4.6) is used to describe the electromechanical properties of porous
ceramics and composites [15]. It is assumed that the relatively small decrease in
the piezoelectric coefficient with increasing porosity m
p
results from a
continuous distribution of the piezo-ceramic framework along the poling axis OX
3
,
and it can take place with the presence of elements in the structure with 13
connectivity. According to the model from ref. [15], the layer of the first type
represents a piezo-passive matrix reinforced by ceramic cylindrical rods being
lengthy and parallel to the poling axis OX
3
. The layer of the second type is
assumed to be ceramic and lengthy in the OX
1
and OX
2
directions. These layers,
which randomly alternate along the OX
3
direction, are the main elements of the
structure of porous ceramic or composite materials (Figure 4.6). Volume fractions
of their piezo-active (m
PZT
) and piezo-passive (m
p
) components are determined by
the formulae m
PZT
= m
s
m
r
+ 1 m
s
and m
p
= m
s
(1 m
r
). An additional effect of
porosity can be taken into account if the polymer matrix in the layers of the first
type is considered as a cellular structure with uniformly distributed spherical air
inclusions, and the volume fraction of these inclusions in the matrix equals m
in
.
*
33
d
Figure 4.6. Schematic diagram showing the structure of the porous piezo-active ceramic or
porous composite based on the piezo-ceramic. The dashed area is the monolithic ceramic
component with the remanent polarisation vector P
R
nn OX
3
, the clear area is the piezo-
passive medium such as air, polymer or porous polymer (reprinted from Topolov et al. [15],
The effective parameters )
*
= )
*
(m
PZT
) of the porous material schematically
shown in Figure 4.6 are determined using formulae for piezo-composites in three
stages. Averaging on m
in
is carried out on the basis of formulae [4] (03
connectivity), averaging on m
r
is based on formulae [45] (13 connectivity) and
averaging on m
s
is performed with formulae [6] (22 connectivity, see also
Appendix). In calculating the effective electromechanical constants of the 13-type
layer, the approximations concerned with the aspect ratio D
r
of the ceramic rods
(see Section 3.3.3) should also be considered. It is assumed that the rods elongated
in porous samples in the OX
3
direction (Figure 4.6) are characterised by values of
the aspect ratio D
r
t 5.
Examples of the )
*
(m
PZT
) dependences calculated for porous materials are
shown in Table 4.1 and Figure 4.7. Table 4.1 also shows the dependences of the
normalised piezoelectric coefficients / on the parameters m
r
and m
s
.
The axes in the graphs of Figure 4.7 are to the same scale as the experimental data
in Figures 4.1 4.5 for direct comparison. The data from Table 4.1 and Figure 4.7
contain the )
*
(m
PZT
) dependences calculated for four different cases. Cases I and II
are related to the PZT-5H / polyethylene and PZT-5H / elastomer composites,
respectively. Case III is related to the PZT-5H / porous elastomer composite with
the regular distribution of spherical pores (m
in
= 0.3) in the polymer matrix. Case
IV is related to the PZT-5H ceramic / air composite (i.e., limiting case of the most
compliant matrix).
*
3l
d
) 5 (
3
H PZT
l
d

4.4 Comparison of Calculated and Experimental Results
The calculated and experimental results can be compared in two ways. First, a
comparison of the normalised parameters )
*
/ )
(PZT-5H)
for cases I and II suggests
that the substitution of the polymer component by a more compliant one (e.g.,
polyethylene o elastomer) results in
(i) a slight increase in (Figure 4.7b),
V
H
*
33
Table 4.1. Calculated normalised piezoelectric coefficients / of composite
materials based on the PZT-5H ceramic (reprinted from Topolov et al. [15], with permission
from Maney Publishing)
*
3l
d
) 5 (
3
H PZT
l
d

m
s
m
r
m
p
m
PZT
*
31
d
/
) 5 (
31
H PZT
d
*
33
d
/
) 5 (
33
H PZT
d

Case I: PZT-5H polyethylene
0.10 0.08 0.092 0.908 0.413 0.724
0.15 0.10 0.135 0.865 0.388 0.745
0.20 0.10 0.180 0.820 0.327 0.720
0.25 0.13 0.218 0.782 0.355 0.771
0.30 0.12 0.264 0.736 0.299 0.738
0.35 0.12 0.308 0.692 0.273 0.728
0.40 0.10 0.360 0.640 0.215 0.672
0.45 0.10 0.405 0.595 0.202 0.667
0.50 0.09 0.455 0.545 0.175 0.634
0.55 0.06 0.517 0.483 0.117 0.519
0.60 0.05 0.570 0.430 0.0984 0.468
0.65 0.04 0.624 0.376 0.0821 0.409
Case II: PZT-5H Elastomer
0.10 0.08 0.092 0.908 0.445 0.872
0.15 0.10 0.135 0.865 0.412 0.885
0.20 0.10 0.180 0.820 0.346 0.872
0.25 0.13 0.218 0.782 0.368 0.900
0.30 0.12 0.264 0.736 0.308 0.883
0.35 0.12 0.308 0.692 0.279 0.878
0.40 0.10 0.360 0.640 0.218 0.847
0.45 0.10 0.405 0.595 0.201 0.843
0.50 0.09 0.455 0.545 0.171 0.823
0.55 0.06 0.517 0.483 0.112 0.744
0.60 0.05 0.570 0.430 0.0913 0.703
0.65 0.04 0.624 0.376 0.0737 0.650
Case III: PZT-5H porous elastomer
0.10 0.08 0.092 0.908 0.460 0.962
0.15 0.10 0.135 0.865 0.422 0.967
0.20 0.10 0.180 0.820 0.354 0.963
0.25 0.13 0.218 0.782 0.373 0.972
0.30 0.12 0.264 0.736 0.311 0.966
0.35 0.12 0.308 0.692 0.280 0.965
0.40 0.10 0.360 0.640 0.217 0.955
0.45 0.10 0.405 0.595 0.199 0.954
0.50 0.09 0.455 0.545 0.166 0.947
0.55 0.06 0.517 0.483 0.106 0.917
0.60 0.05 0.570 0.430 0.0833 0.900
0.65 0.04 0.624 0.376 0.0638 0.876
Case IV: PZT-5H Air
0.10 0.08 0.092 0.908 0.466 1
0.15 0.10 0.135 0.865 0.426 1
0.20 0.10 0.180 0.820 0.358 1
0.25 0.13 0.218 0.782 0.375 1
m
s
m
r
m
p
m
PZT
*
31
d
/
) 5 (
31
H PZT
d
*
33
d
/
) 5 (
33
H PZT
d

0.30 0.12 0.264 0.736 0.313 1
0.35 0.12 0.308 0.692 0.281 1
0.40 0.10 0.360 0.640 0.218 1
0.45 0.10 0.405 0.595 0.199 1
0.50 0.09 0.455 0.545 0.166 1
0.55 0.06 0.517 0.483 0.104 1
0.60 0.05 0.570 0.430 0.0811 1
0.65 0.04 0.624 0.376 0.0607 1
(ii) a considerable increase in | | (Table 4.1) and (Figure 4.7a) and
*
3l
d
*
h
d
(iii) a more significant increase in (Figure 4.7c) at 0 < m
PZT
< 0.5.
*
h
g
This change in the matrix material influences the piezoelectric anisotropy value ]
d
of the composite, primarily due to the change in the elastic properties and Poisson
ratio of the polymer component.
Second, comparing data from cases II, III and IV, it is possible to conclude that
the changes in and are relatively small. However, the , and
values increase significantly when an elastomer is substituted by a more compliant
porous elastomer or by air. Such behaviour is concerned with the weak
electromechanical interaction between piezo-active rods in the OX
1
and OX
2
directions, owing to the piezo-passive matrix, and with conditions favourable to the
rodlayer electromechanical interaction in the OX
3
direction [15]. It is known
[44, 45] that softening the matrix surrounding the piezo-active rods in the 13
composites provides higher values of , , , and even at small volume
fractions m
r
of the rods. This can be understood by considering the application of a
force to the composite along the poling axis. As the stiffness of the polymer matrix
is reduced, more force is experienced by the stiffer piezoelectric material and
hence there is an increase in . An important consequence of the slight (m
s
,
m
r
) dependence at small m
r
values (see cases II and III in Table 4.1) is the equality
V
H
*
33
*
31
d
*
33
d
*
h
d
*
h
g
*
33
d
*
33
g
*
h
d
*
h
g
*
33
d
*
33
d
/ = / (4.3)
*
33
g
) 5 (
33
H PZT
g
2 *
33
) (Q
2 ) 5 (
33
) (
H PZT
Q

which holds true with accuracy up to a few per cent [15].
The correlation between the calculated (Figure 4.7 and Table 4.1) and
experimental (Figures 4.14.5) volume-fraction dependences of the effective
parameters exists in a wide porosity range. According to the model concepts [15],
changes in m
p
are associated with changing the parameters m
s
and m
r
of the layers
of the first type (Figure 4.6). As porosity m
p
increases, m
s
also increases, i.e., at m
p
c
> m
p
the inequality m
s
c > m
s
holds and m
r
is varied within a narrow range. As seen
from Table 4.1, the relation m
p
c / m
p
| m
s
c / m
s
holds at 0.10 < m
p
< 0.60. However,
as m
p
approaches 0.60, a comparison of the calculated and experimental values of
)
*
/ )
(PZT-5H)
becomes problematic [15]. Our calculations for m
p
t 0.60 suggest that
the calculated and experimental parameters would be in good agreement at m
r
<
0
1
2
3
4
5
6
7
8
9
10
0 0.25 0.5 0.75 1
Relative Density
d
*
h
/
d
h
(
P
Z
T
-
5
H
)
Case I
Case II
Case III
Case IV
a
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 0.25 0.5 0.75 1
Relative Density
H
*
3
3
T
/
H
3
3
T
(
P
Z
T
-
5
H
)
Case I
Case II
Case III
Case IV
b
0
20
40
60
80
100
120
0 0.25 0.5 0.75 1
Relative Density
g
*
h
/
g
h
(
P
Z
T
-
5
H
)
Case I
Case II
Case III
Case IV
c
Figure 4.7. Calculated normalised effective parameters )* of porous composites based on
the PZT-5H ceramic: ) = d
h
(a), (b) and g
h
(c) (reprinted from Topolov et al. [15], with
permission from Maney Publishing)
V
H
33
0.10. However, the structure of the porous material with small m
r
values at m
s
> 2/3
(see Figure 4.6) becomes mechanically unstable and may be destroyed by a small
amount of mechanical loading. This highlights the range of 0.10 < m
p
< 0.60 as the
most suitable porosity range wherein the model [15] can be used for interpretation
of experimental data on porous piezo-active materials.
Another example of the application of the model [15] for interpretation of
experimental results is concerned with the work in ref. [35] on porous piezo-active
materials based on the ZTS-19 ceramic. This ceramic also belongs to the Pb(Zr,
Ti)O
3
-type materials and has piezoelectric coefficients satisfying
conditions (4.1) (see Table 2.1). A comparison of the calculated and experimental
[35] parameters (Table 4.2) demonstrates the applicability of the model [15] in a
wide porosity range. It should be added that the effective parameters of the porous
composite based on ZTS-19 are subjected to small changes (about 12%) as
) 19 (
3
ZTS
l
e
Table 4.2. Calculated
a
and experimental (in parentheses, from ref. [35]) values of
normalised effective parameters )
*
/ )
(ZTS-19)
of the porous ZTS-19 ceramic
m
s
m
r
m
p
V
V
H
H
), 19 (
33
*
33
ZTS ) 19 (
31
*
31
ZTS
d
d
) 19 (
33
*
33
ZTS
g
g
2 ) 19 (
33
2 *
33
) (
) (
ZTS
Q
Q
0.160
0.160
0.180
0.180
0.300
0.230
0.330
0.250
0.112
0.123
0.121
0.135
0.130
0.729
0.652
0.733
0.650
(0.736)
0.730
0.653
0.733
0.651
(0.648)
1.37
1.53
1.36
1.53
(1.36)
1.37
1.52
1.36
1.53
(1.36)
0.320
0.320
0.340
0.340
0.230
0.270
0.280
0.240
0.246
0.234
0.245
0.258
0.250
0.483
0.537
0.536
0.484
(0.533)
0.485
0.538
0.534
0.482
(0.488)
2.06
1.85
1.86
2.06
(1.88)
2.04
1.85
1.86
1.99
(1.88)
0.420
0.420
0.440
0.440
0.200
0.210
0.210
0.220
0.336
0.332
0.375
0.343
0.350
0.374
0.388
0.377
0.394
(0.375)
0.376
0.390
0.379
0.391
(0.385)
2.65
2.55
2.63
2.54
(2.67)
2.63
2.53
2.61
2.52
(2.67)
0.500
0.500
0.520
0.520
0.120
0.130
0.130
0.140
0.440
0.435
0.452
0.447
0.450
0.215
0.231
0.224
0.239
(0.234)
0.218
0.234
0.227
0.242
(0.233)
4.56
4.26
4.39
4.12
(4.34)
4.48
4.19
4.32
4.06
(4.42)
0.600
0.600
0.620
0.620
0.110
0.120
0.110
0.140
0.534
0.528
0.552
0.533
0.530
0.172
0.186
0.167
0.209
(0.172)
0.175
0.189
0.171
0.212
(0.211)
5.71
5.28
5.86
4.72
(5.47)
5.59
5.18
5.74
4.64
(5.47)
Note to Table 4.2
a. Calculations were made within the framework of the model of ref. [15]. In the layers of
the first type, ceramic rods (Figure 4.6) are embedded in a porous elastomer matrix. This
matrix contains spherical pores with the volume fraction m
in
= 0.3
porosity m
in
in the elastomer matrix is varied or even as this matrix is replaced by
air. The normalised piezoelectric coefficient (m
s
, m
r
) / undergoes slight
changes (under 5%) as m
s
and m
r
are varied in wide ranges to obey condition m
p
<
0.60. At the same time, (m
s
, m
r
) / and (m
s
, m
r
) /
considerably decrease on increasing m
p
(Table 4.2). As a result of the constant
effective piezoelectric coefficient of the porous material at 0.10 < m
p
< 0.60, an
equality / = / is valid with an accuracy up to a few
per cent (compare with Equation 4.3). In general, the calculated effective
parameters from Table 4.2 follow the experimental values due to both the
monotonic decrease in m
r
and the monotonic increase in m
s
as porosity m
p
increases. A similar trend was shown earlier for / (Table 4.1), but on
decreasing m
r
in the fairly narrow range only.
*
33
d
) 19 (
33
ZTS
d
*
31
d
) 19 (
31
ZTS
d
V
H
*
33
V
H
), 5 (
33
H PZT
*
33
d
*
33
g
) 19 (
33
ZTS
g
2 *
33
) (Q
2 ) 19 (
33
) (
ZTS
Q
*
3l
d
) 5 (
3
H PZT
l
d

Thus, the porous piezo-active composite (ceramic) material can be represented
as a combination of structural elements that are described by either 13 or 22
connectivity (Figure 4.6). These elements play an important role in forming the
effective electromechanical properties and the variable piezoelectric anisotropy in
the range of porosity 0.10 < m
p
< 0.60. The combination of high piezoelectric
activity with high piezoelectric sensitivity in the studied Pb(Zr, Ti)O
3
-based porous
materials promotes their piezotechnical applications as active elements of sensors,
hydrophones and ultrasound transducers. In particular, results [46] demonstrate the
possible use of ferroelectric hard porous ceramics (for example, based on the
PCR-78 composition) in ultrasound transducers with high intensity. An application
of novel porous ceramics of the PCR group in the piezoelectric transducers of
medical devices for therapy and diagnostics was discussed in ref. [47].
4.5 From 33 Connectivity to Porous Materials
A simple view of understanding the large and values of porous piezo-active
composites with 33 connectivity can be developed using the Banno unit cell
(Figure 3.12a) that consists of a perfectly interpenetrating piezo-active (poled
ferroelectric ceramic) and a piezo-passive component (polymer or air). The
piezoelectric ceramic is assumed to be fully poled in the OX
3
direction. Figure 4.8
shows a side view of the Banno unit cell from Figure 3.12a. When considering the
variation of the effective piezoelectric coefficient with the polymer volume
fraction it is necessary to consider a force applied parallel to the poling axis (as in
Figure 4.8a) and the specific piezoelectric elements that contribute to the
generation of charge in response to the applied force. In the case of Figure 4.8a the
majority of the force is experienced by piezoelectric areas 1 and 2, since the
stiffness of the piezo-ceramic component is greater than that of the polymer
component. As a result, the piezoelectric coefficient is dependent on the sizes
(and fraction) of areas 1 and 2. As the polymer volume fraction is increased
*
h
g
*
h
d
*
33
d
*
33
d
Polymer Polymer
Force (d
33
)
Force (d
31
)
1
2
2 3
Polymer
Polymer
Force (d
33
)
1
2
2 3
Force (d
31
)
X
1
X
3
a c
b d
Figure 4.8. Cross-sections of the Banno unit cell of the 33 piezo-composite and vectors of
external forces applied to the sample. The piezoelectric coefficients determined at the direct
piezoelectric effect are given in parentheses
(Figure 4.8b) the fraction of areas 1 and 2 gradually reduce and hence will
reduce.
*
33
d
To interpret the behaviour of the piezoelectric coefficient , we consider the
response of the Banno unit cell with a force applied normal to the poling direction,
as shown in Figure 4.8c. In this case, the high stiffness of the ceramic results in the
majority of the force being experienced by areas 2 and 3. Since area 3 is in series in
the OX
3
direction with a piezo-passive polymer it does not contribute to the
piezoelectric coefficient . Hence, only area 2 contributes to the piezoelectric
coefficient of the composite. As can be seen from Figure 4.8d, as the polymer
volume fraction is increased the fraction of piezoelectric contributing to (area 2
in Figure 4.8d) falls at a much quicker rate than (areas 1 and 2 in Figure 4.8d).
Therefore, while decreases slowly, quickly falls to small values (see
Figures 4.1 and 4.2). This combination of high and low results in the large
hydrostatic piezoelectric coefficient . The reduction in permittivity also results
in large values of the hydrostatic piezoelectric coefficient (see Figures 4.4 and
4.5). The optimum polymer volume fraction for high is ultimately a balance of
having sufficient polymer component to lead to a reduction in , while avoiding
a significant reduction is (which can occur at high polymer volume fractions,
see, for example, a range of low relative density in Figure 4.1).
*
31
d
*
31
d
*
31
d
*
31
d
*
33
d
*
33
d
*
31
d
*
33
d
*
31
d
*
h
d
*
h
g
*
h
d
*
31
d
*
33
d
It should be noted that if the second component is air (i.e., a continuous
medium with very large compliance) then all of the force is experienced by areas 1
and 2 (Figure 4.8) irrespective of the polymer volume fraction of the ceramic
component, and remains constant. This is clearly not the case in some systems
(see, for example, Figure 4.1) and is possibly due to a lack of interconnectivity at
small ceramic volume fraction and/or due to the difficulty in poling highly porous
samples.
*
33
d
Thus, features of volume-fraction dependences of the piezoelectric coefficients
, and of the porous materials are closely connected with certain elements
of microstructure responsible for such piezoelectric properties. It is clear that there
is no unified approach to predict the effective electromechanical properties in
porous materials. Therefore, the different connectivity patterns (03, 13, 22, 30,
31, 33, etc.) can be taken into consideration for an analysis of experimental data
on the electromechanical properties and related parameters of the piezo-active
porous materials.
*
3l
d
*
h
d
*
h
g
References
[1] Banno H, (1987) Effects of shape and volume fraction of closed pores on dielectric,
elastic, and electromechanical properties of dielectric and piezoelectric ceramics a
theoretical approach. Ceramic Bulletin 66:13321337
[2] Wersing W, (1986) Composite piezoelectrics for ultrasonic transducers. In: ISAF86:
Proceedings of the Sixth International Symposium on Applications of Ferroelectrics.
Lehigh University, Bethlehem, PA, USA, 811 June, 1986. IEEE, Piscataway, pp
212223
[3] Wersing W, Lubitz K and Mohaupt J, (1986) Dielectric, elastic and piezoelectric
properties of porous PZT ceramics. Ferroelectrics 68:7797
[4] Dunn M, (1995) Effects of grain shape anisotropy, porosity, and microcracks on the
elastic and dielectric constants of polycrystalline piezoelectric ceramics. Journal of
[5] Gmez Alvarez-Arenas T, Montero de Espinosa Freijo F, (1996) New constitutive
relations for piezoelectric composites and porous piezoelectric ceramics. Journal of
the Acoustical Society of America 100:31043114
[7] Kahn M, (1985) Acoustic and elastic properties of PZT ceramics with anisotropic
pores. Journal of the American Ceramic Society 68:623628
[8] Ting R, (1986) Evaluation of new piezoelectric composite materials for hydrophone
applications. Ferroelectrics 67:143157
[9] Kahn M, Dalzell A, Kovel B, (1987) PZT ceramicair composites for hydrostatic
sensing. Advanced Ceramic Materials 2:836840
[10] Lopatin SS, Lupeiko TG, (1991) Properties of the porous piezoelectric ceramic of the
lead zirconate-titanate type. Neorganicheskie Materialy 27:19481951 (in Russian)
[12] Banno H, (1993) Effects of porosity on dielectric, elastic and electromechanical
properties of Pb(Zr, Ti)O
3
ceramics with open pores: a theoretical approach. Japanese
Journal of Applied Physics Pt 1 32:42144217
[13] Wu T-L, (2000) Micromechanics determining of electroelastic properties of
piezoelectric materials containing voids. Materials Science and Engineering A
280:320327
[14] Kara H, Ramesh R, Stevens R, Bowen CR, (2003) Porous PZT ceramics for receiving
transducers. IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
50:289296
[15] Topolov VYu, Glushanin SV, Bowen CR, (2005) Piezoelectric response of porous
ceramic and composite materials based on Pb(Zr, Ti)O
3
: experiment and modelling.
Advances in Applied Ceramics 104:300305
[16] Madelung O, Nelson DF (eds.) Landolt-Brnstein, (1993) Numerical Data and
Functional Relationships in Science and Technology. New Series. Group III: Solid
State Physics. Vol. 29, Subvol. b: Piezoelectric, Pyroelectric and Related Constants.
Springer, Berlin Heidelberg New York London Paris Tokyo Hong Kong Barcelona
Budapest
[17] Skinner DP, Newnham RE, Cross LE, (1978) Flexible composite transducers.
Materials Research Bulletin 13:599607
[18] Creedon MJ, Schulze WA, (1994) Axially distorted 33 piezoelectric composites for
hydrophone applications. Ferroelectrics 153:333339
[19] Shrout TR, Schulze WA, Biggers JV, (1979) Simplified fabrication of PZT / polymer
composites. Materials Research Bulletin 14:15531559
[20] Rittenmyer K, Shrout T, Schulze WA, Newnham RE, (1982) Piezoelectric 33
[21] Bowen CR, Perry A, Lewis ACF, Kara H, (2004) Processing and properties of porous
piezoelectric materials with high hydrostatic figures of merit. Journal of the European
Ceramic Society 24:541545
[22] Kara H, Perry A, Stevens R, Bowen CR, (2002) Interpenetrating PZT / polymer
composites for hydrophones: models and experiments. Ferroelectrics 265:317332
[23] Hayashi T, Sugihara S, Okazaki K, (1991) Processing of porous 33 PZT ceramics
using capsule-free O
2
-HIP. Japanese Journal of Applied Physics 30:22432246
[24] Praveen Kumar B, Kumar HH, Kharat DK, (2005) Study on pore-forming agents in
processing of porous piezoceramics. Journals of Materials Science: Materials in
Electronics 16:681686
[25] Praveenkumar B, Kumar HH, Kharat DK, (2006) Study on microstructure,
piezoelectric and dielectric properties of 33 porous PZT composites. Journal of
Materials Science: Materials in Electronics 17:515518
[26] Levassort F, Holc J, Ringgaard E, Bove T, Kosec M, Lethiecq M, (2007) Fabrication,
modelling and use of porous ceramics for ultrasonic transducer applications. Journal
of Electroceramics 19:127139
[27] Geis S, Fricke J, Lobmann P, (2002) Electrical properties of PZT aerogels. Journal of
the European Ceramic Society 22:11551161
[28] Cracuin F, Galassi G, Roncari E, Filippi A, Guidarelli G, (1998) Electro-elastic
properties of porous piezoelectric ceramics obtained by tape casting. Ferroelectrics
205:4957
[29] Bast U, Wersing W, (1989) The influence of internal voids with 31 connectivity on
the properties of piezoelectric ceramics prepared by a new planar process.
[30] Lopatin SS, Lupeiko TG, Nesterov AA, Vikhryanova IV, (1988) Properties of lead
zirconate-titanate ceramic with oriented pores. Izvestiya Akademii Nauk SSSR.
Neorganicheskie Materialy 24:12291230 (in Russian)
[31] Hikita K, Yamada K, Nishioka M, Ono M, (1983) Effect of porous structure on
piezoelectric properties of PZT ceramics. Japanese Journal of Applied Physics 22
(Suppl. 2):6466
[32] Bruggeman DAG, (1935) Berechnung verschiedener physikalischer Konstanten von
heterogenen Substanzen. I. Dielektrizittskonstanten und Leitfhigkeiten der
Mischkrper aus isotropen Substanzen. Annalen der Physik 24:636664
[33] Rybjanets AN, (2002) Porous piezo-ceramic materials. Part 1. Structure, technology
of processing and mathematical models. In: Proceedings of the International
Scientific-Practical Conference Basic Problems of Piezoelectric Device Designing
Piezotechnics-2002. 1721 September 2002. Tver State University, Tver, pp 4855
(in Russian)
[34] Rybjanets AN, (2002) Porous piezo-ceramic materials. Part 2. Electrophysical
properties. In: Proceedings of the International Scientific-Practical Conference Basic
Problems of Piezoelectric Device Designing Piezotechnics-2002. 1721 September
2002. Tver State University, Tver, pp 5664 (in Russian)
[35] Vorontsov AA, Glushanin SV, Topolov VYu, Panich AE, (2003) Features of the
piezoelectric effect and piezoelectric sensitivity of Pb(Zr, Ti)O
3
-based porous
ceramics and composites. In: Bolshakov VI (ed.) Aims for Future of Engineering
Science. Collection of Articles. Issue 5. Gaudeamus, Dnepropetrovsk, pp 107114 (in
Russian)
[36] Rybjanets AN, Nasedkin AV, Turik AV, (2004) New micro structural design concept
for polycrystalline composite materials. Integrated Ferroelectrics 63:179182
[37] Kar-Gupta R, Venkatesh TA, (2006) Electromechanical response of porous
piezoelectric materials. Acta Materialia 54:40634078
[38] Kar-Gupta R, Venkatesh TA, (2007) Electromechanical response of porous
piezoelectric materials: Effects of porosity distribution. Applied Physics Letters
91:062904 3 p
[39] Asai M, Takano N, Uetsuji Y, Taki K, (2007) An iterative solver applied to strongly
coupled piezoelectric problems of porous Pb(Zr, Ti)O
3
with nondestructive modelling
of microstructure. Modelling and Simulation in Materials Science and Engineering
15:597617
[40] Crumm AT, Jalloran JW, Silva ECN, Montero de Espinosa F, (2007)
Microconfigured piezoelectric artificial materials for hydrophones. Journal of
Materials Science 42:39443950
[41] Reilly CJ, Halloran JW, Silva ECN, (2007) Piezoelectric hydrophones from optimal
design, Part I: fabrication. Journal of Materials Science 42:48244831
[42] Reilly CJ, Halloran JW, Silva ECN, Montero de Espinosa F, (2007) Piezoelectric
hydrophones from optimal design, II: properties. Journal of Materials Science
42:48104814
[43] Topolov VYu, Turik AV, (1998) Modified layered composite with high anisotropy of
piezoconstants and
*
. Technical Physics Letters 24:441443
*
ij
e
ij
d
[44] Topolov VYu, Turik AV, (2000) Non-monotonic concentration dependence of
electromechanical properties of piezoactive 22 composites. Journal of Physics D:
of Composite Materials 25:5461
[46] Rybjantes AN, (2002) Physical properties and applications of ferroelectrically hard
porous ZTS-based piezo-ceramics. In: Proceedings of the International Scientific-
Practical Conference Basic Problems of Piezoelectric Device Designing
Piezotechnics-2002. 1721 September 2002. Tver State University, Tver, pp 6572
(in Russian)
Device Designing Piezotechnics-2002. 1721 September 2002. Tver State
University, Tver, pp 305308 (in Russian)
[47] Sitalo EI, Smotrakov VG, Eremkin VV, Ivanov PV, Tsikhotsky ES, (2002)
Technology of processing and properties of the porous piezo-ceramic. In: Proceedings
of the International Scientific-Practical Conference Basic Problems of Piezoelectric

