Elastic And Mechanical Properties Of Expanded Perlite And ...
Transcript of Elastic And Mechanical Properties Of Expanded Perlite And ...
Elastic And Mechanical Properties
Of Expanded Perlite And
Perlite/Epoxy Foams
Haleh Allameh Haery
MIEngDes (UPM)
This thesis is submitted for the degree of
Doctor of Philosophy (Mechanical Engineering)
University of Newcastle, Australia
Faculty of Engineering and Built Environment
School of Engineering
February 2017
Statement of Originality
This is to certify that the thesis entitled “Elastic and Mechanical Properties of Expanded
perlite and Perlite/Epoxy foams” submitted by Miss. Haleh Allameh Haery contains no
material which has been accepted for the award of any other degree or diploma in any
university or other tertiary institution and, to the best of my knowledge and belief,
contains no material previously published or written by another person, except where due
reference has been made in the text. I give consent to this copy of my thesis when
deposited in the University Library, being made available for loan and photocopying
subject to the provisions of the Copyright Act 1968.
Haleh Allameh Haery
Statement of Collaboration
I hereby certify that the work embodied in this thesis has been done in collaboration with
other researchers. I have included as part of the thesis a statement clearly outlining the
extent of collaboration, with whom and under what auspices.
Statement of Authorship
I hereby certify that the work embodied in this thesis contains published papers of which
I am a joint author. I have included as part of the thesis a written statement, endorsed by
my supervisor, attesting to my contribution to the joint publications.
Haleh Allameh Haery
Dedication
I dedicate this thesis to my parents, who have always loved and supported me
unconditionally.
Acknowledgments
I would like to express my sincere gratitude to my supervisors, Professor Erich Kisi and
Doctor Thomas Fiedler, for their help and support throughout my period of PhD study. I
would like to take this opportunity to express my appreciation to my supervisor Erich,
whose patience, kind words, complete support and professional advice have been a source
of encouragement and a motivation to explore my ideas to their fullest depth. I would also
like to thank Doctor Jubert Pineda for his mentorship and constructive advice which has
helped me enormously.
I would like to take this opportunity to express my appreciation to the University of
Newcastle for granting me IPRS and Postgraduate Research scholarships during the four
years of my study.
I would also like to gratefully acknowledge the assistance from the staff in the electron
microscope and X-ray unit at the University of Newcastle, Australia.
Last, not the least, I want to thank my parents for their love, support and words of
encouragement throughout studies.
List of Publications and Awards
Awards
1. 2016 Three Minutes Thesis Finalist (Faculty of Engineering and Built
Environment); (University of Newcastle, Australia).
2. Runner Up - 2016 Three Minutes Thesis Faculty competition (Faculty of
Engineering and Built Environment); (University of Newcastle,
Australia).
3. Faculty of Engineering and Built Environment 2015 postgraduate research
prize, Mechanical Engineering (University of Newcastle, Australia).
4. Faculty of Engineering and Built Environment 2014 postgraduate research
prize, Mechanical Engineering (University of Newcastle, Australia).
5. Runner Up - 2015 Three Minutes Thesis Faculty competition (Faculty of
Engineering and Built Environment); (University of Newcastle,
Australia).
6. 2015 Three Minutes Thesis Finalist (Faculty of Engineering and Built
Environment); (University of Newcastle, Australia).
Conference (Poster)
Haleh Allameh-Haery, Erich Kisi, Thomas Fiedler, “Characterization of perlite-epoxy
foam cores for sandwich panels”, International Conference on Advances in Functional
Materials", 29 June-3 July, 2015, Long Island, NY, USA.
Conference Papers
H. A. Haery, and H. S. Kim, "Damage of hybrid composite laminates." SPIE Proceedings,
Volume 8793, Aerospace Composites, 87931F-87931F-13 (2013)
Journal Papers
1. H. Allameh Haery, H. S. Kim, R. Zahari, E. Amini, Tensile strength of notched
carbon/glass/epoxy hybrid composite laminates before and after fatigue loading.
Journal of Industrial Textiles, 2014, 44(2): p. 307 – 331.
2. H. Allameh Haery, and H. S. Kim, Damage in hybrid composite laminates.
Journal of Multifunctional Composites, 2013, 1(2): p.127-137.
3. H. Allameh Haery, E. Kisi, and T. Fiedler, Novel cellular perlite-epoxy foams:
effect of density on mechanical properties. Journal of Cellular Plastics, Doi:
10.1177/0021955x16652110, 2016.
4. H. Allameh-Haery, C.M. Wensrich, T. Fiedler, and E. Kisi, Novel Cellular
perlite-epoxy foams: effects of particle size. Journal of Cellular Plastics, Doi:
10.1177/0021955X16670528, 2016.
5. H. Allameh-Haerya, E. Kisi, J. Pineda, L. P. Suwal, T. Fiedler, Characterization
and prediction of Elastic properties of green compacts of Expanded perlite
particles. Journal of Powder Technology, 25 January 2017.
6. H. Allameh-Haery, E. Kisi, J. Pineda, L. P. Suwal, T. Fiedler, Elastic properties
of perlite-epoxy foams. (Under review).
Table of Contents
Statement of Originality .................................................................................................... 2
Statement of Collaboration ............................................................................................... 3
Statement of Authorship ................................................................................................... 4
Dedication ......................................................................................................................... 5
Acknowledgments ............................................................................................................. 6
List of Publications and Awards ....................................................................................... 7
Conference (Poster) ........................................................................................................... 7
Conference Papers ............................................................................................................. 8
Table of Contents .............................................................................................................. 9
List of Figures ................................................................................................................. 13
List of Tables................................................................................................................... 20
List of Symbols and Abbreviations .................................................................................... I
Abstract ............................................................................................................................. 1
1 Chapter One: Introduction......................................................................................... 4
2 Chapter Two: Literature Review ............................................................................... 9
Introduction ........................................................................................................ 9
The characteristics, manufacturing processes and applications of Perlite ....... 10
2.2.1 Introduction to perlite ................................................................................ 10
2.2.2 Expansion process ..................................................................................... 12
2.2.3 The physical properties of perlite .............................................................. 16
2.2.4 Chemical composition ............................................................................... 19
2.2.5 Applications .............................................................................................. 24
2.2.6 Concluding Remarks ................................................................................. 41
Foams ............................................................................................................... 42
2.3.1 Introduction to foams and syntactic foams ............................................... 42
2.3.2 Manufacturing syntactic foams ................................................................. 46
2.3.3 Particulate composites containing naturally occurring fillers ................... 50
Motivation and Problem Statement .................................................................. 59
Research Objectives and Research Significance .............................................. 60
3 Chapter Three: Methodology .................................................................................. 61
Introduction ...................................................................................................... 61
Material ............................................................................................................ 62
3.2.1 Expanded perlite particles ......................................................................... 62
3.2.2 Epoxy resin ............................................................................................... 63
Sample preparation ........................................................................................... 64
3.3.1 Preparation of expanded perlite (EP) particle samples ............................. 64
3.3.2 Preparation of solid perlite samples .......................................................... 67
3.3.3 Preparation of resin samples ..................................................................... 67
3.3.4 Preparation of EP/epoxy foam samples .................................................... 70
Experimental Setup and Tests .......................................................................... 73
3.4.1 Mechanical testing on the packed beds of EP Particles ............................ 73
3.4.2 Mechanical testing of EP/epoxy foams ..................................................... 73
3.4.3 Elastic wave tests on EP particles ............................................................. 75
3.4.4 Elastic wave tests on EP/epoxy foams and solid perlite ........................... 79
Microstructural analysis ................................................................................... 81
Damage observation ......................................................................................... 82
The theory of dynamic moduli measurement ................................................... 83
4 Chapter Four: Results .............................................................................................. 86
Introduction ...................................................................................................... 86
Elastic properties of sintered solid perlite ........................................................ 88
Properties of packed beds of EP particles ........................................................ 91
4.3.1 Structural characterisation of EP particles ................................................ 91
4.3.2 Measurement of elastic moduli of packed EP particle beds using quasi-static
mechanical tests ...................................................................................................... 96
4.3.3 Measurement of elastic moduli of packed EP particle beds using elastic
waves 98
Mathematical models for prediction for Elastic properties of porous bodies 107
4.4.1 Phani Models ........................................................................................... 109
4.4.2 Nielson Model ......................................................................................... 113
4.4.3 Minimum solid area (MSA) models ....................................................... 117
4.4.4 Gibson and Ashby Model ....................................................................... 121
Properties of epoxy resin ................................................................................ 124
Properties of EP/epoxy foams ........................................................................ 126
4.6.1 Microstructural characterisation of EP/epoxy foams .............................. 126
4.6.2 Volume fraction of the epoxy in EP/epoxy foams .................................. 128
4.6.3 Compressive response of EP/epoxy foams ............................................. 133
Damage analysis ............................................................................................. 141
Measurement of the elastic moduli of EP/epoxy foams using elastic wave speed
147
5 Chapter Five: Discussion ...................................................................................... 159
Introduction .................................................................................................... 159
Manufacturing method ................................................................................... 160
Young’s modulus of packed EP particles ....................................................... 163
Structural characterisation of EP/epoxy foams .............................................. 167
Compressive behaviour and compressive properties ..................................... 172
Damage mechanisms under compressive loading .......................................... 178
6 Chapter Six: Summary .......................................................................................... 182
Conclusions .................................................................................................... 182
Future Research .............................................................................................. 186
References ..................................................................................................................... 189
List of Figures
Figure 1.1. Schematic diagram illustrating the structure and composition of syntactic
foams. ................................................................................................................................ 4
Figure 2.1. (a) Unexpanded perlite (b) Expanded perlite. Both types were supplied by
Ausperl. ........................................................................................................................... 11
Figure 2.2. Horizontal expander and handling system [30]. ........................................... 15
Figure 2.3. Vertical expander and handling system [30]. ............................................... 15
Figure 2.4. Cellular structure formed internally in expanded perlite particles. .............. 17
Figure 2.5. US expanded perlite use by application....................................................... 24
Figure 2.6. Spherical gas tank farm in a petroleum refinery [116]. ................................ 32
Figure 2.7. SEM image showing EP microspheres consisting of one or more microcellular
bubbles [27]. .................................................................................................................... 37
Figure 2.8. (a) Closed cell Polyurethane foam [102], (b) Open cell Polyurethane foam [3].
......................................................................................................................................... 43
Figure 3.1. Durometer (Type D) for hardness test. ........................................................ 63
Figure 3.2. Tapping device for measuring tapped density of EP particles. .................... 65
Figure 3.3. Prepared mould (a) The different parts of the mould (b) Assembled mould.
......................................................................................................................................... 66
Figure 3.4. Solid perlite sample (sintered perlite powders). ........................................... 67
Figure 3.5. Mould for manufacturing epoxy resin samples for (a) compression and
flexural tests and (b) tensile tests. ................................................................................... 68
Figure 3.6. A hand-made device for separating expanded particles from unexpanded and
semi-expanded particles. ................................................................................................. 71
Figure 3.7. Formation of two phases by buoyancy. The top phase is packed perlite
particles and diluted binder and the bottom phase contained only diluted binder. ......... 72
Figure 3.8. Samples prepared using three particle size ranges; from the left, 1 - 2 mm, 2 -
2.8 mm and 2.8 – 4 mm. ................................................................................................. 72
Figure 3.9. Schematic representation of the experimental set-up for measuring wave
velocity in packed beds of EP particles........................................................................... 75
Figure 3.10. Examples of (a) P-wave and (b) S-wave signals in low density (0.12 g/cm3)
EP particle compacts and (c) P-wave and (d) S-waves in high density (0.3 g/cm3) EP
particle compacts. ............................................................................................................ 78
Figure 3.11. Schematic representation of the ultrasonic experimental set-up for measuring
wave velocity in solid perlite and EP/epoxy samples. .................................................... 79
Figure 4.1. Microscope images taken from the polished surface of sintered perlite. The
polishing marks are designated on the picture in order not to be confused with porosity.
......................................................................................................................................... 88
Figure 4.2. SEM images showing the external structure of an EP particle: (a) in the 1 - 2
mm size range; (b) in the 2 - 2.8 mm size range; (c) in the 2.8 - 4 mm size range. ........ 92
Figure 4.3. SEM images showing the internal structure of an EP particle in the: (a) 1 - 2
mm size range; (b) 2 - 2.8 mm size range; (c) 2.8 - 4 mm size range; (d) 2 - 2.8 mm size
range, illustrating natural reinforcement inside EP particle cells. .................................. 93
Figure 4.4. Constrained modulus of packed EP particle beds as a function of compact
density. ............................................................................................................................ 97
Figure 4.5. The (a) Compression wave and (b) Shear wave velocities versus compact
density. ............................................................................................................................ 99
Figure 4.6. Formation of platy and fine particles as a result of the brittle crushing of cell
walls. ............................................................................................................................. 100
Figure 4.7. (a) Percentage of debris at each EP compact density. (b) Cumulative particle
size distribution for entire compacts (EP particles and debris). .................................... 101
Figure 4.8. (a) Particle density versus particle size. This graph also includes debris density
versus debris size; (b) Inter-particle porosity (excluding debris) versus compact density;
(c) Inter-particle space filled by debris versus compact density. .................................. 105
Figure 4.9. (a) Young’s modulus (E) of packed beds of EP particles versus compact
density. (b) Poisson’s ratio of packed beds of EP particles versus compact density. In both
graphs, the upper and right hand scales allow the normalised moduli versus porosity to
be read from the same graphs........................................................................................ 106
Figure 4.10. (a) Normalised Young’s modulus versus porosity (based on experimental
moduli); (b) Normalised Poisson’s ratio versus porosity (based on experimental moduli);
(c) Normalised Young’s modulus versus porosity (based on modified moduli); (d)
Normalised Poisson’s ratio versus porosity (based on modified moduli). ................... 112
Figure 4.11. (a) Shape factor versus compact porosity for the Phani and Neilson models
applied to both the experimental moduli and modified moduli; (b) Modified Phani and
Nielson models for Young’s Moduli as a function of porosity; (c) Modified Phani and
Nielson models for Poisson’s ratio as a function of porosity........................................ 116
Figure 4.12. Hardness versus time curve used to determine the curing period of the diluted
epoxy. ............................................................................................................................ 125
Figure 4.13. Changes in the viscosity of epoxy + hardener diluted with acetone versus
acetone content. ............................................................................................................. 125
Figure 4.14. Micrographs taken from the cross-sections of EP/epoxy foams with a density
of 0.15 g/cm3 made with EP particles in the size ranges: a) 1 - 2mm; b) 2 - 2.8mm; and c)
2.8 - 4mm. ..................................................................................................................... 127
Figure 4.15. (a) Foam density as a function of applied pressure. (b) Particle density as a
function of compaction pressure. .................................................................................. 130
Figure 4.16. Volume fraction of epoxy versus density of the foam. ............................. 131
Figure 4.17. Volume fraction of epoxy binder in EP/epoxy foams of type 2. .............. 133
Figure 4.18. Typical stress-strain curves for the different EP/epoxy foam densities of: (a)
Type 1; (b) Type 2; (c) Type 3. ..................................................................................... 135
Figure 4.19. Properties of manufactured foams of type 1 ( ⃣ ), type 2 (◇) and type 3 (∆):
(a) Maximum stress versus foam density; (b) Effective modulus versus foam density; (c)
Modulus of toughness versus foam density. ................................................................. 137
Figure 4.20. Confined modulus of EP particles ( ⃣ ) and confined modulus of the foams
(◇) as a function of the density of the packed bed of EP particles. ............................. 138
Figure 4.21. Predicted elastic moduli of EP particles using the Voigt and Reuss models.
To quantify the particles’ contributions to the stiffness of the foam, the effective elastic
modulus of the EP/epoxy foams as a function of foam density is presented. ............... 140
Figure 4.22. Schematic representation of failure in EP/epoxy foams. .......................... 141
Figure 4.23. Macroscopic images showing (a) A sample of type 1 compressed to a strain
of 12% (b) A typical remnant of the samples after the test. .......................................... 145
Figure 4.24. SEM images showing (a) Uncrushed EP particles in the wedge-like fractured
side of a failed sample; (b) Cell walls of EP perlite particles fractured along a plane; (c)
Crushed cells in the central region of a uniformly compressed sample. ....................... 146
Figure 4.25. Typical stress-strain curve for different EP/epoxy foam densities. .......... 147
Figure 4.26. The longitudinal wave and shear wave velocities versus foam density. .. 149
Figure 4.27. Young’s modulus of EP/epoxy foams versus foam density determined from
the elastic wave speed (○) and mechanical tests (∆); and Young’s modulus of the EP
particles versus foam density determined from the elastic wave speed (◇). ............... 149
Figure 4.28. Volume fraction of the epoxy binder in the second series of EP/epoxy foams
versus foam density. ...................................................................................................... 151
Figure 4.29. (a) stress-strain curves and (b) Young’s modulus measured from the gradient
of the unloading path in Figure 4.29 (a) at different stress level in cyclic compressive tests
conducted on a sample with density of 0.26 g/cm3. ...................................................... 155
Figure 4.30. Poisson’s ratio of epoxy resin (), foam (○) and packed beds of EP
particles ( ⃣ ) versus the foam density. .......................................................................... 157
Figure 5.1. Young’s moduli of the packed EP particle beds measured using mechanical
tests and estimated using the Voigt and Reuss models. ................................................ 165
Figure 5.2. (a) Volume fraction of EP particles and (b) Porosity within inter-particle space
(by consideration of debris) and total porosity in type 2 EP/epoxy foams. .................. 169
Figure 5.3. A schematic representation of the internal structure of (a) low density and (b)
high density EP/epoxy foams. ....................................................................................... 171
Figure 5.4. Schematic representation of compressive stress-strain curves for a) an
elastomeric, b) an elastic-plastic, c) a brittle foam [11]. ............................................... 172
Figure 5.5. Schematic representation of the compressive stress-strain curve for EP/epoxy
foams. ............................................................................................................................ 175
Figure 5.6. Compressive strength (a) and compressive modulus (b) plotted against density
for currently available foams (Ashby et al., 2000) and results obtained for perlite-based
foams ( )....................................................................................................................... 177
List of Tables
Table 2.1. Typical physical properties of expanded perlite [23, 26, 39]......................... 18
Table 2.2.Chemical composition of perlite (Percent) [22]. ............................................. 20
Table 2.3. Mineral phases detected in perlites of different origins. ................................ 23
Table 2.4. Examples of perlitic sound insulation boards in which starch is used as binder
[69]. ................................................................................................................................. 30
Table 3.1. Chemical composition [198] .......................................................................... 62
Table 3.2. ASTM D-695, ASTM D-638 and ASTM D-790 for measuring the
compressive, tensile and flexural properties of epoxy resin. .......................................... 68
Table 4.1. Elastic properties of solid component of EP particles. For comparison, the
corresponding data for obsidian by Manghnani et al. [213] are presented. ................... 90
Table 4.2. Bulk density and particle density measurements of EP particles ................... 91
Table 4.3. Cell dimensions of EP particles of 1 - 2 mm, 2 - 2.8mm and 2.8 - 4 mm size
range. Each value is an average of 150 measurements. .................................................. 95
Table 4.4. Exponents of Eq. (4.3) and Eq. (4.4) for the three particle size ranges. ........ 96
Table 4.5. Physical parameters of mathematical models applied to the experimental and
modified moduli. The physical parameters of the modified moduli are given in
parenthesis. .................................................................................................................... 111
Table 4.6. Mechanical properties of cured epoxy resin with and without dilution by
acetone........................................................................................................................... 124
Table 4.7. Coefficients in Equations (4.21) and (4.22) for foam density Df and particle
density Dp . .................................................................................................................... 131
Table 4.8. Elastic properties of EP particles, epoxy resin and EP/epoxy foams........... 150
I
List of Symbols and Abbreviations
Item Description
EP Expanded perlite
Cp Compression wave velocity
Cs Shear wave velocity
CL Longitudinal wave velocity
𝜈 Poisson’s ratio
𝜈𝑐𝑟 Critical Poisson’s ratio
𝜈0 Poisson’s ratio at zero porosity
E Young’s modulus
E0 Young’s modulus at zero porosity
E* Constrained elastic modulus
G Shear modulus
G0 Shear modulus at zero porosity
K Bulk modulus
σcomp Compaction pressure
𝜌 Density
𝜌0 Density at zero porosity
ρc Composite density
ρp Particle density
ρg Density of the glass
ρtrue True density of the microspheres
ρe Epoxy density
Df foam density as a function of compaction pressure
II
DP Particle density as a function of compaction pressure
tw Wall thickness of particle
r Average radius of the microspheres
C Cell size
dp EP particle diameter
𝜌𝐸𝑃 EP particle density
𝜌𝑈𝑃 Density of fully dense perlite
𝜑 Angle of internal friction
a Packing geometry dependent parameter
P Porosity
Pcr Critical porosity
tapP Porosity of tapped packing state
greenP porosities of as-poured packing state
0n Constant related to pore morphology
En Pore structure dependent parameter (shape factor)
𝛽 Pore structure dependent parameter (shape factor)
S Shape factor as a function of porosity
CSA Composite sphere assemblage
b Constant related to particle stacking
𝑏′ Constant related to particle stacking
M Young’s, shear or bulk modulus
C Constant of proportionality
C Constant of proportionality
Ø Volume fraction of solid in the cell- struts
III
UpperPE Upper bound of Young’s modulus of EP particle
LowerPE Lower bound of Young’s modulus of EP particle
σhyd Hydrostatic stress
σdev Deviatoric stress
δ Kronecker delta
λ, µ Lamé constants
1
Abstract
Syntactic foams are composite materials made by reinforcing a resinous matrix with
hollow particles called microspheres. These materials are often used as the core material
for sandwich panels, where a combination of low density, high compressive strength, high
compressive deformation and high damage tolerance are required, e.g. in the aerospace,
automotive and marine industries. Syntactic foams have superior mechanical and thermal
properties but they are more expensive and denser than conventionally gas-blown foams.
The higher density and cost of syntactic foams is mainly due to the density and cost of
the synthetically made microspheres. This problem can be mitigated by using light-weight
naturally occurring particles which provide syntactic foams with similar properties but
which are significantly cheaper. In this study, the potential of expanded perlite particles
(EP particles) in manufacturing light-weight syntactic foams is investigated. Perlite is a
glassy volcanic rock of silicic composition and in its expanded form has a high porosity
(>95%), low density (~ 0.18 g/cm3) and offers excellent thermal and acoustical insulating
properties, chemical inertness, physical resilience, fire resistance and water retention
properties. These features, along with the fact that it is abundant and cheap make it a
suitable candidate for manufacturing syntactic foams.
In this study, the structural, microstructural, physical and mechanical properties of EP
particles were investigated. The elastic properties of packed EP particle beds were
characterised by the isotropic elastic moduli Poisson’s ratio and Young’s modulus,
calculated from elastic wave speeds along the axial (compaction) direction for a wide
range of compaction densities. It was observed that during compaction to achieve
different densities, some crushing of particles into smaller particles and platy debris
2
occurred. Consequently, analyses were conducted based on both the raw compaction
densities and densities modified by the removal of debris from consideration, on the
assumption that debris is non-structural. Based on the raw compaction densities, Young’s
moduli of the packed EP particles were found to vary in the range 31.4 - 371.3 MPa, while
based on the modified densities, they varied in the range 31.4 - 152.8 MPa for the compact
densities ranging from 0.1 to 0.375 g/cm3. Poisson’s ratio of packed EP particles did not
show a large variation with compact density in the range 0.1 - 375 g/cm3; Poisson’s ratio
was about 0.3. The equation for Poisson’s ratio is independent of density, hence the values
obtained based on the experimental and modified densities resulted in the same Poisson’s
ratios.
Four analytical models were applied to predict the elastic moduli of packed beds of EP
particles within the porosity range 84 - 95%. Models were assessed on their ability to
successfully predict the elastic moduli of these highly porous bodies from the properties
of solid perlite for both cases: using the raw compaction density and the modified density.
It was found that the Wang (Minimum Solid Area) model was able to estimate Young’s
modulus, while the Gibson and Ashby model was reasonable for the average behaviour
of both of the elastic moduli. The best agreement, however, was obtained through the
Phani model utilising our modified shape factor.
The impact of the research was broadened by dispersing EP particles in a matrix of epoxy
resin and the manufacture of light-weight EP/epoxy foams. Foams were fabricated with
three distinct particle size ranges and, within each size range, the samples covered a
density range 0.15 - 0.45 g/cm3. The effects of particle size and foam density variations
on the compressive strength, effective elastic modulus and modulus of toughness of the
EP/epoxy foams were investigated. The compressive properties of the EP/epoxy foams
3
showed a strong dependence on the foam density, but were almost independent of the
particle size. The compressive strength of the EP/epoxy foams was found to vary linearly
in the range 0.15 - 1.77 MPa with the foam densities ranging from 0.15 to 0.45 g/cm3.
However, the Young’s modulus and modulus of toughness of the EP/epoxy foams varied
parabolically in the ranges 33 - 227 MPa and 0.01 - 0.35 MPa, respectively. The elastic
properties of the EP/epoxy foams were characterised by adopting an isotropic model for
the medium and measuring the elastic wave speeds (i.e. longitudinal and shear wave) in
the axial direction (similar to the measurements of the particle beds). Quasi-static
compressive test results were compared with those obtained by the elastic wave tests.
Both were observed to follow the same qualitative pattern, however the Young’s moduli
measured using elastic waves were more than twice those obtained from the mechanical
tests. Poisson’s ratio showed an increasing trend, ranging from 0.17 to 0.34 over the foam
density range, and appeared to be influenced by the increase in contact surface area
between the particles and the matrix as the foam density increased.
Post-test scanning electron microscopy (SEM), coupled with photogrammetry during the
tests, were used to understand the behaviour of the foams under compressive load. The
observations revealed the presence of three different failure modes for all of the foams,
regardless of their particle size and density, however the strain to activate each mode was
different for each foam type. In addition, the observations showed that the formation of
wedge-like fragments in the foam samples under applied compressive stress were due to
the effects of pure shear (i.e. the deviatoric components of the applied stress). However,
the compressive deformation due to the effect of the hydrostatic components of the
applied stress was concentrated in the middle of the foam samples.
4
1 Chapter One: Introduction
Foams are cellular solids where cells with solid ligaments and membranes pack together
in three dimensions to fill space. These materials are widely found in nature, and have
been in use for more than 5000 years, for example, the wooden artefacts found in Egypt’s
pyramids or cork which has been used for the soles of shoes since Roman times [1]. These
materials inspired engineers to synthesise light-weight, yet strong and stiff, cellular
materials for various structural applications. Syntactic foams are a special class of cellular
materials which are made by embedding hollow microspheres in a polymeric matrix or
binder. They are identified as foams due to their cellular structure created by void space
in both the polymer resin matrix and the hollow microspheres. The schematic structure
of syntactic foams is presented in Figure 1.1.
Figure 1.1. Schematic diagram illustrating the structure and composition of syntactic foams.
Syntactic foams were initially developed in the 1960s as a buoyancy aid material for deep
submergence applications where high hydrostatic pressures are involved [2]. They are
generally used as the core material in sandwich plates and shells as an effective weight-
saving design option for various structural applications [3], and in areas where a
5
combination of low density, high compressive strength, high compressive deformation
and low moisture absorption are required, e.g. in naval, aeronautical, aerospace, civil and
automotive applications [4-6].
There are a wide variety of hollow microspheres available for use in syntactic foams.
They are commonly made from glass, ceramic, carbon or any polymeric materials. The
appropriate choice of microspheres can produce resinous foams with superior strength,
stiffness, damage tolerance and chemical resistance. Preference is generally given to
hollow glass microspheres due to their low density (~ 0.28 - 0.7 g/cm3 [7-9]), mechanical
strength, stiffness, the regularity of the surface and their price, which is lower than other
microspheres [8]. However, compared with conventionally gas-blown foams, syntactic
foams have higher density and cost of production. This is mainly due to the cost of
synthetically made hollow spheres. Cost and density are two main factors in the selection
of materials, especially when production of weight-sensitive structural components in
large quantities is involved. Hence, it would be worthwhile to find a light-weight material
which has the potential to provide syntactic foams with similar specific properties, but
which is significantly cheaper than other commercially available microspheres. One
promising material is expanded perlite particles. Perlite is a glassy volcanic rock of
rhyolitic composition normally comprised of 71-75% SiO2, 12.5-18% Al2O3, 4-5% K2O,
1-4% sodium and calcium oxides, lesser amounts of several metal oxides and 2-5% water
by weight [10]. Upon rapid controlled heating in the range 760 - 1100°C, the combined
water in perlite grains is vaporised, producing pressure that causes expansion of the perlite
to 4 - 20 times its original volume. The expanded form of perlite has a low density (~ 0.18
g/cm3 [11]) and offers excellent thermal and acoustic insulating properties, chemical
inertness, physical resilience, fire resistance and water retention properties [12]. In
6
addition, expanded perlite particles have a cellular structure mainly consisting of closed
cells (sealed bubbles) which provide them with progressive crushing characteristics as
opposed to the one-step crushing of microspheres [13]. This can be beneficial to
preserving the cellular structure of the perlite-reinforced plastic/foam for a longer service
life.
The present study focuses on the development of a light-weight foam material made by
dispersing expanded perlite particles (hence called EP particles) in a matrix of epoxy
resin. The structural, microstructural, and mechanical properties of EP particles and
EP/epoxy foams are investigated. Consequently, the potential of EP/epoxy foams as a
novel class of engineering materials, especially for light-weight structural applications,
are evaluated.
The layout of the remainder of the thesis, a significant portion of which is based on articles
published by the author [14-16] while undertaking this study, are organised as follows:
Chapter 2 delivers an overview of the existing literature related to this work in two main
sections. The first section gives a comprehensive overview of the physical properties,
chemical properties, expansion process and applications of perlite particles. The second
section outlines an overview of foams, syntactic foams and polymeric composites
reinforced with different naturally occurring minerals, including perlite particles.
Subsequently, based on gaps in the current literature, the motivation for conducting the
current project and the key objectives are highlighted.
Chapter 3 includes a description of the various raw materials used in manufacturing
EP/epoxy foams, fabrication techniques and testing procedures (i.e. quasi-static
mechanical tests and elastic wave tests) used for characterisation of the mechanical and
7
physical properties of solid perlite (i.e. virtually pore free), packed beds of expanded
perlite particles, and the novel EP/epoxy foams manufactured.
Chapter 4, firstly, outlines the experimental results related to tests performed on solid
perlite, e.g. microscopy and elastic wave tests. Secondly, the experimental results relating
to microscopy and mechanical and elastic wave tests, on packed beds of EP particles are
discussed. Thirdly, the experimental results relating to mechanical and elastic properties
of cured epoxy resin are presented. Fourthly, four mathematical models for the prediction
of elastic properties of packed EP particle beds are introduced. The predictive ability of
the models is evaluated and modifications are made to adapt to the morphology of the
packed EP particles. Fifthly, the behaviour of the foams under quasi-static compressive
loading, together with the consequences of different parameters (i.e. particle size and
foam density) on the response of the foams, are investigated. Finally, the results of elastic
wave tests on EP/epoxy foams are presented and a comparison with those obtained by the
quasi-static tests is made.
Chapter 5 discusses the significance of the results in a broad context, including an
evaluation of the advantages and disadvantages of the manufacturing method used in this
study and possible alternative methods. Additional factors are discussed, such as: i) the
Young’s modulus of packed EP particle beds measured using the different methods; ii)
the interior structure of EP/epoxy foams based on the experimental results; iii) the
compressive behaviour and properties of EP/epoxy foams; and iv) the damage
mechanisms during compressive tests. Moreover, a comparison is made with other
commercially available foams, on the basis of their compressive properties, compressive
behaviour and damage mechanisms.
8
Chapter 6 presents conclusions which may be drawn from the study. It also provides
several suggestions for future work, which may be taken as a guide by students and
researchers entering this area or wishing to pursue improved EP/epoxy foam manufacture
and/or its properties.
9
2 Chapter Two: Literature Review
Introduction
This chapter is divided into four parts. The first part introduces the perlite material and
provides comprehensive information about the process of expansion, its physical and
chemical properties, as well as current commercial uses of perlite in three areas:
construction, horticulture and industry. In the second part of this chapter, a review is
conducted on previously developed polymer matrix foams and particulate composites
which contained naturally occurring fillers, including perlite particles. The final two parts
provide the motivation and main objectives of the current project based on gaps in the
literature presented in the first and second parts.
10
The characteristics, manufacturing processes and applications of
Perlite
2.2.1 Introduction to perlite
Perlite is a glassy volcanic rock of silicic or rhyolithic composition, typically formed by
the hydration of obsidian. Perlite occurs naturally in the form of block type-lava domes,
restricted in area, which are formed by the extrusion of highly viscous magmas. Perlite
deposits are restricted to Tertiary age, or younger, deposits of rhyolithic composition.
Rhyolithic glasses are unstable and devitrify1 with age into microcrystalline aggregates
of quartz and feldspar, and transform to zeolites and other aluminosilicate minerals.
Therefore, the preservation of perlite is rarely found in rocks older than Tertiary age [17].
The name perlite is a derivation from the German word perlstein and is originally given
to “certain glassy rocks (hyaloliparites, hyalo-rhyolites) characterized with numerous
concentric cracks”, the ‘perl’, referring either to the resemblance of the broken-out
fragments to pearls or to the pearly lustre of the surfaces [18]. Petrologically, the term
‘perlite’ refers to volcanic glasses in which cooling strain resulted in a concentric or onion
structure of fracturing which may be visible to the naked eye or may only be observed
under microscopes. This structure of fracturing is also known as perlitic [19].
Perlite is distinguished from other natural water-bearing glasses (e.g. obsidian, pumice
and pitchstone) by the total water content of 2 to 5 wt% held within the glass structure,
by the presence of a pearly lustre and by its onion skin perlitic fractures. The term,
1. The process in which a glass (noncrystalline or vitreous solid) transforms to a crystalline solid.
11
however, should be expanded to include the less dense textures of perlite which are related
to the classical variety but rarely, or only microscopically, exhibit perlitic fractures, and
which might not exhibit a pearly lustre. Textures of perlite may be commonly found in
deposits ranging from the dense and classically fractured variety to vesiculated and
pumiceous varieties which may not exhibit a pearly lustre or megascopic perlitic fracture
[20]. Commercially, the term ‘perlite’ has been applied to any naturally glass of igneous
origin that will expand or ‘pop’ when heated quickly, forming a light-weight frothy
material. However, this property is also present in some glassy rocks that do not have
perlitic structure, and in some that are not primary in origin but are alteration products of
rhyolite or obsidian [21].
Upon heating within its softening range of 760°C to 1100°C [10], the combined water in
perlite grains is converted to steam and the perlite expands 4 - 20 times its original
volume. Hence, it is called ‘expanded perlite’ (EP). Figure 2.1 illustrates the unexpanded
and expanded forms of perlite.
(a) (b)
Figure 2.1. (a) Unexpanded perlite (b) Expanded perlite. Both types were supplied by Ausperl.
12
Expanded perlite has low thermal conductivity, high sound absorption, high resistance to
heat, chemical inertness, physical resilience and water retention ability. More than half of
the perlite produced is consumed in the construction industry as aggregate in insulation
boards, acoustical ceiling tiles, plaster and concrete [17]. It is also used in horticulture
and other applications like filtration, e.g. in the pharmaceutical and food industries, and
as fillers in various processes and materials [22]. There have been many studies about the
use of perlite in particular industrial and construction applications (e.g. [23], [24], and
[25]). However, few articles have been published which review perlite’s properties and
its applications. Singh [26] conducted a review on the multifarious uses of perlite as a
construction material. Barker and Santini [27] provided an overview of the world perlite
deposits and activities. Therefore, this section of the chapter aims to provide a
comprehensive overview on the characteristics and properties of perlite, as well as its
applications. Hopefully, the information provided has the potential to rapidly fill the void
in our understanding of perlite.
2.2.2 Expansion process
In general, the perlite expansion process involves rapid heating in a kiln to a softening
point, and then rapidly cooling. The size to which the rock is crushed before heating
depends on the composition of the perlite (e.g. water content), furnace design and the end-
use for which the expanded perlite is required. Generally, perlite particles that pass 8-
mesh (2.380 mm) and are retained on 30-mesh (0.595 mm) are suitable to be used as
concrete aggregate; perlite particles that pass 14-mesh (1.410 mm) and are retained on
50-mesh (0.297 mm) are suitable to be used as plaster aggregate; and fine particles that
pass 30-mesh (0.595 mm) are used for the lightest types of expanded product [28]. A
13
typical kiln feed contains perlite grains with diameters from 0.254 to 2.54 mm. It is
customary to eliminate grains smaller than 25 µm, as they tend to produce very fine dust
particles in the expanded product [21].
