Ferroelectric Materials for Energy Harvesting and Storage [1 ed.] 0081028024, 9780081028025

Table of contents :
Front Matter
Copyright
Contributors
Introduction to ferroelectrics and related materials
Ferroelectrics: A chronical journey
Signature of ferroelectricity: A polarization hysteresis loop
Thermodynamics of ferroelectrics
Classification of ferroelectrics
Perovskites
Aurivillius oxides
Tungsten-bronze family
Ilmenite compounds
Polymer ferroelectrics
Other classes of ferroelectrics
Ferroelectric perovskites
Distortions of cubic perovskites to ferroelectric phase
Displacement of B cations inside the oxygen octahedra
Tilt in oxygen octahedra
Distortion of the octahedron
Domain and domain walls in perovskites
Domain switching in perovskites and evolution of P-E loop
Other related phenomena
Piezoelectricity
Pyroelectricity and electrocaloric effect
Flexoelectricity
Crystallographic anisotropy of functional behavior
Characterization of ferroelectrics and related materials
Ferroelectric polarization characterization using Sawyer-Tower circuit
Determination of different piezoelectric coefficients
Resonance-Antiresonance method
Piezoelectric strain coefficient (d)
Electromechanical coupling coefficient (k)
Voltage output constant (g)
Quasi-static method for low-field longitudinal piezoelectric characterization
Piezoresponse force microscopy
Applications of ferroelectrics in energy harvesting
Solar energy harvesting
Mechanical energy harvesting
Magnetic energy harvesting
Thermal energy harvesting
Summary
References
Solar energy harvesting with ferroelectric materials
Introduction
Solar photovoltaics
Fundamentals of physics of solar photovoltaics
The solar spectra
Open circuit voltage, short-circuit current, quantum efficiency, and fill factor
Factors affecting the performance of a conventional solar cell
Photovoltaics with ferroelectrics
Bulk photovoltaic effect
Ferroelectric domain wall model
Schottky-junction effect
Depolarization field model
Design parameters for ferroelectric materials for PV applications
Perovskite photovoltaics
Fabrication of perovskite solar cell
Disadvantages of perovskite solar cell
Stability and limited service life
Noxious material
Tin-based halide perovskite:
Cesium tin iodines:
Methylammonium tin iodide:
Formamidinium tin iodide:
Transition metal oxides
Photochemical conversion of solar energy: Solar water splitting
Basics of solar water splitting
Ferroic materials for photoelectrochemical water splitting: Fundamentals of material requirement
Various ferroelectrics as photoelectrode material for PEC water splitting
Summary
References
Harvesting thermal energy with ferroelectric materials
Introduction
Ferroelectricity
Working principle of ferroelectric thermal energy harvesting
Ferroelectric thermodynamic cycles
Ferroelectric thermal energy harvesters
Other applications
Electrocaloric cooling
Pyroelectric detectors
Summary/future perspective
References
Leveraging size effects in flexoelectric-piezoelectric vibration energy harvesting
Introduction
Direct and converse flexoelectric and piezoelectric effects
Flexoelectric energy harvesting using a centrosymmetric cantilever
Flexoelectrically coupled mechanical equation and modal analysis
Flexoelectrically coupled electrical circuit equation and modal analysis
Closed-form voltage response and vibration response at steady state
Size effects on modal electromechanical coupling coefficient
Case studies and results
Electromechanical coupling coefficient and size effects
Resonant energy harvesting: Electromechanical frequency response and size effects
Size effects in piezoelectric energy harvesting due to flexoelectricity
Flexoelectrically and piezoelectrically coupled mechanical equation and modal analysis
Flexoelectrically and piezoelectrically coupled electrical circuit equation and modal analysis
Closed-form voltage response and vibration response at steady state
Flexoelectric-piezoelectric electromechanical coupling coefficient and size effects
Cases studies and results
Electromechanical coupling coefficient and size effects
Resonant energy harvesting: Electromechanical frequency response and size effects
Conclusions
Acknowledgment
References
Modeling and identification of nonlinear piezoelectric material properties for energy harvesting
Introduction
Representation and implementation of constitutive relations
Direct excitation
Modeling using nonlinear stress and electric displacement constitutive relations
Modeling using electromechanical enthalpy
Reduced-order model: Galerkin discretization
Approximate solution: Method of multiple scales
Parameter identification strategy
Validation of parameter identification strategy
Parametric excitation
Mathematical modeling
Reduced-order model: Galerkin discretization
Approximate solution: Method of multiple scales
Parameter identification strategy
Validation of parameter identification strategy
Conclusions
Appendices
Simplification of weighted residual statement: Direct excitation
Simplification of weighted residual statement: Parametric excitation
References
Sustainable Composites for Lightweight Applications
Copyright
Preface
Key features of this book
Target audiences of this book
Chapter highlights of this book
1. Introduction to composite materials
1.1 Background and context
1.2 Matrices and their types
1.2.1 Types and main functions and the properties of matrices
1.2.1.1 Epoxy resins
1.2.1.2 Polyester resins
1.2.1.3 Vinyl ester resins
1.2.1.4 Phenolic resins
1.2.1.5 Polyethylene
1.2.1.6 Polypropylene
1.2.1.7 Polystyrene
1.2.1.8 Polylactic acid
1.3 Reinforcements and their types
1.3.1 Conventional reinforcements and their types
1.3.1.1 Glass fibres
1.3.1.2 Carbon fibres
1.3.1.3 Ceramic fibres
1.3.2 Natural fibres and their types
1.3.2.1 Advantages and disadvantages of natural fibres
1.4 Main drivers of composite materials
1.5 Application of sustainable composite materials
1.6 Summary
References
Further reading
2. Sustainable natural fibre reinforcements and their morphological structures
2.1 Commonly used sustainable materials (plant-based natural fibres reinforcements in composites)
2.1.1 Hemp fibres
2.1.2 Flax fibres
2.1.3 Jute fibres
2.1.4 Kenaf fibres
2.1.4.1 Advantages of kenaf fibres
2.1.5 Date palm fibres
2.1.6 Sisal fibres
2.1.7 Oil palm fibres
2.1.8 Banana fibres
2.2 Influence of processing and chemical composition on the properties
2.2.1 Importance of fibre processing parameters
2.2.2 Chemical composition and their influences on the properties
2.2.3 Cellulose structure
2.2.3.1 Cellulose
2.2.3.2 Hemicellulose
2.2.3.3 Lignin
2.3 Mechanical, physical and morphological characteristics of plant fibres
2.3.1 Morphological structure of natural fibres
2.3.1.1 Primary and secondary cell walls
2.3.1.2 Lumen
2.3.2 Effects of variable morphological structure and mechanical properties
2.4 Effects of variable morphology on properties
2.5 Physical and mechanical investigation of single fibres and fibre bundles
2.5.1 Importance of single fibre and fibre bundle properties
2.6 Summary
References
Further reading
3. Lightweight composites, important properties and applications
3.1 Lightweight composite materials: requirements and their key features
3.1.1 Lightweight concept
3.1.2 Lightweight drives
3.1.3 Achieving lightweighting potentials
3.1.4 Lightweighting benefits
3.2 Important properties
3.2.1 Mechanical properties of biobased composites
3.2.1.1 Tensile properties
3.2.1.2 Flexural properties
3.2.1.3 Impact properties
Parameters influencing the impact damage characteristics of composites
3.2.1.4 Fatigue properties
3.2.1.5 Creep behaviour
3.3 Thermal stability of biobased composites
3.3.1 Thermal degradation and stability of biobased composites
3.3.2 Flammability behaviour
3.3.2.1 Parameters influencing cone calorimeter performance
3.3.2.2 Ways for improvement of fire properties of natural fibre reinforcements and composites
3.3.3 Thermal conductivity measurements
3.3.3.1 Ways improving the thermal conductivity of polymer matrix composites
3.4 Environmental effects (water absorption) and their influence in different properties
3.4.1 Moisture diffusion mechanisms in composites
3.4.2 Effects of moisture diffusion the mechanical properties
3.5 Numerical modelling of mechanical properties and damage behaviour of natural fibre-reinforced biobased composites
3.5.1 Background
3.5.2 Predicting mechanical and damage behaviour of natural fibres and composites
3.5.2.1 Finite element method
3.5.2.2 Boundary element method
3.5.2.3 Finite difference method
3.5.3 The prediction of static mechanical properties of composites using FEA
3.6 Applications of lightweight natural fibre composites
3.6.1 Automotive application (road vehicles and land transport)
3.6.2 Aerospace and related application
3.6.3 Marine applications
3.6.4 The building construction application
3.6.5 Other applications
3.7 Conclusions
References
4. Design, manufacturing processes and their effects on bio-composite properties
4.1 Introduction and context
4.2 Eco-design and sustainability (design for environment and design for manufacturing)
4.2.1 Eco-design
4.2.2 Sustainability
4.2.3 Design for environment
4.2.3.1 Materials
4.2.3.2 Production
4.2.3.3 Distribution
4.2.3.4 Use
4.2.3.5 Recovery
4.2.4 Design for manufacture
4.3 Manufacturing processes and their influences on properties of bio-composites
4.3.1 Hand and spray lay-ups
4.3.1.1 Hand lay-up
4.3.1.2 Spray lay-up
4.3.2 Vacuum bagging moulding
4.3.3 Injection moulding
4.3.4 Compression moulding
4.3.5 Vacuum resin infusion
4.3.6 Pre-impregnated resin
4.3.7 Extrusion
4.3.8 Resin transfer moulding
4.3.9 Automated fibre placement
4.3.10 Filament winding
4.3.11 Autoclave moulding
4.3.12 Out-of-autoclave moulding
4.3.12.1 Autoclave and out-of-autoclave curing processes
4.3.13 Additive manufacturing
4.3.14 Brief comparison among manufacturing processes
4.4 Key drivers for cleaner production or green manufacturing
4.5 Manufacturing defects
4.5.1 Microcracks and cracks
4.5.2 Temperature effects
4.5.3 Moisture absorption
4.5.4 Inclusions or contamination
4.5.5 Porosity (void or pores)
4.5.6 Other manufacturing defects
4.6 Conclusions
References
5. Testing and damage characterisation of biocomposite materials
5.1 Introduction and context
5.2 Testing methods for damage characterisation and their importance
5.2.1 Visual inspection or testing
5.2.2 Ultrasonic testing
5.2.3 Thermography testing
5.2.4 Radiography testing
5.2.5 Electromagnetic testing
5.2.6 Acoustic emission inspection
5.2.7 Acousto-ultrasonic testing
5.2.8 Shearography testing
5.2.9 Computed tomography scanning
5.2.10 X-ray micro-computed tomography examination
5.2.11 Scanning electron microscopy
5.3 Damage mechanisms and types (key factors for improving damage resistance)
5.3.1 Damage types and mechanisms
5.3.2 Failure or damage modes
5.3.3 Failure or damage mechanisms associated with FRP composites
5.3.4 Damage detection in FRP composite structures
5.3.5 Key factors for improving damage resistance
5.4 Characterisation of damage modes using destructive and non-destructive damage analysis techniques (SEM, X-ray micro CT, AE, ...
5.4.1 Categorisation of NDT methods for FRP composite materials
5.4.2 Contact versus non-contact techniques
5.4.3 Inspection type versus NDT methods
5.4.4 Physical behaviours and structural integrity
5.5 Experimental and numerical modelling of damage modes and mechanisms
5.5.1 Impact damage
5.5.2 Fatigue life model
5.5.3 Thermal effects
5.6 Conclusions
References
6. Sustainable composites and techniques for property enhancement
6.1 The context of sustainability in composites (comparison of sustainability of biocomposites versus conventional composites t ...
6.2 Inherent properties of natural fibres of biocomposite materials
6.3 Improvement of reinforcements and matrices through various treatments and fillers
6.3.1 Fibre treatments
6.3.2 Chemical treatments
6.3.3 Physical treatments
6.3.4 Additive treatments
6.3.5 Biological treatments
6.4 Approaches towards overall property enhancement via hybridisation, pinning, stitching, among others
6.4.1 Stitching
6.4.2 Hybridisation
6.4.3 Pinning
6.4.4 Knitting
6.4.5 Weaving
6.4.6 Braiding
6.4.7 Tufting
6.5 Summary and further evaluation
6.6 Conclusion
References
7. Future outlooks and challenges of sustainable lightweight composites
7.1 Journey of composite materials towards sustainability
7.2 Market outlook and supply chain scenario
7.3 Challenges of achieving properties for lightweight applications
7.3.1 Materials and manufacturing process
7.3.2 Recyclability and end-of- life option
7.3.3 Long-term durability
7.4 Future outlook
References
Further reading
Index
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
R
S
T
U
V
X
Biomechanical energy harvesting with piezoelectric materials
Introduction
Principles of biomechanical energy harvesting
Theoretical background (analysis with few cases)
Heel strike
Lower body parts (ankle, knee, hip) motion
Center of mass (CM) motion
Arm motion
Motions trajectory during human walking
Electrical response of piezoelectric material: Modeling
Design considerations and performance criterion
Cantilevers
Disks: Cymbal
Disk: Diaphragms
Other configurations
Performance criterion of piezoelectric energy harvester
State-of-the-art
Ceramic-based NGs
Polymer and polymer-ceramic composite-based NGs
Summary
Challenges and future outlook
Materials and process issues
Electrical output
Life span of devices
Encapsulation of energy harvesters
Flexibility of devices
Integration issues
Acknowledgments
References
Harvesting stray magnetic field for powering wireless sensors
Introduction
Energy sources for ubiquitous magnetic fields
Overview of a magnetic energy harvester
Piezoelectric materials
Polycrystalline piezoelectrics
Macro-fiber composite (MFC)
Single-crystal piezoelectrics
Single-crystal fiber composite (SFC)
Magnetostrictive materials
Terfenol-D (TbxDy1-xFe2)
Galfenol (FeGa alloy)
Nickel and Metglas
Multiferroic and magnetoelectric materials
Magneto-mechano-electric (MME) generator
Conversion improvement
Harvested energy transfer optimization
Design and applications
Optimization of the device and energy harvesting
Applications: Autonomous wireless sensor networks
Conclusion
Glossary
Acknowledgments
References
Lead-based and lead-free ferroelectric ceramic capacitors for electrical energy storage
Introduction
Energy storage in dielectric capacitors
Dielectric capacitors in pulsed power systems and their applications
Figures of merit for energy storage in dielectric capacitors
Energy storage density
Energy storage efficiency
Fatigue endurance
Thermal stability
Properties of interest for energy storage in dielectric capacitors
Dielectric permittivity and loss
Polarization and hysteresis loss
Leakage current
Dielectric strength or breakdown field
Lead (Pb) containing dielectric ceramic materials
Pb-based ferroelectrics
Pb-based relaxor ferroelectrics
(Pb,La)(Zr,Ti)O3 (PLZT) RFE ceramics and films
Pb-based solid solution RFEs
Pb-based antiferroelectrics
PbZrO3-based AFE ceramics and films
Pure PbZrO3 AFE materials
A-site-doped PbZrO3 AFE materials
A-, B-site co-doped PbZrO3 AFE materials
Pb-based complex perovskite AFEs
Lead (Pb)-free dielectric ceramic materials
Pb-free ferroelectrics
BaTiO3-based FE ceramics and films
(Bi0.5Na0.5)TiO3-based FE ceramics and films
Pb-free relaxor ferroelectrics
BaTiO3-based RFE ceramics and films
BaTiO3-Bi compound solid solution RFEs
BaTiO3-BiMO3 solid solution RFEs
BaTiO3-Bi(M1,M2)O3 solid solution RFEs
BiFeO3-based RFE ceramics and films
BiFeO3-BaTiO3 solid solution RFEs
BiFeO3-SrTiO3 solid solution RFEs
(K,Na)NbO3-based RFE ceramics and films
Pb-free antiferroelectrics
AgNbO3-based AFE ceramics
NaNbO3-based AFE ceramics
(Bi0.5Na0.5)TiO3-based AFE ceramics and films
(Bi0.5Nb0.5)TiO3-NaNbO3 solid solution AFE ceramics
HfO2-based AFE films
Summary and future directions
Acknowledgments
References
Index
A
B
C
D
E
F
G
H
I
L
M
N
O
P
Q
R
S
T
V
W

Citation preview

Ferroelectric Materials for Energy Harvesting and Storage

Woodhead Publishing Series in Electronic and Optical Materials

Ferroelectric Materials for Energy Harvesting and Storage Edited by

Deepam Maurya Center for Energy Harvesting Materials and Systems (CEHMS), Virginia Tech, Blacksburg, VA, United States Department of Materials Science and Engineering, Virginia Tech, Blacksburg, VA, United States

Abhijit Pramanick Department of Materials Science and Engineering, City University of Hong Kong, Hong Kong SAR, China

Dwight Viehland Department of Materials Science and Engineering, Virginia Tech, Blacksburg, VA, United States

An imprint of Elsevier

Woodhead Publishing is an imprint of Elsevier The Officers’ Mess Business Centre, Royston Road, Duxford, CB22 4QH, United Kingdom 50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States The Boulevard, Langford Lane, Kidlington, OX5 1GB, United Kingdom © 2021 Elsevier Ltd. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/ permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library ISBN: 978-0-08-102802-5 (Print) ISBN: 978-0-08-102879-7 (Online) For information on all Woodhead publications visit our website at https://www.elsevier.com/books-and-journals

Publisher: Matthew Deans Acquisitions Editor: Kayla Dos Santos Editorial Project Manager: Emma Hayes Production Project Manager: Vignesh Tamil Cover Designer: Victoria Pearson Typeset by SPi Global, India

Contributors

Venkateswarlu Annapureddy Department of Physics, National Institute of Technology Tiruchirappalli, Tiruchirappalli, Tamil Nadu, India Bharat G. Baraskar Department of Physics, Savitribai Phule Pune University, Pune, Maharashtra, India Ritamay Bhunia Department of Materials Science and Engineering, Indian Institute of Technology Kanpur, Kanpur, India Tulshidas C. Darvade Department of Physics, Savitribai Phule Pune University, Pune, Maharashtra, India Alper Erturk G.W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA, United States Bushara Fatma Department of Materials Science and Engineering, Indian Institute of Technology Kanpur, Kanpur, India Ashish Garg Department of Materials Science and Engineering, Indian Institute of Technology Kanpur, Kanpur, India Raju Kumar Gupta Department of Chemical Engineering; Center for Environmental Science and Engineering, Indian Institute of Technology Kanpur, Kanpur, India Shashaank Gupta Laboratory for Multifunctional Materials, Virginia Tech India Research and Education Forum, Chennai, TN, India Shashikant Gupta Department of Materials Science and Engineering, Indian Institute of Technology Kanpur, Kanpur, India Muhammad R. Hajj Department of Civil, Environmental and Ocean Engineering, Davidson Laboratory, Stevens Institute of Technology, Hoboken, NJ, United States Geon-Tae Hwang Functional Ceramics Group, Korea Institute of Materials Science, Changwon, Republic of Korea Rahul C. Kambale Department of Physics, Savitribai Phule Pune University, Pune, Maharashtra, India

x

Contributors

Ravi Anant Kishore National Renewable Energy Laboratory, Golden, CO, United States Abhishek Kumar School of Minerals, Metallurgical and Materials Engineering, Indian Institute of Technology, Bhubaneswar, Orissa, India Vamsi C. Meesala Department of Biomedical Engineering and Mechanics, Virginia Tech, Blacksburg, VA, United States Adriane G. Moura G.W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA, United States Somdutta Mukherjee Materials Chemistry Department, CSIR-IMMT Bhubaneswar, Bhubaneswar, Orissa, India Haribabu Palneedi Department of Materials Science and Engineering, Pennsylvania State University, State College, PA, United States Mahesh Peddigari Functional Ceramics Group, Korea Institute of Materials Science, Changwon, Republic of Korea Prateek Department of Chemical Engineering, Indian Institute of Technology Kanpur, Kanpur, India Amritendu Roy School of Minerals, Metallurgical and Materials Engineering, Indian Institute of Technology, Bhubaneswar, Orissa, India Jungho Ryu School of Materials Science and Engineering, Yeungnam University, Gyeongsan, Republic of Korea Jose P.B. Silva Centre of Physics of the Universities of Minho and Porto (CF-UM-UP), Braga, Portugal Jayant Sirohi Department of Aerospace Engineering and Engineering, Mechanics, The University of Texas at AustinAustin, TX, United States Ashutosh Upadhyay Functional Ceramics Group, Korea Institute of Materials Science, Changwon, Republic of Korea

Introduction to ferroelectrics and related materials

1

Shashaank Gupta Laboratory for Multifunctional Materials, Virginia Tech India Research and Education Forum, Chennai, TN, India

1.1

Ferroelectrics: A chronical journey

Ferroelectrics are materials that possess nonzero switchable electric polarization in the absence of electric field [1–3]. Switching of ferroelectric polarization from one state to another can be achieved by applying an electric field higher than a threshold value, commonly known as the coercive field. The term “ferroelectric” was derived from its magnetic analog “ferromagnetic,” a phenomenon discovered well before it and stands for a spontaneous magnetization in materials such as iron. The nonzero switchable polarization in ferroelectrics only exists below a certain fixed temperature, above which these materials transform into a paraelectric state and cease to have such nonlinear remnant polar behavior. This transformation temperature is commonly known as Curie temperature or Curie point (TC). The phenomenon of ferroelectricity was first anticipated by the well-known theoretical physicist P Debye in 1912 during his well-known work on dielectrics, where he hypothesized “a permanent electric dipole moment in certain class of molecules, even in the absence of an electric field” [4, 5]. Since no such material was known then, his hypothesis did not attract much attention. The first experimental observation of ferroelectricity as well as its analogy with ferromagnetism was made about a decade later by J. Valasek in 1921 while studying Rochelle salt, a white crystalline material with orthorhombic symmetry [5]. Valasek, a PhD student then, stated during a conference “…. the dielectric displacement D, electric intensity E and electric polarization P are analogous to B, H and I in the case of magnetism…. this would suggest a parallelism between the behavior of Rochelle salt and Fe, for example, as a ferromagnetic substance.” Later, in 1921, he published the very first ferroelectric loop on Rochelle salt [6, 7]. After Rochelle salt, potassium dihydrogen phosphate (KDP) was the second material found exhibiting ferroelectricity in 1932 [2, 8]. At this point of time, it was believed that ferroelectricity was somehow related to polar H–O bonds, but there was no proper explanation for this [4, 8]. Strategic interests in ferroelectrics were triggered during the second world war when researchers desperately looked for a high dielectric constant material to replace mica from sonar systems used for the detection of submarines [4, 9]. Titania (TiO2) with a dielectric constant of >100 was one of the potential candidates and researchers were trying to improve its performance by doping Ferroelectric Materials for Energy Harvesting and Storage. https://doi.org/10.1016/B978-0-08-102802-5.00001-7 © 2021 Elsevier Ltd. All rights reserved.

2

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 1.1 Venn Diagram representing the group-subgroup relationships among different electronic materials. The numbers in the parentheses show the maximum numbers of point groups which may show corresponding phenomenon.

it with different elements [5]. In the year 1941, one such attempt of doping barium into titania resulted in the discovery of BaTiO3, exhibiting a dielectric constant above 1000, a value at least 10 times higher than any known dielectric material at that time. Later, in 1945, BaTiO3 was also found to be ferroelectric in nature, and hence started a journey into an ever-expanding perovskite family of isostructural materials, which showed excellent performance for a variety of applications, such as ferroelectricity and other related phenomena [10–12]. The phenomenon of ferroelectricity is restricted to only crystalline solids with an insulating nature. Besides, there are two more fundamental requirements for a material to exhibit ferroelectric behavior. First, it should have a spontaneous dipole moment in the absence of any external electric field; and secondly, this permanent dipole moment should be switchable between multiple symmetry equivalent states by the application of electric field [13, 14]. The necessity of having a nonzero spontaneous dipole moment is closely associated with the crystallographic nature of materials and is possible only if there exists an absence of a center of symmetry. The Venn diagram in Fig. 1.1 depicts that, out of 32 points groups, only 10 are allowed to exhibit ferroelectricity.

1.2

Signature of ferroelectricity: A polarization hysteresis loop

As described in the previous section, presence of a switchable polarization is an essential requirement for a material to be ferroelectric. The switchability of the ferroelectric polarization can be confirmed by measuring the dielectric displacement current in response to a cyclic electric field. The polarization response of a ferroelectric material is a hysteresis loop, which is characterized by the three parameters, namely, a saturation polarization (Ps), a remnant polarization (Pr), and a coercive field (Ec). Fig. 1.2A and B shows the hysteresis loop for a ferroelectric material and the triangular voltage waveform signal used to drive the sample. Fig. 1.2B–E depicts the stepwise evolution of the hysteresis loop in a ferroelectric material. An electrical circuit employed to measure the hysteresis loop of a ferroelectric material only determines the instantaneous current flowing through it in response to the applied electric field. The integration of the current over time is calculated

3

Ferroelectric polarization (C/m2)

Introduction to ferroelectrics and related materials

PS Pr

EC

(A)

Electric field (V/m)

V

P

DP3

DP2

t

(B)

DP1

t

(E) I

V

i3 i2

v3

v2

i1

v1

(C)

t

(D)

t

Fig. 1.2 (A) P-E hysteresis loop obtained for a ferroelectric material in response to a triangular voltage signal shown in (B), a small portion of which (in red circle) is enlarged in (C). (D) and (E) are the responses of the ferroelectric material in terms of the instantaneous current and its integration with respect to time, i.e., ferroelectric polarization. Adapted from K.M. Rabe, C.H. Ahn, J.-M. Triscone, Physics of Ferroelectrics: A Modern Perspective, Springer, Berlin, New York, 2007.

4

Ferroelectric Materials for Energy Harvesting and Storage

electronically to determine the ferroelectric polarization as a function of electric field. For an ideal ferroelectric material, the current flowing in response to the applied electric signal is the sum of two components, the switching and the nonswitching currents [3, 15]. The nonswitching current arises due to the displacement of ions with respect to each other in the unit cells and has a profile similar to that of a charging current of a capacitor with time. On the other hand, the switching current, which is the major contributor to the ferroelectric polarization, arises due to the realignment of polarization vectors in response to the applied electric field.

1.3

Thermodynamics of ferroelectrics

In 1954, Devonshire developed his famous thermodynamic theory for ferroelectric crystals under applied electric field and mechanical stress [16]. This attempt was inspired by the previous work of Landau and Ginzburg explaining phase transitions in materials [1]. Though the theory of Devonshire is quite general and can accommodate any sort of crystals, it is considered as one of the very first attempts to understand the physics of ferroelectrics pertaining to the relationships among different thermodynamic properties as well as phase transition behavior [1, 17, 18]. Devonshire successfully used his thermodynamic theory to study different aspects of BaTiO3 [17, 18], which was later expanded for more complex materials like PZT [19–22]. Thermodynamically, the state of a crystal can be defined in terms of several state functions such as internal energy, Helmholtz free energy, enthalpy, etc. In his work, Devonshire chose the Gibbs elastic energy (conventionally abbreviated as G1) for this purpose, which can be represented in differential form as follows [12]. dG1 ðP, X, T Þ ¼ SdT 

X

xi dXj 

i, j

X

Ek dPl

(1.1)

k, l

ði, j ¼ 1, 2, … 6 and k, l ¼ 1, 2, 3Þ where S, xi, and Ek represent the entropy, strain, and electric field components, whereas T, Xj, and Pl are temperature, stress, and polarization, respectively. The Gibbs elastic energy can be represented as a Taylor series expansion in terms of polarization. Since G1 should not be dependent on the orientation of the polarization, the expansion series consists of only even powered terms in P. 1 1 1 1 G1 ðP, X, T Þ ¼ Go ðX, T Þ + a1 P2 + a2 P4 + a3 P6 + a4 P8 ::  EP 2 4 6 8

(1.2)

The first term in this equation, Go(X, T), represents all the contributions to G1 independent of the polarization. The ai’s are temperature-dependent coefficients and may have positive or negative values. In the absence of any external stimulus (E ¼ 0), the free energy of the system should attain a minimum with respect to the state variable P, which gives

Introduction to ferroelectrics and related materials

5

Fig. 1.3 Profiles of Gibb’s elastic energy for (A) a1 > 0 and (B) a1 < 0.

 

∂G1 ∂P



  ¼ 0 ) PS a 1 + a 2 P S 2 + a 3 PS 4 ¼ 0

(1.3)

> 0 ) a1 + 3a2 PS 2 + 5a3 PS 4 > 0

(1.4)

P¼PS

∂2 G1 ∂2 P



P¼PS

Since the coefficients ai’s vary in magnitude as well as in sign with temperature, they determine not only the equilibrium state of the crystal (in terms of minimum elastic Gibbs energy) at any temperature, but also the nature of transformation from one state to another, arising due to change in temperature. Eq. (1.3) suggests that the optimum value for G1 can be attained in two scenarios; either for PS ¼ 0 or a1 + a2 PS 2 + a3 PS 4 ¼ 0. First, the possibility of PS ¼ 0 is considered. On substituting PS ¼ 0 into Eq. (1.4), one can achieve an equilibrium state (∂2G1/ ∂2P) > 0 only if a1 > 0, irrespective of the signs of parameters a2 and a3. In this case, the crystal is said to be in the paraelectric state. On the contrary, if a1 < 0, G1 may have an optimum value only for a nonzero value of the polarization (PS 6¼ 0). Therefore, a change in the parameter a1 from positive to negative value results in a transition from zero to nonzero polarization, which is the paraelectric to ferroelectric transition. Fig. 1.3A and B illustrates the elastic Gibbs free energy curves as a function of polarization for positive and negative values of the parameter a1. If we approximate Eq. (1.2) upto quadratic terms only, for a nonzero electric field, the optimum value of G1 corresponds to 

∂G1 ∂P



E 1 ¼ 0 ) a1 ¼ ¼ P χ P¼PS

(1.5)

The signs of the two other ai’s parameters in Eq. (1.2) give rise to two scenarios regarding the nature of ferroelectric to paraelectric transition. Scenario 1: Second order phase transition (a2 > 0, a3 > 0) From Eq. (1.3), the roots other than PS ¼ 0 can be given as PS ¼ 2

a2 

pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi a2 2  4a1 a3 2a3

(1.6)

6

Ferroelectric Materials for Energy Harvesting and Storage

 Among the two roots,

a2 

 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi a2 2  4a1 a3 =2a3 is always negative for a2 > 0,

a3 > 0, and hence, represents the imaginary state of crystal as PS 2 cannot attain a negative value. However, the careful observation of the other root   pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi a2 + a2 2  4a1 a3 =2a3 indicates a positive finite value of PS 2 if a1 < 0. The two values of PS can be given as sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi a2 + a2 2  4a1 a3 PS ¼  2a3

(1.7)

Fig. 1.4A represents the variation of the elastic Gibbs free energy (G1) as a function of polarization (P) for the scenario PS 6¼ 0. Devonshire assumed that the parameter a1, which is the inverse of the susceptibility at low fields, varies linearly with temperature and can be approximated as a linear function of temperature involving two positive constants a1 S and TC. a1 ¼ a1 S ðT  TC Þ

(1.8)

Fig. 1.4 (A) and (C) Evolution of Gibb’s elastic energy (G1) with temperature for second- and first order transitions, respectively. (B) and (D) show the variation of polarization near the Curie point for two types of transitions.

Introduction to ferroelectrics and related materials

7

Eq. (1.8) can be used to calculate approximate value of PS near TC, which can be given as sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi a1 S ð T C  T Þ PS ¼  a2

(1.9)

Eq. (1.9) indicates that the variation of polarization from paraelectric (a1 > 0) to ferroelectric state (a1 < 0) at TC is a continuous second order transition, as shown in Fig. 1.4B. Scenario 2: First order phase transition (a2 < 0, a3 > 0) Considering Eq. (1.3) again, for a negative value of the parameter a2, there is a possibility of achieving two more minima for G1 in the ferroelectric state (a1 < 0) other than the one corresponding to the paraelectric state of the system (a1 > 0). It implies a stable state of the system corresponding to the three polarization values (two nonzero values  PS and a zero) existing simultaneously. For such scenario, G1 takes the qualitative form shown in Fig. 1.4C with varying temperature. According to the Devonshire assumption, the parameter a1 can still be represented as a linear function of temperature, given as a1 ¼ a1 F ðT  To Þ

(1.10)

However, here To is not same as the Curie temperature, but the Curie-Weiss temperature. A first order transition takes place when G1 attains a minimum after an abrupt change to a nonzero polarization value. Simultaneously solving the Eqs. (1.2), (1.3), and (1.10) yields the first order Curie point to occur at a temperature (Fig. 1.4D) TCF ¼ To +

3 a23 16 a1 a3

(1.11)

The polarization values just before the sharp, discontinuous ferroelectric to paraelectric transition can then be given as PFS

1.4

rffiffiffiffiffiffiffiffiffiffiffi 3a2 ¼ 4a3

(1.12)

Classification of ferroelectrics

A number of material systems exhibit ferroelectricity provided they satisfy the crystallographic considerations. They can be classified on the basis of their formula units under the following broad categories.

8

Ferroelectric Materials for Energy Harvesting and Storage

1.4.1 Perovskites The term “perovskite” has been derived from the name of the naturally occurring mineral CaTiO3, the first known isostructural compound named after the noted Russian mineralogist L. A. Pervoski [14]. Among all the known ferroelectrics, perovskites are most important from an applications perspective owing to their excellent performances as compared to alternatives. The term “perovskite” was originally coined for ABO3 type compounds with cubic crystal structure (space group—Pm3m), which are not ferroelectric in nature. However, a small distortion from cubic symmetry may result in ferroelectricity in the ABO3-type structures. For the sake of accuracy, sometimes noncubic ABO3 structures that exhibit ferroelectricity are referred to as “distorted perovskites” rather than “perovskites.” However, in general, both terms are frequently used interchangeably as synonyms for each other, representing the ABO3 type structures. Perovskites are represented by the formula unit ABO3, where both A and B positions are occupied by metal cations, with A having larger ionic size than B [14, 23, 24]. A typical perovskite structure is illustrated in Fig. 1.5A. The A-site cations occupy the

Fig. 1.5 Structure of (A) perovskites (B) aurivillius oxides and (C) tungsten-bronze (a–b plane) ferroelectric. Adapted from R. Whatmore, S. Kasap, P. Capper (Eds.), Springer Handbook of Electronic and Photonic Materials, Springer International Publishing, Cham, 2017.

Introduction to ferroelectrics and related materials

9

eight corners of the cell, whereas the B-site ones are positioned at the center of the cell. The two cations have 12-fold and 6-fold coordinations, respectively. The oxygen anions occupy the six “face centered” positions, forming an octahedron around the B-site cation. The stability of perovskites is often governed by tolerance factor (t), defined in Eq. (1.13a) [11, 24]. For a variety of A- and B-site cations of radius RA and RB, respectively (Ro being the radius of oxygen ion), perovskites form an ideal cubic structure (space group—Pm3m) if the tolerance factor attains a unit value (t  1). A deviation in the tolerance factor from the value of unity results in noncubic distorted perovskites, where the stability of the structure is maintained as long as 0.85 < t < 1.05. In the case when the A- and B-sites are occupied by more than one types of ions, weighed averages of the ionic radii can be used to determine the tolerance factor (Eq. 1.13b) [23, 25]. The stability of the structure with respect to the significantly large range of the tolerance factors provides the flexibility to design a variety of compositions by simultaneously doping multiple ions of different sizes on both A- and B-sites. Furthermore, perovskites can also accommodate cations of variable charge valance co-occupying either or both sites, as long as the overall charge neutrality of the system ([ABO3]m (m is an integer greater than one)) is maintained. High structural stability of perovskites with respect to the incorporation of variety of ions is a great advantage, as far as the tailoring of performance is concerned. R A + RO t ¼ pffiffiffi 2 ð RB + RO Þ

(1.13a)

xRA1 + ð1  xÞRA2 + RO t ¼ pffiffiffi 2ðyRB1 + ð1  yÞRB2 + RO Þ

(1.13b)

1.4.2 Aurivillius oxides Aurivillius oxides are another class of ferroelectrics, which also have the presence of the oxygen octahedra. The structure of aurivillius oxides is illustrated in Fig. 1.5B. It consists of layers of perovskite blocks of composition (Am1BmO3m+1)2 in the x–y planes separated by (M2O2)+2 layers at regular fixed intervals along the z axis [11, 24]. The parameter m in the formula unit represents the number of layers of perovskite units present between the two successive (M2O2)+2 layers. In general, the parameter m acquires the value between 1 and 5, where for m ¼ 1, the structure is reduced to the perovskite ABO3 with the metal cation “M” taking the A sites [23]. Bismuth is one of the most common elements used as the metal ion “M” and the resulting Aurivillius compounds having layers of Bi2O+2 2 are also known as bismuth-layered oxides. Bi4Ti3O12 (m ¼ 3), Bi4BaTiO15 (m ¼ 4), and Bi4Ba2Ti5O18 (m ¼ 5) are some of the well-known Aurivillius (bismuth layered) oxides. Among them, Bi4Ba2Ti5O18 in single crystal form is of particular interest for optical devices and FeRAMs due to its excellent ferroelectric behavior and high Curie point of 650 °C [11].