5
Effective Properties in Novel Piezo-composites Based on
Relaxor-ferroelectric Single Crystals
5.1 Relaxor-ferroelectric Solid Solutions and Engineering
In the last decades, crystallographic and compositional engineering has become an
important tool [1] in the development of advanced piezo-active materials. An
important group of such materials comprises perovskite-type solid solutions with a
general formula (1 x)Pb(B
1
, B
2
)O
3
xPbTiO
3
. The complex perovskites Pb(B
1
,
B
2
)O
3
are disordered compounds containing ions of metals from the following
groups: B
1
= Mg, Zn, Ni, Fe, Sc, Yb, and In (low valence) and B
2
= Nb, Ta and W
(high valence) [13]. The combination of metal ions with low and high valences
results in a series of physical properties that distinguish Pb(B
1
, B
2
)O
3
from
normal (ordered) perovskite-type ferroelectrics such as PbTiO
3
. The complex
perovskites Pb(B
1
, B
2
)O
3
exhibit a broad and frequency-dispersive dielectric
maxima and contain polar nanoregions (with ferro- or antiferroelectric ordering) in
a non-polar phase over a wide temperature range [2, 3]. They are characterised by
the relaxation dielectric polarisation [3] and are called relaxors [2] or ferroelectric
relaxors.
The solid solutions of (1 x)Pb(B
1
, B
2
)O
3
xPbTiO
3
combine physical
properties of the relaxor-type and normal ferroelectric components, and excellent
properties are often observed near the MPB [24]. To achieve a high piezoelectric
activity, the relaxor-ferroelectric solid solutions are engineered by compositional
adjustment with a corresponding decrease in Curie temperature [4]. Of particular
interest for piezotechnical applications are domain-engineered / heterophase single
crystals of (1 x)Pb(Mg
1/3
Nb
2/3
)O
3
xPbTiO
3
(PMNxPT) and (1
y)Pb(Zn
1/3
Nb
2/3
)O
3
yPbTiO
3
(PZNyPT) with molar concentrations x and y,
respectively, taken in the vicinity of the MPB. In these single crystals engineered
non-180q domain structures [5, 6], intermediate ferroelectric phases [710] and
domain switching and rotation [1113] are observed. These single crystals poled
along certain crystallographic directions, for example, along [001] or [011] of the
perovskite unit cell [1416], exhibit very high piezoelectric activity and significant
electromechanical coupling. In [001]-poled polydomain PMN0.33PT single
crystals at room temperature, the piezoelectric coefficient d
33
and
electromechanical coupling factor k
33
can exceed 2800
.
pC / N and 90%,
respectively [14]. According to data in ref. [17], the piezoelectric coefficient d
33
in
[001] poled heterophase PZN0.08PT single crystals can attain values up to 12

000
pC / N at the electric-field-induced phase transition. In [001]-oriented PZN0.08PT
single crystals [13] d
33
becomes as large as 16

350 pC / N under the electric field E
| E
c
and d
33
= 2500 pC / N at E > E
c
, where E
c
is the coercive field (a threshold
field for a non-180q domain-wall motion).
High strains (over 0.5%) [1, 4] are achieved in the [001]-poled polydomain
single crystals with compositions of the 3m point group under an electric field E ||
[001]; although the spontaneous polarisation vectors in domains are parallel to the
[111], [111], [111], and [111] directions [5, 18]. Some interconnections between
the structure and properties in the PMNxPT and PZNyPT solid solutions have
been reviewed in ref. [19]. The high piezoelectric activity and large
electromechanical coupling in these solid solutions are closely connected with the
intermediate monoclinic phases [11, 12, 19] that occur close to the MPB. The high
piezoelectric activity in these solid solutions can be driven by a polarisation
rotation induced by the electric field E [11, 12], and this rotation between the
single-domain states in the tetragonal (4mm) and rhombohedral (3m) phases can be
implemented in different ways. The corresponding rotation paths found in ref. [11]
are the minimum free-energy paths and are associated with different intermediate
ferroelectric phases that would be observed in the studied solid solutions in certain
ranges of molar concentration x or y, electric field E and temperature T.
The aforementioned d
33
and k
33
values are considerably more than those related
to the conventional ferroelectric ceramics based on Pb(Zr
1z
Ti
z
)O
3
[2022], and the
outstanding performance of the PMNxPT and PZNyPT single crystals could lead
to a revolution in electromechanical transducer technology [23, 24] and
piezoelectric materials science. It should be added that the PMNxPT and PZN
yPT single crystals with domain-engineered structures exhibit relatively stable
electromechanical properties at room temperature. Full sets of the
electromechanical constants of these single crystals were measured on samples
(crystal cuts) poled along the [001] (Table 2.2) or [011] direction [16] of the
perovskite unit cell. The outstanding electromechanical properties in these single
crystals near the MPB suggest that these materials would be good candidates for
modern piezoelectric sensors, large-strain actuators, highly sensitive medical
ultrasonic transducers, hydrophones, ultrasonic imaging devices, and other
piezotechnical devices [4, 24].

In the sections below we analyse the possibilities and advantages of using the
relaxor-ferroelectric single crystals as components of the piezo-composites. We
demonstrate how these single crystals improve performance of the novel piezo-
composites, as compared with the performance of the conventional ferroelectric
ceramic / polymer composites, and promote the formation of a series of non-
monotonic volume-fraction dependences of the effective parameters.
Effective Properties in Novel Piezo-composites Based on Relaxor-ferroelectrics 125
5.2 03 Single Crystal / Ceramic Composites
In this section we consider single crystal / ceramic composites with 03
connectivity, where the PMNxPT or PZNyPT single crystals are used as isolated
inclusions with high piezoelectric activity. The motivation to study these
composites is due to a number of factors:

(i) The single crystal / ceramic composites based on PbTiO
3
have been previously
examined [25], and features of their effective electromechanical properties were
studied [26, 27] taking into account the composition, domain structure and
microstructure.
(ii) The 03 Pb(Ni
1/3
Nb
2/3
)O
3
PbTiO
3
single crystal / Pb(Ni
1/3
Nb
2/3
)O
3
Pb(Zr,
Ti)O
3
ceramic composites containing the relaxor-ferroelectric components were
manufactured by advanced hybrid sintering that combines microwave heating and
hot pressing [28] and show good characteristics. These composites have a good
interface state between the single-crystal inclusions and the ceramic matrix so that
neither cracks nor re-solidification of melted material would be observed at the
boundaries separating the components [28].
(iii) The modified PbTiO
3
ceramics are good candidates to be used as matrix
components in the 03 composites. PbTiO
3
ceramics are characterised [29, 30] by
moderate piezoelectric activity (e.g., the piezoelectric coefficient d
33
can be about
50 times less than that of the PMNxPT or PZNyPT single crystals) with a large
anisotropy of the piezoelectric coefficients (i.e., d
33
/ | d
31
| t 10) or
electromechanical coupling factors (i.e., k
33
/| k
31
| t 10 or k
t
/ | k
p
| t 10). The
piezoelectric coefficients e
3j
of these ceramics and single crystals are generally of
equal order of magnitude, but the piezoelectric coefficients e
31
differ in sign (see
Tables 2.1 and 2.2).
(iv) An alternative to the relaxor-ferroelectric single crystals could be textured
ceramics [31] wherein a high degree of crystallographic texture improves the
piezoelectric properties. According to data [31], textured PMNxPT ceramics (0.25
< x < 0.35) are processed by the homo-template grain growth, and a favourable
orientation of templates is achieved during the tape-casting process. An important
feature of such ceramics is the small influence of connectivity and spatial
distribution of the phases on the effective electromechanical properties.

Our further consideration is carried out within the framework of the model of a
03 composite with spheroidal inclusions (Section 3.3). Now, the single-crystal
inclusions are uniformly aligned in the ceramic matrix and form a periodic
structure. The main crystallographic axes X, Y and Z of the single crystals are
parallel to the co-ordinate axes OX
1
, OX
2
and OX
3
, respectively. The semi-axes of
the inclusions are a
1
, a
2
= a
1
and a
3
(Figure 3.7). The single-crystal inclusions are
poled along Z || [001] || OX
3
, and the ceramic matrix is characterised by the
remanent polarisation vector being also parallel to the OX
3
axis. To avoid
confusion between the volume fractions of the inclusions in different composite
groups, hereafter we denote the volume fraction of the single-crystal inclusions
with m
cr
. The volume fraction of the ceramic matrix surrounding these inclusions
equals 1 m
cr
.
The effective properties of the 03 single crystal / ceramic composite [32]
depend on the volume fraction m
cr
and the aspect ratio U = a
1
/ a
3
of the inclusions
and are determined by means of the EFM (see Equations 3.19 and 3.20). Below we
show how the electromechanical constants of the piezo-active components promote
a non-monotonic volume-fraction dependence of the effective piezoelectric
coefficients and , changes in sgn and sgn , along with an increase in
the hydrostatic piezoelectric coefficients
*
3 j
e
*
3 j
d
*
31
e
*
31
d
= + 2 and = + 2 (5.1)
*
h
e
*
33
e
*
31
e
*
h
d
*
33
d
*
31
d
To find a composite with optimum parameters from Equations 5.1, different
combinations of the PMNxPT or PZNyPT single crystals (see Table 2.2) and
PbTiO
3
-type ceramics (see Table 2.1) were analysed on varying the volume
fraction m
cr
from 0 to 0.5 and the aspect ratio U from 0.1 to 10. An example of non-
monotonic behaviour of the effective parameters of the 03 composite (Figure 5.1)
0,1 0,2 0,3 0,4 0,5
-1
0
1
2
3
4
5
12
11
10 9
8
7
6
5
4
3
2
1
e
*
3
1
m
cr
1: U = 0.1
...
5: U = 0.5
...
10: U = 1
11: U = 5
12: U =10
a
Figure 5.1. Volume-fraction dependences of effective piezoelectric coefficients (in C /
m
2
) and (in pC / N) and normalised hydrostatic piezoelectric coefficients / =
( + 2 ) / ( + 2 ) and / = ( + 2 ) / ( + 2
)
), which have
been calculated for the PZN0.08PT single crystal / (Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
(I)
ceramic composite with 03 connectivity. In graphs (a), (b), (e), and (g), curves 1, 2, 3, 4, 5,
6, 7, 8, 9, 10, 11, and 12 are related to the aspect ratios U = 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7,
0.8, 0.9, 1, 5, and 10, respectively. In graph (c), curves 1, 2, and 3 are related to U = 1, 5, and
10, respectively. In graphs (d) and (i), curves 1, 2, 3, 4, 5, and 6 are related to U = 0.1, 0.2,
0.3, 0.4, 0.5, and 1, respectively. In graphs (f) and (h), curves 1, 2, 3, and 4 are related to U =
0.9, 1, 5, and 10, respectively (reprinted from Topolov and Kamlah [32], with permission
from IOP Publishing)
*
3 j
e
*
3 j
d
*
h
e
) (FC
h
e
*
33
e
*
31
e
) 2 (
33
e
) 2 (
31
e
*
h
d
) (FC
h
d
*
33
d
*
31
d
) 2 (
33
d
2 (
31
d
0,1 0,2 0,3 0,4 0,5
5,5
6,0
6,5
7,0
7,5
8,0
8,5
9,0
5
6
7
3
4
2
1
8
e
*
3
3
m
cr

1: U = 0.1
2: U = 0.2
3: U = 0.3
4: U = 0.4
5: U = 0.5
6: U = 0.6
7: U = 0.7
8: U = 0.8
b

0,1 0,2 0,3 0,4 0,5
6,5
7,0
7,5
8,0
8,5
9,0
3
2
1
e
*
3
3
m
cr

1: U = 1
2: U = 5
3: U = 10
c
is considered [32] when the PZN0.08PT single-crystal inclusions are surrounded
by poled (Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
(I) ceramic. This ceramic is characterised
[29] by a large anisotropy of the piezoelectric coefficients (i.e., / | 14.2,
see Table 2.1) and (i.e.,

/ |

11.6).
1

It turns out that the large
) 2 (
3 j
e
) 2 (
33
e
) 2 (
31
e
) 2 (
3 j
d
) 2 (
33
d
) 2 (
31
d

____________
1
Following notations used for the 03 composite in Section 3.3, we denote the
electromechanical constants of the inclusion and the matrix with superscripts (1) and
(2), respectively.

0,1 0,2 0,3 0,4 0,5
0,8
1,0
1,2
1,4
1,6
1,8
2,0
2,2
2,4
6
5
4
3
2
1
e
*
h

/

e
(
F
C
)
h
m
cr

1: U = 0.1
2: U = 0.2
3: U = 0.3
4: U = 0.4
5: U = 0.5
6: U = 1
d

0,1 0,2 0,3 0,4 0,5
-50
-40
-30
-20
-10
0
10
20
8
7
6
5
4
3
2
1
d
*
3
1
m
cr

1: U = 0.1
2: U = 0.2
3: U = 0.3
4: U = 0.4
5: U = 0.5
6: U = 0.6
7: U = 0.7
8: U = 0.8
e
anisotropy of the piezoelectric properties predetermines a non-monotonic volume-
fraction behaviour of both groups of effective piezoelectric coefficients,
(Figures 5.1a and b) and (Figures 5.1e and g). The piezoelectric coefficient
(m
cr
) increases considerably at U d 0.2 (Figure 5.1a), whereas the piezoelectric
coefficient (m
cr
) undergoes a small increase in a restricted range of m
cr
(Figure
5.1b). It is seen that the more distinct changes in the and values take place
at 0.1 d U < 0.5 (curves 15 in Figures 5.1a and b), as the shape of the inclusion is
transformed from a needle (U << 1) to a sphere (U = 1). In this case, the electric
*
3 j
e
*
3 j
d
*
31
e
*
33
e
*
3 j
e
*
3 j
d
0,1 0,2 0,3 0,4 0,5
-50
-40
-30
-20
-10
0
4
3
2
1
d
*
3
1
m
cr

1: U = 0.9
2: U = 1
3: U = 5
4: U = 10
f

0,1 0,2 0,3 0,4 0
30
40
50
60
70
80
90
100
110
120
130
,5
8
7
6
5
4
3
2
1
d
*
3
3
m
cr

1: U = 0.1
2: U = 0.2
3: U = 0.3
4: U = 0.4
5: U = 0.5
6: U = 0.6
7: U = 0.7
8: U = 0.8
g
depolarisation field of the inclusion becomes more intensive and, simultaneously, a
certain relief of the internal mechanical stress occurs. Increasing the aspect ratio U
at m
cr
= const results in moderate changes in the
e
values (Figures 5.1ac).
*
3 j
The studied 03 composite exhibits a unique (U) dependence at m
cr
= const:
for this material the changes monotonically, whereas changes non-
monotonically (compare graphs in Figures 5.1a and c). This behaviour of the
piezoelectric coefficients of 03 composites was initially reported in ref. [32]. The
unusual increase in
e
and decrease in
e
at small m
cr
values and 0 < U d 0.2 (see
curves 1 and 2 in Figures 5.1a and b) is explained by an electromechanical
*
3 j
e
*
31
e
*
33
e
*
31
*
33
0,1 0,2 0,3 0,4 0,5
50
60
70
80
90
100
110
120
130
4
3
2
1
d
*
3
3
m
cr

1: U = 0.9
2: U = 1
3: U = 5
4: U = 10
h

0,1 0,2 0,3 0,4
0,9
1,0
1,1
1,2
1,3
1,4
1,5
1,6
1,7
1,8
0,5
6
5
4
3
2
1
d
*
h

/

d
(
F
C
)
h
m
cr

1: U = 0.1
2: U = 0.2
3: U = 0.3
4: U = 0.4
5: U = 0.5
6: U = 1
i
interaction between the highly piezo-active and highly anisotropic components.
Based on condition 2.13, we can represent the piezoelectric coefficients of the
studied composite as follows:
*
3 j
e
= (
c
+
c
) +
d
and
e
= 2
d
+
d
(5.2)
*
31
e
*
31
d
E *
11
*
12
E *
33
E
c
*
13
*
33
*
31
E
c
*
13
*
33
E
c
*
33
In Equations 5.2 the first terms are small due to condition
d
>> _ _, which is
directly associated with the inequality >> | | being correct for the modified
*
33
*
31
d
) 2 (
33
d
) 2 (
31
d
PbTiO
3
ceramic [29]. The term rises with increasing m
cr
, since a condition
> holds, and a slight decrease in (see curves 1 and 2 in Figure 5.1g)
is compensated due to the ratio / > 4 (see Tables 2.1 and 2.2). In contrast
to these relations, the term from Equations 5.2 decreases on increasing m
cr
due to a decrease in and an inequality < . A further increase of the
aspect ratio U and the volume fraction m
cr
of inclusions changes the balance of the
above-mentioned terms in Equations 5.2: as a consequence, a change in sgn (see
curves 812 in Figure 5.1a) is observed and the shape of the curves changes
(Figures 5.1a and b). As with the piezoelectric coefficients (m
cr
) (Figures 5.1e
and g), the volume-fraction dependences (m
cr
) are also non-monotonic and a
change in sgn takes place at 0.1 d U d 0.2 (Figures 5.1a and b). It is known that
prolate-shaped inclusions cause a slight electric depolarisation field and an
intensive mechanical stress field, and this combination of the internal fields
promotes the non-monotonic volume-fraction dependences of various piezoelectric
coefficients.
*
33
d
E
c
*
13
E
c
), 1 (
13
E
c
), 2 (
13
*
33
d
E
c
), 1 (
13
E
c
), 2 (
13
*
33
d
E
c
*
33
*
33
d
E
c
), 1 (
33
E
c
), 2 (
13
*
31
e
*
3 j
e
*
3 j
d
*
3 j
e
*
31
e
It is of practical interest to study how the electromechanical interaction of the
highly piezo-active and highly anisotropic components influences the hydrostatic
response of the 03 composite. The next example is concerned with the effective
hydrostatic piezoelectric coefficients from Equations 5.1. For a better comparison
with the effective properties, the normalised hydrostatic piezoelectric coefficients
/

and /

(Figures 5.1d and i) are calculated, where hydrostatic
parameters of the ceramic component are = + 2 = (0) and =
+ 2 = (0). As follows from data shown in Figures 5.1a, b, e, and g, the
prolate inclusions with 0.1 d U d 0.2 lead to a considerable increase in both
(m
cr
) and (m
cr
) as well as a small decrease in both (m
cr
) and (m
cr
) at
m
cr
< 0.2. This volume-fraction dependence promotes a distinct enhancement of the
effective hydrostatic piezoelectric properties (see curves 1 and 2 in Figures 5.1d
and i) as compared to those of the ceramic component.
*
h
e
) (FC
h
e
*
h
d
) (FC
h
d
) (FC
h
e
) 2 (
33
e
) 2 (
31
e
*
h
e
) (FC
h
d
) 2 (
33
d
) 2 (
31
d
*
h
d
*
31
e
*
31
d
*
33
e
*
33
d
In the vicinity of max (m
cr
) at U = 0.1, the normalised hydrostatic
piezoelectric coefficients are characterised by a relation / | / |
2. This result can be compared with data on the hydrostatic piezoelectric
coefficient
*
calculated for the 13 PZT-4 ceramic / epoxy composite [33]: the
maximum value of / for this 13 composite does not exceed 2.7 and is
attained at volume fractions of the piezoelectric ceramic rods in a range of 0.8 < m
< 0.9. For purely geometrical reasons, however, values of m > S / 4 for the volume
fractions of cylindrical rods in a square arrangement are impossible. As the volume
fraction m approaches a realistic value of 0.5, / decreases and becomes
*
h
d
*
h
e
) (FC
h
e
*
h
d
) (FC
h
d
h
e
*
h
e
) (FC
h
e
*
h
e
) (FC
h
e
smaller than that calculated for the studied 03 PZN0.08PT-based composite with
0.1 d U d 0.2 (see curves 1 and 2 in Figure 5.1d).
Thus, despite the fairly small value of the hydrostatic piezoelectric coefficient
of the PZN0.08PT single crystal ( + 2
)
| | |, see data from Table 2.2), this
component can play an important role in forming the remarkable hydrostatic
piezoelectric properties in the 03 single crystal / ceramic composite. In the next
section we analyse the role of this material in forming the piezoelectric response in
the 13-type composite.
) 1 (
33
e
1 (
31
e
) 1 (
31
e
5.3 013 Single Crystal / Ceramic / Polymer Composites
In this section we consider 13-type composites that consist of piezo-active 03
single crystal / ceramic rods and a piezo-passive polymer matrix (Figure 5.2). The
cylindrical extended rods are uniformly distributed in the matrix so that a square
arrangement in the (X
1
OX
2
) plane takes place. The volume fraction of the rods and
the surrounding matrix are m and 1 m, respectively, whereas the single-crystal
inclusions in each rod are characterised by a volume fraction m
cr
. It is assumed that
the largest semi-axis a
3
of each inclusion is considerably less than the radius of the
rod and these inclusions are regularly distributed in the rod, i.e., centres of
symmetry of the inclusions are located in the sites of a simple tetragonal lattice.
The main crystallographic axes x, y and z in each single-crystal inclusion are
parallel to the co-ordinate axes OX
1
, OX
2
and OX
3
, respectively. The composite as
a whole has a remanent polarisation vector P
r
|| OX
3
and is described by 013
connectivity.
The electromechanical properties of the rod are determined by means of the
EFM, as described for the 03 composite (see Sections 3.3 and 5.2), and