At the stage of heating perlite to the softening point, the combined water is vaporised,
producing pressure that causes expansion of the perlite to 4 - 20 times its original volume
leading to a significant reduction of its bulk density [29]. The heating may last from one
to three minutes, at temperatures varying between 760°C and 1100°C. It can be carried
out in either a rotary horizontal expander or a stationary vertical expander as both operate
on the same principle. Figures 2.2 and 2.3 show schematics of horizontal and vertical
expanders, respectively. Detailed descriptions of these two types of expander can be
found in [30].
The heating rate for the expansion process may be optimised. If the perlite granules are
heated too slowly, the combined water is removed through pores without much of a
‘popping effect’, leaving a slightly porous but relatively dense product. However, if the
rate of heating is too high, the expansion of the water may be so violent that the perlite
particles shatter and form numerous fines [31]. The optimum temperature at which perlite
expansion occurs may depend on its chemical composition. Variations in the composition
of the glass affect the softening point, the type and degree of expansion, the size of the
bubbles and the wall thickness between them, as well as the porosity of the product [21].
Also, water content in perlite is another factor affecting the optimisation. It has been
found that perlite containing excessive combined water is likely to shatter upon
expansion, whereas perlite containing much less water is likely to produce higher density
products [32]. A typical value of 3.2% to 3.7% combined water in perlite seems to work
best in most conditions [27]. King et al. [33] reported that the water in perlite above 1.2%
14
is loosely held, and therefore is easily removable. However, as dehydration proceeds, the
residual water (below 0.5%) is increasingly more firmly held, which is the effective
portion in producing the expansion of perlite upon heating. Some form of pre-heating is
generally used to reduce the effect of the initial thermal shock, and thus to reduce the
amount of shattered material. It also drives off some of the loosely held water, brings the
larger particles closer to their softening point, and thus reduces the time spent in the kiln
in order to improve the fuel efficiency of the expander [34]. It also can lower the density
of the final product, hence improve the insulating ability of the product. For coarse grades,
in particular, preheating is important as these grades offer less surface area for heat
transfer in a given amount of time, while intermediate and fine grades are not usually
preheated [30].
Although it is possible to achieve some flexibility in the properties of the finished product
properties through adjustment of furnace conditions and through preheating, it should be
noted that all particles do not expand alike [27]. The characteristics of the final product,
such as size, strength and density, depend on the initial granule size, granule water content
and granule composition, together with the design and operation of the kiln and the
heating rate [31]. When expanded perlite is cooled down, it is air classified to separate
any unexpanded particles and fines. Depending on the intended end-use, the expanded
perlite may be further size classified, surface treated or milled. For example, some
expanded perlite is milled down to -100 mesh (0.15 mm) for use as a filter aid, especially
in rotary pre-coat filtration [12]. Micronised perlite particles up to 2 μm can also be
produced by milling for anti-block filler for polymeric film or as a reinforcing filler for
polymers [35].
15
Figure 2.2. Horizontal expander and handling system [30].
Figure 2.3. Vertical expander and handling system [30].
16
2.2.3 The physical properties of perlite
Perlite textures may vary from classical (onion skin-like) to granular and pumiceous [27].
Classical, or onion skin perlite is the densest of the perlite types and is characterised by
well-developed concentric fractures, a pearly-to-resinous lustre and a grey to bluish black
colour. Granular perlite has a sugary or saccharoidal appearance, is microvesicular2 and
highly fractured, with colours ranging from white to grey. Granular perlite is lighter than
classical perlite, but denser than pumiceous perlite. Pumiceous perlite is extremely
lightweight, white to light grey, frothy and commonly friable. All perlite types are mined,
however the most common commercial one is in granular form [36]. It was found that the
grain morphology affects the expansion ratio. Pumiceous perlite has a higher expansion
ratio than granular perlite, but a lower expansion ratio than classical (onion skin-like)
perlite. The granular perlite has a well distributed microvesicular texture that allows water
vapour to escape easily, and thus only a small degree of perlite expansion occurs,
considerably smaller than that of pumiceous and classical perlites. However, in classical
perlite, the water vapour has difficulty in escaping through the successive onion skin-like
layers. Therefore, the sudden vaporisation results in increasing pressure, thus blowing up
the grain until it explodes and greatly expanding the grain [37]. The light-weight porous
material with a cellular interior structure that perlite transforms into upon heating above
760°C, is shown in Figure 2.4.
2. Vesicular texture refers to the presence of small cavities called vesicles, which were originally gas
bubbles in the liquid magma. This gas was initially dissolved in the magma but has come out of solution
because of the pressure decrease during eruption or as the magma rose before eruption. The texture is often
found in extrusive aphanitic, or glassy, igneous rock [Claudia Owen et al., 2010].
17
Figure 2.4. Cellular structure formed internally in expanded perlite particles.
Unexpanded ("raw") perlite has a bulk density ranging from 1.04 to 1.2 g/cm3, while
expanded perlite typically has a bulk density of about 0.032 - 0.4 g/cm3. However, the
true density of perlite does not change as significantly as expected [38]. Expanded perlite
ranges from a fluffy highly porous (85 - 95 vol% [11, 39]) to glazed glassy particles
having a low porosity [19]. Porosity3 provides expanded perlite with a volumetric and
surface absorption capability. The porosity is an advantageous feature for drainage,
aeration and retaining moisture and fertilizers in horticultural uses and landscaping [40].
However, the water absorption in thermal insulation applications is not desirable as the
heat conductivity increases when the pores are filled with water [23]. The expansion
process also creates a brilliant white colour in perlite, which is due to the reflectivity of
the cellular structure. The brilliant white colour is also beneficial in light coloured, visible
3 Porosity is determined as the average ratio between the volume of pores and the total volume of the perlite
grains [Topçu Işikdağ, 2006].
18
surface coatings. Colour is also of concern in filler applications, especially when the
colour of the end product is important [27]. The typical physical properties of expanded
perlite are given in Table 2.1.
Table 2.1. Typical physical properties of expanded perlite [23, 26, 39].
Typical Physical Properties
Colour White
GE brightness % 78
Refractive Index 1.5
Free Moisture, Maximum 0.5%
pH (of water slurry) 6.5 - 8.0
Specific Gravity 2.2 - 2.4
Wet density (kg/m3) 103
Bulk Density (Loose) 0.032 - 0.400 g/cm3
Mesh Size Available 4 - 8 mesh and finer
Softening Point 871 - 1093°C
Fusion Point 1260 - 1343°C
Unit weight 2.2 - 2.4 g/cm3
Specific Heat 0.20 kcal/kg°C
Thermal Conductivity at 75°F (24°C) 0.04 - 0.06 W/m·K
Solubility - Soluble in hot concentrated alkali and HF
- Moderately soluble (<10%) in 1N NaOH
- Slightly soluble (<3%) in mineral acids
(1N)
- Very slightly soluble (<1%) in water or
weak acids
19
2.2.4 Chemical composition
Perlite is chemically inert, having an approximate pH of 7, and is mainly composed of
silica, alumina and lesser amounts of several metal oxides (sodium, potassium, iron,
calcium and magnesium). Table 2.2 gives the composition of perlite ores from different
countries. As can be seen, although perlite from different origins exhibits considerable
variation in composition, it is characterised by a high silica content (≥ 65%) and alumina
content (11 - 18%) with about 7 - 8% of alkaline content [41]. Burriesci et al. [41] found
that there are no traces of carbonate in perlite ore based on thermogravimetric analysis
[41] and another alumina silicate, tuff [42], and comparison between their weight loss
curves. Considering the linearity and flatness of the weight loss curve in the range 700 -
900°C, a temperature range at which the decomposition of carbonate occurs, they
considered that no carbonate should exist in perlite as, if it did, its presence should have
shown a reduction in weight as was shown for tuff. Although this assumption is valid,
referring to the weight loss curve for perlite in [41], the flatness is not well established.
Only three points in this range were used for both curves which does not seem satisfactory
as every line with three points appears linear.
In addition to the above-mentioned compounds, trace amounts of some other elements
were detected in perlite ores, which are usually less than 2% by weight. Typically, these
elements include Arsenic, Boron, Beryllium, Barium, Chlorine, Chromium, Copper,
Gallium, Lead, Lanthanum, Manganese, Molybdenum, Nickel, Niobium, Neodymium,
Sulphur, Titanium, Thorium, Vanadium, Yttrium, Zirconium and Zinc [28-32].
Furthermore, the presence of trace amounts of rare earth elements (La, Ce, Pr, Nd, Sm,
Eu, Gd, Dy, Ho, Er, Yb, Lu) were detected in the analysis of the chemical composition
20
of perlite deposits from Trachilas (Greece) [43], Eastern Rhodope (Bulgaria) [44],
Nathrop (Colorado) and No Agua (New Mexico) [45].
Table 2.2.Chemical composition of perlite (Percent) [22].
Components New Mexico
[29]
Greece
[36]
China
[37]
Turkey
[38]
Iran
[39]
Thailand
[40]
SiO2 74.10 76.10 76.89 68.40 72.32 75.20
Al2O3 13.30 12.16 10.51 15.11 12.62 12.75
Na2O 3.20 3.62 0.80 1.62 2.96 2.85
K2O 4.60 3.03 8.25 4.05 5.02 5.12
CaO 0.60 1.21 0.12 1.72 0.66 0.38
Fe2O3 0.50 1.19 2.48 2.76 0.67 0.99
MgO 0.10 0.27 0.06 0.42 0.21 <0.10
TiO2 0.05 0.14 0.07 n.d. n.d. 0.14
MnO n.d. 0.05 0.04 n.d. 0.66 0.05
P2O5 n.d. 0.03 0.03 n.d. 0.13 <0.05
SO3 0.10 n.d. n.d. n.d. n.d. n.d.
H2O n.d. n.d. 0.31 4.92 n.d. 0.43
LOIa n.d. 2.19 0.64 n.d. 3.75 1.50
a. LOI, loss of ignition.
The chemical composition and physio-chemical properties of perlite (e.g. index of
refraction, specific gravity, etc.) are comparable to those of obsidian and pumice.
However, perlite has a higher water content (generally about 3 - 4%) [41]. Water has an
important role in the expansion process, not only by causing expansion of the grain during
vaporisation, but also by reducing the viscosity of the grain [37]. It is reported that water
21
in perlite is in two forms, viz. molecularly dissolved water and hydroxyl groups. The ratio
of these two species of water is variable in perlite [17]. King et al. [33] made an analogy
between the molal entropy increase between 0 to 298 K in perlite and a number of
substances containing water, either as water of crystallisation (e.g. in MgCl2.H2O) or as
water as hydroxyl groups (e.g. in Mg(OH)2). The molal entropy increase in perlite was
comparable to those compounds containing water of crystallisation. They did not
conclude that the water in perlite is held similarly to the water of crystallisation, however,
as they concluded that it is impossible that the large portion of the water in perlite is bound
as hydroxyl groups. Later, Stolper [46] found that the concentration of hydroxyl groups
increases rapidly with increasing total water content at low total water content (<2 wt%),
but its concentration levels off or even decreases at a total water content higher than 3%
by weight. On the other hand, the concentration of molecular water increases slowly at
low total water content and more rapidly at a total water content higher than 3% by
weight. Moreover, it was found that in water-bearing glasses having total water content
of 4% by weight, the dissolved water is equally divided between the two species of water
but in glasses with total water content of >4% by weight, more water is dissolved as
molecular water. Saisuttichai and Manning [47] investigated the perlite ore from northeast
of Lopburi (Thailand) and conducted Fourier transform infra-red spectroscopy on
samples containing 3.40 wt% water, which showed higher molecular water than hydroxyl
groups, which appeared to be consistent with Stolper’s findings. Roulia et al. [37] also
studied perlite grains containing 3.48 wt% water from Trachilas (Greece) and the results
showed consistency with Stolper’s findings as well. Therefore, the portion of hydroxyl
and molecular groups in perlite ores vary depending on the total water content, and it is
22
not correct to generally suppose a larger portion of molecular water than hydroxyl groups
without considering the total water content as some sources do (e.g. [21]).
Studies on the nature of the perlitic network have shown that perlite is mostly amorphous
in nature, with a small amount of crystalline inclusions (e.g. quartz, feldspar and biotite).
Herskovitch and Lin [48] studied crude perlite from Milos Island4 and reported it to have
89.4% by volume of the amorphous phase and 10.6% by volume of the crystalline phases
(feldspars, biotite, quartz, magnetite and chlorite). The presence of crystals in crude
perlite is correlated to its origin, as represented in Table 2.3. Crystalline silica (e.g. quartz,
cristobalite and tridymite), though it is present in small amounts, is of concern during
production and marketing. Crystalline silica is classified by the International Agency for
Research on Cancer (IARC) as a class 2A (probabale carcinogen) in humans when it is
inhaled but not when it is ingested or contacted. In most commercial perlite deposits,
however, the amount of crystalline silica is very low (<1-3 wt%). This can be even further
lowered during processing and use, e.g. dilution in end product [49]. Though crude perlite
comprises both amorphous and crystalline phases, the expanded perlite has been found to
be clearly amorphous. X-ray diffraction analysis conducted by Barth-Wirsching et al. [50]
and Chakir et al, [51] indicated that expanded perlite is amorphous. Roulia et al. [37] also
found that the expansion of perlite, affected by rapid thermal treatment, only takes place
in the amorphous phase of raw perlite and that crystallites do not expand and are rejected
as unexpanded waste product. Chemical analysis of EP has shown a similar composition
to crude perlite, but with a lower amount of combined water. It was found that the
4 Perlite was supplied by Habonim Industries, Moshav Habonim.
23
expansion of perlite results in a remarkable reduction in hydroxyl group content with a
smaller reduction in molecular water [47].
Table 2.3. Mineral phases detected in perlites of different origins.
Origin Minerals
Socorro, New Mexico[29] feldspar, cristobalite, zeolites, fluoride
Searchlight, Nevada [29] feldspar, quartz, biotite, magnetite, limonite, zeolites
Korea [43] plagioclase, biotite, opaque minerals, traces of hornblende,
apatite, zircon and sanidine
Kawalan, Yemen [44] k-feldspar, plagioclase, pyroxenes, chlorite, quartz, serpentine
Sardinia, Italy [26] α-quartz, mica, feldspatiods
Milos Island, Greece [23] feldspar, quartz, biotite
Lopburi, Thailand [32] plagioclase and alkali feldspar, biotite, opaque minerals
In addition, it was found that, upon expansion, fluorine may be lost from perlite,
presumably as HF, and it becomes increasingly available to water. The increase in the
availability of fluorine to water may limit the use of such perlite in horticulture
applications [47] as the detrimental effects of fluorine to plants are well-known [52]. It
was also found that, upon expansion, the availability of Ca (Calcium) and Mg
(Magnesium) is reduced, but that the extractable K (Potassium) is increased (especially
in highly potassic perlitic rhyolites). However, the increase in the water solubility of K
on expansion can be useful in horticultural applications (e.g. as a possible slow-release
source of potash) [47].
24
2.2.5 Applications
The low density, relatively low price and good insulating properties of perlite have led to
the development of many commercial applications. Figure 2.5 illustrates the
proportionate use of expanded perlite computed by the U.S. Department of the Interior
U.S. Geological Survey [92]. These applications can be classified into three general
categories: construction applications, horticultural applications and industrial
applications.
Figure 2.5. US expanded perlite use by application.
1. Includes acoustic ceiling panels, pipe insulation, roof insulation board, and unspecified formed
products.
2. Includes absorbents, laundries, paint texturisers, and other miscellaneous uses.
3. Estimated and reported data with specific use unknown
25
2.2.5.1 Construction applications
The low density, high heat insulating value, outstanding fire resistance and good sound
absorbing properties of expanded perlite give it a number of construction applications
mainly as formed products, concrete aggregate, masonry and cavity-fill insulation, and
plaster aggregate. Construction uses of perlite account for about 60% - 70% of the total
consumption in major markets. In the United States, the manufacture of formed products
accounts for about 53% of total perlite use (see Figure 2.5 ). However, such a large
demand for the use of perlite in formed products might not be seen in Europe due to
differences in construction methods. In the US, most homes are constructed on a timber
frame with extensive use of gypsum and insulating boards to form internal surfaces. There
is much less use of brick or other masonry products for external walls, in comparison with
Europe. Therefore, there is less use of light-weight insulating plasters, mortars and loose
fill insulation in the US, compared to Europe, which results in a higher percentage of
perlite use for formed products in the US [53]. In the following sections, the different
construction applications in which perlite is currently utilised, or is substituted for a
traditional material, are explained.
Gypsum perlite-based plaster
Perlite, instead of sand, can be mixed with gypsum and water for plastering on a lath in
order to provide a resilient wall or ceiling capable of withstanding stresses and thermal
changes. Indeed, the manufacturing of gypsum-perlite and prefabricated gypsum boards
has become one of the major uses of perlite. The plasters made of expanded perlite are
characterised by their light-weight, ease of application and savings in structural steel.
They are resistant to cracking under stress, highly fire resistant and have excellent sound
26
absorbing properties [21]. It is claimed that gypsum-plaster made of expanded perlite
aggregate weighs about one third of that conventionally made with sand, and has six times
the thermal insulating value [54]. Perlite can also enhance the workability of plasters and
mortars and can be used to create a texture on the finished surfaces [53].
Concrete
Concrete is a composite material and since around 75% of its volume is occupied by
aggregate, the performance of concrete is greatly affected by the properties of the
aggregate. The use of light-weight and porous aggregate as a constituent of concrete
enables the production of light-weight concretes [55]. Light-weight concretes are
economically beneficial, by reducing heat conductivity and unit weight. Reducing unit
weight is especially beneficial in reducing the damage by earthquakes in high buildings
[56]. The low density of EP particles make them suitable for the production of light-
weight concretes [57, 58]. At the same time, EP particles greatly improve thermal
insulation, reduce noise transmission, and they are rot, vermin and termite resistant [59].
The high water absorption capability of EP particles can provide water for the hydration
of cement in the later stages of curing, which is especially advantageous in concrete with
low water/cement ratios [60]. EP particles can also have pozzolanic activity5 and can be
used as a mineral admixture when finely ground [61]. The properties of light-weight
5 A pozzolan is a siliceous or siliceous and aluminous material which itself has little or
no cement-like value but in finely divided form and in the presence of water will react
chemically with calcium hydroxide (slaked lime) at ordinary temperatures and form
compounds exhibiting cement-like properties [61,62].
27
concrete, however, are usually different from normal concretes. Sengul et al. [62] made
light-weight concrete by partially replacing the natural aggregate (e.g. sand) with EP
particles, from 0% to100% with 20% increments. The replacement of natural sand by EP
particles resulted in a reduction in the unit weights by 8.4 - 80.8 % and a reduction in the
compressive strength of the concrete by 40 - 99.7% due to the lower strength of the perlite.
Topçu and Işikdağ [56] also reported a reduction in compressive and splitting tensile
strength of concrete by about 54% and 47%, respectively, for the replacement of the
natural sand by 60% EP particles. The accompanying air entrainment was also effective
in lowering the strength of the concretes. Not only the compressive strength, but also the
modulus of elasticity were adversely affected. For the replacement of natural sand by EP
in the range 20 - 60%, Sengul et al. [62] reported a reduction in the modulus of elasticity
by 34 - 83%. This is due to the fact that the modulus of elasticity of concrete is a function
of the modulus of elasticity of the constituents (i.e. hardened cement paste and concrete).
Concrete made with perlite aggregate may not be as strong as that made with pumice or
heavier materials, however it may be used where compressive strength above about 13.79
MPa is not required [21]. On the other hand, increasing the expanded perlite ratio
improves the thermal insulation of concrete by increasing the total porosity. Porosity is
one of the factors affecting the thermal conductivity of concrete. Introducing enclosed
pores reduces the thermal conductivity due to the low thermal conductivity of air [62].
These materials are beneficial in insulating spaces around heating, steam or coolant pipes;
providing insulating bases for ovens, furnaces; cold storage tanks; and wherever light
structural decks are desired [53].
28
Loose fill insulation
Another important construction application of perlite is as loose fill insulation. Perlite’s
outstanding low density, thermal insulation and fire resistance allow it to be widely used
as loose fill insulation in cavity walls and for filling cores, crevices, mortar areas and the
air holes of masonry blocks. There are several other materials that can alternatively be
used, such as fibre glass; expanded polystyrene foam, beads and panels; and vermiculite.
However, the valuable features of being non-settling and the free flowing behaviour of
perlite make it more preferred [53]. Moreover, the use of perlite as loose fill in masonry
construction has been proven to reduce the transmission of heat by as much as 50 - 70%
when filling the cavities of concrete blocks. As already mentioned, there is a relationship
between thermal conduction and density at various mean temperatures. Based upon the
reported data, the density of 32 - 173 kg/m3 is recommended for loose fill insulation [63].
However, when using perlite as loose fill insulation, its ability to absorb moisture should
be taken into account. Perlite moisture absorption can adversely affect the insulation value
of expanded perlite, or increase settling problems within a conventional wall construction
where perlite is used. Nevertheless, this problem has been alleviated by coating expanded
perlite particles or adding a moisture repellent to the mix before its application into the
wall. Coating the perlite particles with sodium silicate, for example, can improve the
moisture resistance by closing the open pores of perlite and thus reducing the water
retaining property [64]. Alternatively, it can be mixed with silicone (less than 5% by
weight), a natural resin or Bitumen [65, 66]. However, silicon polymers are expensive
and there are hazards associated with their use. Instead, the annealing of expanded perlite
was found to be effective in reducing perlite’s water absorption [67]. In the annealing
process, expaneded perlite is heated to a temperature sufficient to soften the perlite
29
surface (usually ranging from 426 to 537 °C [67]) and to heal many of the surface cracks
and fissures. Sealing the surface fissures results in a reduction of water absorption.
Ceiling tiles
The use of ceiling tiles has had tremendous growth over the last few decades owing to the
advantages it offers, like sound absorption, fire resistance, good thermal insulation, light
reflectance and structural support. The increase in the use of ceiling tile systems has
resulted in an increase in the use of perlite as a component of many kinds of ceiling tiles.
Standard ceiling tiles are typically made from a combination of mineral wool or slag wool,
fibreglass, binders (e.g. starch), and sometimes inorganic fillers (e.g. mica, wollastonite,
silica, calcium carbonate). Expanded perlite is added to the mixture to enhance the
acoustic insulation, thermal insulation and fire resistance properties of the tile. Depending
on the desired properties of the end product, the proportion of expanded perlite can vary
from 0% to75% of the total mixture by weight. Moreover, perlite can be used to create
different textures to the visible surface of the tile [53].
Starch is a well-known binder in ceiling perlitic tiles. The use of starch has been found
useful for acoustic or sound insulation as these tiles do not bleed through or discolour the
material if painted. Baig [68] used starch as a binder in making acoustic perlitic tiles and
produced tiles with acceptable properties, improved drainage time and lower
manufacturing costs. Denning et al. [69] also used starch in manufacturing insulation
boards and reported an unexpected water repellence property. The product was soaked
for three weeks and it was found that material had just slightly softened. It was also
suggested that the addition of a water-repellent substance to achieve greater water
repellence is possible. However, the addition of water repellent substances to acoustical
boards is limited to those that would not discolour the board. Examples of sound
30
insulation boards made using starch are given in Table 2.4. However, the problem with
the use of starch is that under certain environmental conditions, like increased temperature
and humidity, the board loses its dimensional stability. To deal with this problem, latex
binders were found to be useful. Kesky [70] and Deporter et al. [71] used latex binders in
manufacturing perlitic ceiling tiles and reported perlite insulation boards made with latex
binder show a higher MOR (Modulus of Rupture), higher breaking load, lower water
absorption and better dimensional stability than perlitic ceiling tiles made with starch.
Examples of latex binders, having glass transition temperature ranging from about 30° C
to about 110° C, and useful in ceiling tiles include polyvinyl acetate, vinyl acetate/acrylic
emulsion, vinylidene chloride, polyvinyl chloride, styrene/acrylic copolymer and
carboxylated styrene/butadiene [72].
Table 2.4. Examples of perlitic sound insulation boards in which starch is used as binder [69].
Example 1# 2# 3# 4#
Expanded perlite........percent 67.5 62.5 49.6 60.0
Paper pulp fibre.........percent 22.5 27.5 28.9 30.0
Starch.........................percent 10.0 10.0 20.0 10.0
Modulus of rupture ……kPa 357.84 488.84 409.55 536.41
Under floor insulation
Water repellent, dust-suppressed perlite produced in accordance with ASTM: C 549.81
is used for under floor insulation. Perlite under floor insulation can be employed under
floating concrete floors, asphalt floors and floating board floors (thickness 6 - 10 cm). It
is especially effective in the reduction of sound transmission in construction components,
31
from floor to floor, from floor to walls, from footsteps and from under floor piping
systems. The neutral pH of perlite also prevents the development of corrosion in piping
and electrical wiring that may be installed in the under floor area [63].
Fire resistant boards
Perlite’s outstanding properties of light weight and thermal insulation are used for
producing light-weight insulation panels for fire protection where substantial weight
saving is important, especially in the marine, aviation/aerospace and land/rail transport
industries. For example, light-weight insulating materials that have high thermal
insulation and fire resistance are suitable for components of vehicular interiors such as
cabins and cargo holds, partitions and fire doors, or for transporting combustible materials
[73]. The thermal insulation boards discussed in [74] and [73] are representatives of these
types of light-weight perlite-based insulation boards.
2.2.5.2 Horticultural applications
The chemical resistance and porosity of perlite are important features in horticulture
applications when soils are wet for prolonged periods or where drainage, aeration and
optimum moisture retention are required. The high brightness of finely ground perlite
applied to the surfaces of seed blocks reflects light to the underneath of seedlings and
helps with rapid and sturdy growth [27]. Perlite can be used as a soil amendment or alone
as a medium for hydroponics. Studies have shown that outstanding yields can be achieved
when perlite is used as a soil-less growing media in hydroponic systems. When used as
an amendment, perlite’s high permeability, and low water retention can be beneficial and
helps to prevent soil compaction. Other features of perlite advantageous for horticultural
32
applications are its neutral pH and the fact that it is sterile and weed-free. Perlite is also a
good carrier for fertilizer, herbicides and pesticides and for pelletising seed [75].
2.2.5.3 Industrial applications
Perlite has a broad range of applications in industry. These include the use of perlite as
fillers, filter aids, and in cryogenic and high temperature applications. Perlite is also used
as silica-alumina source for the synthesis of zeolites. These applications are briefly
described below.
Cryogenic applications
Expanded perlite particles contain countless glass-sealed particles, which accounts for its
excellent insulating properties. Expanded perlite is a very effective insulator at
temperatures below -100 °C and is widely used in insulated storage tanks containing
liquid hydrogen, helium, oxygen and other gases [76]. Liquefied gas storage tanks (Figure
2.6) are spherical, double-walled vessels in which the space between the inner and outer
wall is evacuated and filled with expanded perlite as the thermal conductivity of expanded
perlite under evacuated conditions is many times lower than under non-evacuated
conditions [53].
Figure 2.6. Spherical gas tank farm in a petroleum refinery [116].
33
Perlite is not as hygroscopic as other siliceous powders and thus it is easier to prepare for
vacuum applications. The low density and correspondingly high porosity (0.85 ≤ ѱ ≤
0.95) of expanded perlite are also two characteristic properties which make the material
suitable for vacuum insulation [77]. Heat transfer through an evacuated porous material
like perlite involves the simultaneous operation of the three transport mechanisms of
thermal radiation, solid conduction and gas conduction. The radiant heat that passes
through an EP particle originates at the hot boundary wall and from the neighbouring
particles. This radiant energy is of a diffuse type, and, as it moves through the particle, it
is scattered and partially absorbed. The absorbed radiation heats the particle. However,
the scattering effect due to the symmetrical shape of typical vessels (e.g. concentric
spheres or cylinders) is not of great importance in the thermal radiation problem of vessels
filled with insulating porous particles [67]. To reduce the heating effect of radiation, the
addition of opacifiers has been found to be effective. Alternatively, metallic powders can
be added to the insulating particles to scatter and absorb radiation. Kropschot and Burgess
[78] found that the addition of 60 wt% aluminium powders (600 µm) can reduce thermal
conductivity of perlite with a density of 64 kg/m3 by 50%.
Thermal energy is also transferred by conduction through the solid fraction of the EP
particles. However, the solid heat conductivity of a porous insulation material is usually
much smaller than the thermal conductivity of the pure solid fraction as the morphology
of the material influences the heat transport [79]. In addition, the particle size was found
to affect the solid conductivity of evacuated perlite beds. When the particle size decreases
at a constant density, thermal conductivity decreases. The contact resistance increases
with decreasing particle size because the greater strength and curvature of small particles
34
decreases the contact area and there are large numbers of contacts per unit length. In
addition, at low temperatures, phonons carrying most of the heat energy have a long wave
length with a long mean free path. If the particles become smaller than the mean free path
of phonons, the scattering and collisions of phonons at the walls of particles becomes
important and reduces the rate of heat transfer [80]. This effect is strengthened by the
internal structure of perlite particles which are cellular. Therefore, reduction of the
particle size increases the resistance to the flow of thermal energy inside each particle.
The morphology of perlite particles is also effective in the significant minimisation of gas
convection. Convection is suppressed inside the particle pores, as the emerging buoyancy
forces do not exceed friction [81]. Though gas convection is almost eliminated, depending
on the level of vacuum there is some residual gas contributing to heat transfer by gaseous
conduction. The small particle size of the powder limits the gaseous heat transfer in
evacuated powders to free molecular gas conduction as the mean free path of the gas
molecules, which describes the average distance a gas molecule travels between
consecutive collisions, is large relative to the distance between surfaces of the adjacent
particles. Aside from this, the reduction of pressure in high vacuum increases the mean
free path of the gas molecules. In such conditions, gas molecules rarely collide with each
other, thus the molecule to molecule conduction is eliminated. The energy is exchanged
between the surface and the colliding molecules as the gas molecules travel unhindered
between the confining walls of the adjacent solid particles. Therefore, the particle
structure imposes an upper limit on the mean-free path within the insulation [82].
In addition to perlite’s low thermal conductivity, its free-flowing nature, the high
strength/compaction resistance of the particles, the absence of shrinkage or slumping and
the ease of handling and installing are features that make expanded perlite ideal for
35
cryogenic applications. Perlite is also very effective in insulating non-evacuated low
temperature storage vessels for oxygen, nitrogen, liquefied natural gas, and similar
products. The effective insulating properties of evacuated perlite are also utilised in some
domestic refrigerators, in which evacuated plastic bags filled with perlite are used for
insulation [83]. Other applications include the insulation of industrial cold boxes, test or
processing equipment, and shipping containers and in food processing [53].
Fillers
One of the main industrial applications of expanded perlite is as fillers, in which its high
surface area, permeability, low density and inertness and reinforcing characteristics can
be very beneficial. Fillers are organic or inorganic particles added to materials (e.g.
plastics, composite materials and concrete) to reduce the consumption of more expensive
binder materials, and thus to improve the economics of the end product, in which case
they are also called extenders. They are also used to improve some properties of the
mixture by developing a beneficial chemical interaction with the host material; these are
called functional fillers [84]. It is useful to distinguish between the three classes of perlite
used in filler applications. One is produced by further milling and classifying expanded
perlite to produce broken particles with jagged interlocking structures. These are, for
example, used as insulating fillers in the manufacture of texture coatings, as anti-block6
6 Anti-block products are additives incorporated in plastic films to decrease the adhesion or blocking
between touching layers of plastic films during fabrication, storage or use. This can be accomplished by
slightly roughening the film surface by surface treatment with wax/polymers or by the inclusion of anti-
block filler products into the plastic films [21].
36
fillers in plastic films, and as reinforcement fillers in polymers [85]. Another class of
perlite used in filler applications is perlite hollow glass microspheres. Perlite
microspheres consist of one or a few inner cells, in contrast with the large number of cells
found in the larger, standardly produced expanded perlite particles. Perlite hollow glass
microspheres are produced under stringent quality control conditions utilising special
technologies. The feedstock for production of microspheres is crude perlite. Upon
expansion, glassy, spherical-shaped hollow particles are produced, as shown in Figure
2.7. Microspheres generally have a diameter smaller than 140 microns, with the average
between 30 and 70 microns. Depending on the desired properties, the microspheres can
be further surface treated with either silicone or silane, which can increase the
hydrophobicity (water repellency) of the microspheres. When perlite microspheres are
used as fillers, a number of important characteristics are imparted into the end-product.
These include improved crack resistance, higher impact, nailing and stapling resistance,
enhanced machinability and sanding features, finer texture with an increased workability,
reduced weight and shrinkage, and a shortened drying time.
Examples of applications in which perlite microspheres as fillers are desired include in
textured and acoustical coating mixes, adhesives and sealants, wall-patching compounds,
thermo-set castings, syntactic foam, sheet moulding (SMC) and bulk moulding
compounds (BMC), rotational moulding, stucco, fibre-reinforced product (FRP), spray
and hand lay-up, specialty coatings and block filler paints [27]. Lastly, the third class is
those which might be called basic or general light-weight perlite fillers, where traditional
expanded perlite with sizes ranging from 500 to 2000 microns is used as fillers in products
and mixes.
37
Figure 2.7. SEM image showing EP microspheres consisting of one or more microcellular
bubbles [27].
Filter aid
Perlite is chemically inert in many environments owing to the high silica content, usually
greater than 70%, and are adsorptive, which make them excellent filter aids [9]. Filter
aids are finely divided solids, chemically inert and light in weight. They help to control
flow and remove finely distributed colloidal suspensoids from liquids. They can be used
in two different ways, ‘precoating’ and ‘body feeding’. In the ‘precoating’ case, filter aids
are applied as a thin layer over the filter before the suspension is filtered. This will
enhance clarity and filter rate in the filtration process, and prevent gelatinous-type solids
from clogging the filter medium, which can cause its resistance to become excessive.
Furthermore, it facilitates the removal of filter cake at the end of the filtration cycle. The
second application method referred to as ‘body feeding’ involves the incorporation of a
certain amount of filter aid to the suspension before filtration. The addition of filter aids
increases the porosity of the slug, reduces the loading of undesirable particulate at the
filter medium, reduces the resistance of the cake during filtering, and hence maintains the
38
desired flow rate. Depending upon the specific separation involved, perlite products may
be used in precoating, body feeding, or both [31, 35].
The production of perlite filter aids involves careful selection of the ore, precise
monitoring of crushing and screening, expansion, light milling to break the closed
particles, followed by further screening and classification. This process produces
fragments of perlite particles with a precisely controlled particle size distribution to give
the desired clarity and flow rate in different applications. In most applications, perlite
filter aids are used to filter solid particles having one micron or larger size [86].
Perlite competes with other filter media like diatomite, diatomaceous earth and cellulose.
However, it has found extensive niches in many applications due to the advantages it
offers, like rapid filtration of viscous liquids, particle size distribution, high surface area,
high porosity, inertness in most mineral and organic acids, low density compared to other
filter media, and relative cheapness [53].
Perlite is particularly suitable for the filtration of viscous liquids such as edible oils, and
this is one of its main applications. It is also used to remove fine solids like algae and
bacteria from beer and filtering out yeasts and other fine solids from wine. Other
applications include the filtering of various chemicals, pharmaceutical solutions, solid
particulates from waste water streams, water for municipal systems and swimming pools,
and the clarification of foodstuffs (e.g. sugar solutions and syrups) [53].
High temperature applications
Perlite insulation is used in a variety of high temperature applications such as in the steel
and foundry industries (like hot topping, ladle topping and risering), in insulating
formulations at a service temperature of up to 1000°C, and in the manufacture of
refractory bricks or castables.
39
Both raw perlite and expanded perlite can be used in ladle topping applications, but the
raw form is more common. As the perlite ore is added to the ladle, it reacts with the slag
and allows easy removal of the slag layer. The expanded perlite can also act as an efficient
insulating layer to maintain the temperature of molten metal during the steel making
process.
In the hot topping of castings and risers, up to 20% perlite by weight is mixed with
exothermic powders. This controls heat loss from the casting, and thereby prevents
shrinkage cavities. Pre-formed boards for hot tops and hollow cylinders for risers also
perform the same function as hot topping and risering powders and compounds.
Perlite is also used as a cushioning agent in foundry cores and moulding sand mixtures to
compensate for the expansion of silica as it goes through phase changes at temperatures
above 540 °C. It helps to minimise casting defects like buckles, veining, fissuring, and
penetration, and to improve the venting of gases by increasing the permeability of core
sand [53].