10

Ferroelectric Materials for Energy Harvesting and Storage

1.4.3 Tungsten-bronze family A number of ferroelectrics with the formula unit AxB2O6 acquire the tungsten-bronze structure. Similar to the perovskites and Aurivillius compounds, tungsten-bronze compounds also consist of a network of oxygen octahedra linked with each other at corners. Fig. 1.5C shows the cross-section view of the tungsten-bronze structure. A typical unit cell consists of 10 octahedra, each one of them occupying a B-site cation position at the center. The A ions occupy the six voids among the octahedra, but all of them may or may not be filled and give rise to multiple formula units and distributions [11, 23]. In their paraelectric state, tungsten-bronze compounds generally acquire a centrosymmetric tetragonal structure and transform to either noncentrosymmetric tetragonal or orthorhombic structure below Tc [11]. PbNb2O6, BaNb2O6, SrNb2O6, and their solid solutions are some of the well-known tungsten-bronze ferroelectrics. In these compounds, five out of six voids for the A ions are occupied, while one is empty [10, 26]. Due to their inferior functional behavior compared to perovskites and difficulties in their synthesis, tungsten-bronze materials find limited applicability in commercial devices. However, their high Curie points as compared to most other ferroelectrics make them a viable alternative for strategic high temperature applications, where a high magnitude of functional behavior may not be essential [11].

1.4.4 Ilmenite compounds Ilmenites are also represented by the formula unit ABO3 like perovskites, but have a different arrangement of ions. The difference in the arrangement of ions in ilmenite and perovskite structures is governed by the size of the A cation [23, 24]. When the size of the A cation is too small, the ilmenite structure is more stable than the perovskite one. The Ilmenite structure consists of a hexagonal closed packing of oxygen ions with anions A and B occupying the octahedral voids between them, in contrast to perovskites where only B cations are present in the octahedral voids. Along the c axis, each alternative octahedron site is vacant, while among the filled ones, A and B cations occupy alternating sites. Since the movement of ions can take place only in one direction, Ilmenite ferroelectrics can have 180o domains only. LiNbO3 and LiTaO3 are two of the well-known ilmenite compounds owing their structure to the small size of lithium at A site. These compositions are characterized by very high Curie temperatures and find application in optical and electro-optic devices [27–29].

1.4.5 Polymer ferroelectrics Polymers, such as polyvinylidene fluoride (PVDF) and many of its copolymers, exhibit ferroelectric behavior due to the presence of electrically switchable dipoles present in a few of its polymorphs [30, 31]. PVDF is a simple polymer (monomer unit CH2F2) –(CH2–CF2)n–. In ideal circumstances, the alternative carbon atoms in the chain should have been attached to the hydrogen and fluorine atoms, respectively, though it is not the case with some of the polymorphs and breaking of the order results in a localized switchable dipole moment [23]. In spite of a feeble magnitude of

Introduction to ferroelectrics and related materials

11

ferroelectricity and other related functional properties with low Curie points, polymer ferroelectrics are of great interest as they simultaneously provide ferroelectricity and flexibility [31], a rare combination, which is not possible to attain in the case of any other family of oxide ferroelectrics described above [32].

1.4.6 Other classes of ferroelectrics Besides the above-described well-known families, there are many other ferroelectrics which cannot be categorized under any of them. These “other” ferroelectrics are of very limited interest due to either a very small magnitude of ferroelectricity or a transition temperature well below the room temperature. Pyrochlores with the formula unit A2B2O7 (Tc < room temperature) [11], Rochelle salt [6, 7], KDP [8], and triglycine sulfates (TGS) [23] are some of ferroelectrics in this never-ending list.

1.5

Ferroelectric perovskites

Among all the ferroelectric families described in Section 1.4, perovskites with an ABO3 type structure are the most studied ones due to their far superior performance and greater flexibility of design. Owing to their technological importance, this chapter keeps the perovskites at the center of discussion hereafter, while exploring different aspects of ferroelectricity and other related phenomena. The structure-property relationship in perovskites is not only essential for a variety of applications, but also fascinating from the point of view of fundamental science. This section provides a basic understanding of the different aspects of perovskites which play an important role in determining their ferroelectric behavior as well as other related phenomena.

1.5.1 Distortions of cubic perovskites to ferroelectric phase As the perovskites are cooled through the Curie temperature, their crystal structure is transformed from paraelectric cubic to a lower symmetry ferroelectric phase. The distortion of the unit cell can be attributed to three types of transformations, which are not mutually exclusive and take place either independently, or in association with each other.

1.5.1.1 Displacement of B cations inside the oxygen octahedra On cooling through the Curie temperature, B cations are displaced from their positions at the center of the octahedra in the cubic crystallographic phase. The displacement of B cations in adjacent unit cells can be either parallel or antiparallel with respect to each other: these two scenarios result in ferroelectric and antiferroelectric states, respectively [33, 34]. In the case of parallel displacement of neighboring B cations, the direction of their off-centered displacement with respect to the crystallographic axes determines the structure of the lower symmetry phase. As shown in Fig. 1.6, shifts along the crystallographic orientations of h100i, h110i, and h111i with respect to

12

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 1.6 Quarter of the perovskite unit cell representing the possible directions of the displacement of B cation, and the resulting crystallographic phases (for simplification, figure depicts only the B cation). For MA and MC, the displacements of B cation may be in any direction between h100ipc  h111ipc and h100ipc  h110ipc, respectively.

the parent cubic unit cell result in tetragonal, orthorhombic, and rhombohedral structures, respectively [14, 35]. Shifts along any arbitrary directions between h100i and h111i or h100i and h110i result in monoclinic crystallographic phases with distinct space groups (MA/MB and MC in Vanderbilt notation) [14, 36, 37].

1.5.1.2 Tilt in oxygen octahedra A tilt of the octahedra takes place to accommodate the variable size of the A site cations in complex perovskites [14, 38]. There could be many ways to represent the tilt in the oxygen octahedra, but the one which is most often used in the field of ferroelectrics is the notation system proposed by Glazer [14, 38–40]. In this scheme, the tilt of the octahedra is measured with respect to the three mutually perpendicular tetrad axes parallel to the lattice vectors a, b, and c of the parent cubic unit cell. The notation representing the tilt of the system consists of three letters with a superscript, each defining the tilt with respect to the lattice vectors a, b, and c. Each of the three letters defines the magnitude of the tilt and its repetition (twice or thrice) means equal amounts of tilt with respect to the other corresponding axes. The superscript of each letter defines the direction of the tilt in successive octahedra along the same direction. The three possibilities—(1) tilt in the same direction, (2) tilt in opposite direction, and (3) no tilt—are denoted by the superscripts “+”, “”, and “0”, respectively. Hence, the tilt notation for cubic perovskites having no tilt along any of its three axes is represented by a0a0a0 (Fig. 1.7A). A detailed crystallographic and group theory analyses suggest that there could be a total of 22 types of tilt systems in distorted perovskites, in addition to a0a0a0 which represents no tilt for the ideal cubic perovskite [38, 40].

Introduction to ferroelectrics and related materials

13

Fig. 1.7 Tilts in oxygen octahedral for (A) a0a0a0, (B) a0a0c+, (C) a0a0c, and (D) a a a systems. Adapted from R.J.D. Tilley, Perovskites—Structure-Property Relationships, Wiley, 2016.

The 22 tilt systems can further be divided into three categories depending upon the numbers of nonzero tilts: i.e., tilt systems with only one tilt, two tilts, and three tilts, respectively. For the sake of comparison, Fig. 1.7B–D depict the arrangements of the octahedra along the z axis in the three tilt systems, a0a0c+, a0a0c, and a a a, respectively.

1.5.1.3 Distortion of the octahedron Distortion in the shape of the oxygen octahedron is due to an unequal length of the six B–O bonds and can be understood in terms of the Jahn-Teller theorem [14, 41]. This theorem suggests that degeneracy and energy minimization cannot occur simultaneously in a nonlinear molecule. In other words, the five d orbitals of the B cation (transition metal ion) in BO6 octahedron need to break the degeneracy to minimize the energy of the system. This disruption in degeneracy results in either the elongation or contraction of the bond lengths for two B–O bonds along one of the crystallographic axes with respect to the remaining four B–O bonds. Jahn-Teller distortions in perovskites lead to elongation or contraction of octahedron along one of the crystallographic axes.

14

Ferroelectric Materials for Energy Harvesting and Storage

1.5.2 Domain and domain walls in perovskites As described in previous section, above their respective Curie points, all perovskites ˚. acquire cubic symmetry (space group—Pm3m) with lattice parameter of about 4 A Due to the presence of a center of symmetry, the cubic perovskites do not exhibit ferroelectric behavior. As the temperature is lowered below the Curie point, a distortion of the unit cells takes place resulting in the loss of some of the symmetry elements to acquire lower symmetry crystallographic phases [42]. As it has been described earlier, if the lower symmetry crystallographic phase belongs to any of the 10 favorable point groups, perovskite crystal may exhibit ferroelectric behavior. To minimize the total energy of the system in the lower symmetry phase, the crystal is divided into multiple volume segments, commonly known as ferroelectric domains or simply domains [43, 44]. Each of the ferroelectric domains in a perovskite crystal is characterized by the unique direction in which all its electric dipoles orient and differentiate it from other neighboring domains with some other directions of their respective dipoles [45]. Besides, each of these polarization directions is equally probable, and hence, acquired by equal volume fractions of crystal, which results in zero net polarization. These domains can be switched by the application of electric field to achieve a nonzero net polarization. While the miniaturization of domains causes a decrease in potential energy, the resultant increase in the density of domain walls (volume fraction) has a converse effect on the total energy of the system. Hence, an optimum domain size (and domain-wall density) is acquired to achieve a minimum energy state. The equilibrium size of the domains is different for different compositions and has a major impact on its functional behavior [46–51]. The division of a crystal into ferroelectric domains results in the formation of boundaries between them, commonly known as domain walls. In general terms, they are the two-dimensional (few unit cell thick) networks in the crystal across which two ferroelectric domains exist with well-defined directions for their polarization vectors. The domain walls in ferroelectrics are characterized by the angles between the directions of the polarization vectors existing on the two sides across the boundary. For example, if the polarization vectors in two adjacent domains are inclined to each other by 180o, the domains are known as 180o domains, and the wall between them is referred to as a 180o domain wall. The symmetry of a crystal restricts the possible orientations of the polarization vectors, and hence, the possible domain walls are restricted to few types and depend upon the crystallographic nature of the perovskite composition. For lower symmetry structures, the number of allowed domain orientations is high. Fig. 1.8 shows the allowed orientations of crystallographic domains with respect to the pseudo-cubic unit cell for tetragonal (T), rhombohedral (R), orthorhombic (O), and monoclinic (MA/MB and MC) symmetries, respectively [45, 52]. For example, in Fig. 1.8A, it can be seen for tetragonal crystals that the domains can be oriented along only h100ipc while making angles of 90o or 180o with respect to each other. Hence, a tetragonal ferroelectric perovskite can have only two types of domains, namely 90o and 180o domains. Similarly, rhombohedral crystals can have only 71o, 109o, and 180o domains, while orthorhombic ones only 60o, 90o, 120o, and 180o domains.

Introduction to ferroelectrics and related materials

15

Fig. 1.8 Schematics representing the possible orientations of polarization vectors for (A) tetragonal, (B) rhombohedral, (C) 0rthorhombic, (D) monoclinic MA/MB, (E) monoclinic MC symmetries in pseudo-cubic unit cells.

In addition to domains having fixed crystallographic orientations, the domain walls separating them also have fixed orientations. The orientations of the domain walls in a crystal are governed by the strain compatibility conditions between the domains [53–56]. During a structural phase transformation, the lower symmetry phase has a higher number of orientation states (OS) as compared to a higher symmetry phase. Using the strain compatibility condition for different possible structural transformations, the orientations of the domain walls can be determined and are listed in Table 1.1 [56]. Since for a particular structural transformation domain walls can only have fixed orientations, the phenomenon has been utilized successfully to reveal the crystallographic nature of the crystals [45, 56].

1.5.3 Domain switching in perovskites and evolution of P-E loop As described in the last section, on cooling through the Curie point, a ferroelectric material is divided into domains in order to minimize the total energy of the system. When subjected to an electric field, the domains oriented along distinct symmetryrelated orientations with respect to crystallographic axes, can switch among each other. This process is called poling and results in a hysteresis loop of the ferroelectric

16

Ferroelectric Materials for Energy Harvesting and Storage

Table 1.1 Possible orientations of ferroelastic domain walls for different structural transformations [56]. Structural transformation

Space groups

Cubic ! Tetragonal

m3m ! 4mm, 432 ! 422 43m ! 42m, m3m ! 4/mmm 422 ! 222, 4mm ! mm2 42m ! 222, 42m ! mm2 4/mmm ! mmm 4 ! 2, 4 ! 2 4/m ! 2/m 622 ! 222, 6mm ! mm2 6m2 ! mm2, 6/mmm ! mmm 32 ! 3 3m ! m 3m ! 2/m

Tetragonal ! Orthorhombic

Tetragonal ! Monoclinic Hexagonal ! Orthorhombic

Trigonal – Monoclinic

Orthorhombic ! Monoclinic

222 ! 2, mm2 ! m mm2 ! 2, mmm ! 2/m

Orientations of permissible domain walls x ¼ y, x ¼ y z ¼ x, z ¼  x y ¼ z, y ¼  z

x ¼ 0, y ¼ 0

x ¼ py, x ¼  y/p where p ¼ b + (a2 + b2)1/2/a x ¼ 0, y ¼ 0 pffiffiffiffi pffiffiffiffi x ¼ 3 y, y ¼  3 x pffiffiffiffi pffiffiffiffi x ¼  3 y, y ¼ 3 x y ¼ 0, z ¼ (a/c)x pffiffiffiffi y ¼ 3 x, a(x + pffiffiffiffi 3 y)2cz ¼ 0 pffiffiffiffi y ¼ 3 x, a(x + pffiffiffiffi 3 y)2cz ¼ 0 x ¼ 0, z ¼ 0

polarization. The resultant domain structure of a ferroelectric crystal depends not only on its crystallographic nature, but also on the orientation of the electric field with respect to the crystallographic axes. Fig. 1.9 shows the domain structure for rhombohedral, orthorhombic, and tetragonal crystals achieved by poling them along the h100ipc, h110ipc, and h111ipc directions, respectively. Poling along an arbitrary direction results in more complex domain structures, although minimization in the overall potential energy remains the main driving force behind domain switching. Poling of the polycrystalline form of perovskites is the most common example of this phenomenon where grains are oriented arbitrarily with respect to applied electric field, resulting in diverse domain structure in each of them [57–59]. The polarization of a ferroelectric is the sum of two components, commonly known as extrinsic and intrinsic contributions [60–62]. Among the two mutually exclusive contributions, the extrinsic one results from domain switching on application of an electric field higher than the coercive field [63]. At the same time, relative motion

Introduction to ferroelectrics and related materials

17

Fig. 1.9 Domain structures for rhombohedral, orthorhombic, and tetragonal crystals poled along h100ipc, h110ipc, and h111ipc, respectively. Green arrows on the top of each column represent the directions of electric field.

among the constituent ions within the unit cell results in a simultaneous change in the dipole moment, which results in an intrinsic contribution to the ferroelectric polarization. Generally, the relative motion of the constituent ions, and hence, the intrinsic contribution to the polarization is a linear and reversible process, whereas the extrinsic contribution shows nonlinear and hysteric behavior [60]. As shown in Fig. 1.10 and expressed mathematically in Eq. (1.14a), the two components arising due to domain switching and the change in dipole moment within the unit cell add up to give the ferroelectric hysteresis (P-E) loop. The addition of two mutually exclusive components

Fig. 1.10 (A) Intrinsic and extrinsic contributions to the polarization as a function of applied electric field (B) The resultant P-E loop obtained by the addition of two components.

18

Ferroelectric Materials for Energy Harvesting and Storage

is not limited to polarization alone, but also extends to the dielectric constant as well as piezoresponse [64]. P ¼ Pint + Pext

(1.14a)

K¼K

(1.14b)

int

+K

ext

d ¼ dint + dext

1.6

(1.14c)

Other related phenomena

As illustrated in Fig. 1.1, there are a number of other physical phenomena which have a close relationship with ferroelectricity. These phenomena are used extensively in different modes of energy harvesting and hence this section describes them in brief.

1.6.1 Piezoelectricity Piezoelectricity is the phenomenon of development of an electric charge on the surface of crystalline insulators in response to an applied mechanical stimulus (Fig. 1.11A) [11, 65]. Piezoelectricity was first observed by the Curie brothers in 1880 in a number of crystals such as quartz, rochelle salt, tourmaline, and cane sugar [9]. A few years later, while thermodynamic aspects of piezoelectricity were under development, a converse phenomenon was anticipated as well—where a mechanical strain in a piezoelectric crystal develops in response to an applied electric field. Later, taking a clue from this thermodynamic prediction, the Curies reexamined the crystals showing piezoelectricity and experimentally confirmed the existence of a converse piezoelectric effect (Fig. 1.11B) [9]. The two effects, direct and converse piezoresponse, can be expressed mathematically, as given in Eqs. (1.15a and b). Pi ¼ dijk :Xjk

(1.15a)

0 xij ¼ dijk :Ek

(1.15b)

Fig. 1.11 Schematics showing the (A) direct and (B) converse piezoelectric effects.

Introduction to ferroelectrics and related materials

19

In these equations, the electric field and polarization (Ek and Pi) are first rank tensors, whereas the mechanical stress and strain (Xjk and xij) are second rank tensors. The proportionality constants, dijk and d’ijk in the two equations are third rank tensors known as piezoelectric coefficient. dijk and d’ijk are mathematically equivalent to each other and represent the ability of the given material to interconvert between mechanical and electrical energies. In general, a third rank tensor dijk can have up to 27 coefficients, but due to symmetry restrictions, the number of independent coefficients is reduced to 18 [13, 66]. In matrix form, the Voigt notations are often used to express different components of the matrices. Since a physically measurable second rank tensor property must be symmetric across its diagonal, the subscript for a two indices (i, j for i and j from 1 to 3) matrix can be represented by one index (i for i ¼ 1–6): by convention 11 ! 1, 22 ! 2, 33 ! 3, 23 or 32 ! 4, 31 or 13 ! 5, 12 or 21 ! 6. For example, the components X32, d333, and d121 in tensor notations can be expressed as X6, d33, and d16, respectively, in matrix notations. Eq. (1.15a) can then be expressed in matrix form as 3 X1 3 6X 7 d16 6 2 7 6X 7 d26 5:6 3 7 6 X4 7 d36 4 5 X5 X6 2

2

3 2 d11 d12 d13 d14 d15 P1 4 P2 5 ¼ 4 d21 d21 d23 d24 d25 P3 d31 d32 d33 d34 d35

(1.15c)

A triclinic crystal with no symmetry restrictions, has 18 independent piezoelectric coefficients. However, for other crystal classes, the number of independent piezoelectric coefficients is decreased by restrictions imposed by the presence of additional symmetry elements. As will be described in a later section, poled polycrystalline piezoelectric ceramics belong to the Curie symmetry group of ∞m and have only three nonzero piezoelectric coefficients, namely, d33, d15, and d31.

1.6.2 Pyroelectricity and electrocaloric effect Certain materials exhibit an instantaneous variation of bound charge on the surface when subjected to a change in temperature (Fig. 1.12). This phenomenon is known as the pyroelectric effect and results due to the variation of spontaneous polarization with change in temperature [67–69]. Since a nonzero spontaneous polarization is a fundamental requirement for pyroelectricity, only materials that are noncentrosymmetric can exhibit the phenomenon. Hence, ferroelectrics with a large spontaneous polarization below the Curie point are a significantly important subset of pyroelectric materials (Fig. 1.1). Pyroelectric materials are of great interest in thermal imaging devices and are also potential candidates for thermal energy harvesting. Mathematically, pyroelectricity is expressed in terms of a pyroelectric coefficient (pi), which is a first rank tensor defined as the rate of change of the polarization with respect to temperature at constant electric field and stress [13, 66].

20

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 1.12 Schematic representing the change in surface polarization in a pyroelectric material in response to a change in temperature.

 pi ¼

∂Pi ∂T

 ε,E

ði ¼ 1, 2, 3Þ

(1.16)

The variation of bound charge on the surface of pyroelectric materials results in a flow of pyroelectric current (ip) in a closed circuit. If the temperature of the sample is varied as a function of time as dT/dt, the resultant pyroelectric current can be given as ip ¼ A:p:

dT dt

(1.17)

where A and p are the area and pyroelectric coefficient of the sample, respectively. Pyroelectricity in a material can be split into two components, i.e., primary and secondary contributions [70, 71]. The primary contribution is owed to a mutual displacement of the ions in the unit cell with varying temperature when there is no change in the unit cell dimensions. This contribution to the pyroelectricity is equivalent to the measurement of pyroelectricity at constant strain, rather than at constant stress as defined in Eq. (1.16). The secondary contribution comes from the change in dimensions of the unit cells as a result of variation in temperature, and is equivalent to the measurement under constant stress. The electrocaloric effect is the phenomena converse to the pyroelectric effect and accounts for a change in the temperature of a material on application of an electric field [72, 73]. The electrocaloric effect has gained significant interests in recent times due to its potential for applications in solid-state refrigeration and thermal management. A number of perovskites exhibit an electrocaloric effect arising due to adiabatic depolarization [74–79]. Pb(Zr0.95Ti0.05)O3 (PZT) thin films have been reported to have the highest electrocaloric effect so far [80]. Fig. 1.13 shows the thermodynamic cycle involved in the cooling of ferroelectrics via adiabatic depolarization. The cooling cycle in a ferroelectric material takes advantage of the fact that the entropy of the system changes with temperature as well as with electric field.

Introduction to ferroelectrics and related materials

21

Fig. 1.13 Adiabatic depolarization in electrocaloric materials to achieve a change in temperature in response to an electric field cycle. Adapted from J.F. Scott, Annu. Rev. Mater. Res. 41 (1) (2011) 229–240.

Generally, indirect methods are used to determine the cooling performance of electrocaloric materials in terms of an achievable variation of temperature (ΔT), defined as [71, 76] 1 △T ¼  ρ

Z

Ef

Ei

  T ∂P C ∂T E

(1.18)

1.6.3 Flexoelectricity Flexoelectricity in an insulating material is the coupling between the electric polarization and strain gradient [81–83]. While defining flexoelectricity, one needs to emphasize the term “strain gradient,” which differentiates it from piezoelectricity, which is a coupling between electric polarization and strain [84]. Unlike piezoelectrics, flexoelectricity does not have any prerequisite of a noncentrosymmetric crystal structure. Hence, a larger number of candidates may exhibit this phenomenon. Fig. 1.14 illustrates the development of nonzero polarization in a rectangular plate of a centrosymmetric flexoelectric material, when subjected to a variable strain (bending). At the scale of the unit cell, the bending of the plate disturbs the charge distribution within it and results in a nonzero dipole moment. The flexoelectric coefficient is a fourth rank tensor (μijkl) defined as the induced electric polarization (Pi) per unit variation in the strain gradient (∂ Sjk/∂ xl) as given in Eq. (1.19a). Like piezoelectricity, flexoelectricity also has a converse equivalence defined as an induced elastic stress across an insulator in response to an applied

22

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 1.14 Flexoelectric effect in a centrosymmetric material. Two-dimensional unit cells at the bottom showing the loss of the center of symmetry due to the bending of the flexoelectric plate.

electric field gradient (Eq. 1.19b). Analogous to piezoelectricity, the two tensors μijkl and μ’ijkl, defining the direct and converse effects, are equal [83]. Pi ¼ μijkl

∂Sjk ∂xl

(1.19a)

Xij ¼ μ0ijkl

∂Ek ∂xl

(1.19b)

Owing to low magnitudes and complexities arising due to the large number of fourth rank tensor components, experimental determination of flexoelectric coefficients has been a Herculean task. In fact, all the elements of the flexoelectric tensor matrix μijkl are not yet known for any material [81–83, 85, 86]. A number of experimental as well as theoretical attempts have been made to determine the flexoelectric coefficients for a number of materials, but they lack any coherence in magnitude as well as in sign. Cross and coworkers have reported the development of experimental setups to determine the longitudinal and transverse flexoelectric coefficients (μ1111 and μ1122) for bulk ceramics [85, 87]. They successfully measured the μ1111 and μ1122 values for PMN and SRT in their cubic regimes, which have only three nonzero coefficients, namely μ1111, μ1122, and μ2323. The phenomenon of flexoelectricity has been known for quite some time, but due to a very low magnitude of resultant electric polarization as compared to the known values for piezoelectric materials, it was not considered for any practical application. The foremost reason for the renewal of interest in ferroelectricity in the last decade has been the recent advancements in the technology for the fabrication of reliable nanoscale devices and their potential applications [81]. Since the achievable strain gradient in nanoscale devices could be orders of magnitude higher than corresponding values in bulk materials, it is possible to achieve large values for electric polarization due to flexoelectric effect, which are close to the corresponding values for piezoelectric

Introduction to ferroelectrics and related materials

23

d33flexo (pC/N)

600

a1

500 400 300 200 100 0

(B) d

(A)

100

200

300

400

500

d (µm)

a2

Fig. 1.15 (A) A flexoelectric device proposed by Cross et al. consisting of an array of truncated pyramids of square-shaped top and bottom surfaces with sides of length a1 and a2 respectively, sandwiched between two metal plates separated by a distance d. (B) Variation of effective longitudinal piezoresponse with the height of the truncated pyramids. Adapted from L.E. Cross, J. Mater. Sci. 41 (1) (2006) 53–63.

materials of the same dimensions [84]. In addition, as described above, the absence of symmetry restrictions provides more freedom in choosing flexoelectric materials, as well as their operational temperature range. To illustrate the size-effect on the resultant piezoresponse via flexoelectricity, let us consider the example of a flexoelectric device proposed by Cross et al., illustrated in Fig. 1.15A [85]. The device consists of arrays of truncated pyramidal-shaped islands of flexoelectric material sandwiched between two metallic plates. If a force F is applied normally to the device, the longitudinal piezoelectric coefficient originating due to the flexoelectric effect (with no contribution from conventional piezoresponse), can be given as Flexo d33

μ ¼ 11 d:C11



a22  a21 a21

 (1.20)

where μ11 and C11 are the longitudinal flexoelectric coefficient and stiffness parameter of the flexoelectric material, respectively. The three parameters, a1, a2, and d, define the geometry of the truncated pyramidal islands. Considering the example of (BaSr)TiO3 in the cubic phase, typical values of μ11 and C11 are 100 μC/m and 1.66  1011 N/m2, respectively. If we fix the ratio of the sides of the square faces of the truncated pyramid to be 1:5 (a1: a2), the variation of the piezoelectric coefficient (dFlexo 33 ) with height (d) can be scaled as shown in Fig. 1.15B. The gradual increase in the longitudinal piezoresponse with decreasing dimensions is in sharp contrast with the conventional piezoelectrics as their performance is severely diminished for thin films due to substrate clamping effect [88, 89]. In this scenario, flexoelectricity has the potential to revolutionize the field of nanoscale devices requiring large piezoelectric coefficients [86, 90–92].

24

1.7

Ferroelectric Materials for Energy Harvesting and Storage

Crystallographic anisotropy of functional behavior

The functional properties such as pyroelectricity and piezoelectricity are anisotropic in nature, and hence, the magnitude of their coefficients strongly depends on the direction of measurement with respect to the crystallographic axes. Mathematically, anisotropic properties are represented by tensors of different ranks. Among the functional properties discussed in this chapter, pyroelectricity and polarization can be represented by a first rank tensors (i.e., vectors), whereas piezoelectricity and flexoelectricity are third and fourth rank tensors, respectively. A tensor of rank k consists of 3k independent elements, which means a maximum of k 3 measurements are required to completely define the specific property. However, the crystallographic symmetry of a crystal, or more precisely the point group it belongs to, can reduce the numbers of nonzero independent elements in the tensor. This section puts forward a brief account of the role of point group symmetry in determining the anisotropic behavior of a physical property of a crystal. The transformation of axes from one coordinate system to another is an important tool to understand how a tensor property of the system is affected by its crystallographic nature (point group symmetry). Mathematically, an orthogonal set of axes Z (Z1, Z2, Z3) can be transformed to another orthogonal set of concentric axes Z 0 ðZ1 0 , Z2 0 , Z3 0 Þ according to Eq. (1.21a) Z 0 ¼ T:Z

(1.21a)

where T is the transformation matrix given by 2

3 a11 a12 a13 T ¼ 4 a21 a22 a23 5 a31 a32 a33

(1.21b)

The nine elements of the transformation matrix T (aij, i, j ¼ 1, 2, 3) are the direction cosines between nine (3  3) possible combinations of the axes in two coordinate systems and are mathematically equal to the cosines of the angles between axes. For example, as shown in Fig. 1.16A, a12 is the cosine of the angle between Z1 0 and Z2 axes and so on. Any symmetry element encountered in a crystal brings about a new set of axes, and hence, can be represented by a unique transformation matrix. To understand this, let us consider the simple example of a two-fold rotation axis along Z2. The two-fold rotation (180o) generates a new coordination system as shown in Fig. 1.16B. The nine direction cosines aij can be determined by considering the angles between old and new axes to give the transformation matrix T2 in Eq. (1.22). Table 1.2 lists some of the most common symmetry elements being encountered in perovskites and transformation matrices for them. 2

3 cos 180o cos 90o cos 90o T2 ¼ 4 cos 90o cos 0o cos 90o 5 cos 90o cos 90o cos 180o

(1.22)

Introduction to ferroelectrics and related materials

25

Fig. 1.16 (A) Two sets of orthogonal axes, related by the transformation matrix having nine direction cosine elements, one of them a12 relating Z1’ and Z2 axes has been shown here. (B) Transformation of axes obtained for the two-fold rotation axis along Z3. Table 1.2 Transformation matrices for few selected symmetry elements commonly found in perovskites. Symmetry element Inversion center

Mirror plane normal to axis Z1

Mirror plane normal to axis Z2

Mirror plane normal to axis Z3

Two-fold rotation axis parallel to Z2

Three-fold rotation axis parallel to axis Z3

Three-fold rotation axis parallel to [111]

Four-fold rotation axis along axis Z3

Transformation matrix 2

3 1 0 0 4 0 1 0 5 0 0 1 2 3 1 0 0 4 0 1 05 0 0 1 2 3 1 0 0 4 0 1 0 5 0 0 1 2 3 1 0 0 40 1 0 5 0 0 1 2 3 1 0 0 4 0 1 0 5 0 0 1 2 3 pffiffiffi 1=2 3=2 0 pffiffiffi 4  3=2 1=2 0 5 0 0 1 2 3 0 1 0 40 0 15 1 0 0 2 3 0 1 0 4 1 0 0 5 0 0 1

26

Ferroelectric Materials for Energy Harvesting and Storage

For the point groups consisting of more than one symmetry elements, one needs to transform the coordinate axes multiple times to reveal the complete effect of a point group symmetry on the tensor property of a crystal [66]. For any of the 32 point groups, one may require at most three transformations to achieve the physical property matrix. The minimum number of symmetry elements required to generate the transformed physical property matrix for a few important point groups are listed in Table 1.3. This table includes the Curie group ∞∞m as well, representing the spherical symmetry of polycrystalline ceramics. Neumann’s principle is an important tool for determining the crystallographically dependent anisotropic behavior of tensor properties. According to Neumann’s principle, the symmetry of a physical property should include the symmetry of the point group of the crystal. In simple words, if more than one directions in a crystal are related by a symmetry element, any physical property measured along these directions should have the same magnitude. However, the opposite is not true, if a physical property has equal magnitude along two directions, it does not necessarily mean that these directions are symmetry-related. Mathematically, any tensor property, X’ijk… calculated by transforming the tensor in the original coordinate system Xmno… using direction cosines should be equivalent to it. 0 Xijk… ¼ aim :ajn :ako :Xmno…

(1.23a)

0 ¼ Xmno… Xijk…

(1.23b)

where aim, ajn,etc. are the direction cosines between two coordinate systems. Neumann’s principle is helpful in determining relationships between tensor property elements and symmetry. It is able to reduce the tensor representation to its simplified form by applying symmetry restrictions particular to a point group. The following example demonstrates how Neumann’s principle can be used to achieve this. Consider the simple case of a 1st rank tensor (vector) property, specifically pyroelectricity. We consider a crystal with the point group symmetry 3m. According to Table 1.3 Minimum symmetry elements required to determine transformation matrix for few selected point groups. Point group

Symmetry elements

mm2 3m 4mm m mmm m3m ∞∞m ∞m

m ? Z1, m ? Z2 3 k Z3, m ? Z1 4 k Z3, m ? Z1 m ? Z2 m ? Z1, m ? Z2, m ? Z3 m ? Z1, 3k [1, 1, 1], m ? [1, 1, 0] ∞ ? Z3, ∞ ? Z1, m ? Z1 ∞ ? Z3, m ? Z1

Introduction to ferroelectrics and related materials

27

Table 1.3, the point group 3m requires two symmetry elements, a three-fold axis along Z3 and a mirror plane perpendicular to Z1 to determine the transformation matrix. The transformation matrices for 3 k Z3 and m ? Z1 can be taken from Table 1.2 to determine the transformation matrix for 3m point group as 2

32 3 pffiffiffi 1 0 0 1=2 3=2 0 pffiffiffi T3m ¼ 4 0 1 0 5:4  3=2 1=2 0 5 0 0 1 0 0 1

(1.24)

hence, according to Eq. (1.23a), 2

3 2 32 32 3 pffiffiffi p01 p1 1 0 0 1=2 3=2 0 p ffiffi ffi 4 p02 5 ¼ 4 0 1 0 5:4  3=2 1=2 0 5 4 p2 5 p03 0 0 1 p3 0 0 1 This gives 2

3

2

pffiffiffi 3 3

1

p01 6 p1 : 2  p2 : 2 7 7 pffiffiffi 4 p02 5 ¼ 6 6 7 4 p : 3  p : 1 5 0 1 2 p3 2 2 p3

(1.25)

application of Neumann’s principle Eq. (1.23b) yields 2

3

2

3

2

1

pffiffiffi 3 3

2

3

p01 p1 p1 6 p1 : 2  p2 : 2 7 7 4 5 pffiffiffi 4 p02 5 ¼ 4 p2 5 ) 6 ¼ p2 6 7 4 p : 3  p : 1 5 1 2 p03 p3 p3 2 2 p3

(1.26)

Solving the equality in Eq. (1.26) gives p1 ¼ p2 ¼ 0. Hence, pyroelectricity has only one nonzero coefficient along Z3 for crystals with 3m point group symmetry. In a similar manner, Neumann’s principle can be applied for other tensor properties as well for crystals belonging to any of the 32 point groups. Table 1.4 lists the piezoelectricity, dielectric constant, and pyroelectricity tensors for some of the common point groups encountered in perovskites.

1.8

Characterization of ferroelectrics and related materials

This section provides a brief review of the characterization techniques used for ferroelectric and other related materials.

28

Ferroelectric Materials for Energy Harvesting and Storage

Table 1.4 Tensors representing pyroelectricity, ferroelectricity, dielectric constant, and piezoelectricity for different point groups after the application of Neumann’s principle. Point group 1

mm2

3m

4mm

∞m

m3m

∞∞m

Pyro/ ferroelectricity

Dielectric constant

Piezoelectricity

2

2

2

3 P1 4 P2 5 P3 2 3 0 405 P3 2 3 0 405 P3 2 3 0 405 P3 2 3 0 405 P3 2 3 0 405 0 2 3 0 405 0

K11 4 K12 K13 2 K11 4 0 0 2 K11 4 0 0 2 K11 4 0 0 2 K11 4 0 0 2 K11 4 0 0 2 K11 4 0 0

3 K12 K13 K22 K23 5 K23 K33 3 0 0 K22 0 5 0 K33 3 0 0 K11 0 5 0 K33 3 0 0 K11 0 5 0 K33 3 0 0 K11 0 5 0 K33 3 0 0 K11 0 5 0 K11 3 0 0 K11 0 5 0 K11

3 d11 d12 d13 d14 d15 d16 4 d12 d22 d23 d24 d25 d26 5 d13 d23 d33 d34 d35 d36 2 3 0 0 0 0 d15 0 4 0 0 0 d24 0 0 5 d31 d32 d33 0 0 0 2 3 0 0 0 0 d15 2d22 4 d22 d22 0 d15 0 0 5 d31 d31 d33 0 0 0 2 3 0 0 0 0 d15 0 4 0 0 0 d15 0 0 5 d31 d31 d33 0 0 0 2 3 0 0 0 0 d15 0 4 0 0 0 d15 0 0 5 d31 d31 d33 0 0 0 2 3 0 0 00 0 0 40 0 0 0 0 05 0 0 00 0 0 2 3 0 0 00 0 0 40 0 0 0 0 05 0 0 00 0 0

1.8.1 Ferroelectric polarization characterization using SawyerTower circuit As described in Section 1.2, ferroelectric materials are characterized by a hysteresis loop obtained by plotting the electric polarization as a function of applied electric field (triangular waveform). The first electronic circuit to measure the ferroelectric polarization was designed in 1930 by C B Sawyer and C H Tower to probe Rochelle-salt crystals, the very first known ferroelectrics [93]. Sawyer-Tower circuit do not measure the polarization of a sample directly, rather it measures the dielectric current flowing through the material in response to the applied electric signal. Integration of the current over time is performed electronically to determine the dipole moment per unit volume of the sample, i.e., the ferroelectric polarization [26, 94, 95]. Zt2 PD¼

i dt t1

(1.27)

Introduction to ferroelectrics and related materials

29

R1

Cferro

R2

Co

Fig. 1.17 Sawyer-Tower circuit for the measurement of ferroelectric loop.