Figure 5.2. Diagram of the 013 composite containing single-crystal (n = 1), ceramic (n =
2) and polymer (n = 3) components. (X
1
X
2
X
3
) is the rectangular co-ordinate system, m and
m
cr
denote the volume fractions of rods and spheroidal inclusions in them, respectively.
Directions of the spontaneous polarisation vector P
s
|| OX
3
in the single-crystal inclusion and
of the remanent polarisation vector P
R
|| OX
3
in the surrounding ceramic medium are
indicated by arrows
characterised by the elastic moduli , piezoelectric coefficients and dielectric
permittivities . The effective electromechanical properties )** in the 013
composite
2
are determined using the same EFM and the Eshelby tensor [34] for the
cylindrical inclusion. The elastic moduli (m, m
cr
, U), piezoelectric coefficients
(m, m
cr
, U) and dielectric permittivities (m, m
cr
, U) of this composite are
regarded as functions of the volume fractions m and m
cr
and the aspect ratio U.
E
ab
c
* *
ij
e
[
H
*
kp
E
ab
c
* *
* *
ij
e
[
H
* *
kp
The 013 PZN0.08PT / (Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
(I) ceramic / polymer
composite is an interesting example of a piezo-active material that is characterised
by a non-monotonic behaviour of the piezoelectric coefficients and changes in their
signs (Figure 5.3). Based on the results [32], two kinds of non-monotonic
behaviour can be described. First, both and non-monotonically depend on
the volume fraction m
cr
(see Figures 5.3a, d, g, and j) at m = const and U = 0.1.
Such a behaviour is observed due to the non-monotonic dependences of the
piezoelectric coefficients (m
cr
) and (m
cr
) of the related 03 single crystal /
ceramic composite (Figures 5.1a, b, e, and g). Second, the piezoelectric coefficient
of the 013 composite passes a minimum (Figures 5.3g, h and i) when
varying the volume fraction m of the rods at constant parameters m
cr
and U of the
single-crystal inclusions. A volume-fraction dependence similar to (m, m
cr
, U)
at m
cr
= const and U = const cannot be found among those determined for the more
conventional 13 Pb(Zr, Ti)O
3
-type ceramic / polymer composites [3537]. In
addition, the piezoelectric coefficient (m, m
cr
, U) of the studied 013
composite at m
cr
= const and U = const changes monotonically and attains
saturation at m | 0.2 (Figures 5.3jl). Such a configuration of the curves is
typical of the 13 ceramic / polymer composites [3436].
* *
3 j
e
* *
3 j
d
*
3 j
e
*
3 j
d
* *
31
d
* *
31
d
* *
33
d
* *
33
d
The unusual behaviour of (m, m
cr
, U) at m
cr
= const and U = const (Figures
5.3gi) may be explained [32] by taking into account the influence of the elastic
and piezoelectric anisotropy of the two piezo-active components on the effective
piezoelectric activity of the composite. It is important to underline, for example,
the significant differences in the / or / ratios where superscripts
n = 1 and 2 are related to the single-crystal and ceramic components, respectively.
Furthermore, an equality sgn

= sgn holds for the chosen single-crystal and
ceramic components. The shape of curves 13 in Figures 5.3gi depends on the
aspect ratio U of the inclusions in the rod. At U = 0.1, the piezoelectric coefficient
has a minimum and passes zero (Figure 5.3g). As m
cr
and U increase, a
saturation trend of (m, m
cr
, U) < 0 is observed (Figures 5.3g and h).
* *
31
d
E n
c
), (
11
E n
c
), (
13
E n
c
), (
33
E n
c
), (
13
) 1 (
31
e
) 2 (
31
e
* *
31
d
* *
31
d
It follows from the results in ref. [32] that the relaxor-ferroelectric component
has a significant influence on (m, 0.4, 1) (see curves 35 in Figures 5.3b and c).
* *
31
e
____________
2
In this section we actually consider a composite in a composite and use two asterisks to
denote the effective electromechanical properties of the studied 013 composite.
0,2 0,4 0,6 0,8 1,0
0
1
2
3
4
5
5
4
3
2
1
e
*
*
3
1
m

U = 0.1
1: m
cr
= 0.1
2: m
cr
= 0.2
3: m
cr
= 0.3
4: m
cr
= 0.4
5: m
cr
= 0.5
a
0,2 0,4 0,6 0,8 1,0
-0,5
-0,4
-0,3
-0,2
-0,1
0,0
0,1
0,2
0,3
0,4
5
4
3
2
1
e
*
*
3
1
m

U = 1
1: m
cr
= 0.1
2: m
cr
= 0.2
3: m
cr
= 0.3
4: m
cr
= 0.4
5: m
cr
= 0.5
b
Figure 5.3. Volume-fraction dependences of the effective piezoelectric coefficients (in
C / m
2
) and (in pC / N) that have been calculated for the 013 PZN0.08PT /
(Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
(I) ceramic / araldite composite. In all graphs, curves 1, 2, 3,
4, and 5 are related to volume fractions of the single-crystal inclusions m
cr
= 0.1, 0.2, 0.3,
0.4, and 0.5, respectively. The aspect ratios of these inclusions are given as follows: U = 0.1
(graphs (a), (d), (g), and (j)), U = 1 (graphs (b), (e), (h), and (k)), and U = 10 (graphs (c), (f),
(i), and (l)) (reprinted from Topolov and Kamlah [32], with permission from IOP
Publishing)
* *
3 j
e
* *
3 j
d
0,2 0,4 0,6 0,8 1,0
-0,8
-0,6
-0,4
-0,2
0,0
0,2
5
4
3
2
1
e
*
*
3
1
m

U = 10
1: m
cr
= 0.1
2: m
cr
= 0.2
3: m
cr
= 0.3
4: m
cr
= 0.4
5: m
cr
= 0.5
c
0,2 0,4 0,6 0,8 1,0
0
1
2
3
4
5
6
7
5
4
3
2
1
e
*
*
3
3
m

U = 0.1
1: m
cr
= 0.1
2: m
cr
= 0.2
3: m
cr
= 0.3
4: m
cr
= 0.4
5: m
cr
= 0.5
d
0,2 0,4 0,6 0,8 1,0
0
1
2
3
4
5
6
7
8
9
5
4
3
2
1
e
*
*
3
3
m

U = 1
1: m
cr
= 0.1
2: m
cr
= 0.2
3: m
cr
= 0.3
4: m
cr
= 0.4
5: m
cr
= 0.5
e
0,2 0,4 0,6 0,8
0
1
2
3
4
5
6
7
8
1,0
5
4
3
2
1
e
*
*
3
3
m

U = 10
1: m
cr
= 0.1
2: m
cr
= 0.2
3: m
cr
= 0.3
4: m
cr
= 0.4
5: m
cr
= 0.5
f
0,2 0,4 0,6 0,8 1,0
-10
-5
0
5
10
15
20
5
4
3
2
1
d
*
*
3
1
m

U = 0.1
1: m
cr
= 0.1
2: m
cr
= 0.2
3: m
cr
= 0.3
4: m
cr
= 0.4
5: m
cr
= 0.5
g
0,2 0,4 0,6 0,8 1,0
-50
-45
-40
-35
-30
-25
-20
-15
-10
-5
0
5
4
3
2
1
d
*
*
3
1
m

U = 1
1: m
cr
= 0.1 2: m
cr
= 0.2
3: m
cr
= 0.3 4: m
cr
= 0.4
5: m
cr
= 0.5
h
0,2 0,4 0,6 0,8 1,0
-50
-40
-30
-20
-10
0
5
4
3
2
1
d
*
*
3
1
m

U = 10
1: m
cr
= 0.1 2: m
cr
= 0.2
3: m
cr
= 0.3 4: m
cr
= 0.4
5: m
cr
= 0.5
i
0,2 0,4 0,6 0,8
0
10
20
30
40
1,0
5
4
3
2
1
d
*
*
3
3
m

U = 0.1
1: m
cr
= 0.1
2: m
cr
= 0.2
3: m
cr
= 0.3
4: m
cr
= 0.4
5: m
cr
= 0.5
j
0,2 0,4 0,6 0,8 1,0
0
10
20
30
40
50
60
70
80
90
100
110
120
130
5
4
3
2
1
d
*
*
3
3
m

U = 1
1: m
cr
= 0.1 2: m
cr
= 0.2
3: m
cr
= 0.3 4: m
cr
= 0.4
5: m
cr
= 0.5
k
0,2 0,4 0,6 0,8 1,0
0
10
20
30
40
50
60
70
80
90
100
110
5
4
3
2
1
d
*
*
3
3
m

U = 10
1: m
cr
= 0.1 2: m
cr
= 0.2
3: m
cr
= 0.3 4: m
cr
= 0.4
5: m
cr
= 0.5
l
At m
cr
t 0.4 the
* *
31
curves assume negative values at large volume fractions m.
The above-given examples of the volume-fraction dependence of the piezoelectric
coefficients (m, m
cr
, U) and (m, m
cr
, U) at m
cr
= const and U = const suggest
that the 013 composite exhibits a piezoelectric anisotropy that can be varied in a
wide range. To describe specific features of piezoelectric anisotropy, one can
introduce the following ratios: /
*
, / and / and we will consider
the PZN0.08PT / (Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
(I) ceramic / polymer composite
with prolate single-crystal inclusions as rods (U = 0.1). The criteria for a large
piezoelectric anisotropy of the composite are written in the following form [32]:

e
* *
3 j
e
* *
3 j
d
* *
33
e
*
31
e
* *
33
d
* *
31
d
* *
33
k
* *
31
k
|
*
/
*
| t 10 (5.3)
*
33
e
*
31
e
| / | t 10 (5.4)
* *
33
d
* *
31
d
| / | t 10 (5.5)
* *
33
k
* *
31
k
Table 5.1 contains data on the volume fractions m
cr
and m for which conditions
5.35.5 hold true. The volume fraction m
cr
of the single-crystal inclusions in the
rods is varied from 0.1 to 0.5. An intersection m
1,opt
d m d m
2,opt
is related to the
optimal volume-fraction ranges, where conditions 5.35.5 hold simultaneously.
The width of this intersection is defined as 'm = m
2,opt
m
1,opt
.
It is found that 'm strongly depends on the elastic properties of the polymer
component and on the volume fraction m
cr
of the single-crystal component (Table
5.1). The softer polymer component, elastomer, leads to a relatively wide range of
optimal volume fractions m
i,opt
: at 0.1 d m
cr
d 0.5 values 'm | 0.2 0.3 are
attained.
The relatively wide ranges of m obeying condition 5.4 (Table 5.1) illustrate the
significant influence of the rodsmatrix composite structure (Figure 5.2, right)
on the volume-fraction behaviour of the piezoelectric coefficients : this
behaviour is similar to that known for the 13 ferroelectric ceramic / polymer
composites [3537]. The wider ranges of m obeying condition 5.5 in comparison
with those related to condition 5.3 (see Table 5.1) are a result of the prolate shape
of the inclusions (Figure 5.2, left). Such inclusion shapes, along with differences in
the elastic moduli of the single-crystal and ceramic components, strongly influence
the elastic properties of the 03 composite (see Section 5.2). An important
consequence is that the piezoelectric anisotropy of the 013 composite depends
on the / ratio, where are the elastic compliances of the 03 single crystal
/ ceramic composite. Finally, the results given in Table 5.1 represent another
example of the non-monotonic volume-fraction behaviour of the effective
parameters. As is shown for the elastomer-containing composite, the calculated
dependence of the range width 'm on the volume fraction m
cr
first decreases with
increasing m
cr
and then increases again. This behaviour is a result of the unusual
* *
3 j
e
E
s
*
11
E
s
*
33
E
ab
s
*
Table 5.1. Ranges of the volume fraction m of rods (U = 0.1, Figure 5.2) in the 013 PZN
0.08PT / (Pb
0.9625
La
0.025
)(Ti
0.99
Mn
0.01
)O
3
(I) ceramic / polymer composite (reprinted from
Topolov and Kamlah [32], with permission from IOP Publishing)
Polymer Volume
fraction
m
cr
Ranges of the volume fraction m obeying conditions
5.3 | 5.4 | 5.5
Araldite 0.1
0.2
0.3
0.4
0.5
0.336 d m d 0.669
0.196 d m d 0.398
0.160 d m d 0.335
0.185 d m d 0.387
0.281 d m d 0.584
0 < m d 0.413
---
---
---
---
0.198 d m d 0.809
0.131 d m d 0.519
0.113 d m d 0.433
0.129 d m d 0.494
0.187 d m d 0.716
Polyur-
ethane
0.1
0.2
0.3
0.4
0.5
0.357 d m d 0.681
0.221 d m d 0.419
0.187 d m d 0.358
0.214 d m d 0.410
0.312 d m d 0.604
0 < m d 0.676
0 < m d 0.333
0 < m d 0.127
0 < m d 0.154
0 < m d 0.346
0.162 d m d 0.861
0.121 d m d 0.598
0.110 d m d 0.510
0.125 d m d 0.573
0.169 d m d 0.781
Elasto-
mer
0.1
0.2
0.3
0.4
0.5
0.461 d m d 0.744
0.313 d m d 0.501
0.276 d m d 0.441
0.310 d m d 0.496
0.422 d m d 0.680
0 < m d 0.918
0 < m d 0.832
0 < m d 0.780
0 < m d 0.787
0 < m d 0.836
0.058 d m d 0.968
0.057 d m d 0.858
0.058 d m d 0.800
0.064 d m d 0.840
0.075 d m d 0.939
Polymer Volume
fraction
m
cr
Range of the
optimal volume
fraction
m
1,opt
d m d m
2,opt
'm = m
2,opt
m
1,opt
Araldite 0.1
0.2
0.3
0.4
0.5
0.336 d m d 0.413
---
---
---
---
0.077
---
---
---
---
Polyur-
ethane
0.1
0.2
0.3
0.4
0.5
0.357 d m d 0.676
0.221 d m d 0.333
---
---
0.312 d m d 0.346
0.319
0.112
---
---
0.034
Elasto-
mer
0.1
0.2
0.3
0.4
0.5
0.461 d m d 0.744
0.313 d m d 0.501
0.274 d m d 0.441
0.310 d m d 0.496
0.422 d m d 0.680
0.283
0.188
0.167
0.186
0.268
volume-fraction dependences of the piezoelectric coefficients (m
cr
) and (m
cr
)
of the related 03 composite (Figures 5.1a, b, e, and g) and by the following
relations between the elastic moduli of the single-crystal (n = 1) and polymer (n =
3) components in the 103 composite: c / | / |

/ and
/
c
| 1 at = .
*
3 j
e
*
3 j
d
E ), 1 ( E
c
), 1 ( E
c
), 1 ( E
c
), 1 ( ) 3 (
c
) 3 (
c
E
c
), 1 ( E ), 1 ( ) 3 (
c
) 3 (
c
11 12 11 13 11 12
11 33 11 33
Thus, as follows from the results analysed in this section, the piezoelectric
properties in the 013 composites containing highly piezo-active (relaxor-
ferroelectric) and highly anisotropic (ferroelectric ceramic) PbTiO
3
-based
components have no analogues among the properties in any existing 13-type
piezo-composites. The important role of the relaxor-ferroelectric single-crystal
component as well as the anisotropy of both the elastic and piezoelectric properties
in the single-crystal and ceramic components has been emphasised. Studying the
effect of the piezo-passive polymer component on the piezoelectric anisotropy
factors from formulae 5.25.4 enables us to conclude that the softer polymer is
preferred for creating a highly anisotropic 013 composite.
Other representatives of the 13-type composites based on the [001]-poled
relaxor-ferroelectric single crystals and the effective electromechanical properties
will be considered in Sections 5.4 and 5.5.
5.4 13 Single Crystal / Polymer Composites
Examples of the application of the relaxor-ferroelectric single-crystal components
of the advanced 13 composites were considered in experimental papers [38, 39].
In both studies the dice-and-fill technique [40,

41] was used to manufacture the 13
single crystal / polymer composites with various volume fractions of the PZN
0.08PT [38] or PMN0.33PT [39] single crystals. Growth of high-quality PMN
xPT [42] and PZNyPT [43] single crystals and the formation of engineered-
domain structures [5, 6] opens up new possibilities for creating piezo-composites
based on these single crystals. With regards to 13 composites, it is of interest to
analyse a variety of possible single crystal / polymer composites. In this section we
consider several volume-fraction dependences of the effective parameters of the 1
3 PMNxPT / araldite and PZNyPT / araldite composites with single-crystal
compositions chosen near the MPB.
It is assumed that the 13 composite sample has been poled under the electric
field E || OX
3
and contains a system of single-crystal rods in the form of
rectangular parallelepipeds (Figure 5.4). These parallelepipeds are continuous in
the OX
3
direction and their lateral faces are parallel to the (X
1
OX
3
) and (X
2
OX
3
)
planes. The main crystallographic axes X, Y and Z of the single crystals are
oriented as follows: X || OX
1
, Y || OX
2
and Z || OX
3
. The single-crystal rods are
regularly distributed in the polymer matrix. A full set of effective
electromechanical constants of the 13 composite shown in Figure 5.4 is
determined using the matrix method [4446] and the 9 u 9 matrices from Equation
3.1. These matrices are now related to the single-crystal (n = 1) and polymer (n =
2) components.
Figure 5.4 shows the 13 structure with planar microgeometry used for the
determination of the effective properties based on averaging the aforementioned
electromechanical constants of the components and taking into account boundary
conditions at x
i
= const (i = 1 and 2). These boundary conditions imply [32, 45, 46]
a continuity of normal components of the mechanical stress V
iv
, tangential
components of the mechanical strain [
rt
, normal components of the electric
displacement D
i
, and tangential components of the electric field E
f
.

An averaging

Figure 5.4. Banno unit cell of the 13 single crystal / polymer composite. The spontaneous
polarisation vector of the single crystal (see the shaded parallelepiped) is denoted by an
arrow. (X
1
X
2
X
3
) is the rectangular co-ordinate system. m and 1 m are volume fractions of
the single-crystal and polymer components, respectively
procedure is implemented in a long-wave approximation that is valid when a
wavelength of an external acoustic field is much longer than the thickness of the
separate rod in the composite. The matrix || C* || of the effective electromechanical
properties in the 13 composite is expressed by Equation 3.2, where m is the
volume fraction of the single-crystal rods and || M || is the matrix related to the
boundary conditions at x
i
= const (i = 1 and 2). According to the results in refs.
[44, 45], || M || depends on elements of || C
(n)
||, and the structure of the || C* ||
matrix is similar to that shown in Equation 3.1. The matrix elements of || C* ||
represent volume-fraction dependences of the effective elastic compliances
(m), piezoelectric coefficients (m) and dielectric permittivities (m) of
the 13 composite. Based on these effective constants and Equations 2.12 2.15,
we evaluate the piezoelectric coefficients (m), (m) and (m) as well as other
effective parameters in a range of 0 < m < 1.
E
ab
s
* *
ij
d
V
H
*
pp
*
ij
e
*
ij
g
*
ij
h
We will now consider the maxima of different effective parameters of the 13
composites [47] based on the PMNxPT or PZNyPT single crystals where x =
0.30, 0.33, 0.42 and y = 0.045, 0.07, 0.08. Some results on maximum values are
shown in Table 5.2 and Figures 5.5 and 5.6 (curve 1 in all graphs). Among the
aforementioned single-crystal components, of particular interest are PMN0.42PT
and PMN0.33PT. In the 13 PMN0.42PT / araldite composite the largest
maximum values are attained for the following parameters [47]: max (m) =
19.3 , max (m) = 23.3 and max (m) = 1.08
)
, where the
piezoelectric coefficients , ,
)
, etc. are listed in Table 5.3. The
piezoelectric response of the 13 PMN0.42PT / araldite composite is closely
*
33
g
) 1 (
33
g
*
h
g
) 1 (
h
g
*
33
h
1 (
33
h
) 1 (
33
g
) 1 (
h
g
1 (
33
h
Table 5.2. Maximum values of effective parameters (in mV

m / N), (in 10
8
V / m),
(in pC / N), (in mV

m / N), , (in 10
12
Pa
1
), and (in 10
12
Pa
1
) that
have been calculated for the 13 relaxor-ferroelectric single crystal / araldite composites
(reprinted from Bezus et al. [47], with permission from IOP Publishing)

*
33
g
*
33
h
*
h
d
*
h
g
*
t
k
2 *
33
) (Q
2 *
) (
h
Q
Single-crystal
component
max
*
33
g
max
*
33
h
max
*
h
d
max
*
h
g
PMN0.30PT 382
(0.013)
a

26.1
(0.414 0.547)
244
(0.434)
103
(0.013)
PMN0.33PT 496
(0.017)
37.8
(0.501 0.532)
274
(0.509)
130
(0.016)
PMN0.42PT 519
(0.024)
56.9
(0.653 0.709)

b
140
(0.024)
PZN0.045PT 317
(0.021)
19.5
(0.311 0.389)
180
(0.438)
81.0
(0.019)
PZN0.07PT 359
(0.023)
23.1
(0.419 0.480)
194
(0.455)
93.9
(0.021)
PZN0.08PT 352
(0.019)
22.3
(0.334 0.390)
206
(0.409)
90.6
(0.018)
Single-crystal
component
max
k
*
t
2 *
33
) Q
2 *
)
h
Q
max[
(
] max[
(
]
PMN0.30PT 0.819
(0.552)
106
(0.160)
6.16
(0.099)
PMN0.33PT 0.861
(0.559)
144
(0.250)
7.45
(0.115)
PMN0.42PT 0.729
(0.640)
46.3
(0.078)
3.46
(0.082)
PZN0.045PT 0.720
(0.532)
108
(0.338)
4.62
(0.141)
PZN0.07PT 0.750
(0.537)
133
(0.460)
5.43
(0.146)
PZN0.08PT 0.765
(0.422)
136
(0.422)
5.38
(0.131)
Notes to Table 5.2
a. The optimal volume fractions m
opt
of the single-crystal component that correspond to the
maximum values (i.e., max)* = )*(m
opt
)) are written in parentheses. In the column
ranges of the volume fractions obey a condition max (m) 10
7
V / m

*
V
H
), 1 (
H
V
H
), 1 (
H
*
g
*
g
33
h
*
33
h
*
33
h
b. For this composite, the monotonic dependence of the hydrostatic piezoelectric coefficient
(m) has been established
*
h
d
associated with the relatively low dielectric permittivity of the PMN0.42PT
single crystal: according to experimental data [48], / = 660, i.e., one order
of magnitude smaller than / of other PMNxPT single-crystals (Table 5.3)
with compositions across the MPB. Despite the fairly low volume fractions 0.01 <
m
opt
< 0.02 corresponding to max (m) and max (m) (see Table 5.2 and

curves
V
H
), 1 (
33
33 0
33 0
33 h
Figure 5.5. Volume-fraction dependences of the piezoelectric coefficients (a, in mV

m
/ N), (b, in mV

m / N) and (c, in 10
8
V / m) that have been calculated for the PMN
0.42PT single crystal / araldite (curve 1) and PMN0.42PT single crystal / porous araldite
(m
p
= 0.3) (curve 2) composites. The porous matrix contains spherical pores (reprinted from
Bezus et al. [47], with permission from IOP Publishing)
*
33
g
*
h
g
*
33
h
Figure 5.6. Volume-fraction dependences of the hydrostatic piezoelectric coefficient (a,
in pC / N), electromechanical coupling factor (b), squared figure of merit ( )
2
(c, in
10
15
Pa
1
), and squared hydrostatic figure of merit ( )
2
(d, in 10
15
Pa
1
) that have been
calculated for the PMN0.33PT single crystal / araldite (curve 1) and PMN0.33PT single
crystal / porous araldite (m
p
= 0.3, spherical pores) (curve 2) composites (reprinted from
Bezus et al. [47], with permission from IOP Publishing)

*
h
d
*
t
k
*
33
Q
*
h
Q



Table 5.3. Experimental values of the room-temperature electromechanical constants of the
relaxor-ferroelectric single crystals (reprinted from Bezus et al. [47], with permission from
IOP Publishing)
Single crystal
) 1 (
33
d
,
pC / N
) 1 (
33
g
,
mVm / N
) 1 (
33
h
,
10
8
V / m
) 1 (
h
d
,
pC / N
) 1 (
h
g
,
mVm / N
PMN0.30PT [48] 1981 28.7 24.6 139 2.1
PMN0.33PT [14] 2820 38.8 33.7 160 2.0
PMN0.42PT [48] 260 44.5 53.2 78 13.4
PZN0.045PT [49] 2000 44.0 17.0 60 2.0
PZN0.07PT [15] 2455 49.3 20.7 47 0.9
PZN0.08PT [38] 2890 42.4 17.7 20 0.2
Single crystal
V
H
), 1 (
33
/
0
) 1 (
t
k
( )
2
,
) 1 (
33
Q
10
12
Pa
1
( )
2
,
) 1 (
h
Q
10
12
Pa
1
PMN0.30PT [48] 7800 0.62 56.9 0.292
PMN0.33PT [14] 8200 0.64 109 0.320
PMN0.42PT [48] 660 0.62 11.6 1.05
PZN0.045PT [49] 5200 0.50 88.0 2.64
PZN0.07PT [15] 5622 0.47 121 0.423
PZN0.08PT [38] 7700 0.45 123 0.004