An additional use for perlite is in the manufacture of refractories by inducing porosity,
reducing density and thermal conductivity where the average temperature does not exceed
1100 °C. At higher temperatures, perlite refractories are used as backup insulating layers
for higher duty refractories [87].
Zeolite synthesis
Zeolites are a well-defined class of crystalline aluminosilicate minerals. They have three-
dimensional structures arising from a framework of tetrahedral units of [SiO4] - and
[AlO4] arranged in a polyhedral structure linked by all corners [88]. Zeolites have been
successfully used in the chemical industry, agriculture, aquaculture, and environmental
protection because of their remarkable physical and chemical properties which include
40
molecular sieving, adsorption, catalytic and ion exchange functions. This has resulted in
a great increase in demand for these products. There are more than 45 natural types of
zeolite; however, only clinoptilolite, mordenite and chabazite and possibly phillipsite are
currently used as commercial products. The use of zeolites in most industrial applications
requires certain specifications. Thus, synthetic zeolites can be good solutions to provide
the strict specifications imposed on adsorption and catalytic processes. Furthermore, it
would be possible to maintain a constant composition of the products during zeolite
synthesis, while natural zeolites usually show considerable chemical and mineralogical
variation [89]. Consequently, zeolite synthesis from low-cost silica-alumina sources has
been the aim of many studies. Synthetic zeolites can be produced from a variety of natural
silica-alumina materials including acidic volcanic glasses, such as natural and expanded
perlite and pumice (e.g. [90, 91]). Experimental investigations into the formation of
zeolites from perlite have shown that this low-cost silica-alumina material is a suitable
material for zeolite synthesis. It has been shown that perlite can be successfully used in
the synthesis of the zeolites ZK-19, W, G, F, Na-Pc, HS, ZSM-5, A, V, Pc, Na-P1, P1,
K-G, Y, Analcime, sodalite octahydrate and calcium zeolite [50, 88, 89, 92-95].
Other industrial applications
Advantage can be taken of the mild abrasiveness of crude perlite in the production of
soaps, cleansers, and polishes. Crude perlite is also used for the stone-washing of fabrics
in the textile industry. Traditionally, pumice was used to age fabrics and to soften the
finish. Perlite, however, has a softer impact on the fabric and produces less wear on
industrial laundry equipment than pumice, which can be excessively abrasive to the fabric
[53].
41
Perlite uses also include its application in oil-gas, water and geothermal well cementing
and grouting, where the perlite prevents lost circulation of drilling lubricants. Moreover,
perlite is used as a porous support for catalysts and reagents in various chemical reactions
[19, 76].
2.2.6 Concluding Remarks
In Section 2.2, the physical and chemical properties of perlite, the process of its expansion
and its application in construction, horticulture and industry were discussed. However,
one of the main properties of this material, its mechanical properties, has been missed.
Although this material has found uses in various applications, to the author’s knowledge,
barely any research has been conducted into the mechanical properties of perlite. Hence,
one of the main gaps identified in the literature is perlite’s mechanical properties, such as
strength, Young’s modulus and Poisson’s ratio in its expanded and unexpanded forms.
42
Foams
2.3.1 Introduction to foams and syntactic foams
A foam is an assembly of cells (i.e. small compartments) packed in three dimensions to
fill space. Foams can be solid or liquid. In liquid foams, the cell walls are interconnected
liquid films and there is gas inside the cells. Solid foams can be considered as a composite
made of a solid and a fluid (i.e. mostly a gas but it can also be a liquid as is the case in
live tissues) [96]. Gibson and Ashby [97] defined solid foams as cellular materials with a
relative density, which is the density of the foam divided by the density of the solid from
which the cells are made, of less than 0.3. For higher relative densities, there would be a
transition from the cellular structure to solids containing isolated pores. If a solid from
which the foam is made is contained in the cell edges only, and hence the fluid phase is
interconnected, the foams are called open-celled. However, if the solid is not only
contained in the cell edges but also in the cell walls, so that each cell is sealed off from
its neighbours [97], the foams are called closed-celled. Of course, some foams contain a
combination of closed and open cells. Examples of a typical open cell foam and a typical
closed cell foam are illustrated in Figure 2.8. Most foams do not have regular packing of
identical cells, and are comprised of cells with different sizes and shapes. The cell
dimensions can vary, from less than one micrometre in microcellular materials up to
millimetres. The thickness of the cell walls and edges can also vary from 1 to 10 µm as
is the case in some polyurethane foams [98]. In foams, the thickness of the cell edge
(struts) is usually larger than that of the cell walls. This is due to the surface tension
drawing more solid material to the cell edges during the foaming process [97]. The shape
43
of cells can also be more or less random due to differences in the cell dimensions, wall
and edge thickness as well as their chemical composition [99-101]. However, a common
pattern may be observed, for example in cork, the cells are usually prismatic and arranged
in parallel columns on staggered bases [96].
(a)
(b)
Figure 2.8. (a) Closed cell Polyurethane foam [102], (b) Open cell Polyurethane foam [3].
Foam-like materials are very common in nature. Examples are cork, sponge, wood and
coral. While natural cellular materials have been used for centuries, such as cork, which
has been used as bungs for wine bottles since Roman times, mankind has made a wide
44
variety of cellular materials using polymers, metals, ceramics, and even composites via
different foam processing techniques [97].
One of the main types of man-made cellular materials is polymeric foams. They are made
by first heating a polymer into the liquid form, then gas bubbles are introduced. These are
allowed to grow and when they have reached the required size the material is ‘solidified’
by cross-linking or cooling [97]. The gaseous phase is introduced by either mechanical
stirring or by using blowing agents. There are two type of blowing agents: chemical
blowing agents and physical blowing agents. Physical blowing agents are inert gases (e.g.
nitrogen and carbon dioxide) forced into solution in the hot polymer at high pressure
which expand into bubbles due to reducing pressure. Alternatively, low melting point
liquids (e.g. methylene chloride) are mixed with a polymer and vaporised into bubbles on
heating. Chemical blowing agents are additives which give-off gases due to either
chemical reactions or thermal decomposition [97, 103]. Each of these processes can yield
open or closed cell foams. It is the rheology and surface tension of the fluids in the melt
which govern the final structure of the foam [97]. Polymeric foams can be categorised
into two main groups, the thermoplastic and thermosetting foams. The thermoplastics can
usually be reprocessed and recycled; however, thermosets are not recyclable as they are
usually heavily cross-linked. Within these categories, the polymeric foams can be further
classified as rigid, semi-rigid, semi-flexible or flexible. Furthermore, a solid polymeric
foam can either be composed of closed (e.g. polystyrene foam used for coffee cups) or
open cells (i.e. the polyurethane foam in seat cushions). Each polymeric foam possesses
unique physical, mechanical and thermal properties, which are attributed to the polymer
matrix, the cellular structure (i.e. cell density, expansion ratio, cell size distribution, cell
geometry and cell integrity) and the gas composition (i.e. the molecular weight of the
45
constituting components) [104, 105]. These foamed polymers have excellent energy
absorption characteristics and are widely used for shock mitigation in vehicles of all
types, in packaging, in cushioning and as an insulation material. They are also widely
used as cores in sandwich panels, where two relatively thin but stiff face sheets are
attached on either side of a thicker light-weight core [3]. In combination with stiff and
strong face sheets, they produce a construction with higher stiffness to weight and
strength to weight ratios than a structure made of a thick single phase material [106].
It has been shown that a foamed resin with higher specific strength and modulus can be
obtained if the resinous matrix is reinforced with different types of fillers, such as solid
particles, hollow particles, short fibres, etc. [107-110]. However, producing low density
core materials has increased interest in the use of low density fillers such as hollow glass
microspheres [111]. Syntactic foams are a special class of closed cell foams made by
embedding hollow particles in a polymer matrix. They have porosity in the form of closed
cells, and are especially useful in applications where a combination of low density, high
compressive strength and high damage tolerance is required [5]. Moreover, the porosity
enclosed by the thin and stiff shells of hollow particles provides syntactic foams with low
moisture absorption, low thermal and electrical conductivity, as well as high dimensional
stability at elevated temperatures [112]. Most of these structural foams are rigid, although
more ductile and even elastomeric syntactic foams can be manufactured using suitable
hollow particles and matrix materials. There are a wide variety of hollow microspheres
for use in syntactic foams. The diameters of the hollow microspheres ranges from 1 to
350 µm, though they are on average about 50 to 100 µm. Hollow microspheres are made
from glass, ceramic or any polymeric material [2, 113]. Most polymers can be used in the
manufacture of syntactic foams. Producing thermoplastic syntactic foams is generally
46
more difficult as these materials are usually melt processed. Syntactic foams are mostly
processed from thermosetting polymers, many of which are available in a castable liquid
'prepolymer' form. The most prevalent thermosetting polymer matrices used in the
manufacture of syntactic foams are the unsaturated polyester and epoxy resins, although
the use of polyurethane, silicones, polyimides, phenol-formaldehyde and urea-
formaldehyde have been reported [113-117]. In the following section, different methods
used to manufacture syntactic foam are explained.
2.3.2 Manufacturing syntactic foams
The manufacturing process used to produce a syntactic foam is important in terms of the
final product’s properties, ease of production and manufacturing cost. Various
manufacturing methods are available for processing syntactic foams. These vary from the
simple blending of components to the coating of particles by resin [113]. The particle and
binder concentration are found to be crucial for ease of manufacturing. In addition, the
choice of process parameters (temperature, mixing time and addition sequence) is
considered influential in the manufacturing process [118].
The most common method of manufacturing syntactic foams involves the impregnation
of the desired quantity of filler particles (e.g. hollow glass microspheres) in a
thermoplastic polymer solution (polymer + solvent) or in liquid thermosetting
prepolymers (which might contain solvent/viscosity control) [117]. This method ensures
the uniform coating of individual particles by resin and hence, a uniform distribution of
resin among particles can be achieved. As the filler particles (e.g. hollow-glass, organic
or carbon microspheres) cannot withstand high pressure, it is practically not suitable to
use extrude or injection-moulding of mixtures (slurry). Instead, the mixture is usually
47
‘cast’ to minimise particle damage. Solvent is evaporated before final curing, though the
entire removal of solvent and minimisation of solvent entrapment is a particular
challenge. There are several impregnation techniques in common use. In one method,
particles of known mass are introduced to a resin solution. Subsequently, the solvent is
removed to get a dough, which is transferred into the mould and allowed to be cured [119-
123]. In another method, the mould is filled with the desired quantity of particles and
sometimes vibrated to facilitate particles packing [113, 124]. Subsequently, a
premeasured amount of the resin solution is poured over the particles. The solution
penetrates the porous regions between particles due to gravity and capillary forces, and
sometimes with the aid of a vacuum. Another method is to first fill the mould completely
with particles to measure particle quantity where different particle sizes provide different
quantities [125]. Subsequently, the measured quantity of particles is mixed with the resin
solution and the prepared mixture is poured evenly into the mould. To maintain the
particles in a well dispersed state, compression is applied on top of the mould lid or with
the aid of a vacuum. This method has the advantage of restricting the spheres from
floating to the surface during the foam production. In this method, as in the previously
explained methods, removal of the solvent before final curing is required. These methods
have many disadvantages. Perhaps the most serious problem with excess solvent is
drawing off resin from the microspheres during the solvent’s evaporation. As the resin
and solvent solution is heavier than the microsphere, it tends to sink when the mixing
action stops. This also promotes separation of the mixture into phases of equal density by
increasing the tendency of microspheres to float to the top of the mixture, and the resin
solution to sink to the bottom; the mixture must be mixed continuously to the point that
it becomes stable [126]. Other problems include the potential environmental hazards and
48
health effects of volatile solvents, the formation of structural non-uniformities as a result
of the solvent’s removal by heat, the difficulty in producing batch-to-batch homogenous
syntactic foams owing to difficulty in the complete removal of excess solvent from the
mixture before moulding, and the additional cost and energy associated with removal of
the solvent and transporting the mixture to the moulding or curing equipment [127, 128].
The processing of syntactic foams from liquid thermosetting resins without using solvent
has also been reported. This method has been found to be suitable for liquid polyesters
and silicones [113, 129]. In this method, solid particles are directly mixed with a heat
curable thermosetting resin, followed by casting into a desired shape and curing to a foam
[129-131]. The difficulty in this method is non-uniform distributions of particles in the
matrix. To mitigate this problem, the mixture is heated to allow thermosetting resin to
flow and to wet the particles and, hence, a more uniform distribution of particles
throughout the matrix is obtained. In this method, it is important to keep the viscosity as
low as possible to prevent the creation of air bubbles in the final product. Due to the
increase in viscosity, the volume fraction of particles is limited (i.e. the higher the particle
concentration, the higher the viscosity) and hence low density foams (ca. 0.3 g/cm3) can
hardly be achieved. On the one hand, as the particle volume fraction increases, their
tendency to float to the surface of epoxy decreases, which is a common problem in a
simple mixture of epoxy and particles. It has been found that a volume fraction of between
35% to 65%, and preferably about 50%, is suitable to minimise the tendency of the
mixture to separate into discrete phases [129]. Yet the volume fraction of hollow spheres
is adjusted based on the composition, size, and shape of the particles.
When the resin is available in a powdered form, a solid mixture of hollow spheres and
powdered thermoplastic (or powdered thermoset) is prepared by gentle shaking and
49
agitating actions [113, 132, 133]. In some cases, a suspending agent is used to facilitate
mixing and is evaporated before moulding [113]. A weighted quantity of the mixture is
then charged into a mould and heat pressed at a temperature range within which the resin
powder melts and curing occurs. The volume fraction of particles is constant for each
sample with the purpose of completely filling the mould with the closest particle packing
possible. By holding the particle volume fraction constant, the density can be dependent
on a single independent variable (either the resin or the voids volume fraction).
Thermoplastic syntactic foams can be manufactured by conventional melt processing
techniques, such as extrusion or injection moulding [117, 134, 135], though many larger
particles may be damaged by the high stresses and hydrostatic pressures associated with
melt processing. Depending on the type of resin, sometimes post-curing after the
moulding operation is required.
Another manufacturing method for syntactic foams consists of resin coating, vacuum
filtration and polymer precipitation [136, 137]. In the coating step, particles are
introduced into a resin solution (e.g. epoxy diluted with acetone) and left there for a
measured amount of time to ensure good adsorption of resin on the surface of the
particles. The mixture is subsequently vacuum filtered and rinsed with liquids while on
the filter in order to precipitate epoxy on particle surfaces and to remove the solvent
simultaneously by leaching. The result is a wet-sand like material which forms into finely
divided coated particles after drying. These particles have a free-flowing behaviour and
are used as a moulding powder. The resin film to particle ratio in the dry powder is
determined by factors such as solution concentration (i.e. dilution ratio), particle/resin
solution ratio, time of contact between the resin solution and the particles, and
temperature. A predetermined quantity of dry coated spheres, calculated to produce a
50
close-packed syntactic foam, is loaded into the mould cavity, pressed to a desired volume,
and cured. This method enables the production of different sample densities by changing
the resin concentration while the particles are closely packed in a mould [2].
Syntactic foams have also been produced using spray-up equipment. In this method, resin
(or resin solution) and particles are sprayed using separate adjustable streams, meet and
mix at some point while still in the air and accumulate as a syntactic foam on any desired
surface. In this method, the density of the foam is controlled by changing the flow rate of
each component [113, 138].
Last, but not least, is manufacturing syntactic foam based on the microsphere buoyancy
disclosed in AU Patent No. 2003205443 [139]. This technique involves mixing
microspheres with diluted epoxy resin. The mixture is subsequently left stationary for the
self-packing of microspheres by buoyancy resulting in two phases, i.e. the top phase is
composed of packed microspheres and diluted binder; and the bottom phase contains only
diluted binder. The bottom phase is then drained from the bottom of the mould and the
remaining top phase is allowed to be cured. This method has the advantage of ease of
mixing of the microspheres with the epoxy.
2.3.3 Particulate composites containing naturally occurring fillers
Though these hollow particles provide the syntactic foams with a high strength to weight
ratio, it is of interest to find a material which provides similar properties but which is
significantly cheaper and has a lower carbon footprint. One type of promising material is
naturally occurring volcanic glasses.
There are several studies which have investigated the mechanical behaviour of
composites containing naturally occurring inorganic porous aggregates in a resinous
51
matrix. One naturally occurring volcanic aggregate is pumice, which is mainly comprised
of silica (~ 60 wt%) and alumina (~ 16%) [140]. Pumice is formed during volcanic
eruptions when molten rock is ejected into the air. The rapid depressurisation due to the
violent expulsion of magma into the atmosphere causes the release of dissolved gases and
the formation of magma froth. The simultaneous cooling and depressurisation freezes the
bubbles in the matrix and turns the frothy-like magma into a porous ceramic [141, 142].
Fleischer and Zupan [142] used these aggregates in two sizes (4mm and 8mm) in a matrix
of epoxy resin to manufacture a rigid cellular core material. Compressive tests were
conducted and the mechanical response of the pumice–epoxy structure as a function of
relative density was evaluated. However, due to the higher density of pumice (i.e. 1.056
- 1.075 g/cm3), the pumice/epoxy composites were quite dense. Sahin et al. [143, 144]
manufactured PPS (Polyphenylene sulphide) composites reinforced with pumice powder
(<45 µm) at different mass concentrations (1, 5, and 10 wt%) and obtained considerable
improvement in the thermal and mechanical properties of PPS reinforced with 1wt%
pumice powder. It showed increases of 5.6%, 9.8% and 22.5% in the glass transition
temperature, tensile strength and relative degree of crystallinity, respectively. Ramesan
et al. [145] prepared a composite with various loadings of pumice powder (~ 60 µm)
added to a polyvinyl alcohol (PVA) - polyvinyl pyrrolidone (PVP) blend, which is a
biodegradable, water-soluble polymer, and for composites containing 10 wt% pumice
powder they obtained higher AC electrical conductivity, dielectric constant and dielectric
loss, thermal decomposition temperature, and of particular relevance, a higher tensile
strength (~ 29.4 %) than the pure PVA/PVP blend. Alvarado et al. [146, 147] reinforced
PHBV, which is a naturally occurring biopolymer, with 12.8 wt% pumice particles
(~ 147μm) and manufactured PHBV-pumice composites which were stiffer (12.35%)
52
and more brittle than PHBV, experiencing almost no plastic deformation, in order to
reduce the cost and density of the current PHBV-based materials for structural and non-
structural applications. Examples of other studies focused on pumice-reinforced
composites can be found in [148-151].
Another mineral aggregate of natural occurrence is vermiculite. Vermiculite resembles
mica in appearance and mainly consists of silica (37 - 42 wt%), magnesium oxide (14 -
12 wt%), alumina (10 - 13 wt%), ferric oxide (5 - 17 wt%), water (8 - 18 wt%) and a
lesser amount of ferrous oxide (1 - 3 wt%) [152]. When subjected to high temperatures
(>900° C), the water situated in the interlayer space is converted to steam which causes
expansion of vermiculite particles to twenty or thirty times their original volume, giving
so-called exfoliated or expanded vermiculite. Consequently, a highly porous material that
is highly thermally insulating is formed which finds use in various industrial applications
[153, 154]. Jun et al. [155] used the fire retardant properties of expanded vermiculite, in
combination with the low thermal conductivity of phenolic resin, in order to develop
flame retardant insulating composites with a density range 0.12 - 0.22 g/cm3 and thermal
conductivity of about 0.045 - 0.069 W/(mK). Verbeek and Du Plessis [156] used
expanded vermiculite particles in combination with phenol formaldehyde resin and
phosphogypsum to manufacture composites with a density as low as 0.986 g/cm3 and
flexural strength as high as 3.04 MPa for building applications. Yu et al. [157] treated
expanded vermiculite powders with benzyldimethyl-octadecyl-ammonium (ODBA) and
mixed it with phenolic resin (98 wt%) to form a composite with better thermal properties
than that of pristine phenolic resin. The thermal decomposition temperature of the
phenolic/treated expanded vermiculite composite reached 482.6 °C in air, which was 48.7
°C higher than that of pristine phenolic resin. Zheng et al. [158] modified natural
53
powdered vermiculite (<54 µm) by cation exchange with quaternary phosphonium salts
and blended it with polyethylene terephthalate (PET), which is a semicrystalline
thermoplastic polymer, to form (PET)/phosphonium vermiculite composites with better
mechanical properties than pure PET. The (PET)/phosphonium vermiculite composites
containing 3% modified vermiculite showed higher tensile strength (17.4%) and stiffness
(35%) than pure PET. Patro et al. [159] dispersed natural powdered vermiculite (<25
µm) in either isocyanate or polyether based polyols to synthesise rigid
polyurethane/vermiculite composites and achieved better mechanical and thermal
properties than those of a rigid polyurethane foam without vermiculite powder. When 2.3
wt% vermiculite was dispersed in isocyanate, the polyurethane/vermiculite
nanocomposites showed increases of 40% and 34% in their compressive strength and
modulus, respectively, and a 10% decrease in their thermal conductivity. Qian et al. [160]
modified natural powdered vermiculite (<4µm ) by cation exchange with long-chain
quaternary alkylammonium salts and then dispersed it in polyether based polyols to
synthesise a polyurethane/vermiculite nanocomposite with better mechanical and thermal
properties than pure polyurethane foams having 30% rigidity. The composites containing
5.3 wt % vermiculite showed a >270% increase in tensile modulus, >60% increase in
tensile strength, and a 30% reduction in N2 permeability than pure polyurethane foams.
There are several other studies in the literature focused on vermiculite-reinforced polymer
composites with the same general results [161-166].
Last, but not least, another example of a naturally occurring mineral is volcanic ash.
Volcanic ash is made of tiny fragments of jagged rock, minerals and volcanic glass which
is created during volcanic eruptions and deposited at the surface. These deposits are
abundant, readily accessible, and are rich in silica and alumina. Volcanic ash is recognised
54
as a mesoporous material having significant porosity, an appropriate pore structure and
high surface area, which increases the effect of surface adhesion when they are used as
fillers in composites [167]. Bora et al. [168, 169] treated volcanic ash particles with 3-
aminopropyltriethoxysilane (3-APTS) at various concentrations (1, 3 and 5 vol.%) and
used these particles at two concentrations (i.e. 10% and 15%) to reinforce PPS
composites. Treated volcanic ash/PPS composites which showed considerable
improvement in their tensile and flexural properties with 3 vol.% coupling silane agent in
comparison to untreated volcanic ash/PPS composites. Trinidad et al. [170] manufactured
two polymeric matrix composites by mixing volcanic ash with epoxy resin and polyester
resin, individually, and achieved slightly better mechanical properties for volcanic
ash/polyester than volcanic ash/epoxy composites. The volcanic ash/polyester composites
showed 9%, 2.3%, 5.6% and 6.5% higher compressive strength, hardness, modulus of
elasticity, and modulus of rupture, respectively, than those of the volcanic ash/epoxy
composites. Fidan [171] reported manufacturing polyvinyl chloride (PVC) composites
reinforced with volcanic ash at various concentrations (5 - 25 wt%) which resulted in
better thermal stability for PVC composites within the temperature range 25 - 600 °C.
Bora [172] fabricated PVC/volcanic ash composites by reinforcing PVC with volcanic
ash at various mass contents (5 - 25 wt%) which led to an increase in the tensile and
flexural modulus but a decrease in the tensile and flexural strength. All the properties
showed significant reduction as a result of the temperature increase from -10°C to 50 °C.
Avcu et al. [173] outlined the manufacture of volcanic ash filled polyphenylene sulphide
(PPS) composites containing different volcanic ash concentrations (2.5 - 20 wt%). They
reported improvement in the thermal, mechanical and residual mechanical properties of
the PPS composite by an increase in the volcanic ash content, although erosion resistance
55
was decreased markedly. Cernohous [174] outlined the melt processing of polypropylene
(PP) with a blowing agent and volcanic ash particles at different mass concentrations (40
- 60 wt%) and produced thermoplastic foamed composites with a density range 0.65 -
0.78 g/cm3 and a flexural modulus in the range 3350 - 4120 MPa. Examples of other
studies which have focused on composite materials made from a polymeric matrix
reinforced with volcanic ash can be found in [175-180]. There are, however, few studies
investigating the use of perlite particles in a resinous matrix. Lukosiute et al. [181]
produced two composites, one was made of an epoxy resin matrix filled with unexpanded
perlite particles at different volume fractions and another one made of an epoxy resin
matrix reinforced with different volume fractions of plasticised PVC particles and
unexpanded perlite particles. The unexpanded perlite particles used in this study had a
mean size of 80 µm, specific weight of 2.37 g/cm3, density of 1.60 g/cm3 and a porosity
of 32%. Perlite/epoxy composites have shown higher tensile strength, elastic modulus
and bending strength but lower impact strength than those containing dispersed PVC
particles. In addition, it was found that an increase in the perlite particle volume fraction
from 7% to 41% increased the stiffness of the perlite/PVC/epoxy composites due to the
higher stiffness of the perlite particles (i.e. 27.5 GPa [181]) compared with the epoxy
matrix (i.e. 2.6 GPa). Sherman and Cameron [182] produced cores for building boards
containing a high proportion of expanded perlite particles (85% by weight) in different
matrices of acrylic resin, epoxy resin and urea-formaldehyde. They found that epoxy resin
provided the core with a higher modulus of rupture and density than the other two resins.
However, the results were not comprehensive as their conclusions were only based on the
two above-mentioned properties. Of all of the available options, perlite is considered the
most promising filler.
56
There are several studies which have investigated the use of EP particles in composites
for structural and non-structural applications. Lu et al. [183] manufactured a form-stable
composite by direct impregnation of expanded perlite in paraffin, resulting in good
thermal energy storage, thermal stability and thermal reliability for composites containing
60 wt% paraffin. Later, Lu et al. [184] manufactured a phase change material by
depositing graphene oxide films on the surface of the expanded perlite/paraffin
composite. The heat storage/release performance test results showed that the heat
storage/release rate of the expanded perlite/paraffin composite with 0.5 wt% graphene
oxide was twice as fast as that of the expanded perlite/paraffin composites because of the
enhanced thermal conductivity. Karaipekli et al. [185] also manufactured a form-stable
expanded perlite/paraffin composite material for latent heat thermal energy storage and
found that adding 5 wt% expanded graphite could increase the thermal conductivity of
the form-stable composite by about 46%. Shastri and Kim [186] used a pre-mould
process, which was initially used by Kim [187] in manufacturing glass
microsphere/epoxy syntactic foams and disclosed in AU Patent No. 2003205443 [139],
for consolidation of expanded perlite particles with starch binder as a building material.
Using this method, they manufactured perlite/starch composites in the density range 0.1
- 0.4 g/cm3 and characterised the compressive strength and modulus to be in the range 0.1
- 1.5 MPa and 10 - 85 MPa, respectively, for this range of foam density. Arifuzzaman and
Kim [188] also produced composite foams for building applications using a similar
method of manufacture, based on the buoyancy principle by dispersing expanded perlite
particles in a matrix of sodium silicate. They manufactured these foam composites in the
density range 0.2 - 0.5 g/cm3 and characterised the compressive strength and modulus to
be about 2.8 MPa and 175 MPa, respectively, for the highest foam density containing
57
0.35 g/ml sodium silicate. Furthermore, they investigated the effect of expanded perlite
particle size in three ranges, 1 - 2mm, 2 - 3mm and 3 - 4mm, on the compressive strength
of the perlite/sodium silicate foams and found that the particle size did not significantly
affect the compressive strength.
The light weight and porous structure of expanded perlite has also received attentions in
manufacturing syntactic metal foams. Taherishargh et al. [11, 189, 190] produced low
density syntactic foams as an effective energy absorber by the counter-gravity infiltration
of expanded perlite particles with molten aluminium. They investigated the effect of
expanded perlite particle size, particle shape and foam density, on the resulting properties
of foams containing 62 - 65 vol% expanded perlite. Smaller expanded perlite particles
resulted in a metallic foam with a refined cell wall microstructure, smaller cell size, fewer
defects and an equiaxed dendritic morphology. In addition, the homogeneity of the cell
geometry was found to increase with a decrease in particle size, resulting in more uniform
plastic deformation. As a result, foams with smaller EP particles showed a smoother and
steeper stress–strain curve and a higher plateau stress and energy absorption capacity.
Regardless of particle size, the mechanical properties (e.g. compressive strength, modulus
of elasticity, energy absorption efficiency and plateau end strain) were found to increase
with an increase in the foam density. This has also been reported in the literature for a
variety of metallic foams [191-197]. The shape of expanded perlite particles was also
found to affect the structural and mechanical properties of expanded perlite/aluminium
syntactic foam [190]. The irregular shape of raw EP particles was turned into nearly
spherical particles using a rotary tumbling machine. Use of rounded EP particles resulted
in an increase in the modulus of elasticity, 1% offset yield stress, plateau stress and energy
absorption by 38%, 24%, 14% and 19%, respectively. Image processing and
58
micro-computed tomography (µCT) revealed strut misalignment, missing cell walls, and
surface roughness inside the pores for foams containing the raw particles, while a
homogenous distribution of nearly spherical pores was observed for foams with rounded
particles. Hence, the superior mechanical properties of the foams with rounded EP
particles were ascribed to their homogenous distribution of pores and fewer structural
defects.
In summary, EP perlite is considered to be a promising material in manufacturing polymer
matrix syntactic foams with regards to its properties (i.e. chemical and physical) and
applications, as explained in the first part of this chapter (see Section 2.2). Perlite, in its
expanded and unexpanded forms, has been used in the manufacture of different
composites; however very few studies have been conducted on the combination of EP
particles with a resinous matrix, especially epoxy. Hence, there is a need for further
investigation of the properties of foams made from perlite particles in a resinous matrix.
59
Motivation and Problem Statement
Despite the fact that hollow microsphere/epoxy syntactic foams are known to offer
desirable properties for structural applications (i.e. high strength to weight ratio and high
stiffness to weight ratio), the high cost of the synthetically made filler particles makes
their production expensive. Considering the low density, abundance and low cost of
expanded perlite particles, it would be beneficial to develop new light-weight foam cores
for sandwich structures to replace conventional hollow microsphere/epoxy syntactic
foams at a significantly lower price. In addition, considering the wide use of expanded
perlite particles in different sectors of construction and industry, it would be worthwhile
to investigate the elastic properties of these particles, which have received little attention
in previous research.
60
Research Objectives and Research Significance
The main objectives of the current study are as follows:
i. To investigate the mechanical properties of packed beds of expanded perlite
particles.
ii. To introduce suitable models capable of predicting the elastic properties of
packed beds of expanded perlite particles.
iii. To introduce and characterise an economical, light-weight syntactic foam.
iv. To investigate the effects of different parameters such as the filler particle sizes
and foam density on the structure, microstructure and mechanical properties of
the new syntactic foams.
v. To investigate damage mechanisms that occur during mechanical tests and
understand the material’s behaviour under applied loads.
The significance of the current research study are as follows:
i. Producing a syntactic foam having properties comparable with other foams
available in the market but significantly cheaper.
ii. Producing a syntactic foam which can be moulded to accommodate different
shapes for different structural applications.
61
3 Chapter Three: Methodology
Introduction
This chapter begins with an introduction to the materials used in this study, followed
by discussion of how these materials were used to prepare packed EP particle beds,
sintered solid perlite, resin and EP/epoxy foam samples. Subsequently, all of the
different tests used to characterise the sample properties (e.g. mechanical and elastic
wave speeds) are explained in detail. Furthermore, the apparatus and procedure used
for the microstructural and damage analysis are explained. Lastly, the theory of
dynamic moduli measurement using the elastic wave speed is described.
62
Material
3.2.1 Expanded perlite particles
Expanded perlite particles (EP), shown in Figure 2.1 (b), were supplied by Industrial
Processors Limited (INPRO) and were sieved into the size ranges 1 - 2 mm, 2 - 2.8 mm
and 2.8 - 4mm. The chemical composition of the perlite provided by the supplier is
presented in Table 3.1.
Table 3.1. Chemical composition [198]
Constituent Percentage present
Silica 74%
Aluminium Oxide 14%
Ferric Oxide 1%
Calcium Oxide 1.3%
Magnesium Oxide 0.3%
Sodium Oxide 3.0%
Potassium Oxide 4.0%
Titanium Oxide 0.1%
Heavy Metals Trace
Sulphate Trace
63
3.2.2 Epoxy resin
The resin system used for binding perlite particles consisted of West System 105 Epoxy
Resin with 206 Slow Hardener, at a ratio of 5.36:1 by weight. To enable the
manufacturing method, ease the distribution of perlite particles in the binder and control
the binder content in the foam, the epoxy was diluted with acetone (Septone, ASA20).
Acetone was considered to be a suitable solvent as it has a less detrimental effect on the
mechanical properties of cured epoxy in comparison with other solvents such as toluene,
tetrahydrofuran, N-dimethylformamide (DMF) and ethanol [199, 200]. Fourier transform
infrared spectroscopy (FTIR) also showed that the use of acetone does not seem to change
the chemical structure of the liquid epoxy [199, 201]. The dilution with acetone at this
ratio reduced the mix viscosity, which was measured using a RST-Brookfield rheometer.
In addition, dilution with acetone increased the curing time of the epoxy, which was
investigated by monitoring hardness change using a Type D durometer, as shown in
Figure 3.1.
Figure 3.1. Durometer (Type D) for hardness test.
64
Sample preparation
3.3.1 Preparation of expanded perlite (EP) particle samples
3.3.1.1 Samples for structural characterisation of EP particles
The tapped density of EP particle beds of known mass were measured using a tapping
device with a graduated measuring cylinder of 100 ml, shown in Figure 3.2. After every
20 taps, the cylinder was rotated to minimise any possible separation of the mass during
tapping down. Five hundred taps were conducted for each density measurement, and the
average of five measurements was considered.
Particle density was also measured using the wax-immersion method (ASTM C914-95).
It should be noted that due to the EP particles’ structure having both open and closed
pores, measurements with an air pycnometer resulted in the wrong density. As expected,
the value given by the air pycnometer lay between the particle density and the true density
of the material. As the wax does not penetrate into the perlite pores, the wax-immersion
method is considered to give the most accurate measurement of the perlite particle
density.
Scanning electron microscopy was used to study the internal structure of perlite particles
by cross–sectioning EP particles carefully. For this purpose, a number of particles were
embedded in epoxy resin poured in a small mould of 30 mm diameter and 30 mm height.
Samples were ground and lapped to a 1m diamond finish. The ground samples were
transferred into an ultrasonic cleanser to remove perlite dust created during the cutting
and polishing. All samples were coated with gold before microscopy to avoid charging.
65
Figure 3.2. Tapping device for measuring tapped density of EP particles.
3.3.1.2 Samples for confined quasi-static compressive tests (oedometer tests)
A cylindrical compaction mould made of steel was manufactured to produce samples 35
mm in height and 30 mm in diameter for measuring the properties of packed EP particle
beds. Figures 3.3 (a) and (b) show the constituent parts of the mould and the assembled
form of the mould, respectively. Samples for each particle size were prepared for the
density range 0.10 - 0.40 g/cm3 and three samples were prepared for each density. To this
end, EP particles of known mass were transferred into the mould and compressive loading
was transferred to the EP particles through a plunger moving vertically into the mould.
The compaction process was carried out with a constant displacement rate of 1.0 mm/min
in a computer-controlled 5kN Shimadzu testing machine. The target compaction densities
were achieved by controlling the mass of EP particles within a constant volume. To ensure
homogeneity, specimens were compacted in three layers. To minimise friction between
66
the piston and the wall of the mould, lubricant oil was sprayed on the wall of the mould
and the plunger.
(a) (b)
Figure 3.3. Prepared mould (a) The different parts of the mould (b) Assembled mould.
3.3.1.3 Samples for elastic wave tests on packed beds of EP particles
An aluminium cylindrical compaction mould was manufactured to produce samples of 45
mm ± 0.015 mm in height and 80 mm in diameter for measuring the elastic properties of
packed beds of EP particles. The target compaction densities were achieved by controlling
the EP particles’ mass within a constant volume. To ensure homogeneity and minimise
variations in density, the specimens were compacted in several layers (up to five). The
compaction process was carried out in a 5000kN computer-controlled Shimadzu testing
machine using a constant displacement rate of 1.0 mm/min. Lubricant oil was used to
minimise friction between the piston and the walls of the mould. Samples were prepared
with a density range 0.10 - 0.35 g/cm3 and three samples were prepared for each density.