A block diagram of a Sawyer-Tower circuit is shown in Fig. 1.17 [93]. A ferroelectric material in the form of a cuboidal slab (Cferro) is connected to a reference capacitor (CO) in series, so that the charge flowing through them is equal [95]. A triangular voltage signal is applied to Cferro through a resistive potential divider. The cathode ray oscilloscope (CRO) used in the original Sawyer-Tower circuit plots two signals on the x–y plane as a function of time. On the abscissa is the voltage applied to the ferroelectric sample Cferro through the resistive divider, whereas on the ordinate is the voltage across the reference capacitor multiplied by its capacitance value to give the charge (Q ¼ CO. VO) flowing through it. Generally, a high capacitance value is preferred for CO so that the larger fraction of the applied voltage falls across Cferro. The plot on the oscilloscope is in the form of a closed loop known as ferroelectric hysteresis loop. Modern instruments used for ferroelectric polarization characterization are much more complex than that is shown in the block diagram shown in Fig. 1.17, but the basic principle remains same. Modern day ferroelectric test systems are designed to minimize the effect of many artifacts arising in the circuitry. The leakage current through different resistors in the circuit, capacitive contributions of cables, and leakage current through the ferroelectric sample are a few of the most common artifacts to mention [94]. The complex electronics used in modern ferroelectric test systems are mainly employed to achieve more accurate determination of the charge accumulated on the ferroelectric sample for variable ranges of applied voltages with variable frequencies. Furthermore, modern ferroelectric testers do not possess CRO, the applied voltage and ferroelectric polarization values are recorded as a function of time independently and plotted electronically.

1.8.2 Determination of different piezoelectric coefficients 1.8.2.1 Resonance-Antiresonance method According to the IEEE and IEC standards on piezoelectrics, the resonance-antiresonance method is the prescribed technique to characterize piezoelectric coefficients [96–98]. Any mechanical system can be characterized by a set of natural vibration frequencies,

30

Ferroelectric Materials for Energy Harvesting and Storage

which is dependent not only on size, but also on the elastic stiffness. When an external alternating force is applied to the system, the modulation in the vibrational amplitude is maximum if the frequency of the applied force is similar to one of its natural frequencies. The phenomenon is called resonance, and the frequency at which it takes place is called the resonance frequency. In addition, there also exists a phenomenon converse to resonance, designated as the antiresonance, which refers to the minimal transfer of stimulus to the system [11]. The corresponding frequency is called antiresonance frequency. Due to the electromechanical coupling in piezoelectric materials, they can be excited to the resonance-antiresonance mode by means of electrical stimulus as well, a method preferred over its mechanical equivalence due to the ease of application [94]. To determine the frequencies at which the two phenomena of resonance and antiresonance take place, a two terminal piezoelectric sample is swept with a small alternating signal with variable frequency to determine the frequency-dependent impedance of the system. Owing to the electromechanical coupling in piezoelectrics, the resonance and antiresonance phenomenon results in the highest and lowest flow of currents, respectively, and hence, local minima and maxima in the impedance values are obtained. Depending upon the mode of vibration of the sample, the resonanceantiresonance frequencies of the system can be used to determine the electromechanical coefficients. A few of the important piezoelectric coefficients are defined below, and Table 1.5 lists the mathematical formulae to calculate them for the different modes of vibrations in piezoelectric samples. To obtain different modes of vibrations, piezoelectric samples need to be shaped according to the IEEE and IEC standards [96, 97]. Different parameters can be calculated by determining the resonance and antiresonance frequencies for the samples by activating the corresponding modes by an electrical stimulus.

1.8.2.1.1 Piezoelectric strain coefficient (d) As defined in Section 1.6.1, the piezoelectric coefficient (d) is the charge generated in a material for per unit applied force. For the converse effect, it is defined as the strain produced in a sample per unit applied electric field. Mathematically, the two coefficients are equal in magnitude.

1.8.2.1.2 Electromechanical coupling coefficient (k) The square of the electromechanical coupling coefficient (k) is a measure of the efficiency of a piezoelectric material in converting a mechanical signal to an electric one, and vice versa.

1.8.2.1.3 Voltage output constant (g) The voltage output constant is defined as the electric field generated by a piezoelectric material on the application of per unit stress.

Introduction to ferroelectrics and related materials

31

Table 1.5 Shapes, dimensions, and poling directions required to activate different modes of vibrations in piezoelectric ceramics and corresponding values of electromechanical coupling coefficients, piezoelectric strain coefficients, and voltage output constants. Material constant symbol Vibration mode

Shape

Radial mode

t

E P

k

d

kp 2 ¼ 2:529

sffiffiffiffiffiffiffiffiffi d E33 T g31 ¼ 31 d31 ¼ k31 ε33 T E Y11

fa  fr fr

d

g

d > 15t

Length mode

k31 2 π fa ¼ 1  k31 2  2 fr π fa cot 2 fr

l

E

P

h

b

l > 4b; b > 3h

b

π fr k33 2 ¼ a  2 f π fr cot 2 fa

l

Longitudinal mode

sffiffiffiffiffiffiffiffiffi d E33 T g31 ¼ 31 T d31 ¼ k31 ε 33 Y11 E

sffiffiffiffiffiffiffiffiffi d E33 T g33 ¼ 33 d33 ¼ k33 ε33 T E Y33

h

h > 2.5b, 2.5l E P

Thickness mode

π kt 2 ¼ 2 π cot 2

l

E P

h

b

Shear mode

l

E

P

h

b

fr fa fr fa

k15 2 ¼ π2 ffar   π fr cot 2 fa

sffiffiffiffiffiffiffiffiffi d E33 T g33 ¼ 33 T d33 ¼ k33 ε 33 Y33 E

sffiffiffiffiffiffiffiffiffi d E11 T g15 ¼ 15 T d15 ¼ k15 ε 11 Y44 E

l>b>h

1.8.2.2 Quasi-static method for low-field longitudinal piezoelectric characterization Quasi-static method of determining the longitudinal piezoelectric coefficient (d33) of an electrically poled sample was first employed by Berlincourt, and hence, is known as the Berlincourt method. Instead of determining the piezoelectric coefficient in

32

Ferroelectric Materials for Energy Harvesting and Storage

Static preload force

Sample

Piezo reference

AC loading system

Amplifier

Fig. 1.18 Schematic diagram representing the working principle of quasi-static d33 meter. Adapted from M.G. Cain, Characterisation of Ferroelectric Bulk Materials and Thin Films, Springer, Dordrecht, 2014.

terms of the resonance-antiresonance frequencies, this method employs the fundamental definition of piezoelectric coefficient (charge produced per unit force) to determine its magnitude. In this methodology, a small signal alternating stress is applied to the sample and the resultant charge appearing on the surface is measured to determine the longitudinal piezoresponse (Fig. 1.18). Due to the ease of measurement and flexibility in terms of sample size and shape, this technique is used widely to determine the low-field longitudinal piezoresponse of a poled sample. While measuring the piezoelectric coefficient by the quasi-static method, one needs to take certain precautions about the magnitude and frequency of the applied stress. The higher magnitude of stress is advantageous in producing larger amounts of charge, and hence, higher signal-to-noise ratio, but can also drive the sample into its nonlinear regime. Hence, there is an optimum magnitude of stress which should be applied to the sample while measuring the piezoelectric coefficient. Similarly, the lower limit of the frequency of the alternating stress signal has also been limited by the electronics which integrate the charge during the cycle, while, on the other hand, one should make sure that the frequency of the signal is much lower than the fundamental resonance modes of the sample. These days, a number of commercial d33 meters are available which take all these factors into consideration. As mentioned earlier, samples with any geometrical shape can be measured by Berlicourt method, as long as they can be clamped between the two directly opposing contacts (Fig. 1.18) with conducting surfaces. With increasing lateral size of the sample, one needs to make sure that it is centered well between the contacts to avoid any artifact in piezoelectric coefficient due to the bending of the sample. Similarly, small sample thicknesses may also give a false value of piezoresponse due to the exertion of a shear force giving rise to the imperfect alignment of the clamping electrodes.

Introduction to ferroelectrics and related materials

33

Fig. 1.19 Working principle of piezoresponse force microscopy (PFM). Adapted from S.V. Kalinin, A. Gruverman, Scanning Probe Microscopy: Electrical and Electromechanical Phenomena at the Nanoscale, Springer, New York, 2007.

1.8.3 Piezoresponse force microscopy Piezoresponse Force Microscopy (PFM) is a variant of contact mode scanning probe microscopy (SPM), in which a sharp conductive tip is scanned across the surface of a ferroelectric material to map the orientation of the local ferroelectric polarization [94, 99–102]. Since nearly all ferroelectric materials are notably piezoelectric, an applied electrical signal through the tip results in the displacement of the surface, leading to its vertical and torsional deflections (Fig. 1.19). These deflections are recorded via a photodetector through a laser beam reflected from the top of the tip with respect to the applied signal. The positions of the photodetector, tip and the direction of the laser beam are chosen in such a way that pure vertical and pure lateral movements of the ferroelectric surface, respectively, result in the vertical and horizontal movement of the laser beam on the detector [103]. Mathematically, the vertical deflection of the tip is given by the distance Z ¼ Z0 + Z1 cos (ωt + φ) in response to an applied voltage signal of V ¼ Vdc + Vac cos (ωt) [101]. The first harmonic displacement of Z1 in this equation is proportional to the longitudinal piezoresponse (d33) of the sample. Similar to the longitudinal piezoresponse, the two lateral responses, X and Y, are proportional to piezoelectric coefficients d35 and d34, respectively [101]. For very low amplitude of the applied voltage, the ferroelectric polarization within a domain can be considered directly proportional to the piezoresponse as d33 ¼ ε. Q P, where ε and Q are the dielectric constant and electrostriction coefficient, respectively. To understand the working principle of PFM, let us consider the example of LiNbO3 crystal, as it has only up and down (180o) domains. For the up domains, the positive voltage of the tip results in the contraction of the surface (Fig. 1.20A) and vice versa; hence, the deflection Z and applied voltage V are out of phase (φ ¼ 180 ° ). On the other hand, the domain with downward polarization expands on applying the positive voltage and results in the in-phase movement of the tip (Fig. 1.20B). For the more complex orientations of the polarization vectors, vertical as well as torsional deflections of the tip need to be considered under the two schemes commonly known as vertical and lateral piezoresponse force microscopy (VPFM and LPFM, respectively). Besides revealing the domain structure of a ferroelectric surface by scanning the surface, switching spectroscopy (SS-PFM) is another effective tool of PFM to

34

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 1.20 Deflection of conducting tip in response to the (A) up and (B) down domains in a ferroelectric material.

determine the local switching behavior of a ferroelectric material. In contrast to bulk characterization, this technique provides the opportunity to determine the switching parameters like coercive voltages (V+ and V), piezoresponse (R+o and R o ), nucleation biases (V+c and V c ), and energy required for switching (area under the hysteresis loop) of an individual domain. The working principle of SS-PFM is demonstrated in Fig. 1.21 [104]. In this technique, a voltage signal V ¼ Vdc + Vac Sin ωt is applied to the AFM tip and the piezoresponse is measured for each incremental change in Vdc for a triangular wave Vac Sin ωt [104]. To avoid any nonlinearity in the piezoresponse due to high dc bias, the piezoresponse is measured at the Vdc ¼ 0 V state after application of each step of bias voltage [104]. Local switching measurements are performed within a single domain, and hence, are less affected by factors like domain widths [100].

Fig. 1.21 (A) SS-PFM data collection at different points along the predetermined positions. (B) Electric signal applied to the tip to probe each point. (C) Hysteresis piezo loop obtained at a point. Adapted from S.V. Kalinin, A. Gruverman, Scanning Probe Microscopy: Electrical and Electromechanical Phenomena at the Nanoscale, Springer, New York, 2007.

Introduction to ferroelectrics and related materials

1.9

35

Applications of ferroelectrics in energy harvesting

In addition to their application in several sensing and actuation devices, the last one decade has seen an enormous interest in ferroelectrics for energy harvesting and storage solutions [105, 106]. These technologies utilize ferroelectricity and other related phenomena described in Section 1.6 to harvest energy from different sources of energy. Ferroelectric solar cells, piezoelectricity-based mechanical energy harvesting, and thermal energy harvesting via pyroelectricity are some of the common examples. Ferroelectrics are considered as potential candidate for energy storage as well [107– 109]. This section provides a brief account on how ferroelectrics and related materials can be utilized for several modes of energy harvesting. Subsequent chapters of this book provide a detailed account of different modes of energy harvesting and storage using ferroelectrics.

1.9.1 Solar energy harvesting Solar energy is one of the cleanest forms of energy available to the mankind with more than sufficient availability for its need. Silicon-based solar cells with low-cost and ease of bulk manufacturing have dominated the commercial sphere and their total installed capacity has crossed over 25 GW worldwide [110, 111]. As the power conversion efficiency of silicon solar cells is approaching the theoretical limit of 30%, scientific community has started to look for the alternative technologies with efficiency higher than the theoretical limit of silicon-based solar cells [112]. The theoretical efficiency of the silicon-based solar cells is proportional to the product of the open circuit voltage (VOC) and short circuit current (ISC). In a semiconductor, open circuit voltage is always limited by its bandgap (VOC < Eg). The higher bandgap materials may provide the enhanced open circuit voltage, but at the same time result in a lower short circuit current as only photons from solar spectrum having energy higher than the bandgap, can generate an electron-hole pair [113]. Ferroelectric materials having internal electric bias due to the nonzero spontaneous polarization provide VOC much higher than its band gap. Also, multiple ferroelectric domains stack up as a series combination of voltage sources between the external electrodes to give high value of VOC, proportional to the domain-wall density [114, 115]. The current research on ferroelectric solar cells is focused on decreasing the bandgap of the ferroelectrics to enhance the short circuit current in order to achieve an optimum efficiency [110].

1.9.2 Mechanical energy harvesting The ability of piezoelectric materials to convert mechanical energy into an electric stimulus provides an attractive mechanism to harvest abandoned mechanical energy from different sources. There can be countless sources of stray mechanical energy which can be tapped for energy harvesting. Moving parts of machines, movement of humans and other living species, and vibrations in civil structures are some of the common sources which have been explored for energy harvesting via

36

Ferroelectric Materials for Energy Harvesting and Storage

piezoelectrics [105, 116–118]. A two-pronged approach derives the research in the area of piezoelectric energy harvesting. First approach focuses on the development of the piezoelectric materials with enhanced figures of merit such as electromechanical coupling coefficient, piezoelectric strain coefficient, and piezoelectric voltage constant. This research includes developing new piezoelectric compositions as well as using domain engineering to achieve higher performance in existing compositions [118–122]. In last decade, several crystallographically textured ceramics and single crystals have been developed to achieve higher performance in existing piezoelectric materials. The second approach is more specific to the source of stray mechanical energy and relies on optimizing the design of the energy harvesting device to achieve optimum efficiency [105]. The most challenging part of piezoelectric energy harvesting is the high resonance frequency of the harvesting devices as compared to the natural frequencies of vibration for most of the stray mechanical energy sources. This mismatch in the frequencies results in a very low efficiency of piezoelectric energy harvesters. In spite of this drawback, piezoelectric energy harvesters are quite useful in providing wireless, on-site power to sensors located at remote locations. With decreasing power consumption for different sensing devices used in structural health monitoring and other applications, there has been a constant thrust in recent times to make them energetically self-reliant and wireless by providing an on-site source of renewable energy. In addition to the small sources of stray mechanical energy described above, piezoelectrics have also been considered for harvesting large-scale energy from ocean waves and water currents in rivers [123]. There have also been several attempts to use piezoelectrics in multimodal energy harvesting devices to achieve higher efficiency of energy harvesting at locations where more than one types of energies are available for harvesting [124, 125].

1.9.3 Magnetic energy harvesting Similar to the phenomenon of piezoelectric energy harvesting from the stray mechanical energy, magnetoelectric materials can be used to harvest stray magnetic fields generated at different sources of varying magnetic field [126–128]. In general, magnetoelectric materials used for energy harvesting are two phase composites involving piezoelectric and magnetostrictive materials. A change in magnetic field at the magnetostrictive material results in a strain, which is transferred to the piezoelectric phase and converted to electric energy [128–133]. Analogous to the approach employed for piezoelectric energy harvesting, the research strategy in the area of magnetoelectric energy harvesting is also two-pronged, developing high performance composite materials as well as optimizing the design of harvesting devices to achieve optimum performance.

1.9.4 Thermal energy harvesting In addition to their application in thermal imaging, pyroelectric materials can also be used for thermal energy harvesting [73, 74, 134–136]. As the pyroelectric current flowing through a material is proportional to the rate of change of temperature with

Introduction to ferroelectrics and related materials

37

time rather than its absolute value (Eq. 1.17), pyroelectric energy harvesting is more suitable for the sources having a continuous variation in temperature. Besides, pyroelectric current being proportional to area of the harvesting device, thin films are the preferred form of pyroelectric materials for energy harvesting. In addition to the efforts to achieve high pyroelectric coefficient materials, a significant effort in this area has been focused on devising high efficiency electric polarization-electric field cycle to achieve optimum energy harvesting from a heat source [73, 135, 137].

1.10

Summary

In summary, the present chapter provides a brief account of the different aspects of ferroelectrics and related materials. Thermodynamic stability of multiple ferroelectric states has been discussed along with the crystallographic necessities to access them. A brief classification of the different types of ferroelectric materials has been followed by a somewhat detailed discussion on perovskites, the most appropriate class of ferroelectrics for energy harvesting applications. Piezoelectricity and pyroelectricity, inextricable properties of ferroelectrics, have been discussed owing to their significance for mechanical and thermal energy harvesting, respectively. A relatively new class of materials called flexoelectrics, considered to be potential candidates to provide much higher piezoelectric-like behavior at small scales, has been introduced as well. Different techniques used to characterize ferroelectrics have been discussed in brief. The chapter also discusses the anisotropic behavior of ferroelectrics, an essential factor to consider while optimizing the energy harvesting performance of these materials in different modes of operation.

References [1] M.E. Lines, A.M. Glass, Principles and Applications of Ferroelectrics and Related Materials, Clarendon Press, Oxford (England), 1977. [2] G.H. Haertling, J. Am. Ceram. Soc. 82 (4) (1999) 797–818. [3] K.M. Rabe, C.H. Ahn, J.-M. Triscone, Physics of Ferroelectrics: A Modern Perspective, Springer, Berlin, New York, 2007. [4] J.C. Slater, J. Chem. Phys 9 (1941) 16–33. [5] R. E. N. L E Cross, W.D. Kingery, E. Lense, S. American Ceramic and Meeting (Eds.), High Technology Ceramics: Past, Present, and Future: The Nature of Innovation and Change in Ceramic Technology, American Ceramic Society, Westerville, OH, 1987. [6] J. Valasek, Phys. Rev. 19 (5) (1922) 478–491. [7] J. Valasek, Phys. Rev. 17 (4) (1921) 475–481. [8] G. Busch, Ferroelectrics 71 (1) (1987) 43–47. [9] W.P. Mason, J. Acoust. Soc. Am. 70 (6) (1981) 1561–1566. [10] K. Uchino, Ferroelectric Devices, second ed., CRC Press, 2009. [11] B. Jaffe, W.R. Cook, H.L. Jaffe, Piezoelectric Ceramics, Academic Press, London; New York, 1971. [12] Y. Xu, Ferroelectric Materials and Their Applications, North-Holland, Amsterdam; New York, NY, USA, 1991.

38

[13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48]

Ferroelectric Materials for Energy Harvesting and Storage

J.F. Nye, Physical Properties of Crystals, Clarendon Press, Oxford, 1957. R.J.D. Tilley, Perovskites—Structure-Property Relationships, Wiley, 2016. Y.M. Kang, et al., Jpn. J. Appl. Phys 41 (2002) 694. A.F. Devonshire, Adv. Phys. 3 (10) (1954) 85–130. A.F. Devonshire, Philos. Mag. 42 (333) (1951) 1065. A.F. Devonshire, Philos. Mag. 40 (309) (1949) 1040. M.J. Haun, Z.Q. Zhuang, E. Furman, S.J. Jang, Ferroelectrics 99 (1) (1989) 45–54. Z.Q. Zhuang, M.J. Haun, S.J. Jang, L.E. Cross, IEEE Trans. Ultrason. Ferroelectr. Freq. Control 36 (4) (1989) 413–416. M.J. Haun, E. Furman, S.J. Jang, L.E. Cross, Ferroelectrics 99 (1989) 13–25. M.J. Haun, 8818001, The Pennsylvania State University, 1988. R. Whatmore, S. Kasap, P. Capper (Eds.), Springer Handbook of Electronic and Photonic Materials, Springer International Publishing, Cham, 2017. B. Stojanovic, Magnetic, Ferroelectric, and Multiferroic Metal Oxides, Elsevier, 2018. J.A. Brehm, J.W. Bennett, M.R. Schoenberg, I. Grinberg, A.M. Rappe, J. Chem. Phys. 140 (22) (2014) 224703. Y.T. Tsui, P.D. Hinderaker, F.J. McFadden, Rev. Sci. Instrum. 39 (10) (1968) 1423–1424. T. Volk, M. W€ohlecke, Lithium Niobate: Defects, Photorefraction and Ferroelectric Switching, Springer, Berlin, 2008. D.L. Staebler, J.J. Amodei, IEEE Trans. Sonics Ultrason. 19 (2) (1972) 107–114. A.M. Glazer, N. Zhang, A. Bartasyte, D.S. Keeble, S. Huband, P.A. Thomas, I. Gregora, F. Borodavka, S. Margueron, J. Hlinka, J. Appl. Crystallogr. 45 (5) (2012) 1030–1037. R.G. Kepler, R.A. Anderson, Adv. Phys. 41 (1) (1992) 1–57. B. Neese, B. Chu, S.-G. Lu, Y. Wang, E. Furman, Q.M. Zhang, Science 321 (5890) (2008) 821. H.S. Nalwa, Ferroelectric Polymers: Chemistry, Physics, and Applications, M. Dekker, Inc., New York, 1995. E. Sawaguchi, J. Phys. Soc. Jpn. 8 (5) (1953) 615–629. S.K. Mishra, N. Choudhury, S.L. Chaplot, P.S.R. Krishna, R. Mittal, Phys. Rev. B 76 (2) (2007) 024110. S. Gupta, V. Petkov, S. Priya, Appl. Phys. Lett. 105 (23) (2014) 232902. D. Vanderbilt, M.H. Cohen, Phys. Rev. B 63 (9) (2001) 094108. A.M. Glazer, P.A. Thomas, K.Z. Baba-Kishi, G.K.H. Pang, C.W. Tai, Phys. Rev. B 70 (18) (2004) 184123. P.M. Woodward, Acta Crystallogr. Sect. B Struct. Sci. 53 (1) (1997) 44–66. A. Glazer, Acta Crystallogr. Sect. B 28 (11) (1972) 3384–3392. P.M. Woodward, Acta Crystallogr. Sect. B Struct. Sci. 53 (1) (1997) 32–43. B.N. Figgis, M.A. Hitchman, Ligand Field Theory and Its Applications, Wiley-VCH, New York, 2000 [2000]. F. Jona, G. Shirane, Ferroelectric Crystals, Pergamon Press, Oxford; New York, 1962. G. Arlt, J. Mater. Sci. 25 (6) (1990) 2655–2666. G. Arlt, H. Dederichs, R. Herbiet, Ferroelectrics 74 (1) (1987) 37–53. S. Gupta, S. Huband, D.S. Keeble, D. Walker, P. Thomas, D. Viehland, S. Priya, CrystEngComm 15 (34) (2013) 6790–6799. J.J. Liu, Y.C. Zhou, A.K. Soh, J.Y. Li, Appl. Phys. Lett. 88 (3) (2006) 032904. P. Paruch, T. Tybell, J.-M. Triscone, Appl. Phys. Lett. 79 (4) (2001) 530–532. A. Pramanick, A.D. Prewitt, J.S. Forrester, J.L. Jones, Crit. Rev. Solid State Mater. Sci. 37 (4) (2012) 243–275.

Introduction to ferroelectrics and related materials

39

[49] T. Hoshina, K. Takizawa, J. Li, T. Kasama, H. Kakemoto, T. Tsurumi, Jpn. J. Appl. Phys. 47 (9) (2008) 7607–7611. [50] S. Wada, K. Muraoka, H. Kakemoto, T. Tsurumi, H. Kumagai, Jpn. J. Appl. Phys. Part 1 43 (9B) (2004) 6692–6700. [51] R. Ahluwalia, T. Lookman, A. Saxena, W. Cao, Phys. Rev. B 72 (1) (2005) 014112. [52] S. Wada, A. Seike, T. Tsurumi, Jpn. J. Appl. Phys 40 (2001) 5690. [53] K. Aizu, J. Phys. Soc. Jpn. 28 (3) (1970) 706–716. [54] K. Aizu, Phys. Rev. B 2 (3) (1970) 754–772. [55] K. Aizu, Phys. Rev. 146 (2) (1966) 423–429. [56] A.K. Tagantsev, L.E. Cross, J. Fousek, Domains in Ferroic Crystals and Thin Films, Springer, 2010. [57] F.X. Li, R.K.N.D. Rajapakse, Acta Mater. 55 (19) (2007) 6472–6480. [58] F.X. Li, R.K.N.D. Rajapakse, Acta Mater. 55 (19) (2007) 6481–6488. [59] J.Y. Li, R.C. Rogan, E. Ustundag, K. Bhattacharya, Nat. Mater. 4 (10) (2005) 776–781. [60] G. Arlt, Ferroelectrics 76 (1) (1987) 451–458. [61] N.A. Pertsev, G. Arlt, A.G. Zembilgotov, Microelectron. Eng. 29 (1–4) (1995) 135–140. [62] G. Arlt, Presented at the Ultrasonics Symposium, 1990. Proceedings., IEEE 1990, 1990 (unpublished). [63] S. Gupta, S. Priya, Appl. Phys. Lett. 102 (1) (2013) 012906. [64] Q.M. Zhang, H. Wang, N. Kim, L.E. Cross, J. Appl. Phys. 75 (1) (1994) 454–459. [65] W.G. Cady, Piezoelectricity: An Introduction to the Theory and Applications of Electromechanical Phenomena in Crystals, Dover Publications, New York, 1964. [66] R.E. Newnham, Properties of Materials: Anisotropy, Symmetry, Structure, Oxford University Press, Oxford; New York, 2005. [67] S.P. Alpay, J. Mantese, S. Trolier-McKinstry, Q. Zhang, R.W. Whatmore, MRS Bull. 39 (12) (2014) 1099–1111. [68] Ferroelectric Devices, second ed., CRC Press, 2009, pp. 145–160. [69] C.R. Bowen, J. Taylor, E. LeBoulbar, D. Zabek, A. Chauhan, R. Vaish, Energy Environ. Sci. 7 (12) (2014) 3836–3856. [70] A.S. Bhalla, R.E. Newnham, Phys. Status Solidi A 58 (1) (1980) K19–K24. [71] J.L. Wentz, L.Z. Kennedy, J. Appl. Phys. 35 (6) (1964) 1767–1770. [72] R. Pirc, Z. Kutnjak, R. Blinc, Q.M. Zhang, J. Appl. Phys. 110 (7) (2011) 074113. [73] J.F. Scott, Annu. Rev. Mater. Res. 41 (1) (2011) 229–240. [74] X.-S. Qian, H.-J. Ye, Y.-T. Zhang, H. Gu, X. Li, C.A. Randall, Q.M. Zhang, Adv. Funct. Mater. 24 (9) (2014) 1300–1305. [75] G. Singh, I. Bhaumik, S. Ganesamoorthy, R. Bhatt, A.K. Karnal, V.S. Tiwari, P.K. Gupta, Appl. Phys. Lett. 102 (8) (2013) 082902. [76] L. Luo, X. Jiang, Y. Zhang, K. Li, J. Eur. Ceram. Soc. 37 (8) (2017) 2803–2812. [77] J. Karthik, L.W. Martin, Appl. Phys. Lett. 99 (3) (2011) 032904. [78] B. Asbani, J.-L. Dellis, A. Lahmar, M. Courty, M. Amjoud, Y. Gagou, K. Djellab, D. Mezzane, Z. Kutnjak, M.E. Marssi, Appl. Phys. Lett. 106 (4) (2015) 042902. [79] N. Novak, R. Pirc, Z. Kutnjak, Phys. Rev. B 87 (10) (2013) 104102. [80] A.S. Mischenko, Q. Zhang, J.F. Scott, R.W. Whatmore, N.D. Mathur, Science 311 (5765) (2006) 1270–1271. [81] T.D. Nguyen, S. Mao, Y.-W. Yeh, P.K. Purohit, M.C. McAlpine, Adv. Mater. 25 (7) (2013) 946–974. [82] P.V. Yudin, A.K. Tagantsev, Nanotechnology 24 (43) (2013) 432001. [83] P. Zubko, G. Catalan, A.K. Tagantsev, Annu. Rev. Mater. Res. 43 (1) (2013) 387–421. [84] W. Ma, Phys. Status Solidi B 247 (1) (2010) 213–218.

40

[85] [86] [87] [88]

[89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100]

[101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113]

Ferroelectric Materials for Energy Harvesting and Storage

L.E. Cross, J. Mater. Sci. 41 (1) (2006) 53–63. X. Jiang, W. Huang, S. Zhang, Nano Energy 2 (6) (2013) 1079–1092. W. Ma, L.E. Cross, Appl. Phys. Lett. 78 (19) (2001) 2920–2921. R. Steinhausen, T. Hauke, W. Seifert, V. Mueller, H. Beige, S. Seifert and P. Lobmann, Presented at the Applications of Ferroelectrics, 1998. ISAF 98. Proceedings of the Eleventh IEEE International Symposium on, 1998 (unpublished). G. Tan, K. Maruyama, Y. Kanamitsu, S. Nishioka, T. Ozaki, T. Umegaki, H. Hida, I. Kanno, Sci. Rep. 9 (1) (2019) 7309. W. Huang, S.-R. Kwon, S. Zhang, F.-G. Yuan, X. Jiang, J. Intell. Mater. Syst. Struct. 25 (3) (2013) 271–277. X. Yan, W. Huang, S. Ryung Kwon, S. Yang, X. Jiang, F.-G. Yuan, Smart Mater. Struct. 22 (8) (2013) 085016. W. Huang, X. Yan, S.R. Kwon, S. Zhang, F.-G. Yuan, X. Jiang, Appl. Phys. Lett. 101 (25) (2012) 252903. C.B. Sawyer, C.H. Tower, Phys. Rev. 35 (3) (1930) 269–273. M.G. Cain, Characterisation of Ferroelectric Bulk Materials and Thin Films, Springer, Dordrecht, 2014. S. Hong, Nanoscale Phenomena in Ferroelectric Thin Films, Kluwer Academic Publishers, Boston, 2004. Proc. IRE 45 (3) (1957) 353–358. Proc. IRE 46 (4) (1958) 764–778. D. Berlincourt, H.I. Metz, E. Mittelmann, L.H. Mwontgomery, S.M. Morrison, Proc. IRE 9 (1961) 1161. S.V. Kalinin, B.J. Rodriguez, S. Jesse, J. Shin, A.P. Baddorf, P. Gupta, H. Jain, D. B. Williams, A. Gruverman, Microsc. Microanal. 12 (03) (2006) 206–220. S.V. Kalinin, B.J. Rodriguez, S. Jesse, Y.H. Chu, T. Zhao, R. Ramesh, S. Choudhury, L. Q. Chen, E.A. Eliseev, A.N. Morozovska, Proc. Natl. Acad. Sci. U. S. A. 104 (51) (2007) 20204–20209. A. Gruverman, S.V. Kalinin, Frontiers of Ferroelectricity, Springer US, 2007, pp. 107–116. S.V. Kalinin, A. Gruverman, Scanning Probe Microscopy: Electrical and Electromechanical Phenomena at the Nanoscale, Springer, New York, 2007. W. Yousuke and S. Toshio, Advances in Scanning Probe Microscopy, Springer, 2000. S. Jesse, A.P. Baddorf, S.V. Kalinin, Appl. Phys. Lett. 88 (6) (2006) 062908. D.J. Inman, S. Priya, Energy Harvesting Technologies, Springer, New York, London, 2009. H. Huang, J.F. Scott, Ferroelectric Materials for Energy Applications, Wiley-VCH, Weinhaim, 2018. L. Wu, X. Wang, L. Li, RSC Adv. 6 (17) (2016) 14273–14282. P. Zhao, H. Wang, L. Wu, L. Chen, Z. Cai, L. Li, X. Wang, Adv. Energy Mater. 9 (17) (2019) 1803048. A. Kumar, S.H. Kim, M. Peddigari, D.-H. Jeong, G.-T. Hwang, J. Ryu, Electron. Mater. Lett. 15 (3) (2019) 323–330. Y. Liu, S. Wang, Z. Chen, L. Xiao, Sci. China Mater. 59 (10) (2016) 851–866. K.T. Butler, J.M. Frost, A. Walsh, Energy Environ. Sci. 8 (3) (2015) 838–848. S.K. Kurinec, Emerging Photovoltaic Materials: Silicon & Beyond, John Wiley & Sons, 2018. B.G. Streetman, S. Banerjee, Solid State Electronic Devices, Pearson/Prentice Hall, Upper Saddle River, N.J., 2006.

Introduction to ferroelectrics and related materials

41

[114] S.Y. Yang, L.W. Martin, S.J. Byrnes, T.E. Conry, S.R. Basu, D. Paran, L. Reichertz, J. Ihlefeld, C. Adamo, A. Melville, Y.-H. Chu, C.-H. Yang, J.L. Musfeldt, D. G. Schlom, J.W.A. III, R. Ramesh, Appl. Phys. Lett. 95 (6) (2009) 062909. [115] S.Y. Yang, J. Seidel, S.J. Byrnes, P. Shafer, C.H. Yang, M.D. Rossell, P. Yu, Y.H. Chu, J. F. Scott, J.W. Ager Iii, L.W. Martin, R. Ramesh, Nat. Nanotechnol. 5 (2010) 143. [116] H.S. Kim, J.-H. Kim, J. Kim, Int. J. Precis. Eng. Manuf. 12 (6) (2011) 1129–1141. [117] S. Adhikari, M.I. Friswell, D.J. Inman, Smart Mater. Struct. 18 (11) (2009) 115005. [118] F. Li, D. Lin, Z. Chen, Z. Cheng, J. Wang, C. Li, Z. Xu, Q. Huang, X. Liao, L.-Q. Chen, T. R. Shrout, S. Zhang, Nat. Mater. 17 (2018) 349–354. [119] Y. Yan, K.-H. Cho, D. Maurya, A. Kumar, S. Kalinin, A. Khachaturyan, S. Priya, Appl. Phys. Lett. 102 (4) (2013) 042903. [120] D. Maurya, Y. Zhou, Y. Yan, S. Priya, J. Mater. Chem. C 1 (11) (2013) 2102–2111. [121] S. Gupta, A. Belianinov, M. Baris Okatan, S. Jesse, S.V. Kalinin, S. Priya, Appl. Phys. Lett. 104 (17) (2014) 172902. [122] B. Narayan, J.S. Malhotra, R. Pandey, K. Yaddanapudi, P. Nukala, B. Dkhil, A. Senyshyn, R. Ranjan, Nat. Mater. 17 (5) (2018) 427–431. [123] G.W. Taylor, J.R. Burns, S.A. Kammann, W.B. Powers, T.R. Welsh, IEEE J. Ocean. Eng. 26 (4) (2001) 539–547. [124] Y. Tadesse, Z. Shujun, S. Priya, J. Intell. Mater. Syst. Struct. 20 (5) (2008) 625–632. [125] H. T€oreyin, E. Topal, H. K€ulah, Proc. Eng. 5 (2010) 1176–1179. [126] Z. Yang, J.W. Zu, L. Guo, IEEE Trans. Magn. 48 (11) (2012) 3344–3347. [127] S. Dong, J. Zhai, J.F. Li, D. Viehland, S. Priya, Appl. Phys. Lett. 93 (10) (2008) 103511. [128] S. Chul Yang, K.-H. Cho, C.-S. Park, S. Priya, Appl. Phys. Lett. 99 (20) (2011) 202904. [129] A. Singh, V. Pandey, R.K. Kotnala, D. Pandey, Phys. Rev. Lett. 101 (24) (2008) 247602. [130] M.M. Kumar, A. Srinivas, G.S. Kumar, S.V. Suryanarayana, J. Phys. Condens. Matter 11 (1999) 8131. [131] M. Feng, J.-j. Wang, J.-M. Hu, J. Wang, J. Ma, H.-B. Li, Y. Shen, Y.-H. Lin, L.-Q. Chen, C.-W. Nan, Appl. Phys. Lett. 106 (7) (2015) 072901. [132] A. Singh, A. Gupta, R. Chatterjee, Appl. Phys. Lett. 93 (2) (2008) 022902. [133] Y.Y. Liu, R.K. Vasudevan, K. Pan, S.H. Xie, W.I. Liang, A. Kumar, S. Jesse, Y.C. Chen, Y.H. Chu, V. Nagarajan, S.V. Kalinin, J.Y. Li, Nanoscale 4 (10) (2012) 3175–3183. [134] G.G. Guzma´n-Verri, P.B. Littlewood, APL Mater. 4 (6) (2016) 064106. [135] G. Sebald, E. Lefeuvre, D. Guyomar, IEEE Trans. Ultrason. Ferroelectr. Freq. Control 55 (3) (2008) 538–551. [136] A. Cuadras, M. Gasulla, V. Ferrari, Sens. Actuators A 158 (1) (2010) 132–139. [137] S. Pandya, G. Velarde, L. Zhang, J.D. Wilbur, A. Smith, B. Hanrahan, C. Dames, L. W. Martin, NPG Asia Mater. 11 (1) (2019) 26.