1 in Figures 5.5a and b) as well as an inequality (m
opt
) << , the PMN
xPT-based composites (x = 0.33 or 0.42) are of practical interest at volume
fractions m = 0.050.10. In this range the ratios (m) / | 813 and (m) /
| 1015 are attained. In contrast to (m), the condition (m) | const holds
for 0.1 < m < 0.9 (curve 1 in Figure 5.5c). This constancy is associated with the
boundary conditions for the electric-field and mechanical-stress components and,
as a consequence, a relation
e
(m) /
)
e
| (m) / holds true for all the 13
PMNxPT / polymer or PZNyPT / polymer composites studied in ref. [47].
V
H
*
33
V
H
), 1 (
33
*
g
) 1 (
g
*
g
) 1 (
g
*
g
*
h
* 1 ( [
H
* [
H
), 1 (
33 33 h
h 33 33
33 33 33 33
The 13 PZN0.08PT / araldite composite is characterised by the large
max[ value (Table 5.2) that is primarily the result of the large and
values of the single-crystal component (see Table 5.3). The optimal volume
fraction m
opt
corresponding to max[ is nearly equal to that found for
max (m), but this closeness is not typical of the other 13 composites listed in
Table 5.2. In addition, the PZN0.08PT / araldite composite does not possess large
values of max (m), max (m) and max (m), but the ratio max[ /
max[ | 25 is the highest value for the composites listed in Table 5.2. Such
features of the composite are associated with the elastic and piezoelectric
properties of the PZN0.08PT single crystal.
2 *
33
)] (m Q
) 1 (
33
d
V
H
), 1 (
33
2 *
33
)] (m Q
*
h
d
*
33
g
*
h
d
*
h
g
2 *
33
)] (m Q
2 *
)] (m Q
h
Maximum values of the effective parameters of the PZNyPT-based
composites are located in relatively narrow ranges (see Table 5.2). Such behaviour
is a result of the monotonic dependences [47] of the electromechanical properties
of the PZNyPT single crystals [15, 38, 49] at 0.045 d y d 0.08. As is seen from
Table 5.2, the high values of max (m) = 1.71 , max (m) = 1.25 ,
max[ = 1.31( )
2
, and max[ = 21.1( )
2
are attained in the
PMN0.33PT-based 13 composite. The PMN0.33PT single crystal is
characterised [14] by the largest values of the elastic compliances | |,
piezoelectric coefficients | | and dielectric permittivity in the range of
molar concentrations 0.30 x 0.42. These outstanding properties strongly
influence the effective electromechanical properties and their maxima in the related
composite.
*
h
d
) 1 (
h
d
*
t
k
) 1 (
t
k
2 *
33
)] (m Q
) 1 (
33
Q
2 *
)] (m Q
h
) 1 (
h
Q
E
a
s
), 1 (
3
) 1 (
3 j
d
V
H
), 1 (
33
A further comparison of the calculated [47] and experimental [39] parameters
of the PMN0.33PT-based composite (Table 5.4) suggests that the good agreement
between these parameters is attained in a wide m range. Some differences between
the calculated and measured values in Table 5.4 can be associated with the finite
sizes of the single-crystal rods [50] and their piezoelectric and dielectric properties.
As is known from data in ref. [39], the measured and
V
values of the PMN
0.33PT single crystal are less than those determined in experimental work [14] and
used in calculations [47].
) 1 (
33
d H
), 1 (
33
Thus, the results represented in this section show how the PMNxPT and PZN
yPT single crystals actively influence the effective electromechanical properties in
Table 5.4. Calculated and experimental (in parentheses, from ref. [39]) values of normalised
parameters
a
of the 13 PMN0.33PT single crystal / epoxy composite (reprinted from Bezus
et al. [47], with permission from IOP Publishing)
m = 0.37 m = 0.48 m = 0.62 m = 0.77
*
t
k
/
) 1 (
t
k
1.32 (1.30) 1.34 (1.36) 1.35 (1.38) 1.31 (1.34)
D
c
*
33
/
D
c
), 1 (
33
0.295 (0.300) 0.371 (0.332) 0.474 (0.431) 0.605 (0.598)
Note to Table 5.4
a. The / and / ratios have been calculated using the room-temperature
electromechanical constants of the PMN0.33PT single crystal [14] and epoxy [39]

*
t
k
) 1 (
t
k
D
c
*
33
D
c
), 1 (
33
13 single crystal / polymer composites. A further increase in specific effective
parameters can be attained by use of a porous structure of the matrix surrounding
the single-crystal rods. Different examples of high-performance 13-type
composites comprising ferroelectric ceramic rods and the porous polymer matrix
were considered in refs. [5153]. In the next section we consider how the porous
polymer matrix would improve the piezoelectric sensitivity of the 103
composites based on the relaxor-ferroelectric single crystals.
5.5 103 Single Crystal / Porous Polymer Composites
In this section we study a 103 composite that contains single-crystal rods in the
form of rectangular parallelepipeds (Figure 5.7). These rods are regularly
distributed in a lengthy porous polymer matrix, wherein the shape of closed air
pores is described by an equation (x
1
/ a
1
)
2
+ (x
2
/ a
1
)
2
+ (x
3
/ a
3
)
2
= 1 relative to the
axes of the rectangular co-ordinate system (X
1
X
2
X
3
), where either an equality a
1
=
a
3
or an inequality a
1
> a
3
holds for the spherical or oblate spheroidal pores,
respectively. The centres of these pores are assumed to occupy the sites of a simple
cubic (at a
1
= a
3
) or tetragonal (at a
1
> a
3
) lattice. The semi-axis a
1
is considerably
less than the length of the side of the square being the intersection of the rod by the
(X
1
OX
2
) plane (Figure 5.7). In this connection one can regard the porous matrix as
a medium with effective properties [52] that depend on porosity m
p
and the aspect
ratio U = a
1
/ a
3
of the pores. The single-crystal component is characterised by the
volume fraction m, and the volume fraction of the porous matrix is 1 m. The
composite sample is poled by an electric field E || OX
3
. Based on results in ref.
[54], the poling procedure can be carried out after inserting the aligned rods into
the matrix and covering the composite sample with electrodes parallel to the
(X
1
OX
2
) plane. Below, we consider two examples of volume-fraction dependences
of the effective electromechanical properties in the 103 composite.
The first example is concerned with a composite containing spherical pores
distributed regularly in the matrix. The effective parameters )* = )*(m, m
p
) of
such a composite based on PMNxPT (x = 0.33 or 0.42) are represented in Figures
5.5 and 5.6 (see curves 2 in all the graphs), and the benefits of using a porous
matrix are clearly demonstrated. Using a porous matrix (0 < m
p
d 0.3)

rather than a

Figure 5.7. Schematic diagram of the 13-type composite. The spontaneous polarisation
vector of the single crystal is denoted by an arrow. (X
1
X
2
X
3
) is the rectangular co-ordinate
system. A spheroidal air pore is shown in the inset (reprinted from Bezus et al. [47], with
permission from IOP Publishing)
monolithic one (m
p
= 0) results in only a modest change in (m, m
p
) (compare
curves 1 and 2 in Figure 5.5c). This independence on m
p
is closely connected with
interrelations (m, m
p
) / | (m, m
p
) / and << that remain
true in a wide m range. The 103 PMN0.33PT / porous araldite (m
p
= 0.3)
composite is of interest because of the following maximum values [47] of the
effective parameters: max (m, 0.3) = 1.32 , max (m, 0.3) =

2.66 ,
max[ =

2.19( )
2
,

and max[ = 72.6( )
2
. These
effective parameters produce new possibilities for creating advanced piezo-active
composites containing the relaxor-ferroelectric components. The maximum values
of the above-mentioned parameters are higher than those measured on 13 piezo-
composites containing ceramic or single-crystal rods. For example, the
experimental values of max are 0.83, 0.80 and 0.70 for the PZN0.08PT single
crystal / polymer [38], PMN0.33PT single crystal / epoxy [39] and PZT-7A
ceramic / polymer [40] composites, respectively. The experimental max value
in the PZT-7A ceramic / Spurr epoxy composite [40] does not exceed 120 pC / N.
*
33
h
*
33
e
) 1 (
33
e
[
H
*
33
[
H
), 1 (
33
[
H
*
33
[
H
), 1 (
33
*
t
k
) 1 (
t
k
*
h
d
) 1 (
h
d
2 *
33
)] 3 . 0 , (m Q
) 1 (
33
Q
2 *
)] 3 . 0 , (m Q
h
) 1 (
h
Q
*
t
k
*
h
d
An increase in polymer porosity m
p
leads to an increase in , , ,
( )
2
, and ( )
2
in a wide m range as compared with the case of m
p
= 0 (see
curves 1 and 2 in Figures 5.5a, b, and 5.6a, c and d). Increasing (m, m
p
) in
comparison with (m, 0) (see Figure 5.6a and Table 5.5) correlates with changes
in the c
11,m
/ c
12,m
ratio of the elastic moduli of the porous matrix. The piezoelectric
*
33
g
*
h
g
*
h
d
*
33
Q
*
h
Q
*
h
d
*
h
d
coefficient (m, m
p
) of the studied 103 composite increases faster than | (m,
m
p
) | on increasing the volume fraction m. This behaviour stems from the fact that a
piezoelectric interaction between the rods in the (X
1
OX
2
) plane is reduced to a
greater degree by a porous matrix than by a monolithic one. The balance of in
this composite favours an increase in max (m, m
p
) at m
p
= const, as follows from
Table 5.5. A change in the balance between (m, m
p
) and (m, m
p
) is due to
changes in c
11,m
/ c
12,m
and the elastic stiffness of the matrix. In particular, as is seen
from Table 5.5, increasing c
11,m
results in the highest volume fractions m
opt

corresponding to max , and such a shift takes place for both the 13 and 103
composites. As follows from data on the 103 composites (Table 5.5), the higher
m
opt
value corresponds to a lower c
11,m
/ c
12,m
ratio. The correlation between max
and c
11,m
/ c
12,m
[47] allows prediction of the hydrostatic piezoelectric activity of the
103 composites based on relaxor-ferroelectric single crystals.
*
33
d
*
31
d
*
3 j
d
*
h
d
*
33
d
*
31
d
*
h
d
*
h
d
The second example is concerned with a similar PMN0.33PT-based
composite containing spheroidal pores distributed regularly in the polymer matrix.
The volume fraction of the single-crystal rods m, porosity m
p
and the aspect ratio U
of the pores are now regarded as three independent arguments in a function )*(m,
m
p
, U) of the effective parameters of the 103 composite. The 103 PMN
0.33PT / porous araldite composite at m
p
= 0.3 is characterised by the following
features of the piezoelectric response [47].
Even at relatively small pore aspect ratios (2 d U d 5), points of max (m, 0.3,
U) and max (m, 0.3, U) appear at 0.005 < m < 0.01. The microgeometry of such a
porous matrix displaces the maximum points of [ (m, 0.3, U)]
2
and [ (m, 0.3,
U)]
2
to lower m values: e.g., for ( )
2
from 0.088 at U = 2 to 0.060 at U = 5. The
*
33
g
*
h
g
*
33
Q
*
h
Q
*
33
Q
Table 5.5. Values of max (in pC / N), m
opt
a
(at m
p
= const)
b
and c
11,m
/ c
12,m
that have
been calculated for the 13-type composites based on PMN0.33PT single crystals
(reprinted from Bezus et al. [47], with permission from IOP Publishing)
*
h
d
Components
max
*
h
d
m
opt
c
11, m
/ c
12, m
Polyethylene (m
p
= 0)
Elastomer (m
p
= 0)
Spurr epoxy (m
p
= 0)
Araldite (m
p
= 0)
Porous araldite (m
p
= 0.10)
Porous araldite (m
p
= 0.15)
Porous araldite (m
p
= 0.20)
Porous araldite (m
p
= 0.25)
Porous araldite (m
p
= 0.30)
174
225
266
274
323
348
373
399
426
0.497
0.256
0.519
0.509
0.456
0.434
0.414
0.395
0.376
1.17
1.22
1.75
1.77
1.92
1.99
2.06
2.12
2.18
Notes to Table 5.5
a. From equation (m
opt
, m
p
) = max
*
h
d
*
h
d
b.

m
p
is the volume fraction of the spherical pores distributed regularly in the polymer matrix

Table 5.6. Effective piezoelectric coefficients (in pC / N), , (in mV

m / N),
squared figures of merit (
Q
)
2
, (
Q
)
2
(in 10
12
Pa
1
), and electromechanical coupling
factors and that have been calculated for the 103 PMN0.33PT single crystal /
porous araldite composite at m
p
= 0.30 (reprinted from Bezus et al. [47], with permission
from IOP Publishing)

*
h
d
*
33
g
*
h
g
* *
*
k
*
k
33 h
33 t

a
m
d,hyd
b

*
h
d

*
33
g

*
h
g

10 0.252 998 149 74.1
100 0.120 1780 317 252
200 0.096 1920 399 335
300 0.083 1980 459 396
400 0.077 2010 495 432
500 0.073 2040 522 459

a

( )
2

*
33
Q
( )
2

*
h
Q
*
33
k

*
t
k

10 300 73.9 0.941 0.920
100 710 449 0.946 0.940
200 910 644 0.946 0.942
300 1050 783 0.946 0.942
400 1140 869 0.947 0.943
500 1210 934 0.947 0.943
Notes to Table 5.6
a. The aspect ratio of the spheroidal pores in the polymer matrix

b
The volume fraction m
d,hyd
of the single-crystal rods obeys condition max (m, 0.3, ) =
(m
d,hyd
, 0.3, ) at the fixed value
*
h
d
*
h
d
maximum points of (m, 0.3, U) also shift to lower m values, but this
displacement to lower m becomes appreciable in the wide range of U. These results
suggest that for piezotechnical applications one can propose the 103 composites
that possess the maximum hydrostatic piezoelectric activity in addition to
considerable piezoelectric sensitivity and electromechanical coupling. Taking the
aspect ratio U from a wide range, a series of effective parameters that are expected
at max (m, 0.3, U) (Table 5.6) have been calculated. It can be seen that an
increase in the U value results in an increase in , , , ( )
2
, and ( )
2
at
almost constant and large values of electromechanical coupling factors and
k
.
The effect of the aspect ratio on the above-given parameters is mainly associated
with a change in the c
11,m
/ c
33,m
ratio of the elastic moduli of the porous matrix.
Increasing the value corresponds to softening of the polymer matrix in the OX
3

direction, and this softening enables the control of the piezoelectric response of the
103 composite at m
p
= const. It should be added that the values of max (m,
0.3, ) from Table 5.6 are larger than those evaluated recently [55] for the 0022
composites containing one or two relaxor-ferroelectric components (for example,
PMN0.33PT and PZN0.045PT) and two monolithic polymer components. This
*
h
d
*
d
*
g
*
g
*
d
*
Q
*
Q
*
k
*
*
d
h
33 h h 33 h
33 t
h
advantage, especially at 5 < < 10, is associated with the important role of the
porous matrix with the variable c
11,m
/ c
33,m
ratio.
It is of interest to consider the possible mechanical properties of the porous
polymers examined in the work. The porosity of the polymer matrix was varied
[47] in the range 0 < m
p
< 0.3. With regard to the typical mechanical properties of
the porous polymer with m
p
= 0.20.3, it is observed that the typical decrease in
stiffness and strength of such materials is about one order of magnitude [56, 57],
relative to the pore-free material. For a sensor material consisting of a porous
polymer with a high stiffness piezo-active material (e.g., the single-crystal rod in
the studied 13-type composites, see Figures 5.4 and 5.7), most of any applied
mechanical load will be experienced by the stiffer piezo-active component. As a
result, the polymer strength is unlikely to be a major limitation. Similarly, for an
actuator material consisting of a porous polymer and piezoelectric ceramic, the
relatively small piezoelectric strain is unlikely to exceed the failure strain of
polymer in the matrix.
5.6 High Performance in Wide Ranges
In this chapter we have considered the effective electromechanical properties and
related parameters of novel piezo-composites based on single crystals of relaxor-
ferroelectric solid solutions, such as PMNxPT or PZNyPT, with compositions
chosen near the MPB. A combination of the single-crystal and ceramic or single-
crystal and polymer components in these composites with different microgeometry
and connectivity provides a series of examples of non-monotonic volume-fraction
behaviour of all four types of the piezoelectric coefficients (see Figures 5.1, 5.3
and 5.5). Moreover, the piezoelectric properties, electromechanical coupling
factors and figures of merit of these composites clearly indicate their advantages
over conventional 03 or 13 ceramic-based composites. In general, the results
discussed in this chapter can aid in the design and selection of single-crystal and
ceramic components as well as a search for proper microgeometry of the porous
matrix.
It is believed that the development of high-quality and high-performance
composites based on the relaxor-ferroelectric single crystals opens up new
possibilities of controlling the effective parameters in wide volume-fraction ranges
(see, for example, data from Tables 5.1, 5.2, 5.4 5.6). Particularly high values of
the effective piezoelectric coefficients , and , electromechanical coupling
factor and squared figures of merit ( )
2
and ( )
2
in certain m ranges as well
as relatively large piezoelectric coefficients | | in the wide m range suggest that
the related composites are suitable for hydrophone, sensor, actuator, and transducer
applications. An electromechanical interaction between the relaxor-ferroelectric
single crystals and modified PbTiO
3
ceramics in the 03 composites also provides
an opportunity to vary the piezoelectric anisotropy over wide ranges.
*
33
g
*
h
g
*
h
d
*
t
k
*
33
Q
*
h
Q
*
3 j
d
References
[1] Park S-E, Hackenberger W, (2000) High performance single crystals piezoelectrics:
Applications and issues. Current Opinion in Solid State and Materials Science 6:11
18
[2] Cross LE, (1993) Ferroelectric ceramics: tailoring properties for specific applications.
In: Setter N, Colla EL (eds.) Ferroelectric Ceramics (Tutorial Reviews, Theory,
Processing, and Applications). Birkhuser Verlag, Basel Boston Berlin, pp 185
[3] Smolensky GA, Bokov VA, Isupov VA, Krainik NN, Pasynkov RE, Sokolov AI,
Yushin NK, (1985) Physics of Ferroelectric Phenomena. Nauka, Leningrad (in
Russian)
[4] Park S-E, Shrout TR, (1997) Relaxor based ferroelectric single crystals for electro-
mechanical actuators. Materials Research Innovations 1:2025
[5] Yin J, Cao W, (2000) Domain configurations in domain engineered
0.955Pb(Zn
1/3
Nb
2/3
)O
3
0.045PbTiO
3
single crystals. Journal of Applied Physics
87:74387441
[6] Liu D, Li JY, (2004) Domain-engineered Pb(Mg
1/3
Nb
2/3
)O
3
PbTiO
3
crystals:
Enhanced piezoelectricity and optimal configurations. Applied Physics Letters
84:39303932
[7] Noheda B, Cox DE, Shirane G, Gao J, Ye Z-G, (2002) Phase diagram of the
ferroelectric relaxor (1 x)PbMg
1/3
Nb
2/3
O
3
xPbTiO
3
. Physical Review B 66:054104
10 p
[8] La-Orauttapong D, Noheda B, Ye Z-G, Gehring PM, Toulouse J, Cox DE, Shirane G,
(2002) Phase diagram of the relaxor ferroelectric (1 x)Pb(Zn
1/3
Nb
2/3
)O
3
xPbTiO
3
.
Physical Review B 65:144101 7 p
[9] Singh AK, Pandey D, (2003) Evidence for M
B
and M
C
phases in the morphotropic
phase boundary region of (1 x)[Pb(Mg
1/3
Nb
2/3
)O
3
] xPbTiO
3
: A Rietveld study.
Physical Review B 67:064102 12 p
[10] Shuvaeva VA, Glazer AM, Zekria D, (2005) The macroscopic symmetry of
Pb(Mg
1/3
Nb
2/3
)
1-x
Ti
x
O
3
in the morphotropic phase boundary region (x = 0.25 0.5).
Journal of Physics: Condensed Matter 17:57095723
[11] Fu H, Cohen RE, (2000) Polarization rotation mechanism for ultrahigh
electromechanical response in single-crystal piezoelectrics. Nature 403:281283
[12] Noheda B, Cox DE, Shirane G, Park S-E, Cross LE, Zhong Z, (2001) Polarization
rotation via a monoclinic phase in the piezoelectric 92%PbZn
1/3
Nb
2/3
O
3
8%PbTiO
3
.
Physical Review Letters 86:38913894
[13] Wada S, Tsurumi T, (2002) Domain switching properties in PZNPT single crystals
with engineered domain configurations. Key Engineering Materials 214215:914
1/3
Nb
2/3
)O
3
0.33PbTiO
3
Physics 90:34713475
[15] Jiang W, Zhang R, Jiang B, Cao W, (2003) Characterization of piezoelectric materials
with large piezoelectric and electromechanical coupling coefficients. Ultrasonics
41:5563
[16] Zhang R, Jiang B, Jiang W, Cao W, (2006) Complete set of elastic, dielectric, and
piezoelectric coefficients of 0.93Pb(Zn
1/3
Nb
2/3
)O
3
0.07PbTiO
3
single crystals poled
along [001]. Applied Physics Letters 89:242908 3 p
[17] Topolov VYu, Turik AV, (2002) An intermediate monoclinic phase and
electromechanical interactions in xPbTiO
3
(1 x)Pb(Zn
1/3
Nb
2/3
)O
3
crystals. Solid
State Physics 44:13551362
[18] Liu T, Lynch CS, (2003) Ferroelectric properties of [110], [001] and [111] poled
relaxor single crystals: Measurements and modeling. Acta Materialia 51:407416
[19] Noheda B, (2002) Structure and high-piezoelectricity in lead oxide solid solutions.
Current Opinion in Solid State and Materials Science 6:2734
[20] Xu Y, (1991) Ferroelectric Materials and Their Applications. North-Holland,
Amsterdam London New York Toronto
[21] Jaffe B, Cook WR, Jaffe H, (1971) Piezoelectric Ceramics. Academic Press, London
New York
Piezoelectric Device-making. Vol. 1: Physics of Ferroelectric Ceramics.
[23] Peng J, Luo H, He T, Xu H, Lin D, (2005) Elastic, dielectric, and piezoelectric
characterization of 0.70Pb(Mg
1/3
Nb
2/3
)O
3
0.30PbTiO
3
single crystal. Materials
Letters 59:640643
[24] Oakley CG, Zipparo MJ, (2000) Single crystal piezoelectrics: A revolutionary
development for transducers. In: IEEE Ultrasonic Symposium: Proceedings. October
2225, 2000. Caribe Hilton San Juan, Puerto Rico. IEEE, Piscataway, pp 11571167
[25] Smotrakov VG, Eremkin VV, Alyoshin VA, Tsikhotsky ES, (2000) Preparation and
study of single-crystalceramic composite. Izvestiya Akademii Nauk. Seriya
Fizicheskaya 64:12201223 (in Russian)
[26] Topolov VYu, Glushanin SV, (2002) Effective electromechanical properties of
ferroelectric piezoactive composites of the crystalceramic type based on
(Pb
1x
Ca
x
)TiO
3
. Technical Physics Letters 28:279282
[27] Glushanin SV, Topolov VYu, Turik AV, (2003) Prediction of piezoelectric properties
of crystal-ceramic composites. Crystallography Reports 48:491499
[28] Takahashi H, Tukamoto S, Qiu J, Tani

J, Sukigara

T, (2003) Property of composite
ceramics composed of single crystals and ceramic matrix using hybrid sintering.
Japanese Journal of Applied Physics. Pt 1 42:60556058

3
ceramics
modified PbTiO
3
[31] Levassort F, Thi MP, Hemery H, Marechal P, Tran-Huu-Hue L-P, Lethiecq M, (2006)
Piezoelectric textured ceramics: Effective properties and application to ultrasonic
transducers. Ultrasonics 44:e621e626
[32] Topolov VYu, Kamlah M, (2004) Piezoelectric properties of PbTiO
3
-based 03 and
013 composites. Journal of Physics D: Applied Physics 37:15761585

Royal Society (London),
Pt A 443:265287
76:484489
[37] Topolov VYu, Turik AV, (1999) Nonmonotonic concentration dependence of
piezoelectric coefficients of 13 composites. Journal of Applied Physics 85:372379
[39] Cheng KC, Chan HLW, Choy CL, Yin Q, Luo H, Yin Z, (2003) Single crystal PMN
0.33PT / epoxy 13 composites for ultrasonic transducer applications. IEEE
[41] Savakus HP, Klicker KA, Newnham RE, (1981) PZT epoxy piezoelectric
transducers: A simplified fabrication procedure. Materials Research Bulletin 16:677
680
[42] Yin ZW, Luo HS, Wang PC, Xu GS, (1999) Growth characterization and properties
of relaxor ferroelectric PMNPT single crystals. Ferroelectrics 231:97102
[43] Bertram R, Reck G, Uecker R, (2003) Growth and correlation between composition
and structure of (1 x)Pb(Zn
1/3
Nb
2/3
)O
3
xPbTiO
3
single crystals near the
morphotropic phase boundary. Journal of Crystal Growth 253:212220