67
3.3.2 Preparation of solid perlite samples
Solid perlite was prepared by grinding 27 g of perlite particles with a mortar and pestle
and sieving to <250 µm. The powder was transferred into a mould and uniaxially pressed
into a pellet at 146 MPa. The pellets, resting on a bed of alumina powder, were heated at
10 °C/min to the sintering temperature of 1100 °C which was held constant for 12 hours.
Subsequently the samples were cooled at 10 °C/min to 450 °C and held for an hour.
Cooling continued at the same rate to 300 °C, at which temperature the furnace was turned
off and the sample was left to cool slowly. The sintered pellet was cut and polished into
samples 30 mm in diameter and 16 mm high, as shown in Figure 3.4. The densities of the
prepared sintered pellets were measured at room temperature using the Archimedes
method.
Figure 3.4. Solid perlite sample (sintered perlite powders).
3.3.3 Preparation of resin samples
To measure the compressive, tensile and flexural properties of epoxy resin, two groups
of samples were prepared, one group for the undiluted epoxy resin and another group for
epoxy resin diluted with acetone according to the ratio of 1 to 7 for epoxy + hardener to
68
acetone, respectively. Samples were prepared by pouring the epoxy resin into two moulds
made of Acetal, as shown in Figure 3.5.
(a) (b)
Figure 3.5. Mould for manufacturing epoxy resin samples for (a) compression and flexural tests
and (b) tensile tests.
The moulds, shown in Figure 3.5, were designed to manufacture samples according to
ASTM D-695, ASTM D-638 and ASTM D-790 standards for measuring the compressive,
tensile and flexural properties of epoxy resin, respectively, as shown in Table 3.2.
Table 3.2. ASTM D-695, ASTM D-638 and ASTM D-790 for measuring the compressive,
tensile and flexural properties of epoxy resin.
ASTM
D-695
69
T (Thickness) = 3.2±0.4 mm, W (Width of narrow section) = 12.7 mm,
LO (Length Overall) = 79.4 mm, WO (Width Overall) = 19 mm,
G (Gage length) = 38.1 mm, R (Radius of fillet) = 38.1 mm
ASTM
D-638
T (Thickness) = 3.2±0.4 mm, W (Width of narrow section) = 13 mm,
L (Length of narrow section) = 57 mm, WO (Width Overall) = 19 mm,
LO (Length Overall) = 165 mm, G (Gage length) = 50 mm, D (Distance between
grips) = 115mm, R (Radius of fillet) = 76 mm
ASTM
D-790
T (Thickness) = 3.2±0.4 mm, W (Width of narrow section) = 12.7 mm,
LO (Length overall) = 79.4 mm
70
3.3.4 Preparation of EP/epoxy foam samples
The basic principle for manufacturing EP/epoxy foam composites was based on the
buoyancy method explained in AU Patent No. 2003205443 [139] for manufacturing glass
microsphere/epoxy syntactic foams. The perlite particles provided by Industrial
Processors Limited (INPRO) contained some unexpanded and semi-expanded particles;
thus a device was designed to separate the expanded particles, based on the winnowing
technique. This device, shown in Figure 3.6, was made of two computer fans, which
provided enough air pressure with respect to the density of EP particles, a sieve of 500
µm and a piece of PVC pipe, which was attached to the sieve. The air blown by the fans
passed through the sieve mesh and lifted the less dense expanded perlite particles out of
the PVC pipe, leaving the denser unexpanded particles behind. Separated EP particles,
the epoxy + hardener and acetone were mixed at the ratio of 1:2:14 respectively, in a
screw cap plastic container through gentle agitating and tumbling actions. This particular
mixing ratio was adopted due to the buoyancy method and to producing light-weight
foams containing a thin film of epoxy around the particles. The mixing container was
then left stationary until particles floated to the surface resulting in two phases, i.e. the
top phase of packed perlite particles and diluted binder, and the bottom phase containing
only diluted binder, as shown in Figure 3.7. The top phase was separated by straining and
transferred into a cylindrical steel mould which is shown in Figure 3.3. The same mould
was used in preparing packed beds of EP particles (explained in Section 3.3.1.2) and
EP/epoxy foam samples in order for the dimensions to be comparable and to correctly
measure the crushing strength of the EP particles during sample manufacturing (to be
explained later in Section 4.6.2). The mould was designed to produce samples 35 mm
high and 30 mm in diameter, according to ASTMC365/C365M – 11a.
71
Figure 3.6. A hand-made device for separating expanded particles from unexpanded and semi-
expanded particles.
Three types of EP/epoxy foam samples were manufactured for the density range 0.15 -
0.45 g/cm3 by changing the mass inside the mould while keeping the volume constant.
To achieve a constant volume, a universal testing machine was used to compress the
mixtures of EP particles and the binder to the target height of 35mm. Figure 3.8 illustrates
the foam samples having the same density but using three different particle size ranges.
Foams containing one particle size range were considered to be of one type. Foams of
type 1, type 2 and type 3 were defined as foams made with particles in the size ranges of
1 - 2 mm, 2 - 2.8 mm and 2.8 - 4 mm, respectively. Five to eight samples were
manufactured for each density and type of foam. The densities of the prepared foams were
measured at room temperature after the curing time elapsed.
72
Figure 3.7. Formation of two phases by buoyancy. The top phase is packed perlite particles and
diluted binder and the bottom phase contained only diluted binder.
Figure 3.8. Samples prepared using three particle size ranges; from the left, 1 - 2 mm, 2 - 2.8
mm and 2.8 – 4 mm.
73
Experimental Setup and Tests
3.4.1 Mechanical testing on the packed beds of EP Particles
To measure the properties of the EP particles, oedeometric tests (confined compressive
tests) were conducted on packed EP particle beds. As explained in Section 3.3.1.2,
compaction was conducted in several layers, where each layer had virtually the same
density and almost the same compaction load. Hence, the compaction load was in the
same range for the whole of each sample. During the tests, experimental data were
recorded by built-in aquisition software (Trapezium 2) in the form of force-displacement
curves for each sample. These were subsequently converted to compaction stress-density
curves based on the cross-sectional area of the plunger and the final height of the sample,
measured using callipers after removing the compacted sample from the mould.
In addition, the compacted samples containing EP particles in the range 2 – 2.8 mm were
used to measure the confined elastic modulus of the particles by cyclic loading-unloading
of the pre-compacted particles to 10%, 15% and 20% of the maximum compaction force.
Three to four tests were conducted for each density.
3.4.2 Mechanical testing of EP/epoxy foams
Uni-axial (unconfined) compression tests were conducted on a computer-controlled 5 kN
Shimadzu testing machine. Samples were compressed at a crosshead speed of 0.5
mm/min at room temprature, which ranged from 23 to 25°C. During the tests,
experimental data were recorded by the built-in aquisition software (Trapezium 2) in the
form of force-displacement curves for each sample and these were then converted to
engineering stress-strain curves, based on the intial cross-sectional area and height of the
74
test sample. To investigate the porosity dependent rigidity, the elastic gradient of the
material was measured based on ISO 13314. In this standard, the elastic gradient is
determined by loading and unloading between 20% and 70% of the maximum stress (σ20
and σ70). The unloading gradient, or so-called effective elastic modulus, was measured to
alleviate two potential problems affecting the accuracy of the initial loading modulus.
Firstly, there might be some localised plasticity in the sample where cells yield at very
low stress levels. This may result in contributions from both elastic and plastic
deformation to the resulting modulus. Secondly, it is very likely that the sample’s surfaces
touching the loading plates are not perfectly parallel. However, applying load at low
strains will flatten the surfaces until complete contact is established (settling).
Consequently, when the load is released, the unloading gradient is purely elastic and not
affected by localised plastic deformation and the settling effect [11]. To this end, one
initial sample must be tested to obtain the maximum compressive stress (max). Then the
next four samples were loaded to 70% of the maximum stress (σ70) and unloaded to 20%
of the maximum stress (σ20).
Furthermore, a series of confined compression tests was conducted on cured EP/epoxy
foams. For this purpose, samples were tightly enclosed in the cylindrical steel mould
shown in Figure 3.3, with an inner diameter equal to the diameter of the sample. This
mould was the same mould used for manufacturing the EP/epoxy foams. The diameter of
the samples after curing were slightly bigger (30.1 ± 0.045 mm) than before curing (30
mm) which helped in tightly enclosing the foams.
75
3.4.3 Elastic wave tests on EP particles
To measure the properties of EP particle compacts, a compaction rig was designed to
include transducers and particles. It consisted of a pair of aluminium rams located inside
the top and bottom of the cylindrical aluminium die enclosing the compacted EP particles.
The rams were designed to have the same height as the transducer and a groove held the
transducer tightly with the ram surface flush with the active element of the transducer. A
schematic representation of the set-up is shown in Figure 3.9.
Figure 3.9. Schematic representation of the experimental set-up for measuring wave velocity in
packed beds of EP particles.
Elastic wave velocity measurements were carried out at a constant contact pressure equal
to the average compaction stress at each density (see Section 3.4.1) using a computer-
controlled Shimadzu testing machine. Two pairs of piezoelectric transducers (S-wave and
P-wave) with a nominal frequency of 1MHz were used to measure the compression, Cp,
and shear wave, Cs, velocities in the axial direction throughout the sample. Each pair of
transducers was placed in the rams and a thin layer of silicone grease was used to improve
76
the coupling between the specimen and the transducers. The experimental setup employed
here is similar to the system used by Arroyo et al. for testing clayey rocks [202]. It
included a programmable function generator (Thurlby Thandar® TTi – TG4001) to
generate and transmit elastic waves as well as a digital oscilloscope (Tektronix® TDS
1001C-EDU) to acquire both input and output signals. Signal stacking was applied to
both input and output waves to minimise random noise. Sine pulses with apparent
frequency ranging from 5 to 40 kHz (100 V peak-to-peak) were used as the input signal.
Elastic wave velocities were estimated as the ratio of the travel length to the travel time.
The recorded input and output signals were used to estimate the travel time by considering
the time delay between the start of the input signal and the start of the output signal
(details explained in [203]). The travel length is the height of the samples and was
determined using callipers.
Examples of input and output signals for the low and high density EP particle compacts
are given in Figure 3.10, where the arrival times for the P-waves (tP) and S-waves (tS) are
indicated by vertical dashed lines. The arrival time tP was clearly identified as the
‘common first break time’ in the output signals, irrespective of the sample density. Two
sets of input frequencies were used to determine the arrival time tS depending on the
density of the EP particle compacts. Input frequencies up to 40 kHz were used in dense
samples, whereas good quality output signals were obtained in low density samples (<
150 g/cm3) using frequencies up to 15 - 20 kHz. A step pulse of 10 Hz was also applied
which, in combination with the sine pulses, helped to check the arrival time of the S-
wave. The determination of the arrival time tS required extra care due to the influence of
low-energy P-waves, which travel faster than S-waves, on the output signals. This
phenomenon, known as the near-field effect [202, 204], is strongly frequency-dependent
77
and may lead to misleading estimates of tS if the arrival time of the P-wave is not
available. From inspection of Figure 3.10 (b) and (d), it is clear that the arrival time of the
S-wave corresponds to the first important break time, whereas the small disturbance
(‘bump’) observed earlier in the output signals (clearer in the low density samples) is
consistent with the arrival time tP previously estimated in Figure 3.10 (a) and (c).
(a) (b)
78
(c) (d)
Figure 3.10. Examples of (a) P-wave and (b) S-wave signals in low density (0.12 g/cm3) EP
particle compacts and (c) P-wave and (d) S-waves in high density (0.3 g/cm3) EP particle
compacts.
79
3.4.4 Elastic wave tests on EP/epoxy foams and solid perlite
A frame was designed to measure elastic wave velocities in EP/epoxy foams as well as
the solid perlite samples. The frame consisted of a mobile top beam guided using a
frictionless roller bearing, a stationary bottom beam and a pair of aluminium rams within
which the sample was placed. Rams were designed to hold the transducers tightly while
keeping the active element of the transducers at the same level as their surface. A
schematic representation of the set-up is shown in Figure 3.11.
Figure 3.11. Schematic representation of the ultrasonic experimental set-up for measuring wave
velocity in solid perlite and EP/epoxy samples.
To improve the coupling between the transducer and the sample, a constant pressure was
applied by adding a 5 kg dead weight to the mobile top beam; this will in turn enhance
the quality of the output signals. Two pairs of piezoelectric transducers with a nominal
frequency of 1MHz were used to measure the longitudinal, CL, and shear wave, Cs,
velocities in the axial direction throughout the sample. Each pair of transducers was
placed in the rams and acoustically coupled by a thin layer of silicone grease. Similar to
80
the experimental setup used for measuring the elastic properties of EP particles (see
Section 3.4.3), the experimental setup included a programmable function generator
(Thurlby Thandar® TTi – TG4001), a digital oscilloscope (Tektronix® TDS 1001C-
EDU) and an ultrasonic pre-amplifier OLYMPUS® (Panametrics 5656C, with a gain of
40dB). Signal stacking was applied to both the input and output waves to minimise
random noise. Sine pulses with apparent frequency ranging from 10 - 50 kHz (100 V
peak-to-peak) were used as the input signal. Elastic wave velocities were estimated in the
same way explained in Section 3.4.3.
81
Microstructural analysis
Scanning electron microscopy (SEM) was used to study the internal structure of perlite
particles and perlite-based foam samples before undergoing compression tests. The
microstructure of expanded perlite particles was studied by carefully cross-sectioning the
particles. For this purpose, a number of particles were embedded in epoxy resin poured
in a small mould of 30 mm diameter and 30 mm height. After 24 hours curing, the sample
was demoulded and the surface of the sample was carefully ground using 240, 320, 600,
800 and 1200 grit silicon carbide papers. A smooth, mirror like surface was subsequently
achieved by polishing with 1µm diamond suspension in distilled water sprayed onto a
polishing cloth. Perlite particles and epoxy resin are both non-conducting materials; hence
the samples were coated with gold before microscopy. The internal structure of the
perlite-based foams was studied by cutting a thickness of 3mm from the sample using a
low speed silicon carbide saw. The thin slices were carefully ground by 800 and 1200 grit
silicon carbide paper. To remove perlite dust created during cutting and polishing the
sample, the thin slices were put into an ultrasonic cleaner for 10 minutes. They were then
dried using a blow dryer and left in an oven at 35 °C for 2 hours. Lastly, the thin slices
were coated with gold and were then ready for microstructural study by SEM.
82
Damage observation
Failure mechanisms of the newly developed foams were investigated using both
macroscopic and microscopic observations. During the course of the mechanical testing
experiments, a high resolution camera (Nikon D800E) was used to record an image of the
compressed sample every 10s. This helped to closely follow the damage sequence and
the failure process in the foam samples. Scanning electron microscopy was used after
finishing the test to examine the fractured samples and to determine the failure modes. As
the samples were mostly made up of perlite particles having a brittle nature, the fracture
of the foam under compressive loading generated a large amount of broken pieces of
particles (e.g. perlite dust). These loose particles can cause charging during electron
microscopic observations. Thus, the fractured surface of the foam sample was exposed to
a gentle blow of air before coating with gold. As the fractured samples had uneven
surfaces with a porous structure, coating was also conducted from different directions and
angles in several steps to avoid charging effects in the microscope.
83
The theory of dynamic moduli measurement
Elastic stress waves propagate in materials by inducing infinitesimal elastic deformation.
Hence, wave propagation equations which are based on elasticity theory can be used to
measure the elastic moduli of a material if the elastic wave velocities and densities are
measured independently [205, 206]. Two wave types propagate in extended elastic solids:
(i) compression waves in which the displacements are in the direction of the wave
propagation, and (ii) shear waves in which the displacements are perpendicular to the
direction of the wave propagation. In an isotropic linear elastic body, in which the
propagating wave does not interact with the boundary of the medium, the compression
wave or P-wave (CP) and shear wave or S-wave (CS) velocities are given by:
SC (3.1)
2
MCP
(3.2)
where M is the constrained modulus, ρ is the density of the material, λ and µ are the Lamé
constants. The Lamé constants are related to the stiffness tensor, 𝐶𝑖𝑗𝑘𝑙, by:
kljkiljlikklijijkl )]+ (+ [=C (3.3)
where 𝛿 is the Kronecker delta function defined as:
𝛿𝑖𝑗 = {0 𝑓𝑜𝑟 𝑖 ≠ 𝑗1 𝑓𝑜𝑟 𝑖 = 𝑗
, 𝑖, 𝑗 = 1, 2, 3.
Young’s modulus (E) and Poisson’s ratio (𝜈) can be expressed as a function of the density,
compression wave velocity and shear wave velocity of the material as follows:
84
22
22
2 43
SP
SP
SCC
CCCE
(3.4)
)(2
222
22
SP
SP
CC
CC
(3.5)
When the characteristic dimensions of a body are not large compared with the
wavelength, these equations no longer hold, and the velocities become dependent upon
the frequency [207]. In the case of an unconstrained cylindrical bar (i.e. uniaxial
conditions) with a radius much smaller than the wave length, the compression wave has
a different velocity called the longitudinal velocity (or rod velocity) CL, given by [207]:
ECL (3.6)
However, the stiffness that controls shear waves in cylindrical bars is the same as the one
in an infinite medium (i.e. Eq. (3.1)) [204]. Consequently, Poisson’s ratio is expressed as:
1
2
1
S
L
C
C (3.7)
In the current study, the above equations are used to characterise the elastic properties of
packed EP particle beds and sintered perlite by assuming that the pores are air-filled voids
which are randomly distributed and oriented throughout the medium. Hence, the
commonly adopted assumption that the porous solid body exhibits isotropic elastic
behaviour in a statistical sense will be adhered to.
With regards to the sample size and wave length, Eqs. (3.4) and (3.5) were used to
characterise the elastic properties of packed beds of EP particles, whereas Eqs. (3.6) and
(3.7) were used to characterise the elastic properties of sintered perlite samples.
Moreover, Eqs. (3.6) and (3.7) were used to characterise the elastic properties of
EP/epoxy foams in the longitudinal direction (i.e. z-axis) by assuming that the particles
85
were randomly distributed and oriented throughout the foam and, thus, the EP/epoxy
foams macroscopically exhibit isotropic elastic behaviour.
86
4 Chapter Four: Results
Introduction
In this chapter, the properties of solid perlite (pore free), packed EP particle beds, epoxy
resin, and EP/epoxy foams are presented. The elastic properties of low porosity solid
perlite and packed EP particle beds were characterized by means of elastic wave
propagation along the the cylinder axis of the specimens. By adopting an isotropic model,
the Young’s modulus and Poisson’s ratio were used to characterise the elastic response
of the medium. Moreover, elasticity theory was used to estimate the properties of pore
free solid perlite. Using these properties for solid perlite, porosity-elastic moduli relations
were investigated using four analytical models. The models were compared with the
experimental data from this study and their ability to estimate the elastic properties at
different porosity levels was evaluated.
Mechanical properties of cured epoxy resin (before and after dilution with acetone)
measured using quasi-static mechanical tests (according to ASTM standards) and elastic
wave tests are also provided and discussed. In addition, the results from hardness and
viscosity measurements for epoxy resin are presented and the effect of dilution on the
curing time reported.
Following the above characterisation of the resin, the compressive properties of EP/epoxy
foams characterised by quasi-static mechanical tests and the effects of density and particle
size are reported and discussed. Post compression microscopic and macroscopic
observations taken during the test are presented in order to understand the damage
87
mechanisms that occurred during the test and give additional insight into the material
behaviour under compressive loading. Moreover, based on the results of mechanical tests,
estimates for the contribution of the particle elastic modulus to the resulting foam
effective elastic modulus are shown.
The elastic properties of Ep/epoxy foams were then characterised using elastic wave tests
along the cylinder axis of the specimens in terms of Young’s modulus and the Poisson’s
ratio are presented. Subsequently, a comparison is made between the properties obtained
by mechanical and elastic wave tests. Furthermore, based on the results of the elastic
wave tests, the contribution of the constituents’ properties to the resulting foam properties
is discussed.
88
Elastic properties of sintered solid perlite
Using microscope images (shown in Figure 4.1) taken from the polished surface of the
samples and an image processing technique (MATLAB, MathWorks, MA), the porosity
of the sintered perlite samples was determined to be about 4%.
Figure 4.1. Microscope images taken from the polished surface of sintered perlite. The
polishing marks are designated on the picture in order not to be confused with porosity.
It has been shown that for porosity less than 5%, there is a linear relationship between the
elastic moduli and the porosity [208-211]. Within this range (0 - 5%), elasticity theory
successfully predicts the zero-porosity properties from the elastic moduli of porous solids,
by the following relations [212]:
89
)1(0 PEE E (4.1a)
)1(0 PGG c (4.1b)
)1(0 P (4.1c)
)1(0 PCC LLL (4.1d)
)1(0 PCC SSS (4.1e)
where )1129(18/1 0 E
3/5c
)18/(118/119/5 00 c
1)21()1()1()2(22/1 000020 EL
3/1S
and E, G, ν correspond to Young’s modulus, shear modulus, and Poisson’s ratio,
respectively, and subscript 0 refers to the zero porosity values of the properties. The
resulting zero porosity values of the density ρ0, longitudinal wave velocity CL0, shear
wave velocity CS0, Poisson’s ratio ν0, shear modulus G0 and Young’s modulus E0 of
sintered perlite are shown in row 2 of Table 4.1. For comparison, the corresponding values
from solid obsidian, measured by Manghnani et al. [213], are also presented. It can be
seen that the measured wave velocity for perlite with 4% porosity is closer to obsidian
than those for pore-free perlite. This might be due to the obsidian structure which is
typically vesicle-poor [214] (i.e. <2.5% porosity [215]), though the percentage of the
porosity was not reported by Manghnani et al. [213]. The comparison with obsidian
90
provides a benchmark for confidence in the measured values for the sintered perlite,
especially Poisson’s ratio.
Table 4.1. Elastic properties of solid component of EP particles. For comparison, the
corresponding data for obsidian by Manghnani et al. [213] are presented.
𝜌
(g/cm3)
𝐶𝐿
(m/s)
𝐶𝑆
(m/s)
𝐺
(GPa)
𝐸
(GPa)
𝜈
4% porosity Perlite 2.312 5616.44 3655.1 30.9 72.93 0.181
0% porosity Perlite 2.411 5836.75 3704.5 33.1 78.33 0.183
Obsidian [213] 2.357 5775 3608 30.7 75.16 0.180
91
Properties of packed beds of EP particles
4.3.1 Structural characterisation of EP particles
The bulk and particle densities of EP particles were measured according to the method given
in Section 3.3.1.1 and an average of five measurements for each particle size range is
presented in Table 4.2.
Table 4.2. Bulk density and particle density measurements of EP particles
particle size 1 - 2 mm 2 - 2.8 mm 2.8 - 4 mm
Bulk Density (g/cm3) 0.0703 0.0857 0.0939
Standard Deviation(g/cm3) 0.0003 0.0033 0.0007
Particle Density (g/cm3) 0.181 0.183 0.186
Standard Deviation (g/cm3) 0.016 0.010 0.004
Scanning electron microscopy (SEM) was used to characterise the microstructure of EP
particles. SEM images taken from the outer surface of an EP particle at each size range are
shown in Figure 4.2. As can be seen, the outer surface of the perlite particles is covered with
both closed and open pores and has a froth-like structure.
92
(a) (b)
(c)
Figure 4.2. SEM images showing the external structure of an EP particle: (a) in the 1 - 2 mm size
range; (b) in the 2 - 2.8 mm size range; (c) in the 2.8 - 4 mm size range.
93
(a) (b)
(c) (d)
Figure 4.3. SEM images showing the internal structure of an EP particle in the: (a) 1 - 2 mm size
range; (b) 2 - 2.8 mm size range; (c) 2.8 - 4 mm size range; (d) 2 - 2.8 mm size range, illustrating
natural reinforcement inside EP particle cells.
Examination of the different microscope images shows that the outer surface of the EP
particles in the size range 1 - 2 mm contains more open pores than the other two size ranges.
Figures 4.3 (a) - (c) reveal that the interior of expanded perlites in all three size ranges has a
three-dimensional cellular structure mainly composed of closed cells. Examination of the
94
micrographs also shows that EP particles in the size range 2.8 - 4mm are made up of
polyhedral cells, mostly possessing pentagonal dodecahedron and tetrakaidecahedron
geometries. However, the cells composing the structures of EP particles in the size range 1 -
2 mm and 2 - 2.8 mm do not appear to have a preferred geometrical shape; some are
polyhedral and some are almost spherical. Micrographs taken from EP particles in the size
ranges 2 - 2.8 mm and particularly 1 - 2 mm reveal the presence of inner cell reinforcement;
fibre-like reinforcements spanning the walls of a cell (see Figure 4.3 (d)). This phenomenon
was not observed in EP particles in the size range 2.8 - 4 mm. The governing conditions and
formation mechanism of the fibrous cell reinforcement are not known at this stage.
Examination of the micrographs taken from EP particles of each size range, cell size, wall
thickness and edge cell thickness7 were measured and are presented in Table 4.3. The
measured values show larger thicknesses for the edge cells than for the walls. This can be
ascribed to surface tension effects during the perlite expansion process which draws solid
into the cell edges and leaves a thin wall framed by a thicker edge [11, 216]. It can be seen
that the measured values are highest for particles in the size range 1 - 2 mm. This could
explain the higher crushing resistance of this particle size range, illustrated later in Section
4.6.2.1.
7 Edge thickness: as defined by Gibson and Ashby [17]
95
Table 4.3. Cell dimensions of EP particles of 1 - 2 mm, 2 - 2.8mm and 2.8 - 4 mm size range. Each
value is an average of 150 measurements.
Cell Size Wall Thickness Edge Cell Thickness
1 - 2 mm 84.497 µm 0.532 µm 3.067 µm
Standard Deviation 13.437 µm 0.100 µm 0.725 µm
2 - 2.8 mm 61.107 µm 0.512µm 1.493 µm
Standard Deviation 9.397 µm 0.092 µm 0.438 µm
2.8 - 4 mm 53.494 µm 0.516 µm 2.425 µm
Standard Deviation 8.047 µm 0.092 µm 0.472 µm
Fine et al. [217] suggested a geometric relationship given by Eq. (4.2) for calculating the wall
thickness of a glass hollow microsphere from its relative density (true density of a
microsphere divided by the density of glass).
3
1
11
g
truew
r
t
(4.2)
with 𝑡𝑤= wall thickness of the microspheres, r = average radius of the microspheres, ρtrue =
true density of the microspheres, and ρg = density of the glass. Using the results presented in
Table 4.3 and measured particle densities presented in Table 4.2, Eq. (4.2) may be modified
to adapt to the perlite morphology. Two Eqs. (4.3) and (4.4) are suggested for calculating
the ratio of the wall thickness (tw) and cell size (C) to the EP particle diameter (dp) in terms
of their relative density:
96
x
UP
EP
p
w
d
t
5.15.1 (4.3)
y
UP
EP
pd
C
5.15.1 (4.4)
where 𝜌𝐸𝑃 is the EP particle density and 𝜌𝑈𝑃 is the density of fully dense perlite. The
exponents x and y for the three particle size ranges obtained by fitting the experimental results
(Table 4.3) to Eq. (4.3) and Eq. (4.4) are presented in Table 4.4.
Table 4.4. Exponents of Eq. (4.3) and Eq. (4.4) for the three particle size ranges.
x y
1 - 2 mm 1.25 1.075
2 - 2.8 mm 1.20 1.15
2.8 - 4 mm 1.20 1.17
4.3.2 Measurement of elastic moduli of packed EP particle beds using quasi-static
mechanical tests
The elastic moduli of EP particles as a function of density were measured by conducting
several loading-unloading compressive tests on packed beds of EP particles (explained in
detail in Section 3.4.1). Considering the correlation coefficient (R2) of unloading gradients
and the deviation between them, it was possible to determine the local yield point for that
compacted sample. Below the determined yield point was considered to be the elastic region
97
where the confined elastic modulus could be measured. The results as a function of compact
density are given in Figure 4.4.
Figure 4.4. Constrained modulus of packed EP particle beds as a function of compact density.
Since the particles used in this test were confined, the calculated unloading gradient was not
identical to the elastic modulus of the packed bed of particles but a function of both their
elastic modulus (EP) and Poisson’s ratio (𝜈). Considering 휀𝑦 = 휀𝑧 = 0, the confinement
introduced by the wall of the cylinder, and 𝜎𝑦 = 𝜎𝑧, the unloading gradient (𝜎𝑥 /휀𝑥) or
constrained elastic modulus E* is defined as [218]:
)21()1(
)1(*
PEE (4.5)
There are various methods to determine Poisson’s ratio (𝜈) in the geotechnical literature. A
common one [219] is given by Eq. (4.6):
98
k
k
1 ;
245tan2
k (4.6)
where 𝜑 denotes the angle of internal friction. However, it was found that the Poisson’s ratio
(𝜈) calculated from Eq. (4.6) overestimates the real value of this coefficient. Therefore, it is
not applicable where the methods of theoretical elasticity are applied [220]. Sawicki and
Świdziński [220] proposed a method for determining the elastic moduli of particulate
materials using confined compressive tests, with additional measurement of the lateral
stresses. Although their method provides more exact solutions for the elastic moduli, it
requires the conduct of biaxial compressive tests. Performing such tests was not possible in
this study due to limitations of the testing machine. As a result, the study was extended to
measure the elastic moduli of packed beds of EP particles using elastic waves, as explained
in the next section.
4.3.3 Measurement of elastic moduli of packed EP particle beds using elastic waves
The compression and shear wave velocities of packed beds of EP particles at different
densities are shown in Figure 4.5. As can be seen, both the compression and shear wave
velocities reach a plateau for densities higher than 0.2 g/cm3. At lower densities, particles
have more space available to move or rotate. At higher densities, the relative movement of
particles is restrained as the particles reach an optimal structural arrangement (i.e. the particle
fabric). Further compression reduces pore volumes between the EP particles by fracturing
particles which produces both smaller cellular EP particles and fine platy debris such as that
shown in Figure 4.6. Because of the closed cellular EP structure, microscopic analysis
99
conducted on EP particles in a previous study [15], indicated that debris does not form inside
EP particles i.e. the debris is restricted to the inter-particle spaces.
(a)
(b)
Figure 4.5. The (a) Compression wave and (b) Shear wave velocities versus compact density.
100
Figure 4.6. Formation of platy and fine particles as a result of the brittle crushing of cell walls.
The use of static density in Eqs. (3.1), (3.2), (3.4) and (3.6) can be justified only if the debris
are mechanically linked to the particles [221]. If the debris particles are free moving, they
should not be taken into account in the evaluation of the dynamic compact density.
Consequently, further investigation of the debris was undertaken. The debris concentration
was investigated by separating the EP particles from the debris and measuring their relative
masses at each compact density. Mechanical separation was ineffective without causing more
particle fracture and debris. Therefore, the EP particle compacts were immersed in water for
24 hours to disperse. The dispersion resulted in two phases: a top phase comprising EP
particles floating on the surface and sediment. The sediment was considered to be 100%
debris due to its higher density (note the solid density of perlite in Table 4.1). Beakers
containing the EP particles and sediment were placed in an oven at 120 °C for 24 h. The
debris size was measured and found to be less than 212 µm which is used here to discriminate
between particles and debris. To be sure that all the debris had separated from the EP
particles, the samples were sieved and the fraction bellow 212 µm was added to the debris.
Figure 4.7 (a) illustrates the weight percentage of debris as the EP particle compact density
101
increases. This is also demonstrated by the particle size distribution measured after
compaction which is shown in Figure 4.7 (b).
(a)
(b)
Figure 4.7. (a) Percentage of debris at each EP compact density. (b) Cumulative particle size
distribution for entire compacts (EP particles and debris).
102
It can be seen that as the compact density increases due to the breakage of particles, the
percentage of larger particles decreases and the amount of debris increases. This post
compaction investigation revealed that the mass of EP particles that resisted crushing at
densities higher than 0.2 g/cm3 remained relatively constant. This was very strongly
correlated with the plateau region of the compression wave and shear wave velocity versus
density curves, suggesting that only the EP particles contribute to sound wave propagation
and the debris plays little role. This was further validated by investigating the evolution of
the inter-particle porosity as compact density increases. To this end, the density of different
EP particles size ranges was measured using the wax immersion method. In this method, wax
was melted in a 5ml measuring cylinder and kept at a constant temperature for each
measurement. EP particles of known mass and size were immersed in the melted wax and
their particle density was measured through the change in the wax’s volume. The results are
shown in Figure 4.8 (a). Using the results shown in Figure 4.7 (b) and Figure 4.8 (a), the
inter-particle porosity (excluding debris) within the total sample volume was calculated and
is presented in Figure 4.8 (b). It can be seen that the inter-particle porosity does not change
significantly as the compact density increases. This can be explained by the breaking of the
original EP particles into smaller EP particles as the compact density increases. The smaller
particles created new inter-particle porosity the increase of which was offset by the
simultaneous production of debris. Considering the mass fraction of the debris at each
compact density (see Figure 4.7 (a)) and the results shown in Figure 4.8 (a), the volume
percentage of debris with respect to the total sample volume as well as with respect to the
inter-particle porosity volume was calculated (Figure 4.8c). As can be seen, the debris does
103
not occupy a significant portion of the total sample volume and the inter-particle porosity
volume due to the very high density of the debris compared with the density of EP particles.
Therefore, the raw compact densities were modified by subtracting the debris mass from the
total mass of the EP particles and new densities were calculated. The modified results are
plotted along with the experimental results in Figure 4.5.
It should be noted that both of these results are useful. In practical situations, it is not usually
feasible to separate the debris from the EP particles. However, the raw or experimental results
are readily measured and reflect the effective state of the granular body. Hence, these results
are of most practical interest. Otherwise, if the properties of pure EP particles are desired for
more in-depth studies, the modified results are appropriate.
Young’s moduli of packed beds of EP particles, corresponding to the wave velocities in
Figure 4.5, calculated based on both raw compact density and modified density using Eq.
(3.4) is shown in Figure 4.9 (a). Figure 4.9 (a) also illustrates the normalised value of Young’s
modulus (E/E0) versus the compact porosity. Both curves, calculated from the experimental
and modified wave velocity results show that the Young’s modulus of packed beds of EP
particles increases as the density of the compact increases. In addition, these two curves
provide an upper and lower bound for Young’s modulus of EP particle compacts. In the
authors’ view, the lower bound is closer to the true value considering the minor mechanical
role of the debris.
Figure 4.9 (b) shows Poisson’s ratio of EP particle compacts as a function of compact density
as well as the normalised Poisson’s ratio (𝜈/𝜈0) as a function of porosity. As the calculation
of Poisson’s ratio is independent of density (Eq. (3.5)), the values obtained from the
104
experimental and modified wave velocities resulted in the same Poisson’s ratios. Poisson’s
ratio does not show as large a variation with density as Young’s modulus does within this
range of compact density (or porosity). It increases up to density 0.25 g/cm3 followed by a
decreasing trend however the changes are small (<10% over the whole range). There is no
general consensus among researchers about the influence of porosity on Poisson’s ratio.
While some consider Poisson’s ratio as a constant [222], others claim it to be a function of
porosity [223-225]. Such debates may arise from the fact that the Poisson’s ratio is usually
derived as a function of two other elastic moduli (i.e. Young’s and shear moduli) whose
relative dependence on porosity may intensify or supress the dependence of Poisson’s ratio
on porosity. The evolution of Poisson’s ratio in the low density compacts seems to be affected
by the pore character rather than the solid phase material properties. The compacted bed of
EP particles has a double-porosity structure; intra-particle porosity within the individual
particles and inter-particle porosity between the perlite particles. Thus, there are two
phenomena which contribute to Poisson’s ratio of the samples: Poisson’s effect of unit cells
within individual particles and Poisson’s effect of the whole particle. This suggests that in
the low density samples (ρ < 0.15 g/cm3), the bulk elastic response may be dominated by the
Poisson effect of cells as well as the compliance of the inter-particle contact region. The
initial contact areas are relatively small and well separated, hence localised lateral
deformation (Poisson effect) has a greater probability of intruding into the porous region
between particles rather than straining the adjacent particle. Therefore, the overall lateral
deformation is reduced, resulting in suppression of the Poisson’s ratio. Further increase in
the density reduced the porosity between particles and increased the contact size along with
105
increasingly more solid phase material being involved in load transfer throughout the bulk.
Consequently, the lateral deformation in both unit cells and individual particles contributes
to the observed Poisson’s ratio.
(a)
(b) (c)
Figure 4.8. (a) Particle density versus particle size. This graph also includes debris density versus
debris size; (b) Inter-particle porosity (excluding debris) versus compact density; (c) Inter-particle
space filled by debris versus compact density.