Solar energy harvesting with ferroelectric materials

2

Abhishek Kumara, Somdutta Mukherjeeb, and Amritendu Roya a School of Minerals, Metallurgical and Materials Engineering, Indian Institute of Technology, Bhubaneswar, Orissa, India, bMaterials Chemistry Department, CSIR-IMMT Bhubaneswar, Bhubaneswar, Orissa, India

2.1

Introduction

With depleting reserves of fossil fuels and growing environmental concerns over global warming, search for renewable and clean energy technologies has been one of the key areas of interest for the survival of human civilization in coming centuries. This is reflected by the sharp increase in the number of research articles in this area over the last decade or so (Fig. 2.1). Research in various avenues, such as solar, wind, river and ocean current, geothermal, and so forth, demonstrated feasibility of generation of electricity from diverse resources available in nature which otherwise remain unused. Perhaps, the most promising of these is the solar energy. Sun, thanks to the nuclear fission reaction within it, produces a gigantic amount of energy and radiates, approximately, 1.74
1017 W of solar power to earth alone [1]. If converted to electrical energy, it is estimated that the requirement of human civilization for an entire year could be fulfilled in about 1 h [1]. Efficient means to transform solar energy to usable electrical energy has been the subject of scientific research in the field of renewable energy harvesting. The mechanism to transform solar energy to usable electrical energy depends upon the characteristic of the device for the conversion process and the material used in it. Solar energy absorbed by a device can either raise the kinetic energy of the atoms and electrons (internal energy) of the material in the absorbed layer or can increase the potential energy of the electrons or may be a combination of both. The prevalence of a particular mechanism depends upon the absorbing material and how it is connected to the external world. Accordingly, solar energy can be converted to usable electrical energy through at least three distinct technologies: (a) Solar thermal, (b) Solar Photovoltaic, and (c) Photochemical energy conversion. Solar thermal technology uses the entire range of solar wavelengths, including the infrared and ultraviolet regimes, wherein the material in the device is chosen to be heated up easily. In a solar thermal converter, the absorbed solar energy normally increases the internal energy vis-a`-vis the temperature of the device. The device is thermally insulated from the atmosphere and provides a sizeable temperature difference. The resulting temperature difference allows the device to function as a heat engine and to do work, for example, to rotate a steam turbine or to run a thermoelectric Ferroelectric Materials for Energy Harvesting and Storage. https://doi.org/10.1016/B978-0-08-102802-5.00002-9 © 2021 Elsevier Ltd. All rights reserved.

44

Ferroelectric Materials for Energy Harvesting and Storage

7000

No. of research papers

6000 5000 4000 3000 2000 1000 0 1990

1995

2000

2005

2010

2015

Year

Fig. 2.1 Year-wise publications of original research articles with the key word “renewable energy”. Data collected from SCOPUS.

and/or a pyroelectric converter to generate electricity. A number of ferroelectric systems demonstrate efficient thermoelectric as well as pyroelectric conversion behavior [2–4]. The concurrent physics and other technical details of some solar thermal conversion would be discussed in other chapters of the present book. Unlike solar thermal conversion, solar photovoltaic technology relies on increase in the potential energy of the electrons upon absorption of solar radiation. In PV systems, solar energy raises the electrochemical potential of the system. Electrons in the occupied states, upon absorption of photons with requisite energy, migrate to excited states at higher energies. In order to minimize the total energy of the system, the excited electrons tend to emit the additional energy and return to lower energy states. To prevent (or reduce) such spontaneous relaxation of the excited electrons, the excited electronic state needs to be separated from the ground state by an energy gap. This energy gap should be larger than the thermal vibrational energy, KBT, where KB is the Boltzmann’s constant and T is temperature in the absolute scale. Therefore, for PV material, the primary requirement is that: (i) the energy band structure should contain two or more energy bands, and (ii) these bands should be separated by energies greater than KBT. Thus, semiconductors became the potential class of materials for PV technology. In semiconductors, the gap between the valence and conduction bands serves as the buffer between the excited electrons with higher energies and the ground-state electrons in the valence band for sufficient duration before they are collected. In a two band system, the difference between the excited-state and ground-state electrons is expressed in terms of Gibb’s free energy, N Δ μ, where N is the number of electrons promoted to a particular energy band and Δ μ is the electrochemical potential difference between the particular excited state and the ground state. Δ μ results from the absorption of photons by the electrons in the ground state and is often termed as

Solar energy harvesting with ferroelectric materials

45

chemical potential of radiation [5]. The electrochemical potential energy extraction is most efficient when the ground state is fully occupied and the excited state is completely empty. This is the primary reason why semiconductors are the most suitable materials for PV technology. It is important to note that, in solar photovoltaic, only those photons of solar radiation are absorbed which have energy greater than the bandgap of the material. Photovoltaic effect (PV) in ferroelectric material was discovered in 1970s [6–8]. Anomalous PV effect was reported in ferroelectrics with photovoltage reaching tens of multiples of the bandgap of the material, unlike the classical PV materials, where the observed photovoltage is a fraction of the bandgap of the material [9]. However, the observed current in PV-Ferroelectrics is far too small to yield a noticeable conversion efficiency [10, 11]. It was explained and quite rightly so that ferroelectrics with large bandgap are not capable of absorbing solar spectra efficiently, resulting in generation of fewer electron-hole pairs. However, with modern design methodologies, bandgap-engineered ferroelectrics have been found to yield as high as 8% conversion efficiency [12, 13]. Section 2.2 discusses the current state of the art of photovoltaic technology based on ferroelectrics along with required physics. Photochemical conversion of solar energy is similar to photovoltaic conversion in the sense that both the processes involve quantum energy conversion, i.e., the chemical potential of the systems is raised upon absorption of solar energy [14]. However, the crucial difference lies in the fact that, unlike PV process, a photochemical conversion increases the chemical potential permanently rather than generating electric power. The photoexcited electrons in PC conversion drive a chemical reaction, for instance, dissociation of water into hydrogen and oxygen instead of driving an electric current in the external circuit. We discuss the relevant technology in Section 2.3.

2.2

Solar photovoltaics

As discussed, solar photovoltaic (PV) is a single-step conversion technology wherein light energy is converted into electrical energy. When light is incident on a material, the electrons within the material absorb the quanta of light energy, viz., photons, and migrate to the higher energy states. However, the excited electrons, in general, very soon lose their energy and relax back to their ground state. A photovoltaic device (Fig. 2.2) prevents the relaxation of electrons. A built-in asymmetry in PV devices drives the electrons from the higher energy states before they can relax and direct them to an external circuit. The additional energy of the excited electrons creates an electromotive force that drives the electrons into the electrical circuit to perform work. The efficacy of the PV device depends on the selection of the light-absorbing material and the electrical design of the device. Initial research in photovoltaics was conducted on single crystalline silicon. Further research continued in Group II–VI and Group III–V semiconductors [15]. Due to the advances in silicon technology in the microelectronics industry, silicon-based photovoltaic solar cell has remained the most prominent photovoltaic material, which already witnessed large-scale commercialization.

46

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 2.2 Schematic of a conventional p-n junction solar cell.

Even after such prolong research and development, commercial silicon-based solar cells are still expensive in comparison to fossil fuel-based electricity production [15]. In this regard, to reduce production cost and improve efficiency, second- and third-generation PV cell technology with amorphous silicon thin-film solar cells, dye-sensitized solar cells, quantum dot solar cells, organic solar cells, and perovskite solar cells are under extensive study to cut the material and production costs and improve efficiency. However, stability of the materials used during service and their environmental impact often raises questions about the future of these new generation PV materials [16, 17]. Ferroelectrics, especially oxide-based ferroelectrics, are exciting class of materials with novel photovoltaic properties. When a ferroelectric material is connected to two electrodes in an M-FE-M configuration and illuminated, based on the bandgap, the ferroelectric material absorbs solar energy and as a result photoexcited charge carriers are formed. Noncentrosymmetric distribution of the atoms within the ferroelectric unit cell leads to unequal transition probabilities of electronic jump from a state with momentum k to k0 and back. Such asymmetric momentum distribution of the photogenerated charge carriers translates to a steady photocurrent. Macroscopically, it is believed that ferroelectric polarization-induced internal electric field is capable of separating the photogenerated charge carriers, giving rise to the generation of built-in electric field across the ferroelectric/electrode interface. Certain ferroelectrics are known to develop large photovoltages [6]. The observed photovoltage is an order of magnitude larger than the bandgap of these ferroelectrics. However, the photocurrent developed is, in general, very small. Consequently, the power conversion efficiency for these materials is insignificant in comparison to traditional photovoltaic materials. Recent works, however, demonstrated tangible improvement in the conversion efficiency owing to much lowered bandgap in band-engineered ferroelectrics and multiferroic materials. On the basis of the above success, low cost, stable, and environment-friendly next-generation PV technology

Solar energy harvesting with ferroelectric materials

47

with sizeable conversion efficiency based on band-engineered ferroelectric and multiferroic materials may possibly be realized for practical applications in coming years. However, such a technological development would require an in-depth understanding of the materials as well as device physics triggering ferroelectric photovoltaic phenomena. In this connection, it is crucial to begin with the review of fundamental physics relevant to photovoltaics.

2.2.1 Fundamentals of physics of solar photovoltaics A solar photovoltaic cell can be described as a two terminal device which behaves like a diode in dark and generates a photovoltage when exposed to solar radiation (Fig. 2.2). Under illumination, a traditional semiconductor PV cell is capable of producing a small photovoltage of 0.5–1.0 V and a decent photocurrent of few tens of milliamperes per cm2 [18]. To generate useful voltage, the cells are joined in series. To prevent loss, antireflection coating is applied on the cell which makes the cell appear dark blue or black. To understand the working principles of solar PV vis-a`-vis materials designing for such application, one needs to understand the physics of the technology. In this regard, the first thing to understand is the solar spectrum.

2.2.1.1 The solar spectra It is observed that solar irradiance (radiant energy from sun per unit area per unit time), at a point outside earth’s atmosphere, is highest in the visible range, λ  300–800 nm. The solar radiation in the outer atmosphere can be modeled as a black body radiation at 5760 K. The energy emitted by a black body with a characteristic temperature, Ts, follows a certain distribution. The normal component of spectral photon flux with energy E radiated from the surface of a black body is given by [18]: 2 Ibb ðEÞ ¼

3

2 7 2FS 6 6
E 7 4 5 3 2 E h c e kB T S  1

(2.1)

where FS is a geometrical factor, FS ¼ πSin2θsun, wherein θsun is the half angle subtended by the blackbody to the point where the flux is measured. As seen from the earth, θsun ¼ 0.26o. In the above expression, the temperature of the blackbody is considered to be the same, TS. At the surface of sun, the emitted power density is estimated to be 62 MW/m2, while at a point just outside the earth’s atmosphere, the solar radiation is reduced to 1353 W/m2 [18]. When solar radiation passes through the earth’s atmosphere, a fraction of the same gets absorbed. Particles in the atmosphere further scatter part of the light. As a result, the solar spectrum reaching the surface of the earth are both attenuated as well as changed in shape. Infrared frequencies are absorbed by atomic and molecular oxygen, carbon dioxide, and water molecules creating dips in

48

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 2.3 When sun is at angle of elevation, γ Sun, the light from sun travels a distance, d
Cosec γ Sun,to reach the earth’s surface, where the depth of the atmosphere is d. The optical depth is expressed in terms of ηAir Mass.

the solar spectrum that is received at earth’s surface. Attenuation of solar spectrum by atmosphere is quantified by “Air Mass” defined as: ηAir Mass ¼

Optical Path length to Sun ¼ cosec γ sun Optical path length if Sun is directly overhead

where γ sun is the angle of elevation of the sun as shown in Fig. 2.3. The standard spectrum, Air Mass 1.5 or AM 1.5, corresponds to the sun being at an angle of elevation of 42° (Fig. 2.4). AM 1.5 corresponds to a mean irradiance of 900 W/m2.

2.2.1.2 Open circuit voltage, short-circuit current, quantum efficiency, and fill factor In the two terminal photovoltaic cell (Fig. 2.5), when the terminals are isolated, the photovoltage generated under illumination is called the open circuit voltage, Voc. When the terminals are shorted, the current drawn from the device is termed as “short-circuit current,” Isc. Under load, the device develops an intermediate voltage (V) between 0 and Voc and delivers a current, I, such that I ¼ RVL  ISC . The current is further dependent on the current-voltage characteristics of the device under illumination. Since the current is approximately proportional to illuminated surface area of the device, short-circuit current density (JSC) would be the more appropriate term for quantitative evaluation of a PV cell. The photocurrent density at short-circuit, under illumination (JSC), is a function of the incident light and the materials property of the absorber layer of the PV device ð JSC ¼ e ISun ðEÞηQE ðEÞdE

(2.2)

where e is the charge on electrons, ISun(E) is the incident solar photon flux density, i.e., the number of photons with energy in the range of E and E + dE, which falls on unit

Solar energy harvesting with ferroelectric materials

49

1.8

Solar spectral irradiance, W/m2· nm

1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 0

500

1000

1500

2000

2500

3000

3500

4000

wavelength (λ) (nm)

Fig. 2.4 Solar irradiance plotted as a function of wavelength for AM 1.5 G solar spectra (ASTM G-173-03).

Fig. 2.5 Equivalent circuit of ideal solar cell.

area per unit time. ISun(E) is an important parameter in designing efficient PV cells. Above the atmosphere of earth, the solar radiation intensity is 1.353 kW/m2 [18]. As the solar radiation traverses through the earth’s atmosphere, it gets absorbed, partially. As mentioned above, the air mass is a measure of how the atmosphere absorbs the spectral content and intensity of the solar radiation reaching the earth’s surface. Air mass is defined as ηAir Mass ¼ cosec γ sun; where γ sun is the angle of incidence. γ sun ¼ 90° when sun is directly overhead. A widely used standard for comparing solar cell performance is the AM 1.5 spectrum normalized to a total power density of approximately 1 kW/m2. ηQE(E) is the quantum efficiency of the cell, defined as the probability that an incident photon of energy E would lift an electron to the external circuit. ηQE(E) is dependent on the absorption coefficient of the absorber material

50

Ferroelectric Materials for Energy Harvesting and Storage

of the PV cell as well as efficiency with which the charges are separated and collected in the device. Thus, ηQE(E) is one of the key material properties in designing and selection of PV absorber material. It should be noted that, in photovoltaics, the sign convention of current and voltage is such that the photocurrent is positive. When there is a load in the external circuit, a potential difference is developed across the two terminals of the cell. As a result, a current (current density) flows in the opposite direction to the photocurrent. The net current density through the circuit is thus reduced from the JSC value. The reverse current density under a bias voltage, V, that flows through the circuit is called the dark current density, Jdark(V). A defining feature of a PV cell is its rectifying nature. In dark, it allows a large current flow through the circuit under forward bias (V > 0) in comparison to reverse bias (V < 0). Such asymmetric diode behavior is the basis of separation of the photogenerated charge carriers. The dark current density, Jdark(V), for an ideal diode can be expressed as:
eV  Jdark ðV Þ ¼ JO eKB T  1

(2.3)

where KB is the Boltzmann’s constant, T is the temperature in degrees Kelvin, and JO is a constant. The overall current-voltage behavior of the PV cell can be approximately expressed as the sum of the short-circuit photocurrent and the dark current, J ðV Þ ¼ JSC  Jdark ðV Þ

(2.4)

For an ideal diode PV cell,
eV  K T J ðV Þ ¼ JSC  JO e B  1

(2.5)

When the load in the external circuit is withdrawn, the potential difference across the two terminal is maximum and is termed as the open circuit voltage, VOC  . The above definition allows us to describe the VOC for an ideal diode, VOC ¼ KeB T ln

JSC JO

+ 1 . The

operational range of the PV cell is from 0 to VOC, over which the cell delivers power. The power density of the cell is given by P ¼ JV. P reaches its maximum at a certain voltage, Vm, and the corresponding current density is Jm, as schematically shown in the Fig. 2.6. The fill factor defined as, FF ¼ Jm
Vm/(JSC
VOC), represents the squareness of the J-V curve shown in Fig. 2.6. The conversion efficiency of the cell, η, is formally defined as the ratio of the power density delivered by the cell at the operating point and incident light power density: ¼ Jm
Vm/PSun ¼ JSC
VOC
FF/PSun; the four quantities in the above equation, JSC,VOC, FF, and η, are the fundamental performance indicators of a PV cell. Ideal diode behavior assumed above is rarely observed in real PV cell. Therefore, to account the non-ideality from the diode behavior, ideality factor, m, is introduced in the current density equation,

Solar energy harvesting with ferroelectric materials

51

Fig. 2.6 Representative current-voltage (blue) characteristics of an ideal cell.


eV  mK T B J ðV Þ ¼ JSC  JO e 1

(2.6)

It should be noted that electrical losses due to contact resistance of the metal electrodes and leakage current through the edges of the cell modify the current density expression.

2.2.1.3 Factors affecting the performance of a conventional solar cell The fundamental quantities related to the performance of a photovoltaic cell have been identified as JSC,VOC, FF, and η. Photocurrent is the outcome of photon absorption by the absorber layer, resulting in migration and consequent collection of electrons. If the migrated electron has a probability, ηc(E), to be collected, the photocurrent density at short-circuit, obtained from the detailed balance ð JSC ¼ e ISun ðEÞηc ðEÞf1  RðEÞgaðEÞdE

(2.7)

where a(E) is the probability of absorption of a photon of energy, E, and R(E) is the probability of photon reflection. It is seen that Eqs. (2.1) and (2.6) are identical where quantum efficiency, ηQE(E), is expressed as the product of absorption and collection efficiency. In the case of an ideal solar cell material with perfectly absorbing as well as nonreflecting character, each incident photon with energy, E > Eg is absorbed to promote one electron migrating from the valence band to the conduction band. Multiple carrier generation, i.e., promotion of more than one electron by a single absorbed

52

Ferroelectric Materials for Energy Harvesting and Storage

photon, is neglected. A situation allowing perfect charge separation, i.e., all migrated electrons survive radiative recombination and collected at the negative terminal of the solar cell, further yields maximum photocurrent and  ηQE ðEÞ ¼ aðEÞ ¼

1 E  Eg 0 E < Eg

As a result, JSC ¼ e

ð∞

ISun ðEÞdE

(2.8)

Eg

Eq. 2.8 predicts that photocurrent is only a function of the bandgap and the incident solar radiation in an ideal scenario. It also suggests that, with lower Eg, the JSC will be greater. Eq. 2.8 also suggests that the solar radiation needs to be standardized in order to define any statement of solar conversion efficiency. Based on the above, it can be stated that the power conversion efficiency of the ideal two band photo-converter is only a function of bandgap, Eg, and the incident solar spectrum. For a standard solar radiation (AM 1.5 solar spectrum), the conversion efficiency, η ( ¼ V
J(V)/ Ð∞ EI (E)dE ), is therefore only a function of bandgap of the material. In the present 0 Sun theory, VOC is always smaller than Eg. At small Eg, η is small due to small V; Eqs. (2.5) and (2.7) further suggest that at large values of Eg, J(V) will be small leading to small η. For AM 1.5 solar spectrum, the optimum bandgap for which η maximizes (33%) is 1.4 eV. Therefore, optimization of the performance of an ideal solar photovoltaic cell lies basically on the judicious selection of the absorbing material. The entire analysis presented above is based on the following important assumptions: (i) The PV material possesses completely filled valence band and an unoccupied conduction band separated by an energy bandgap, Eg. (ii) All incident photon with energy, E  Eg, are absorbed. (iii) Each absorbed photon excites one electron resulting in one electron-hole pair. (iv) Excited electron-hole pairs do not recombine except radiatively. Excited electron-hole pairs are completely separated. (v) Electrons and holes are transported to the external circuit without loss.

In real systems, deviations from ideality are common. However, it is desired that real materials would be close to the ideal situation. Therefore, design criteria of materials toward efficient PV solar cell application, in general, include: (a) Optimum bandgap close to 1.4 eV. (b) Large absorption coefficient, so that absorption of light with E  Eg can be ensured. With increasing thickness of the sample, most semiconductors are able to demonstrate perfect absorption with few tens or hundreds of microns thick films. With increasing thickness, the material consumption, production cost and overall economy of the technology go up. Further, increasing thickness of the material increases the concentration of various types of defects and phenomena associated with them. However, with large absorption coefficient, the thickness can be reduced to an optimum level reducing the materials and fabrication cost.

Solar energy harvesting with ferroelectric materials

53

(c) Long carrier diffusion length which translates to long recombination lifetime and small electron-hole binding energy. (d) High carrier mobility to ensure that the photogenerated charge carriers are quickly transported and collected at the electrodes.

2.2.2 Photovoltaics with ferroelectrics While certain parameters for designing photovoltaic materials laid down in Section 2.2.1.3 are valid for all materials, the basic mechanism for conventional photovoltaics based on silicon technology (Fig. 2.7A) is fundamentally different from that envisaged for photovoltaic effect in ferroelectrics (Fig. 2.7B and C). While the photogenerated carriers are separated by the potential generated by asymmetry of the p-n junction and the principle of detailed balance is generally valid, in the latter case the charge separation is presumed to take place due to the asymmetry offered by the noncentrosymmetric polar nature of the ferroelectric materials, which violates the principle of detailed balance. Early work on ferroelectric photovoltaics speculated that the inherent electric field due to the spontaneous polarization of ferroelectrics not only separates out the photogenerated excitons into free charges of opposite polarities, but transports the free charges into opposite directions to reduce the recombination rates. Since the spontaneous polarization of a ferroelectric is a switchable property, PV response in these materials is supposedly tuned by tailoring the polarization of the ferroelectrics [19]. In fact, the photovoltage in a FE-PV device, as shown in Fig. 2.7B and C, depends on a host of factors including electrical conductivity [20], spontaneous polarization [21], crystallographic orientation [22], size of ferroelectric domains and domain wall character of the ferroelectric absorber layer [23, 24], incident light intensity [19], interelectrode separation [25], and nature of ferroelectric-electrode interface [26]. However, majority of the above studies report a small photocurrent and poor conversion efficiencies in traditional ferroelectrics. This has impacted the research interest on ferroelectrics for PV application beyond the academia. Recent discovery of organic-inorganic halide hybrid perovskites with polar crystal symmetry yielding very high photocurrent and conversion efficiency has resulted in new a focus area of study for ferroelectric photovoltaic materials [27, 28]. Extraordinary PV performance of halides, indeed, demonstrates the potential of polar materials in designing PV materials and devices [27, 28]. While the large photocurrent and high PV efficiency of perovskite halides has been attributed to favorable optoelectronic properties together with bulk photovoltaic effect [29], anomalous photovoltage output in traditional FE-PV devices for certain ferroelectrics has been variously explained in terms of asymmetric scattering of the charge carriers, nonlinear dielectric model, and shift-current model [20, 30, 31]. However, shift-current mechanism has been most successful in consistently explaining the photovoltaic effect in ferroelectric materials [32]. The fundamental feature of all the above models is that the output photovoltage is generated in the bulk of the material. In contrast, the other mechanism emphasizes on the critical role of ferroelectric domain walls stacked in tandem, to explain the origin of observed photovoltaic effect [24].

54

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 2.7 Basic functioning of (A) conventional p-n junction solar cell, (B) and (C) ferroelectric solar cell.

2.2.2.1 Bulk photovoltaic effect Observation of photovoltages in pristine single crystal ferroelectrics under illumination is variously termed as bulk photovoltaic effect (BPE), photogalvanic effect, or nonlinear photonics [20]. Single crystal BaTiO3 was the first system to be reported to demonstrate steady state photovoltage with the magnitude of the photocurrent

Solar energy harvesting with ferroelectric materials

55

depending on the magnitude and direction of the spontaneous polarization of the system [7, 8]. Similar effects were subsequently reported in LiTaO3 [33], LiNbO3 [6, 34], and more recently in BiFeO3 [24, 26]. Observation of BPE is the direct violation of the principle of the detailed balance which suggests that the probability, φkk0 , of electron transition from a state of momentum k0 to a state with momentum k is equal to the reverse transition: φkk0 ¼ φk0 k. In centrosymmetric crystals, the above equality holds and the observed PV effect is the consequence of device asymmetry, which is discussed in Section 2.2.1. In noncentrosymmetric systems, however, φkk0 6¼ φk0 k; such inequality of the transition probability results in asymmetric momentum distribution of the photogenerated electrons (and holes) yielding a steady state photocurrent in homogeneous crystals under uniform illumination [20]. A number of mechanisms have been proposed to explain the microscopic origin which may independently or collectively contribute to the bulk photovoltaic effect in noncentrosymmetric crystals. These are listed below. (i) elastic scattering of the charge carriers from asymmetric scattering centers giving rise to asymmetric scattering, resulting in asymmetric momentum distribution, and therefore, nonzero photocurrent (Fig. 2.8A). (ii) Excitation of absorbing centers having asymmetric potential well, and therefore, unequal momentum distribution. This is explained in Fig. 2.8B. Upon incidence and subsequent absorption of light, the electrons are excited from the ground state to energies proportional to the energy of the incident light. If the electron energy in the excited state is less than the barrier height of the potential well, the electrons remain trapped. Absorption of light, corresponding to such situation, is not favorable.

k k′

–k

k –k

–k

–k′

k

(A)

(B)

Wex (–R)

Wex (R)

(C)

R

Fig. 2.8 Schematic representation of elementary processes of asymmetry, (A) Scattering, (B) photoexcitation from asymmetric potential well, and (C) interstate transition of photoexcited carriers.

56

Ferroelectric Materials for Energy Harvesting and Storage

However, if the potential well is asymmetric such that the barrier heights in either side of the well (in a 1D case) are different ϕ1 and ϕ2; ϕ1 6¼ ϕ2 , then it is possible for an electron to absorb light with energy, ϕ1 < E < ϕ2. In such cases, the excited electron will naturally move in the direction along which the barrier height is less than ϕ1. Nevertheless, possibility of electron tunneling cannot be ruled out along the other direction where the barrier width is small. (iii) Hopping of asymmetrically distributed charges in the crystals under illumination resulting in a net photocurrent. (iv) Shift current originating from the asymmetry of the electronic wavefunctions. Noncentrosymmetric materials with long range magnetic order, in addition, may give rise to magneto-optic effect-driven photovoltaic mechanism. Spin-orbit splitting of the valence band can give rise to asymmetric excitation of electrons in presence of circularly polarized light, and thus, net photocurrent is yielded [20]. Further, changing the direction of circular polarization of the incident light can change the direction of photocurrent. Independent of the above microscopic mechanisms, the BPE can be explained in terms of distribution of nonthermalized electrons. While for centrosymmetric crystals the momentum distribution of the photoexcited electrons in the conduction band is symmetric, noncentrosymmetric crystals provide asymmetric distribution. The resulting photocurrent density can be expressed as: ji ¼ βijk pj p∗k I0

(2.9)

Here, I0 represents the intensity of incident (absorbed) light, pn is the polarization of the medium in n-direction, and βijk, the photoferroic coefficient, which is a third-rank tensor and can be written as: βijk ¼

eδk χφ ħω

(2.10)

where δk is the mean free path of the photoexcited charge carriers comparable to the bulk exciton-bohr radius, e is the electronic charge, χ is the excitation asymmetry, φ is the quantum yield, and ħω is the photon energy. Photoconversion efficiency can be further expressed as: η ¼ βijkE where the electric field (E) created by the flow of ji is given by, E ¼ ji/σ d + σ pv, in which σ pv is the photoconductivity and σ d is the dark conductivity of the material [35]. The physical manifestation of photoferroic effect on the dielectric and ferroelectric nature of the material is complex. For instance, spontaneous electric polarization, which is attributed to the separation of photogenerated charges, is reduced in presence of increased concentration of photogenerated charge carriers due to enhanced screening effect. Large concentration of photogenerated charge carriers is further responsible for structural deformation, lowering of Curie temperature, and modification of effective permittivity. Such interdependence of physical parameters renders complexity to the mechanism of the observed photovoltaic effect in ferroelectric materials. The phenomenological treatment of BPE typically predicts a conversion efficiency limit, based on the lifetime of thermalized 12 s and nonthermalized electrons, η ¼ ττnon ¼ 10  106 : Thus, recent observation of 106 s th

Solar energy harvesting with ferroelectric materials

57

higher quantum efficiency in thin-film ferroelectric BiFeO3 photovoltaic cell may possibly have contributions not considered in the macroscopic model [24, 36]. By far, shift-current mechanism [30, 37] has been most consistent explanation for the origin of bulk photovoltaic effect (BPVE) in perovskite halides and ferroelectric oxides such as BaTiO3 [38], and BiFeO3 [39]. Generation of shift current is a second-order nonlinear optical phenomena [30]. The model suggests that the position of the electron wave packet shifts in the real space upon interband optical transition subsequent to absorption of the photon. The shift vector is a function of the difference of the Berry connection of the Bloch functions of the corresponding bands. The shift vector assumes a nonzero value upon collapse of spatial inversion symmetry. While large spontaneous polarization in ferroelectrics need not necessarily lead to large shift current vis-a`-vis PV response, polarization in ferroelectrics and shift current have close correlation. According to the modern theory of polarization [40], the total spontaneous polarization in a ferroelectric has two components. While ionic displacement within the ferroelectric phase constitutes the ionic component (Pion), asymmetry within the wavefunction forming partial covalent bonding gives rise to the electronic polarization (Pel). According to the modern theory of polarization, Pel can be calculated by integrating the Berry phase over the valence state. The shift vector is also connected to the Berry phase and it has the form of difference in Pel between the valence and conduction bands upon optical transition. Thus, a material possessing large Pel is expected to demonstrate large shift current, and therefore, enhanced PV response [41]. It has been further demonstrated that materials with polar rhombohedral and orthorhombic symmetries demonstrate better shift-current response in comparison to other symmetries [32].

2.2.2.2 Ferroelectric domain wall model Yang et al. [24] studied BiFeO3 ferroelectric films with ordered domain strips having two different device configurations as shown in Fig. 2.9A and B, to examine the photovoltaic effect in the system. It was observed that the magnitude of photovoltage in

Fig. 2.9 Schematics of the (A) perpendicular domain wall and (B) parallel domain wall-based ferroelectric photovoltaic devices.

58

Ferroelectric Materials for Energy Harvesting and Storage

the film increased constantly with increase in the total number of domain walls between the electrodes along the direction of net polarization [24]. However, the device with perpendicular geometry showed very poor photovoltage response; therefore, it was proposed that anomalous photovoltage ought to be mainly because of domain walls [24]. It was also suggested that there is no photovoltaic effect along the perpendicular direction to the direction of net polarization. It was concluded that each domain wall serves as nanogenerator, which are connected in series, such that the produced photovoltages accumulate toward the direction of net polarization and the photocurrent is continuous. This theory is akin to the principle of the tandem solar cell, in which the overall generated photovoltage is represented by the sum of the voltages of each cell. However, few studies also suggest that the domain wall is responsible for the source of current, whereas the total output voltage was calculated considering the conductivity of the ferroelectric film and the electrode spacing [42]. It was proposed that the anomalous photovoltaic effect is due to the exciton, which is created inside the domain, and the bulk photovoltaic effect was insignificant because of rapid recombination of the excitons generated outside the domain wall. However, Alexe et al. analyzed BiFeO3 single crystal with a photoelectric atomic force microscopy and piezoresponse atomic force microscopy and observed practically equivalent photocurrent inside and outside the domain wall representing weak recombination of the excitons in the bulk of the BiFeO3 domains [43]. According to the above domain wall theory, photocurrent is independent of the direction of the polarization, although a few studies demonstrated the dependence of photocurrent on the direction of light polarization in BiFeO3 film [13, 44]. First-principles study uncovered the origin of the zero photocurrent in the direction parallel to the ordered domain wall; it is predominantly because of rare geometry such that the bulk photovoltaic effect in each domain wall was counterbalanced by the neighboring domains [39]. It was additionally showed that the photocurrent due to the bulk photovoltaic effect and domain wall effect was neutralized by each other.

2.2.2.3 Schottky-junction effect According to this model, presence of Schottky contact at the interface of ferroelectric semiconductor and metal electrode leads to the formation of photocurrent under illumination due to the local electric field caused by the band bending close to metal electrodes. The generated photocurrent is a function of the depth of the depletion region and the Schottky barrier height (SBH) [45]. The generated photovoltage due to Schottky-junction effect is equivalent to the bandgap of the FE semiconductor material. However, the photovoltage reported here is very small as compared to the reported anomalous photovoltage (APV) in bulk FE materials. This effect is obviously observed in ferroelectric photovoltaics where generated photovoltage is of small magnitude. In general, photocurrent due to Schottky contact is negligible in ferroelectric photovoltaic devices having similar electrodes due to inverse polarization in the two Schottky-junctions. In general, the effect does not depend on the direction of the polarization; however, there are few conditions where photovoltage of the Schottkyjunction ferroelectric photovoltaic devices switch with change in the direction of

Solar energy harvesting with ferroelectric materials

59

spontaneous polarization. Yi et el. at first affirmed that it was because of the bulk photovoltaic effect in the BiFeO3 film; however, then they uncovered that the BiFeO3/Au Schottky contacts switched between Schottky contact and Ohmic contact at the time of poling operation because of the transfer of the oxygen vacancies [46].

2.2.2.4 Depolarization field model Poled FE films that are not screened have high surface charge density, and therefore, prompt a colossal electric field in the ferroelectric film. Ferroelectric, BaTiO3 with a polarization of 26 μC/cm2 have an internal electric field of 300 MV/cm when unscreened [47]. The presence of electrodes (metallic or semiconducting) on the surface of ferroelectric thin films causes inadequate screening of polarization surface charges by the charges in the electrodes. Normally, semiconducting electrodes lead to a higher depolarization field than metallic electrodes, whereas superconducting electrodes tend to completely compensate the depolarization field. For most part, inadequate screening of the surface charge is because of the mismatch between the centers of the surface charge and the free compensation charge. Electric field developed due to the incomplete screening of surface charges around FE films is named as the depolarization field. The depolarization field significantly influences the physical properties of FE, as it would, in general, affect spontaneous polarization. Depolarization field is inversely related to the thickness of the FE films or distance between the electrodes. It was proposed that anomalous photovoltage (APV) is closely dependent on the amount of screening of the spontaneous electric polarization [19, 22, 48]. Depolarization field might be the factor behind the separation of photogenerated charge carriers. Distribution of screening charge mainly depends on the properties of both FE materials and electrodes, namely, dielectric constant, free charge density, and remnant polarization [49]. High photovoltage output around 0.25 V was observed when aluminum-doped zinc oxide was used as the electrode instead of gold, which was because of the larger depolarization field [50]. Chen et el. used indium tin oxide (ITO) as the electrode due to large expected depolarization field on the thick PZT film and found a drop in photovoltage output, which was attributed to a decrease in depolarization field in the PZT film [51]. But dependence of generated photovoltage on the depolarization field has not been rationalized [50, 51]. However, the effect of depolarization field on the photocurrent is more pronounced than the effect on the photovoltage.

2.2.2.5 Design parameters for ferroelectric materials for PV applications Section 2.2.1.3 explained the requirement of optimum bandgap (1.4 eV), large absorption coefficient, superior carrier mobility, exciton dissociation efficiency, and long recombination lifetime and large diffusion length (mean free path) of the charge carriers for an efficient PV absorbing layer operating in conventional mechanism. Ferroelectrics, especially transition metal oxide ferroelectrics, although demonstrate a novel PV response, often do not meet the above conditions, simultaneously.

60

Ferroelectric Materials for Energy Harvesting and Storage

In general, transition metal oxide ferroelectrics are associated with large optical bandgap (PbTiO3, 3.5 eV; BaTiO3, 3.2 eV; Bi4Ti3O12, 3.25 eV). In most transition metal oxides, the top of the valence band is predominantly occupied by O 2p states. Presence of transition metal (TM) d-states with possibilities of hybridization with O 2p states is also common. The conduction band minima, on the other hand, is chiefly occupied by transition metal d-states. Large optical bandgap in these materials stems from the transition metal-oxygen (TM-O) bond character which includes bond length, hybridization energy, and Coulomb energy as well as the large difference in electronegativities (large ionic character of the bond) of them. A general observation of inverse relationship between bond length and bandgap suggests that small TM-O bond length results in larger bandgap [52]. Such large bandgap translates into poor absorption of solar spectrum in the visible range, 1.65–3.1 eV, and poor PV response as absorber layer. Therefore, to improve PV response in transition metal oxide ferroelectrics, it is very important to reduce the bandgap close to the ideal value, 1.4 eV. Fig. 2.10 plots bandgap as well as the spontaneous polarization of oxide ferroelectrics. It is observed that there are very few ferroelectric oxides with optimum bandgap as well as superior ferroelectric properties. Tailoring bandgaps in ferroelectrics requires modification of TM-O bond length as well as interaction energies (hybridization energy and Coulomb repulsion) between TM and O. To engineer the optical bandgap in these materials, several strategies have been adapted. The most significant of them are: (a) doping with iso- or aliovalent ions at cation sites [53, 54], (b) substrate misfit 1.0

KBNNO BaTiO3 SrTiO3

Ideal FE for PV application

0.8

Polarization (C/m2)

PbTiO3 LiNbO3 LiTaO3

0.6

PVDF NaNbO3 KNbO3 KN-BLZN BiFeO3 (K0.53Na0.47)NbO3 KN-SLZN KN-PLZN KN-BBZN KN-SBZN KN-PBZN BaZrO3 LaScO3 WO3

0.4

KTaO3 PIN-BIB PSN-BSB CaBiO3 CdBiO3

0.2

ZnBiO3 MgBiO3 BFCO

1

2

3

4 Band gap (eV)

5

Fig. 2.10 Bandgap versus the spontaneous polarization of oxide ferroelectrics.