4473
[47] Bezus SV, Topolov VYu, Bowen CR, (2006) High-performance 13-type composites
based on (1 x) Pb(A
1/3
Nb
2/3
)O
3
xPbTiO
3
single crystals (A = Mg, Zn). Journal of
[48] Cao H, Schmidt VH, Zhang R, Cao W, Luo H, (2004) Elastic, piezoelectric, and
dielectric properties of 0.58Pb(Mg
1/3
Nb
2/3
)O
3
0.42PbTiO
3
single crystal. Journal of
[49] Yin J, Cao W, (2002) Effective macroscopic symmetries and materials properties of
multidomain 0.955Pb(Zn
1/3
Nb
2/3
)O
3
0.045PbTiO
3
Physics 92:444448
[51] Gibiansky LV, Torquato S, (1997) On the use of homogenization theory to design
optimal piezocomposites for hydrophone applications. Journal of the Mechanics and
Physics of Solids 45:689708
[52] Sigmund O, Torquato S, Aksay IA, (1998) On the design of 13 piezocomposites
using topology optimization. Journal of Materials Research 13:10381048
[54] Richard C, (1992) tude exprimentale et thorique de composites pizolectriques de
connectivit 1.3.1 pour hydrophone: Dr. thse acoustique. Lyon: Institut National
des Sciences Appliques de Lyon
[55] Topolov VYu, Glushanin SV, Panich AE, (2004) Features of the piezoelectric
response for a novel four-component composite structure. Ferroelectrics 308:5365
[57] Pal R, (2005) Porosity-dependence of effective mechanical properties of pore-solid
composite materials. Journal of Composite Materials 39:11471158