106
(a)
(b)
Figure 4.9. (a) Young’s modulus (E) of packed beds of EP particles versus compact density. (b)
Poisson’s ratio of packed beds of EP particles versus compact density. In both graphs, the upper and
right hand scales allow the normalised moduli versus porosity to be read from the same graphs.
107
Mathematical models for prediction for Elastic properties of porous
bodies
A large number of empirical and semi-empirical models have been developed to explain the
porosity dependence of material properties. The majority of these models are successful in
predicting elastic properties within a porosity range of less than 38% (e.g. [222, 226, 227])
and some assume the vanishing of elastic properties at porosity of 50% (i.e. as a result of an
implicit assumption of symmetry between pore and solid phase [228-231]). However, this is
not valid as many granular and porous materials have measurable elastic properties for
porosities greater than 50% (e.g. bodies made of hollow spherical particles [232]).
Consequently, models have been developed which cover the other end of the porosity
spectrum (>50%). In this study, four of these models were used to predict the elastic
properties of EP particle compacts of different densities. It is noteworthy that when choosing
a model, three criteria were considered: (i) interpretation of the physical parameters of the
model; (ii) whether standard relations of linear elasticity hold between all moduli; and (iii)
the ability of the model to satisfy the boundary conditions. The boundary conditions include
the prediction of the solid phase properties at zero porosity and the presence of a critical
porosity, or percolation limit, at which particles no longer form a continuous network. Hence
stiffness (i.e. E, G, K) goes to zero and the Poisson’s ratio approaches an asymptotic value.
This value has been found to be independent of the Poisson’s ratio of the solid phase but is
dependent on the geometry of the solid phase at the critical porosity [233]. Experimental and
numerical investigations have shown that, regardless of the solid phase Poisson’s ratio, the
Poisson’s ratio of a porous body asymptotically approaches a fixed value which has been
108
identified differently to be 0.2 [234-236] or 0.25 [223, 237]. Dunn [225] analytically
investigated four different pore shapes as a function of the aspect ratio of a spheroid and
found pore shape is another factor affecting the asymptotic behaviour of Poisson’s ratio. It
was found that the asymptotic value for spherical and needle shaped pores is 0.2 while for
disk-shaped pores and penny shaped cracks it is 0. With respect to experimental and
analytical findings, the critical Poisson’s ratio is considered to lie in the range 25.00 cr
at critical porosity, crP . For a granular material, crP falls within the range of gravity induced
packing states Ptap < Pcr < Pgreen, where Ptap and Pgreen are the porosities of the ‘tapped’ and
‘as-poured’ packing states [238]. In the evaluation of the models, the value of Pcr was
considered to lie within these bounds in order to preserve its physical interpretation.
It should be noted that particle crushing and the production of debris is not taken into account
in any of the models’ assumptions. Hence, it would be more appropriate to apply these
models to the modified values. However, to investigate the effect of particle debris on the
physical features of the models, the models were applied to moduli determined using both
the raw compact densities (hereafter the experimental moduli) and moduli calculated using
the modified densities, hereafter the modified moduli. In the following, each model and its
physical features are explained, and the applicability of these models to the EP particle
compaction data is discussed.
109
4.4.1 Phani Models
Phani and Niyogi [239] derived a semi-empirical relation based on a simple model of
applying a uniform state of stress on a porous body of constant cross-sectional area and
constant length. This model has been found to agree well with the data from many
polycrystalline brittle solids over a wide range of porosity [239-242] and is expressed by:
EnaPEE )1(0 (4.7)
where E and E0 are Young’s modulus at volume fraction porosity P and zero, respectively,
and a and En are packing geometry and pore structure dependent parameters, respectively.
This model satisfies the boundary conditions: E = E0 at P = 0, and E= 0 at P = Pcr = 1/a.
Rice [243] theoretically derived the values of Pcr to be 0.785, 0.964 and 1.0 for the cubic
staking of cylindrical, spherical and hexagonal pores, respectively. Knudsen [244] studied
the contact area as a function of bulk density and calculated the value of Pcr for rhombohedral,
orthorhombic and cubic packing of spherical particles to be 0.26, 0.397 and 0.476,
respectively. From the theoretically derived values of Pcr and the existing relationship
between Pcr and parameter a (i.e. Pcr = 1/a), the value of a lies in the range 85.31 a . For
random packing with isolated spherical pores, the constants a and En were found to be about
1 and 2, respectively [239]. However, the increase in En corresponds to the transition of
pores from being spherical to being more interconnected.
Phani and Sanyal [245] derived a relation between the shear modulus and Young’s modulus
of an isotropic porous solid based on the Mori–Tanaka mean-field approach, in which the
110
elastic properties are obtained by subjecting the inclusion (i.e. pores) to an effective stress or
strain field. This relation is given by:
0
0
0
0
0
0
3
21)1(
3
2G
E
E
E
EG
n
(4.8)
where G is the shear modulus, E is predicted from Eq. (4.7) and 0n is a constant for a given
data set of a porous material related to pore morphology. The value of 0n can be evaluated
using the experimentally measured values of E and G at a single porosity value. This model
satisfies the boundary conditions: G = G0 when E = E0 at P = 0, and G = 0 when E = 0 at P
= Pcr.
Predicted values of Young’s moduli from Eq. (4.7) along with the ones for shear moduli from
Eq. (4.8) were used to evaluate Poisson’s ratio by,
12
G
E (4.9)
as a function of porosity. The required data for pore-free solid phase (i.e. E0, G0 and ν0) were
obtained from Table 4.1. Non-linear regression coefficients to fit the model to the
experimental and modified moduli are presented in Table 4.5. The regression value for nE in
both cases (i.e. based on the experimental and modified moduli) was indicative of the
compacts’ pore structure being non-spherical and partially interconnected. The regression
value of the constant a (i.e. Pcr = 1/a) gives a Pcr of 1.0. For packed beds of EP particles, the
range Ptap < Pcr < Pgreen was experimentally determined to be 0.96 < Pcr < 0.98. Thus, the
estimated value of Pcr by Eq. (4.7) is slightly higher than the expected range.
111
Table 4.5. Physical parameters of mathematical models applied to the experimental and modified
moduli. The physical parameters of the modified moduli are given in parenthesis.
Model Name Physical Parameters
Phani a = 1 (1); nE = 2.8 (3.095); n0 = 1.104 (1.053); Pcr = 1 (1)
Nielsen β = 0.24 (0.12); Pcr = 0.965 (0.979)
Rice1 bʹE = 0.036 (0.02); bʹG = 0.0327 (0.02); Pcr = 0.976 (0.978)
Wang b = 3 (3.63); c = 3.1(3.08); d = 0.95 (1); Pcr = 1 (1)
Gibson-Ashby C1 = 0.1 (0.068); C1ʹ = 0.1 (0.072); C2 =0.037 (0.026);
C2ʹ = 0.039 (0.0276); Pcr = 1 (1)
1. The subscript E in bʹE refers to Young’s modulus and subscript G in bʹG refers to shear modulus
Application of the Phani model to both the experimental and modified moduli is illustrated
in Figures 4.10 (a) - (d). In both cases, Young’s modulus values predicted by Eq. (4.7) show
a considerable deviation (see Figures 4.10 (a) and (c)). Although the predicted Poisson’s
ratios (Figures 4.10 (b) and (d)) do not follow the same trend as the experimentally measured
values, they show slightly better agreement with the ones obtained based on the modified
densities. It should be noted that Poisson’s ratio is a small quantity dependent on the
differences of other elastic moduli and is hence very sensitive to errors in them [212]. The
considerable deviation from the experimental moduli of values predicted by the Phani model
might be ascribed to the effect of the shape factor nE as a single parameter to reflect the
112
change in porosity as compaction proceeds. This effect is discussed in more detail in Section
4.4.2.
(a) (b)
(c) (d)
Figure 4.10. (a) Normalised Young’s modulus versus porosity (based on experimental moduli); (b)
Normalised Poisson’s ratio versus porosity (based on experimental moduli); (c) Normalised
Young’s modulus versus porosity (based on modified moduli); (d) Normalised Poisson’s ratio
versus porosity (based on modified moduli).
113
4.4.2 Nielson Model
Nielsen [246] proposed a model for predicting the elastic moduli of porous materials based
on the composite sphere assemblage (CSA) approach. CSA considers a porous body to be
constituted of congruent composite elements consisting of a spherical pore embedded in a
concentric spherical shell of matrix material. It is assumed that the composite spheres are
available in an infinite range of sizes and they are distributed in a way where smaller
composite spheres fill all the interstices between larger spheres. This model was originally
introduced by Hashin and assumed the applied stress on the assembled body is uniformly
distributed (hydrostatic) around each inclusion so that strains can be calculated to obtain the
elastic moduli for the body. Nielsen set forth to predict the elastic moduli for porous materials
by assuming two types of composite elements: one as described above and the other one
made of a spherical matrix embedded in a concentric spherical shell of pore. These two types
of CSAs defined two bounds of continuous solid phase (isolated pores) and continuous pore
phase (isolated solid phase) where transition between them (different mixtures of them)
results in porous material with a wide range of porosity of any geometry. The Nielsen model
for Young’s modulus and Poisson’s ratio are expressed as:
)1315)(1())(75(2
))(75)(1(2
000
00
PPPP
PPPEE
crcr
cr (4.10)
)1315)(1())(75(2
)35)(1())(75(2
000
0000
PPPP
PPPP
crcr
crcr (4.11)
where 𝛽 is a shape factor characterising the ability of the material to transfer stress and
providing information on the shape of the pores and their inter-connectivity. The shape factor
114
𝛽 is in the range 0< β <1. The lower bound, 𝛽 = 0, is when the pore phase totally surrounds
the solid phase (i.e. particle) while the upper bound, 𝛽 = 1, corresponds to when the solid
phase completely surrounds the pore phase (i.e. isolated pores). Hence, the lower value of 𝛽
correlates with highly interconnected pores and smaller contact areas between particles,
while higher values of 𝛽 indicates that pores are becoming increasingly more isolated.
The Nielsen model satisfies the boundary conditions at zero porosity and correctly predicts
both of the solid phase elastic moduli (i.e. E = E0, 𝜈 = 𝜈0). It also predicts the presence of
the critical porosity at which Young’s modulus goes to zero and Poisson’s ratio is given by
1315
35
0
0
Cr . Given that Poison’s ratio for an isotropic elastic material lies in the
physically realistic range 0 ≤ 𝜈0 ≤ 0.5, this model predicts critical Poisson’s ratio to lie in
the range 0.09 ≤ 𝜈𝑐𝑟 ≤ 0.23, which does not cover the expected range 0 ≤ 𝜈𝑐𝑟 ≤ 0.25.
This limited range may have a significant effect on the quality of predicted values for
Poisson’s ratio.
Non-linear regression analysis was conducted to fit the Nielson model to the experimental
and modified moduli and the results are shown in Table 4.5. The predicted values of Pcr for
both the raw compact and modified densities were within the specified range 0.96 < Pcr <
0.98 and the low value of 𝛽 is representative of a high-volume fraction of pores compared to
the solid phase. Figures 4.10 (a) and (b) show that the agreement between the Neilson model
and the experimental moduli is not very good. In the case of Poisson’s ratio, the discrepancy
can be attributed to the low value of νcr (i.e. 0.09 ≤ 𝜈𝑐𝑟 ≤ 0.23) and the shape factor β.
Similar to nE in the Phani model, the shape factor β in the Nielson model is used as a single
115
value to accommodate microstructural change during the compaction of compacts of
different densities and with a wide range of pore shapes and geometries. This problem might
be alleviated by solving Eqs. (4.7) and (4.10) point by point. The results of this calculation
for both of the shape factors (nE and β) as a function of porosity are presented in Figure 4.11
(a). As can be seen, both shape factors change as a function of porosity and the results show
the opposite of what was expected. At lower porosity 𝛽 approaches zero while at successively
higher porosities, where the structure is open and interconnected, it approaches 0.5 for the
experimental moduli and exceeded the upper bound of unity for the modified moduli. Similar
to 𝛽, the evolution of the shape factor nE with porosity does not follow the expected behaviour
defined by Phani and Niyogi. To accommodate these differences, the shape factors were
expressed as a function of porosity P, given by:
134
2321 ....)( n
n PPPPPPS (4.12)
where S is a shape factor, either nE or β and 𝜂𝑛 (n = 1, 2, 3, ...) are empirical constants.
Applying the modified form of the shape factor in Eq. (4.12) to the experimental and
modified moduli up to n = 4, both Nielson’s and Phani’s models were replotted and are
presented in Figures 4.11 (b) and (c). It can be seen that both Nielson’s and Phani’s models
for Young’s modulus show close agreement with the experimental moduli when the shape
factor was defined as a function of porosity. In the case of Poisson’s ratio, the modified form
of the shape factor improved the results of the Phani model but not those from the Neilson
model (see Figure 4.11 (c)). Therefore, as the Nielson model does not exhibit good agreement
with both the Young’s modulus and Poisson’s ratio, it is not considered to be a good model
for the prediction of the elastic properties of EP particle compacts.
116
(a) (b)
(c)
Figure 4.11. (a) Shape factor versus compact porosity for the Phani and Neilson models applied to
both the experimental moduli and modified moduli; (b) Modified Phani and Nielson models for
Young’s Moduli as a function of porosity; (c) Modified Phani and Nielson models for Poisson’s
ratio as a function of porosity.
117
4.4.3 Minimum solid area (MSA) models
The minimum solid area (MSA) model assumes that the macroscopic elastic response is
related to the load-bearing area of the solid phase material. The key assumption in MSA
models is that normalised elastic moduli scale according to the minimum area of solid phase
per unit area in a plane perpendicular to the direction of the applied load, i.e. MSAE
E
0
. In
this context, a common approach is to assume an idealised microstructure, including the
regular stacking of solid particles in a void matrix, and the regular stacking of pores (spheres,
cylinders, etc.) in a solid matrix. The initial and major development of such models was
Knudsen’s model [244]. Knudsen investigated the ideal microstructure of three particle
stackings (i.e. simple cubic, orthorhombic and rhombohedral) and found that the stacking of
particles, and hence the pores, is significant. In addition, plotting of the resultant models as
a function of porosity on semi-log plots for low to intermediate porosity was approximated
by:
bPeEE 0 (4.13)
where b is related to particle stacking and hence is a function of pore shape, geometry and
alignment with respect to the stress axis. This equation has been criticised [227] for not
satisfying the boundary condition that E = 0 at P = 1. However, such criticism is only valid
when critical porosity occurs, at about unity. Anderson [247] analytically proved the
existence of such exponential dependence on porosity from strain analysis of isolated
ellipsoidal pores. He stated that Eq. (4.13) becomes invalid at very high pore fractions where
constant b does not reflect pore interactions occurring as the concentration of pores increases.
118
Rice [243] discussed that this relationship (Eq. (4.13)) provides a good approximation for
effective properties up to 1/3-1/2 of 𝑃/𝑃𝑐𝑟 and holds true for all elastic moduli. In attempts
to estimate the minimum solid area for higher porosity, Rice and Wang proposed two
equations. Rice [248] suggested that the role of pores and solid phase material can be
exchanged (i.e. stacking of pores rather than particles) leading to the theoretical equation:
)1()/1(
0
'crPPb
eMM
(4.14)
where M is the Young’s, shear or bulk moduli and 𝑏′ is related to the pore stacking. Rice
discussed that 𝑏′ decreases as b increases. For instance, for spherical pores in cubic stacking,
the value of b is 2.7 while the value of 𝑏′ is 0.5. For other pore shapes and stackings, the
values of b and 𝑏′ can be obtained from Figure 2 in [248]. This model predicts zero stiffness
as the porosity goes to unity, but it does not satisfy the zero porosity boundary condition
since (1 − exp (−𝑏′(1 − 𝑃/𝑃𝑐𝑟)) does not go to unity as P goes to zero. Rice explained this
by expressing that Eq. (4.14) is a continuation of Eq. (4.13) for higher porosity values and
no extrapolation between them is required.
Wang [222] used the ideal model of spherical particles in a simple cubic stacking developed
by Knudsen [244] and modified it for real microstructures. The modifications account for
misalignment of the uniaxial applied stress, which induces shear and hinge effects at the neck
(the small contact area between two particles). Wang derived a complex relation between
porosity and Young’s modulus, and proposed an approximate solution to his exact solution
which could satisfactorily cover a wide range of porosity, as given by:
...)](exp[ 32 dPcPbPEE S (4.15)
119
where b, c and d are empirical constants. Brown et al. [249] analytically derived the value of
b for different idealised pore geometries and orientations and proposed a similar equation to
Wang’s model, but in the context of strength, as given by:
j
i
i
ii
i
ii
i
ii
i
i Pbj
PbPbPb )(!
1)(
!3
1)(
!2
1)/ln( 32
0 (4.16)
where ib is related to the pore geometry and orientation of the ith kind of pore and iP is the
contribution made by pores of the ith kind to the total porosity. Therefore, the constants of
Eq. (4.15) can be expressed in terms of the constants of Eq. (4.16). Wang has shown that the
proposed equation with a quadratic exponent can predict the Young’s modulus of a porous
material up to a porosity of 38% [222] and additional higher order terms can be included for
higher porosities.
The Wang model with a cubic exponent was used for the prediction of elastic moduli of EP
particle compacts. Non-linear regression analysis was conducted to fit the Wang and Rice
models to the experimental and modified moduli. The results are presented in Table 4.5. The
critical porosity value predicted by the Wang model is slightly higher than the specified range
0.96 < Pcr < 0.98, while the value predicted by the Rice model (for both shear and Young’s
moduli) is within the range. As shown in Figures 4.10 (a) and (c), the Rice model shows very
close agreement with the experimental moduli, however, it did not show such agreement with
the modified moduli. Similarly, the Wang model shows better agreement with the
experimental moduli than with the modified Young’s modulus values (see Figures 4.10 (a)
and (c)). The regression parameters in Wang’s model for the experimental Young’s moduli
in terms of the constants in Eq. (4.16), correspond to; b - cylindrical pores aligned
120
perpendicular to the loading direction; c - oblate pores (with a ratio of about 0.8) in a cubic
stacking; and d - a combination of about 70% cubical pores in <100> orientation with 30%
cylindrical pores parallel to the loading direction [250-252], respectively. The interpretation
of the regression parameters in Wang’s model for the modified Young’s moduli is similar to
those for the experimental moduli, except for constant b (3.63) which corresponds to a
combination of about 77% cubical pores in <110> direction with 23% cubical pores in <111>
direction. This approximation seems reasonable with respect to the cell shape in EP particles
(Figure 4.3) and the fact that pores can have different directions in a real packing of particles.
The predicted Young’s and shear moduli based on both experimental and modified moduli
were found to lie within the simple cubic and rhombohedral packing of pores. For
comparison, a similar range of values can be acquired from Figure 2 in [248]. Rice discussed
how porous bodies are better modelled by combining two or more idealised pore structures,
as opposed to just one. Consequently, the values of 𝑏′ (Table 4.5) can be ascribed to different
combinations of these two pore packing geometries (e.g. simple cubic and rhombohedral)
and their interaction as porosity evolves. Poisson’s ratio values were calculated using the
predicted Young’s and shear moduli values from Eq. (4.14) and then combining them using
Eq. (4.9). Poisson’s ratio values predicted by the Rice model based on raw compact density
show close agreement with the experimentally measured Poisson’s ratio values. However,
the deviation of Young’s and shear moduli values predicted by the Rice model based on
modified moduli resulted in poor predictive results for Poisson’s ratio (see Figures 4.10 (b)
and (d)). As a whole, the Rice model was shown to be successful in predicting the elastic
properties of EP particle compacts only when the raw compact density is considered
121
4.4.4 Gibson and Ashby Model
Gibson and Ashby derived a semi-imperial model for the relationship between elastic moduli
and porosity by employing dimensional arguments (using standard beam theory) for a
cellular solid. For simplicity, they assumed cell struts and walls of uniform dimensions, while
in reality these will usually be tapered toward, and thinner in, the centre (i.e. due to the effect
of surface tension during the foaming process). The correction for this tapering was
implemented by arranging the cubic cells in a staggered stacking, so that their members meet
at their midpoint. This model has been shown by many authors (e.g. [253-258]) to
successfully predict the elastic moduli of closed cell foams, which are given by the following
expression:
)1)(1(1 '
1
22
1
0
PCPCE
E
(4.17)
)1)(1(1 '
2
22
2
0
PCPCE
G
(4.18)
Where 𝐶 and 𝐶′ are constants of proportionality, Ø is the volume fraction of solid in the cell-
struts ,which can be obtained from the relative area of the struts and faces on a plane section
as explained in [259], and the remaining fraction (1− Ø) is the solid contained in the cell
faces. The quadratic term describes the contribution of the cell struts bending to the modulus
and is the same for open cell foams. The linear term corresponds to the cell walls’ lateral
stretching. Gibson and Ashby suggested 𝐶1 and 𝐶1′ should be about unity and 𝐶2 and 𝐶2
′
should be about 3/8 for porous materials with 𝜈0 = 0.33, where these satisfy the boundary
conditions at P = 0 and P = Pcr = 1. However, the values of 𝐶1 and 𝐶1′ will change based on
122
the volume fraction of solid in the cell struts, the variable geometries of the foam and
uncertainty in the value of the solid Young’s modulus, E0 [97]. As a result, the values of these
constants depend on the type of foam and vary from one foam to another.
The Gibson–Ashby relation provides a good approximation for the cellular structure of the
EP particles shown in Figure 4.3. For a given porosity and solid fraction of cell struts (Ø =
0.78), the regression analysis yielded the values 𝐶1 and 𝐶1′ in Eq. (4.17) for both the
experimental and modified Young’s moduli (see Table 4.5). Application of the Gibson and
Ashby model to the experimental and modified Young’s moduli is presented in Figures 4.10
(a) and (c). This model shows good agreement with the experimental Young’s modulus
values. In the case of the modified Young’s moduli, the Gibson and Ashby model
demonstrates fair agreement with the moduli for porosity higher than 0.87, while for lower
porosity a considerable deviation is observed.
Gibson and Ashby suggested that if 𝐶1 and 𝐶1′ are about unity and
8
3
E
G, which holds true
for polycrystalline metallic materials [260], Poisson’s ratio of the foam is about 0.33. Since
constants 𝐶1 and 𝐶1′ are not unity (see Table 4.5), this relation does not hold and Poisson’s
ratio should be calculated using two elastic moduli. For this purpose, the Young’s moduli
from Eq. (4.17) along with the shear moduli predicted from Eq. (4.18) were used to calculate
Poisson ratios by Eq. (4.9). For a given porosity and solid fraction of cell struts (Ø = 0.78),
the regression analysis yielded the values of 𝐶2 and 𝐶2′ in Eq. (4.18) for both the experimental
and modified shear moduli (see Table 4.5). It is noteworthy that although 𝐶1 and 𝐶1′ are not
unity, the ratio of the shear to Young’s moduli is almost 3/8, which gives implicit satisfaction
of the boundary condition (explained above). The results for Poisson’s ratio are presented in
123
Figures 4.10 (b) and (d). As can be seen, the Poisson’s ratio values predicted by the Gibson
and Ashby model show good agreement with both the experimental and modified Poisson’s
ratios. Overall, the Gibson and Ashby model gives a satisfactory means of predicting the
elastic properties of EP particle compacts based on the experimental and modified moduli.
124
Properties of epoxy resin
The mechanical properties of the cured epoxy resin, with and without dilution with acetone,
were measured according to the specified ASTM standards and are presented in Table 4.6.
Table 4.6. Mechanical properties of cured epoxy resin with and without dilution by acetone.
Properties Epoxy/Hardener Epoxy/Hardener/Acetone
Hardness (Shore D) ASTM D-2240 82 57
Compression yield ASTM D-695 103 MPa 3.40 MPa
Compressive modulus ASTM D-695 3170 MPa 1002 MPa
Tensile strength ASTM D638 45.40 MPa 7.50 MPa
Tensile elongation ASTM D-638 7.3 % 3.20%
Tensile Modulus ASTM D-638 2910 MPa 1834.7 (2022.4) 1 MPa
Flexural strength ASTM D-790 184 MPa 35.70 MPa
Flexural Modulus ASTM D-790 7365 MPa 1370 MPa
Poisson’s ratio 0.39 (0.37) 0.41 (0.38)2
Density 1.18 1.14 (g/cm3)
1 & 2. The measured properties using elastic wave tests are given in parenthesis.
Dilution with acetone changed the curing time of the epoxy which was investigated by
monitoring hardness change using a durometer (type D). The results presented in Figure 4.12
are indicative of a minimum curing time of 45 days for diluted epoxy as no change in
hardness was observed after this period. The dilution with acetone at this ratio reduced the
125
mix viscosity of epoxy resin+hardner from 1.80 Pas (with no acetone) to 0.0067 Pas (with
90 wt% acetone); viscosity was measured, shown in Figure 4.13, using RST-Brookfield
rheometer and the data were subsequently analysed using the Bingham model [261]. The
point for 100% epoxy + hardener is not shown in the graph due to scale limitations.
Figure 4.12. Hardness versus time curve used to determine the curing period of the diluted epoxy.
Figure 4.13. Changes in the viscosity of epoxy + hardener diluted with acetone versus acetone
content.
126
Properties of EP/epoxy foams
4.6.1 Microstructural characterisation of EP/epoxy foams
Scanning electron microscopy was used to examine the microstructure of foam samples
before undergoing compression tests. Figure 4.14 shows micrographs taken from cross-
sections of the three foam types with the same density. It can be seen that the epoxy has filled
the space between the particles and did not penetrate into the perlite particles, which explains
the low density of the foams. The SEM images also show good bonding between the perlite
particles and the epoxy binder as few gaps were seen at their interfaces. However, this may
not have a significant effect on the resulting properties (mechanical, physical, etc.). This can
be due to the porous structure as well as the size and volume fraction of the perlite particles.
The porous structure of perlite inhibits a continuous layer of adhesion between the particle
and matrix, compared with if there is a solid continuous surface. Regarding the particle size,
it has been reported [262] that particles of smaller size have a higher surface area to volume
ratio leading to better interfacial interaction with the matrix. The maximum of such
interactions for spherical and near-spherical particles was achieved at the size range of
nanometer and submicrometer particles [120, 263-265]. It was found that the particle
contents in the range 25 - 30 vol% provided the maximum interfacial interaction between the
filler (particle) and the matrix. At higher particle content, the overlap of interfacial adhesion
between adjacent particles reduced the overall filler-matrix interfacial interaction [263]. In
the present study, however, the size of the particles ranged from 1 to 2, 2 to 2.8 and 2.8 to 4
mm and the volume fraction ranged from 46 to 58% (to be explained in Section 5.4).
127
Therefore, adhesion between the particles and the epoxy cannot be considered to have had a
notable effect on the resulting mechanical properties. This will be investigated further in
Section 4.6.3.
(a) (b)
(c)
Figure 4.14. Micrographs taken from the cross-sections of EP/epoxy foams with a density of 0.15
g/cm3 made with EP particles in the size ranges: a) 1 - 2mm; b) 2 - 2.8mm; and c) 2.8 - 4mm.
In addition, an examination of the micrographs shows fewer gaps (marked on Figure 4.14)
between the epoxy binder and the EP particles in the foams of type 1 and 2 than foams of
128
type 3. Hence, it is expected that cracks initiate and develop faster under compressive loading
in the foams of type 3 than in the other two types of foam. This will be investigated in detail
in Section 4.7.
4.6.2 Volume fraction of the epoxy in EP/epoxy foams
In the following, the volume fraction of the epoxy binder in EP/epoxy foams is quantified by
employing an estimation method as well as the measured experimental values. However, the
comparison of the results and a discussion of the ability of the estimation method to quantify
this parameter are held over until Section 5.4.
4.6.2.1 Estimating the volume fraction of epoxy in EP/epoxy foams from the foam
density
The volume fraction of epoxy Ve in the foams was calculated theoretically from the rule of
mixtures given in Eq. (4.19):
)/()( PePceV (4.19)
where ρc is the composite density, ρp is the particle density and ρe is the epoxy density. It
should be noted that the particle densities presented in Table 4.2 cannot be used in Eq. (4.19)
since these densities only refer to undeformed particles. During the sample manufacturing
process, the compaction pressure effectively changed the density and density distribution of
the EP particles (see Figure 4.8 (a)). Therefore, a modified version of Eq. (4.19) was used to
estimate the volume fraction of epoxy in the manufactured foams:
PePfe DDDV / (4.20)
129
where Df and Dp are the foam density and particle density as a function of compaction
pressure (σcomp), respectively:
)( compf fD ; )( compp fD
To obtain the particle density as a function of compaction pressure, a group of samples, which
only contained perlite particles was prepared. To this end, EP particles of known mass were
compacted in the same mould used for manufacturing the EP/epoxy foam samples. As the
compaction proceeded, the change in height, and hence density, were calculated as a function
of compaction pressure. Results illustrating the particle density versus compaction pressure
are shown in Figure 4.15 (a).
Using the experimental results and the assumption that there was little or no ingress of epoxy
into the EP particles, the foam density Df and particle density Dp as a function of compaction
pressure (𝜎𝑐) can be expressed by the following equations:
32
2
1 CCCD CCf
(4.21)
32
2
1 CCPD (4.22)
The coefficients Ci and αi for the foam density Df and particle density Dp were determined
using least squares fitting. The coefficients for the foam density and particle density
containing EP particles of three particle size ranges are given in Table 4.7.
130
(a)
(b)
Figure 4.15. (a) Foam density as a function of applied pressure. (b) Particle density as a function of
compaction pressure.
131
Table 4.7. Coefficients in Equations (4.21) and (4.22) for foam density Df and particle density Dp .
Particle
size range 𝐶1 𝐶2 𝐶3 𝑅2 𝛼1 𝛼2 𝛼3 𝑅2
1-2mm -0.0114 0.1268 0.1549 0.9876 -0.0078 0.1188 0.125 0.9908
2-2.8 mm -0.0129 0.1313 0.1456 0.9759 -0.0174 0.1574 0.0874 0.9807
2.8-4 mm -0.0146 0.1405 0.1329 0.9792 -0.0066 0.1065 0.1016 0.9908
Using Eqs. (4.21) and (4.22), the particle density and the corresponding foam density at each
pressure level were calculated, substituted in Eq. (4.20) and subsequently the volume
percentage of epoxy (Ve) was calculated. Figure 4.16 illustrates the volume percentage of
epoxy as a function of foam density for the three types of foam.
Figure 4.16. Volume fraction of epoxy versus density of the foam.
132
The results show that foams of type 3 have the highest volume percentage of epoxy. In
contrast, foams of type 1 show the lowest volume percentage of epoxy versus density. The
volume percentage of epoxy increases with the foam density in the range of 0.15 - 0.4 g/cm3
for foams of type 3 but reaches a maximum earlier at 0.38 g/cm3 for type 1 foams. At higher
densities, though, a gradual decrease in the volume fraction of epoxy is observed, which will
be discussed in Section 5.4.
4.6.2.2 Calculating the volume fraction of epoxy in EP/epoxy foams
The volume fraction of epoxy in each foam sample was calculated from the difference
between the mass of the foam and the mass of the EP particles in that foam. Next, the mass
of the epoxy was divided by the density of the cured epoxy binder (i.e. 1.14 g/cm3) to get the
volume of the epoxy. Dividing the volume of the epoxy by the total sample volume gives the
volume fraction of the epoxy binder in that foam sample. Following this procedure, the
volume fraction of the epoxy binder in the three types of EP/epoxy foams were calculated
and are presented in Figure 4.17. Figure 4.17 shows that the volume percentage of epoxy
within the EP/epoxy foams was between 1.9% and 6% and was mostly in the range 2% - 5%.
Though it seems the volume fraction is almost constant, a decreasing trend in the type 1 and
2 EP/epoxy foams and a polynomial trend in the type 3 foams can be noticed. Overall, the
results presented in Figure 4.17 show that epoxy occupies a very small portion of the
EP/epoxy foams which explains their low density.
133
Figure 4.17. Volume fraction of epoxy binder in EP/epoxy foams of type 2.
4.6.3 Compressive response of EP/epoxy foams
Compressive stress-strain curves for the three types of EP/epoxy foams in the density range
0.15 - 0.48 g/cm3 are illustrated in Figure 4.18. The stress-strain curves show a similar trend,
with the maximum stress occurring in the strain range of 0.03 - 0.07, followed by strain-
softening. The stress-strain curves were used to evaluate the maximum stress, effective
elastic modulus, and modulus of toughness for different foam densities and these are
presented in Figure 4.19. Maximum stress is the peak value in the stress-strain diagram;
effective elastic modulus is the unloading gradient (details explained in Section 3.4.2); and
the modulus of toughness is the area under the stress-strain diagram which indicates the
strain-energy density absorbed by the material before it fractures [266]. The compressive
stress shows a linearly increasing trend with density (Figure 4.19 (a)), while the effective
elastic modulus (Figure 4.19 (b)) and the modulus of toughness (Figure 4.19 (c)) show (𝜌)1
2
and 𝜌2 trends, respectively. In addition, the results show that the compressive stress and the
134
effective elastic modulus of EP/epoxy foams are independent of the particle size range.
However, a small deviation was observed for the modulus of toughness (Figure 4.19 (c)).
Similar behaviour was observed in Perlite/sodium silicate [188] (explained in Section 2.3.3).
Foams of type 1 and 2 show a slightly higher energy absorption capacity. This can be seen
by comparing the stress-strain curves for the three types of foams (Figure 4.18). The stress-
strain curves for foams made with EP particles of 2 - 2.8 mm (Figure 4.18 (b)) show a plateau
for a significant range of strain and thus a higher energy absorption capacity. This can be
ascribed to the morphology of EP particles in the range of 1 - 2 mm and 2 - 2.8 mm which
have a larger edge cell thickness (Table 4.3) and contain internal fibre reinforcement (Figure
4.3). Other factors contributing to the increase in the modulus of toughness are better bonding
of smaller particles with epoxy when used in foams (Figure 4.14) and slightly higher epoxy
volume fractions in the type 1 and 2 foams (Figure 4.17).
(a)
135
(b)
(c)
Figure 4.18. Typical stress-strain curves for the different EP/epoxy foam densities of: (a)
Type 1; (b) Type 2; (c) Type 3.
136
(a)
(b)
137
(c)
Figure 4.19. Properties of manufactured foams of type 1 ( ⃣ ), type 2 (◇) and type 3 (∆): (a)
Maximum stress versus foam density; (b) Effective modulus versus foam density; (c) Modulus of
toughness versus foam density.
Considering the high volume fraction of EP particles in the manufactured foams, the EP’s
properties could have a significant contribution in the resulting foam properties. To measure
this contribution as a function of density, several loading-unloading confined compressive
tests were conducted on packed beds of EP particles as explained in detail in Section 4.3.2.
As discussed earlier, however, the moduli obtained from these tests were not equal to the
Young’s modulus; they were a function of both the Young’s modulus and Poisson’s ratio.
Therefore, to find the stiffness contribution of EP particles to the effective stiffness of
EP/epoxy foams, a series of additional confined tests was conducted on the EP/epoxy foams,
as explained in Section 3.4.2. As the experimental results have already proven that the
138
effective elastic moduli of the EP/epoxy foams are independent of particle size, the confined
tests were only conducted on the foams of type 2. The measured confined modulus of the EP
particles and the EP/epoxy foams as a function of the particle density are presented in Figure
4.20.
Figure 4.20. Confined modulus of EP particles ( ⃣ ) and confined modulus of the foams (◇) as a
function of the density of the packed bed of EP particles.
As can be seen in Figure 4.20, for foams having the same combined mass of EP particles per
unit volume as the packed bed samples of EP particles, the effective elastic modulus is
generally lower. This is due to friction and interaction between the EP particles. The inclusion
of the epoxy as a matrix reduced the friction between EP particles; however, this reduction
could not be quantified. The measured effective elastic moduli of the EP/epoxy foams are
much closer to the elastic moduli of EP particles than to that of epoxy (1002 MPa). This is
clearly indicative of the significant role of EP particles in the stiffness of the EP/epoxy foams
139
and can be ascribed to the high-volume fraction of the EP particles (88 - 94 Vol%) in these
foams (i.e. type 2). Though the stiffness provided by the EP particles in the EP/epoxy foams
cannot be measured using quasi-static mechanical tests, an upper 𝐸𝑃𝑈𝑝𝑝𝑒𝑟 and a lower 𝐸𝑃
𝐿𝑜𝑤𝑒𝑟
bound can be predicted using the Voigt and Reuss models given by Eq. (4.23) and Eq. (4.24).