6

7

Solar energy harvesting with ferroelectric materials

61

strain or epitaxial strain [55, 56], (c) controlling octahedral rotation [57], (d) replacing transition metal ion with d0 state with those having d10 states [58], and (e) using cations with small ionic radii in ABiO3 type of oxides [52]. A number of reports over the last decade have focused on bandgap engineering in ferroelectric oxides for prospective PV applications [59] which demonstrated that site-specific substitution by Mott insulator lanthanum cobaltite could reduce the bandgap of ferroelectric bismuth titanate by as high as 1 eV. Similar approach of bandgap tuning by means of doping and solid solution formation has resulted in a number of narrow bandgap TMOs such as KBiFe2O5 [60], Bi3.25La0.75Ti3O12 [61], Bi(Fe,Cr)O3 [13], and (KNbO3)1x(BaNi0.5Nb0.5O3δ)x [62]. Gap-state engineering by means of introducing defect states within the bandgap is another approach for improving optical absorption [63]. While there are several reports of improvement of PV performance in ferroelectric materials by tailoring the electronic structure [61, 63, 64], there are several other conditions as outlined above that need to be fulfilled for a high performing solar cell. Any systematic material design strategy in this regard has not been observed in literature. While few studies in halide perovskites attribute their improved performance to large shift current coupled with favorable optoelectronic properties, no corresponding literature in oxide ferroelectrics is available. Generally, ferroelectric oxide materials have a small absorption coefficient that corresponds to a small photocurrent [65, 66]. This, in turn, reduces the material consumption and processing cost. Absorption coefficient in a solid is related to the imaginary part of the dielectric function and the refractive index using equation [67], ε00i, ∞ ¼

αnr c ω

(2.11)

Here, ε00i, ∞ is the imaginary part of the dielectric function, α is the absorption coefficient, nr is the refractive index, c is the velocity of light, and ω is the angular frequency. Therefore, high absorption coefficient requires large dielectric constant. A simple harmonic oscillator model can result in the expression of the dielectric function as follows: Ni q2i ε∗i, ∞ ¼ 1 + ε0 mi

(

) 1  2  ωo,i  ω2 + iγ i ω

(2.12)

From Eq. (2.12), one can find out the imaginary component of the dielectric constant: ε00i, ∞

Ni q2i ¼ ε0 mi

(

γω  2 2 ωo,i  ω2 + γ 2i ω2

) (2.13)

From the above equation, it is observed that the imaginary part of the dielectric constant of the materials is dependent on the mass of the charged particles, the charge on electron/ion, and the plasma frequency indicating the type and strength of the bonds in the material.

62

Ferroelectric Materials for Energy Harvesting and Storage

One of the significant parameters of the materials for PV application is the dissociation efficiency of the excitons. In order to have high photocurrent, dissociation efficiency of the excitons ought to be high. The binding energy of the electron-hole pairs is normally conversely proportional to the dielectric constant of the materials, Ebind ¼ 13.6
(μ/ε2r mo); here, μ is the effective mass of the charge carriers such that μ1 ¼ m1e + m1h ; me and mh are the mass of electrons and holes, respectively. εr is the dielectric constant and mo is the rest mass of electron. Practically, majority of the ferroelectric materials have larger dielectric constant than the semiconductors; therefore, binding energy between electron and hole in ferroelectric materials is small indicating relatively easy dissociation of the exciton. Bulk exciton Bohr radius, aB ¼ a0
(εrmo/μ) and a0 ¼ 0.528 A˚, which represents the separation of the electrons and holes could be large for ferroelectrics. Therefore, it is believed that FE-PV device allows larger length-scale for efficient separation of the charge carriers. Typically, excited electrons move along the direction of polarization in noncentrosymmetric ferroelectric PV devices [20]. In this way, for higher photocurrent in the ferroelectric PV devices, materials should be designed in such a way that it has higher polarization in the crystal. Higher polarization in the ferroelectric materials can be achieved by modifying the chemical composition and/or the structural symmetry of the material, which can potentially be attributed to high photovoltaic response [68–70]. The role of polarization in the photovoltaic response in ferroelectrics is not unambiguous. As mentioned above, the PV response in ferroelectrics is best explained by shift-current model within bulk photovoltaic effect. The model suggests that the position of the electron wave packet shifts in the real space upon interband optical transition during absorption of light. According to the modern theory of polarization, electronic contribution to the overall polarization, Pel, is also related to the Berry phase over the valence state. The shift current is also related to the Berry phase and it related to the electronic contribution, Pel, between the valence and conduction bands upon optical transition. Thus, a material possessing large Pel is expected to demonstrate large shift current, and therefore, enhanced PV response [41]. However, the above rationale has not been confirmed across different ferroelectric systems. Charge collection efficiency is a significant aspect that needs to be considered so as to improve the photocurrent response in photovoltaic devices. The charge collection efficiency relies on carrier mobility, carrier lifetime, and electric field. The lifetime of the nonthermalized electron was in picoseconds (1012 s) [35] and recombination lifetime was estimated to be in microseconds [71]. In general, charge collection efficiency is a strong function of thickness of the layer, as demonstrated experimentally by Ref. [23, 72]. PLZT films of 4 μm thickness showed a 100-fold increase in photocurrent in comparison to that of a 2.4 mm thick film. However, the photovoltage was reported to decrease with decreasing film thickness. Another method to improve charge collection efficiency is to increase the collecting electric field as [73], worked on BiFeO3-related devices in which ITO electrode was replaced by nitric acid-treated graphene that was attributed to the improved photocurrent. Ref. [43] investigated a tip-enhanced PV effect in the BiFeO3 devices and found high charge collection efficiency by the atomic force microscope (AFM) tips.

Solar energy harvesting with ferroelectric materials

63

Depolarization field and charge collection efficiency are clearly controlled by the degree of screening between the ferroelectric film and the electrode as well as the film thickness. Charge collection efficiency and exciton dissociation efficiency are increased by reducing the degree of screening and the film thickness. Theoretical calculation by Qin et al. confirmed that photocurrent in PLZT should be improved by reducing the film thickness and replacing the metallic electrode with semiconducting electrode, i.e., lower screening effect [25, 48]. In addition to the above stated requirements of broken inversion symmetry and large electronic polarization, other material property requirements for BPVE has not been well-defined. However, it would be quite natural to consider that material property requirement of conventional PV devices which are equally important for BPVE-based PV materials. Therefore, the general design criteria of ferroelectric materials toward efficient PV solar cell application should include: (a) Optimum bandgap close to 1.4 eV. (b) Large absorption coefficient. Absorption of light with E  Eg can be ensured. With increasing thickness of the sample, most semiconductors can be ensured to demonstrate perfect absorption with few tens or hundreds of microns thick. With increasing thickness, the material consumption, production cost vis-a`-vis overall economy of the technology go up. Further, increasing thickness of the material increases the concentration of various types of defects and phenomena associated with them. However, with large absorption coefficient, the thickness can be reduced to an optimum level reducing the materials and fabrication cost. (c) Long carrier diffusion length which translates to long recombination lifetime and small electron-hole binding energy. (d) Sufficient carrier mobility to ensure that the photogenerated charge carriers are quickly transported and collected at the electrodes. (e) Polar crystal symmetry with large contribution of electronic polarization to the spontaneous polarization.

2.2.3 Perovskite photovoltaics Perovskite photovoltaic is a rising innovation that is considered as one of the most attractive approaches to convert solar energy into electricity. Perovskite photovoltaics belong to third-generation photovoltaic technology, i.e., thin-film solar cells which involve a perovskite-structured compound, most often a hybrid organic-inorganic lead or tin halide-based material. Third-generation technologies intend to improve low output efficiency of the second generation (thin-film) technologies, while keeping up with the exceptionally low production costs [15]. Perovskites are defined as crystalline materials with chemical formula ABX3. Transition metal oxide perovskites show excellent physical properties, such as ferroelectricity, magnetism, and superconductivity, to name a few. In perovskites, A and B are cations have different sizes; generally “A" atoms are larger than “B” atoms, whereas X is an anion. Generally, in perovskite cubic unit cell, “A” atom occupies corner positions, i.e., (0, 0, 0), “B” atom occupies body-centered position (1/2, 1/2, 1/2), and oxygen atoms are at face-centered positions (1/2, 1/2, 0) as shown in Fig. 2.11 [27].

64

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 2.11 Schematic of an ideal perovskite structure.

Some major advantages of these perovskites include low manufacturing/fabrication cost, high absorption coefficient allowing the use of thin, flexible, and lightweight solar modules and higher power conversion efficiency (PCE) [74–77]. The excellent conversion efficiency at competitive price renders perovskite solar cell an edge over the earlier photovoltaic technologies such as silicon with a PCE of 20%, GaAs (PCE  18.4%), CdTe (PCE  19.6%), and copper indium gallium selenide/sulfide (PCE  19.6%) [78–80]. In 2009, Miyasaka and coworkers first used perovskite as the solar cell material, based on the configuration of dye-sensitized solar cell with a thin layer of perovskite cell on titanium oxide (TiO2) and found very low power conversion efficiency of 3.8%, which was attributed to the presence of liquid electrolyte [81]. In 2011, Im et al. improved the power conversion efficiency from 3.8% to 6.5% by using quantum dots for the nanocrystalline materials [82]. Lee et al. studied the perovskite solar cells based on meso-superstructured organometallic halide and reported a low-cost solar cell with power conversion efficiency of 10.9% [75]. In 2013, many developments were observed on both the sensitized and planar architectures. Burschka et al. reported power conversion efficiency of almost 15% based on deposition technique [74]. Zhou et al. achieved power conversion efficiency of 19.1% while adopting planar architecture and suppressing the recombination of carriers [77]. As of 2016, power conversion efficiency of 22.1% was reported by the researchers from University of Science and Technology, Korea, and UNIST (South Korea) [83]. It is noticeable that the maximum power conversion efficiency of perovskite solar cells increased from 3.8% to 22.1% in the last 7 years. The power conversion efficiency of perovskite solar cell is shown in Fig. 2.12. This improvement in power conversion efficiency is due to the excellent optoelectronic properties of perovskites, particularly a tunable energy bandgap, high absorption coefficient, long carrier lifetimes, and high effective masses of electron/hole and exciton binding energy [84–91]. Perovskite systems are mainly categorized into two broad systems: inorganic oxide perovskite systems and halide perovskite systems. The entire list of classification of perovskite systems is shown in Fig. 2.13.

Solar energy harvesting with ferroelectric materials

65

25

Power conversion efficiency (%)

22.1 19.3

20

20.1

15 15

10

9 6.5

5

3.8

0 2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

Year Fig. 2.12 Power conversion efficiency of perovskite solar cell.

Perovskite system

Halide perovskite

Organo metal perovskite

Alkali halide perovskite

Inorganic oxide perovskite

Doped perovskite

Intrinsic perovskite

Fig. 2.13 Classification of perovskite system.

Halide perovskite systems have very high performance with affordable cost. Halide perovskite systems are further classified into organometallic perovskite systems and alkali halide perovskite systems. In common halide perovskite (ABX3), univalent organic cations, for example, formamidinium (NH2CH ¼ NH+2 or FA+) and methylammonium (CH3NH+3 or MA+), or inorganic alkali metal cation sits at A-site, whereas halogen anion sits at the X-site, which increases the optoelectronic properties and stability of the perovskite solar cell. Divalent metal ions such as Pb2+ sit at the Bsite that enhances the properties of perovskites, such as superior electrical movability, tunable energy bandgap, high temperature stability, mechanical stability, magnetic and dielectric transition, and mechanical plasticity.

66

Ferroelectric Materials for Energy Harvesting and Storage

2.2.3.1 Fabrication of perovskite solar cell Techniques used to deposit perovskite layer include spin coating, spray coating, screen coating, dip coating, vapor deposition, and dual source evaporation. Advantages and disadvantages of these fabrication techniques are listed in Table 2.1. Spin coating is an economical, popular, and quick technique used for developing thin films on different types of substrates. This method involves deposition of perovskite solution upon a specific substrate followed by continuous rotation at high velocity and subsequent drying, resulting in the development of a thin film. The thickness of the perovskite thin film is managed by various factors such as temperature, rotational velocity, and density of the solution. After drying, annealing is used to remove residual solvent and increase crystallinity. Spray pyrolysis is another technique for the fabrication of perovskite thin film at low cost. In this technique, deposition of the thin film is done by spraying perovskite solution onto a substrate and consequent chemical reaction. Ultrasonic spray deposition and electrostatic spray deposition are mostly used for spray coating. Substrate temperature should be taken care of because the low temperature might induce cracking in the deposited film and high temperature cause porous and rough films. Screen printing is one of the efficient and cheap techniques for the fabrication of thin films. This technique requires high viscous and low volatility perovskite solutions accompanied with large wet film thickness. A steel or synthetic material screen having printing pattern first imperviously contacts with a viscous solution followed by with the substrate and consequent dragging from one side to other, which allows the solution to discharge through uncovered parts of the mesh

Table 2.1 Advantages and disadvantage of various fabrication methods. Methods

Advantages

Disadvantages

Spin coating

▪ ▪ ▪ ▪ ▪

▪ Uneven film thickness ▪ Material loss ▪ Have defects

Spray coating

▪ ▪ ▪ ▪

Dip coating

▪ ▪

Screen printing

▪ Simple mechanism

▪ Slow

▪ Easy adjustment of layer thickness ▪ Cheap

▪ Screen blocking

Cheap Quick processing Multilayer deposition Controllable film thickness Less quantity of solution required Cheap Simple setup Consistent film thickness Diluted perovskite solution is used Uncomplicated Cheap

▪ Nearly uncontrollable process parameter

▪ Uneven film thickness ▪ Slow

Solar energy harvesting with ferroelectric materials

67

printing-required pattern. Dip coating is a very effective and low-cost technique to have perovskite film onto the substrate by simply immersing the substrate in the coating solution for a specific span of time followed by withdrawing of the wetted substrate and consequent solvent evaporation, resulting in the formation of a thin film. Fabrication of perovskite thin film by this technique requires consideration of the governing forces, such as viscous drag, inertia, surface tension, and gravitational force. Single dipping of the substrate into the coating solution can produce a dense and defect-free thin layer of thickness in between 15 and 60 μm. Thermal evaporation technique involves the fabrication of perovskite thin film in which deposition of material onto the substrate takes place straight from the vapor phase. This process demands the temperature of the substrate below the freezing point of the material, but high temperatures to induce vaporization. Materials are thermally discharged from the surface and move toward the substrate, resulting in the formation of film coating. Furthermore, there are two proper techniques to fabricate perovskite films, which are single source evaporation and dual source evaporation. In dual source evaporation technique, both organic and inorganic salts are evaporated at a same time from two different sources, whereas in single source, evaporation of inorganic part is rapid because of high temperature, but abstains from thermal decomposition of the organic part.

2.2.3.2 Disadvantages of perovskite solar cell Stability and limited service life Economy, efficiency, and stability are the key obvious parameters required for successful commercialization of perovskite solar cells. As discussed earlier, efficiency of the newly synthesized perovskite solar cell has now increased significantly to 22%; however, their efficiency decreases rapidly during operation [92]. Therefore, from the perspective of commercialization, it is important to work on issues, such as low stability and limited service lifetime. The perovskite materials degrade quickly due to exposure to water, oxygen, humidity, and heat [92].

Noxious material Presence of lead (Pb) in perovskite solar cell materials poses potential environmental hazard. Therefore, it is necessary to replace lead-containing PV materials with alternative element(s) that must fulfill features such as low-cost, excellent optoelectronic properties, flexibilities, long-term stability, etc. In this respect, different attempts have been made: (i) Homovalent lead substitutions

Tin (Sn2+) and germanium (Ge2+) cations are alternatives that are environmentfriendly. Perovskite solar cells fabricated using tin and germanium have been successfully tested with a maximum power conversion efficiency of 6.4% for MASnI3 [93] and 0.2% for MAGeI3 [94]. By considering bandgap and stability, Mg2+, Ni2+, Cd2+, V2+, Ga2+, In2+, Hg2+, Ge2+, and Sn2+ have been suggested as potential substitutes of lead in perovskite solar cell material. The bandgap of 1.7, 1.5, and 0.9 eV have been found in Mg-based perovskites, namely, CsMgI3, MAMgI3, and FAMgI3,

68

Ferroelectric Materials for Energy Harvesting and Storage

respectively [95]. Ca2+, Cd2+, and Sr2+ are the other possible substitutes toward reducing the lead-related toxicity [96]. (ii) Heterovalent lead substitutions

Generally, A2BB3+ X6 double perovskites exhibit excellent bandgap, magnetic, ferroelectric and multiferroic properties. The first developed A2BB3+ X6 double perovskites have B sites occupied by monovalent Cu, Ag, and Au, whereas B3+ is occupied by trivalent metals Bi and Sb, which maintain tunable energy bandgaps, lower than 2.7 eV, as well as retain total charge neutrality [97]. Both homovalent and heterovalent substitutions of lead in halide perovskites, which consists of various cations, 7 for A-site, 8 for B-site, and 34 for B3+-site as well as anions of 5 for X-sites, are shown in Fig. 2.14 [98]. Tin-based halide perovskite: Tin and lead have almost identical external electron ˚ ) than the Pb shell arrangement, whereas the smaller ionic radius of Sn (1.35 A ˚ (1.49 A) leads to an easy replacement of Pb without any notable deviation in the lattice structure. However, there is a difference in semiconducting properties and chemical bonding as compared to lead-based perovskite. Some useful properties of tin-based perovskites are high optical absorption coefficient, a low exciton binding energy, high charge mobility, and narrow bandgap. The drawback of Sn-based perovskite is their low power conversion efficiency, sensitivity to ambient atmosphere and difficulty in forming pinhole-free films. Cesium tin iodines: In cesium tin iodide (CsSnI3), holes are the majority charge carriers. The material has low energy bandgap of 1.3 eV, high photoluminescence, better electrical conductivity, high charge carrier density (107 cm3), high hole

LI

3

Na

11

K

19

Rb

37

Cs

55

NH4+

A2

Sc Y

21

Ti

22

V

23

Cr

24

3+

B

Mn

B

25

Fe

26

X6

Co

27

Ni

28

39

57

La

29

Cu Ag

47

Au

79

Al

13

Ga

31

In

49

Sb

51

Tl

81

Bi

Ho

67

F

9

Cl

17

Br

35

I

53

83

CN–

CH3NH3+

Ce

58

Pr

59

Nd

60

62

Sm

Eu

63

Pu 94 Am95

Gd

64

Tb

65

Dy

66

Er

68

69

Tm

Yb

70

Bk 97

Fig. 2.14 Halide double perovskites forming element having composition A2BB3+ X6.

Lu

71

Solar energy harvesting with ferroelectric materials

69

mobility (585 cm2 ∙ V1 ∙ s1), low exciton binding energy (18 meV), and a large absorption coefficient. These properties confirm its appropriateness as optically active materials [98, 99]. Methylammonium tin iodide: Methylammonium tin iodide (MASnI3) has small optical energy bandgap, 1.3 eV, large electron mobility, 2000 cm2 ∙ V1 ∙ s1, and high charge carrier density, 1014 cm3 [100]. It also has excellent electrical conductivity, 5
102 S ∙ cm1, at ambient temperature and thermoelectric sensitivity of 260 m ∙ V∙K1, which are good to be compared with lead halide perovskites [93]. Formamidinium tin iodide: Formamidinium tin iodide (FASnI3) possesses excellent properties, such as low energy bandgap, 1.41 eV, decent film quality and stability at over a wide temperature range because of the inflexible perovskite structure due to the electrostatic attraction between the Sn2+ and HC(NH2)+2 cations. Although formamidinium tin iodide is stable over a wide temperature range, methylammonium tin iodide shows a low phase transition temperature of 329 K [101, 102]. Several other lead substitute perovskites such as Sn-based perovskites, Ge-based halide perovskites, Bi-based halide perovskite, Sb-based halide perovskite, In-based halide perovskites, and transition metal namely Cu, Cr, Mn, and Fe-based halide perovskites are also available in literature. Among all these perovskite solar cell materials, only the tinbased lead-free perovskites have a favorable power conversion efficiency, which is still less from the efficiency essential for commercial application. Therefore, fabrication of new lead-free perovskites for commercial applications still offers open challenges with regard to improvement of conversion efficiency.

2.2.4 Transition metal oxides Transition metal oxide perovskites are regarded as a promising material class to complement silicon-based conventional solar cells. Research in photovoltaic based on transition metal oxide has nowadays increased because of their nontoxicity, abundance, and higher chemical stability. Large bandgap in common ferroelectric perovskites is mainly because of the basic characteristics of the metal-oxygen bonds (AdO and BdO). In most transition metal oxides, the top of the valence band is predominantly occupied by oxygen (O) 2p states. The conduction band minima, on the other hand, are chiefly occupied by transition metal d-states. A simple and generic band structure representation of transition metal oxide (TMO) is shown in Fig. 2.15. Large bandgap (3–5 eV) in these materials is due to the large difference in electronegativity between the transition metal and oxygen. However, multiferroics with partially filled d-states of the transition metal yield comparative small bandgap. Multiferroic BiFeO3 demonstrates an energy bandgap of 2.7 eV, which triggered its potential use in photovoltaic applications [59, 103]. BiFeO3-based photovoltaic devices demonstrate further interesting characteristics, such as high photovoltage and switchable photodiode behavior. However, photocurrent and power conversion efficiencies are limited for commercial application. Moreover, it absorbs only 20% of the solar spectrum. Poor PV response in BiFeO3 is further attributed to the presence of imperfections such as oxygen and cation vacancies. These defects increase the recombination rate,

70

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 2.15 Simple and generic band structure representation of transition metal oxide (TMO).

which can be attributed to the short carrier diffusion length. Several strategies have been adapted so far to tailor the bandgap of TMOs for potential photovoltaic applications. Wang et al. explored merging the mesostructured carrier transporters NiO and TiO2 together with BiFeO3. This nanoheterostructure design similar to the DSSC can effectively increase the power conversion efficiency [104]. Furthermore, the bandgap of BiFeO3 can be reduced by doping with Cr that changes the structure of BFO to double perovskite Bi2FeCrO6. Nechache et al. fabricated the double perovskite BFCO epitaxial thin films on the niobium-doped SrTiO3 substrate with help of pulse laser deposition (PLD) technique. They used red laser having wavelength and intensity of 635 nm and 1.5 mW/cm2, respectively, to investigate the photovoltaic properties and found that the magnitude of photovoltage, photocurrent, and power conversion efficiency were 0.6 V, 1 mA/cm2, and 6%, respectively [105]. Afterward, Nechache et al. demonstrated yet another technique to enhance the power conversion efficiency of Bi2FeCrO6, through tuning the bandgap by modifying the Fe and Cr cationic ordering. They first synthesized single layer photovoltaic devices with indium tin oxide as top electrodes and SrRuO3 as bottom electrodes by pulse laser deposition techniques with different deposition rates. They revealed gain in power conversion efficiency with decreasing deposition rates, and reported the power conversion efficiency, photocurrent, fill factor, and photovoltage of 3.3%, 11.7 mA/cm2, 36%, and 0.79 V, respectively. Although single layer photovoltaic devices are not enough to scavenge the whole solar radiation spectrum, they stack three layers in one device and reported the better power conversion efficiency, photocurrent, fill factor, and photovoltage of 8.1%, 20.6 mA/cm2, 47%, and 0.84 V, respectively [13]. Zenkevich et al. studied the typical perovskite BaTiO3 films of thickness in between 20 and 50 nm. They used pulse laser deposition technique to deposit epitaxial Pt/BaTiO3 above the monocrystalline MgO coated with platinum substrate and found highest photovoltaic field of 300 kV/cm, a low photocurrent, and a photovoltage of 0.62 V near UV in the 20 nm thin film [106]. Power conversion efficiency in the perovskite solar cells has been mainly restricted by the charge recombination. One of the efficient methods to reduce charge recombination is the interface modification. Qin et el. used spin coating technique to successfully synthesize an extremely thin BaTiO3 modification layer on mesoporous TiO2 with different concentrations of barium salt solution (0–2.5%) in

Solar energy harvesting with ferroelectric materials

71

order to optimize the performance of the solar cells, which significantly reduced interfacial charge recombination. Out of the various concentrations of barium nitrate solution used, 0.9% concentration showed maximum power conversion efficiency, open circuit voltage, fill factor, and short-circuit current density of 17.87%, 1.10 V, 71.73%, and 22.52 mA/cm2, respectively [107]. As we know, inorganic perovskite oxide usually has low light absorption properties because of the wide bandgaps, which can be generally modified by the doping with a metal or nonmetal. Yang et al. systematically studied the effects of single element doping on the performance of barium titanate and found that the single element doping introduced the impurity levels. These impurity levels act as the recombination sites for electrons and holes, corresponding to the decrease in power conversion efficiency. Co-doping with two distinct types of elements might reduce some of the unwanted features of single doping, such as less solubility, formation of deep levels that are not suitable for carrier transitions, and spontaneous origination of imperfections. They also studied the electronic properties of V-M00 co-doped BaTaO3 and Nb-M00 co-doped BaTiO3 where M00 represent 3d or 4d transition metal by first-principle method using DFT calculations. They suggested that, for tailoring the bandgap of BaTiO3, V-M00 co-doping is more acceptable than Nb-M00 co-doping because of formation of deep levels around the energy band by the Nb-M00 co-doping, whereas VdCr doping reduced the energy bandgap to almost 0.4 eV [108].

2.3

Photochemical conversion of solar energy: Solar water splitting

Although solar photovoltaic is the most exciting technology toward sustainable energy solution, the process of conversion of solar energy to usable energy involves costly and hazardous chemicals and processes that may produce toxic by-products. Further, the challenge of transportation and storage of electricity is also of major concern. From the point of view of storability, transportability, and sustainability, hydrogen energy and related applications are considered as the most efficient renewable energy technologies. Energy can be extracted efficiently from hydrogen fuel cells (HFCs). Hydrogen-powered fuel cells are more than two times efficient compared to conventional combustion power plants without emitting toxic greenhouse gas carbon dioxide. These attributes make hydrogen a prominent candidate for a futuristic energy carrier. Being classified as an environment-friendly renewable energy source, an efficient cost-effective method for carbon-free green production of hydrogen is highly desirable. In the present scenario, hydrogen is produced primarily from fossil fuels, e.g., biomass, thermal decomposition of natural gas or oil, and coal gasification which are sources of carbon emission that have severe effect on environment. In comparison to these processes of hydrogen production from fossil fuels, photovoltaic-based water splitting using solar energy is being considered as the most eco-friendly technique. This process of hydrogen gas generation is an indirect way of conversion of sunlight, apart from conventional conversion of solar energy to solar photovoltaic and thermal energy [109, 110]. In this section, we will discuss solar water splitting technology such

72

Ferroelectric Materials for Energy Harvesting and Storage

as solar photocatalysis, photoelectrochemical water splitting using visible solar spectrum. We will also discuss the possibility of exploiting the ferroelectric/piezoelectric properties of ferroic materials to design a photocatalytic cycle that enables water splitting into hydrogen and oxygen under sunlight.

2.3.1 Basics of solar water splitting In solar water splitting, semiconductors are used as active material in electrolytes. Sunlight is directly absorbed in the depletion layer of semiconductor where light energy is converted to exciton (electron-hole pair) and used to drive chemical reaction to produce hydrogen at the interface of electrolyte and semiconductor. Solar energy is directly utilized to dissociate water molecule into hydrogen and oxygen through three kinds of water splitting systems [106, 109–113]: (1) particulate photocatalysis; (2) photoelectrochemical (PEC); and (3) photovoltaic photoelectrochemical (PVPEC). In solar water splitting process, photocatalysis plays an important role. The early concept of hydrogen generation through photoelectrochemical process was to use a photocatalyst semiconductor such as rutile TiO2 as the cathode and platinum (Pt) as the anode material. Photocatalytic water splitting reaction involves direct absorption of visible solar light in semiconductor-based photocatalyst through charge carrier generation, separation, transport, and transfer where bandgap and band position of photocatalytic semiconductor play important role. Fig. 2.16 describes the principle of photocatalytic water splitting reaction where photogenerated electrons and holes reduce and oxidize water, respectively, at the surface of semiconducting cathode and anode. Catalytic water oxidation occurs at the anode in the presence of photogenerated holes, while photoelectrons combine with H+ in the metal cathode releasing H2. Cathode: 4H+ + 4e (water reduction); Anode: 4OH + 4h+ (water oxidation of H2O). Water splitting reaction is thermodynamically nonspontaneous reaction and needs an extra energy of 237 kJ/mol (Gibbs free energy). That means that since a photocatalyst has a bandgap larger than energy

Fig. 2.16 Schematic of photocatalytic water splitting hydrogen generation process. Taken from J.S. Jang, et al., Heterojunction semiconductors: a strategy to develop efficient photocatalytic materials for visible light water splitting, Catal. Today 185 (1) (2012) 270–277.

Solar energy harvesting with ferroelectric materials

73

required for water splitting 1.23 eV, therefore its conduction and valence band edge should have oxidation and reduction potential of water. Therefore, O2/H2O redox potential difference is 1.23 eV and the requirement of bandgap is larger than the energy of water splitting. In other words, the valence band potential should be more positive than O2/H2O redox potential of 1.23 eV vs. normal hydrogen electrode (NHE) to allow water oxidation, and the conduction band must be more negative than the H+/H2 redox potential of 0 V vs. NHE to permit water reduction [109, 110, 113].

2.3.2 Ferroic materials for photoelectrochemical water splitting: Fundamentals of material requirement The PEC activity of water splitting is decided by the choice of appropriate photoelectrode material. Conventional wide bandgap semiconducting photoelectrode materials exhibit restricted photoabsorption in solar spectrum, though they are reasonably stable in electrolytic solution. On the other hand, narrow bandgap materials are efficient in light absorption in the visible solar spectrum, but may have issues related to stability in electrolytic solution. The choices of photoelectrode materials for water splitting are restricted by following criteria [112–114]: (i) Electrode materials must have a bandgap 2 eV at least more than 1.23 eV (the O2/H2O redox potential difference). More precisely, both bandgap and band edge positions are important to achieve high efficiency of PEC water splitting process. Fig. 2.17 shows bandgap and band edge positions of a few selected photoelectrode materials. Electrolysis of water is composed of two reactions occurring at two electrodes. Hydrogen is generated through water reduction at the photocathode with H+/H2 redox potential 0 V vs normal hydrogen electrode (NHE). Oxygen is produced by oxidation of water with O2/H2O redox potential 1.23 eV vs NHE at the photo-anode. Considering the thermodynamic energy loss

Fig. 2.17 Bandgap and band edge position of a few semiconductor and ferroelectric materials with respect to NHE. Taken from S. Kim, et al., Ferroelectric materials: a novel pathway for efficient solar water splitting, Appl. Sci. 8 (9) (2018) 1526.

74

Ferroelectric Materials for Energy Harvesting and Storage

(0.3–0.4 eV) and an overpotential of  (0.4–0.6 eV), we conclude that the photon energy must exceed 2 eV to start the water splitting process. Since maximum limit of visible solar spectrum is 400 nm, i.e., 3.2 eV, the choice of bandgap for semiconducting materials is limited within the range of 2–3.2 eV for efficient solar water splitting. Along with the bandgap of photoelectrode material, the band edges should have specified positions. The conduction band (CB) edge must position more negatively compared to 0 V H2/H+ redox potential vs. NHE to satisfy H2 production condition. Similarly, the valence band (VB) edge position should be more positive than 1.23 eV O2/H2O redox potential vs. NHE for oxygen generation at the photo-anode. When a photoelectrode material satisfies both bandgap and band edge condition, the visible light is absorbed that excites electron from valence band to conduction band leaving holes in the valence band. (ii) Efficient charge generation and transfer: The efficiency of PEC is severely affected by the problem of charge recombination and inadequate charge transportation. Intrinsic properties like charge (hole and electron) mobility as well as the structural properties like crystallinity and surface to volume ratio in nanostructures affect the separation and transport of charge carrier. (iii) There exists energy barrier for charge transfer that causes energy loss in the PEC reaction. Fast surface reaction kinetics can pass such energy barriers and avoid electron-hole recombination. In PEC, hydrogen is produced from H+ (in acidic) and from H2O in base environment, whereas oxygen is generated from H2O (in acidic) and OH– (in basic) environment. In long-term PEC cells, the strong acidic or basic condition leads to photocorrosion and fluctuation in photocurrent within short interval of time. On the other hand, maintaining extreme environment (a high or low pH condition) in a PEC is important to minimize Ohmic losses, localized pH gradient overpotential in separate electrode PEC cells. Ferroic materials with their intrinsic stability in corrosive electrolytic condition are suitable for using in PEC cells. Further, favorable band edge alignment of electrode material with respect to decomposition potential promotes stable and efficient PEC devices.

2.3.2.1 Various ferroelectrics as photoelectrode material for PEC water splitting Photoelectrode materials for visible light photocatalysis water splitting should satisfy the requirements of band gap 1.8 eV < Eg < 2.2 eV and band positions as discussed in the previous section. The other requirements of photoelectrode materials are facile scalable fabrication, stability in aqueous electrolyte solution, high crystallinity with little defects, high carrier mobility, and efficient charge transport and separation. Generally, p-type semiconductors serve as cathode material to generate the required cathodic photocurrent for water reduction while n-type materials work as photoanodes for oxidation in PEC water splitting. Metallic oxide semiconductors, such as Cu2O, TiO2, CuO, and NiO are the well-known conventional photoelectrodes used for solar water splitting due to their cost-effective production and a long-term stability in aqueous electrolytic solution. However, these semiconducting electrode materials are limited by some major deficiencies such as corrosion in electrolyte medium [115], inappropriate position of valence and conduction band [116, 117] and wide band gap (3.6–4 eV) resulting in inadequate photocurrent, low stability, and thereby restricted PEC activity. In this regard, combining ferroelectrics with conventional

Solar energy harvesting with ferroelectric materials

75

oxide-based semiconducting electrodes opens up possibilities of better PEC performance. Suitable control of polarization induced internal electric field results in efficient charge separation as well as transfer at the electrode-electrolyte interface leading to improved hydrogen generation in PEC devices. PbTiO3 or PTO is well-known room temperature ferroelectric oxide having band gap 2.75 eV with suitable band structure that helps in charge separation and transfer. BTO is used as photocathode in PEC devices due to its p-type semiconducting behavior. Further, transition metal doping and or grafting has been utilized to improve photogenerated electron transfer at the electrode-electrolyte interface to obtain enhanced photocurrent [118]. A few innovative synthesis techniques have been used such as microwave-assisted synthesis to achieve improved photocurrent. Quaternary metal oxide such as (PbZrTi)O3 thin film with prominent ferroelectric behavior functions as effective photocathode for water splitting when deposited on ITO coated quartz glass [119]. Double perovskite ferroelectric Bi2FeCrO6 also plays role of photocathode with narrow band gap 1.9–2 eV [120]. Ferroelectric BaTiO3 or BTO with n-type semiconducting behavior is utilized in photooxidation of water as photoanode material in heterojunction-based PEC devices at the TiO2/BTO interface [121]. BiFeO3 or Bismuth ferrite is the mostly studied multiferroic oxide that exhibits room temperature ferroelectricity with significant spontaneous polarization. Its band gap ranges from 2.2 to 2.7 eV with suitable band positions makes it an appropriate material for PEC water splitting. The robust spontaneous polarization in BiFeO3 has a crucial effect on band bending and hence plays a significant role in separation of photoexcited charge carriers within the space charge region. Depending on the processing technique BiFeO3 can be synthesized as n-type or p-type material and hence can be used as anode or cathode in PEC water splitting, respectively [122, 123]. Apart from the aforementioned perovskite ferroelectric oxides, CdS is the mostly used ferroelectric chalcogenide in PEC water splitting due to its narrow band gap and suitable band structure, i.e., band edge position. However, it suffers from internal charge recombination that can be overcome by formation of heterojunctions with carbon or ZnO nanostructures.

2.4

Summary

This chapter reviews the fundamentals of solar energy harvesting using ferroelectric materials. Among all the possibilities, solar photovoltaics and photochemical energy conversion using ferroelectrics have been discussed. The chapter begins with the discussion on the fundamentals of solar cell physics and material requirement for conventional solar cell applications. Prerequisite for conventional solar cell materials includes consideration of materials properties, such as, optimum bandgap ( 1.4 eV), large absorption coefficient, long carrier diffusion length which translates into long recombination lifetime and small electron-hole binding energy, and high carrier mobility to ensure that the photogenerated charge carriers are quickly transported and collected at the electrodes. Physics of photovoltaic effect in ferroelectric materials was discussed primarily in light of bulk photovoltaic effect.