[56] Ashby MF, (2006) The properties of foams and lattices. Philosophical Transactions:
Mathematical, Physical and Engineering Sciences (Series A) 364:1530
6
Comparison of Results on Two-component
Piezo-composites
6.1 The Pros and Cons of Different Methods
This chapter presents data on the effective electromechanical properties determined
by different methods or averaging procedures. It is well known that determining
the effective properties of composites offers challenges and difficulties, even in the
case of relatively simple two-component piezo-composites with well-defined
microgeometry. These difficulties are closely connected with functions that would
describe structural inhomogeneities and their spatial distributions [1, 2].
Calculations of the effective electromechanical constants of piezo-composites are
based on methods that are approximate and require the internal fields affected by
numerous inclusions in a continuous matrix to differ from the electric and elastic
fields caused by a single inclusion in the same matrix. Moreover, any deviation
from a regular distribution of the inclusions in the matrix influences the effective
properties [2]. Methods based on direct averaging of the properties over a
macroscopic volume (analytical methods and FEM), methods based on
regularisation of structure, stochastic differential equations or virial expansion, as
well as a self-consistent method [16] are often applied to determine the effective
properties of the matrix of piezo-composites. In these composites the inclusions
can be approximated by ellipsoids or spheroids and electromechanical interactions
such as inclusion inclusion and inclusion matrix are taken into account in
the averaging procedures [1, 46]. The spheroidal shape of the inclusion enables
one to consider transitions from an oblate spheroid to a layer or from a prolate
spheroid to a circular cylindrical rod. The matrix method based on coupled
electroelastic Eshelby tensors [7] is applicable to the 03 composite with
spheroidal inclusions (Section 3.3.1). This method is associated with different
schemes of averaging [46], for example, the dilute approach, EFM (Mori
Tanaka method generalised for piezoelectric media), effective medium method,
differential scheme, etc. It is obvious that the electromechanical interaction
between components, along with the piezoelectric activity of at least one of the
composite components, influences its effective properties and coupled effects. A
comparison of the effective properties evaluated using different schemes or
methods would be of value when studying the advanced piezo-composites and
their performance. In the following sections we consider a series of examples
whereby a comparison is made related to ceramic / polymer composites with 03,
13 or 22 connectivity. As in Chapters 25, we assume that the effective
electromechanical properties are determined in the long-wave approximation, i.e.,
when wavelengths of external acoustic or electric fields are much longer than the
sizes of inclusions in a composite sample.
6.2 03 Versus 22 or 13
The model of a 03 composite (Figure 3.7) with spheroidal inclusions that are
regularly distributed in a matrix is important for further comparison with 22 and
13 structures. By varying the aspect ratio U or semi-axes a
i
of the inclusion
described by Equation 3.18, it is possible to consider the volume-fraction
behaviour of the effective properties in limiting cases, at U >> 1 or U << 1. The 03
composite with highly oblate inclusions (U >> 1) resembles a 22 series-connected
composite (Figure 3.1a) whereas the 03 composite with highly prolate inclusions
(U << 1) resembles a 13 composite with continuous rods parallel to the poling axis
OX
3
.
Examples of volume-fraction behaviour of the properties in the related 03 (U
>> 1) and 22 composites with one or two piezoelectric components are given in
Tables 6.1 and 6.2. Calculations were carried out within the framework of the EFM
(03 connectivity, Equation 3.20) or the matrix method (22 connectivity,
Equation 3.2). The electromechanical constants of the components were taken from
Tables 2.1 and 2.2 and experimental data [8]. It is known that the EFM is
applicable to the 03 composite (Figure 3.7) provided that its matrix is either
transversely isotropic or isotropic [47]. In the matrix method [911], which is
applied to the piezo-composites with planar interfaces, no restrictions regarding the
symmetry of the components are introduced.
The volume-fraction dependences calculated for the 22 ceramic / polymer and
single crystal / ceramic composites (Tables 6.1 and 6.2) suggest that their effective
properties satisfy an inequality
|)
ij
*
U = 10
2
)
ij
*
U = 10
3
| < |)
ij
*
U = 10
3
)
ij
*
22
| (6.1)
where )* = e*, d* and H*
V
. Condition 6.1 holds true in different cases, for
instance, when absolute values of the piezoelectric coefficients | (m)| and
| (m)| increase monotonously (Table 6.1) or when | (m)| and | (m)| show a
monotonic decrease and increase, respectively (Table 6.2). Inequality 6.1 is due to
slight changes in the aspect ratio U of the oblate inclusions and by corresponding
monotonic changes in the Eshelby tensors [7]. Increasing U implies that interfaces
separating the components in the composite approach the plane x
*
3 j
e
*
3 j
d
*
3 j
e
*
3 j
d
3
= const (Figure
3.1a) and both the components are distributed continuously in the OX
1
and OX
2
Comparison of Results on Two-component Piezo-composites 161
Table 6.1. Volume-fraction dependences of the piezoelectric coefficients (in C / m
*
3 j
e
2
) and
(in pC / N) and relative dielectric permittivity / H
*
d
V
H
*
3 j 33
0
, which have been calculated for
the 03 and 22 PZT-5 ceramic / araldite composites by means of the EFM
m
a
10
2
*
31
e
(U = 10
2
)
10
3
*
33
e
(U = 10
2
)
*
31
d
(U = 10
2
)
*
33
d
(U = 10
2
)
V
H
*
33
/ H
0
(U = 10
2
)
0.05
0.238
0.207
0.190
0.233 4.21
0.1
0.508
0.463
0.283
0.362 4.45
0.2
1.16
1.18
0.401
0.554 5.01
0.3
2.03
2.31
0.502
0.744 5.74
0.4
3.20
4.19
0.614
0.973 6.70
0.5
4.87
7.51
0.759
1.28 8.04
m
a
10
2
*
31
e
(U = 10
3
)
10
3
*
33
e
(U = 10
3
)
*
31
d
(U = 10
3
)
*
33
d
(U = 10
3
)
V
H
*
33
/ H
0
(U = 10
3
)
0.05
0.292
0.229
0.211
0.257 4.21
0.1
0.617
0.509
0.302
0.384 4.44
0.2
1.38
1.28
0.414
0.568 5.00
0.3
2.37
2.49
0.509
0.751 5.71
0.4
3.66
4.46
0.618
0.974 6.66
0.5
5.45
7.88
0.759
1.28 7.99
m
a
10
2
*
31
e
(22
composite)
10
3
*
33
e
(22
composite)
*
31
d
(22
composite)
*
33
d
(22
composite)
V
H
*
33
/ H
0
(22
composite)
0.05
0.300
0.232
0.214
0.261 4.21
0.1
0.632
0.515
0.305
0.387 4.44
0.2
1.42
1.29
0.415
0.569 5.00
0.3
2.41
2.51
0.510
0.752 5.71
0.4
3.72
4.49
0.618
0.975 6.66
0.5
5.53
7.93
0.759
1.27 7.98
Note to Table 6.1
a. Volume fraction of the ceramic component
*
3 j
e
2
) and
(in pC / N) and relative dielectric permittivity / H
*
d
V
H
*
3 j 33
0
, which have been calculated for
the 03 and 22 PMN0.33PT single crystal / PCR-7M ceramic composites by means of the
EFM
m
a
*
31
e
(U = 10
2
)
*
33
e
(U = 10
2
)
*
31
d
(U = 10
2
)
*
33
d
(U = 10
2
)
V
H
*
33
/ H
0
(U = 10
2
)
0.05
9.00
30.1
345
750 4980
0.1
8.54
29.2
344
750 4790
0.2
7.70
27.6
349
760 4480
0.3
6.97
26.2
361
787 4260
0.4
6.33
25.0
383
833 4140
0.5
5.78
23.9
417
904 4120
m
a
*
31
e
(U = 10
3
)
*
33
e
(U = 10
3
)
*
31
d
(U = 10
3
)
*
33
d
(U = 10
3
)
V
H
*
33
/ H
0
(U = 10
3
)
0.05
8.99
30.1
345
750 4980
0.1
8.53
29.2
344
748 4780
0.2
7.69
27.6
348
757 4460
0.3
6.96
26.2
360
783 4250
0.4
6.33
25.0
381
828 4120
0.5
5.77
23.9
414
899 4090
m
a
*
31
e
(22
composite)
*
33
e
(22
composite)
*
31
d
(22
composite)
*
33
d
(22
composite)
V
H
*
33
/ H
0
(22
composite)
0.05
9.00
30.1
345
749 4980
0.1
8.53
29.2
344
748 4780
0.2
7.69
27.6
348
757 4460
0.3
6.96
26.2
359
783 4240
0.4
6.33
25.0
380
828 4120
0.5
5.77
23.9
414
898 4090
Note to Table 6.2
a. Volume fraction of the single-crystal component
directions. Such an arrangement leads to considerable mechanical stress relief in
comparison with that in a 03 composite containing, for example, spherical
inclusions (U = 1).
A transition from the 03 composite with highly prolate inclusions (U << 1) to
the 13 composite also results in relatively small changes in the electromechanical
properties (Table 6.3). These changes are expressed by a formula
|)
ij
*
U = 10
2
)
ij
*
U = 10
3
| < |)
ij
*
U = 10
3
)
ij
*
13
| (6.2)
It is worth noting that the similarity of conditions 6.1 and 6.2 occurs at two
different transformations of the 03 structure. A decrease of the piezoelectric
activity of the 03 composite on increasing the aspect ratio U (Table 6.3) is a result
of an increase in the depolarisation field in the ferroelectric ceramic inclusions (see
Figure 3.7). It is clear that the 13 composite with the structure shown in Figure
3.2 exhibits some advantages over the 03 composite at U o 0 due to the
continuous distribution of the piezo-active component along the poling axis OX
3
.
Another interesting case of the transition from 03 to 13 connectivity is
concerned with the (Pb
0.75
Ca
0.25
)TiO
3
ceramic / araldite composite. Its ceramic
component is characterised by the piezoelectric coefficients > 0 and large
anisotropy of the piezoelectric coefficients at sgn = sgn < 0 (see Table
3.3). The presence of this component causes a small increase of and a decrease
of | | (Table 6.4) as the aspect ratio U increases at m = const. On increasing U the
) 1 (
3 j
e
) 1 (
3 j
d
) 1 (
31
d
) 1 (
33
d
*
31
e
*
31
d
*
3 j
e
2
)
and (in pC / N) and relative dielectric permittivity / H
*
d
V
H
*
3 j 33
0
, which have been calculated
for the 03 and 13 PCR-7M ceramic / araldite composites by means of the EFM
m
a
*
31
e
(U = 10
2
)
*
33
e
(U = 10
2
)
*
31
d
(U = 10
2
)
*
33
d
(U = 10
2
)
V
H
*
33
/ H
0
(U = 10
2
)
0.05
0.0293
1.58
74.8
203 121
0.1
0.0621
3.19
112
298 276
0.2
0.141
6.50
154
398 631
0.3
0.243
9.91
182
458 1030
0.4
0.378
13.4
205
504 1470
0.5
0.565
17.1
227
545 1950
m
a
*
31
e
(U = 5u10
3
)
*
33
e
(U = 5u10
3
)
*
31
d
(U = 5u10
3
)
*
33
d
(U = 5u10
3
)
V
H
*
33
/ H
0
(U = 5u10
3
)
0.05
0.0331
1.81
90.5
246 145
0.1
0.0697
3.62
138
368 338
0.2
0.156
7.27
190
493 779
0.3
0.266
10.9
222
561 1260
0.4
0.409
14.6
247
606 1760
0.5
0.604
18.3
267
640 2290
m
a
*
31
e
(U = 10
3
)
*
33
e
(U = 10
3
)
*
31
d
(U = 10
3
)
*
33
d
(U = 10
3
)
V
H
*
33
/ H
0
(U = 10
3
)
0.05
0.0348
1.91
98.5
268 157
0.1
0.0731
3.81
151
405 370
0.2
0.163
7.61
210
544 857
0.3
0.276
11.4
244
614 1380
0.4
0.423
15.2
267
657 1910
0.5
0.620
18.9
286
686 2450
m
a
*
31
e
(13
composite)
*
33
e
(13
composite)
*
31
d
(13
composite)
*
33
d
(13
composite)
V
H
*
33
/ H
0
(13
composite)
0.05
0.0349
1.91
99.0
269 158
0.1
0.0733
3.82
152
407 372
0.2
0.163
7.63
211
547 862
0.3
0.277
11.4
245
618 1380
0.4
0.424
15.2
269
660 1920
0.5
0.621
18.9
287
689 2460
Note to Table 6.3
a. Volume fraction of the ceramic component
*
3 j
e
2
),
(in pC / N),
g
(in mV
*
d
* *
h
3 j 3 j
m / N), and (in 10
3 j
9
the 03 (U << 1) and 13 (Pb
0.75
Ca
0.25
)TiO
3
ceramic / araldite composites by means of the
EFM
a
and FEM
b
Method
and U
m
c
10
2 *
31
e
*
33
e
*
31
d
*
33
d
*
31
g
*
33
g
10
2 *
31
h
*
33
h
EFM,
U = 0.05
0.1
0.2
0.3
0.4
0.5
0.999
2.24
3.80
5.83
8.56
0.263
0.556
0.884
1.25
1.67
4.85
5.14
4.82
4.31
3.72
16.2
19.7
21.5
22.7
23.8
32.4
19.1
12.3
8.24
5.64
108
73.3
54.7
43.5
36.1
6.86
8.67
10.1
11.8
13.8
1.80
2.15
2.36
2.53
2.69
EFM,
U = 0.01
0.1
0.2
0.3
0.4
0.5
0.972
2.16
3.66
5.61
8.23
0.440
0.885
1.34
1.79
2.26
6.58
6.54
5.90
5.12
4.30
21.5
24.6
25.8
26.5
26.9
38.0
21.0
13.1
8.66
5.88
124
78.7
57.1
44.8
36.8
5.94
7.45
8.78
10.3
12.2
2.69
3.05
3.20
3.29
3.36
EFM,
U = 0
0.1
0.2
0.3
0.4
0.5
0.960
2.14
3.63
5.56
8.17
0.459
0.919
1.38
1.84
2.31
6.73
6.65
5.99
5.19
4.35
22.0
25.0
26.1
26.8
27.2
38.4
21.1
13.1
8.69
5.89
126
79.3
57.3
44.9
36.9
5.82
7.31
8.63
10.1
12.1
2.78
3.14
3.28
3.36
3.42
FEM,
U = 0
0.1
0.2
0.3
0.4
0.5
0.960
2.14
3.63
5.59
8.31
0.459
0.919
1.38
1.84
2.31
6.73
6.65
5.99
5.18
4.30
22.0
25.0
26.1
26.8
27.2
38.4
21.1
13.1
8.67
5.83
126
79.1
57.3
44.8
36.8
5.82
7.31
8.63
10.2
12.3
2.78
3.14
3.28
3.36
3.42
Notes to Table 6.4
a. Calculations performed using Equation 3.20 and the Eshelby tensor from ref. [7]
b. Calculations performed using the COMSOL package [12]
c. Volume fraction of the ceramic component
larger increase in the piezoelectric coefficient in comparison with (Table
6.4) is a result of the effective dielectric permittivity being about one order of
magnitude smaller than of the above-considered PCR-7M ceramic / araldite
composites. With reference to 13 connectivity, Table 6.4 shows that values of all
the piezoelectric coefficients calculated by means of the EFM are in good
agreement with the values obtained using the FEM. Some difference between the
values of the piezoelectric coefficients )
*
31
h
*
31
e
[
H
*
[ *
33
H
33
31
* () = e, d, g, or h) is observed at m t
0.4 and may be associated with features of the electromechanical interaction
between the piezoelectric rods in the (X
1
OX
2
) plane. At large volume fractions (m)
the interaction is more intensive than that predicted on the basis of the EFM. This
is similar to discussions by Berger et al. [13] in work considering the prediction of
the effective electromechanical properties in 13 PZT-5 ceramic / polymer
composites. In Section 6.3 we will consider a series of examples of the volume-
fraction dependences determined for the 13 ceramic / polymer composites by
means of different methods.
6.3 Data on 13 Composites
The shape, size and arrangement of the ceramic rods in the 13 piezo-composites
influence their effective electromechanical properties [1316]. The volume-
fraction behaviour of these properties is often considered for composites that
contain ferroelectric ceramic rods in the form of either circular cylinders or
rectangular parallelepipeds and are characterised by a regular distribution of these
rods.
6.3.1 Composites with Cylindrical Ceramic Rods
The effective electromechanical constants , and of the 13 ceramic /
polymer composite shown in Figure 3.2 were calculated [13] by means of the
asymptotic homogenisation method (analytical solution) and the FEM. The FEM
was applied to the 13 structure with a specific arrangement of the rods, and
examples of the volume-fraction dependence of the effective electromechanical
properties are shown in Figure 6.1. It should be noted that Berger et al. [13] used
superscript eff to denote the effective constants (this superscript is equivalent to
an asterisk in our notation) and fibre volume fraction in Figure 6.1 denotes the
volume fraction m of the ceramic component (rods) in our notations. The , ,
, and curves show a good correlation in the wide volume-fraction range
(Figures 6.1c, fh). In addition, according to results from ref. [13], the ,
and calculated by the FEM are also in good agreement with the similar
E
ab
c
* *
ij
e
[
H
*
pp
E
c
*
13
*
31
e
*
15
e
[
H
*
11
E
c
*
33
*
33
e
[
H
*
33
a
Figure 6.1. Volume-fraction dependence of elastic moduli (i.e., in our notations,
graphs a e), piezoelectric coefficients (i.e., in our notations, graphs f and g) and
dielectric permittivity (i.e., in our notations, graph h) which have been calculated
for the 13 PZT-5 ceramic / polymer composite by either the FEM or the asymptotic
homogenisation method (AHM). In the case of FEM calculations the square (Squ) and
hexagonal (Hex) arrangements of the ceramic rods in the polymer matrix have been
considered (reprinted from Berger et al. [13], with permission from Elsevier)
eff
ab
c
E
ab
c
*
eff
kl
e
*
kl
e
eff
11
H
[
H
*
11
b
c
d
e
f
g
h
parameters calculated by the asymptotic homogenisation method in the same
volume-fraction range.
A complex study on the effective electromechanical properties of 13
composites has recently been carried out [17]. Kar-Gupta and Venkatesh
developed an analytical model suitable for prediction of the effective properties in
the general case, i.e., when the 13 composite consists of two piezo-active and
elastically anisotropic components. A scheme for determining the effective
electromechanical constants (up to 45 constants) has been proposed to characterise
the properties in the wide range of volume fractions of the ceramic component and
to take into account symmetry features of the components. This scheme has been
applied to the following classes of the 13 composites: longitudinal composites
with components poled on the longitudinal direction (OX
3
) and transverse
composites wherein the matrix is poled on the transverse direction (OX
1
) and the
rods are poled on the longitudinal direction (OX
3
). These two classes provide 20
material combinations with different symmetry and properties. Kar-Gupta and
Venkatesh compared the effective electromechanical properties determined within
the framework of the analytical model with those predicted using the FEM. Greater
agreement between FEM and analytical results [17] is achieved for 13 composites
where the dielectric properties of the piezo-active rods are dominant. Less
agreement is achieved where the dominant dielectric properties are related to the
matrix of the 13 composites.
6.3.2 Composites with Parallelepiped-shaped Ceramic Rods
In this section it is assumed that the 13 ceramic / polymer composite contains rods
in the form of a rectangular parallelepiped aligned parallel to the poling axis OX
3
with square bases in the (X
1
OX
2
) plane [14]. The rods are regularly distributed in a
continuous matrix so that a square pattern of rods is apparent in the (X
1
OX
2
) plane.
The effective electromechanical properties of the 13 composite with
parallelepiped-shaped rods that are continuous on the OX
3
direction can be
determined using the matrix method [10, 11, 14] applied to the 1E (E = 1, 2 and 3)
and 22 composites (see Section 3.1.1). The averaging procedure involves full sets
of the elastic compliances , piezoelectric coefficients and dielectric
permittivities of the components (n = 1 and 2). Averaging is implemented on
the volume fraction m of the ceramic component at 0 < m < 1 due to the planar
microgeometry of this 13 composite. Two examples of the volume-fraction
behaviour of the effective piezoelectric coefficients and hydrostatic parameters of
the 13 composites examined are shown in Table 6.5.
E n
ab
s
), ( ) (n
kl
d
V
H
), (n
pp
1
The ceramics used as
composite components are distinguished (Table 2.1) by the signs of the
piezoelectric coefficients , by values of dielectric permittivity and by the
anisotropy of the elastic moduli . The effective parameters of the 13
composites with the same components have been calculated by means of the FEM
[15] and within the framework of the differential medium theory [16] (see the two
types of curves in Figure 6.2). The 13 composite considered in ref. [15] contains
ceramic rods in the form of a rectangular parallelepiped (Figure 6.3) with a finite
height h. A maximum difference between the similar effective parameters
calculated using the matrix method (Table 6.5, h o f) and the FEM (Figure 6.2)
for a mesh shown in Figure 6.3 does not exceed 19% at volume fractions
) 1 (
31
e
[
H
), 1 (
33
E
ab
c
), 1 (
____________
1
Hydrostatic electromechanical coupling factor of the 13 composite studied is defined [10,
15] as = ( )
*
h
k
*
h
d
E
h
s
* V
H
*
33
1/2
, where = + + is the hydrostatic piezoelectric
coefficient, = + + + + is the hydrostatic elastic compliance and
is the dielectric permittivity.
*
h
d
*
31
d
*
32
d
*
33
d
E
h
s
* E
s
*
11
E
s
*
12
E
s
*
13
E
s
*
33
V
H
*
33
Table 6.5. Volume-fraction dependences of the piezoelectric coefficients , and
(in pC / N), squared hydrostatic figure of merit
(
(in 10
*
31
d
*
33
d
*
h
d
2 *
) Q
*
k
h
15
Pa
1
), and hydrostatic
electromechanical coupling factor , which have been calculated for the 13 ceramic /
Stycast polymer composites with parallelepiped-shaped rods by means of the matrix method
h
Ceramic
component
m
*
31
d
*
33
d
*
h
d
2 *
) (
h
Q
*
h
k
PZT-5A 0.05
0.1
0.2
0.3
0.4
0.5
0.6
0.7
26.3
45.7
73.6
94.1
111
124
136
147
87.1
146
221
266
297
319
336
349
29.6
47.6
65.1
70.2
69.5
65.6
60.0
53.4
1750
2080
1740
1260
888
614
418
279
0.121
0.138
0.136
0.125
0.114
0.104
0.094
0.087
(Pb, Ca)TiO
3
0.05
0.1
0.2
0.3
0.4
0.5
0.6
0.7
8.42
11.7
13.4
12.8
11.5
9.93
8.29
6.68
28.0
40.9
53.2
59.1
62.5
64.8
66.5
67.7
13.6
20.9
30.1
36.6
42.0
46.7
51.0
54.9
1700
2280
2460
2430
2400
2370
2360
2340
0.120
0.145
0.161
0.173
0.185
0.200
0.218
0.242
a
Figure 6.2. Effective parameters calculated for the 13 ceramic / polymer composites within
either the framework of the FEM (bold line) or the differential effective medium theory
(dotted line): piezoelectric coefficients , and (in pC / N) of the PZT-5A ceramic
/ Stycast composite (a) and (Pb, Ca)TiO
*
31
d
*
33
d
*
h
d
3
ceramic / Stycast composite (b), squared
hydrostatic figure of merit (c, in 10
2 *
) (
h
Q
15
Pa
1
) and hydrostatic electromechanical
coupling factor (d, in per cent) of both the composites (reprinted from Nelli Silva et al.
[15], with permission from Springer)
*
h
k
b
c
d
Figure 6.3. 13 composite unit cell discretised by the 3 u 3 u 3 finite element mesh
(reprinted from Nelli Silva et al. [15], with permission from Springer)
0.05 d m d 0.7. The h value mainly affects the piezoelectric coefficient and, to
a lesser degree, the hydrostatic parameters listed in Table 6.5. It should be added
that for calculations of the effective parameters (Figure 6.2), Nelli Silva et al. [15]
used the electromechanical constants , and of the components,
whereas our averaging procedure based on formulae [10, 11] involves ,
and . On comparing data shown in Table 6.5 and Figure 6.2, the difference
between the effective parameters is less pronounced in the PZT-5A-based
composite at large m values and in the (Pb, Ca)TiO
*
31
d
E n
ab
c
), ( ) (n
kl
e
[
H
), (n
pp
E n
ab
s
), ( ) (n
kl
d
V
H
), (n
pp
3
-based composite at small m
values. Such volume-fraction behaviour can be considered to be due to the
differing anisotropy of the ceramic components in the 13 composites studied.
6.3.3 Variational Bounds for Effective Properties in 13 Composites with
Cylindrical Ceramic Rods
Effective parameters of a 13 ceramic / polymer composite with cylindrical rods
are calculated within the framework of the EFM (matrix formalism) [36], the
FEM [12, 13] or different micromechanical models (analytical formulae are given,
for example, in refs. [1, 2, 8, 1719, 23, 24]). The effective electromechanical
constants calculated using these methods are in good agreement over a wide
volume-fraction range.
The 13 ceramic / polymer composite is also of interest due to rigorous
variational bounds [20, 21] determined for the effective electromechanical
properties. These bounds were derived on the basis of variational principles
proposed by Hashin and Shtrikman [22] and generalised by Bisegna and Luciano
[20, 21] for the case of piezo-active heterogeneous media. Bisegna and Luciano
first applied the generalised principles [20] to 13 piezo-composites with a regular
arrangement (square or circular) of ceramic rods in a polymer matrix. The
generalised principles enabled the prediction of tighter bounds (in comparison with
the Hashin Shtrikman and other bounds) and was initially demonstrated for the
electromechanical constants of a PZT-type ceramic / polymer composite [20, 21].
The lower and upper bounds can be determined [20, 21] for the following sets of
electromechanical constants of the 13 composite:
(i) elastic moduli , piezoelectric coefficients and dielectric
impermeabilities (first set) and
D
ab
c
* *
kl
h
[
E
*
pp
(ii) elastic compliances , piezoelectric coefficients and dielectric
impermeabilities (second set).
E
ab
s
* *
kl
d
V
H
*
pp
The rigorous lower and upper bounds [12] are compared with the effective
electromechanical constants listed in Table 6.6. For this comparison a composite
based on the PbTiO
3
-type ceramic was chosen due to the large anisotropy of its
piezoelectric coefficients . It is seen that the effective constants | |, and
of the 13 composite studied attain the upper bound at various m values
*
3l
d
*
31
d
*
33
d
V
H
*
33
Table 6.6. Elastic moduli (in 10
D
a
c
*
3
9
Pa), piezoelectric coefficients (in 10
*
33
h
9
V / m) and
(in pC / N), dielectric impermeability (in 10
*
d
[
E
* V
H
*
3l 33
9
m / F), dielectric permittivity / H
33
0
,
and rigorous bounds on these properties in the 13 (Pb
0.75
Ca
0.25
)TiO
3
ceramic / araldite
composite with cylindrical rods
a
Evaluations m
d
D
c
*
13
D
c
*
33
*
33
h
| |
*
31
d
*
33
d
[
E
*
33
V
H
*
33
/ H
0
Using EFM
b
0.1 4.76 25.0 2.78 6.73 22.0 6.06 19.8
Lower bound
c
0.1 4.76 25.0 2.78 3.63 21.5 6.06 19.7
Upper bound
c
0.1 6.52 25.8 2.91 6.73 22.0 6.08 19.8
Using EFM 0.2 5.21 42.4 3.14 6.65 25.0 3.42 35.6
Lower bound 0.2 5.21 42.4 3.14 2.74 24.6 3.41 35.6
Upper bound 0.2 8.97 44.1 3.29 6.65 25.0 3.43 35.7
Using EFM 0.3 5.78 59.9 3.28 5.99 26.1 2.38 51.5
Lower bound 0.3 5.78 59.9 3.28 2.06 25.9 2.38 51.5
Upper bound 0.3 11.8 62.6 3.45 5.99 26.1 2.39 51.6
Using EFM 0.4 6.52 77.5 3.36 5.19 26.8 1.82 67.5
Lower bound 0.4 6.52 77.5 3.36 1.59 26.6 1.82 67.4
Upper bound 0.4 15.1 81.3 3.54 5.19 26.8 1.83 67.5
Using EFM 0.5 7.52 95.2 3.42 4.35 27.2 1.48 83.4
Lower bound 0.5 7.51 95.2 3.41 1.24 27.0 1.48 83.3
Upper bound 0.5 18.9 100 3.61 4.35 27.2 1.49 83.4
Using EFM 0.6 8.94 114 3.46 3.51 27.4 1.24 99.3
Lower bound 0.6 8.94 113 3.46 0.977 27.3 1.24 99.3
Upper bound 0.6 23.5 120 3.67 3.51 27.4 1.25 99.3
Notes to Table 6.6
a. Square arrangement of the ceramic rods is assumed to be in the (X
1
OX
2
) plane
b. Electromechanical constants determined by means of the EFM (see Equation 3.20) and
constitutive piezoelectric equations (see Equations 2.122.19)
c. Electromechanical constants determined on the basis of bounding [12, 20, 21]
d. Volume fraction of the ceramic component
(Table 6.6). The effective constants , , , and of the same
composite attain the lower bound. This feature of the volume-fraction dependence
of the effective electromechanical properties is closely connected with the structure
of the 13 composite (Figure 3.2) that contains ceramic rods aligned and poled on
the OX
D
c
*
13
D
c
*
33
*
33
h
[
E
*
33
3
direction. The structure promotes, for example, a fulfilment of conditions
(m) =
)
(see curve 2 in Figure 3.4a) and (m) = [8, 23, 24] with an
accuracy to a few per cent at m > 0.1. The rigorous bounds shown in Table 6.6
exhibit relatively narrow ranges of the electromechanical properties of the
composite (e.g., , , , , and ).
*
33
d
1 (
33
d
*
31
d
) 1 (
31
d
D
c
*
33
*
33
h
*
33
d
[
E
*
33
V
H
*
33
As follows from the data shown in Tables 6.4 and 6.6, the EFM is undoubtedly
a reliable approach to predict the electromechanical properties in the 13 ceramic /
polymer composite in the wide volume-fraction range.
6.3.4 Variational Bounds for Effective Properties in 13 Composites with
Parallelepiped-shaped Ceramic Rods
Another case of developing bounds of the effective electromechanical properties is
related to the 13 composite with ceramic rods in the form of a rectangular
parallelepiped. It is assumed that these rods are continuous in the OX
3
direction and
regularly distributed in the polymer matrix so that a square arrangement takes place
in the (X
1
OX
2
) plane. The composite as a whole is poled along the OX
3
direction,
and their effective properties can be determined by different methods (see Section
6.3.2). Next we consider the effective electromechanical constants calculated using
the matrix method [10, 11, 14] and the rigorous lower and upper bounds found [25]
for the piezoelectric and dielectric constants (Table 6.7) of the 13 composite
based on the PbTiO
3
-type ceramic. It is seen that values of the piezoelectric
coefficients are slightly more than the lower bounds (Table 6.7) found in the
wide m range. As with the data in Table 6.6, the rigorous bounds on dielectric
impermeability form narrow ranges, and the values calculated by the
matrix method are also within these bounds (see Table 6.7). Similar trends are
observed in 13 composites with different ceramic (Pb
*
3 j
h
[
E
*
33
[
E
*
33
1x
Ca
x
)TiO
3
compositions
(Table 3.3) with a variety of piezoelectric anisotropies.
The results of the rigorous lower and upper bounds determined for the 13
ceramic / polymer composites can lead to a further comparison of the effective
electromechanical properties of piezo-composites with different microgeometries.
Such a comparison, for example, would be important when predicting the effective
electromechanical properties of composites with one or two piezo-active
components (e.g., ceramic and polymer or single crystal and ceramic) with
inclusions of variable aspect ratio.
Table 6.7. Piezoelectric coefficients (in pC / N), 10
*
3l
d
2
and
*
31
e
*
33
e
(in C / m
2
),
*
3l
g
(in
mV m / N), 10
2
*
31
h
*
33
h
[
E
*
33
H
and (in 10
9
V / m), dielectric impermeability (in 10
9
m / F),
V *
33 (in 10
10
F / m),
a
and rigorous bounds [25],
b
which have been
calculated for the 13 (Pb
1x
Ca
x
)TiO
3
ceramic / araldite composite with rods in the form of
the rectangular parallelepiped
x m
c
*
31
d
*
33
d
10
2
*
31
e
*
33
e
*
31
g
0.20 0.1 5.91 19.3 0.809 0.402 39.3
0.2 6.02 21.9 1.59 0.803 22.6
0.3 5.59 23.0 2.55 1.21 14.6
0.4 5.01 23.5 3.79 1.61 10.1
0.5 4.39 23.9 5.47 2.01 7.15
0.6 3.75 24.1 7.87 2.42 5.14
0.7 3.12 24.3 11.6 2.83 3.69
x m
c
*
33
g
10
2
*
31
h
*
33
h
[
E
*
33
V
H
*
33
0.20 0.1 128 5.66
(4.41, 25.5)
2.81
(2.80, 2.89)
7.00
(6.99, 7.01)
1.51
0.2 82.5 6.40
(5.65, 31.4)
3.23
(3.23, 3.33)
4.02
(4.02, 4.03)
2.66
0.3 60.2 7.20
(6.71, 35.4)
3.40
(3.40, 3.52)
2.82
(2.82, 2.83)
3.82
0.4 47.3 8.24
(7.92, 39.2)
3.50
(3.49, 3.62)
2.18
(2.17, 2.18)
4.97
0.5 39.0 9.67
(9.46, 43.2)
3.56
(3.56, 3.70)
1.77
(1.77, 1.77)
6.13
0.6 33.1 11.7
(11.6, 47.7)
3.61
(3.60, 3.75)
1.49
(1.49, 1.49)
7.29
0.7 28.8 14.9
(14.8, 53.0)
3.65
(3.65, 3.80)
1.29
(1.29, 1.29)
8.45
x m
c
*
31
d
*
33
d
10
2
*
31
e
*
33
e
*
31
g
0.25 0.1 6.55 22.0 1.23 0.460 37.5
0.2 6.55 25.0 2.43 0.920 20.8
0.3 5.94 26.1 3.89 1.38 13.0
0.4 5.16 26.8 5.78 1.84 8.65
0.5 4.34 27.2 8.35 2.31 5.88
0.6 3.51 27.5 12.0 2.78 3.99
0.7 2.69 27.7 17.7 3.26 2.63
x m
c
*
33
g
10
2
*
31
h
*
33
h
[
E
*
33
V
H
*
33
0.25 0.1 126 7.48
(5.81, 34.2)
2.79
(2.78, 2.91)
6.06
(6.06, 6.08)
1.75
0.2 79.1 8.29
(7.30, 41.2)
3.14
(3.14, 3.29)
3.42
(3.41, 3.43)
3.15
0.3 57.3 9.25
(8.62, 46.2)
3.29
(3.28, 3.45)
2.38
(2.38, 2.39)
4.56
0.4 44.9 10.6
(10.1, 51.0)
3.37
(3.36, 3.54)
1.82
(1.82, 1.83)
5.97
0.5 36.8 12.4
(12.1, 56.2)
3.42
(3.41, 3.61)
1.48
(1.48, 1.49)
7.38
x m
c
*
33
g
10
2
*
31
h
*
33
h
[
E
*
33
V
H
*
33
0.25 0.6 31.2 15.0
(14.8, 62.1)
3.46
(3.46, 3.67)
1.25
(1.24, 1.25)
8.79
0.7 27.1 19.0
(18.9, 69.1)
3.60
(3.50, 3.72)
1.08
(1.07, 1.08)
10.2
Notes to Table 6.7
a. Electromechanical constants determined by means of the matrix method [10, 11, 14]
b. The rigorous lower and upper bounds given for and in parentheses are published
as a private communication, with permission of Prof. Dr. P. Bisegna (University of Rome
Tor Vergata, Italy)
*
33
h
[
E
*
33
c. Volume fraction of the ceramic component
6.4 Data on 03 Composites
The effective electromechanical properties of the 03 ceramic / polymer composite
(Figure 3.7) with a square arrangement of spheroidal inclusions in the (X
1
OX
2
)
plane can be determined through different averaging procedures involving the
Eshelby tensor S. These procedures are often concerned with the dilute approach,
the EFM (generalised Mori Tanaka approach) or the effective medium (self-
consistent) method [4, 6, 17]. The matrix of the effective electromechanical
properties is written in the general form [4, 6, 26] as
|| C* || =|| C
(2)
|| +m(|| C
(1)
|| || C
(2)
||) || A|| (6.3)
where || A || is the mechanical strain electric field concentration matrix that is
expressed in terms of the matrices of the electromechanical properties of the
components || C
(n)
|| (n = 1 and 2) and of the Eshelby tensor || S || [7] and m is the
volume fraction of the ceramic component (inclusions). The || C
(n)
|| matrices from
Equation 6.3 have a form similar to that in Equation 3.19, and n = 1 relates to the
ceramic inclusion and n = 2 relates to the polymer matrix.
The dilute approach neglects the electromechanical interaction between the
components. As a consequence, the average electroelastic field in the composite is
approximated by a field that appears in an isolated inclusion surrounded by a
continuous matrix [6, 26, 27]. In this case Equation 6.3 is represented as
|| C* || =|| C
(2)
|| +m(|| C
(1)
|| || C
(2)
||) [|| I || + || S || || C
(2)
||
-1
(|| C
(1)
|| || C
(2)
||)]
1
(6.4)
where || I || is the identity matrix. The EFM is based on an assumption [4, 6, 26]
that || A || from Equation 6.3 is determined from a solution of a micromechanical
problem for a single piezo-active inclusion in a continuous matrix subjected to an
applied electroelastic field that is equal to an average field in the matrix of the 03
composite. Its effective electromechanical properties are determined using
Equation 3.20. The effective medium method [4, 6, 26, 27] enables one to consider
a single inclusion in a homogenised or effective medium (i.e., a matrix with similar
inclusions), and the effective properties of this system are determined by taking
into account the electromechanical interaction between the inclusion and the
surrounding medium. In this case the || A || matrix from Equation 6.3 is written in
the form
|| A || = [|| I || + || S || || C* ||
-1
(|| C
(1)
|| || C* ||)]
1
(6.5)
where || C* || characterises the effective properties of the composite. The averaging
procedure becomes self-consistent and needs iterations to solve Equation 6.3,
wherein the || A || and || S || matrices (see Equation 6.5) depend on the elements of
the || C* || matrix.
In the section below, numerical results predicted using the aforementioned
averaging procedures are compared. The volume-fraction dependences of the
effective parameters from Table 6.8
2
agree with those shown in Figure 6.4. It
should be mentioned that refs. [6, 7] devoted to piezo-composites comprise
different sets of formulae for the Eshelby tensor S calculated for the spheroidal
inclusion. Nevertheless, the difference between the calculated values from Table
6.8 and Figure 6.4 remains less than 5%. A good correlation is observed in wide m
and c / a ranges. The effective piezoelectric coefficients of the studied 03
composite obey conditions | |
*
3 j
d DIL
< | |
*
3 j
d MT
< | |
*
3 j
d SC
and | |
*
3 j
e DIL
< | |
*
3 j
e MT
<
| |
*
3 j
e SC
(Figure 6.4) where subscripts DIL, MT and SC denote the dilute
approach, the EFM (Mori Tanaka approach) and the effective medium (or self-
consistent) method, respectively. A transition from the non-interacting components
to interacting components results in higher piezoelectric activity in the 03
composite. We see that the electromechanical interaction between the components
is more appreciable at c / a o 1 (Table 6.8 and Figure 6.4) and leads to small
differences between the constants evaluated by different methods at c / a >> 1.
This distinction is closely connected with the internal electric and mechanical
fields that are highly dependent on the aspect ratio of the ceramic inclusions. The
case of c / a = 1 corresponds to the composite with the spherical inclusions wherein
the depolarisation and stress fields cannot be neglected. In the case of c / a = 10
3
the presence of the needle-shaped inclusions poled along the OX
3
axis leads to a
removal of the depolarisation field, however, the stress field remains essential and
plays an important role in forming the piezoelectric properties in the composite.
Apparently, such a stress field would not affect equalities | |
*
3 j
d
DIL
| | |
*
3 j
d
MT
|
| |
*
3 j
d
SC
and | |
*
3 j
e
DIL
| | |
*
3 j
e
MT
| | |
*
3 j
e
SC
that hold at c / a = 10
3
(see data from Table
6.8, Figure 6.4 and ref. [6]).
Agreement between the piezoelectric properties determined through different
procedures is also observed when comparing the aspect-ratio dependences shown
in Table 6.9 and Figure 6.5. The piezoelectric coefficients and = + 2
*
33
e
*
h
e
*
33
e
*
31
e
____________
2
The aspect ratio c / a introduced by Fakri et al. [6] equals U
-1
= a
3
/ a
1
, where a
3
= c and
a
1
= a are semi-axes of the spheroidal inclusion shown in Figure 3.7.
Table 6.8. Volume-fraction dependences of the piezoelectric coefficients (in pC / N)
and (in C / m
*
31
d
*
e
*
e
) 1 (
*
e
) 1 (
31
2
), normalised piezoelectric coefficient /
e
and normalised hydrostatic
piezoelectric coefficient /
e
,
33 33
h h
which have been calculated for the 03 PZT-type ceramic
/ epoxy composite with spheroidal inclusions
Ceramic
component
m
*
31
d
(c / a = 1)
a
*
31
d
(c / a = 10)
a
*
31
d
(c / a = 10
3
)
a
PZT-5 0.05
0.1
0.2
0.3
0.4
0.5
0.132
c
0.239
0.405
0.526
0.619
0.692
5.36
d
8.12
11.0
12.6
13.6
14.3
50.2
e
76.3
104
119
129
136
Ceramic
component
m
*
31
d
(c / a = 1)
b
*
31
d
(c / a = 10)
b
*
31
d
(c / a = 10
3
)
b
PZT-5 0.05
0.1
0.2
0.3
0.4
0.5
0.144
f
0.286
0.576
0.893
1.27
1.74
5.70
g
9.14
14.0
18.1
22.5
27.9
50.3
h
76.6
105
122
133
142
Ceramic
component
m
*
31
e
(c / a = 10
3
)
a
*
33
e
/
) 1 (
33
e
(c / a = 10
3
)
a
*
h
e
/
) 1 (
h
e
(c / a = 10
3
)
a
PZT-4 0.05
0.1
0.2
0.3
0.4
0.5
0.0189
0.0377
0.0754
0.113
0.151
0.189
0.0608
i
0.122
0.243
0.365
0.487
0.608
0.188
j
0.375
0.750
1.12
1.50
1.88
Ceramic
component
m
*
31
e
(c / a = 10
3
)
b
*
33
e
/
) 1 (
33
e
(c / a = 10
3
)
b
*
h
e
/
) 1 (
h
e
(c / a = 10
3
)
b
PZT-4 0.05
0.1
0.2
0.3
0.4
0.5
0.0198
0.0416
0.0926
0.157
0.240
0.352
0.0608
k
0.122
0.242
0.363
0.483
0.601
0.187
l
0.373
0.740
1.10
1.45
1.7
Notes to Table 6.8
a. Calculated within the framework of the dilute approach (see Equation 6.4)
b. Calculated by means of the EFM (see Equation 3.20)
c. The calculated (m) dependence is to be compared with that shown in Figure 6.4a (DIL
curve)
*
31
d
d. The calculated (m) dependence is to be compared with that shown in Figure 6.4b (DIL
curve)
*
31
d
e. The calculated (m) dependence is to be compared with that shown in Figure 6.4c (DIL
curve)
*
31
d
f. The calculated (m) dependence is to be compared with that shown in Figure 6.4a (MT
curve)
*
31
d
g. The calculated (m) dependence is to be compared with that shown in Figure 6.4b (MT
curve)
*
31
d
h. The calculated (m) dependence is to be compared with that shown in Figure 6.4c (MT
curve)
*
31
d
i. The calculated (m) / dependence is to be compared with that shown in Figure 6.4d
(lower DIL curve)
*
33
e
) 1 (
33
e
j. The calculated (m) / dependence is to be compared with that shown in Figure 6.4d
(upper DIL curve)
*
h
e
) 1 (
h
e
k. The calculated (m) / dependence is to be compared with that shown in Figure 6.4d
(lower MT curve)
*
33
e
) 1 (
33
e
l. The calculated (m) / dependence is to be compared with that shown in Figure 6.4d
(upper MT curve)
*
h
e
) 1 (
h
e
a
Figure 6.4. Effective piezoelectric coefficients calculated for the 03 PZT-5 ceramic /
epoxy (a c) and PZT-4 ceramic / epoxy (d) composites within the framework of the dilute
approach (DIL curves), the EFM (MT curves) and the effective medium method (SC
curves): (in pC / N) at c / a = 1 (a), 10 (b) and 10
*
31
d
3
(c), (m) / and (m) / at c
/ a = 10
*
33
e
) 1 (
33
e
*
h
e
) 1 (
h
e
3
(d). In graph (a) bold dots correspond to experimental data by Furukawa et al. [28]
(reprinted from Fakri et al. [6], with permission from Elsevier)
b
c
d
Table 6.9. Normalised piezoelectric coefficients / and /
*
33
e
) 1 (
33
e
*
h
e
) 1 (
h
e
that have been
calculated for the 03 PZT-4 ceramic / epoxy composite with spheroidal inclusions
c / a
*
33
e
/
) 1 (
33
e
(m = 0.2)
a
*
h
e
/
) 1 (
h
e
(m = 0.2)
a
*
33
e
/
) 1 (
33
e
(m = 0.5)
a
*
h
e
/
) 1 (
h
e
(m = 0.5)
a
1
2
5
7
10
20
50
70
100
200
500
700
1000
2.93u10
-4 c
9.66u10
-4
6.70u10
-3
1.40u10
-2
2.95u10
-2
9.16u10
-2
0.189
0.211
0.225
0.238
0.243
0.243
0.243
1.26u10
-4 d
2.15u10
-3
1.94u10
-2
4.18u10
-2
8.90u10
-2
0.281
0.582
0.649
0.694
0.733
0.747
0.749
0.750
7.32u10
-4 e
2.42u10
-3
1.67u10
-2
3.51u10
-2
7.36u10
-2
0.229
0.473
0.527
0.563
0.595
0.607
0.608
0.608
3.16u10
-4 f
5.38u10
-3
4.85u10
-2
0.104
0.223
0.702
1.46
1.62
1.73
1.83
1.87
1.87
1.87
c / a
*
33
e
/
) 1 (
33
e
(m = 0.2)
b
*
h
e
/
) 1 (
h
e
(m = 0.2)
b
*
33
e
/
) 1 (
33
e
(m = 0.5)
b
*
h
e
/
) 1 (
h
e
(m = 0.5)
b
1
2
5
7
10
20
50
70
100
200
500
700
1000
4.48u10
-4 g
1.46u10
-3
9.61u10
-3
1.95u10
-2
3.88u10
-2
0.107
0.198
0.216
0.228
0.238
0.242
0.242
0.243
2.01u10
-4 h
3.21u10
-3
2.75u10
-2
5.73u10
-2
0.116
0.326
0.603
0.659
0.695
0.727
0.738
0.740
0.741
2.70u10
-3 i
8.38u10
-3
4.85u10
-2
9.00u10
-2
0.160
0.348
0.528
0.559
0.578
0.595
0.600
0.601
0.602
1.33u10
-3 j
1.79u10
-2
0.135
0.257
0.466
1.03
1.56
1.66
1.71
1.76
1.78
1.78
1.78
Notes to Table 6.9
a. Calculated within the framework of the dilute approach (see Equation 6.4)
b. Calculated by means of the EFM (see Equation 3.20)
c. The calculated (c / a) / dependence is to be compared with that shown in Figure
6.5a (lower DIL curve)
*
33
e
) 1 (
33
e
d. The calculated (c / a) / dependence is to be compared with that shown in Figure
6.5a (upper DIL curve)
*
h
e
) 1 (
h
e
e. The calculated (c / a) / dependence is to be compared with that shown in Figure
6.5b (lower DIL curve)
*
33
e
) 1 (
33
e
f. The calculated (c / a) / dependence is to be compared with that shown in Figure
6.5b (upper DIL curve)
*
h
e
) 1 (
h
e
g. The calculated (c / a) / dependence is to be compared with that shown in Figure
6.5a (lower MT curve)
*
33
e
) 1 (
33
e
h. The calculated (c / a) / dependence is to be compared with that shown in Figure
6.5a (upper MT curve)
*
h
e
) 1 (
h
e
i. The calculated (c / a) / dependence is to be compared with that shown in Figure
6.5b (lower MT curve)
*
33
e
) 1 (
33
e
j. The calculated (c / a) / dependence is to be compared with that shown in Figure
6.5b (upper MT curve)
*
h
e
) 1 (
h
e
Figure 6.5. Normalised piezoelectric coefficients / and / of the 03 PZT-4
ceramic / epoxy composite at ceramic volume fractions m = 0.2 (a) and 0.5 (b). Calculations
were made within the framework of the dilute approach (DIL curves), the EFM (MT curves)
and the effective medium method (SC curves) (reprinted from Fakri et al. [6], with
permission from Elsevier)
*
33
e
) 1 (
33
e
*
h
e
) 1 (
h
e
of the 03 ceramic / polymer composite are now studied when the aspect ratio U is
varied from 1 (spherical inclusions with c / a = 1) to 10
3
(needle-shaped inclusions
with c / a = 10
3
). The comparison of the dependences (Table 6.9 and Figure 6.5)
shows that a difference between the parameters calculated using different formulae
for the Eshelby tensor S does not exceed 5% at 1 d c / a d 10
3
, i.e., 10
3
d U d 1 in
our notations. This conclusion is valid for the parameters calculated on the basis of
the dilute approach (Equation 6.4) and the EFM (Equation 3.20) at m = const. It is
assumed that the ceramic volume fraction m is taken from a range of 0 < m d 0.5,
i.e., low and moderate volume fractions can be considered.
Thus, our results on the 03 ceramic / polymer composites with spheroidal
inclusions suggest that the EFM can be reliably used to predict the
electromechanical properties in the wide m and U ranges. Moreover, the
transformation of the 03 structure shown in Figure 3.7 leads to the 22 or 13
composite (see Section 6.2). The effective properties in the 22 or 13 composites
are determined in limiting cases, for the aspect ratios U o 0 or U o f,
respectively. It should be noted that no jumps are observed in the aspect-ratio
dependences of the piezoelectric and dielectric properties (Tables 6.16.4) at m =
const when approaching U o 0 or U o f. In general, based on data on asymptotic
behaviour of the effective properties in the 03 ceramic / polymer composites, one
can conclude that the EFM is a reliable approach for determining the volume-
fraction and aspect-ratio dependences of the effective electromechanical properties
in these materials. A further comparison of data on the effective electromechanical
constants of 03 piezo-composites can be made using the FEM.
References
[3] Luchaninov AG, (2002) Piezoelectric Effect in Non-polar Heterogeneous
Ferroelectric Materials. Volgograd State Academy of Architecture and Construction,
Volgograd (in Russian)
30:161175
superlattices. Journal of Applied Physics 64:4469 4473
[11] Topolov VYu, Glushanin SV, (2002) Evolution of connectivity patterns and links
[12] Topolov VYu, Bisegna P, Krivoruchko AV, (2008) Features of electromechanical
properties of 13 composites based on PbTiO
3
-type ceramics. Journal of Physics D:
Applied Physics 41:035406 8 p
[16] Avellaneda M, Swart PJ, (1994) The role of matrix porosity and Poissons ratio in the
design of high-sensitivity piezocomposite transducers. In: Garcia E, Cudney H, Das
Gupta A (eds.) Adaptive Structures and Composite Materials: Analysis and
Applications. American Society of Mechanical Engineers, New York, AD 45:5966
[17] Kar-Gupta R, Venkatesh T A, (2007) Electromechanical response of 13 piezoelectric
composites: An analytical model. Acta Materialia 55:10931108
[18] Smith WA, (1993) Modeling 13 composite piezoelectrics: hydrostatic response.
IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control 40:4149
76:484489
[20] Bisegna P, Luciano R, (1996) Variational bounds for the overall properties of
piezoelectric composites. Journal of the Mechanics and Physics of Solids 44:583602
[21] Bisegna P, Luciano R, (1997) On methods for bounding the overall properties of
periodic piezoelectric fibrous composites. Journal of the Mechanics and Physics of
Solids 45:13291356
[22] Hashin Z, Shtrikman S, (1962) On some variational principles in anisotropic and
nonhomogeneous elasticity. Journal of the Mechanics and Physics of Solids 10:335
342
[25] Bisegna P, (2007) Private communication
[26] Dunn M, Micromechanics of coupled electroelastic composites: Effective thermal
expansion and pyroelectric coefficients. Journal of Applied Physics 73:51315140
[28] Furukawa T, Fujino K, Fukada E, (1976) Electromechanical properties in the
composites of epoxy resin and PZT ceramics. Japanese Journal of Applied Physics
15:21192129
7
Conclusions
Eigentlich weiE man nur, wenn man wenig weiE, mit dem
Wissen wchst der Zweifel.
J.W. von Goethe
From a practical point of view, the physical world only matters
in so far as it affects us, and the intrinsic nature of what goes on in
our absence is irrelevant, provided we can predict the effects upon
ourselves. This we can do, just as a person can use a telephone
without understanding electricity. Only the most abstract knowledge
is required for practical manipulation of matter. But there is a grave
danger when this habit of manipulation based upon mathematical
laws is carried over into our dealings with human beings, since they,
unlike the telephone wire, are capable of happiness and misery,
desire and aversion.
B. Russell
7.1 From Components to Effective Electromechanical Properties
This book has demonstrated to the reader that the analysis of the effective
electromechanical properties must take into account a number of factors
(components, connectivity, microgeometry, etc.) and the variety of averaging
procedures that are available. These procedures become complicated due to
electromechanical coupling in the presence of piezo-active components. The role
of the ferroelectric components has been considered in detail, demonstrating that
their remarkable piezoelectric properties are fundamental in forming the
piezoelectric response of the composites with various connectivity patterns and / or
different combinations of the components. Our study enables us to conclude that
both the ferroelectric ceramic and single-crystal components within piezoelectric
composites lead to considerable piezoelectric activity, sensitivity, anisotropy, and
hydrostatic piezoelectric response in specific volume-fraction ranges. These
properties have been determined for specific composite architectures, connectivity
and components. In a number of cases, conventional ferroelectric ceramic /
polymer composites with 13, 22 or 03 connectivity have electromechanical
properties that can be further improved by using the latest relaxor-ferroelectric
single crystals with engineered-domain structures and very high piezoelectric
activity. Additional possibilities of improving the performance of the above-
mentioned ceramic / polymer composites are concerned with porous structures [1],
auxetic polymer components [2, 3] or the addition of a third component, e.g.,
ferroelectric polymer, ceramic or single crystal with specific electromechanical
properties. Furthermore, advanced methods of manufacturing (rapid prototyping
[4], selective laser sintering, etc.) also often lead to better performance of the
piezo-composites and enable us to control their microgeometry and sizes of several
microstructural elements more effectively.
The main steps of our analysis are schematically shown in Figure 7.1.
Figure 7.1. Outline of the study on the piezo-composites
The diagram will be reinforced by new important items associated with future
research results and applications of advanced piezo-active and other composites. In
this connection we would like to finish by quoting a review paper [5]: As niche
applications become more prevalent in the future, composites and displacement-
amplifying techniques and materials will proliferate in a continuing effort to widen
the forcedisplacement envelope of performance. These devices, too, will become
smarter and smarter as the applications demand.
References
[2] Evans KE, Alderson A, (2000) Auxetic materials: functional materials and structures
from lateral thinking. Advanced Materials 12:617628
[3] Topolov VYu, Bowen CR, (2008) Characteristics of 13-type ferroelectric ceramic /
auxetic polymer composites. Modelling and Simulation in Materials Science and
Engineering 16:015007 12 p
[5] Haertling GH, (1999) Ferroelectric ceramics: history and technology. Journal of the
Appendix
Formulae for Effective Electromechanical Constants of
22 and 13 Piezo-composites
22 series-connected composites with two piezo-active transversely isotropic
components (Figure 3.1a) distributed stochastically on the OX
3
direction are
characterised [1] by the effective elastic moduli , piezoelectric coefficients
and dielectric susceptibilities as follows:
E
ij
c
* *
kl
e
[
F
*
pp
E
c
*
11
= m + m (m
E
c
), 1 (
11
1
E
c
), 2 (
11
2 1
m
2
/')[( )
E
c
), 1 (
13
E
c
), 2 (
13
2
( m + m )
+ 2(
c
)(
e
e
)(
e
m +
e
m
[
F
), 2 (
33
1
[
F
), 1 (
33
2
E
c
), 1 ( E ), 2 ( ) 1 ( ) 2 ( ) 2 ( ) 1 ( ) 1 ( ) 2 ( E ), 2 ( E
c
), 1 (
13 13 31 31 33
1
33
2
) (
e
e
)
31 31
2
(
c
m + m )],
33
1
33
2
E
c
*
12
= m
E
c
), 1 (
12
1
+ m
E
c
), 2 (
12
2
(m
1
m
2
/')[( )
E
c
), 1 (
13
E
c
), 2 (
13
2
( m
[
F
), 2 (
33
1
+ m
[
F
), 1 (
33
2
)
+ 2( )( )( m
E
c
), 1 (
13
E
c
), 2 (
13
) 1 (
31
e
) 2 (
31
e
) 2 (
33
e 1
+ m
) 1 (
33
e 2
) ( )
) 1 (
31
e
) 2 (
31
e
2
( m
E
c
), 2 (
33
1
+ m
E
c
), 1 (
33
2
)],
E
c
*
13
= m
E
c
), 1 (
13
1
+ m
E
c
), 2 (
13
2
(m
1
m
2
/'){( )[( )( m
E
c
), 1 (
13
E
c
), 2 (
13
E
c
), 1 (
33
E
c
), 2 (
33
[
F
), 2 (
33
1
+
[
F
), 1 (
33
m
2
) + ( )( m
) 1 (
33
e
) 2 (
33
e
) 2 (
33
e 1
+ m
) 1 (
33
e 2
)] ( )[( )( m
) 1 (
31
e
) 2 (
31
e
) 1 (
33
e
) 2 (
33
e
E
c
), 2 (
33
1
+ m
E
c
), 1 (
33
2
) ( )( m
E
c
), 1 (
33
E
c
), 2 (
33
) 2 (
33
e 1
+ m
) 1 (
33
e 2
)]},
E
c
*
33
= m
E
c
), 1 (
33
1
+ m
E
c
), 2 (
33
2
(m
1
m
2
/')[( )
E
c
), 1 (
33
E
c
), 2 (
33
2
( m
[
F
), 2 (
33
1
+ m
[
F
), 1 (
33
2
)
+ 2( )( )( m
E
c
), 1 (
33
E
c
), 2 (
33
) 1 (
33
e
) 2 (
33
e
) 2 (
33
e 1
+ m
) 1 (
33
e 2
) ( )
) 1 (
33
e
) 2 (
33
e
2
( m
E
c
), 2 (
33
1
+ m
E
c
), 1 (
33
2
)],
E
c
*
44
= m
E
c
), 1 (
44
1
+ m
E
c
), 2 (
44
2
m
1
m
2
[( )
E
c
), 1 (
44
E
c
), 2 (
44
2
/ ( m
E
c
), 2 (
44
1
+ m
E
c
), 1 (
44
2
)],
*
31
e
= m
) 1 (
31
e 1
+ m
) 2 (
31
e 2
(m
1
m
2
/'){( )[( )( m
) 1 (
33
e
) 2 (
33
e
E
c
), 1 (
13
E
c
), 2 (
13
[
F
), 2 (
33
1
+ mB
2
[
F
), 1 (
33
) + m
[
F
), 2 (
33
1
+ m
[
F
), 1 (
33
2
) + ( )( m
) 1 (
31
e
) 2 (
31
e
) 2 (
33
e 1
+ m
) 1 (
33
e 2
) + (
)[( )( m
[
F
), 1 (
33
[
F
), 2 (
33
) 1 (
31
e
) 2 (
31
e
E
c
), 2 (
33
1
+ m
E
c
), 1 (
33
2
) ( )( m
E
c
), 1 (
13
E
c
), 2 (
13
) 2 (
33
e 1
+ m
) 1 (
33
e 2
)]},
*
33
e
= m
) 1 (
33
e 1
+ m
) 2 (
33
e 2
(m
1
m
2
/'){( )[( )( m
) 1 (
33
e
) 2 (
33
e
E
c
), 1 (
33
E
c
), 2 (
33
[
F
), 2 (
33
1
+ m
[
F
), 1 (
33
2
) + ( )( m
) 1 (
33
e
) 2 (
33
e
) 2 (
33
e 1
+ m
) 1 (
33
e 2
)] + ( )[( )( m
[
F
), 1 (
33
[
F
), 2 (
33
) 1 (
33
e
) 2 (
33
e
E
c
), 2 (
33
1
+ m
E
c
), 1 (
33
2
) ( )( m
E
c
), 1 (
33
E
c
), 2 (
33
) 2 (
33
e 1
+ m
) 1 (
33
e 2
)]},
188 Appendix
= m
*
15
e
) 1 (
15
e 1
+ m
) 2 (
15
e 2
m
1
m
2
[( )(
) 1 (
15
e
) 2 (
15
e
E
c
), 1 (
44
) / ( m
E
c
), 2 (
44
E
c
), 2 (
44
1
+ m
E
c
), 1 (
44
2
)],
= m
[
F
*
11
[
F
), 1 (
11
1
+ m
[
F
), 2 (
11
2
+ m
1
m
2
[( )
) 1 (
15
e
) 2 (
15
e
2
/ ( m
E
c
), 2 (
44
1
+ m
E
c
), 1 (
44
2
)], and
= m
[
F
*
33
[
F
), 1 (
33
1
+ m
[
F
), 2 (
33
2
(m
1
m
2
/')[( )
[
F
), 1 (
33
[
F
), 2 (
33
2
( m
E
c
), 2 (
33
1
+ m
E
c
), 1 (
33
2
)
+ 2( )( )( m
[
F
), 1 (
33
[
F
), 2 (
33
) 1 (
33
e
) 2 (
33
e
) 2 (
33
e 1
+ m
) 1 (
33
e 2
) ( )
) 1 (
33
e
) 2 (
33
e
2
( m
[
F
), 2 (
33
1
+ m
[
F
), 1 (
33
2
)],
where m
n
, , , and are the volume fraction, elastic moduli,
piezoelectric coefficients, and dielectric susceptibilities of the nth component,
respectively, n = 1 and 2, m
E n
ij
c
), ( ) (n
kl
e
[
F
), (n
pp
2
= 1 m
1
, and ' = ( m
E
c
), 2 (
33
1
+ m
E
c
), 1 (
33
2
)( m
[
F
), 2 (
33
1
+
m
[
F
), 1 (
33
2
) + ( m
) 2 (
33
e 1
+ m
) 1 (
33
e 2
)
2
. Dielectric susceptibility and dielectric
permittivity of the strained nth component are interrelated as follows:
= 1. The like relation, = 1, also holds for the composite.
[
F
), (n
pp
[
H
), (n
pp
[
F
), (n
pp
[
H
), (n
pp
[
F
*
pp
[
H
*
pp
13 composites with two piezo-active transversely isotropic components (i.e.,
long cylindrical rods distributed regularly in a continuous matrix, Figure 3.2) are
characterised [2] by the effective elastic moduli , piezoelectric coefficients
and and dielectric permittivities as follows:
E
ij
c
* *
kl
d
*
kl
e
V
H
*
pp
E
c
*
11
+ = {( ) < + > + ( + )( + )}c
E
c
*
12
E
c
), 2 (
11
E
c
), 2 (
12
E
c
11
E
c
12
E
c
), 1 (
11
E
c
), 1 (
12
E
c
), 2 (
11
E
c
), 2 (
12
1
,
E
c
*
13
= {2m + (1 m) ( + + )}c
E
c
), 1 (
13
E
c
), 2 (
11
E
c
), 2 (
13
E
c
), 1 (
11
E
c
), 1 (
12
E
c
), 2 (
11
E
c
), 2 (
12
1
,
E
c
*
33
= < > 2m(1 m)( )
E
c
33
E
c
), 1 (
13
E
c
), 2 (
13
2
c
1
,
E
c
*
44
= = {( + < >)( + [ ]) (1 m)
E
c
*
55
E
c
), 2 (
55
E
c
), 1 (
55
E
c
55
V
H
), 2 (
11
V
H
11
2
E
c
), 1 (
55
E
c
), 2 (
55
(
)
) 1 (
15
d
) 2 (
15
d
2
}{ + [ ])( + [ ]) (1 m)
E
c
), 2 (
55
E
c
55
V
H
), 2 (
11
V
H
11
2
E
c
), 1 (
55
E
c
), 2 (
55
( )
) 1 (
15
d
) 2 (
15
d
2
}
1
,
*
31
d
= {(2 m
) 1 (
31
d
E
s
), 2 (
11
+ ( + + )(1 m))( m
) 2 (
31
d
E
s
), 1 (
11
E
s
), 1 (
12
E
s
), 2 (
11
E
s
), 2 (
12
E
s
), 2 (
33