The Voigt and Reuss models have been shown to set the upper and lower bounds of the elastic
moduli for most particulate micro- and nano-composites [262]. It has been shown that all
experimental and predicted values (using models like Kerner [267], Counto [268], Guth [269]
and Paul [270]) fall between these bounds:
e
eecLower
pV
VEEE
1 (4.23)
ece
eecUpper
pVEE
VEEE
1 (4.24)
where E is the elastic modulus and the subscripts p, e and c stand for particle, epoxy and
composite, respectively. Figure 4.21 shows an upper and a lower bound for the elastic
modulus of EP particles at each foam density. To quantify the contribution of EP particles in
the resulting foam stiffness, the effective elastic modulus of the foams as a function of foam
density is illustrated. Note that the Reuss model assumes a state of uniform stress and the
Voigt model a state of uniform strain. In solid materials, a perfectly elastic 100% dense
system of interacting grains, where each grain is anisotropic and under the Reuss condition,
suffers strain (or displacement) singularities or gaps at the grain boundaries. Likewise, the
Voigt condition would lead to stress singularities in the grain boundaries [271]. With regards
to the results presented in Figure 4.21, two things can be inferred. Firstly, the EP/epoxy foams
show Reuss-like behaviour similar to metals, but are atypical of non-plastic materials. In
140
metals, a Reuss-like state can be maintained because deformation which would lead to
displacement singularities can be accommodated through localised plastic deformation. In a
similar manner, in EP/epoxy foams the accommodation is through local EP cell wall collapse
at the boundaries between particles. The local failure associated with the cell wall collapse
releases the stress concentration without affecting the whole structure of the EP particle,
which on average behaves in a Reuss-like fashion. Secondly, the particles make a significant
contribution to the resulting elastic modulus (or stiffness) of the EP/epoxy foams.
Figure 4.21. Predicted elastic moduli of EP particles using the Voigt and Reuss models. To
quantify the particles’ contributions to the stiffness of the foam, the effective elastic modulus of the
EP/epoxy foams as a function of foam density is presented.
141
Damage analysis
The macroscopic observations taken during the tests coupled with post-test microscopic
observations were used to understand the deformation mechanisms of the EP/epoxy foams
under compressive loading. Figure 4.22 shows a schematic representation of the damage
sequence in the EP/epoxy foam samples, which is irrespective of their density and type.
However, the strain of occurrence was different for each type of foam, as denoted by the
strain range at which each damage step occurred.
Type 1 ε =5.0-6.8% ε =7.9-10.0 % ε =11.0-13.0 % ε =28.0-40.0 %
Type 2 ε =5.6-7.2% ε =8.0-11.0% ε =12.0-13.0% ε =30.0-46.0%
Type3 ε =3.8-4.2% ε =5.3-6.2% ε = 8.9-10.0% ε =27.0-28.0 %
Figure 4.22. Schematic representation of failure in EP/epoxy foams.
As can be seen, damage started and developed faster in the type 3 foam samples. This can be
explained by two factors: interfacial bonding between the particles and the epoxy binder, and
the volume fraction of epoxy. Microscopic observations (explained in Section 4.6.1) showed
a) b) c) d)
142
that there is better interfacial adhesion between the particles and the binder as the EP particles
get smaller in size. Foams of type 2 and 1 have the higher resistance to the formation and
propagation of cracks than foams of type 3. This may be due to the better interfacial adhesion
between the particles and the binder and the higher volume fraction of epoxy in comparison
with the type 3 foams. According to the macroscopic observations, cracks started to develop
in the foam when the stress reaches its maximum, at strains of about 3.8 to 7.0% (Figure 4.22
(a)). Cracks on the wall of the sample developed in the direction of the applied load, while
shear type cracks originated from the upper and lower rims. After reaching the maximum
stress, inelastic deformation occurred as the sides of the sample barrelled outward under
compressive loading. The barrelling effect became more prominent in the samples with larger
particle size. For foams of the same type, the barrelling effect reduced as the density
increased. As the compressive load increased, the cracks on the sample wall joined together
and caused the formation of longitudinal cracks (see Figure 4.22 (b)) due to secondary tensile
stresses. These secondary tensile stresses are normal to the applied load and are generated
due to transverse deformation (Poisson effect) [272]. Longitudinal cracks resulted in
longitudinal splitting along the rim of the sample. At the same time, shear type cracks
propagated through the foam and converged at the central part of the sample where horizontal
cracks were generated due to the pure compression (see Figure 4.22 (c)). This type of shear
failure is common in brittle materials such as ceramics and concrete during compression
testing due to the restraining effect of friction on the upper and lower surfaces. The eventual
joining of the longitudinal splitting and shear cracks in the sample gave rise to the formation
of wedge-like fragments on the sides of the samples (see Figure 4.22 (d)). The development
143
and formation of the wedge-like fragments can be better understood by analysing the uniaxial
compressive state of stress, assuming frictionless compression:
000
000
00
ij (4.25)
where σ is the applied stress. The state of stress can be decomposed into hydrostatic (σhyd)
and deviatoric (σdev) stress components:
300
03
0
003
2
300
03
0
003
devhydij (4.26)
The deviatoric stress components can be further divided into two parts, both representing a
state of pure shear [124]:
300
03
0
000
000
03
20
003
2
(4.27)
As can be seen, the hydrostatic stress component produces pure compression while the
deviatoric component gives rise to shear stresses in the material. The effect of the deviatoric
and hydrostatic components of applied stress on the damage of EP/epoxy foams can be
observed in the macroscopic pictures such as Figure 4.23, taken from a type 1 sample
compressed to a strain of 12%. In these pictures, the propagation of shear cracks from the
144
upper rim and the formation of longitudinal splitting on the wall of the sample can be
observed. Figure 4.23 (b) displays the remnants of the sample, which is typical of all of the
samples after the tests, showing the significant role of shear and compressive stresses on the
failure modes. The shear stress (the deviatoric component) is responsible for the formation
of shear cracks in the foams while the compressive stress (hydrostatic component) gives rise
to the formation of horizontal cracks. Post-test micrographs taken from the failed samples
provide further evidence for the effects of shear and compressive stresses on the failure
modes. Figure 4.24 shows the SEM micrographs taken from the wedge-like side of the
sample. As can be seen, the particles are uncrushed (Figure 4.24 (a)) and cell walls seem to
be fractured along a plane (Figure 4.23 (b)) which could be due to shearing in the sample.
On the other hand, micrographs from the centre of the sample (Figure 4.24 (c)), where
horizontal cracks dominate, show a significant difference. A large proportion of the cell walls
are crushed which is a typical feature of failure under compression.
(a)
145
(b)
Figure 4.23. Macroscopic images showing (a) A sample of type 1 compressed to a strain of 12%
(b) A typical remnant of the samples after the test.
(a)
(b)
146
(c)
Figure 4.24. SEM images showing (a) Uncrushed EP particles in the wedge-like fractured side of a
failed sample; (b) Cell walls of EP perlite particles fractured along a plane; (c) Crushed cells in the
central region of a uniformly compressed sample.
147
Measurement of the elastic moduli of EP/epoxy foams using elastic
wave speed
Elastic properties of the EP/epoxy foams were characterised by means of elastic wave
propagation (compression and shear) along the cylinder axis of the specimen. In addition, a
new series of quasi-static compressive tests were conducted on the EP/epoxy foams in
parallel, as the EP particles previously used (Sections 4.6.1 - 4.6.3) were from a different
batch. Having all of the samples made from the same particle batch ensured that a more
reliable comparison could be made between the properties obtained from the mechanical and
elastic wave tests. Compressive stress-strain curves for the new EP/epoxy foams with
different densities are illustrated in Figure 4.25.
Figure 4.25. Typical stress-strain curve for different EP/epoxy foam densities.
148
Stress-strain curves were used to evaluate Young’s modulus of the foams from the unloading
gradient (explained in detail in Section 3.4.2). The results for Young’s modulus obtained
using the quasi-static tests are illustrated as a function of foam density in Figure 4.27 and the
average values are tabulated in Table 4.8. The Young’s moduli obtained for the new series
of EP/epoxy foams were very close to the ones manufactured from another batch of EP
particles, presented in previous Section 4.6.3. Determination of Poisson’s ratio of the foam
by installation of regular strain gauges was not possible. The strain gauges had a tendency to
debond during the compressive test (un-confined one) due to minimal surface contact
afforded by voids on the surface of the foam. Therefore, the work was extended to
characterise the complete elastic properties of EP/epoxy foams by measuring the velocity of
the passage of elastic waves (i.e. longitudinal and shear waves) in the axial direction for a
wide range of foam densities. The longitudinal and shear wave velocities of EP/epoxy foam
at different foam densities are shown in Figure 4.26. Similar to Figure 4.5 for packed EP
particles, both velocities reached a plateau for densities higher than 0.2 g/cm3. This was
investigated in packed EP particle beds (details explained in Section 4.3.3) and was ascribed
to the relatively constant mass of EP particles which resisted crushing at densities higher than
0.2 g/cm3. In addition, the inter-particle porosity was found to be about 50 Vol%, excluding
debris, and in the range 40 - 50 Vol% including debris for compact densities in the range 0.1
- 0.375 g/cm3. In EP/epoxy foams the inter-particle porosity is filled with 3 - 5 Vol% epoxy,
as shown in Figure 4.28. A very similar behaviour to the wave velocities in the EP particle
compacts (Figure 4.5) is indicative of the dominant role of the EP particles in wave
propagation.
149
Figure 4.26. The longitudinal wave and shear wave velocities versus foam density.
Figure 4.27. Young’s modulus of EP/epoxy foams versus foam density determined from the
elastic wave speed (○) and mechanical tests (∆); and Young’s modulus of the EP particles versus
foam density determined from the elastic wave speed (◇).
150
Table 4.8. Elastic properties of EP particles, epoxy resin and EP/epoxy foams.
Density
(g/cm3)
EP particles EP/epoxy foams
Elastic wave tests1 Mechanical tests2 Elastic wave tests
𝐸 (MPa) 𝜈 𝐸 (MPa) 𝜈 𝐸 (MPa) 𝜈
0.15 35.6 0.277 54.7 - 115.2 0.172
0.20 130.3 0.287 102.2 - 246.7 0.194
0.25 210.3 0.295 145.2 - 359.7 0.219
0.30 275.6 0.300 175.6 - 454.5 0.245
0.35 326.4 0.301 199.7 - 531.6 0.274
0.40 362.5 0.300 223.7 - 591.6 0.305
0.42 372.8 0.299 253.9 - 610.8 0.318
Density Epoxy resin (Diluted with acetone)
(g/cm3) Mechanical tests Elastic wave tests
1.14 𝐸(MPa) 𝜈 𝐸(MPa) 𝜈
1834.7 0.42 2022.4 0.38
1. Elastic properties of the EP particles are given as a function of the corresponding EP/epoxy foam
density.
2. Elastic properties of the EP/Epoxy foams are given as a function of the EP/epoxy foam density
151
Figure 4.28. Volume fraction of the epoxy binder in the second series of EP/epoxy foams versus
foam density.
Isotropic elastic properties of the foams were investigated in terms of two elastic constants E
and ν, given by Eqs. (3.6) and (3.7), and illustrated as a function of foam density in Figures
4.27 and 4.30, respectively. For comparison, the Young’s modulus results obtained from
elastic wave tests are plotted with the ones determined from quasi-static mechanical tests
(Figure 4.27). Both plots follow the same qualitative pattern and show that the Young’s
modulus increases with the foam density. However, the measured Young’s moduli from the
elastic wave speed are more than twice the values obtained by mechanical tests. Williams
and Johnson [273] also reported dynamic moduli (measured using elastic wave tests) were
about twice the static moduli for cancellous cell composites. They ascribed this discrepancy
to a difference in the strain rate. The difference in strain rate can influence the results for a
continuous material by about 10 - 15% [274], which is consistent with the measured values
for epoxy resin (Table 4.6), and for granular material by up to about 40% [275]. Nevertheless,
152
the observed discrepancy (≥ 100%) cannot be attributed only to the strain rate difference.
There are three other possible hypotheses to explain such a discrepancy. The first is the
deformation mechanism of the EP particles during unloading. The foams contain a high-
volume fraction of EP particles, consequently not all of the interstices between the particles
are filled with epoxy and some particles are in direct contact with each other. In this case (in
the absence of epoxy between particles), Hardin [276] explained that both elastic and plastic
deformation occurs during the unloading of particulate materials and strain cannot be
separated into discrete elastic and plastic components. This will in turn adversely affect the
Young’s modulus measured using quasi-static mechanical tests.
The second possible explanation for this discrepancy could be the strain level at which
Young’s modulus was measured by the mechanical and elastic wave tests. It was
hypothesised that loading to 70% of the maximum stress (ISO 13314), which was about 3%
strain, could deteriorate the microstructure of EP/epoxy foams by local failure of EP particle
cell walls which leads to lowering the stiffness of the material. On the contrary, the material
displacements produced by the passage of elastic waves are in the order of a few angstrom
[277], and hence Young’s modulus was measured at very low strains and not affected by
microstructural deterioration. To test this hypothesis, one sample with a density 0.26 g/cm3
was loaded to 70% of the maximum load and while the load was held constant, longitudinal
elastic waves were propagated through the sample and Young’s modulus was measured. The
result, which is designated by a filled black circle in Figure 4.27, not only does not show a
lower Young’s modulus value in comparison with previously measured ones but a slightly
higher value than the average ones. This result invalidates the above hypothesis and shows
153
that loading to 70% of the maximum load does not adversely affect the Young’s modulus
and should even provide slightly higher moduli by compressing the foam to a higher density.
This is also demonstrated by conducting cyclic compressive tests on a sample with a density
of 0.26 g/cm3. The specimen was loaded to 2.5% of the maximum load, and then unloaded
at the same crosshead speed to 2N load. The loading-unloading continued to 70% of the
maximum load with a load increment of 2.5%, as shown in Figure 4.29 (a). It can be seen
that as the stress increases, the gradient of the unloading path gets steeper which is measured
and presented in Figure 4.29 (b). It clearly shows that loading to 70% of the maximum load
does not detoriorate the internal microsturcture of EP/epoxy foam but makes the foam stiffer.
Nevertheless, these results show that ISO 13314 cannot give a unique Young’s modulus for
EP/epoxy foams, but only the particular elastic modulus of the material at 70% of the
maximum load.
The third hypothesis relates to the deformation of the epoxy ligaments, as well as the
deformation of the perlite cell struts. In mechanical testing where the effective Young’s
modulus (EMech) is determined from the unloading gradient between 70% and 20% of the
maximum load, the unloading displacement is proportional to the compliance MechE
1. As
Young’s modulus measured by the elastic waves (EWave) is bigger than the Young’s modulus
measured by mechanical tests EWave > EMech, the mechanical testing case is more compliant
MechWave EE
11 . If the true compressive modulus is taken as
WaveE , then a larger unloading
displacement (mechanically) is observed than was expected from the pure compressive value,
154
which would be proportional to WaveE
1. One simple way for a mechanical system to show a
large elastic displacement is in bending or buckling. The epoxy ligaments are oriented so as
to respond by bending/buckling (n = 1, both ends of a column pinned). However, wave
velocity measurements use small displacements so that direct compressive displacement is
the same as the bending displacement. A similar situation occurs within EP particle struts.
This is supported by the experimental results in Sections 4.3.2 and 4.3.3 where the
constrained modulus of packed beds of EP particles was measured using quasi-static
mechanical and elastic wave tests. That result also showed that for the density range 0.17 -
0.37 g/cm3, the constrained modulus measured using elastic waves was higher than that
measured by the mechanical tests by up to about 25%. Therefore, the combined effects of the
strain rate difference, yielding of EP particles during unloading, and the buckling
deformation in both EP particles and epoxy ligaments may result in a much lower
‘mechanical test’ modulus than the one measured by elastic wave tests.
Notwithstanding, both these properties (i.e. static and dynamic moduli) are useful. For
applications where the material is subjected to compressive load, the stiffness of the material
is characterised by Young’s modulus obtained using quasi-static mechanical tests. On the
other hand, when they are used in a vibrating structure, for example, for sound absorption in
buildings, inside the fuselages of aeroplanes and in machinery enclosures, the true elastic
properties measured by the elastic wave tests are of importance.
155
(a)
(b)
Figure 4.29. (a) stress-strain curves and (b) Young’s modulus measured from the gradient of the
unloading path in Figure 4.29 (a) at different stress level in cyclic compressive tests conducted on a
sample with density of 0.26 g/cm3.
To evaluate the relative contribution of EP particles to the stiffness of EP/epoxy foams,
Young’s modulus of the EP particles measured in Section 4.3.3 were plotted with those of
the EP/epoxy foams in Figure 4.27. For foam densities of 0.15 g/cm3 and 0.2 g/cm3, the EP
156
particles contribute to the stiffness of the EP/epoxy foams by 12% and 49%, respectively.
For higher foam densities, however, this contribution reached a constant value of about 62%.
There are three main factors contributing to the stiffness of EP/epoxy foams: i) epoxy binder,
ii) EP particles, and iii) the interaction between the EP particles and epoxy binder. The role
of the epoxy binder in the stiffness of EP/epoxy foams seems to be almost constant with
respect to the constant range (3 - 5%) of the epoxy volume fraction in EP/epoxy foams, as
shown in Figure 4.28. However, the contribution of EP particles and the interaction between
the EP particles and epoxy binder appear to be variable. This might be ascribed to the number
of contact points and the contact size (or contact surface area) among the EP particles as well
as the contact size between the EP particles and the epoxy binder. As particles are crushed
into smaller particles during foam densification, the number of contact points per unit volume
as well as the contact size will increase among the particles, and also between the EP particles
and the binder. These result in more mechanical bonds between the constituents and hence
an increase in the stiffness of EP/epoxy foams. In low density foams (< 0.2 g/cm3), a greater
portion of large EP particles exist (see Figure 4.7) and thus a smaller number of contact points
and smaller contact size are present in comparison to foams at higher densities. Beyond 0.2
g/cm3, the intra-particle volume becomes constant. This will be discussed in Section 5.4.
The Poisson’s ratio of EP/epoxy foams, corresponding to the wave speed measurements in
Figure 4.26, were calculated using Eq. (3.7) and are presented in Figure 4.30. It can be seen
that Poisson’s ratio of the foams monotonically and almost linearly increases with respect to
the foam density. This is indicative of the material’s increasing resistance to change in
volume rather than to distort under uniaxial load.
157
Figure 4.30. Poisson’s ratio of epoxy resin (), foam (○) and packed beds of EP particles ( ⃣ )
versus the foam density.
To investigate the effect of the constituent materials on the foam Poisson’s ratio, Poisson’s
ratio of EP particles and the cured diluted epoxy resin are also plotted on the same figure.
Poisson’s ratio of the foams begins lowest, intersects with the Poisson’s ratio of EP particles
at 0.38 g/cm3, and approaches the Poisson’s ratio of the epoxy binder with increase in the
foam density. Conversely, Poisson’s ratio of the packed EP particles is almost constant with
increase in the foam density. However, when the particles and epoxy are combined, the
resultant Poisson’s ratio is far from constant. Without epoxy, rotation can occur about the
points of contacts between the particles – the inclusion of the epoxy elastically constrains
these motions. This situation is similar to the assumptions considered in [278] concerning the
elastic behaviour of granular materials as assemblies of randomly distributed particles. Figure
158
5 in [278] shows that Poisson’s ratio increases by decreasing ks/ kn ratio, where ks is the
resistance of a contact between two particles to shear displacement and kn is the resistance
of a contact to compression along a line connecting the centre of the two particles. The
decrease in ks/ kn can be due to: i) increase in kn, ii) decrease in ks, or iii) concurrent increase
in kn and decrease in ks. At this stage, it is suggested that the decrease in ks/ kn is responsible
for the observed increase in Poisson’s ratio with increase in foam density, however it is not
clear which case (i, ii or iii) is active. This will be investigated by numerical simulation in a
future study.
159
5 Chapter Five: Discussion
Introduction
In Chapter 4, there was a very focused discussion of each specific result. In this chapter, the
results from the previous chapters will be discussed in a broader context, which includes: a)
Comparison of the manufacturing methods for EP/epoxy foams with other syntactic foams;
b) Comparison of Young’s moduli of packed EP particle beds found experimentally using
the oedometric tests and elastic wave tests, as well as the ones predicted using the Voigt and
Reuss models; c) Analysis of the method for estimation of the volume fraction of epoxy
binder in the EP/epoxy foams; d) Comparison of the compressive properties of EP/epoxy
foams with the foams discussed in the literature; e) Comparison of the damage modes in
EP/epoxy foams under compressive loading with a variety of existing foams.
160
Manufacturing method
In this study, EP/epoxy foams were manufactured by the buoyancy method (AU Patent No.
2003205443 [139]), as described in Section 3.3.4. Alternative methods for manufacturing
foams were reviewed in Section 2.3.2. In comparison with those methods, the buoyancy
method used here has several advantages. These include avoiding breakage of the fragile EP
particles during mixing with the epoxy prior to pressing. This is essential for the very low
density foams but is also useful for the high density foams. In addition, this method offers
the homogeneous distribution of EP particles in the matrix, the feasibility of manufacturing
foams with density as low as 0.15 g/cm3 and ease of manufacture.
On the other hand, this method has some drawbacks. In this method, the mixture of EP
particles and diluted epoxy binder creates a two-phase system where just the top phase is
used and the bottom phase is disposed of. Even though 2 - 5 vol% of the epoxy binder is
present in the EP/epoxy foams, a large amount of acetone is required to dilute the epoxy
binder (the mixing ratio is explained in Section 3.3.4). This is a large waste of materials as
the residual bottom phase, including a large amount of acetone, is thrown away. The second
problem is the curing time which is significantly longer (40 days) than the competing
methods mentioned in Section 2.3.2 which require from less than 2 days and up to 7 days.
The third problem is compressing EP particles of different mass within a constant volume to
obtain different densities. The compression causes breakage and crushing of EP particles and
thus reduces the properties which are pertinent to the cellular structure of EP particles, such
as the thermal and acoustic insulating properties.
161
Of the alternative methods explained in Section 2.3.2, two methods are considered suitable
for EP/epoxy foams. The first one involves the filling of a mould with a desired quantity of
particles and pouring a pre-measured amount of epoxy over the particles. The other is to fill
a mould completely with particles and then measure their mass. This is followed by mixing
the particles with an epoxy solution and then transferring the mixture to a mould. Both
methods have the advantage of using a small amount of acetone which is contrary to the
buoyancy method. In addition, the first method has the advantages of avoiding particle
fracture and having accurate information regarding the volume fraction of particles and
binder. The second method also has the advantage of a homogeneous distribution of particles
within the foam as well as restricting the particles from floating to the surface during foam
production. These methods, however, have several disadvantages in addition to those
mentioned in Section 2.3.2. The first method has the potential for an inhomogeneous
distribution of particles owing to the separation of phases due to differences in the constituent
densities. Moreover, there is a possibility of binder concentrating in the lower part of the
foam due to gravity, especially as the particles are free to move and do not completely fill
the mould. The second method has a high potential to break particles during mixing with the
epoxy binder before the moulding process, which is especially of concern for manufacturing
of the lowest density foams.
Taking these factors into consideration, the first method, with the constraint of filling the
mould completely, has the greater potential for EP/epoxy manufacture. Here, different foam
densities could be achieved by using different particle sizes, which have different densities
162
in Figure 4.8 (a), within a constant volume. In addition, using a low viscosity epoxy is
suggested rather than diluting standard epoxy with acetone.
163
Young’s modulus of packed EP particles
Considering the high-volume fraction of EP particles in the EP/epoxy foams, the Young’s
modulus of packed beds of EP particles were required to be determined. For this purpose,
oedometer tests (confined quasi-static compressive tests) were conducted on packed EP
particle beds, as detailed in Section 4.3.2. However, since the particles in this test were
confined, the calculated unloading gradient was equal to the constrained modulus of the
packed particle beds and a function of both their Young’s modulus (EP) and Poisson’s ratio
(𝜈), given by Eq. (4.5). As it was not possible to obtain Poisson’s ratio by oedometer tests,
attempts were made to estimate the Young’s modulus of packed EP particle beds using the
Voigt and Reuss models (i.e. Eqs. (4.23) and (4.24)) which give an upper 𝐸𝑝𝑈𝑝𝑝𝑒𝑟
and a lower
𝐸𝑝𝐿𝑜𝑤𝑒𝑟 bound, shown in Figure 4.18. To characterise the complete elastic properties, the
work was extended to investigate the elastic properties of packed EP particle beds by
measuring the velocity of the passage of elastic waves (i.e. compression and shear) in the
axial direction for a wide range of compaction densities.
Compression and shear wave velocity measurements and Eqs. (3.4) and (3.5) were used to
measure the Young’s modulus and Poisson’s ratio of a packed bed of EP particles at each
compact density (see Figure 4.9). Having determined the Poisson’s ratio of the EP particle
beds, the constrained moduli obtained earlier using the oedometer tests were converted to
Young’s modulus, hereafter referred to as the quasi-static Young’s modulus. However, this
Young’s modulus is an estimate as the Poisson’s ratios obtained by the elastic wave tests
would be slightly different from the ones that could be obtained from quasi-static testing;
164
Poisson’s ratio from a quasi-static test is expected to be slightly higher. It has been found that
Young’s modulus is dependent on the strain rate or frequency of the tests [1]. Generally,
Young’s modulus increases with strain rate or frequency of the tests. This will in turn result
in a lower Poisson’s ratio, as shown for the epoxy in Table 4.6, which was attributed to the
difference in the deformation rate of the specimens [2, 3].
For comparison, the estimated Young’s modulus of packed EP particles from the quasi-static
tests were re-plotted with the ones predicted using the Voigt and Reuss models as a function
of the foam density, as shown in Figure 5.1. In addition, Young’s moduli obtained using the
elastic wave tests, in terms of both the experimental compact density and the modified
density, are shown in Figure 5.1. As the density is not involved in the calculation of Young’s
modulus from the mechanical results, the mechanical results cannot be modified to reflect
the correction of density excluding debris.
For the most part, the mechanical test results lie in between the two wave speed determined
moduli which could be considered as upper and lower bounds for constrained moduli of EP
particle compacts. For a compacted density of 0.1 g/cm3, the constrained moduli determined
from mechanical tests is higher than those determined by elastic wave tests. This indicates
that an additional resistance to displacement is present in the mechanical tests, i.e. the
unloading response observed in the previous work was not truly elastic. It has been widely
discussed in the literature that particulate materials yield during unloading, and consequently
the true elastic behaviour is restricted to infinitesimal increments of unloading [4, 5].
Therefore, the measurement of elastic properties, especially in low density (high porosity)
compacts where the size of inter-particle contacts is very small is best performed by elastic
165
wave velocity measurements. In addition, Young’s modulus for a density of 0.1 g/cm3
measured by quasi-static tests do not fall within the bounds of the Voigt and Reuss models
which are a simple test of validity. The predicted values of Young’s modulus for most
particulate micro- and nano-composites [6] are found to fall within the identified upper and
lower bound by the Voigt and Reuss models. The predicted values which fall within these
bounds are considered to be valid. Hence, with regards to the elastic wave Young’s moduli
and Voigt and Reuss estimations, the quasi-static Young’s moduli for the compacts with a
density higher than 0.15 g/cm3 are considered to be valid.
Figure 5.1. Young’s moduli of the packed EP particle beds measured using mechanical tests and
estimated using the Voigt and Reuss models.
Within this range (0.15 - 0.37 g/cm3), the Young’s moduli determined by elastic waves are
higher than those determined by the mechanical tests. As these two types of measurement
impose strains that differ by 3 - 5 orders of magnitude and occur at widely differing strain
rates, they explore different aspects of a material’s behaviour [7, 8] which can influence the
166
results for a continuous material by about 10 - 15% [7], and for a granular material by up to
40% [9]. Nevertheless, the lower and upper bounds determined by the experimental and
modified Young’s moduli using elastic-wave tests appear to provide a reasonable range in
which the Young’s moduli of packed EP particles can be explored.
167
Structural characterisation of EP/epoxy foams
The manufactured EP/epoxy foams had a structure of a packed EP particle bed with the epoxy
resin partially filling the interstices. Therefore, the experimental results on the packed EP
particle beds presented in Section 4.3.3 helped to understand the structure of packed EP
particles and hence the EP/epoxy foams. As mentioned in Section 4.6.3, since the properties
of EP/epoxy foams show almost no dependency on particle size, type 2 foams were chosen
and the rest of the tests, including the elastic wave speed test, were conducted on this type of
foam. Hence, this discussion on the structure of the EP/epoxy foam is based on experiments
conducted on only the type 2 foams, with the assumption that other foam types behave
similarly.
Using the calculated inter-particle porosity shown in Figures 4.8 (b) and (c), the volume
fraction of EP particles in EP/epoxy foams was calculated and is presented in Figure 5.2 (a).
It can be seen that the volume fraction of EP particles increase with the foam density,
reaching a peak at a density of 0.375 g/cm3, and then start to decrease for higher foam
densities. This can be explained by fracturing of more EP particles into smaller particles and
debris, as shown in Figure 4.7, which results in a reduction of the volume of EP particles and
hence their fraction in the EP/epoxy foams at high densities (> 0.375 g/cm 3).
In addition, the percentage of porosity at each foam density was calculated as the sum of the
inter-particle porosity and the intra-particle porosity. The inter-particle porosity at each foam
density was calculated by subtracting the sum of the volume fraction of the EP particles
(Figure 5.2) and the epoxy binder (Figure 4.17) from the total sample volume:
168
sample
epoxyParticlesample
V
VVVPorositycleInterparti
)( (5.1)
Intra-particle porosity was calculated by:
i
perlitesolid
EPi fPsizeparticleoneinsidePorosity *1
(5.2)
n
i
iPPorositycleIntraparti1
(5.3)
where ρEP is the density of an EP particle of a certain size presented in Figure 4.8 (a),
𝜌solid perlite is the density of the pore free perlite given in Table 4.1, and fi is the fraction of
that particle size within a compact. The calculated inter-particle porosity from Eq. (5.1) and
the intra-particle porosity from Eq. (5.3) were added together and the total porosity of the
EP/epoxy foams at each density was calculated and is presented in Figure 5.2 (b).
The volume percentage of epoxy in the EP/epoxy foams was estimated using a method
introduced in Section 4.6.2.1, and the results for three types of EP/epoxy foams were
presented in Figure 4.16. In addition, the epoxy volume fractions in three types of EP/epoxy
foams were measured based on experimental data (explained in detail in Section 4.6.2.2) and
are shown in Figure 4.17. Both the results showed that the volume fraction of epoxy is
influenced by the EP particle size. However, while the estimated values show significant
change in epoxy volume fraction with density, the experimentally measured ones show only
a slight change in epoxy volume fraction with density. Moreover, the experimentally
measured values show that the measured epoxy volume fractions for the type 1 foams were
slightly higher than the type 2 and type 3 foams, and that the measured values for the type 2
foams were slightly higher than the type 3 ones. This is reasonable given the fact that smaller
169
particles have a bigger surface area per unit volume and thus a larger amount of epoxy is
required to cover those surfaces.
(a)
(b)
Figure 5.2. (a) Volume fraction of EP particles and (b) Porosity within inter-particle space (by
consideration of debris) and total porosity in type 2 EP/epoxy foams.
170
The estimated values (Figure 4.16) were reasonable for the average behaviour of type 1 and
2 foams of, however they show a significant deviation with the experimentally measured
values for type 3 foams. This indicates that there is a particle size limitation, and that the
estimation method does not give a reasonable trend for the foams made with EP particles
larger than 2.8 mm.
Considering the epoxy volume fraction, the inter-particle and intra-particle porosity as well
as the investigation of SEM images, it can be concluded that EP/epoxy foams have a structure
which is schematically drawn in Figure 5.3. The porosity inside the EP particles is not drawn
but the figure shows how the inter-particle porosity is distributed. In addition, it shows that
the epoxy resin is drawn to the EP particles by surface tension while keeping the inter-particle
space relatively open. Hence, it can be seen that low density EP/epoxy foams have a similar
interior structure to high density foams, but at a different scale. Moreover, it should be noted
that though the particle volume fraction in different density foams varies in the range 46 -
58%, the resultant foam density is mainly caused by the density of the EP particles. The
higher the foam density, the larger the portion of smaller particles and debris that were shown
to have higher densities (Figure 4.8 (a)). Accordingly, the total foam porosity decreases from
91 Vol% to 79 Vol% while the inter-particle porosity reduces from 50 Vol% to 37 Vol%.
This clearly shows that the intra-particle porosity is constant for foams of different density
( 41 - 42 Vol%). Therefore, while the experimental results on the packed EP particles showed
(Figure 4.8 (b)) that the inter-particle porosity was almost constant excluding debris, the real
compact structures, which included the debris, had a constant intra-particle porosity. Overall,
the porous structure of the EP particles contributed to the porosity of the EP/epoxy foams by
171
as much as about 42% and the rest of the foam porosity is related to the packing of EP
particles and the void spaces between particles.
(a) (b)
Figure 5.3. A schematic representation of the internal structure of (a) low density and (b) high
density EP/epoxy foams.
172
Compressive behaviour and compressive properties
A schematic representation of the compressive stress-strain curves for elastomeric, elastic-
plastic, and brittle single-phase foams (i.e. foams that are manufactured from one type of
material) is shown in Figure 5.4. They show linear elasticity at low stress levels, followed by
a long stress plateau regime truncated by a regime of densification where the stress rises
steeply. The gradient of this regime (the initial slope of the stress-strain curve) is considered
as Young’s modulus. The linear elastic region is followed by a plateau region which
characterises the energy absorption capacity of a foam through the deformation of cells. The
final regime of densification is associated with the touching of opposite collapsed cell walls
and if compression continues to a higher strain (called densification) the stress-strain curve
should reach the elastic response of the solid material from which the cell walls are made.
(a) (b) (c)
Figure 5.4. Schematic representation of compressive stress-strain curves for a) an elastomeric, b) an
elastic-plastic, c) a brittle foam [11].
173
As can be seen, no drop after the peak stress has usually been observed in the stress-strain
curves for elastomeric foams (Figure 5.4 (a)), while a small drop in elastic-plastic foams
(Figure 5.4 (b)) and a significant drop in the brittle foams or foams with little plasticity
(Figure 5.4 (c)) has been observed. It has also been found that increasing the foam density
increases the Young’s modulus and plateau stress, but shortens the strain at which
densification occurs [10, 11].
Similar to the single-phase foams explained above, the stress-strain curves for syntactic
foams start with an approximately linear region corresponding to the elastic behaviour of the
foam until the peak stress is reached. Peak stress designates the point of crack initiation. In
some stress-strain curves, the peak stress is followed by a sharp stress drop while in others
there is only a slight decrease in the stress level. The sharper stress drop suggests that it is
more difficult for cracks to initiate in the matrix which is more prominent as the matrix
volume fraction and the strain rate increases [12]. On the contrary, increasing the volume
fraction of hollow microspheres decreases the stress drop after the peak stress [13]. These
imply that the failure initiation in syntactic foams does not depend on the strength of the
hollow microspheres, but is mainly related to the properties of the polymeric matrix. After
this decrease, a plateau region starts at which the stress becomes nearly constant for further
compression. The plateau region corresponds to the energy absorption in the process of the
collapse of cells comprising a microsphere and its surrounding epoxy resin. Syntactic foams
with higher volume fractions of hollow microspheres show a larger plateau region. When a
significant portion of the microspheres are crushed, further compression results in the
densification of the foam which is visible as an upward trend in the stress-strain curve
174
following the plateau region. This point is considered the failure point for syntactic foams
because at this point most of the load bearing microspheres have been crushed. It is worth
noting that these three distinct regions are found to be influenced by the strain rate of the test
and the geometry of the samples. As the strain rate increases, the plateau stress level and
energy dissipation capacity of the foam increases [14, 15]. Gupta [12, 16] investigated the
effects of the specimen aspect ratio on the fracture behaviour and compressive properties. It
was found that the peak compressive strength depends on the mechanical properties of the
hollow microspheres and matrix resin, and is independent of the specimen aspect ratio.
However, the specimen behaviour during compression shows remarkable differences with
respect to the aspect ratio. Specimens with low aspect ratio show no drop after the peak stress
and show a shorter plateau region, while specimens with high aspect ratio either show a small
decrease after peak stress and a larger horizontal plateau region [16] or a large drop and no
plateau regime [12].
A schematic representation of the stress-strain curves in the current study is presented in
Figure 5.5. The damage mechanisms causing these features will be discussed in Section 5.6.