76

Ferroelectric Materials for Energy Harvesting and Storage

Generated photovoltage in the ferroelectric materials depends upon many parameters such as electrical conductivity, orientation of the crystal, remnant polarization, intensity of light, film thickness, interface between the electrode and ferroelectric film, and domain walls. Available literature on the mechanisms of photovoltaic effect in ferroelectrics to explain the origin of the large photovoltage in ferroelectric photovoltaic devices was presented. Early literature on large photovoltage in ferroelectric photovoltaic devices explained the phenomena by various models such as nonlinear dielectric model, depolarization field model, and Schottky barrier model. According to the nonlinear dielectric model, the generated photovoltage is due to the nonlinear relationship between polarization density and the electric field of the incident light. According to the depolarization field model, the internal depolarization field of the ferroelectric is attributed to large photovoltage. Depolarization field is oppositely related to the thickness of the FE film; therefore, depolarization filed theory is applicable only when the FE films are thin. Schottky-junction effect is also supposed to give rise to the photovoltage in FE thin films. The generated photovoltage due to Schottky junction is still equivalent to the bandgap of the FE semiconductor materials, which is much less than anomalous photovoltage in bulk ferroelectrics. In contrast to the above models, domain wall theory suggests that the observed photovoltage is due to the domain structure of ferroelectrics wherein the magnitude of photovoltage increases with the total number of domain walls between the electrodes along the direction of net polarization, while there is no photovoltaic effect along direction perpendicular to the direction of net polarization. More recently, shift-current model has been proposed to explain the observed photovoltaic effect in ferroelectric materials. In this model, the photovoltaic effect is due to the asymmetry in the wavefunction of the cation-anion partial covalent bonds. Shift-current model predicts enhanced PV response in materials with rhombohedral and orthorhombic symmetries. On the basis of the physics of conventional as well as bulk photovoltaic effect, material design strategies for ferroelectric-photovoltaic devices have been proposed as follows: efficient ferroelectric materials must have an optimum bandgap, large absorption coefficient, superior carrier mobility, high exciton dissociation efficiency, and long recombination lifetime along with polar symmetry with rhombohedral or orthorhombic crystal classes. The most significant techniques used to modify the bandgap in the ferroelectric material are controlling octahedral rotation, doping with iso or aliovalent ions at cation sites, substrate misfit strain or epitaxial strain, replacing transition metal ion with d0 state with those having d10 states, and using cations with small ionic radii in ABiO3 type of oxides. Large absorption coefficients in the ferroelectric materials correspond to their high photocurrent; moreover, the absorption coefficient is directly related to the imaginary part of the dielectric function. Dissociation efficiency of the excitons also directly influences the photocurrent in the ferroelectrics. Noncentrosymmetry in the materials is related to the higher dissociation efficiency which can be increased by modifying the chemical composition. Charge collection efficiency also affects the photocurrent in the photovoltaic materials which further depends on carrier mobility, carrier lifetime, degree of screening, and film thickness. Recently, importance of bulk photovoltaic effect in solar cell application is realized in perovskite solar cells based on halides. Perovskite solar cells (PSC) which are third-generation emerging

Solar energy harvesting with ferroelectric materials

77

technology have come up as an alternative to silicon-based photovoltaic materials because of their distinctive features and low cost. Regardless of sudden improved efficiency from 3.8% to 22.1% in the perovskite solar cell, the key problem with the PSC is their less stability, limited service life, and material toxicity. Considering the material toxicity, several lead substitution efforts have been made and suitable materials have been designed toward high power conversion efficiency, low cost, and more stability. Some of the materials which may complement the lead-based PV are a tin-based halide, cesium tin iodides, methylammonium tin iodides, formamidinium tin iodides, etc. Transition metal oxides with their superior stability and susceptibility to external stimuli toward tailored functionalities are also coming up in big way as potential ferroelectric photovoltaic materials. In the next section, photochemical conversion of solar energy from the perspective of dissociation of water for production of hydrogen is discussed. As far as materials requirement for efficient photoelectrochemical solar water splitting is concerned, both bandgap and band edge positions are important. For efficient solar water splitting, bandgap for semiconducting materials should be within the range, 2–3.2 eV. In addition, the conduction band (CB) edge must be located at more negative to 0 V H2/H+ redox potential vs normal hydrogen electrode (NHE). Similarly, the valence band (VB) edge position should be more positive than 1.23 eV O2/H2O redox potential vs NHE for oxygen generation at the photo-anode. Ferroic materials with their intrinsic stability in corrosive electrolytic condition along with favorable band edge alignment of electrode material with respect to decomposition potential promote stable and efficient photoelectrochemical devices.

References [1] N. Armaroli, V. Balzani, The future of energy supply: challenges and opportunities, Angew. Chem. Int. Ed. 46 (1–2) (2007) 52–66, https://doi.org/10.1002/anie. 200602373. [2] A.A. Golestaneh, Study of thermoelectric conversion using some ferroelectric materials, Ferroelectrics 38 (1) (1981) 984, https://doi.org/10.1080/00150198108209601. [3] S. Pandya, J. Wilbur, J. Kim, R. Gao, A. Dasgupta, C. Dames, L.W. Martin, Pyroelectric energy conversion with large energy and power density in relaxor ferroelectric thin films, Nat. Mater. 17 (5) (2018) 432–438, https://doi.org/10.1038/s41563-018-0059-8. [4] T. Saito, Y. Kawai, T. Ono, Thermoelectric power generator based on a ferroelectric material combined with a bimetal thermostat, in: 2013 Transducers & Eurosensors XXVII: The 17th International Conference on Solid-State Sensors, Actuators and Microsystems (Transducers & Eurosensors XXVII), 2013, pp. 2280–2283, https://doi. org/10.1109/Transducers.2013.6627260. [5] P. Wurfel, The chemical potential of radiation, J. Phys. C Solid State Phys. 15 (18) (1982) 3967–3985, https://doi.org/10.1088/0022-3719/15/18/012. [6] A.M. Glass, D. von der Linde, T.J. Negran, High-voltage bulk photovoltaic effect and the photorefractive process in LiNbO3, Appl. Phys. Lett. 25 (4) (1974) 233–235, https://doi. org/10.1063/1.1655453. [7] W.T.H. Koch, R. Munser, W. Ruppel, P. W€urfel, Bulk photovoltaic effect in BaTiO3, Solid State Commun. 17 (7) (1975) 847–850, https://doi.org/10.1016/0038-1098(75) 90735-8.

78

Ferroelectric Materials for Energy Harvesting and Storage

[8] W.T.H. Koch, R. Munser, W. Ruppel, P. W€urfel, Anomalous photovoltage in BaTiO3, Ferroelectrics 13 (1) (1976) 305–307, https://doi.org/10.1080/00150197608236596. [9] M. Simon, S. Wevering, K. Buse, E. Kr€atzig, The bulk photovoltaic effect of photorefractive: Fe crystals at high light intensities, J. Phys. D. Appl. Phys. 30 (1) (1997) 144–149, https://doi.org/10.1088/0022-3727/30/1/018. [10] K.T. Butler, J.M. Frost, A. Walsh, Ferroelectric materials for solar energy conversion: photoferroics revisited, Energy Environ. Sci. 8 (3) (2015) 838–848, https://doi.org/ 10.1039/C4EE03523B. [11] Y. Yuan, Z. Xiao, B. Yang, J. Huang, Arising applications of ferroelectric materials in photovoltaic devices, J. Mater. Chem. A 2 (17) (2014) 6027–6041, https://doi.org/ 10.1039/C3TA14188H. [12] T.-Y. Cai, S.-C. Liu, S. Ju, C.-Y. Liu, G.-Y. Guo, Multiferroic double perovskites ScFe 1-x Cr x O3 (1/6 x  5/6) for highly efficient photovoltaics and spintronics, Phys. Rev. Appl. 8 (3) (2017) 34034, https://doi.org/10.1103/PhysRevApplied.8.034034. [13] R. Nechache, C. Harnagea, S. Li, L. Cardenas, W. Huang, J. Chakrabartty, F. Rosei, Bandgap tuning of multiferroic oxide solar cells, Nat. Photonics 9 (1) (2014) 61–67, https://doi.org/10.1038/nphoton.2014.255. [14] V. Balzani, A. Credi, M. Venturi, Photochemical conversion of solar energy, ChemSusChem 1 (1–2) (2008) 26–58, https://doi.org/10.1002/cssc.200700087. [15] D. Ginley, M.A. Green, R. Collins, Solar energy conversion toward 1 terawatt, MRS Bull. 33 (4) (2008) 355–364, https://doi.org/10.1557/mrs2008.71. [16] K. Branker, M.J.M. Pathak, J.M. Pearce, A review of solar photovoltaic levelized cost of electricity, Renew. Sust. Energ. Rev. 15 (9) (2011) 4470–4482, https://doi.org/10.1016/j. rser.2011.07.104. [17] B. Parida, S. Iniyan, R. Goic, A review of solar photovoltaic technologies, Renew. Sust. Energ. Rev. 15 (3) (2011) 1625–1636, https://doi.org/10.1016/j.rser.2010.11.032. [18] J. Nelson, The Physics of Solar Cells, first ed., ICP, London, 2003. [19] P.S. Brody, Large polarization-dependent photovoltages in ceramic BaTiO3 + 5 wt.% CaTiO3, Solid State Commun. 12 (7) (1973) 673–676, https://doi.org/10.1016/00381098(73)90310-4. [20] V.M. Fridkin, Photoferroelectrics, Springer, 1979. [21] P.S. Brody, Semiconductor-ferroelectric nonvolatile memory using anomalous high photovoltages in ferroelectric ceramics, Appl. Phys. Lett. 38 (3) (1981) 153–155, https://doi.org/10.1063/1.92283. [22] M. Ichiki, H. Furue, T. Kobayashi, R. Maeda, Y. Morikawa, T. Nakada, K. Nonaka, Photovoltaic properties of (Pb,La)(Zr,Ti)O3 films with different crystallographic orientations, Appl. Phys. Lett. 87 (22) (2005) 222903, https://doi.org/10.1063/1.2128479. [23] M. Ichiki, Y. Morikawa, Y. Mabune, T. Nakada, Electrical properties of photovoltaic lead lanthanum zirconate titanate in an electrostatic-optical motor application, J. Eur. Ceram. Soc. 24 (6) (2004) 1709–1714, https://doi.org/10.1016/S0955-2219(03)00474-6. [24] S.Y. Yang, J. Seidel, S.J. Byrnes, P. Shafer, C.-H. Yang, M.D. Rossell, … R. Ramesh, Above-bandgap voltages from ferroelectric photovoltaic devices, Nat. Nanotechnol. 5 (2010) 143. Retrieved from, https://doi.org/10.1038/nnano.2009.451. [25] M. Qin, K. Yao, Y.C. Liang, S. Shannigrahi, Thickness effects on photoinduced current in ferroelectric (Pb0.97La0.03)(Zr0.52Ti0.48)O3 thin films, J. Appl. Phys. 101 (1) (2007) 14104, https://doi.org/10.1063/1.2405732. [26] W. Ji, K. Yao, Y.C. Liang, Bulk photovoltaic effect at visible wavelength in epitaxial ferroelectric BiFeO3 thin films, Adv. Mater. 22 (15) (2010) 1763–1766, https://doi. org/10.1002/adma.200902985.

Solar energy harvesting with ferroelectric materials

79

[27] M.A. Green, A. Ho-Baillie, H.J. Snaith, The emergence of perovskite solar cells, Nat. Photonics 8 (2014) 506. Retrieved from, https://doi.org/10.1038/nphoton.2014.134. [28] H.S. Jung, N.-G. Park, Perovskite solar cells: from materials to devices, Small 11 (1) (2015) 10–25, https://doi.org/10.1002/smll.201402767. [29] F. Zheng, H. Takenaka, F. Wang, N.Z. Koocher, A.M. Rappe, First-principles calculation of the bulk photovoltaic effect in CH3NH3PbI3 and CH3NH3PbI3–xClx, J. Phys. Chem. Lett. 6 (1) (2015) 31–37, https://doi.org/10.1021/jz502109e. [30] J.E. Sipe, A.I. Shkrebtii, Second-order optical response in semiconductors, Phys. Rev. B 61 (8) (2000) 5337–5352, https://doi.org/10.1103/PhysRevB.61.5337. [31] F. Wang, I. Grinberg, L.A.I. Jiang, S.M. Young, P.K. Davies, A.M. Rappe, Materials design of visible-light ferroelectric photovoltaics from first principles, Ferroelectrics 483 (2015) 1–12, https://doi.org/10.1080/00150193.2015.1058096. [32] A. Bihari Swain, D. Murali, B.R.K. Nanda, P. Murugavel, Large bulk photovoltaic response by symmetry-breaking structural transformation in ferroelectric [Ba (Zr0.2Ti0.8)O3] 0.5[(Ba0.7Ca0.3)TiO3]0.5, Phys. Rev. Appl. 11 (4) (2019) 44007, https://doi.org/10.1103/PhysRevApplied.11.044007. [33] S.I. Karabekian, S.G. Odoulov, Bulk photovoltaic effect in LiTaO3: Fe crystals, Phys. Status Solidi B 169 (2) (1992) 529–535, https://doi.org/10.1002/pssb.2221690224. [34] G. Chanussot, A.M. Glass, A bulk photovoltaic effect due to electron-phonon coupling in polar crystals, Phys. Lett. A 59 (5) (1976) 405–407, https://doi.org/10.1016/0375-9601 (76)90427-8. [35] V.M. Fridkin, Bulk photovoltaic effect in noncentrosymmetric crystals, Crystallogr. Rep. 46 (2001) 654–658. [36] A. Melville, Y. Chu, C. Yang, J.L. Musfeldt, D.G. Schlom, J.W. Ager, R. Ramesh, Photovoltaic effects in BiFeO3, Appl. Phys. Lett. 95 (2009) 062909, https://doi.org/ 10.1063/1.3204695. [37] R. Von Baltz, W. Kraut, Theory of the bulk photovoltaic effect in pure crystals, Phys. Rev. B 23 (10) (1981) 5590–5596, https://doi.org/10.1103/PhysRevB.23.5590. [38] S.M. Young, A.M. Rappe, First principles calculation of the shift current photovoltaic effect in ferroelectrics, Phys. Rev. Lett. 109 (11) (2012) 1–5, https://doi.org/10.1103/ PhysRevLett.109.116601. [39] S.M. Young, F. Zheng, A.M. Rappe, First-principles calculation of the bulk photovoltaic effect in bismuth ferrite, Phys. Rev. Lett. 109 (23) (2012) 1–5, https://doi.org/10.1103/ PhysRevLett.109.236601. [40] R.D. King-Smith, D. Vanderbilt, Theory of polarization of crystalline solids, Phys. Rev. B 47 (3) (1993) 1651–1654, https://doi.org/10.1103/PhysRevB.47.1651. [41] M. Nakamura, S. Horiuchi, F. Kagawa, N. Ogawa, T. Kurumaji, Y. Tokura, M. Kawasaki, Shift current photovoltaic effect in a ferroelectric charge-transfer complex, Nature Commun. 8 (1) (2017) 1–5, https://doi.org/10.1038/s41467-017-00250-y. [42] J. Seidel, D. Fu, S.-Y. Yang, E. Alarco´n-Llado´, J. Wu, R. Ramesh, J.W. Ager, Efficient photovoltaic current generation at ferroelectric domain walls, Phys. Rev. Lett. 107 (12) (2011) 126805, https://doi.org/10.1103/PhysRevLett.107.126805. [43] M. Alexe, D. Hesse, Tip-enhanced photovoltaic effects in bismuth ferrite, Nat. Commun. 2 (2011) 255–256, https://doi.org/10.1038/ncomms1261. [44] T. Choi, S. Lee, Y.J. Choi, V. Kiryukhin, S.-W. Cheong, Switchable ferroelectric diode and photovoltaic effect in BiFeO3, Science 324 (5923) (2009) 63–66, https://doi.org/ 10.1126/science.1168636. [45] J. Zhang, X. Su, M. Shen, Z. Dai, L. Zhang, X. He, W. Cheng, Enlarging photovoltaic effect: combination of classic photoelectric and ferroelectric photovoltaic effects, Sci. Rep. 3 (2013) 1–6, https://doi.org/10.1038/srep02109.

80

Ferroelectric Materials for Energy Harvesting and Storage

[46] H.T. Yi, T. Choi, S.G. Choi, Y.S. Oh, S. Cheong, Mechanism of the Switchable Photovoltaic Effect in Ferroelectric BiFeO3, Adv. Mater. 23 (2011) 3403–3407. [47] C.H. Ahn, K.M. Rabe, J. Triscone, Ferroelectricity at the nanoscale: local polarization in oxide thin films and heterostructures, Science 303 (5657) (2004) 488–491. [48] M. Qin, K. Yao, Y.C. Liang, Photo induced current in (Pb0.97La0.03) (Zr 0.52Ti0.48)O3 thin films of different thicknesses, Integr. Ferroelectr. 88 (1) (2007) 58–67, https://doi. org/10.1080/10584580601098647. [49] M. Qin, K. Yao, Y.C. Liang, Photovoltaic mechanisms in ferroelectric thin films with the effects of the electrodes and interfaces, Appl. Phys. Lett. 95 (2) (2009) 22912, https://doi. org/10.1063/1.3182824. [50] W. Dong, Y. Guo, B. Guo, H. Liu, H. Li, H. Liu, Photovoltaic properties of BiFeO3 thin film capacitors by using Al-doped zinc oxide as top electrode, Mater. Lett. 91 (2013) 359–361, https://doi.org/10.1016/j.matlet.2012.10.031. [51] B. Chen, Z. Zuo, Y. Liu, Q.-F. Zhan, Y. Xie, H. Yang, … R.-W. Li, Tunable photovoltaic effects in transparent Pb(Zr0.53, Ti0.47)O3 capacitors, Appl. Phys. Lett. 100 (17) (2012) 173903, https://doi.org/10.1063/1.4709406. [52] J. He, C. Franchini, J.M. Rondinelli, Ferroelectric oxides with strong visible-light absorption from charge ordering, Chem. Mater. 29 (6) (2017) 2445–2451. [53] Y. Jiang, H. Ning, J. Yu, Optical bandgap tuning of ferroelectric semiconducting BiFeO3-based oxide perovskites via chemical substitution for photovoltaics, AIP Adv. 8 (12) (2018) 125334, https://doi.org/10.1063/1.5053634. [54] C. Nayek, M. Al-Akhras, I.M. Obaidat, Tuning of the optical band-gap of rare earth doped BiFeO3 submicron particles for solar cell applications, in: 2018 5th International Conference on Renewable Energy: Generation and Applications (ICREGA), 2018, pp. 147–149, https://doi.org/10.1109/ICREGA.2018.8337582. [55] R.J. Cottier, N.A. Steinle, D.A. Currie, N. Theodoropoulou, Band gap tuning of epitaxial SrTiO3-δ/Si(001) thin films through strain engineering, Appl. Phys. Lett. 107 (22) (2015) 221601, https://doi.org/10.1063/1.4936608. [56] W. Zhou, Y. Liu, Y. Yang, P. Wu, Band gap engineering of SnO2 by epitaxial strain: experimental and theoretical investigations, J. Phys. Chem. C 118 (12) (2014) 6448–6453, https://doi.org/10.1021/jp500546r. [57] R. Prasanna, A. Gold-Parker, T. Leijtens, B. Conings, A. Babayigit, H.-G. Boyen, … M. D. McGehee, Band gap tuning via lattice contraction and octahedral tilting in perovskite materials for photovoltaics, J. Am. Chem. Soc. 139 (32) (2017) 11117–11124, https://doi. org/10.1021/jacs.7b04981. [58] Y. Liang, G. Shao, First principles study for band engineering of KNbO3 with 3d transition metal substitution, RSC Adv. 9 (13) (2019) 7551–7559, https://doi.org/10.1039/ C9RA00289H. [59] W.S. Choi, M.F. Chisholm, D.J. Singh, T. Choi, G.E. Jellison Jr., H.N. Lee, Wide bandgap tunability in complex transition metal oxides by site-specific substitution, Nature Commun. 3 (2012) 689. [60] G. Zhang, H. Wu, G. Li, Q. Huang, C. Yang, F. Huang, … J. Lin, New high Tc multiferroics KBiFe2O5 with narrow band gap and promising photovoltaic effect, Sci. Rep. 3 (2013) 1265. [61] H. An, J.Y. Han, B. Kim, J. Song, S.Y. Jeong, C. Franchini, … S. Lee, Large enhancement of the photovoltaic effect in ferroelectric complex oxides through bandgap reduction, Scientific Reports 6 (2016) 28313. [62] I. Grinberg, D.V. West, M. Torres, G. Gou, D.M. Stein, L. Wu, … A.M. Rappe, Perovskite oxides for visible-light-absorbing ferroelectric and photovoltaic materials, Nature 503 (2013) 509.

Solar energy harvesting with ferroelectric materials

81

[63] H. Matsuo, Y. Noguchi, M. Miyayama, Gap-state engineering of visible-light-active ferroelectrics for photovoltaic applications, Nature Commun. 8 (1) (2017) 207, https:// doi.org/10.1038/s41467-017-00245-9. [64] S. Pal, A.B. Swain, P.P. Biswas, D. Murali, A. Pal, B.R.K. Nanda, P. Murugavel, Giant photovoltaic response in band engineered ferroelectric perovskite, Sci. Rep. 8 (1) (2018) 8005, https://doi.org/10.1038/s41598-018-26205-x. [65] S.R. Basu, L.W. Martin, Y.H. Chu, M. Gajek, R. Ramesh, R.C. Rai, Photoconductivity in BiFeO3 thin films, Appl. Phys. Lett. 92 (2008) 091905, https://doi.org/10.1063/ 1.2887908. [66] A. Dhar, A. Mansingh, Optical properties of reduced lithium niobate single crystals, J. Appl. Phys. 68 (11) (1990) 5804–5809, https://doi.org/10.1063/1.346951. [67] M.S. Dresselhaus, Solid State Physics Part II Optical Properties of Solids, vol. 6, 2001, http://citeseerx.ist.psu.edu/viewdoc/summary?doi¼10.1.1.691.9232. [68] A. Dogan, A.V. Prasadarao, K. Uchino, P. Poosanaas, S. Komarneni, Photostriction of sol–gel processed PLZT ceramics, J. Electroceram. 1 (1) (1997) 105–111, https://doi. org/10.1023/A:1009906600737. [69] P. Poosanaas, K. Uchino, Photostrictive effect in lanthanum-modified lead zirconate titanate ceramics near the morphotropic phase boundary, Mater. Chem. Phys. 61 (1) (1999) 36–41. [70] S. Shannigrahi, K. Yao, Effects of WO3 dopant on the structure and electrical properties of Pb0.97La0.03(Zr0.52Ti0.48)O3 thin films, Appl. Phys. Lett. 86 (9) (2005) 92901, https://doi.org/10.1063/1.1870125. [71] M. Alexe, Local mapping of generation and recombination lifetime in BiFeO3 single crystals by scanning probe photoinduced transient spectroscopy, Nano Lett. 2012 (12) (2012) 2193–2198, https://doi.org/10.1021/nl300618e. [72] M. Ichiki, R. Maeda, Y. Morikawa, Y. Mabune, T. Nakada, K. Nonaka, Photovoltaic effect of lead lanthanum zirconate titanate in a layered film structure design, Appl. Phys. Lett. 84 (2004) 395. (2004), https://doi.org/10.1063/1.1641528. [73] Y. Zang, D. Xie, X. Wu, Y. Chen, Y. Lin, M. Li, … D. Plant, Enhanced photovoltaic properties in graphene/polycrystalline BiFeO3/Pt heterojunction structure, Appl. Phys. Lett. 99 (13) (2011) 2011–2014, https://doi.org/10.1063/1.3644134. [74] J. Burschka, N. Pellet, S.-J. Moon, R. Humphry-Baker, P. Gao, M.K. Nazeeruddin, M. Gr€atzel, Sequential deposition as a route to high-performance perovskite-sensitized solar cells, Nature 499 (2013) 316. Retrieved from, https://doi.org/10.1038/nature12340. [75] M.M. Lee, J. Teuscher, T. Miyasaka, T.N. Murakami, H.J. Snaith, Efficient hybrid solar cells based on meso-superstructured organometal halide perovskites, Science 338 (6107) (2012) 643–647, https://doi.org/10.1126/science.1228604. [76] A. Mei, X. Li, L. Liu, Z. Ku, T. Liu, Y. Rong, … H. Han, A hole-conductor–free, fully printable mesoscopic perovskite solar cell with high stability, Science 345 (6194) (2014) 295–298, https://doi.org/10.1126/science.1254763. [77] H. Zhou, Q. Chen, G. Li, S. Luo, T. Song, H.-S. Duan, … Y. Yang, Interface engineering of highly efficient perovskite solar cells, Science 345 (6196) (2014) 542–546, https://doi. org/10.1126/science.1254050. [78] J. Cui, H. Yuan, J. Li, X. Xu, Y. Shen, H. Lin, M. Wang, Recent progress in efficient hybrid lead halide perovskite solar cells, Sci. Technol. Adv. Mater. 16 (3) (2015) 36004, https://doi.org/10.1088/1468-6996/16/3/036004. [79] Z. Shi, J. Guo, Y. Chen, Q. Li, Y. Pan, H. Zhang, … W. Huang, Lead-free organic–inorganic hybrid perovskites for photovoltaic applications: recent advances and perspectives, Adv. Mater. 29 (16) (2017) 1605005, https://doi.org/10.1002/adma.201605005.

82

Ferroelectric Materials for Energy Harvesting and Storage

[80] S. Yang, W. Fu, Z. Zhang, H. Chen, C.-Z. Li, Recent advances in perovskite solar cells: efficiency, stability and lead-free perovskite, J. Mater. Chem. A 5 (23) (2017) 11462–11482, https://doi.org/10.1039/C7TA00366H. [81] A. Kojima, K. Teshima, Y. Shirai, T. Miyasaka, Organometal halide perovskites as visible-light sensitizers for photovoltaic cells, J. Am. Chem. Soc. 131 (17) (2009) 6050–6051, https://doi.org/10.1021/ja809598r. [82] J.-H. Im, C.-R. Lee, J.-W. Lee, S.-W. Park, N.-G. Park, 6.5% efficient perovskite quantum-dot-sensitized solar cell, Nanoscale 3 (10) (2011) 4088–4093, https://doi.org/ 10.1039/C1NR10867K. [83] W.S. Yang, B.-W. Park, E.H. Jung, N.J. Jeon, Y.C. Kim, D.U. Lee, … S.I. Seok, Iodide management in formamidinium-lead-halide–based perovskite layers for efficient solar cells, Science 356 (6345) (2017) 1376–1379, https://doi.org/10.1126/science. aan2301. [84] X. Dai, Y. Zhang, H. Shen, Q. Luo, X. Zhao, J. Li, H. Lin, Working from both sides: composite metallic semitransparent top electrode for high performance perovskite solar cells, ACS Appl. Mater. Interfaces 8 (7) (2016) 4523–4531. [85] J.H. Han, X.W. Yin, H. Nan, Y. Zhou, Z.B. Yao, J.B. Li, … H. Lin, Enhancing the performance of perovskite solar cells by hybridizing SnS quantum dots with CH 3 NH 3 PbI 3, Small 13 (2017) 1700953, https://doi.org/10.1002/smll.201700953. [86] E. Li, Y. Guo, T. Liu, W. Hu, N. Wang, H. He, H. Lin, Preheating-assisted deposition of solution-processed perovskite layer for an efficiency-improved inverted planar composite heterojunction solar cell, RSC Adv. 6 (37) (2016) 30978–30985, https://doi.org/ 10.1039/C5RA27434F. [87] Q. Luo, H. Ma, Y. Zhang, X. Yin, Z. Yao, N. Wang, … H. Lin, Cross-stacked superaligned carbon nanotube electrodes for efficient hole conductor-free perovskite solar cells, J. Mater. Chem. A 4 (15) (2016) 5569–5577, https://doi.org/10.1039/ C6TA01715K. [88] Q. Luo, Y. Zhang, C. Liu, J. Li, N. Wang, H. Lin, Iodide-reduced graphene oxide with dopant-free spiro-OMeTAD for ambient stable and high-efficiency perovskite solar cells, J. Mater. Chem. A 3 (31) (2015) 15996–16004, https://doi.org/10.1039/ C5TA02710A. [89] Z. Yao, W. Wang, H. Shen, Y. Zhang, Q. Luo, X. Yin, … J. Li, CH 3 NH 3 PbI 3 grain growth and interfacial properties in meso-structured perovskite solar cells fabricated by two-step deposition, Sci. Technol. Adv. Mater. 18 (1) (2017) 1–10, https://doi.org/ 10.1080/14686996.2017.1298974. [90] X. Yin, Z. Yao, Q. Luo, X. Dai, Y. Zhou, Y. Zhang, … H. Lin, High Efficiency inverted planar perovskite solar cells with solution- processed NiO, ACS Appl. Mater. Interfaces 9 (3) (2017) 2439–2448, https://doi.org/10.1021/acsami.6b13372. [91] X. Zhao, H. Shen, Y. Zhang, X. Li, X. Zhao, M. Tai, … H. Lin, Aluminum-doped zinc oxide as highly stable electron collection layer for perovskite solar cells, ACS Appl. Mater. Interfaces 8 (12) (2016) 7826–7833, https://doi.org/10.1021/acsami.6b00520. [92] Q. Wang, A. Abate, Strategies toward stable perovskite solar cells, Adv. Mater. Interfaces 5 (22) (2018) 1800264, https://doi.org/10.1002/admi.201800264. [93] N.K. Noel, S.D. Stranks, A. Abate, C. Wehrenfennig, S. Guarnera, A. Haghighirad, … H. J. Snaith, Lead-free organic-inorganic tin halide perovskites for photovoltaic applications, Energy Environ. Sci. 7 (2014) 3061–3068, https://doi.org/10.1039/c4ee01076k. [94] T. Krishnamoorthy, H. Ding, C. Yan, L. Leong, T. Baikie, Lead-free germanium iodide perovskite materials for photovoltaic applications, J. Mater. Chem. A 3 (2015) 23829–23832, https://doi.org/10.1039/c5ta05741h.

Solar energy harvesting with ferroelectric materials

83

[95] M.R. Filip, F. Giustino, Computational screening of homovalent lead substitution in organic inorganic halide perovskites, J. Phys. Chem. 120 (1) (2016) 166–173, https://doi.org/10.1021/acs.jpcc.5b11845. [96] J. Navas, A. Sa´nchez-Coronilla, J.J. Gallardo, N. Cruz Herna´ndez, J.C. Pin˜ero, R. Alca´ntara, … J. Martı´n-Calleja, New insights into organic–inorganic hybrid perovskite CH3NH3PbI3 nanoparticles. An experimental and theoretical study of doping in Pb2+ sites with Sn2+, Sr2 +, Cd2 + and Ca2 +, Nanoscale 7 (14) (2015) 6216–6229, https://doi.org/10.1039/C5NR00041F. [97] G. Volonakis, M.R. Filip, A.A. Haghighirad, N. Sakai, B. Wenger, H.J. Snaith, F. Giustino, Lead-free halide double perovskites via heterovalent substitution of noble metals, J. Phys. Chem. Lett. 7 (7) (2016) 1254–1259, https://doi.org/10.1021/acs. jpclett.6b00376. [98] F. Giustino, H.J. Snaith, Toward lead-free perovskite solar cells, ACS Energ. Lett. 1 (6) (2016) 1233–1240, https://doi.org/10.1021/acsenergylett.6b00499. [99] Q. Zhang, F. Hao, J. Li, Y. Zhou, Y. Wei, H. Lin, Perovskite solar cells: must lead be replaced—and can it be done? Sci. Technol. Adv. Mater. 19 (1) (2018) 425–442, https://doi.org/10.1080/14686996.2018.1460176. [100] Q. Zhang, F. Hao, J. Li, Y. Zhou, Y. Wei, H. Lin, Perovskite solar cells: must lead be replaced—and can it be done? Sci. Technol. Adv. Mater. 19 (1) (2018) 1–18, https:// doi.org/10.1080/14686996.2018.1460176. [101] A. Amat, E. Mosconi, E. Ronca, C. Quarti, P. Umari, K. Nazeeruddin, … F.D. Angelis, Cation-induced band-gap tuning in organohalide perovskites: interplay of spin-orbit coupling and octahedra tilting, Nano Lett. 14 (6) (2014) 3608–3616, https://doi.org/10.1021/ nl5012992. [102] T.M. Koh, K. Fu, Y. Fang, S. Chen, T.C. Sum, N. Mathews, Formamidinium-containing metal-halide: an alternative material for near-IR absorption perovskite solar cells, J. Phys. Chem. 118 (30) (2014) 16458–16462, https://doi.org/10.1021/jp411112k. [103] S.R. Basu, L.W. Martin, Y.H. Chu, M. Gajek, R. Ramesh, R.C. Rai, … J.L. Musfeldt, Photoconductivity in BiFeO3 thin films, Appl. Phys. Lett. 92 (9) (2008) 91905, https://doi.org/10.1063/1.2887908. [104] L. Wang, H. Ma, L. Chang, C. Ma, G. Yuan, J. Wang, T. Wu, Ferroelectric BiFeO3 as an oxide dye in highly tunable mesoporous all-oxide photovoltaic heterojunctions, Small 13 (1) (2017) 1602355, https://doi.org/10.1002/smll.201602355. [105] R. Nechache, C. Harnagea, S. Licoccia, E. Traversa, A. Ruediger, A. Pignolet, F. Rosei, Photovoltaic properties of Bi2FeCrO6 epitaxial thin films, Appl. Phys. Lett. 98 (20) (2011) 202902, https://doi.org/10.1063/1.3590270. [106] A. Zenkevich, Y. Matveyev, K. Maksimova, R. Gaynutdinov, A. Tolstikhina, V. Fridkin, Giant bulk photovoltaic effect in thin ferroelectric BaTiO3 films, Phys. Rev. B 90 (16) (2014) 161409, https://doi.org/10.1103/PhysRevB.90.161409. [107] J. Qin, Z. Zhang, W. Shi, Y. Liu, H. Gao, Y. Mao, Enhanced performance of perovskite solar cells by using ultrathin BaTiO3 interface modification, ACS Appl. Mater. Interfaces 10 (2018) 36067–36074, https://doi.org/10.1021/acsami.8b16358. [108] F. Yang, L. Yang, C. Ai, P. Xie, S. Lin, C.-Z. Wang, X. Lu, Tailoring bandgap of perovskite BaTiO3 by transition metals Co-doping for visible-light photoelectrical applications: a first-principles study, Nanomaterials 8 (7) (7) (2018) 455. [109] R.D. Tentu, S. Basu, Photocatalytic water splitting for hydrogen production, Curr. Opin. Electrochem. 5 (1) (2017) 56–62, https://doi.org/10.1016/j.coelec.2017.10.019. [110] A. Vilanova, T. Lopes, C. Spenke, M. Wullenkord, A. Mendes, Optimized photoelectrochemical tandem cell for solar water splitting, Energy Storage Mater. 13 (2018) 175–188, https://doi.org/10.1016/j.ensm.2017.12.017.

84

Ferroelectric Materials for Energy Harvesting and Storage

[111] A.A. Ismail, D.W. Bahnemann, Photochemical splitting of water for hydrogen production by photocatalysis: a review, Sol. Energy Mater. Sol. Cells 128 (2014) 85–101, https://doi.org/10.1016/j.solmat.2014.04.037. [112] C. Jiang, S.J.A. Moniz, A. Wang, T. Zhang, J. Tang, Photoelectrochemical devices for solar water splitting—materials and challenges, Chem. Soc. Rev. 46 (15) (2017) 4645–4660, https://doi.org/10.1039/C6CS00306K. [113] L. Li, P.A. Salvador, G.S. Rohrer, Photocatalysts with internal electric fields, Nanoscale 6 (1) (2014) 24–42, https://doi.org/10.1039/C3NR03998F. [114] K. Sivula, R. van de Krol, Semiconducting materials for photoelectrochemical energy conversion, Nat. Rev. Mater. 1 (2016) 15010. Retrieved from, https://doi.org/10. 1038/natrevmats.2015.10. [115] Hara, M.; Kondo, T.; Komoda, M.; Ikeda, S.; Shinohara, K.; Tanaka, A.; Kondo, J.N.; Domen, K. Cu2O as a photocatalyst for overall water splitting under visible light irradiation, Chem. Commun. 1998, 357–358. [116] Chauhan, D.; Satsangi, V.R.; Dass, S.; Shrivastav, R. Preparation and characterization of nanostructured CuO thin films for photoelectrochemical splitting of water, Bull. Mater. Sci. 2006, 29, 709–716. [117] Gupta, R.K.; Ghosh, K.; Kahol, P.K. Fabrication and characterization of NiO/ZnO p-n junctions by pulsed laser deposition, Physica E 2009, 41, 617–620. [118] Reddy, K.H.; Parida, K. Fabrication, characterization, and photoelectrochemical properties of Cudoped PbTiO3 and its hydrogen production activity, ChemCatChem 2013, 5, 3812–3820. [119] Zhang, M.; Jiang, C.X.; Dong, W.; Zheng, F.G.; Fang, L.; Su, X.D.; Shen, M.R. Composition dependence of the photochemical reduction of Ag+ by as-grown Pb (ZrxTi1x) O3 films on indium tin oxide electrode, Appl. Phys. Lett. 2013, 103, 1–4. [120] Glass, A.M.; von der Linde, D.; Auston, D.H.; Negran, T.J. Excited state polarization, bulk photovoltaic effect and the photorefractive effect in electrically polarized media, J. Electron. Mater. 1975, 4, 915–943. [121] Yang, W.G.; Yu, Y.H.; Starr, M.B.; Yin, X.; Li, Z.D.; Kvit, A.; Wang, S.F.; Zhao, P.; Wang, X.D. Ferroelectric polarization-enhanced photoelectrochemical water splitting in TiO2 -BaTiO3 core-shell nanowire photoanodes, Nano Lett. 2015, 15, 7574–7580. [122] Yilmaz, P.; Yeo, D.; Chang, H.; Loh, L.; Dunn, S. Perovskite BiFeO3 thin film photocathode performance with visible light activity, Nanotechnology 2016, 27, 1–9. [123] Song, J.; Kim, T.L.; Lee, J.; Cho, S.Y.; Cha, J.; Jeong, S.Y.; An, H.; Kim, W.S.; Jung, Y.S.; Park, J.; et al. Domain-engineered BiFeO3 thin-film photoanodes for highly enhanced ferroelectric solar water splitting, Nano Res. 2018, 11, 642–655.