+ (1 m)) 2( m
E
s
), 1 (
33
E
s
), 2 (
13
+ (1 m))( m
E
s
), 1 (
13
) 1 (
31
d
E
s
), 2 (
13
+ (1 m)) + (
)(( + + ) 2 )m(1 m)}A
) 2 (
31
d
E
s
), 1 (
13
) 1 (
33
d
) 2 (
33
d
E
s
), 1 (
11
E
s
), 1 (
12
E
s
), 2 (
11
E
s
), 2 (
12
E
s
), 2 (
13
E
s
), 2 (
11
E
s
), 1 (
13
1
,
*
33
d
= {(( + )m
E
s
), 2 (
11
E
s
), 2 (
12
+ ( + )(1 m) + )( m
E
s
), 1 (
11
E
s
), 1 (
12
E
s
), 2 (
11
E
s
), 2 (
12
) 1 (
33
d
E
s
), 2 (
33
+
) 2 (
33
d
E
s
), 1 (
33
(1 m)) 2( m
E
s
), 2 (
13
+ (1 m))( m
E
s
), 1 (
13
) 1 (
33
d
E
s
), 2 (
13
+ (1 m))
+ 2( )( )m(1 m)}A
) 2 (
33
d
E
s
), 1 (
13
) 1 (
31
d
) 2 (
31
d
E
s
), 1 (
33
E
s
), 2 (
13
E
s
), 2 (
33
E
s
), 1 (
13
1
,
*
15
d
= + 4m ( ){(
) 2 (
15
d
E
c
), 1 (
55
V
H
), 2 (
11
) 1 (
15
d
) 2 (
15
d
E
c
), 1 (
55
+ < >)( + [ ]) (1
m)
E
c
55
V
H
), 2 (
11
V
H
11
2
E
c
), 1 (
55
E
c
), 2 (
55
( )
) 1 (
15
d
) 2 (
15
d
2
}
1
,
*
31
e
= {2 m +
) 1 (
31
e
E
c
), 2 (
11
) 2 (
31
e
( + + )(1 m)}c
E
c
), 1 (
11
E
c
), 1 (
12
E
c
), 2 (
11
E
c
), 2 (
12
1
,
*
33
e
= m
) 1 (
33
e
+ (1 m) 2m(1 m)( )( )c
) 2 (
33
e
) 1 (
31
e
) 2 (
31
e
E
c
), 1 (
13
E
c
), 2 (
13
1
,
V
H
*
11
= { + < >)( + < >) (1 m
V
H
), 2 (
11
V
H
), 1 (
11
V
H
11
E
c
), 1 (
55
E
c
55
2
) (
)
E
c
), 1 (
55
E
c
), 2 (
55
) 1 (
15
d
) 2 (
15
d
2
}{ + [ ])( +
V
H
), 1 (
11
V
H
11
E
c
), 1 (
55
< >) (1 m
E
c
55
2
) ( )
E
c
), 1 (
55
E
c
), 2 (
55
) 1 (
15
d
) 2 (
15
d
2
}
1
, and
Appendix 189
V
H
*
33
= m + (1 m) m(1 m){( )s
V
H
), 1 (
33
V
H
), 2 (
33
) 1 (
33
d
) 2 (
33
d c
+ 2( )P
2
) 1 (
31
d
) 2 (
31
d
[ ]
4( )( )[ ]A
E
s
33
) 1 (
31
d
) 2 (
31
d
) 1 (
33
d
) 2 (
33
d
E
s
13
1
where , , , , and are the elastic moduli, elastic
compliances, piezoelectric stress coefficients, piezoelectric charge coefficients, and
dielectric permittivities of the nth component, respectively, m is the volume
fraction of the rods, [X] = mX
E n
ij
c
), ( E n
ij
s
), ( ) (n
kl
e
) (n
kl
d
V
H
), (n
pp
(2)
+ (1 m)X
(1)
, <Y> = mY
(1)
+ (1 m)Y
(2)
, A = ([
E
s
11
+
] + )[ ] 2[ ]
E
s
12
E
s
), 2 (
11
E
s
), 2 (
12
E
s
33
E
s
13
2
, c = + m(
E
c
), 2 (
11
E
c
), 2 (
12
E
c
), 2 (
11
+ ) + (1
m)( + ), n = 1 corresponds to the rods, and n = 2 corresponds to the
matrix.
E
c
), 2 (
12
E
c
), 1 (
11
E
c
), 1 (
12
References
of Piezo-active Composite Materials. Kiev: Naukova Dumka (in Russian)
List of Abbreviations
AHM, asymptotic homogenisation method
APC, perovskite-type piezoelectric ceramic
BURPS, burned-out plastic spheres (technique)
CS, corn starch
DIL, dilute approach
EFM, effective field method
FC, ferroelectric ceramic
FEM, finite-element method
Hex, hexagonal arrangement of rods
MPB, morphotropic phase boundary
MT, effective field method (Mori Tanaka method applied to piezo-active media)
PCR, piezoelectric ceramic from Rostov-on-Don (Russia)
PEO, polyethylene oxide (PEO)
PKI, perovskite-type piezoelectric ceramic
PMMA, polymethyl methacrylate
PMNxPT, relaxor-ferroelectric solid solutions of (1 x)Pb(Mg
1/3
Nb
2/3
)O
3

xPbTiO
3
PVDF, polyvinylidene fluoride
PZ, perovskite-type piezoelectric ceramic
PZNxPT, relaxor-ferroelectric solid solutions of (1 x)Pb(Zn
1/3
Nb
2/3
)O
3