Similar to the foams explained above, there is a linear elastic region ending in a peak stress.
The peak stress is followed by a significant drop in stress i.e. strain softening. Similar
behaviour was noticed in the compressive stress–strain curves of expanded perlite/starch
foams [17], cenosphere/epoxy syntactic foams [12] and pumice/epoxy composites [18].
Following the strain softening, the stress becomes nearly constant for further compression in
the plateau region. Contrary to the single-phase foams and syntactic foams discussed
previously, there is no densification regime in the stress-strain curves of EP/epoxy foams.
175
The point at the end of the plateau where the stress started to decrease continuously is the
failure point (designated in Figure 5.5) for EP/epoxy foams. At this point, the major portion
of the structure has collapsed and the remnants are incapable of bearing substantial amounts
of load. The compressive stress-strain curves (see Figure 4.18) also show that an increase in
a foam’s density increases the peak stress, Young’s modulus, the energy absorption capacity
and hence the modulus of toughness of the EP/epoxy foam. This behaviour is similar to the
single phase foams and syntactic foams discussed above.
Figure 5.5. Schematic representation of the compressive stress-strain curve for EP/epoxy foams.
To compare the compressive properties of the newly manufactured perlite-based foams with
currently available foams, Ashby diagrams of the average compressive strength and modulus
of the foams as a function of density are presented in Figure 5.6. The perlite-based foams are
illustrated by small red ellipses which set them apart from the other foams. A line with a
slope of one is drawn in both Figures 5.6 (a) and (b). All materials lying on lines with this
176
slope have the same specific strength (𝜎𝐶
𝜌) or specific modulus (
𝐸𝐶
𝜌), respectively. Specific
properties provide a better comparison of different foams having different densities. As
shown in Figure 5.6 (a), the specific compressive stress of the newly developed foams in the
density range 300 - 440 kg/m3 (0.300 - 0.440 g/cm3) is comparable with that of alumina
foam (0.745 g/cm3), aluminium-silicon carbide foam (0.27 g/cm3), closed cell Phenolic foam
(0.035 g/cm3), and closed cell polypropylene (PP) foam (0.030 g/cm3). The specific modulus,
however, is lower than alumina foam and aluminium-silicon carbide foam and higher than
the other two above-mentioned foams. In addition, the specific compressive stress of the
EP/epoxy foams in the density range of 300 - 440 kg/m3 (0.300 - 0.440 g/cm3) is comparable
with EP/starch foam (0.3 g/cm3 and 0.375 g/cm3) [17] and EP/sodium silicate foam
containing 0.20 g/ml sodium silicate (0.3 g/cm3 and 0.4 g/cm3) [19] while both foams have
lower specific compressive modulus than that of EP/epoxy foams. Additionally, Figure 5.6
(b) does shows that the specific compressive modulus of perlite-based foams in the density
range of 155 - 440 kg/m3 (0.155 - 0.440 g/cm3) is comparable with that of rigid closed cell
polyurethane foam (0.24 - 0.6 g/cm3), closed cell polyethylene terephthalate (PET) foam
(0.108 - 0.15 g/cm3), closed cell Phenolic foam (0.12 g/cm3), hollow glass microsphere/epoxy
syntactic foams (0.139 and 0.27 g/cm3), and closed cell polystyrene (PS) foam (0.05 g/cm3).
Therefore, the newly developed perlite-based foams may be adaptable for applications where
similar or better properties can be achieved using this economically advantageous material.
177
(a)
(b)
Figure 5.6. Compressive strength (a) and compressive modulus (b) plotted against density for
currently available foams (Ashby et al., 2000) and results obtained for perlite-based foams ( ).
178
Damage mechanisms under compressive loading
Post-test microscopy observations, coupled with macroscopic observations taken during the
tests, revealed the presence of three different failure modes (i.e. longitudinal splitting, shear
failure, and compression failure) for all of the EP/epoxy foams regardless of their particle
size and density. However, the strain to activate each mode was different for each foam type.
In addition, these observations helped to understand the damage sequence and its correlation
with the peak stress, strain softening, plateau region and failure point of the stress-strain
curves for EP/epoxy foams, shown in Figure 5.5. The peak stress is the point at which cracks
started to appear in the specimens, due to shear stresses in the form of shear type cracks
originating from the sample corners, by secondary tensile stresses in the form of longitudinal
cracks, and by compressive stresses in the form of horizontal cracks in the sample centre (as
explained in Section 4.7). The strain softening might be due to the combined effects of the
very low density of the EP/epoxy foams, the small EP cell wall thickness (0.512 - 0.532 µm),
the brittle nature of the EP particles, and the brittleness and low volume fraction of the epoxy
matrix. Moreover, the development and propagation of longitudinal cracks in the samples
causes disintegration of the sample structure which was even more prominent in samples
with lower density due to easier lateral deformation under secondary tensile stresses. These
together make the foam incapable of bearing a substantial amount of compressive load as the
strain increases. The plateau region corresponds to where: i) shear cracks propagated and
meet horizontal cracks in the sample centre; ii) longitudinal cracks meet and form
longitudinal splitting; and iii) the remaining part of the sample, which is in the shape of an
179
hourglass, underwent uniform compressive deformation. The failure point which led to the
end of the plateau region corresponds to the formation of wedge-like fragments by the
intersection of the longitudinal splitting and shear cracks in the sample and the completion
of the collapse of the EP particles along with the surrounding resin matrix.
It is worth noting that similar modes of failure were observed in cenosphere/epoxy [12] as
well as in hollow glass microspheres/epoxy syntactic foams under compressive load [16, 20].
On the other hand, in perlite/A356 syntactic foams with densities lower than 1.06 g/cm3
similar modes of failure to the EP/epoxy foams were observed [21]. At higher densities,
however, cells deformed layer by layer (e.g. by buckling or bending of the cell walls), in a
uniform manner and no shear or wedge-like failure modes were observed [22]. This was
explained by the brittle behaviour of the cell walls in foams with densities lower than 1.06
g/cm3. In higher density foams, strain hardening occurred in deformed cell walls which
resulted in the successful transfer of stress from the distorted layer of cells to the adjacent
areas and thus uniform growth of damage through the whole sample block. Shear-like failure
has also been commonly observed in metal foams showing ductile behaviour [23]. For
example, in cenosphere/aluminium syntactic foams [24] during compression, deformation
was observed to initiate from the upper and lower corners of the sample and developed along
the plane of shear at an angle of approximately 45° towards the centre of the sample where
compression failure was dominant due to the concentration of the compressive stress in this
region. The sheared zone grew layer by layer during deformation and the maximum
deformation was observed to occur within the sheared zones and in the centre of the samples.
180
However, the fully developed wedge-like failure observed in the EP/epoxy foams was absent
[22].
In metal foams with little plasticity, e.g. Zn−22Al alloy foam [25], the deformation
mechanisms mainly consist of brittle crushing and crumbling of cell walls, cell wall tearing
and large shear fracture. In all these foams after an initial linear elastic region, the stress-strain
curve is characterised by a significant drop in load resistance at the onset of plastic
deformation and jagged, oscillatory plateau behaviour associated with the progressive
crushing of brittle cell walls (see schematic representation in Figure 5.4 (c)). This is the
opposite of what has been observed in ductile metal foams which show little or no drop in
load at the beginning of plasticity and smooth plateau behaviour before densification [11].
Polymeric viscoelastic foams have been shown to behave differently [11, 26-29]. These may
be further categorised as elastomeric, elastic-plastic or brittle foams. In these materials, linear
elastic deformation is controlled by cell struts bending and if they contain closed cells, by
cell wall stretching. Under compressive loading, the elastic deformation is followed by the
collapse of the cells by elastic buckling in elastomeric foams (e.g. EPDM foam [30]), plastic
buckling of cell walls in elastic-plastic foams (e.g. rigid polyvinyl chloride foam [31]), and
brittle crushing of the cell walls and formation of shear fractures in brittle foams (e.g. rigid
polyurethane [32] and epoxy-based polymeric foams [33]). Deformation mechanisms
associated with elastomeric and elastic-plastic foams, particularly the elastic buckling
deformation of cell walls and the formation of plastic hinges, was not observed in the
EP/epoxy foams. However, the crushing of cells in the EP/Epoxy foams is similar to what
has been observed in other brittle foams.
181
The role of the different structural components in the failure is also of interest. X-ray
microtomography was recently used in conjunction with finite element modelling in order to
study the progressive collapse of glass hollow sphere reinforced epoxy samples under
compressive loads [34]. In that study, it was clearly demonstrated that the glass spheres near
the centre of the sample are the first to collapse and transfer load to the surrounding epoxy
and spheres. Following this, some sphere collapse along 45 degree planes, analogous to our
shear planes, was also evident. It is possible that the perlite particles within the samples in
the current study underwent a similar collapse sequence, commencing in the central region
leading to the shear cracks and splitting, essentially the reverse of the sequence described in
Section 4.7. However, as the core of the sample is not visible in our work, we are unable to
clarify this at present. In contemplating the damage sequence, it is necessary to note some
fundamental differences between the tomographic study and the work reported here. The
volume fraction of glass cenospheres used was quite low (0.3), the cenospheres had quite
thick walls (15 microns) compared with perlite (0.32 microns), and perlite is a multi-cellular
particle. Therefore, perlite particles under load begin to crush far more easily than
cenospheres, however as multi-cellular particles, they can exist in a variety of intermediate
states between uncrushed and crushed i.e. they can crush progressively, cell by cell.
These comparisons serve to highlight that failure by a combination of shear, compression
and longitudinal splitting observed in this work is characteristic of porous structures
consisting of brittle cell walls, although these may be represented to different degrees in
different systems.
182
6 Chapter Six: Summary
Conclusions
This study investigated the potential of expanded perlite particles (EP particles) in the
manufacture of light-weight syntactic foams. The structural, microstructural and physical
properties of EP particles were characterised. Two geometrical relationships were suggested
for relating the micro-structure of EP particles to their macroscopic properties (Section 4.3.1).
In section 4.3.2, the elastic modulus of packed EP particles beds were measured using quasi-
static mechanical tests. However, as the test was conducted in a confined situation, the
calculated unloading gradient was not identical to the Young’s modulus of the packed bed of
particles but a function of both their Young’s modulus and Poisson’s ratio. Hence in section
4.3.3, the elastic properties of packed EP particles beds, in terms of two isotropic elastic
moduli (Young’s modulus and Poisson’s ratio), were characterised using elastic wave speed
measurements along the axial direction. Particle distributions within EP particle compacts
after compaction showed that as the compact density increases, the percentage of larger
particles decreases due to breakage into smaller particles accompanied by the production of
debris. Despite this process, it was found that the inter-particle porosity remains relatively
constant with increasing compact density. The mass % of debris rises to quite high levels
owing to the high mass density of the debris particles, however, the volume of debris never
exceeded 20% of the available inter EP-particle space. Due to the formation of debris, the
183
experimental compact densities were modified by deducting the debris mass from the total
mass of EP particles and new densities were calculated. Young’s moduli calculated based on
the experimental and modified density values showed a decrease with porosity. However,
Poisson’s ratio was relatively independent of porosity.
The properties of solid perlite were investigated by sintering powdered perlite into low
porosity solid samples and the application of elasticity theories (Section 4.2). The elastic
properties of solid perlite were found to be very close to those for solid obsidian reported by
Manghnani et al. [20]. Using these properties of solid perlite, the porosity-elastic moduli
relations were investigated in Section 4.4. Four analytical models predicting the elastic
moduli of packed beds of EP particles from the properties of the parent material were
investigated. Although these models were not developed to explicitly handle debris, the
Wang, Rice, and Gibson and Ashby models showed reasonable agreement with the
experimental moduli. However, none of the original models predicted the modified moduli
well although the Wang model and the Gibson and Ashby models gave a reasonable average
trend. It was shown that modifying the Phani model shape factor and expressing it as a
function of porosity can provide a satisfactory means of predicting the elastic properties of
EP particle compacts based on both the experimental and the modified densities.
The ultimate purpose of the EP particles was for the formation of syntactic foams. Thus, the
research was extended in Section 4.6 by producing light-weight foams synthesised by
dispersing EP particles in a matrix of epoxy. Three types of foams containing different sized
particles were fabricated for a density range of 0.15 - 0.45 g/cm3. Quasi-static compressive
tests were conducted on the EP/epoxy foams and the effects of particle size and foam density
184
variation on the compressive properties were investigated. The compressive stress and
effective elastic modulus were shown to be independent of the particle size. However, a slight
variation was observed in modulus of toughness as foams made with EP particles in the size
range 2 - 2.8 mm showed a slightly higher energy absorption capacity. On the contrary,
EP/epoxy foams showed a strong dependence of the compressive properties on the foam
density. The strength increased linearly, peaking at 1.77 MPa, whereas the effective elastic
modulus and modulus of toughness increased at parabolically increasing and decreasing
rates, respectively. In Section 4.7, post-test SEM observations coupled with photogrammetry
during the tests revealed the presence of three different failure modes for all of the foams,
regardless of their particle size and density. However, the strain to activate each mode was
different for each foam type.
The specific compressive stress of perlite–epoxy foams in the density range of 0.3 - 0.44
g/cm3 was found to be comparable with that of foams such as alumina (0.745 g/cm3),
aluminium–silicon carbide (0.27 g/cm3), closed cell phenolic foam (0.035 g/cm3), and closed
cell PP foam (0.030 g/cm3). The specific compressive modulus, however, was found to be
lower than alumina foam and aluminium–silicon carbide foam but higher than closed cell
phenolic foam and closed cell PP foam. In addition, the specific compressive stress of the
EP/epoxy foams in the density range 0.300 - 0.440 g/cm3 is comparable with EP/starch foam
(0.3 g/cm3 and 0.375 g/cm3) [186] and EP/sodium silicate foam containing 0.20 g/ml sodium
silicate (0.3 g/cm3 and 0.4 g/cm3) [188] while both foams have lower specific compressive
modulus than that of EP/epoxy foams. Nevertheless, the specific compressive modulus of
perlite–epoxy foams in the density range 0.155 - 0.44 g/cm3 was found to be comparable with
185
those of a number of rigid closed cell foams made from polyurethane (0.4 g/cm3), PET (0.15
g/cm3), phenolic (0.12 g/cm3) and PS (0.05 g/cm3) materials.
In the additional work presented in Section 4.8, the elastic properties of EP/epoxy foams
were characterised by means of elastic wave propagation (compression and shear) along the
the cylinder axis of the specimens. By adopting an isotropic model, the Young’s modulus
and Poisson’s ratio were used to characterise the elastic response of the medium. The young’s
moduli obtained using quasi-static compressive tests were compared with those obtained
using elastic wave tests. Both followed the same qualitative pattern, however the Young’s
moduli measured using elastic waves were more than twice the values obtained by
mechanical tests. This discrepancy was explained by the combined effects of the strain rate
difference, strain level, yielding of EP particles during unloading, and buckling deformation
in both EP particles and the epoxy ligaments. Poisson’s ratio showed an increasing trend with
the foam density and appeared to be influenced by the increase in contact surface area
between the particles and the matrix as the foam density increased.
In summary, all of the objectives of the current research listed in Section 2.5, have been
addressed:
vi. EP particles were introduced as a cost-efficient porous filler to produce light-weight
syntactic foams.
vii. Structural, microstructural, physical, and mechanical properties of EP particles were
characterised.
viii. Three models suitable for predicting the elastic properties of packed beds of EP
particles were introduced. Two of which (i.e. Wang, and Gibson and Ashby) were
186
considered to give an average reasonable trend and the other one (i.e. Phani) with our
modified shape factor was introduced to give the best agreement with the data
generated in this study.
ix. An economical light-weight EP/epoxy foam was manufactured.
x. Microstructural, structural, and mechanical properties of EP/epoxy foams were
characterised using the optical and scanning electron microscopy, quasi-static
compressive tests and elastic wave speed tests.
xi. The effects of the EP particle sizes and foam density on mechanical properties and
behaviour of the foam was investigated.
xii. Damage mechanisms occurring during the tests and behaviour of EP/epoxy foams
under compressive loads were investigated.
Future Research
The mechanical properties (e.g. compressive strength and stiffness) of EP/epoxy foams will
be improved by improving the bonding between the EP particles and the epoxy binder. Since
inorganic fillers such as EP particles generally have a poor affinity with organic resins, they
cannot be chemically bonded and the bonding between the EP particles and the resin is
through the mechanical retention in their irregular pores. To improve the adhesion, the EP
particles could be initially coated with organic compounds and then embedded in the resin
matrix. Organic compounds which have three or more ethylenically unsaturated groups, such
as acrylates or methacrylates of polyhydric alcohols, have been found to be very effective in
bonding inorganic fillers to organic resins [150]. Since not all of the ethylenically unsaturated
187
groups in the organic compounds are consumed during the coating step of the EP particles,
the coated EP particles would be very reactive and could be chemically bonded to organic
resins. As a result of concurrent mechanical and chemical bonds between the EP particles
and the epoxy resin, significant improvement in the compressive strength, compressive
stiffness and bending strength of the EP/epoxy foams is expected. The mixing ratio of the
organic compounds to the EP particles varies in the range 20:80 - 80:20 and the optimum
value should be found by experimentation. To this end, the coated EP particles need to be
embedded in the matrix of epoxy. However, a different manufacturing method rather than
the buoyancy method used in the current study, is recommended. Due to problems associated
with the buoyancy method such as an increased curing time, and the breakage and crushing
of EP particles at the compression step, it would be well worth investigating other methods.
One promising method would be the filling of a mould completely with the closest particle
packing possible and then pouring a pre-measured amount of resin solution over the particles.
However, instead of diluting the epoxy with acetone which degrades the mechanical
properties and greatly increases the curing time (see Section 4.5), a low viscosity resin such
as CRACKBOND® SLV-302 can be used. This epoxy resin has a viscosity of 195 cP at 24°C
and a curing time of 7 days. This manufacturing method would have the advantage of a
quicker curing time and a significant reduction in the breakage and crushing of EP particles.
In addition, it would be easy to calculate the volume fraction of the epoxy and EP particles
which was a challenge in the current study. Such EP/epoxy foams should be fabricated with
different particle size ranges (e.g. 0.710 – 1 mm, 1 - 1.18 mm, 1.8 - 1.40 mm and 1.40 - 2.0
mm) to cover a density range of 0.15 - 0.45 g/cm3. The structural, microstructural, mechanical
188
and thermal properties of the foams would be investigated, and the optimal mixing ratio of
the organic compounds to EP particles, in terms of the resultant properties, would be
determined. Moreover, as one of the main applications of syntactic foams is to use as the core
material in sandwich plates and shells, additional work to attach a stiff skin to the top and
bottom of the core should be undertaken. The skin itself could be a composite of one of three
types: i) glass fibre/epoxy, ii) carbon fibre/epoxy, and iii) hybrid glass/carbon/epoxy. The
prepared panels would then be subjected to flexural and fatigue tests, and the results would
be analysed and the best sandwich panels determined.
189
References
1. Gibson, L.J., Modelling the mechanical behavior of cellular materials. Materials Science and
Engineering: A, 1989. 110: p. 1-36.
2. Lee, S.M., Pocessing and Preparation of Syntactic foams, in Handbook of composite
reinforcements. 1992, VCH Publishers and VCH vch verlagsgesellschaft mbH. p. 257.
3. Gong, L., Kyriakides, S., and Jang, W.Y., Compressive response of open-cell foams. Part I:
Morphology and elastic properties. International Journal of Solids and Structures, 2005.
42(5–6): p. 1355-1379.
4. Gupta, N., Woldesenbet, E., hore, K., and Sankaran, S., Response of Syntactic Foam Core
Sandwich Structured Composites to Three-Point Bending. Journal of Sandwich Structures &
Materials, 2002. 4(3): p. 249-272.
5. Gupta, N., Karthikeyan, C.S., Sankaran, S., and Kishore, Correlation of Processing
Methodology to the Physical and Mechanical Properties of Syntactic Foams With and
Without Fibers. Materials Characterization, 1999. 43(4): p. 271-277.
6. Bardella, L. and Genna, F., Elastic design of syntactic foamed sandwiches obtained by filling
of three-dimensional sandwich-fabric panels. International Journal of Solids and Structures,
2001. 38(2): p. 307-333.
7. Peroni, L., Scapin, M., Avalle, M., Weise, J., and Lehmhus, D., Dynamic mechanical
behavior of syntactic iron foams with glass microspheres. Materials Science and
Engineering: A, 2012. 552: p. 364-375.
8. Shutov, F.A., Syntactic polymer foams, in Chromatography/Foams/Copolymers. 1986,
Springer Berlin Heidelberg: Berlin, Heidelberg. p. 63-123.
9. Ghosh, D., Wiest, A., and Conner, R.D., Uniaxial quasistatic and dynamic compressive
response of foams made from hollow glass microspheres. Journal of the European Ceramic
Society, 2016. 36(3): p. 781-789.
10. Alkan, M. and Doğan, M., Adsorption of Copper(II) onto Perlite. Journal of Colloid and
Interface Science, 2001. 243(2): p. 280-291.
11. Taherishargh, M., Belova, I.V., Murch, G.E., and Fiedler, T., Low-density expanded perlite–
aluminium syntactic foam. Materials Science and Engineering: A, 2014. 604(0): p. 127-134.
12. Ciullo, P.A., Industrial minerals and their uses: a handbook and formulary. 1996, Westwood,
N.J. : Noyes Publications.
13. Broxtermann, S., Taherishargh, M., Belova, I.V., Murch, G.E., and Fiedler, T., On the
compressive behaviour of high porosity expanded Perlite-Metal Syntactic Foam (P-MSF).
Journal of Alloys and Compounds, 2017. 691: p. 690-697.
14. Allameh-Haery, H., Kisi, E., and Fiedler, T., Novel cellular perlite–epoxy foams: Effect of
density on mechanical properties. Journal of Cellular Plastics, 2016.
15. Allameh-Haery, H., Wensrich, C.M., Fiedler, T., and Kisi, E., Novel Cellular perlite-epoxy
foams: effects of particle size. Journal of Cellular Plastics, 2016.
16. Allameh-Haery, H., Kisi, E., Pineda, J., Suwal, L.P., and Fiedler, T., Elastic properties of
green expanded perlite particle compacts. Powder Technology, 2017. 310: p. 329-342.
17. Moavenzadeh, F., ed. Concise Encyclopedia of Building and Construction Materials.
Advances in Materials Science and Engineering. 1990, MIT Press: Cambridge, MA. 698.
18. Bush, A.L., Lightweight Aggregate, in United states mineral resources, D.A. Brobst and
W.P. Pratt, Editors. 1973, United States government printing office: Washington, WA. p.
333-355.
190
19. Otis, L.M., Perlite, in Mineral facts and problems, United States Bureau of Mines, Editor.
1960, United states Government Printing Office: Washington, WA. p. 581-587.
20. Breese, R.O.Y. and Barker, J.M. Perlite. in Industrial Minerals and Rocks. 1994. Society for
Mining, Metallurgy and Exploration, Littleton, CO, United States.
21. Bates, R.L., Geology of the Industrial Rocks and Minerals. 1960, New York, NY: Harper &
Brothers.
22. Doğan, M. and Alkan, M., Removal of methyl violet from aqueous solution by perlite. Journal
of Colloid and Interface Science, 2003. 267(1): p. 32-41.
23. Topçu, İ.B. and Işıkdağ, B., Manufacture of high heat conductivity resistant clay bricks
containing perlite. Building and Environment, 2007. 42(10): p. 3540-3546.
24. Sarı, A., Tuzen, M., Cıtak, D., and Soylak, M., Adsorption characteristics of Cu(II) and
Pb(II) onto expanded perlite from aqueous solution. Journal of Hazardous Materials, 2007.
148(1–2): p. 387-394.
25. Alkan, M., Karadaş, M., Doğan, M., and Demirbaş, Ö., Adsorption of CTAB onto perlite
samples from aqueous solutions. Journal of Colloid and Interface Science, 2005. 291(2): p.
309-318.
26. Singh, M. and Garg, M., Perlite-based building materials — a review of current applications.
Construction and Building Materials, 1991. 5(2): p. 75-81.
27. Barker, J.M. and Santini, K., Industrial Minerals and Rocks - Commodities, Markets, and
Uses J.E. Kogel, N.C. Trivedi, J.M. Barker, and S.T. Krukowski, Editors. 2006, Society for
Mining, Metallurgy, and Exploration (SME). p. 685-702.
28. Johnstone, S.J. and Johnstone, M.G., Minerals for the chemical and allied industries 2nd ed.
1961, London, UK: Chapman and Hall
29. Gunning , D.F. and Ltd., M.A., Perlite Market Study for British Columbia. 1994. p. 100.
30. Jackson, F.L., Processing perlite for use in insulation applications. Transactions of the
American Institute of Mining, Metallurgical, and Petroleum Engineers, Society, 1986. 280(pt
A): p. 40-45.
31. Saunders, E., Agglomerating particulate perlite. 1979: US.
32. Kadey, F.L., Jr. , Perlite, in Industrial Minerals and Rocks, S.J. Lefond, Editor. 1983,
American Institute of Mining, Metallurgical, and Petroleum Engineers. p. 997-1015.
33. King, E.G., Todd, S.S., and Kelley, K.K., Perlite: Thermal data and energy required for
expansion. 1948, United States Bureau of Mines: Washington, DC, United States. p. 15.
34. Pettifer, L., Perlite; diversification the key to overall expansion. Industrial Minerals
(London), 1981. 171: p. 55-75.
35. Wang, B., Smith, T.R.U.S., Masters, A.L.U.S., and Danvers, N.J.K., Micronized perlite filler
product, C. Advanced Minerals, Editor. 2009: US.
36. AUSTIN, G.S. and BARKER, J.M., Commercial perlite deposits of New Mexico and North
America. Mack, GH. Austin, GS, & Barker, JM (eds) New Mexico Geological Society,
Guidebook, 1998. 49: p. 271-277.
37. Roulia, M., Chassapis, K., Kapoutsis, J.A., Kamitsos, E.I., and Savvidis, T., Influence of
thermal treatment on the water release and the glassy structure of perlite. Journal of
Materials Science, 2006. 41(18): p. 5870-5881.
38. Doğan, M. and Alkan, M., Some physicochemical properties of perlite as an adsorbent.
Fresenius Environmental Bulletin, 2004. 13(36): p. 251-257.
39. Cheremisinoff, N.P., Handbook of Water and Wastewater Treatment Technologies. Elsevier.
106-120.
40. Afshari, H., Ashraf, S., Ghaffar Ebadi, A., Jalali, S., Abbaspour, H., Sam Daliri, M., Toudar,
S.R., and Study of the effects irrigation water salinity and pH on production and relative
191
absorption of some elements nutrient by the tomato plant. American Journal of Applied
Sciences, 2011. 8(8): p. 766-772.
41. Burriesci, N., Arcoraci, C., Antonucci, P., and Polizzotti, G., Physico-chemical
characterization of perlite of various origins. Materials Letters, 1985. 3(3): p. 103-110.
42. Sersale, R., Burriesci, N., Pino, L., and Bart, J.C.J., Iron distribution in some Italian tuffs.
Materials Letters, 1984. 3(1–2): p. 51-57.
43. Koukouzas, N., Rare earth elements in volcanic glass: A case study from Trachilas perlite
deposit, Greece. Chemie Der Erde, 1997. 57: p. 351-362.
44. Yanev, Y. and Yordanov, Y., REE, Th, U, Hf, Sc, Co and Ta in the cesium-bearing perlites,
adularites and zeolitites from the Eastern Rhodopes. Comptes rendus de l'Académie bulgare
des Sciences, 1991. 44(4): p. 71-74.
45. Zielinski, R., Lipman, P., and Millard, H., Minor-element abundances in obsidian, perlite,
and felsite of calc-alkalic rhyolites. American Mineralogist, 1977. 62(5-6): p. 426-437.
46. Stolper, E., Water in silicate glasses: an infrared spectroscopic study. Contributions to
Mineralogy and Petrology, 1982. 81(1): p. 1-17.
47. Saisuttichai, D. and Manning, D.A., Geochemical characteristics and expansion properties
of a highly potassic perlitic rhyolite from Lopburi, Thailand. Resource geology, 2007. 57(3):
p. 301-312.
48. Herskovitch, D. and Lin, I.J., Upgrading of raw perlite by a dry magnetic technique.
Magnetic and Electrical Separation, 1996. 7(3): p. 145-161.
49. Scott, P.W. and Bristow, C.M., eds. Industrial Minerals and Extractive Industry Geology.
2002, The Geological Society Bath, Uk.
50. Barth-Wirsching, U., Höller, H., Klammer, D., and Konrad, B., Synthetic zeolites formed
from expanded perlite: Type, formation conditions and properties. Mineralogy and
Petrology, 1993. 48(2-4): p. 275-294.
51. Chakir, A., Bessiere, J., Kacemi, K.E.L., and Marouf, B.b., A comparative study of the
removal of trivalent chromium from aqueous solutions by bentonite and expanded perlite.
Journal of Hazardous Materials, 2002. 95(1–2): p. 29-46.
52. Fornasiero, R.B., Phytotoxic effects of fluorides. Plant Science, 2001. 161(5): p. 979-985.
53. Kendall, T., No sign of the bubble bursting- Perlite uses & Markets. Industrial Minerals,
2000(393): p. 51-53.
54. Johnstone, S.J. and Johnstone, M.G., Minerals for the chemical and allied industries. 1961:
Chapman and Hall.
55. Sengul, O., Tasdemir, C., and Tasdemir, M.A., Influence of aggregate type on mechanical
behavior of normal- and high-strength concretes. ACI Materials Journal, 2002. 99(6): p. 528-
533.
56. Topçu, İ.B. and Işıkdağ, B., Effect of expanded perlite aggregate on the properties of
lightweight concrete. Journal of Materials Processing Technology, 2008. 204(1–3): p. 34-38.
57. ASTM, Standard Specification for Lightweight Aggregates for Structural Concrete. 2014,
ASTM International: West Conshohocken.
58. American Standards for Testing Materials, Standard specification for lightweight aggregates
for insulating concrete. 2009, ASTM International: West Conshohocken, PA.
59. Pevzner, E. and Gontmakher, V., Method for manufacture of foamed perlite material. 2002:
US.
60. Lura, P., Bentz, D.P., Lange, D.A., Kovler, K., and Bentur, A. Pumice aggregates for internal
water curing. in International RILEM Symposium on Concrete Science and Engineering: A
Tribute to Arnon Bentur. 2004. RILEM Publications SARL.
192
61. Yu, L.-H., Ou, H., and Lee, L.-L., Investigation on pozzolanic effect of perlite powder in
concrete. Cement and Concrete Research, 2003. 33(1): p. 73-76.
62. Sengul, O., Azizi, S., Karaosmanoglu, F., and Tasdemir, M.A., Effect of expanded perlite on
the mechanical properties and thermal conductivity of lightweight concrete. Energy and
Buildings, 2011. 43(2): p. 671-676.
63. Singh, M. and Garg, M., Perlite-based building materials—a review of current applications.
Construction and Building Materials, 1991. 5(2): p. 75-81.
64. Yilmazer, S. and Ozdeniz, M.B., The effect of moisture content on sound absorption of
expanded perlite plates. Building and Environment, 2005. 40(3): p. 311-318.
65. Berge, B., The ecology of building materials. 2nd ed. 2009, Oxford, UK: Elsevier’s Science
& Technology.
66. Englert, M.H.U.S., Acoustical tile containing treated perlite, I. Usg Interiors, Editor. 1999:
US.
67. Nelson, C.R.U.S., Expanded perlite annealing process, C. United States Gypsum, Editor.
2008: US.
68. Baig, M.A.U.S., Acoustic ceiling tiles made with paper processing waste, L.L.C. Usg
Interiors, Editor. 2012: US.
69. Denning Paul, S., Insulating product and its manufacture. 1962: US.
70. Keskey, W., Camisa, J.D., and Meath, K.R., Composite board, containing certain latex
binder compositions, C. The Dow Chemical, Editor. 1992: EP.
71. DePorter, C.D., Dawson, S.D., Battaglioli, M.V., and Sandoval, C.P., Perlite-based
insulation board, I. Johns Manville International, Editor. 2000: US.
72. Izard, D.G.U.S. and Englert, M.H.U.S., Method for manufacturing a mineral wool panel, I.
Usg Interiors, Editor. 1993: US.
73. Fernando, J.A.U.S. and Rioux, R., Ultra low weight insulation board, I.L. Unifrax, Editor.
2013: US.
74. Greve, D.R.U.S. and Richards, T.W.U.S., Fire door core, C. Georgia-Pacific, Editor. 1979:
US.
75. Mukherjee, S., Applied Mineralogy: Applications in Industry and Environment
2011, New York, NY: Springer
76. Dickson, T.W., Perlite. The mineral industry of New South Wales. Vol. 31. 1968, Sydney,
Australia: Dept. of Mines, Geological Survey of New South Wales
77. Beikircher, T. and Demharter, M., Heat Transport in Evacuated Perlite Powders for Super-
Insulated Long-Term Storages up to 300 °C. Journal of Heat Transfer, 2013. 135(5): p.
051301-051301.
78. Kropschot, R.H. and Burgess, R.W. Perlite for cryogenic insulation. in Proceedings of the
Cryogenic Engineering Conference, 1962. 1962.
79. Demharter, M., Heat transport in evacuated perlite powder insulations and Its application in
long-term
hot water storages, in Physics. 2011, Technische Universität München.
80. Fulk, M.M., Evacuated powder insulation for low temperatures, in Progress in Cryogenics,
K. Mendelssohn, Editor. 1959, Heywood &Company: London, UK. p. 65-84.
81. Demharter, M., Heat Transport in Evacuated Perlite Powder Insulations and Its Application
in Long-Term Hot Water Storages. Master's Thesis, Technische Universität München,
Faculty of Physics, 2011.
82. Black Igor, A. and Fowle Arthur, A., Insulating assembly, U.S. Patent, Editor. 1964.
83. Dubé, W.P., Sparks, L.L., and Slifka, A.J., Thermal conductivity of evacuated perlite at low
temperatures as a function of load and load history. Cryogenics, 1991. 31(1): p. 3-6.
193
84. Elias, H.G., Plastics, general survey. Ullmann's Encyclopedia of Industrial Chemistry, 2000.
85. Dedeloudis, C. and Karalis, T., Milled expanded volvanic glass as lamellar filler, B.I.M. S.A,
Editor. 2012: EP.
86. Wang, B., Roulston, J.S.U.S., Palm, S.K.U.S., and Hayward, C., Perlite products with
controlled particle size distribution, C. Advanced Minerals, Editor. 2002: US.
87. Agioutantis, Z., Agioutantis, Z.G., and Komnitsas, K. Book of proceedings: advances in
mineral resources management and environmental geotechnology. 2006. Heliotopos
Conferences.
88. Kongkachuichay, P. and Lohsoontorn, P., Phase diagram of zeolite synthesized from perlite
and rice husk ash. SCIENCEASIA, 2006. 32: p. 13-16.
89. Christidis, G.E., Paspaliaris, I., and Kontopoulos, A., Zeolitisation of perlite fines:
mineralogical characteristics of the end products and mobilization of chemical elements.
Applied Clay Science, 1999. 15(3–4): p. 305-324.
90. Ottanà, R., Saija, L.M., Burriesci, N., and Giordano, N., Hydrothermal synthesis of zeolites
from pumice in alkaline and saline environment. Zeolites, 1982. 2(4): p. 295-298.
91. Antonucci, P.L., Crisafulli, M.L., Giordano, N., and Burriesci, N., Zeolitization of perlite.
Materials Letters, 1985. 3(7–8): p. 302-307.
92. Wang, P., Shen, B., Shen, D., Peng, T., and Gao, J., Synthesis of ZSM-5 zeolite from expanded
perlite/kaolin and its catalytic performance for FCC naphtha aromatization. Catalysis
Communications, 2007. 8(10): p. 1452-1456.
93. Khodabandeh, S. and Davis, M.E., Alteration of perlite to calcium zeolites. Microporous
Materials, 1997. 9(3–4): p. 161-172.
94. Steenbruggen, G. and Hollman, G.G., The synthesis of zeolites from fly ash and the properties
of the zeolite products. Journal of Geochemical Exploration, 1998. 62(1–3): p. 305-309.
95. Christidis, G.E. and Papantoni, H., Synthesis of FAU type zeolite Y from natural raw
materials: hydrothermal SiO2-Sinter and Perlite glass. The Open Mineralogy Journal, 2008.
2: p. 1-5.