Harvesting thermal energy with ferroelectric materials

3

Ravi Anant Kishore National Renewable Energy Laboratory, Golden, CO, United States

3.1

Introduction

Pyroelectric energy harvesting is based on phenomenon called pyroelectric effect which originates due to interaction between electrical polarization and temperature change in pyroelectric materials. All ferroelectric materials are pyroelectric, but not all pyroelectric materials are ferroelectric. However, since the most common working materials used for pyroelectric thermal energy harvesting are ferroelectric, the terms “ferroelectric energy harvesting” and “pyroelectric energy harvesting” are often used interchangeably [1]. Pyroelectric energy harvesting is fundamentally different from thermoelectric energy harvesting, which is currently the most popular solid-state thermal to electric energy conversion technique. Thermoelectricity originates due to Seebeck effect, where a voltage difference is generated in response to the temperature difference between two ends of a thermoelectric couple, comprising pand n-types of thermoelectric materials. The thermoelectric energy conversion, therefore, relies on spatial thermal gradient. Pyroelectric energy harvesting, on the other hand, is based on pyroelectric effect, which occurs due to temperature-dependent change in the surface-charge density, temperature-dependent variation in dielectric permittivity, thermally induced strain in piezoelectric materials, or flexoelectric effects (electric polarization induced by a strain gradient) from thermal gradients in certain materials [2]. Pyroelectric energy harvesting, therefore, relies on temporal thermal gradients (time-dependent variation in temperature), making it suitable for energy conversion using heat sources whose temperature continuously fluctuates. Fig. 3.1 schematically depicts the basic differences between thermoelectric and pyroelectric energy harvesting. Outline of this chapter is as follows. Section 3.1 provides a brief introduction of the topic and highlights the key differences between pyroelectric and thermoelectric energy harvesting. Section 3.2 explains the physics behind ferroelectricity, followed by the working principle of ferroelectric thermal energy harvesting in Section 3.3. Sections 3.4 and 3.5, respectively, describe various ferroelectric thermodynamic cycles and devices proposed in the literature. In Section 3.6, I focus on other applications of ferroelectricity, and lastly in Section 3.7, I conclude the discussion by providing summary and future scope.

Ferroelectric Materials for Energy Harvesting and Storage. https://doi.org/10.1016/B978-0-08-102802-5.00003-0 © 2021 Elsevier Ltd. All rights reserved.

86

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 3.1 Fundamental difference between thermoelectric and pyroelectric energy harvesting.

Fig. 3.2 Various forms of polarization: (A) dielectric polarization, (B) paraelectric polarization, and (C) ferroelectric polarization.

3.2

Ferroelectricity

Ferroelectric materials have a spontaneous electric polarization, i.e., they naturally possess dipole moments that add up in the direction normal to a flat surface to provide a net electrical polarization. While most materials can be polarized by applying an external electric field, not all have a spontaneous polarization (nonzero polarization even when the applied field is zero). When induced polarization is linearly proportional to the applied external electric field, it is called dielectric polarization (Fig. 3.2A) and the materials that exhibit this behavior are called dielectrics. The slope of the polarization curve is defined as dielectric permittivity of the material. In some

Harvesting thermal energy with ferroelectric materials

87

materials, however, the relationship between the induced polarization and the external electric field is nonlinear (Fig. 3.2B). Such materials are termed as paraelectric. The electric permittivity of paraelectric materials is a function of the external electric field as the slope of the polarization curve varies with change in electric field. Ferroelectric materials, on the other hand, demonstrate the peculiar polarization curve, which is characterized by a hysteresis loop, as shown in Fig. 3.2C. It can be noted that the electric polarization of such materials is dependent not only on the direction and magnitude of the instantaneous electric field, but also on the history of polarization. The electric polarization of ferroelectric materials has nonzero value when the applied electric field is zero. Also, the direction of the polarization in ferroelectric materials switches when the direction of the applied alternating electric field is reversed. However, reversal is often accompanied by some hysteresis, which leads to the phenomenon of ferroelectric hysteresis, as is shown in Fig. 3.2C. Most ferroelectric materials exhibit a transition temperature (called Curie point), where the spontaneous polarization of a ferroelectric material drops to zero. Increasing the temperature above the Curie point causes the ferroelectric material to transition into a nonferroelectric or paraelectric phase. Fig. 3.3A depicts polarization vs electric field curves of a ferroelectric material before and after the ferroelectric to paraelectric phase transition. Fig. 3.3B shows change in polarization of a ferroelectric material with respect to temperature at different applied electric fields. Blue curve depicts polarization vs temperature relationship, when no electric field is applied. Increasing the temperature reduces the polarization, and when temperate is equal to the Curie temperature, polarization reduces to zero. Curie temperature of the ferroelectric materials is of special interest in the field of ferroelectric thermal energy harvesting because the pyroelectric coefficient (gradient of polarization vs temperature, dP dT ) is high near Curie temperature [3].

Fig. 3.3 (A) Polarization (P) vs applied electric field (E) responses of a ferroelectric material above and below the Curie temperature, TC. (B) Variation of polarization with respect to temperature at different applied electric fields. Figures reconstructed from S.P. Alpay, J. Mantese, S. Trolier-McKinstry, Q. Zhang, R.W. Whatmore, Next-generation electrocaloric and pyroelectric materials for solid-state electrothermal energy interconversion, MRS Bull. 39 (2014) 1099–111.

88

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 3.4 Conceptual model of ferroelectric thermal energy harvesting. Figure reconstructed from S.B. Lang, Pyroelectricity: from ancient curiosity to modern imaging tool, Phys. Today. 58 (2005) 31.

3.3

Working principle of ferroelectric thermal energy harvesting

Fig. 3.4 illustrates the working principle of ferroelectric thermal energy harvesting. The key components include a ferroelectric material, which is placed between the two conductive electrodes of a capacitor. When the ferroelectric material is first placed between the two electrodes, the capacitor charges until the surface charge on the ferroelectric material is neutralized. If the temperature of the material is held constant and the capacitor is connected to an external electric circuit, there is no current in the circuit in the steady state condition (Fig. 3.4A). However, if the ferroelectric material is heated, increase in temperature of the material causes the polarization to decrease, as shown in Fig. 3.4B. Likewise, as shown in Fig. 3.4C, if the ferroelectric material is cooled, the consequent decrease in temperature of the material causes the polarization to increase. The change in temperature of the material, therefore, alters the quantity of bound charges. Consequently, to compensate the change in bound charges, the redistribution of free charges occurs, which results in a current flow, which is termed as ferroelectricity. It can be concluded that an alternating current can be generated by cyclic heating and cooling of a ferroelectric material. This phenomenon is called ferroelectricity and can be used for generating electricity for thermal energy harvesting. The pyroelectric coefficient of a ferroelectric material is defined as the rate of change of spontaneous polarization with respect to the change in temperature, under constant stress and electric field conditions. Mathematically, the pyroelectric coefficient p is expressed as [4]:
 dPs p¼ dT σ,E

(3.1)

where Ps and T denote spontaneous polarization and temperature of the ferroelectric material, respectively. Subscript σ and E, respectively, denote a constant stress and constant electric field conditions.

Harvesting thermal energy with ferroelectric materials

89

If the ferroelectric material is homogeneous, i.e., the pyroelectric coefficient is constant throughout the material, the ferroelectric current ip generated due to temperature change is given as [4, 5]: ip ¼ pA

dT dt

(3.2)

where p denotes pyroelectric coefficient, A is surface area of the ferroelectric material, and dT dt is rate of temperature change. Using Eq. (3.2), the net charge developed on the electrodes of capacitor can be calculated as: ð Q ¼ ip dt ¼ pAΔT

(3.3)

where ΔT is the temperature change. The equivalent capacitance (C), the open-circuit voltage (VOC), and the total stored energy (TE) of the ferroelectric capacitor are given as [6]: C¼

Aεσ33 h

Voc ¼

Q phΔT ¼ σ C ε33

1 2 1 p2 AhðΔT Þ2 TE ¼ CVoc ¼ 2 2 εσ33

(3.4) (3.5)

(3.6)

where, εσ33 is the permittivity parallel to the polarization direction at constant stress and h is thickness of the ferroelectric material.

3.4

Ferroelectric thermodynamic cycles

Fig. 3.5 depicts Carnot cycle and Ericson cycle for ferroelectric thermal energy harvesting. Carnot cycle is an ideal cycle, which consists of two isothermal processes and two adiabatic processes. In Fig. 3.5A, state 1 represents the initial condition, when no external electric field is applied. The material is then exposed to an electric field and the electric field strength is increased isothermally from state 1 to state 2. Increase in electric field strength increases the polarization in the ferroelectric material, causing a decrease in the electrical entropy. This causes release in internal energy of the material in the form of heat, which needs to be absorbed by a heat sink in order to maintain a constant temperature (process 1–2 is isothermal). The applied electric field is further increased adiabatically from state 2 to state 3. Again, increase in electric field strength leads to an increase in electric polarization, causing reduction in the

90

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 3.5 (A) Thermodynamic cycles for ferroelectric thermal energy harvesting. (A) Carnot cycle (B) Ericsson cycle.

internal energy. Since there is no heat interaction with the surrounding, temperature of the working material increases. Therefore, process 2–3 can be called an adiabatic heating process. Next, the applied electric field is isothermally reduced from state 3 to state 4. As electric polarization decreases, the ferroelectric material absorbs heat from a heat source, maintaining a constant temperature. Lastly, the applied electric field is further reduced until it reaches zero (state 1), causing a decrease in the temperature of the material (adiabatic cooling). Electrical work output of the ferroelectric Carnot cycle is given as: þ þ þ WCarnot ¼ P dE ¼ E dP ¼ E p dT

(3.7)

Since processes 1–2 and 3–4 are isothermal and assuming constant pyroelectric coefficient p, WCarnot can be expressed as: WCarnot ¼ p

ð3

ð1 E dT + p

2

E dT

(3.8)

4

The efficiency of ferroelectric Carnot cycle operating between a heat source at temperature Th and a heat sink at temperature Tc is given as: ηcarnot ¼

WCarnot Qin  Qout Tc ¼ ¼1 Qin Qin Th

(3.9)

where, Qin and Qout are the heat in and heat out of the ferroelectric material during heating and cooling processes, respectively. Th and Tc are the temperature of the heat source and heat sink, respectively. The Carnot cycle provides the maximum possible efficiency of a ferroelectric thermal energy harvester, which is used to evaluate relative efficiency of the other

Harvesting thermal energy with ferroelectric materials

91

thermodynamic cycles, such as ferroelectric Ericsson cycle. The ferroelectric Ericsson cycle is depicted in Fig. 3.5B. This cycle consists of two isothermal processes and two constant electric field processes. State 1 represents the initial condition when the ferroelectric material is under zero electric field and near a heat source at constant temperature Th. In the process 1–2, the material is cooled under a zero electric field condition until temperature Tc is achieved. As temperature reduces, the electric polarization increases due to a decrease in the thermal agitation in the material. Electric field is then applied, and it is gradually increased at fixed temperature Tc, until the electric field strength increases to Emax (process 2–3). Increase in electric field further increases the polarization of the material to Pmax. The isothermal process 2–3 is followed by an isofield heating process 3–4, where the ferroelectric material is heated from a lower temperature Tc to a higher temperature Th under a constant electric field (Emax). Increases in temperature of the material reduce its polarization due to increased thermal agitation. Lastly, in the process 4–1, the electric field is isothermally diminished from Emax to zero, while the temperature of the material is maintained at fixed temperature Th. The electric work output of a ferroelectric Ericsson cycle is given as: þ þ þ WEricsson ¼ P dE ¼ E dP ¼ E p dT WEricsson ¼ p

ð2

ð3 E dT + p

1

(3.10)

ð4 E dT + p

ð1 E dT + p

2

3

E dT

(3.11)

4

Since processes 2–3 and 4–1 are isothermal, ð3 p

E dT ¼ 0 ¼ p

2

ð1 E dT

(3.12)

4

Also, since electric field is constant during processes 1–2 and 3–4, WEricsson ¼ p Emin

ð2 dT + p Emax 1

ð2

ð4 dT

(3.13)

3

dT ¼ Tc  Th

(3.14)

dT ¼ Th  Tc

(3.15)

1

ð4 3

WEricsson ¼ p ðEmax  Emin ÞðTh  Tc Þ

(3.16)

Heat inflow from the heat source can be obtained as: Qin ¼ Qout + WEricsson

(3.17)

92

Ferroelectric Materials for Energy Harvesting and Storage

Heat outflow, Qout, is given as Qout ¼ c ðTh  Tc Þ +

ð Tc se dT

(3.18)

Th

where, c is the specific heat of the ferroelectric material at zero electric field and se is the electrical entropy. The first term Ð c (Th  Tc) denotes heat interaction due to sensible cooling and the second term ThTc se dT denotes heat interaction due to changes in electrical entropy. When the applied electric field is small and temperature change is large, the second term can be ignored in comparison to the first term. Qout  c ðTh  Tc Þ

(3.19)

Therefore, Qin  c ðTh  Tc Þ + p ðEmax  Emin ÞðTh  Tc Þ

(3.20)

Thermodynamic efficiency of an Ericsson cycle can then be determined as: ηEricsson ¼

WEricsson p ðEmax  Emin Þ ¼ c + p ðEmax  Emin Þ Qin

(3.21)

If the material properties of the ferroelectric material are temperature-dependent, WEricsson and ηEricsson are given as: WEricsson ¼ ðEmax  Emin Þ

ð Th p dT

(3.22)

Tc

ðEmax  Emin Þ

ηEricsson ¼ ð Th Tc

ð Th

p dT ð Th c dT + ðEmax  Emin Þ p dT Tc

(3.23)

Tc

Fig. 3.6 shows other types of ferroelectric thermodynamic cycle proposed in the literature. Fig. 3.6A shows the simplest one where a ferroelectric material is alternately heated and cooled between temperatures Tc and Th, and the resulting electric current is passed through an electric resistor. This cycle is, thus, termed as a ferroelectric resistive cycle [7–9]. The cycle shown in Fig. 3.6B is a modified resistive cycle where a pair of diodes is used to control the flow of electric current through the load resistor [7, 10]. The cycle shown in Fig. 3.6C is called ferroelectric Stirling cycle as it includes two isothermal and two constant polarization processes [7, 11]. The ideal ferroelectric in an energy harvesting cycle discussed above does not consider hysteresis losses. Fig. 3.7 shows the hysteresis loop of a ferroelectric material at

Harvesting thermal energy with ferroelectric materials

93

Fig. 3.6 Additional ferroelectric thermodynamic cycles, (A) resistive cycle, (B) modified resistive cycle, (C) Stirling cycle.

Fig. 3.7 Olsen cycle for ferroelectric thermal energy harvesting. Figure adapted from R. Kishore, S. Priya, A review on low-grade thermal energy harvesting: materials, methods and devices. Materials 11 (2018) 1433.

heat source temperature Th and heat sink temperature Tc. Area under the curve 10 –20 — 3–4 signifies the electric work using Ericsson cycle. However, it can be observed that, during an energy conversion cycle, the ferroelectric material practically follows the curve 1–2–3–4, instead of 10 –20 —3–4. The thermodynamic cycle defined by curve 1–2–3–4 is termed as modified Ericsson cycle or Olsen cycle [12]. Evidently, Olsen cycle is more realistic than the ideal Carnot and Ericsson cycles. Table 3.1 shows the electrical work calculated for different materials using Olsen cycle. The electric work output of the ferroelectric thermodynamic cycle is directly proportional to the changes in temperature and electrical field. This indicates that the

94

Ferroelectric Materials for Energy Harvesting and Storage

Table 3.1 Available work output, WOlsen, based on Olsen cycle for different materials [3, 6].

PNZST ceramic PNZST ceramic PMN-PT 90/10 ceramic PLZT 8/65/35 ceramic KNTM ceramic BNLT ceramic BNKT ceramic BNK-BST ceramic YBFO thin PZST PZST PZST PZST PZN-4.5PT PZN-5.5PT PMN-10PT PMN-32PT P(VDF-TrFE) 73/27 P(VDF-TrFE) 60/40 P(VDF-TrFE) 60/40 P(VDF-TrFE) 60/40 P(VDF-TrFE) 60/40 P(VDF-TrFE-CFE) 61.3/29.7/9

Tc (°C)

Th (°C)

Elow (MVm21)

Ehigh (MVm21)

Wcycle (Jm23/cycle)

158 145 35

170 175 85

0.4 0.8 0.5

2.8 3.2 3.5

95 300 186

25

160

0.2

7.5

888

140 25 25 20 258 157 145 146 110 100 100 30 80 23 58 67 25 25 0

160 120 110 160 27 177 178 159 170 160 190 80 170 67 77 81 110 120 25

0.1 0.1 0.1 0.1 0.1 0.4 1.2 0 0 0 0 0 0 23 4.1 20.3 20 20 0

5 11.2 5.2 4 4 3.2 3.2 2.9 2.8 2 1.2 3.5 0.9 53 47.2 37.9 50 60 25

629 1146 1986 1523 7570 131 130 100 0.4 217 150 186 100 30 52 130 521 900 50

Data taken from [12–29].

work output of the ferroelectric thermal energy harvester can be enhanced either by increasing the applied electric field (as shown in Fig. 3.8A) or by increasing the temperature change (as shown in Fig. 3.8B). As shown in Fig. 3.8, the electrical work output of the ferroelectric harvester, indicated by the shaded region, can be increased by either increasing the applied electric field or increasing the temperature change.

3.5

Ferroelectric thermal energy harvesters

The concept of ferroelectric thermal energy harvesting was demonstrated several decades ago; however, there are limited number of working devices reported in the literature. Fig. 3.9 depicts an early version of ferroelectric thermal to electric energy converter (or ferroelectric harvester), proposed by Olsen in 1985 [13]. The cyclic

Harvesting thermal energy with ferroelectric materials

95

Fig. 3.8 Electric work output of the ferroelectric thermal energy harvester increases by increasing (A) the applied electric field or (B) temperature change.

Fig. 3.9 Ferroelectric generator proposed by Olsen in 1985. Figure reconstructed from R.B. Olsen, D.A. Bruno, J.M. Briscoe, Pyroelectric conversion cycles, J. Appl. Phys. 58 (1985) 4709–16.

variation in the temperature of a ferroelectric material was achieved by employing an electric heater, a heat exchanger, and a pump drive. The electric heater generates heat, which is eventually rejected into a cooling unit via a heat exchanger. This causes a temperature gradient from top to bottom portions of the ceramic stack and the surrounding fluid. The piston driven by pump causes an oscillatory motion of the fluid, which results in a cyclic variation in temperature of the stacked ferroelectric materials. The ferroelectric harvesters demonstrated by Olsen and coworkers exhibited power output between 1 and 40 mW and conversion efficiency of around 0.4% [21, 30, 31].

96

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 3.10 Resonating bimaterial cantilever beam-shaped ferroelectric harvester. Figure reconstructed from S.R. Hunter, N.V. Lavrik, S. Mostafa, S. Rajic, P.G. Datskos, Review of Pyroelectric Thermal Energy Harvesting and New MEMs-Based Resonant Energy Conversion Techniques. International Society for Optics and Photonics. p. 83770D.

Fig. 3.10 shows another ferroelectric harvester, which consists of a resonating capacitor structure [32]. A ferroelectric material is sandwiched between two metal electrodes and the resulting capacitor structure is clamped at one end like a cantilever beam. The other end of the cantilever beam is attached with two small proof masses to increase the thermal mass of the structure and to enhance thermal contact with the hot and cold surfaces. When the cantilever beam oscillates, the temperature of the ferroelectric material fluctuates causing cyclic variation in polarization. An electric resistor can be attached to the harvester to obtain an alternating current or electric work. Fig. 3.11 shows the electrical characteristics of a ferroelectric harvester that comprised of a commercial transducer (P-876.A11 DuraAct Patch Transducer) as the working material. In this study, a heat lamp was used as the heat source and natural convection with air was utilized as the cooling system. It was noted that when the working material was heated from 310 to 340 K, a positive voltage of 1.5 V was obtained; however, when the material was slowly cooled down to the initial temperature, a negative voltage of 1 V was obtained. Fig. 3.11E depicts the short-circuit current, fluctuating between 1.5 and 0.5 μA. Fig. 3.11F shows that the response time of the demonstrated ferroelectric harvester was around 121 ms. Researchers have also demonstrated ferroelectric thermal energy harvesting using boiling water as the heat source and ambient air as the heat sink [33]. As shown in Fig. 3.12, an output voltage of 0.4–1.6 V was obtained when the temperature of the ferroelectric material was varied from 313 to 333 K. Fig. 3.13 shows the output voltage for a ferroelectric thermal energy harvester powered by body heat. It was observed that when the harvester (kept at room temperature of 300 K) was touched with a finger (at 305 K), a positive peak output voltage of about 0.1 V was obtained, whereas when the finger was removed, a negative output voltage peak of 0.05 V was obtained. It can be noted that the electric power generated by a ferroelectric harvester is dependent on the rate of cyclic variation in temperature of the ferroelectric material. The heat transfer rate between the ferroelectric material and the heat source and the sink often limits the frequency of operation. To improve the heat transfer rate and the oscillating frequency, few researchers have proposed thin film-based ferroelectric

Harvesting thermal energy with ferroelectric materials

97

Temperature (K)

350

Heat lamp

DPT device

340 330 320 310 300

(A)

0

(B)

50

100

150 200 Time (s)

100

150 200 Time (s)

250

300

8 2 Voltage (V)

dT/dt (K/s)

6 4 2

1

0

0 –2

–1 0

50

100

(C)

150 200 Time (s)

250

300

50

250

2

Current (mA)

2

Current (mA)

0

(D)

1

0

1

0 121 ms

–1

(E)

0

50

100

150 200 Time (s)

250

300

56.0

(F)

56.5 Time (s)

57.0

Fig. 3.11 Ferroelectric harvester made up of a commercial transducer as the working material and a heat lamp as the heat source. Figure taken from A. Sultana, M.M. Alam, T.R. Middya, D. Mandal, A pyroelectric generator as a self-powered temperature sensor for sustainable thermal energy harvesting from waste heat and human body heat, Appl. Energy 221 (2018) 299–307.

98

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 3.12 Ferroelectric thermal energy harvesting using water vapor as the heat source. Figure taken from A. Sultana, M.M. Alam, T.R. Middya, D. Mandal, A pyroelectric generator as a self-powered temperature sensor for sustainable thermal energy harvesting from waste heat and human body heat, Appl. Energy 221 (2018) 299–307.

harvester, making it suitable for deploying flexible thermal energy harvesting devices for applications such as wearable devices for human body. Fig. 3.14 shows the ferroelectric harvester based on polyvinylidene fluoride film reported by Leng et al. [34]. The ferroelectric film was oscillated between hot and cold-water flow, resulting in a

Harvesting thermal energy with ferroelectric materials

99

(A)

(C) 0.2

Voltage (V)

Voltage (V)

0.10

0.05

0.00

0.1

0.0

–0.05

(B)

0

20

40

60

80

Time (s)

100

120

0

(D)

20

40

60

80

100

120

Time (s)

Fig. 3.13 Ferroelectric thermal energy harvester powered by body heat. Figure taken from A. Sultana, M.M. Alam, T.R. Middya, D. Mandal, A pyroelectric generator as a self-powered temperature sensor for sustainable thermal energy harvesting from waste heat and human body heat, Appl. Energy 221 (2018) 299–307.

maximum output open-circuit voltage of 192 V and a short-circuit current of 12 mA under a temperature change of 80 °C. The highest output power density was noted to be 14 μW/cm2 or 1.08 W/cm3. Researchers have also reported hybrid thermal harvesters incorporating pyroelectric and piezoelectric effects [35]. Since all ferroelectric materials are piezoelectric, the integration between two effects is relatively easy. One such nanoscale piezoelectric–pyroelectric hybrid device was recently demonstrated by Lee at el. [35]. As shown in Fig. 3.15, the reported nano-generator was comprised of a micro-patterned piezoelectric P(VDF-TrFE) polymer, micropatterned PDMS carbon nanotube (CNTs) composite, and graphene nanosheets. While PDMS-CNT makes the device flexible and stretchable, graphene allows fast thermal response due to its high thermal conductivity. The hybrid nanogenerator was found to exhibit high robustness with up to 30% stretching and generated quite stable piezoelectric and pyroelectric power outputs due to micropattern designing. Fig. 3.16 shows another flexible PMN-PT ribbon-based piezoelectric-ferroelectric hybrid energy harvester to scavenge the mechanical and thermal energy from human body [36]. Such flexible hybrid harvester can be used as wearable sensor to measure human body motions, body temperature, and acoustic sounds.

100

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 3.14 A ferroelectric generator based on the PVDF film. Figure taken from Q. Leng, L. Chen, H. Guo, J. Liu, G. Liu, C. Hu, et al., Harvesting heat energy from hot/cold water with a pyroelectric generator, J. Mater. Chem. 2 (2014) 11940–7.

3.6

Other applications

3.6.1 Electrocaloric cooling Electrocaloric cooling is based on electrocaloric effect, which is the reverse of pyroelectric effect. As the name suggests, electrocaloric effect is used to produce temperature change in a ferroelectric material by passing an external electric current into the material. Fig. 3.17 shows the difference between a pyroelectric and an electrocaloric

Harvesting thermal energy with ferroelectric materials

101

Fig. 3.15 A flexible hybrid ferroelectric-piezoelectric energy harvester. Figure taken from J.H. Lee, K.Y. Lee, M.K. Gupta, T.Y. Kim, D.Y. Lee, J. Oh, et al., Highly stretchable piezoelectric-pyroelectric hybrid nanogenerator, Adv. Mater. 26 (2014) 765–9.

cycle. While the four states in the cycle remain the same, the direction of the cycle determines the direction of the heat flow. During pyroelectric cycle, heat flows from a high temperature reservoir to a low temperature reservoir, generating some electrical work. During electrocaloric cycle, however, an electrical work is performed on the system to allow heat flow from a low temperature reservoir to a high temperature reservoir, causing electrocaloric cooling. Electrocaloric cooling, in general, is not a widely explored subject in the literature. Despite the principle of electrocaloric cooling being known for the last several decades, electrocaloric cooler never became a commercial reality. This is because of the poor cooling capacity and coefficient of performance, making the electrocaloric refrigerator economically unviable. In addition, there are several technical challenges that need to be resolved. Fundamentally, electrocaloric cooling requires cyclic variation in temperature. This infers that the working ferroelectric material must make and break thermal contact between the heat source and heat sink. Due to the large thermal contact resistances, the operation of such system is restricted by the operating frequency, leading to poor cooling capacity. For instance, the cooling capacity of a heat-switch-based electrocaloric cooler reported by Wang et al. was noted to be around 0.018 W/g, which is orders of magnitude smaller than the theoretically calculated values based on intrinsic material properties [37, 38]. Alternatively, cyclic cooling and heating in the ferroelectric material can be achieved by circulating a thermal fluid. Such designs, however, require fluid pumps to circulate fluid, adding additional complexity, mass, and cost to the system [39]. A recent study by Ma et al.,

102

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 3.16 PMN-PT ribbon-based piezoelectric-ferroelectric hybrid attached on the wrist skin. Figure taken from Y. Chen, Y. Zhang, F. Yuan, F. Ding, O.G. Schmidt, A flexible PMN-PT ribbon-based piezoelectric-pyroelectric hybrid generator for human-activity energy harvesting and monitoring. Adv. Electron. Mater. 3 (2017) 1600540; A. Thakre, A. Kumar, H-C. Song, DY. Jeong, J. Ryu, Pyroelectric energy conversion and its applications—flexible energy harvesters and sensors, Sensors 19 (2019) 2170.

however, has reported an electrocaloric cooler having the specific cooling capacity of 2.8 watts per gram and a COP of 13, which is higher than the most solid-state cooling technologies [38].

3.6.2 Pyroelectric detectors In addition to the thermal energy harvesting and electrocaloric cooling, ferroelectric materials are also being used to make sensors/detectors. It is based on the same principle as pyroelectric effect, where temperature fluctuations in a ferroelectric material generate variations in surface change density, resulting in electrical current or signal that can be detected or sensed. The temperature gradient can be caused by either a direct heat source or an indirect heat source, resulting in contact-type pyroelectric

Harvesting thermal energy with ferroelectric materials

103

Fig. 3.17 Pyroelectric energy harvesting cycle vs electrocaloric cooling cycle.

Fig. 3.18 Working principle of pyroelectric thermal detector/sensors.

thermal detector or noncontact-type pyroelectric infrared detector, respectively. Fig. 3.18 shows the working principle of the pyroelectric detectors/sensors. Thermal radiation from a heat source is absorbed using an absorber plate/coating. Heat is then transferred to a thin ferroelectric piece that generates electric current in response to temperature change. The sensitivity of a pyroelectric sensor depends on various factors. Some of them are outlined below [40]: l

l

Higher pyroelectric coefficient results in higher pyroelectric current, and thus, better sensing. Higher dielectric constant determines higher capacitance, thereby affecting the noise. In voltage operation, a larger capacitance is desired.

104 l

l

Ferroelectric Materials for Energy Harvesting and Storage

Higher specific heat of the working materials determines its temperature change in response to the absorbed radiation. A low specific heat infers a large temperature increase, and thus, a better signal. Low AC resistance and dielectric losses are desired to generate better signal.

3.7

Summary/future perspective

Research in the area of pyroelectric thermal energy harvesting is certainly limited. There are several technical challenges that still need to be resolved before this technology becomes commercially viable. One of the key challenges is the poor thermal to electrical energy conversion efficiency. Few recent studies, however, have shown encouraging results that possibly will set a new pathway for solid-state thermal energy harvesting using pyroelectric effect. Pyroelectric thermal energy harvesting is particularly advantageous in situations where heat source is freely available. Some of such examples are waste heat generated from industrial or domestic discharge and natural heat from solar or geothermal energy.

References [1] H.R. Jo, Ferroelectric Phase Transformations for Energy Conversion and Storage Applications, (2016). [2] S. Pandya, G. Velarde, L. Zhang, J.D. Wilbur, A. Smith, B. Hanrahan, et al., New approach to waste-heat energy harvesting: Pyroelectric energy conversion, NPG Asia Mater. 11 (2019) 1–5. [3] R. Kishore, S. Priya, A review on low-grade thermal energy harvesting: materials, methods and devices, Materials 11 (2018) 1433. [4] A. Cuadras, M. Gasulla, V. Ferrari, Thermal energy harvesting through pyroelectricity, Sensors Actuators A Phys. 158 (2010) 132–139. [5] P. Mane, J. Xie, K.K. Leang, K. Mossi, Cyclic energy harvesting from pyroelectric materials, IEEE Trans. Ultrason. Ferroelectr. Freq. Control 58 (2011) 10–17. [6] C.R. Bowen, J. Taylor, E. LeBoulbar, D. Zabek, A. Chauhan, R. Vaish, Pyroelectric materials and devices for energy harvesting applications, Energy Environ. Sci. 7 (2014) 3836–3856. [7] S.P. Alpay, J. Mantese, S. Trolier-McKinstry, Q. Zhang, R.W. Whatmore, Nextgeneration electrocaloric and pyroelectric materials for solid-state electrothermal energy interconversion, MRS Bull. 39 (2014) 1099–1111. [8] A. Van Der Ziel, Solar power generation with the pyroelectric effect, J. Appl. Phys. 45 (1974) 4128. [9] J.A. Gonzalo, Ferroelectric materials as energy converters, Ferroelectrics 11 (1976) 423–429. [10] W.H. Clingman, R.G. Moore Jr., Application of ferroelectricity to energy conversion processes, J. Appl. Phys. 32 (1961) 675–681. [11] J.D. Childress, Application of a ferroelectric material in an energy conversion device, J. Appl. Phys. 33 (1962) 1793–1798. [12] R.B. Olsen, D. Evans, Pyroelectric energy conversion: hysteresis loss and temperature sensitivity of a ferroelectric material, J. Appl. Phys. 54 (1983) 5941–5944.

Harvesting thermal energy with ferroelectric materials

105

[13] R.B. Olsen, D.A. Bruno, J.M. Briscoe, Pyroelectric conversion cycles, J. Appl. Phys. 58 (1985) 4709–4716. [14] G. Sebald, S. Pruvost, D. Guyomar, Energy harvesting based on Ericsson pyroelectric cycles in a relaxor ferroelectric ceramic, Smart Mater. Struct. 17 (2007) 015012. [15] F.Y. Lee, S. Goljahi, I.M. McKinley, C.S. Lynch, L. Pilon, Pyroelectric waste heat energy harvesting using relaxor ferroelectric 8/65/35 PLZT and the Olsen cycle, Smart Mater. Struct. 21 (2012) 025021. [16] G. Vats, A. Chauhan, R. Vaish, Thermal energy harvesting using bulk Lead-free ferroelectric ceramics, Int. J. Appl. Ceram. Technol. 12 (2015). [17] A. Chauhan, S. Patel, G. Vats, R. Vaish, Enhanced thermal energy harvesting using li, Kdoped Bi0.5 Na0.5TiO3 lead-free ferroelectric ceramics, Energy Technol. 2 (2014) 205–209. [18] G. Vats, R. Vaish, C.R. Bowen, An analysis of lead-free (Bi0.5Na0.5) 0.915-(Bi0.5K0.5) 0.05Ba0.02Sr0.015TiO3 ceramic for efficient refrigeration and thermal energy harvesting, J. Appl. Phys. 115 (2014) 013505. [19] G. Vats, H.S. Kushwaha, R. Vaish, Enormous energy harvesting and storage potential in multiferroic epitaxial thin film hetrostructures: an unforeseen era, Mater. Res. Express 1 (2014) 015503. [20] R. Olsen, D. Bruno, J. Briscoe, J. Dullea, Cascaded pyroelectric energy converter, Ferroelectrics 59 (1984) 205–219. [21] R.B. Olsen, J.M. Briscoe, D.A. Bruno, W.F. Butler, A pyroelectric energy converter which employs regeneration, Ferroelectrics 38 (1981) 975–978. [22] A. Khodayari, S. Pruvost, G. Sebald, D. Guyomar, S. Mohammadi, Nonlinear pyroelectric energy harvesting from relaxor single crystals, IEEE Trans. Ultrason. Ferroelectr. Freq. Control 56 (2009). [23] R. Kandilian, A. Navid, L. Pilon, The pyroelectric energy harvesting capabilities of PMN– PT near the morphotropic phase boundary, Smart Mater. Struct. 20 (2011) 055020. [24] R.B. Olsen, D.A. Bruno, J.M. Briscoe, E.W. Jacobs, Pyroelectric conversion cycle of vinylidene fluoride-trifluoroethylene copolymer, J. Appl. Phys. 57 (1985) 5036–5042. [25] M. Ikura, Conversion of low-grade heat to electricity using pyroelectric copolymer, Ferroelectrics 267 (2002) 403–408. [26] H. Nguyen, A. Navid, L. Pilon, Pyroelectric energy converter using co-polymer P (VDFTrFE) and Olsen cycle for waste heat energy harvesting, Appl. Therm. Eng. 30 (2010) 2127–2137. [27] A. Navid, L. Pilon, Pyroelectric energy harvesting using Olsen cycles in purified and porous poly (vinylidene fluoride-trifluoroethylene)[P (VDF-TrFE)] thin films, Smart Mater. Struct. 20 (2011) 025012. [28] R. Olsen, D. Bruno, Pyroelectric conversion materials, in: IECEC’86; Proceedings of the Twenty-first Intersociety Energy Conversion Engineering Conference, 1986, pp. 89–93. [29] H. Zhu, S. Pruvost, P. Cottinet, D. Guyomar, Energy harvesting by nonlinear capacitance variation for a relaxor ferroelectric poly (vinylidene fluoride-trifluoroethylenechlorofluoroethylene) terpolymer, Appl. Phys. Lett. 98 (2011) 222901. [30] R.B. Olsen, Ferroelectric conversion of heat to electrical energy—a demonstration, J. Energy 6 (1982) 91–95. [31] R. Olsen, D. Brown, High efficiency direct conversion of heat to electrical energy-related pyroelectric measurements, Ferroelectrics 40 (1982) 17–27. [32] S. Hunter, N. Lavrik, S. Rajic, P. Datskos, Pyroelectric thermal energy harvesting using MEMS based resonant structures, in: Energy Harvesting and Storage: Materials, Devices, and Applications II, Proc SPIE 8035, 2012.