xPbTiO
3
PZT, piezoelectric ceramic of the Pb(Zr, Ti)O
3
type
SC, effective medium (self-consistent) method
Squ, square arrangement of rods
SR, self-raising flour
ZTS, ZTSSt, ZTStBS, piezoelectric ceramics of the Pb(Zr, Ti)O
3
type
About the Authors
Vitaly Yuryevich Topolov was born in Rostov-on-Don, Russia (former USSR) on
November 8th, 1961. He received the qualification Physicist. Educator (honours
degree) in 1984 along with the degrees Candidate of Sciences (Physics and
Mathematics) and Doctor of Sciences (Physics and Mathematics) in 1987 and
2000, respectively, all from the Rostov State University, Russia. From 1987 to
1991, he worked as a research scientist of the Institute of Physics at the Rostov
State University. From 1991 to 2000, he worked as a Senior Lecturer (19911992)
and an Associate Professor (19922000) of the Department of Physics at the
Rostov State University. Since 2000, he has been Professor of the same
Department. Since December 2006, after reforming the Rostov State University, he
is Professor of the Department of Physics at the Southern Federal University
(Rostov-on-Don, Russia).
Dr. V. Yu. Topolov was also a visiting scientist at the Moscow State
University, Russia (former USSR, 1989), University of Saarland, Germany (1994
1995), Aachen University of Technology RWTH Aachen, Germany (1998),
Karlsruhe Research Center, Germany (2002 and 20032004), University of Bath,
UK (2006 and 2007), and University of Rome Tor Vergata, Italy (2008). His
research interests include heterogeneous ferroelectrics, smart materials, domain
and heterophase structures, as well as electromechanical effects in ferroelectrics
and related materials. He earned the special award from the International Science
Foundation (1993) and the Soros Associate Professor title and awards from the
International Soros Science-Educational Program and the Open Society Institute
(1997, 1998, 2000, and 2001). He presented the best poster at the International
Symposium on Ferroelectric Domains (China, 2000) and the best training aids in
the Department of Physics, Southern Federal University (Russia, 2006). He is the
author of about 280 scientific papers, reviews, conference proceedings, and
abstracts. Dr. V. Yu. Topolov has been included in the list of Active Russian
Scientists whose papers have been cited over 100 times in periodicals from 2000 to
2006. The biographical sketch by Dr. V. Yu. Topolov was first published in the
26th edition of Dictionary of International Biography (IBC, Cambridge, 1998),
p. 465.
194 About the Authors
Christopher Rhys Bowen was born on January 18th, 1968 and grew up in
Beddau, South Wales (UK). He earned a BSc (First Class) in Materials Science at
the School of Materials, University of Bath, UK in 1990 and worked on his DPhil
thesis in ceramics under the supervision of Prof. Brian Derby in the Department of
Materials, University of Oxford, UK in 19901993 (PhD awarded in 1994). In
1993 1994, he worked as a Researcher in the Advanced Ceramics Group at the
Technical University of Hamburg-Harburg (TUHH), Germany under the
supervision of Prof. Nils Claussen. From 1994 to 1996, he was research fellow at
the School of Materials, University of Leeds, UK. From 1996 to 1998, he was a
Senior Scientist at DERA, Functional Materials Group, UK. He joined the
University of Bath in August 1998 and is now a Reader at the same University.
The research interests of Dr. C. R. Bowen are concerned with functional
ceramics, including ferroelectric ceramics and composites for modern sensors and
actuators, as well as with manufacturing and characterisation of these materials. He
earned the Thornton and Hazelwood prizes for academic work (19861990),
Institute of Materials National Lecture Competition award (1993), SET award
(2002), and John Willis award for excellence in research and teaching (2003).
Dr. C. R. Bowen has published approximately 80 scientific papers, conference
proceedings, and abstracts.
Index
anisotropy, 3, 28, 45, 58, 61, 132,
134, 141, 143, 174
dielectric, 71
elastic, 50, 61, 143, 170, 171
piezoelectric, 3, 16, 17, 18, 48, 49,
50, 52, 54, 58, 61, 62, 66, 67,
70, 71, 73, 75, 83, 89, 90, 102,
109, 110, 113, 116, 125, 131,
139, 140, 142, 154
large, 3, 17, 18, 49, 50, 54, 62,
66, 67, 71, 73, 74, 76, 102,
125, 127, 131, 141, 162,
172
applications, 1, 14, 25, 26, 34, 43, 46,
52, 54, 93, 124, 143, 153, 154,
186
acoustic antenna, 29
actuator, 1, 14, 29, 46, 53, 124,
154
hydroacoustic, 46, 62, 109
hydrophone, 8, 29, 46, 53, 63, 85,
105, 109, 116, 124, 154
mechanical engineering, 29
medical device, 46, 53, 62, 116,
124
non-destructive testing, 53
piezotechnical, 17, 18, 29, 34, 35,
62, 105, 116, 123, 124, 153
sensor, 1, 14, 29, 46, 62, 63, 85,
116, 124, 156
signal generation, 92
transducer, 1, 3, 29, 46, 53, 116,
124, 154
araldite, 48, 49, 50, 55, 56, 66, 68, 70,
73, 134, 141, 142, 143, 144, 145,
146, 149, 151, 152, 153, 161, 162,
163, 164, 172, 174, 175, 177
porous, 145, 146, 151, 152, 153
arrangement, 4, 5, 16, 26, 29, 30, 31,
48, 53, 54, 86, 102, 105, 162, 165
irregular, 63
rods, 53, 54, 142, 165
layers, 77
regular, 30, 51, 102, 171
hexagonal, 165
square, 54, 131, 132, 172, 173,
175
quasi-continuous, 76
aspect ratio, 64, 65, 66, 70, 73, 77,
78, 88, 111, 126, 129, 133, 150,
152, 153, 160, 162, 173, 176, 182
averaging procedure, 17, 22, 25, 27,
28, 29, 34, 63, 75, 77, 78, 82, 101,
111, 142, 159, 168, 171, 175, 176,
185
Banno unit cell, 26, 27, 30, 63, 82,
87, 104, 116, 117, 143
boundary conditions, 3, 25, 27, 29,
46, 47, 60, 82, 142, 143, 149
196 Index
ceramic, 1, 2, 3, 5, 17, 29, 43, 45, 46,
47, 49, 53, 58, 61, 63, 64, 65, 67,
71, 73, 76, 77, 79, 82, 83, 84, 86,
87, 89, 93, 94, 101, 102, 114, 116,
123, 131, 165, 174, 185
BaTiO
3
, 3, 16, 17, 18, 19, 24, 72
ferroelectric, 1, 2, 3, 5, 6, 7, 8, 14,
15, 16, 17, 18, 19, 20, 22, 25,
26, 27, 28, 29, 30, 34, 35, 43,
45, 48, 52, 62, 72, 101, 102,
115
hard PZT, 15
magnetostrictive, 24
modified PbTiO
3
, 19, 89, 92, 125,
131
(Pb, A)(Ti, B)O
3
-type, 66
Pb(Ca, Ti)O
3
, 19, 63, 67, 71, 72,
73, 76, 162, 164, 169
(Pb
1x
Ca
x
)TiO
3
, see Pb(Ca, Ti)O
3
(Pb
1x
Me
x
)TiO
3
, 66
PbTiO
3
-type, 17, 18, 48, 51, 54,
65, 70, 71, 75, 81, 83, 102,
126, 172, 173
Pb(Zr, Ti)O
3
-type, see PZT-type
Pb(Zr
1x
Ti
x
)O
3
, 3, 15, 18, 20, 43,
45, 46, 48, 53, 54, 61, 62, 72,
76, 89, 90
piezoelectric, 1, 15, 16, 20, 24, 26,
86, 88, 89, 90, 110, 111
porous, 8, 26, 27, 101, 102, 103,
104, 105, 109, 110, 111
classification, 105
pyroelectric, 1
PZT-type, 15, 22, 43, 45, 46, 48,
61, 65, 75, 81, 86, 101, 102,
103, 105, 106, 107, 108, 109,
110, 115, 133, 172, 177
soft PZT, 15
textured, 125
combination properties, 22, 23, 24
component, 1, 2, 3, 4, 5, 6, 9, 13, 22,
23, 24, 25, 26, 27, 28, 29, 30, 31,
32, 33, 34, 43, 44, 45, 46, 47, 52,
53, 54, 57, 58, 60, 61, 62, 63, 64,
65, 71, 75, 76, 77, 78, 79, 81, 82,
83, 85, 86, 87, 88, 89, 90, 92, 101,
106, 107, 108, 110, 111, 113, 116,
117, 118, 123, 124, 125, 126, 130,
131, 132, 133, 134, 141, 142, 143,
144, 145, 152, 153, 154, 161, 162,
164, 165, 167, 168, 169, 170, 171,
173, 175, 176, 177, 185, 186
highly anisotropic, 66, 71, 130,
131, 142
piezo-active, 17, 24, 25, 29, 30,
35, 43, 47, 54, 58, 61, 62, 63,
73, 76, 79, 87, 88, 89, 91, 101,
110, 116, 117, 131, 132, 134,
142, 155, 161, 164, 170, 172,
176, 185
piezo-passive, 5, 22, 24, 26, 28,
43, 47, 60, 61, 63, 65, 66, 78,
79, 81, 88, 110, 113, 116, 117,
132, 142
tensor, 25, 26, 65
vector, 25, 26, 60
composite, 1, 2
architecture, 92, 185
cable, 63
ceramic / ceramic, 29
ceramic / polymer, 7, 24, 27, 29,
30, 34, 35, 43, 45, 46, 48, 52,
53, 54, 61, 62, 63, 64, 65, 66,
70, 71, 72, 73, 75, 76, 77, 81,
83, 85, 86, 87, 89, 90, 92, 93,
104, 124, 132, 132, 133, 140,
141, 151, 160, 164, 165, 168,
169, 171, 172, 173, 175, 182,
185
classification, 4, 5
film, 63
high-performance, 154
interpenetrated, 4, 85, 87, 116
matrix, 4, 6
micrographs, 7, 8
modified layered, 110
piezo-active, see piezo-composite
piezoelectricpiezomagnetic, 29
porous, 102, 103, 111, 114, 115
single crystal / ceramic, 5, 29,
125, 126, 132, 140, 160
single crystal / polymer, 35, 142,
143, 150, 151
structure, 4, 5, 26, 29, 30, 31, 32,
Index 197
45, 46, 53, 54, 60, 63, 76, 77,
78, 81, 82, 85, 86, 87, 89, 110,
111, 115, 160, 162, 165, 173,
182
chain-like, 54, 63
complicated, 4, 32, 85
fibrous, 4
granular, 4
irregular, 5
laminar, 4
periodic, see regular
regular, 5, 26, 28, 30, 34, 51,
54, 63, 64, 82, 87, 88, 111,
132, 142, 150, 152, 159, 160,
165, 168, 171, 173
stochastic, 5, 26
three-component, 6, 110, 133, 151
two-component, 5, 30, 35, 43, 61,
159
classification, 5, 6
03, 5, 6, 7, 22, 26, 27, 28,
29, 30, 62, 63, 64, 65, 66,
68, 70, 71, 72, 73, 75, 76,
77, 78, 79, 80, 81, 82, 83,
84, 85, 86, 92, 93, 125,
126, 129, 131, 132, 133,
140, 154, 160, 161, 163,
164, 175, 176, 178, 180,
181, 185, 187
11, 30, 31, 32, 34, 93
13, 6, 7, 22, 24, 27, 29, 30,
31, 34, 52, 53, 54, 55, 57,
58, 59, 60, 61, 62, 63, 65,
75, 76, 77, 78, 81, 82, 83,
86, 89, 92, 93, 110, 111,
113, 116, 118, 131, 133,
134, 140, 142, 145, 149,
150, 151, 152, 160, 162,
163, 164, 165, 167, 168,
169, 171, 172, 173, 182,
185
formulae, 188
longitudinal, 167
transverse, 168
22, 8, 26, 27, 30, 31, 32, 34,
43, 44, 45, 46, 47, 48, 50,
52, 53, 57, 65, 78, 110,
110, 111, 116, 160, 161,
162, 168, 182, 185
formulae, 187
parallel-connected, 44, 45,
46, 48, 49, 50, 51, 57
series-connected, 43, 44,
45, 48, 51, 52, 65, 78,
110, 160
composition structure properties,
1, 2
conductivity, 24, 34, 51, 52, 64
connectivity, 2, 4, 5, 6, 22, 26, 27, 30,
31, 32, 35
changes, 30, 32
mixed, 29, 75
pattern, 5, 6, 26, 30, 31, 32, 33,
34, 35, 75, 82, 92, 102, 104,
118, 185
evolution, 30, 31, 32, 35
junction, 30, 31
DE, 5, 6, 30
DEJ, 6
composite in composite, 133
constituent, 25
copolymer, 72, 73, 78, 79, 80
coupled effect, 29, 159
cracking, 105, 109
Curie group, 2
cymbal, 46
dependence of properties, 22, 24, 43,
49, 52, 61, 73, 79, 101, 104, 106,
107, 108, 126, 143
monotonic, 22, 24, 165
non-monotonic, 24, 43, 45, 49, 52,
57, 61, 67, 70, 71, 87, 94, 124,
126, 131, 132, 134, 153
dielectric permittivity, 12, 15, 18, 19,
21, 22, 24, 25, 28, 34, 43, 46, 48,
49, 52, 54, 57, 67, 76, 78, 83, 86,
87, 88, 91, 93, 101, 107, 109, 133,
134, 143, 144, 149, 161, 163, 164,
165, 168, 172, 174
normalised, 75, 107, 109
differential scheme, 29, 159
dilute concentration limit, 29, 159,
175, 176, 182
198 Index
dilute concentration solution, see
dilute concentration limit
domain, 12, 15, 16, 17, 31, 67
reorientations, 15, 16
structure, 3, 16, 17, 18, 30, 34,
123, 142
engineered, 3, 18, 124, 142,
185
180q, 16, 17
non-180q, 16, 17, 123
domain-wall displacements, 12, 16,
17
effective medium, 16, 17, 159, 169,
175, 176, 178, 181
eigenstrain, 28
elastic compliance, 13, 15, 16, 28, 34,
46, 51, 67, 82, 83, 87, 88, 89, 118,
140, 143, 149, 168, 172
hydrostatic, 168
elastic modulus, 12, 13, 19, 21, 22,
25, 43, 54, 57, 66, 67, 68, 70, 78,
101, 133, 140, 141, 151, 153, 165,
168, 172
elastomer, 62, 84, 111, 112, 113, 115,
116, 140, 152
electric conductivity, see conductivity
electric displacement, 3, 11, 13, 43,
47, 142
electromechanical constants, 14, 15,
16, 17, 18, 20, 22, 25, 27, 28, 29,
34, 46, 47, 51, 57, 61, 65, 67, 75,
76, 78, 81, 82, 101, 104, 105, 111,
124, 126, 142, 148, 150, 159, 160,
165, 167, 171, 172, 173, 175, 182
approximation for porous piezo-
ceramics, 103
formulae for 13 composites, 188,
189
formulae for 22 composites, 187,
188
scheme for determining, 167
electromechanical coupling factor,
15, 24, 43, 45, 52, 66, 77, 101,
124, 125, 146, 153, 154, 155, 169
hydrostatic, 168
thickness, 43, 45, 52
electromechanical interaction, 3, 6, 7,
16, 27, 29, 34, 63, 78, 81, 104,
109, 113, 154, 159, 164, 175, 176
electrostriction, 11, 24
epoxy, 7, 10, 86, 131, 150, 151, 152,
177, 180
Spurr, 8, 151, 152
Eshelby tensor, 28, 29, 30, 57, 65,
133, 159, 160, 175, 176, 178, 182
experimental data, 3, 12, 16, 18, 19,
21, 22, 48, 60, 67, 70, 71, 72, 73,
74, 76, 78, 79, 80, 81, 86, 103,
104, 105, 106, 107, 108, 111,
113, 115, 148, 150, 160, 178
ferroelectric, 1, 2, 3, 5, 6, 12, 14, 15,
16, 17, 18, 22, 24, 25, 29, 31, 34,
35
relaxor, see relaxor-ferroelectric
fibre, 4, 7, 52, 165
field, 12, 13, 15, 16, 24, 27, 28, 29,
34, 47, 48, 52, 64, 105, 132, 161,
162
coercive, 12, 15, 124
effective, 28, 54, 63
elastic, see mechanical field
electric, 2, 3, 11, 12, 13, 14, 15,
16, 18, 24, 28, 34, 43, 47, 48,
52, 53, 61, 63, 105, 109, 124,
142, 150, 175
electroelastic, 63, 159, 175
mechanical, 3, 6, 13, 15, 18, 25,
27, 28, 29, 46, 47, 48, 52, 61,
67, 71, 86, 110
figure of merit, 45, 52, 61, 62, 83, 90,
92, 101, 146, 154, 169
hydrostatic, 52, 83, 90, 92,
146, 169
fixture, 53
heterophase, 3, 5, 18, 26, 123, 124
hierarchy-of-properties chain, 3, 17
homogenisation, 26, 165, 167
Hookes law, 12
inclusion, 4, 5, 6, 26, 27, 28, 29, 30,
52, 63, 64, 65, 66, 67, 68, 70, 71,
Index 199
73, 75, 76, 77, 80, 81, 82, 83, 84,
101, 105, 108, 109, 110, 125, 126,
127, 128, 129, 131, 132, 133, 134,
140, 159, 160, 173, 175, 176, 177,
182
aligned, 27, 28, 53, 54, 63, 64, 71,
109, 125, 150, 168
cylindrical, 133
distribution, 27, 28, 30, 64, 77
ellipsoidal, 28, 30, 64
fibre-like, 83
hierarchy, 75, 77, 81
hybrid, 63
needle-shaped, 128, 176, 182
oblate, 159, 160
parallelepiped-shaped, 27, 63, 75,
82, 142, 150, 165, 168, 169,
173, 174
penny-shaped, 28, 30, 65
plate-like, 4, 83
problem, 28
prolate, 63, 67, 70, 71, 73, 75, 81,
85, 140, 159, 162
rod-shaped, 52, 65
shape, 26, 27, 28, 30, 52, 53, 64,
128, 131
spherical, 63, 67, 70, 75, 76, 77,
104, 162, 176, 182
spheroidal, 27, 28, 29, 63, 64, 66,
70, 71, 81, 125, 159, 160, 175,
176, 177, 180, 182
interface, 1, 5, 18, 28, 30, 31, 32, 35,
46, 47, 48, 52, 61, 64, 82, 104,
125, 160
isotropic, 28, 47, 65, 104, 160
layer, 4, 8, 26, 30, 31, 33, 34, 43, 45,
46, 47, 48, 49, 51, 52, 76, 77, 78,
81, 93, 110, 111, 159
parallel-connected, 44, 45, 46, 49,
50, 51, 57
series-connected, 43, 44, 45, 48,
51, 52, 65, 78, 110, 160
03-type, 76, 77, 78
13-type, 77, 78, 110, 111
long-wave approximation, 47, 143,
160
manufacturing, 3, 18, 27, 34, 43, 45,
52, 53, 62, 63, 82, 86, 92, 93, 102,
103, 105, 108, 109, 125, 142, 185
method, see method for
manufacturing
porous material, 102, 103, 105,
108
03 composite, 63, 93
11 composite, 93
22 composite, 45, 46, 93
material, 1, 2, 4, 5
electrostrictive, 1
foam-based, 109
heterogeneous, 1, 24, 25, 26, 28,
32, 101, 104
intelligent functional, 1
magnetostrictive, 1
microinhomogeneous, 1
porous, 101, 102, 103, 104, 105,
106, 107, 108, 109, 110, 111,
112, 113, 114, 115, 116, 118
smart, 1
matrix, 4, 5, 26, 28, 30, 53, 54, 57,
58, 59, 61, 62, 63, 64, 65, 70, 75,
76, 77, 78, 81, 82, 83, 86, 102,
104, 110, 111, 113, 115, 116, 125,
132, 142, 145, 150, 151, 152, 153,
154, 159, 160, 165, 168, 173, 175
piezo-active, 28, 57, 78, 104
piezo-passive, 28, 61, 65, 66, 78,
81, 110, 113, 132
matrix approach, see matrix method
for evaluations
measurement conditions, 13
mechanical strain, 3, 11, 12, 13, 14,
24, 25, 28, 43, 46, 47, 124, 142,
154, 175
mechanical stress, 3, 11, 12, 13, 14,
15, 18, 28, 47, 61, 88, 89, 129,
131, 142, 149, 162, 176
method for evaluations, 16, 18, 25,
26, 27, 28, 29, 30
analytical, 22, 51, 54, 159, 171
asymptotic homogenisation, 165,
170
200 Index
conditional moments, 29
differential medium theory, 168
dilute approach, see dilute
concentration limit
effective field, 28, 29, 57, 63, 64,
77, 104, 126, 133, 138, 159,
160, 161, 163, 164, 171, 172,
176, 178, 180, 182
effective medium, 16, 17, 28, 67,
159, 175, 176, 178, 181
equivalent inclusion, 28, 63
finite element, 26, 29, 34, 159,
164, 165, 168, 171
matrix, 27, 30, 34, 46, 82, 142,
159, 160, 168, 169, 173
Mori Tanaka, 28, 63, 104, 159,
175, 176
regularisation of structure, 26, 159
self-consistent, 28, 67, 159, 175,
176
method for manufacturing, 93
aerogel process, 102
burned-out plastic spheres, 86, 93,
102, 103
co-extrusion, 45, 53, 93
coral replamine, 102
dice and fill, 45, 53, 93, 142
dielectrophoresis, 53, 63, 93
distorted reticulated ceramics, 86,
93
hot-isostatic pressing, 102, 109
hot rolling, 63, 93
injection moulding, 45, 53, 93
jet machining, 45, 93
lost mould, 53, 93
lost wax, 86, 93
pick and place, 53, 93
polymeric sponge, 102
rapid prototyping, 45, 53, 82, 86,
93, 185
relic processing, 86, 93
replamine, 102
robocasting, 86, 93
tape casting, 102, 125
tape lamination, 45, 53, 93
uniaxial pressing, 105
microgeometry, 3, 5, 25, 30, 43, 45,
53, 63, 75, 82, 83, 85, 90
planar, 82, 142, 168
pore, 101, 102, 105
microstructure, 2, 3, 5, 7, 17, 18, 22,
63, 75, 76, 102, 103, 104, 105,
108, 118, 125
model, 17, 29, 63, 88, 104, 108, 110,
115, 167, 168, 171
Banno cube, 26, 104
finite element, 75, 105, 168
Kar-Gupta and Venkatesh, 167
Levassort et al., 104
Marutake, 104
modified layered composite, 110
Wagner, 104
03 composite, 64, 76, 82, 125,
160
moony, 46
morphotropic phase boundary, 3, 5,
18, 101, 123, 124, 142, 144, 154
non-monotonic behaviour, see non-
monotonic dependence of
properties
pearl chains, 7
perovskite, 3, 15, 16, 18, 62, 66, 102,
123
unit cell, 124
piezo-composite, 1, 2, 3, 5, 6, 11, 22,
24, 25, 26, 27, 28, 29, 30, 32, 34,
35
piezoelectric, 1, 2, 3, 5, 6, 11, 12, 13,
14, 15
activity, 3, 43, 52, 60, 86, 109,
116, 123, 124, 125, 133, 152,
153, 159, 162, 176, 185
hydrostatic, 109, 152, 153
charge coefficient, 3, 13, 189
coefficient, 3, 12, 13, 14, 15, 16,
17, 18, 19, 21, 22, 23, 25, 28,
34, 43, 44, 46, 47, 48, 49, 50,
51, 52, 54, 57, 60, 61, 62, 65,
66, 67, 68, 70, 71, 72, 73, 75,
78, 82, 83, 85, 87, 88, 92, 101,
102, 108, 110, 115, 116, 117,
118, 124, 125, 126, 127, 128,
Index 201
129, 130, 131, 133, 134, 140,
141, 143, 149, 151, 154, 160,
162, 164, 168, 171, 172, 172,
173, 176
hydrostatic, 45, 52, 61, 70, 89,
92, 109, 117, 126, 132,
147, 171
normalised, 58, 59, 71, 73, 75,
76, 106, 111, 112, 114,
115, 116, 126, 131, 177,
180, 181
equations, 12, 13
effect, 2, 11, 12, 16, 29
direct, 11, 13, 14, 117
converse, 11, 13, 14
response, 6, 11, 48, 66, 67, 75, 80,
89, 101, 132, 152, 153, 185
sensitivity, 52, 57, 61, 70, 71, 83,
85, 86, 116, 150, 153
strain coefficient, 13
stress coefficient, 13
texture, 2, 3
voltage coefficient, 13
piezoelectricity, 3
piezorubber, 63
point group, 2, 124
polarisation, 11, 12, 13, 64, 105, 123
remanent, 16, 64, 111, 125, 132
spontaneous, 2, 15, 16, 17, 124,
132, 143
poling direction, 32, 46, 48, 78, 87,
105, 109, 117
polyethylene, 66, 111, 152
polymer, 3, 5, 7, 22, 27, 43, 45, 46,
47, 48, 52, 53, 54, 62, 63, 64, 65,
66, 67, 71, 76, 77, 78, 79, 82, 83,
86, 87, 88, 89, 92, 93, 105, 110,
111, 116, 117, 132, 140, 141, 142,
143, 154, 173, 185
porous, 62, 111, 150, 154
polyurethane, 53, 70, 71, 72, 73
pore-forming agent, 102
porosity, 101, 102, 103, 104, 105,
108, 109, 110, 113, 115, 116, 150,
152, 154
pattern, 104
properties, 1, 2, 3
combination, see combination
properties
diagram of changes, 57, 58, 59
dielectric, see dielectric
permittivity
effective, 3, 7, 17, 22, 23, 24, 26,
28, 29, 30, 32, 34, 35, 47, 51,
52, 57
variational bounds, 171, 173,
174, 176
Bisegna and Luciano, 171
Hashin and Shtrikman,
171, 172
13 composite with
cylindrical rods, 172, 173
13 composites with
parallelepiped-shaped rods,
173, 174, 175
elastic, see elastic compliance and
elastic modulus
electromechanical, 14, 15, 16, 17,
18, 25, 26, 27, 28, 29, 32, 34
ferroelectric, 2, 3
modelling, see model
physical, 1, 2, 22
piezoelectric, see piezoelectric
coefficient
product, 22, 24
sum, 22, 23
pyroelectric coefficient, 12
redistribution of fields, 3, 61, 70, 83,
110
relaxor-ferroelectric, 1, 3, 18, 123,
124, 125, 133, 142, 150, 151, 152,
153, 154, 185
representative volume element, 26,
27, 29, 30
response actions, 12
rod, 28, 30, 52, 53, 54, 57, 61, 76, 77,
78, 86, 93, 110, 111, 113, 131,
132, 133, 140, 141, 142, 143, 149,
150, 151, 152, 154, 159, 160, 164,
165, 168, 171, 172, 173
finite, 77, 78
infinite, 77, 78
rod-like element, 77, 78
202 Index
single crystal, 2, 5, 11, 12, 14, 15, 16,
17, 18, 22, 25
acentric, 2, 11, 12
ferroelectric, 2, 3
(1 x)Pb(Mg
1/3
Nb
2/3
)O
3

xPbTiO
3
, 123, 124, 125, 126, 142,
143, 144, 148, 149, 150, 151,
152, 153, 154
(1 y)Pb(Zn
1/3
Nb
2/3
)O
3

yPbTiO
3
, 123, 124, 125, 126, 142,
143, 144, 148, 149, 154
solid solution, 3, 18, 62, 66, 67, 123,
124, 154
spatial distribution, 2, 29, 125, 159
structure
cellular, 110
chain-like, 53, 63
hierarchy of inclusions, 75, 77, 81
irregular, 5, 63
negative, 53
regular, 5, 16, 26
structure processing property, 34
synthesis, 2, 32
thermodynamic function, 11, 12
threshold value, 15
transversely isotropic, 28, 47, 54, 61,
65, 72, 104, 160, 187, 188
vinylidene fluoride and
trifluoroethylene, 72, 73, 78, 79
volume fraction, 2
dependence, see dependence of
properties
gradient, 8, 46

Arama Motoru Optimizasyon Baslangic Rehberi 2

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Arama Motoru Optimizasyon Baslangic Rehberi 2

Загружено:

Авторское право:

Доступные форматы

Engineering Materials and Processes

Вам также может понравиться