96. Rosa, M.E., An introduction to solid foams. Philosophical Magazine Letters, 2008. 88(9-10):
p. 637-645.
97. Gibson, L.J. and Ashby, M.F., Cellular solids: structure and properties. 1999: Cambridge
university press.
98. Smits, G.F., Effect of Cellsize Reduction on Polyurethane Foam Physical Properties. Journal
of Thermal Insulation and Building Envelopes, 1994. 17(4): p. 309-329.
99. Rosa, M.E., Pereira, H., and Fortes, M., Effects of hot water treatment on the structure and
properties of cork. Wood and Fiber Science, 1990. 22(2): p. 149-164.
100. Pereira, H., Rosa, M.E., and Fortes, M., The cellular structure of cork from Quercus suber L.
IAWA Journal, 1987. 8(3): p. 213-218.
101. Fortes, M.A., Rosa, M.E., and Pereira, H., A cortiça. 2004: IST Press Lisboa.
102. Cateto, C.A., Barreiro, M.F., Ottati, C., Lopretti, M., Rodrigues, A.E., and Belgacem, M.N.,
Lignin-based rigid polyurethane foams with improved biodegradation. Journal of Cellular
Plastics, 2013. 50(1): p. 81-95.
103. Khemani, K.C., Polymeric foams: an overview, in Polymeric Foams: Science and
Technology, K.C. Khemani, Editor. 1997, American Chemical Society: Washington, DC.
104. Martelli, F., in Twin Screw Extruders. 1983, Van Nostrand Reinhold Co. : New York.
105. Lee, S.-T., Park, C.B., and Ramesh, N.S., Polymeric foams: science and technology. 2006:
CRC Press.
106. Fatt, M.S.H. and Chen, L., A viscoelastic damage model for hysteresis in PVC H100 foam
under cyclic loading. Journal of Cellular Plastics, 2015. 51(3): p. 269-287.
194
107. Leidner, J. and Woodhams, R.T., The strength of polymeric composites containing spherical
fillers. Journal of Applied Polymer Science, 1974. 18(6): p. 1639-1654.
108. Nicolais, L. and Mashelkar, R.A., The strength of polymeric composites containing spherical
fillers. Journal of Applied Polymer Science, 1976. 20(2): p. 561-563.
109. Spanoudakis, J. and Young, R.J., Crack propagation in a glass particle-filled epoxy resin.
Journal of Materials Science, 1984. 19(2): p. 473-486.
110. Alonso, M.V., Auad, M.L., and Nutt, S., Short-fiber-reinforced epoxy foams. Composites
Part A: Applied Science and Manufacturing, 2006. 37(11): p. 1952-1960.
111. Gupta, N. and Nagorny, R., Tensile properties of glass microballoon-epoxy resin syntactic
foams. Journal of Applied Polymer Science, 2006. 102(2): p. 1254-1261.
112. Gupta, N., Zeltmann, S.E., Shunmugasamy, V.C., and Pinisetty, D., Applications of Polymer
Matrix Syntactic Foams. JOM, 2014. 66(2): p. 245-254.
113. Puterman, M., Narkis, M., and Kenig, S., Syntactic Foams I. Preparation, Structure and
Properties. Journal of Cellular Plastics, 1980. 16(4): p. 223-229.
114. Okuno, K. and Woodhams, R.T., Mechanical Properties and Characterization of Phenolic
Resin Syntactic Foams. Journal of Cellular Plastics, 1974. 10(5): p. 237-244.
115. Luoma, E.J. and Watkins, L., Syntactic Foams, in Modern Plastics Encyclopedia. 1979,
McGraw-Hill: New York. p. 273.
116. Lubin, G., Handbook of fiberglass and advanced plastics composites. 1975: RE Krieger
Publishing Company.
117. Lee, S.M., Handbook of composite reinforcements. 1992: John Wiley & Sons.
118. Suh, K., Klempner, D., and Frisch, K., Handbook of Polymeric Foams and Foam Technology.
Hanser, Munich, 2004: p. 189-232.
119. Landrock, A.H., 6. Solvent Cementing and Adhesive Bonding of Foams, in Handbook of
Plastic Foams. William Andrew Publishing/Noyes.
120. Yung, K.C., Zhu, B.L., Yue, T.M., and Xie, C.S., Preparation and properties of hollow glass
microsphere-filled epoxy-matrix composites. Composites Science and Technology, 2009.
69(2): p. 260-264.
121. Gupta, N., Woldesenbet, E., and Mensah, P., Compression properties of syntactic foams:
effect of cenosphere radius ratio and specimen aspect ratio. Composites Part A: Applied
Science and Manufacturing, 2004. 35(1): p. 103-111.
122. Kim, H.S. and Plubrai, P., Manufacturing and failure mechanisms of syntactic foam under
compression. Composites Part A: Applied Science and Manufacturing, 2004. 35(9): p. 1009-
1015.
123. Gupta, N. and Maharsia, R., Enhancement of Energy Absorption in Syntactic Foams by
Nanoclay Incorporation for Sandwich Core Applications. Applied Composite Materials,
2005. 12(3): p. 247-261.
124. Gupta, N., Kishore, Woldesenbet, E., and Sankaran, S., Studies on compressive failure
features in syntactic foam material. Journal of Materials Science, 2001. 36(18): p. 4485-4491.
125. Samsudin, S., Ariff, Z., Zakaria, Z., and Bakar, A., Development and characterization of
epoxy syntactic foam filled with epoxy hollow spheres. Exp. Pol. Lett, 2011. 7: p. 653-660.
126. Mudge, R.S., Process for making a low density syntactic foam product and the resultant
product. 1988, US Patent.
127. Matthews, A.M., Impact resistant thermoplastic syntactic foam composite and method. 1994,
US Patent.
128. Dodiuk, H. and Goodman, S.H., Handbook of thermoset plastics. 2013: William Andrew.
129. Du Pont, P.S., Freeman, J.E., Ritter, R.E., and Wittmann, A., Fiber-reinforced syntactic foam
composites and method of forming same. 1986, US Patent.
195
130. Harrison, E.S., Bridges, D.J., and Melquist, J.L., Thermosetting syntactic foams and their
preparation. 2000, US Patent.
131. Kishore, Shankar, R., and Sankaran, S., Gradient syntactic foams: Tensile strength, modulus
and fractographic features. Materials Science and Engineering: A, 2005. 412(1–2): p. 153-
158.
132. Narkis, M., Puterman, M., and Kenig, S., Syntactic Foams II. Preparation and
Characterization Of Three-Phase Systems. Journal of Cellular Plastics, 1980. 16(6): p. 326-
330.
133. Ashton-Patton, M.M., Hall, M.M., and Shelby, J.E., Formation of low density
polyethylene/hollow glass microspheres composites. Journal of Non-Crystalline Solids,
2006. 352(6–7): p. 615-619.
134. Patankar, S.N., Das, A., and Kranov, Y.A., Interface engineering via compatibilization in
HDPE composite reinforced with sodium borosilicate hollow glass microspheres.
Composites Part A: Applied Science and Manufacturing, 2009. 40(6–7): p. 897-903.
135. Patankar, S.N. and Kranov, Y.A., Hollow glass microsphere HDPE composites for low
energy sustainability. Materials Science and Engineering: A, 2010. 527(6): p. 1361-1366.
136. Cravens, T.E., Syntactic Foams Utilizing Saran Microspheres. Journal of Cellular Plastics,
1973. 9(6): p. 260-267.
137. Narkis, M., Gerchcovich, M., Puterman, M., and Kenig, S., Syntactic Foams III. Three-Phase
Materials Produced from Resin Coated Microballoons. Journal of Cellular Plastics, 1982.
18(4): p. 230-232.
138. Seamark, M.J., Innovative Use of Syntactic Foam in GRP Sandwich Construction for Wind
Generator Nacelles, in Composite Structures 4: Volume 1 Analysis and Design Studies, I.H.
Marshall, Editor. 1987, Springer Netherlands: Dordrecht. p. 333-341.
139. Kim, H.S., Syntactic foam. 2003,
Patents: Australia.
140. Hossain, K.M.A., Development of volcanic pumice based cement and lightweight concrete.
Magazine of Concrete Research, 2004. 56(2): p. 99-109.
141. Taherishargh, M., Belova, I.V., Murch, G.E., and Fiedler, T., Pumice/aluminium syntactic
foam. Materials Science and Engineering: A, 2015. 635: p. 102-108.
142. Fleischer, C.A. and Zupan, M., Mechanical Performance of Pumice-reinforced Epoxy
Composites. Journal of Composite Materials, 2010. 44(23): p. 2679-2696.
143. Sahin, A., Yildiran, Y., Avcu, E., Fidan, S., and Sinmazcelik, T., Mechanical and Thermal
Properties of Pumice Powder Filled PPS Composites. Acta Physica Polonica A, 2014.
125(2): p. 518-520.
144. Sahin, A., Karsli, N.G., and Sinmazcelik, T., Comparison of the mechanical,
thermomechanical, thermal, and morphological properties of pumice and calcium
carbonate-filled poly(phenylene sulfide) composites. Polymer Composites, 2016. 37(11): p.
3160-3166.
145. Ramesan, M.T., George, A., Jayakrishnan, P., and Kalaprasad, G., Role of pumice particles
in the thermal, electrical and mechanical properties of poly(vinyl alcohol)/poly(vinyl
pyrrolidone) composites. Journal of Thermal Analysis and Calorimetry, 2016. 126(2): p. 511-
519.
146. Alvarado, A., Morales, K., Srubar, W., and Billington, S., Effects of natural porous additives
on the tensile mechanical performance and moisture absorption behavior of PHBV-based
composites for construction. Stanford Undergraduate Research Journal, 2011. 10.
147. Srubar, W.V. and Billington, S.L., PHBV/ground bone meal and pumice powder engineered
biobased composite materials for construction. 2013, Google Patents.
196
148. Han, B., Sun, Z., Chen, Y., Tian, F., Wang, X., and Lei, Q. Space charge distribution in Low-
density Polyethylene (LDPE)/Pumice composite. in 2009 IEEE 9th International Conference
on the Properties and Applications of Dielectric Materials. 2009.
149. Ramesan, M.T., Jose, C., Jayakrishnan, P., and Anilkumar, T., Multifunctional ternary
composites of poly (vinyl alcohol)/cashew tree gum/pumice particles. Polymer Composites,
2016: p. n/a-n/a.
150. Masuhara, E., Nakabayashi, N., Nagata, K., and Takeyama, M., Composite filler and dental
composition containing the same. 1982, US Patents.
151. Ahmetli, G., Dag, M., Deveci, H., and Kurbanli, R., Recycling studies of marble processing
waste: Composites based on commercial epoxy resin. Journal of Applied Polymer Science,
2012. 125(1): p. 24-30.
152. Rashad, A.M., Vermiculite as a construction material – A short guide for Civil Engineer.
Construction and Building Materials, 2016. 125: p. 53-62.
153. Straaten, H.P.V., Rocks for Crops: Agrominerals of Sub-Saharan Africa. 2002, Nairobi,
Kenya: ICRAF 338.
154. Kogel, J.E., Trivedi, N.C., Barker, J.M., and Krukowski, S.T., Industrial Minerals and Rocks
- Commodities, Markets, and Uses (7th Edition). Society for Mining, Metallurgy, and
Exploration (SME).
155. JUN, S., XIAOLIN, F., YONG, R., BOWEN, C., and XINYU, R., Preparation method of
phenolic resin/ expanded vermiculite composite flame retardant heat preservation material.
2016, Chinese Patents.
156. Verbeek, C.J.R. and du Plessis, B.J.G.W., Density and flexural strength of phosphogypsum–
polymer composites. Construction and Building Materials, 2005. 19(4): p. 265-274.
157. Yu, J., He, J., and Ya, C., Preparation of phenolic resin/organized expanded vermiculite
nanocomposite and its application in brake pad. Journal of Applied Polymer Science, 2011.
119(1): p. 275-281.
158. Zeng, X., Cai, D., Lin, Z., Cai, X., Zhang, X., Tan, S., and Xu, Y., Morphology and thermal
and mechanical properties of phosphonium vermiculite filled poly(ethylene terephthalate)
composites. Journal of Applied Polymer Science, 2012. 126(2): p. 601-607.
159. Patro, T.U., Harikrishnan, G., Misra, A., and Khakhar, D.V., Formation and characterization
of polyurethane—vermiculite clay nanocomposite foams. Polymer Engineering & Science,
2008. 48(9): p. 1778-1784.
160. Qian, Y., Lindsay, C.I., Macosko, C., and Stein, A., Synthesis and Properties of Vermiculite-
Reinforced Polyurethane Nanocomposites. ACS Applied Materials & Interfaces, 2011. 3(9):
p. 3709-3717.
161. Xu, J., Meng, Y.Z., Li, R.K.Y., Xu, Y., and Rajulu, A.V., Preparation and properties of
poly(vinyl alcohol)–vermiculite nanocomposites. Journal of Polymer Science Part B:
Polymer Physics, 2003. 41(7): p. 749-755.
162. Tjong, S.C., Meng, Y.Z., and Hay, A.S., Novel Preparation and Properties of
Polypropylene−Vermiculite Nanocomposites. Chemistry of Materials, 2002. 14(1): p. 44-51.
163. Mittal, V., Epoxy—Vermiculite Nanocomposites as Gas Permeation Barrier. Journal of
Composite Materials, 2008. 42(26): p. 2829-2839.
164. Zhang, Y., Han, W., and Wu, C.-F., Preparation and Properties of Polypropylene/Organo-
Vermiculite Nanocomposites. Journal of Macromolecular Science, Part B, 2009. 48(5): p.
967-978.
165. Harikrishnan, G., Lindsay, C.I., Arunagirinathan, M.A., and Macosko, C.W., Probing
Nanodispersions of Clays for Reactive Foaming. ACS Applied Materials & Interfaces, 2009.
1(9): p. 1913-1918.
197
166. Li, X., Lei, B., Lin, Z., Huang, L., Tan, S., and Cai, X., The utilization of organic vermiculite
to reinforce wood–plastic composites with higher flexural and tensile properties. Industrial
Crops and Products, 2013. 51: p. 310-316.
167. Avcu, E., Çoban, O., Bora, M.Ö., Fidan, S., Sınmazçelik, T., and Ersoy, O., Possible use of
volcanic ash as a filler in polyphenylene sulfide composites: Thermal, mechanical, and
erosive wear properties. Polymer Composites, 2014. 35(9): p. 1826-1833.
168. Bora, M.Ö., Çoban, O., Fİdan, S., Kutluk, T., and Sinmazçelİk, T., Surface Modification
Effect of Volcanic Ash Particles Using Silane Coupling Agent on Mechanical Properties of
Polyphenylene Sulfide Composites. Acta Physica Polonica A, 2016. 129(4): p. 495-497.
169. Bora, M.Ö., Çoban, O., Kutluk, T., Fİdan, S., and Sinmazçelİk, T., The influence of heat
treatment process on mechanical properties of surface treated volcanic ash
particles/polyphenylene sulfide composites. Polymer Composites, 2016.
170. Trinidad, S., Luis, J., Vite, J., Franco, A., Mendoza Nuñez, M., and Gutiérrez Torres, C.d.C.
Mechanical behavior of ceramic and polymer composites reinforced with volcanic ashes. in
Key Engineering Materials. 2010. Trans Tech Publ.
171. Fidan, S., Volcanic Ash Reinforcement Concentration Effect on Thermal Properties of
Polyvinyl Chloride Composites. Acta Physica Polonica A, 2015. 127(4): p. 1002-1003.
172. Bora, M.Ö., Evaluation of Volcanic Ash Concentration Effect on Mechanical Properties of
Poly(Vinyl Chloride) Composites. Acta Physica Polonica A, 2015. 127(4): p. 1004-1006.
173. Avcu, E., Çoban, O., Özgür Bora, M., Fidan, S., Sınmazçelik, T., and Ersoy, O., Possible use
of volcanic ash as a filler in polyphenylene sulfide composites: Thermal, mechanical, and
erosive wear properties. Polymer Composites, 2014. 35(9): p. 1826-1833.
174. Cernohous, J.J., Compositions and methods for producing high strength composites. 2012,
Google Patents.
175. Çoban, O., Özgür Bora, M., Kutluk, T., Fidan, S., and Sinmazçelik, T., Heat treatment effect
on thermal and thermomechanical properties of polyphenylene sulfide composites reinforced
with silane-treated volcanic ash particles. Polymer Composites, 2016: p. n/a-n/a.
176. Çoban, O., Bora, M.Ö., Kutluk, T., Fİdan, S., and Sinmazçelİk, T., Effect of Silane as
Coupling Agent on Dynamic Mechanical Properties of Volcanic Ash Filled PPS Composites.
Acta Physica Polonica A, 2016. 129(4): p. 492-494.
177. Vite-Torres, M., Vite, J., Laguna-Camacho, J.R., Castillo, M., and Marquina-Chávez, A.,
Abrasive wear on ceramic materials obtained from solid residuals coming from mines. Wear,
2011. 271(9–10): p. 1231-1236.
178. Vite-Torres, J., Vite-Torres, M., Laguna-Camacho, J.R., Escalante-Martínez, J.E., Gallardo-
Hernández, E.A., and Vera-Cardenas, E.E., Wet abrasive behavior of composite materials
obtained from solid residuals mixed with polymer and ceramic matrix. Ceramics
International, 2014. 40(7, Part A): p. 9345-9353.
179. Baker, C.H. and Smail, V., Polyurethane composite matrix material and composite thereof.
2010, Google Patents.
180. Clay, E.L. and Baker, J.L., Natural sandwich of filled polyurethane foam. 1979, Google
Patents.
181. Lukosiute, I., Levinskas, R., Sapragonas, J., and Kviklys, A., Transition Layer in Composites
with a Plasticized Filler and Its Influence on the Strength Properties. Mechanics of
Composite Materials, 2004. 40(2): p. 151-158.
182. Sherman, N. and Cameron, J.H., Method of manufacturing improved mineral board. 1981.
183. Lu, Z., Xu, B., Zhang, J., Zhu, Y., Sun, G., and Li, Z., Preparation and characterization of
expanded perlite/paraffin composite as form-stable phase change material. Solar Energy,
2014. 108: p. 460-466.
198
184. Lu, Z., Hou, D., Xu, B., and Li, Z., Preparation and characterization of an expanded
perlite/paraffin/graphene oxide composite with enhanced thermal conductivity and leakage-
bearing properties. RSC Advances, 2015. 5(130): p. 107514-107521.
185. Karaıpeklı, A., Sarı, A., and Kaygusuz, K., Thermal Characteristics of Paraffin/Expanded
Perlite Composite for Latent Heat Thermal Energy Storage. Energy Sources, Part A:
Recovery, Utilization, and Environmental Effects, 2009. 31(10): p. 814-823.
186. Shastri, D. and Kim, H.S., A new consolidation process for expanded perlite particles.
Construction and Building Materials, 2014. 60: p. 1-7.
187. Kim, H., Syntactic foam. 2005, Google Patents.
188. Arifuzzaman, M. and Kim, H.S., Novel mechanical behaviour of perlite/sodium silicate
composites. Construction and Building Materials, 2015. 93: p. 230-240.
189. Taherishargh, M., Sulong, M.A., Belova, I.V., Murch, G.E., and Fiedler, T., On the particle
size effect in expanded perlite aluminium syntactic foam. Materials & Design (1980-2015),
2015. 66, Part A: p. 294-303.
190. Taherishargh, M., Belova, I.V., Murch, G.E., and Fiedler, T., The effect of particle shape on
mechanical properties of perlite/metal syntactic foam. Journal of Alloys and Compounds,
2017. 693: p. 55-60.
191. Jiang, B., Wang, Z., and Zhao, N., Effect of pore size and relative density on the mechanical
properties of open cell aluminum foams. Scripta Materialia, 2007. 56(2): p. 169-172.
192. Ramamurty, U. and Paul, A., Variability in mechanical properties of a metal foam. Acta
Materialia, 2004. 52(4): p. 869-876.
193. Wen, C.E., Yamada, Y., Shimojima, K., Chino, Y., Hosokawa, H., and Mabuchi, M.,
Compressibility of porous magnesium foam: dependency on porosity and pore size. Materials
Letters, 2004. 58(3–4): p. 357-360.
194. Kolluri, M., Karthikeyan, S., and Ramamurty, U., Effect of Lateral Constraint on the
Mechanical Properties of a Closed-Cell Al Foam: I. Experiments. Metallurgical and
Materials Transactions A, 2007. 38(9): p. 2006-2013.
195. Luong, D.D., Strbik Iii, O.M., Hammond, V.H., Gupta, N., and Cho, K., Development of high
performance lightweight aluminum alloy/SiC hollow sphere syntactic foams and compressive
characterization at quasi-static and high strain rates. Journal of Alloys and Compounds,
2013. 550: p. 412-422.
196. Li, K., Gao, X.L., and Subhash, G., Effects of cell shape and cell wall thickness variations on
the elastic properties of two-dimensional cellular solids. International Journal of Solids and
Structures, 2005. 42(5–6): p. 1777-1795.
197. Sugimura, Y., Meyer, J., He, M.Y., Bart-Smith, H., Grenstedt, J., and Evans, A.G., On the
mechanical performance of closed cell Al alloy foams. Acta Materialia, 1997. 45(12): p.
5245-5259.
198. http://www.australianperlite.com/perlite/. Perlite. 2016 [cited 2016 18/09/2016]; Available
from: http://www.australianperlite.com/perlite/.
199. Lau, K.-t., Lu, M., Chun-ki, L., Cheung, H.-y., Sheng, F.-L., and Li, H.-L., Thermal and
mechanical properties of single-walled carbon nanotube bundle-reinforced epoxy
nanocomposites: the role of solvent for nanotube dispersion. Composites Science and
Technology, 2005. 65(5): p. 719-725.
200. Hong, S.G. and Wu, C.S., DSC and FTIR Analyses of The Curing Behavior of
Epoxy/dicy/solvent Systems on Hermetic Specimens. Journal of Thermal Analysis and
Calorimetry, 2000. 59(3): p. 711-719.
201. Loos, M.R., Coelho, L.A.F., Pezzin, S.H., and Amico, S.C., The effect of acetone addition on
the properties of epoxy. Polímeros, 2008. 18: p. 76-80.
199
202. Arroyo, M., Pineda, J., and Romero, E., Shear Wave Measurements Using Bender Elements
in Argillaceous Rocks. 2010.
203. Suwal, L.P. and Kuwano, R., Disk shaped piezo-ceramic transducer for P and S wave
measurement in a laboratory soil specimen. Soils and Foundations, 2013. 53(4): p. 510-524.
204. Santamarina, J.C., Klein, A., and Fam, M.A., Elastic waves in the continuum, in Soils and
waves: Particulate materials behavior, characterization and process monitoring
2001, John Wiley and Sons Ltd.
205. Auld, B.A., Acoustic fields and waves in solids. 1973: Рипол Классик.
206. Krautkrämer, J. and Krautkrämer, H., Ultrasonic testing of materials. third ed. 1983, New
York: Springer-Verlag Berlin Heidelberg New York.
207. Wachtman, J.B., Mechanical and Thermal Properties of Ceramics: Proceedings. 1969: US
Department of Commerce, National Bureau of Standards.
208. Rice, R.W., Microstructure dependence of mechanical behavior of ceramics, in Treatise on
materials science and technology, R.K. MacCrone, Editor. 1977, Academic Press: New
York. p. 199-381.
209. Rossi, R.C., Prediction of the Elastic Moduli of Composites. Journal of the American
Ceramic Society, 1968. 51(8): p. 433-440.
210. Dutta, S.K., Mukhopadhyay, A.K., and Chakraborty, D., Assessment of Strength by Young's
Modulus and Porosity: A Critical Evaluation. Journal of the American Ceramic Society,
1988. 71(11): p. 942-947.
211. Hasselman, D.P.H. and Fulrath, R.M., Effect of Small Fraction of Spherical Porosity on
Elastic Moduli of Glass. Journal of the American Ceramic Society, 1964. 47(1): p. 52-53.
212. Panakkal, J.P., Willems, H., and Arnold, W., Nondestructive evaluation of elastic parameters
of sintered iron powder compacts. Journal of Materials Science, 1990. 25(2): p. 1397-1402.
213. Manghnani, M.H., Schreiber, E., and Soga, N., Use of ultrasonic interferometry technique
for studying elastic properties of rocks. Journal of Geophysical Research, 1968. 73(2): p.
824-826.
214. Castro, J. and Cashman, K.V., Constraints on rheology of obsidian lavas based on
mesoscopic folds. Journal of Structural Geology, 1999. 21(7): p. 807-819.
215. Richnow, J., Eruptional and post-eruptional processes in rhyolite domes. 1999.
216. Kramar, D. and Bindiganavile, V., Impact response of lightweight mortars containing
expanded perlite. Cement and Concrete Composites, 2013. 37(0): p. 205-214.
217. Fine, T., Sautereau, H., and Sauvant-Moynot, V., Innovative processing and mechanical
properties of high temperature syntactic foams based on a thermoplastic/thermoset matrix.
Journal of Materials Science, 2003. 38(12): p. 2709-2716.
218. Wood, D.M., Soil mechanics: a one-dimensional introduction. 2009: Cambridge University
Press.
219. Selvadurai, A.P., Elastic analysis of soil-foundation interaction. 2013: Elsevier.
220. Sawicki, A. and Świdziński, W., Elastic moduli of non-cohesive particulate materials.
Powder Technology, 1998. 96(1): p. 24-32.
221. Mei, J., Liu, Z., Wen, W., and Sheng, P., Effective dynamic mass density of composites.
Physical Review B, 2007. 76(13): p. 134205.
222. Wang, J.C., Young's modulus of porous materials. Journal of Materials Science, 1984. 19(3):
p. 801-808.
223. Ramakrishnan, N. and Arunachalam, V.S., Effective elastic moduli of porous solids. Journal
of Materials Science, 1990. 25(9): p. 3930-3937.
224. Arnold, M., Boccaccini, A.R., and Ondracek, G., Prediction of the Poisson's ratio of porous
materials. Journal of Materials Science, 1996. 31(6): p. 1643-1646.
200
225. Dunn, M.L. and Ledbetter, H., Poisson's ratio of porous and microcracked solids: Theory
and application to oxide superconductors. Journal of Materials Research, 1995. 10(11): p.
2715-2722.
226. Spriggs, R.M., Expression for Effect of Porosity on Elastic Modulus of Polycrystalline
Refractory Materials, Particularly Aluminum Oxide. Journal of the American Ceramic
Society, 1961. 44(12): p. 628-629.
227. Hasselman, D.P.H., On the Porosity Dependence of the Elastic Moduli of Polycrystalline
Refractory Materials. Journal of the American Ceramic Society, 1962. 45(9): p. 452-453.
228. Tai Te, W., The effect of inclusion shape on the elastic moduli of a two-phase material.
International Journal of Solids and Structures, 1966. 2(1): p. 1-8.
229. Martin, L.P., Dadon, D., and Rosen, M., Evaluation of Ultrasonically Determined Elasticity-
Porosity Relations in Zinc Oxide. Journal of the American Ceramic Society, 1996. 79(5): p.
1281-1289.
230. Hill, R., A self-consistent mechanics of composite materials. Journal of the Mechanics and
Physics of Solids, 1965. 13(4): p. 213-222.
231. Budiansky, B., On the elastic moduli of some heterogeneous materials. Journal of the
Mechanics and Physics of Solids, 1965. 13(4): p. 223-227.
232. Andersen, O., Waag, U., Schneider, L., Stephani, G., and Kieback, B., Novel Metallic Hollow
Sphere Structures. Advanced Engineering Materials, 2000. 2(4): p. 192-195.
233. Schwartz, L.M., Feng, S., Thorpe, M.F., and Sen, P.N., Behavior of depleted elastic
networks: Comparison of effective-medium and numerical calculations. Physical Review B,
1985. 32(7): p. 4607-4617.
234. Wojciechowski, K.W. and Novikov, V.V., Negative Poisson’s ratio and percolating
structures. Task Quarterly, 2001. 5: p. 5-11.
235. Arns, C.H., Knackstedt, M.A., and Pinczewski, W.V., Accurate Vp:Vs relationship for dry
consolidated sandstones. Geophysical Research Letters, 2002. 29(8): p. 44-1-44-4.
236. Markov, K. and Preziosi, L., Heterogeneous Media: Micromechanics Modeling Methods and
Simulations. Meccanica, 2001. 36(2): p. 239-240.
237. Ashkin, D., Haber, R.A., and Wachtman, J.B., Elastic Properties of Porous Silica Derived
from Colloidal Gels. Journal of the American Ceramic Society, 1990. 73(11): p. 3376-3381.
238. Hentschel, M.L. and Page, N.W., Elastic properties of powders during compaction. Part 3:
Evaluation of models. Journal of Materials Science, 2006. 41(23): p. 7902-7925.
239. Phani, K.K. and Niyogi, S.K., Young's modulus of porous brittle solids. Journal of Materials
Science, 1987. 22(1): p. 257-263.
240. Phani, K.K. and Niyogi, S.K., Elastic modulus-porosity relationship for Si3N4. Journal of
Materials Science Letters, 1987. 6(5): p. 511-515.
241. Zhang, L., Gao, K., Elias, A., Dong, Z., and Chen, W., Porosity dependence of elastic
modulus of porous Cr3C2 ceramics. Ceramics International, 2014. 40(1, Part A): p. 191-198.
242. Dı́az, A. and Hampshire, S., Characterisation of porous silicon nitride materials produced
with starch. Journal of the European Ceramic Society, 2004. 24(2): p. 413-419.
243. Rice, R.W., Porosity of Ceramics: Properties and Applications. 1998: CRC Press.
244. Knudsen, F.P., Dependence of Mechanical Strength of Brittle Polycrystalline Specimens on
Porosity and Grain Size. Journal of the American Ceramic Society, 1959. 42(8): p. 376-387.
245. Phani, K.K. and Sanyal, D., The relations between the shear modulus, the bulk modulus and
Young's modulus for porous isotropic ceramic materials. Materials Science and Engineering:
A, 2008. 490(1–2): p. 305-312.
246. Nielsen, L.F., Elastic properties of two-phase materials. Materials Science and Engineering,
1982. 52(1): p. 39-62.
201
247. Andersson, C.A., Derivation of the Exponential Relation for the Effect of Ellipsoidal Porosity
on Elastic Modulus. Journal of the American Ceramic Society, 1996. 79(8): p. 2181-2184.
248. Rice, R.W., Extension of the Exponential Porosity Dependence of Strength and Elastic
Moduli. Journal of the American Ceramic Society, 1976. 59(11-12): p. 536-537.
249. Brown, S.D., Biddulph, R.B., and Wilcox, P.D., A Strength–Porosity Relation Involving
Different Pore Geometry and Orientation. Journal of the American Ceramic Society, 1964.
47(7): p. 320-322.
250. Rice, R.W., Comparison of physical property-porosity behaviour with minimum solid area
models. Journal of Materials Science, 1996. 31(6): p. 1509-1528.
251. Rice, R.W., Comparison of stress concentration versus minimum solid area based
mechanical property-porosity relations. Journal of Materials Science, 1993. 28(8): p. 2187-
2190.
252. Rice, R.W. The porosity dependence of physical properties of materials: a summary review.
in Key Engineering Materials. 1996. Trans Tech Publ.
253. Maji, A.K., Schreyer, H.L., Donald, S., Zuo, Q., and Satpathi, D., Mechanical properties of
polyurethanefoam impact limiters. Journal of Engineering Mechanics, 1995. 121(4): p. 528-
540.
254. Hohe, J.r. and Becker, W., Effective stress-strain relations for two-dimensional cellular
sandwich cores: Homogenization, material models, and properties. Applied Mechanics
Reviews, 2001. 55(1): p. 61-87.
255. Kanny, K., Mahfuz, H., Carlsson, L.A., Thomas, T., and Jeelani, S., Dynamic mechanical
analyses and flexural fatigue of PVC foams. Composite Structures, 2002. 58(2): p. 175-183.
256. Yin, B., Li, Z.-M., Quan, H., Yang, M.-B., Zhou, Q.-M., Tian, C.-R., and Wang, J.-H.,
Morphology and Mechanical Properties of Nylon-1010-filled Rigid Polyurethane Foams.
Journal of Elastomers and Plastics, 2004. 36(4): p. 333-349.
257. Sciamanna, V., Nait-Ali, B., and Gonon, M., Mechanical properties and thermal conductivity
of porous alumina ceramics obtained from particle stabilized foams. Ceramics International,
2015. 41(2): p. 2599-2606.
258. Raj, R.E. and Daniel, B.S.S., Structural and compressive property correlation of closed-cell
aluminum foam. Journal of Alloys and Compounds, 2009. 467(1–2): p. 550-556.
259. Reitz, D.W., Schuetz, M.A., and Glicksman, L.R., A Basic Study of Aging of Foam Insulation.
Journal of Cellular Plastics, 1984. 20(2): p. 104-113.
260. Zwikker, C. and Smoluchowski, R., Physical properties of solid materials. Physics Today,
1955. 8: p. 17.
261. Barnes, H.A., Hutton, J.F., and Walters, K., An introduction to rheology. Vol. 3. 1989:
Elsevier.
262. Fu, S.-Y., Feng, X.-Q., Lauke, B., and Mai, Y.-W., Effects of particle size, particle/matrix
interface adhesion and particle loading on mechanical properties of particulate–polymer
composites. Composites Part B: Engineering, 2008. 39(6): p. 933-961.
263. Todd, M.G. and Shi, F.G., Molecular Basis of the Interphase Dielectric Properties of
Microelectronic and Optoelectronic Packaging Materials. IEEE Transactions on
Components and Packaging Technologies, 2003. 26(3): p. 667-672.
264. Todd, M.G. and Shi, F.G., Characterizing the interphase dielectric constant of polymer
composite materials: Effect of chemical coupling agents. Journal of Applied Physics, 2003.
94(7): p. 4551-4557.
265. Douce, J., Boilot, J.-P., Biteau, J., Scodellaro, L., and Jimenez, A., Effect of filler size and
surface condition of nano-sized silica particles in polysiloxane coatings. Thin Solid Films,
2004. 466(1–2): p. 114-122.
202
266. Hibbeler, R.C., Mechanics of materials. 2008, Singapore: Prentice Hall.
267. Kerner, E.H., The Elastic and Thermo-elastic Properties of Composite Media. Proceedings
of the Physical Society. Section B, 1956. 69(8): p. 808.
268. Counto, U.J., The effect of the elastic modulus of the aggregate on the elastic modulus, creep
and creep recovery of concrete. Magazine of Concrete Research, 1964. 16(48): p. 129-138.
269. Guth, E., Theory of Filler Reinforcement. Journal of Applied Physics, 1945. 16(1): p. 20-25.
270. Paul, B., PREDICTION OF ELASTIC CONSTANTS OF MULTIPHASE MATERIALS.
Technical Report No. 3, in Other Information: Orig. Receipt Date: 31-DEC-59. 1959. p.
Medium: X; Size: Pages: 22.
271. Kisi, E.H. and Howard, C.J., Stress and elastic constants, in Applications of neutron powder
diffraction. 2008, Oxford University Press: Oxford. p. 486.
272. Gupta, N. and Maharsia, R., Enhancement of Energy Absorption in Syntactic Foams by
Nanoclay Incorporation for Sandwich Core Applications. Applied Composite Materials,
2005. 12(3-4): p. 247-261.
273. Williams, J.L. and Johnson, W.J.H., Elastic constants of composites formed from PMMA
bone cement and anisotropic bovine tibial cancellous bone. Journal of Biomechanics, 1989.
22(6): p. 673-682.
274. Chavez, H.L., Alonso-Guzmán, E., Graff, M., and Arteaga-Arcos, J., Prediction of the Static
Modulus of Elasticity Using Four non Destructive Testing. Revista de la Construcción.
Journal of Construction, 2014. 13(1).
275. KURAMA, S. and Elif, E., Characterization of Mechanical Properties of Porcelain Tile
Using Ultrasonics. Gazi University Journal of Science, 2012. 25(3): p. 761-768.
276. Hardin, B.O. and Blandford, G., Elasticity of Particulate Materials. Journal of Geotechnical
Engineering, 1989. 115(6): p. 788-805.
277. Matikas, T., Karpur, P., and Shamasundar, S., Measurement of the dynamic elastic moduli of
porous titanium aluminide compacts. Journal of materials science, 1997. 32(4): p. 1099-1103.
278. Hicher, P.Y. and Chang, C.S., Evaluation of two homogenization techniques for modeling the
elastic behavior of granular materials. Journal of Engineering Mechanics, 2005. 131(11): p.
1184-1194.