106

Ferroelectric Materials for Energy Harvesting and Storage

[33] A. Sultana, M.M. Alam, T.R. Middya, D. Mandal, A pyroelectric generator as a selfpowered temperature sensor for sustainable thermal energy harvesting from waste heat and human body heat, Appl. Energy 221 (2018) 299–307. [34] Q. Leng, L. Chen, H. Guo, J. Liu, G. Liu, C. Hu, et al., Harvesting heat energy from hot/ cold water with a pyroelectric generator, J. Mater. Chem. A 2 (2014) 11940–11947. [35] J.H. Lee, K.Y. Lee, M.K. Gupta, T.Y. Kim, D.Y. Lee, J. Oh, et al., Highly stretchable piezoelectric-pyroelectric hybrid nanogenerator, Adv. Mater. 26 (2014) 765–769. [36] Y. Chen, Y. Zhang, F. Yuan, F. Ding, O.G. Schmidt, A flexible PMN-PT ribbon-based piezoelectric-pyroelectric hybrid generator for human-activity energy harvesting and monitoring, Adv. Electron. Mater. 3 (2017) 1600540. [37] Y.D. Wang, S.J. Smullin, M.J. Sheridan, Q. Wang, C. Eldershaw, D.E. Schwartz, A heatswitch-based electrocaloric cooler, Appl. Phys. Lett. 107 (2015) 134103. [38] R. Ma, Z. Zhang, K. Tong, D. Huber, R. Kornbluh, Y.S. Ju, et al., Highly efficient electrocaloric cooling with electrostatic actuation, Science 357 (2017) 1130–1134. [39] M. Ozˇbolt, A. Kitanovski, J. Tusˇek, A. Poredosˇ, Electrocaloric refrigeration: Thermodynamics, state of the art and future perspectives, Int. J. Refrig. 40 (2014) 174–188. [40] Lasercomponents, n.d. https://www.lasercomponents.com/. Pyroelectric Detectors: Materials, Applications, and Working Principle. https://www.lasercomponents.com/us/news/ pyroelectric-detectors-materials-applications-and-working-principle/. Accessed on Nov 26, 2019.

Leveraging size effects in flexoelectric-piezoelectric vibration energy harvesting

4

Adriane G. Moura and Alper Erturk G.W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA, United States

4.1

Introduction

Microelectromechanical systems (MEMS) and nanoelectromechanical systems (NEMS) have received growing attention in the last decade for various applications including mechanical energy harvesting at very small scales [1–4]. The mechanical energy in this context spans from structure-borne vibrations [5] and waves [6, 7] to rigid-body motions [8, 9], acoustic energy [10–12], as well as aeroelastic [13, 14] and hydroelastic [15, 16] vibrations. In harvesting various forms of mechanical energy, piezoelectricity remains arguably the most widely studied transduction method with examples ranging from PZT-based (lead zirconate titanate) ferroelectric thin films [3] to piezoelectric nanowires [4] employing nonferroelectrics, such as ZnO (zinc oxide). It is well-known that the electromechanical coupling [17] of piezoelectric materials diminishes dramatically in thin films [18] and polymers [19]. Piezoelectric polymers, such as PVDF (polyvinylidene fluoride), are environmentally benign as compared to ceramics, but they are poor power generators due to low electromechanical coupling. Bulk piezoelectric ceramics (such as PZT-5A and PZT-5H) are relatively brittle and less reliable for powering sensor systems in harsh environments. Moreover, lead content in most piezoelectric ceramic compositions is a major environmental issue [20]. Furthermore, several of the high electromechanical coupling materials lose their piezoelectricity at moderate to high temperatures, where self-powered sensors are most needed. Recent efforts at small scales [21] suggest that the effective electromechanical properties of elastic dielectrics can be enhanced dramatically under nonuniform strain fields due to an entirely different phenomenon called flexoelectricity [22–27]. Flexoelectricity describes the generation of electric polarization in elastic dielectrics by the application of a mechanical strain gradient [23, 25, 28]. The phenomenon of flexoelectricity in solids is a higher-order effect and is expected to be rather weak except for very small (submicron) dimensions, making the concept of interest mainly for potential MEMS and especially NEMS applications. Following the early efforts by Mashkevich and Tolpygo [29], Kogan [30], and Indenbom et al. [31], the first comprehensive theoretical discussion of the Ferroelectric Materials for Energy Harvesting and Storage. https://doi.org/10.1016/B978-0-08-102802-5.00004-2 © 2021 Elsevier Ltd. All rights reserved.

108

Ferroelectric Materials for Energy Harvesting and Storage

flexoelectric effect was presented by Tagantsev [22]. The research field of flexoelectricity has been active for liquid crystals [32] and biological matter [33] for decades. However, it was only in early 2000s that the flexoelectric effect in solids has received suddenly growing attention, especially after the experiments by Ma and Cross [34–39] on elastic dielectrics, specifically high-K materials such as ferroelectric perovskites (see the review article by Cross [40]). In addition to experimental efforts by Ma and Cross [34–39] and others [41–43] for samples with high dielectric constants, atomistic simulations [28] were presented to extract flexoelectric coefficients. Importantly, substantial difference (several orders of magnitude) was reported between the simulated and identified flexoelectric coefficients [39]. The experimental samples [34–39] (of
mm thickness) used in the identification efforts were typically far from the thickness levels of interest (
nm) in flexoelectricity. A comprehensive article on the flexoelectric effect in solids by Yudin and Tagantsev [25] presents a detailed discussion on the subject matter along with a historical account. It is of no surprise that, with its promise of increased electromechanical coupling at small scales, flexoelectricity is of great interest for submicron level energy harvesting [44, 45] to power next-generation nanoscale sensors and other extremely low-power small electronic components. Other than the mismatch in the order of magnitude of flexoelectric coupling between atomistic simulations [28] and experimental measurements [34–39], one of the issues in flexoelectric transduction and energy conversion has been the lack of a clear understanding and modeling of the converse effect, as the subject has created confusion since the converse effect is represented by a polarization gradient [46–48]. In a recent work focusing on finite samples, Tagantsev and Yurkov [49] presented a consistent and symmetric converse effect representation and its justification. In the present work, we combine the direct effect of flexoelectricity and this symmetric converse effect within a distributed-parameter electroelastodynamic framework and provide a modal analysis solution for vibration energy harvesting from base excitation of dielectric cantilevers. In addition to closed-form expressions for the electromechanically coupled voltage across the electrical load and the shunted vibration response (that accounts for the effect of the electrical load), the size-dependent flexoelectric coupling coefficient is extracted analytically, and a figure of merit is identified. The modeling framework is then extended to noncentrosymmetric configurations to understand the interaction between flexoelectricity and piezoelectricity for different thickness levels from mm-scale to nm-scale. Both purely flexoelectric energy harvesting [50] and combined flexoelectric-piezoelectric energy harvesting [51] scenarios are discussed based on the authors’ recent efforts.

4.2

Direct and converse flexoelectric and piezoelectric effects

In elastic dielectrics, piezoelectricity is the response of polarization to applied mechanical strain and vice versa. Piezoelectric coupling is controlled by a third-rank tensor and is allowed only in materials that are noncentrosymmetric. Flexoelectricity,

Flexoelectric-piezoelectric vibration energy harvesting

109

on the other hand, is the generation of electric polarization by the application of a nonuniform mechanical strain field, i.e., a strain gradient, and is expected to be pronounced at submicron thickness levels, especially at the nanoscale. Flexoelectricity is controlled by a fourth-rank tensor and is, therefore, allowed in materials of any symmetry, i.e., a piezoelectric material also exhibits the flexoelectric effect at very low thickness levels. As a gradient effect, flexoelectricity is size-dependent, while piezoelectric coupling has no size dependence. The polarization including the direct flexoelectric and piezoelectric effects for the transverse mode can be written as P3 ¼ χ 33 E3 + e311 S11 + μ1133

∂S11 ∂x3

(4.1)

where P3 is the polarization in thickness direction (3-direction is the thickness direction and 1-direction is the axial direction), E3 is the electric field, e311 is the piezoelectric constant, S11 is the axial strain, χ 33 is the dielectric susceptibility (which has the units of F/m, and should not be confused with the dimensionless electric susceptibility form χ 33 , χ 33 ¼ χ 33 ε0 where ε0 is the vacuum permittivity), and μ1133 is the transverse flexoelectric coefficient. The mechanical stress accounting for the converse piezoelectric and flexoelectric effects, for the transverse mode, can be expressed as T11 ¼ c1111 S11 + e311 E3 + f1133

∂P3 ∂x3

(4.2)

∂E3 ∂x3

(4.3)

or alternatively T11 ¼ c1111 S11 + e311 E3 + μ1133

where T11 is the axial stress, c1111 is the elastic modulus of the piezoelectric material (under short-circuit condition of the electrodes), and f1133 is the “flexocoupling coefficient” (f1133 ¼ χ 1 33 μ1133). Note that the above form of flexoelectric coupling is suitable for basic “exogenous” strain gradients, such as those due to mechanical bending; but would be limited for “endogenous” ones, such as those due to domain boundaries and interfaces, which are beyond the scope of this work [25]. The approximate flexocoupling coefficient order of magnitude [24] is 1–10 V based on Kogan’s estimate [30] f  q/(4πε0a), where q and a are the lattice charge and spacing, respectively (this order of magnitude is in agreement with recently presented upper bounds by Yudin et al. [52]). For a centrosymmetric (nonpiezoelectric) beam, Eqs. (4.1) and (4.2) simplify to include only the transverse mode flexoelectric effect, yielding the following constitutive equations: P3 ¼ χ 33 E3 + μ1133

∂S11 ∂x3

(4.4)

110

Ferroelectric Materials for Energy Harvesting and Storage

T11 ¼ c1111 S11 + f1133

∂P3 ∂x3

(4.5)

The form of Eqs. (4.1) and (4.2) contains both the flexoelectric and piezoelectric effects, and therefore, it is applicable to a wide range of geometric scales.

4.3

Flexoelectric energy harvesting using a centrosymmetric cantilever

First, we consider the problem of a centrosymmetric thin cantilever under mechanical base excitation (Fig. 4.1) for linear transverse (bending) vibrations, i.e., linear electroelastic material behavior and geometrically small oscillations are assumed in this continuum framework. The surface electrodes of the cantilever are shunted to a resistive electrical load to quantify the electrical power output in the harvester model. The sample geometry justifies thin beam assumptions, such that the width (b) and the thickness (h) of the rectangular cross section are much smaller than the overhang length (L). “Static” flexoelectricity [25] is applicable since the thickness (smallest dimension) of the beam is much smaller than the wavelength at vibration frequencies of interest in this work for the first few bending modes. In the following, the focus is placed on static bulk flexoelectricity, and therefore, the surface effects [24, 25] are excluded.

4.3.1 Flexoelectrically coupled mechanical equation and modal analysis The partial differential equation governing the forced vibration of a uniform cantilevered centrosymmetric thin dielectric beam under transverse base excitation (Fig. 4.1) is

Fig. 4.1 Base-excited centrosymmetric dielectric cantilever with surface electrodes (that are perpendicular to the thickness direction) connected to a resistive electrical load for energy harvesting, and a cross sectional view. The transverse displacement of the beam relative to the moving base is wrel and the voltage output across the resistive load is v.

Flexoelectric-piezoelectric vibration energy harvesting

∂2 Mðx1 , tÞ ∂5 wrel ðx1 , tÞ ∂wrel ðx1 , tÞ ∂2 wrel ðx1 , tÞ + m + c I + c s a ∂x1 2 ∂t ∂t2 ∂x41 ∂t 2 d wb ðtÞ ¼ m dt2

111



(4.6)

where wb(t) is the base excitation (in the form of displacement), wrel(x1, t) is the transverse displacement of the beam (reference surface, or neutral axial level) relative to its base and M(x1, t) is the internal bending moment at position x1 and time t, ca is the viscous air damping coefficient (as a mass-proportional dissipative term), cs is the strain-rate damping coefficient (as a stiffness-proportional dissipative term), and m is the mass per unit length of the beam (m ¼ ρbh where ρ is the mass density of the beam material). In the same vein as cantilevered piezoelectric energy harvester counterparts [53, 54], the linear damping operators satisfy the proportional damping condition so that the mode shapes of the corresponding undamped system can be used in modal analysis (implementation of nonlinear intrinsic and extrinsic damping mechanisms is beyond the scope of this work—see Leadenham and Erturk [55, 56], among others, for the resonant modeling of quadratic solid [55] and fluid [56] damping). The internal bending moment in Eq. (4.6) is the first moment of the axial stress field over the cross section: Z Mðx1 , tÞ ¼ b

h=2

h=2

T11 x3 dx3

(4.7)

The axial strain component is due to bending only and it can be expressed as S11 ðx1 , x3 , tÞ ¼ x3

∂2 wrel ðx1 , tÞ ∂x21

(4.8)

It is clear from Eqs. (4.4) and (4.8) that the strain gradient ∂ S11/∂ x3 in this model is nothing but the negative curvature of the uniform Euler-Bernoulli beam (assuming the effect of the gradient ∂ S11/∂ x1 to be negligible). Following Tagantsev and Yurkov [49], for a finite sample in which the polarization in the thickness direction varies continuously from its bulk value to zero at the top and bottom surfaces of the cantilever at x3 ¼ h/2 and x3 ¼  h/2 (based on the blocking boundary condition assumption—see fig. 12A in Yudin and Tagantsev [25]), the second right-hand-side term can be evaluated using integration by parts to identify the role of this term in the moment equation: Z bf 1133

∂P3 x3 dx3 ¼ bf 1133 h=2 ∂x3 h=2

Z

h=2

h=2

P3 dx3 ¼ bf 1133 hhP3 i

(4.9)

112

Ferroelectric Materials for Energy Harvesting and Storage

where hP3i is the average polarization induced by the electric field in the beam. The spatial scale of the polarization variation at the dielectric–electrode interface is much smaller than that in the overall thickness h; therefore this average polarization is approximately the polarization in the bulk [25, 49]: hP3 i  χ 33 E3

(4.10)

and it is useful to recall from Eqs. (4.2) and (4.3) that the dielectric susceptibility is 1133 . χ 33 ¼ μf1133 The electric field E3 can be given in terms of the voltage (v) across the surface electrodes and the electrode spacing as E3 ¼  v/h (where it is assumed that the electrode thickness is negligible). The flexoelectric term in Eq. (4.9) is only a function of time, and therefore, we multiply it by [H(x1)  H(x1  L)] (where H(x1) is the Heaviside function), to ensure the survival of this term when the bending moment is substituted into Eq. (4.6) for full electrode coverage (Fig. 4.1) from the clamped end (x1 ¼ 0) to the free end (x1 ¼ L). The internal bending moment is then Mðx1 , tÞ ¼ YI

∂2 wrel ðx1 , tÞ + μ1133 bvðtÞ½H ðx1 Þ  H ðx1  LÞ ∂x1 2

(4.11)

where the flexural rigidity YI for the rectangular cross section (under short-circuit condition) is YI ¼

c1111 bh3 12

(4.12)

The flexoelectrically coupled centrosymmetric Euler-Bernoulli beam equation for transverse vibrations can then be obtained from Eq. (4.6) as ∂4 wrel ðx1 , tÞ ∂5 wrel ðx1 , tÞ ∂wrel ðx1 , tÞ ∂2 wrel ðx1 , tÞ + m + c + c I s a ∂x1 4 ∂x1 4 ∂t ∂t ∂t2   dδðx1 Þ dδðx1  LÞ d2 wb ðtÞ μ1133 bvðtÞ  ¼ m dx1 dx1 dt2 YI

(4.13)

where δ(x1) is the Dirac delta function that satisfies the following equation for a smooth test function γ(x1): Z

∞ ∞

dðnÞ δðx1  pÞ ðnÞ

dx1

γ ðx1 Þdx1 ¼ ð1Þn

dγ ðnÞ ðpÞ ðnÞ

dx1

(4.14)

Flexoelectric-piezoelectric vibration energy harvesting

113

The vibration response (transverse displacement of the reference surface) relative to the moving base in Fig. 4.1 can be expressed as wrel ðx1 , tÞ ¼

∞ X

ϕr ðx1 Þηr ðtÞ

(4.15)

r¼1

Here, ηr(t) is the modal mechanical coordinate and ϕr(x1) is the mass-normalized eigenfunction (obtained from the short-circuit problem) for the rth vibration mode: ϕr ð x 1 Þ ¼

rffiffiffiffiffiffiffi   1 λr x1 λr x1 sin λr  sinh λr λr x1 λr x1 cos  cosh +  sinh sin mL L L cos λr + cosh λr L L (4.16)

where the eigenvalues (λr > 0, r ¼ 1, 2, …) are the roots of the characteristic equation (for the short-circuit and clamped-free boundary conditions): 1 + cos λcosh λ ¼ 0

(4.17)

The mass-normalized eigenfunctions in Eq. (4.15) satisfy the following orthogonality conditions: Z

L

Z mϕr ðx1 Þϕs ðx1 Þdx1 ¼ δrs ,

0

0

L

YIϕr ðx1 Þ

d 4 ϕs ðx1 Þ dx1 ¼ δrs ω2r dx41

(4.18)

where δrs is the Kronecker delta and ωr is the undamped natural frequency of the rth vibration mode under short-circuit conditions (Rl ! 0): ωr ¼ λ2r

rffiffiffiffiffiffiffiffiffi YI mL4

(4.19)

which can also be denoted by ωsc r , the short-circuit natural frequency of the rth mode. The mechanical equation in modal coordinates can be obtained after substituting Eq. (4.15) into Eq. (4.13) (then multiplying the latter by the mode shape, integrating over the beam length, and applying the orthogonality conditions) as d2 ηr ðtÞ dη ðtÞ + 2ζ r ωr r + ω2r ηr ðtÞ  θr vðtÞ ¼ fr ðtÞ 2 dt dt

(4.20)

where the modal electromechanical coupling term due to flexoelectricity is θr ¼ μ1133 b

 dϕr ðx1 Þ dx1 x1 ¼L

(4.21)

114

Ferroelectric Materials for Energy Harvesting and Storage

and mechanical forcing function can be expressed as Z d2 wb ðtÞ L fr ðtÞ ¼ m ϕr ðx1 Þdx1 dt2 0

(4.22)

4.3.2 Flexoelectrically coupled electrical circuit equation and modal analysis In order to obtain an electrical circuit equation in the presence of a finite electrical load impedance, it is useful to obtain the electric displacement that is compatible with the polarization form of Eq. (4.4) through the well-known dielectric relationship D 3 ¼ P3 + ε 0 E 3

(4.23)

Substituting Eq. (4.4) into Eq. (4.23), the nonzero electric displacement component for transverse vibrations of the thin beam configuration with surface electrodes shown in Fig. 4.1 becomes D3 ¼ ε33 E3 + μ1133

∂S1 ∂x3

(4.24)

where ε33 is the dielectric permittivity ε33 ¼ ε0 + χ 33 ¼ ð1 + χ 33 Þε0 (note that for high-K materials, which are of interest in flexoelectricity, χ 33 ≫ 1, and ε33  χ 33). In the presence of a finite resistive load connected across the electrodes of the beam, the flexoelectrically coupled circuit equation can be obtained from the integral Z  d vðtÞ D  ndA ¼ (4.25) dt A Rl where D is the vector of electric displacement components, n is the unit outward normal of the electrodes, and the integration is performed over the electrode area A. The only contribution to the inner product of the integrand is from D3 given by Eq. (4.24). Using Eq. (4.24) in Eq. (4.25), the following circuit equation (current balance) is obtained: dvðtÞ vðtÞ + C + μ1133 b dt Rl

Z 0

∂ wrel ðx1 , tÞ dx1 ¼ 0 ∂x1 2 ∂t

L 3

(4.26)

where the capacitance (C) is C¼

ε33 bL h

(4.27)

Note that, it is possible to introduce dielectric losses by changing the real-valued capacitance to C(1  j tan δ) where tanδ is the loss tangent.

Flexoelectric-piezoelectric vibration energy harvesting

115

Eq. (4.15) can be substituted into Eq. (4.26) to obtain C

∞ dvðtÞ vðtÞ X dη ðtÞ + + θr r ¼ 0 dt Rl dt r¼1

(4.28)

Here, the modal electromechanical coupling (θr) due to the direct flexoelectric effect is the same as Eq. (4.21) that was obtained for the converse effect, which further confirms the symmetry in the fully coupled governing electroelastodynamic equations, which are Eqs. (4.13) and (4.26) in physical coordinates, or Eqs. (4.20) and (4.28) in modal coordinates. Flexoelectric power generation as a result of voltage output across the resistive load is due to Eq. (4.26), and simultaneously the voltage output sends a feedback to the mechanical domain due to the voltage term in Eq. (4.13), as a manifestation of the thermodynamic consistency resulting from the two-way coupling.

4.3.3 Closed-form voltage response and vibration response at steady state For harmonic base excitation with wb(t) ¼ W0ejωt, the modal forcing function given by Eq. (4.22) can be expressed as fr(t) ¼ Frejωt, where the amplitude Fr is Z

L

Fr ¼ ω mW 0 2

ϕr ðx1 Þdx1

(4.29)

0

Then, the steady state modal mechanical coordinate of the beam and the steady state voltage response across the resistive load are also harmonic at the same frequency as ηr(t) ¼ Hrejωt and v(t) ¼ Vejωt, respectively, where the amplitudes Hr and V are complexed valued. Therefore, Eqs. (4.20) and (4.28) yield 

ω2r  ω2 + j2ζ r ωr ω Hr  θr V ¼ Fr



(4.30)

 ∞ X 1 + jωC V + jω θr Hr ¼ 0 Rl r¼1

(4.31)

where ζ r is the modal mechanical damping ratio (due to purely mechanical dissipation) that can easily be related to cs and ca as 2ζ rωr ¼ csIω2r /YI + ca/m. The steady state voltage response is obtained from Eqs. (4.30) and (4.31) as ∞ X

vðtÞ ¼

jωθr Fr + j2ζ r ωr ω

ω 2  ω2 r¼1 r ∞ X

1 + jωC + Rl

jωθ2r 2 ω  ω2 + j2ζ r ωr ω r¼1 r

ejωt

(4.32)

116

Ferroelectric Materials for Energy Harvesting and Storage

Once the voltage across the electrical load is obtained, the current and power output can be calculated easily. For the case of a real-valued electrical load (i.e., resistive load), the current delivered to the load is i(t) ¼ v(t)/Rl and the instantaneous power output is P(t) ¼ v2(t)/Rl. The steady state modal mechanical response of the beam (that accounts for the converse effect) can be obtained as 3 1 jωθr Fr 7 C B ∞ 6 X ω2  ω2 + j2ζ r ωr ω 7 C 6B ϕr ðx1 Þejωt r¼1 r 7 C 6 B wrel ðx1 , tÞ ¼ ∞ C ω2  ω2 + j2ζ ωr ω7 6BFr  θr 1 2 X jωθ r 5 A r r r¼1 4@ + jωC + 2  ω2 + j2ζ ω ω ω Rl r r r r¼1 20

∞ X

(4.33)

4.3.4 Size effects on modal electromechanical coupling coefficient One direct measure of energy conversion is the electromechanical coupling coefficient “k” as commonly used in piezoelectricity [17, 57]. It is possible to analytically extract the transverse mode electromechanical coupling coefficient due to bulk flexoelectricity for the centrosymmetric cantilever of Fig. 4.1. A dynamic definition of the modal electromechanical coupling coefficient can be obtained based on the difference between the open- and short-circuit natural frequencies [17, 57]:  k2 ¼

ωoc r

2

 2  ωsc r  2 oc ωr

(4.34)

where k is the flexoelectric coupling coefficient for the rth vibration mode (the focus in the simulations of this work will be placed on the fundamental mode, r ¼ 1). Square of the coupling coefficient, as well-known from piezoelectric energy conversion problems, is a measure of how much of the mechanical work is converted to electrical energy, or vice versa in electrical actuation problems. A similar argument and an analogous expression can be given in terms of the open- and short-circuit stiffness terms in quasistatic conditions [17, 57]. In order to express the coupling coefficient using Eq. (4.34), recall that the undamped short-circuit natural frequency of the rth vibration mode is ωsc r

¼ ωr ¼ λ2r

rffiffiffiffiffiffiffiffiffi YI mL4

(4.35)

Then, for modal vibrations (dominated by the rth mode) and under open-circuit conditions, Eq. (4.28) can be reduced to

Flexoelectric-piezoelectric vibration energy harvesting

vðtÞ ¼

θr ηr ðtÞ , Rl ! ∞ C

117

(4.36)

Substituting Eq. (4.36) into the modal mechanical equation, Eq. (4.20), the undamped open-circuit natural frequency of the rth vibration mode becomes  θ2r 1+ 2 ωr C

(4.37)

 θ2r θ2 ω2 C k2 ¼  r 2  ¼ 2 r 2 θ ωr C + θ r 1 + 2r ωr C

(4.38)



2 ωoc r

¼



2 ωsc r



yielding 

Here, the capacitance is given by Eq. (4.27) and the modal coupling term can be expressed as  dϕr ðx1 Þ 1 λr θr ¼ μ1133 b ¼ μ1133 b pffiffiffiffiffiffiffi αr  dx1 x1 ¼L mL L where αr ¼  sin λr  sinh λr +

(4.39)

sin λr  sinh λr ð cos λr  cosh λr Þ. cos λr + cosh λr

Eq. (4.38) then becomes k2 ¼

1 c1111 ε33 λ2r 2 h +1 μ21133 12α2r

(4.40)

which clearly captures the thickness dependence of the modal flexoelectric coupling coefficient. Note that the fundamental (first) bending vibration mode (r ¼ 1) is typically of interest for energy harvesting using a linear cantilever under base excitation [53, 54], yielding λ1 ¼ 1.87510407 and α1 ¼  1.46819102 for the simulations in this work (the first mode shape is shown in Fig. 4.2A). It is worth mentioning that energy harvesting at higher vibration modes requires using segmented electrodes to avoid charge cancellation [53, 54]. The fundamental bending mode results in no cancellation for continuous electrode coverage since the curvature is in phase throughout the length of the beam (Fig. 4.2C), i.e., there are no inflection points. Note that, 85% of the electric charge is produced by the first half of the cantilever, i.e., 0  x1  L/2, according to Fig. 4.2B (since the integral of curvature is related to the electric charge according to current balance Eq. (4.26) and the curvature is maximum near the clamped end in Fig. 4.2C). From the electromechanical coupling standpoint, 85% of the modal

1 0.9

f1 [Normalized]

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0

0.1

0.2

0.3

0.4

0.5 x1/L

0.6

0.7

0.8

0.9

1

0

0.1

0.2

0.3

0.4

0.5 x1/L

0.6

0.7

0.8

0.9

1

0

0.1

0.2

0.3

0.4

0.5 x1/L

0.6

0.7

0.8

0.9

1

(A) 1 0.9 df1/dx1 [Normalized]

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

(B) 1

d2 f1/dx12 [Normalized]

0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

(C)

Fig. 4.2 (A) Displacement, (B) slope, and (C) curvature distributions of a thin cantilever for the fundamental bending vibration mode (r ¼ 1). The maximum curvature is near the clamped end. The region 0  x1  L/2 produces 85% of the modal electromechanical coupling.

Flexoelectric-piezoelectric vibration energy harvesting

119

electromechanical coupling θ1 (which determines the coupling coefficient k) is due to the region 0  x1  L/2, in view of Eq. (4.21) and Fig. 4.2B. In terms of the size effect, Eq. (4.40) shows that, with decreased thickness (h), the coupling coefficient (k) increases. This equation also shows the effect of the material properties on the coupling coefficient. The Figure of Merit (FoM) in flexoelectric energy conversion is FoM ¼

μ21133 c1111 ε33

(4.41)

or μ21133/c1111χ 33, and as FoM ! ∞, k2 ! 1 which is the limit of 100% mechanical-toelectrical energy conversion within the structure (note that it is not the percentage energy delivered to the electrical load, hence not an overall efficiency). Since μ1133 ¼ χ 33 f1133 and FoM ∝ f21133 χ 33/c1111, the alternative FoM forms are f21133 χ 33/ c1111 or f21133 χ 33s1111 (where s1111 is the short-circuit elastic compliance). According to FoM ∝ f21133 χ 33s1111, for a fixed flexocoupling coefficient (e.g., f  q/(4πε0a) based on Kogan’s estimate [30]), softer materials with high dielectric permittivity should be preferred for increased FoM. For resonant energy harvesting purposes, the quality factor (Q) of the material is also important, and k2Q should be used for comparing flexoelectric energy harvesting performance of different materials (and soft materials tend to be more lossy, yielding low Q values as a trade-off). However, to explore size effects in the same single material, the coupling coefficient alone is sufficient assuming size-independent loss characteristics for simplicity (in fact, favorably, the intrinsic quality factor is expected to increase with reduced thickness according to recent molecular dynamics simulations [58] at the nanoscale).

4.3.5 Case studies and results In this section, simulations are performed to show the effect of thickness on the electromechanical coupling and the frequency response behavior of a flexoelectric uniform beam under bending vibrations for energy harvesting. The simulations are for Strontium Titanate (STO) using the atomistic value [28] of μ1133 ¼ 3.75  109 C/m along with the relevant material properties [59, 60]: c1111 ¼ 318 GPa, ε33 ¼ 2.66 nF/m, and ρ ¼ 5116 kg/m3.

4.3.5.1 Electromechanical coupling coefficient and size effects The electromechanical coupling coefficient due to flexoelectric energy conversion, or simply the transverse mode flexoelectric coupling coefficient, k, is plotted for a range of FoM values and cantilever thicknesses in Fig. 4.3. The focus is placed on the fundamental bending vibration mode (r ¼ 1), and the beam thicknesses in the simulations range from 1 mm to 1 nm. As stated previously based on Eq. (4.40), the coupling coefficient increases with decreased thickness. This is now illustrated graphically in Fig. 4.3. The coupling coefficient also increases with increased flexoelectric FoM defined by Eq. (4.41). As a specific instance, for typical Strontium

120

Ferroelectric Materials for Energy Harvesting and Storage

k

1 0.5

0 10–3 10–10 0

10

10–20

h [mm] 103

m21133/(c1111 å33) [m2]

Fig. 4.3 Transverse mode flexoelectric coupling coefficient vs cantilever thickness and figure of merit in flexoelectric energy conversion (for the fundamental bending vibration mode).

Titanate (STO) elastic [59, 60] and dielectric [24, 61] properties, the resulting FoM is around μ21133/c1111ε33 ¼ 1.66  1020 m2, yielding negligible coupling coefficient values except at the nanoscale according to Fig. 4.3. Typical atomistic simulations [28] result in flexoelectric coefficient values on the order of 109 C/m, while the experimentally identified values (by Cross et al. [36, 39, 62] for mm-scale samples) are as high as 104 C/m. Consider the experimental value of μ1133 ¼ 100  106 C/m (the authors of the original paper [36] reported a positive value) for Barium Strontium Titanate (BST) from the experiments by Ma and Cross [36] for mm-thick samples. The elastic modulus and permittivity values of BST were reported in another work by the same group [62] as c1111 ¼ 166 GPa and ε33 ¼ 0.1594 μF/m, respectively. The thickness dependence of the transverse mode coupling coefficient for these properties of BST is shown in Fig. 4.4 (solid line), along with that of STO (dashed line) based on the aforementioned atomistic values by Maranganti and Sharma [28] to demonstrate the order of magnitude difference between available experiments and atomistic simulations, although the materials are not identical. As expected, with decreased thickness, the coupling coefficient gradually approaches unity, indicating increased energy conversion with reduced thickness. Importantly, Fig. 4.4 reveals that, for μ1133 ¼ 100  106 C/m, the coupling coefficient is nearly unity for all submicron thickness levels, which makes the validity of this value (identified from mm-thick samples [36]) as a bulk flexoelectric coefficient questionable. Such an order of magnitude in bulk flexoelectric coefficient (
104 C/m) suggests very high conversion even for micron-thick nonpiezoelectric cantilevers, which, obviously, is not the case. This observation definitely encourages rigorous experiments at much smaller scales (ideally less than 10 nm thickness) and conditions under which the effects other than bulk flexoelectricity can be eliminated or controlled. The trend in the second curve (dashed line) based on atomistic simulations [28] of STO (with μ1133 ¼  3.75  109 C/m) is more reasonable, as it reveals that

Flexoelectric-piezoelectric vibration energy harvesting

121

100

k

10–2

10–4

10–6

m1133 = 10 mC/m, BST (Ma and Cross, 2002) m1133 = –3.75 nC/m, STO (Maranganti and Sharma, 2009)

10–9

10–6 h [m]

10–3

Fig. 4.4 Flexoelectric coupling coefficient (k) vs thickness (h) plots obtained using sample flexoelectric coefficient (μ1133) values identified by Ma and Cross [36] for BST (experimental) and calculated by Maranganti and Sharma [28] for STO (atomistic). The order of magnitude of the experimentally identified value for BST results in unrealistically high values of the coupling coefficient for all submicron thickness levels, suggesting that this coefficient identified in mm-thick samples is probably not static bulk flexoelectricity and is not valid for small scales.

the coupling coefficient exceeds 0.1 only when the cantilever thickness is a few nanometers. Overall, reducing the thickness from 1 mm to 1 nm increases the flexoelectric coupling coefficient by nearly 6 orders of magnitude in the STO cantilever.

4.3.5.2 Resonant energy harvesting: Electromechanical frequency response and size effects In this section, the electromechanical frequency response of a cantilevered flexoelectric energy harvester under base excitation is simulated with a focus on the first bending mode (r ¼ 1) for a broad range of electrical load resistance values. Three different geometric scales are explored, spanning from mm-scale to nm-scale thickness. For each case, the length/width/thickness aspect ratio is fixed at 100/5/1. The cantilever is made of STO and has perfectly conductive surface electrodes on the faces that are perpendicular to the direction of transverse base excitation (Fig. 4.1). A mechanical quality factor (Q) of
50 is assumed, yielding an approximate modal mechanical damping ratio of 1% of the critical damping (i.e., ζ 1 ffi 1/2Q ¼ 0.01 for the first bending mode). Three cases with thicknesses of 1 mm, 1 μm, and 1 nm are analyzed to explore the effect of thickness, while keeping the L/b/h aspect ratio fixed at 100/5/1. The mechanical excitation is harmonic base acceleration, d2wb(t)/dt2 ¼  ω2W0ejωt. Therefore, the results are given in the form

122

Ferroelectric Materials for Energy Harvesting and Storage

of frequency response magnitude maps normalized by the base acceleration quantified in terms of the gravitational acceleration (g ¼ 9.81 m/s2). A wide range of electrical load resistance values spanning from short- to open-circuit conditions (100 Ω to 1 GΩ) are simulated for each case study to capture the optimal load in power generation and the respective trends with changing load. The voltage output (per base acceleration) frequency response map (obtained from Eq. (4.32) via jv(t)/  ω2W0ejωt j as the magnitude form) for the 1 mm-thick STO beam (100 mm  5 mm  1 mm) is shown in Fig. 4.5A. The base excitation frequency is normalized with respect to the fundamental short-circuit natural frequency. With increased electrical load resistance, the voltage output increases monotonically at all frequencies, as a typical trend in energy harvesting [54]. It is shown that the resonance frequency for the 1 mm-thick STO cantilever is unaffected by the change in resistive load, i.e., the frequency of peak magnitude does not change as the electrical load resistance value is swept from short- to open-circuit conditions. This indicates very low electromechanical coupling such that the feedback in the mechanical domain due to induced low voltage is negligible. The flexoelectric coupling coefficient for the 1 mm thickness level and STO material property combination is obtained from Eq. (4.40) or Fig. 4.4 as k  3.5  107, confirming negligible electromechanical coupling. The beam thickness is then decreased to 1 μm while keeping the same aspect ratio (i.e., the dimensions are now 100 μm  5 μm  1 μm) and the analysis is repeated. The voltage output frequency response map for this case is shown in Fig. 4.5B. As with the 1 mm thickness case, the 1 μm-thick STO cantilever shows no noticeable change in the fundamental resonance frequency with changing load resistance. The flexoelectric coupling coefficient of this simulation case is k  3.5  104, which, again, indicates very weak electromechanical coupling. Next, the beam thickness is further decreased to 1 nm and the analysis is repeated for a 100 nm  5 nm  1 nm sample. As shown by the voltage output frequency response map in Fig. 4.5C, this nm-thick beam exhibited a certain shift in resonance frequency from short- to open-circuit conditions, which is a manifestation of significant electromechanical coupling according to Eq. (4.34). Decreasing the thickness of the cantilever (while keeping the same volumetric aspect ratio) results in increased electromechanical coupling; hence, increased mechanical to electrical energy conversion. The electromechanical coupling for this thickness level is k  0.33, which is in agreement with nearly 5.6% difference between the values of the fundamental short- and open-circuit resonance oc frequencies in Fig. 4.5C. This can easily be confirmed using Eq. (4.34) (for ωsc 1 , ω1 , and k relationship), since the resonance frequencies (frequencies of peak forced response magnitude) are very close to the natural frequencies in the lightly damped setting with ζ 1 ¼ 0.01. The electric current flowing to the resistive load is simply obtained from the voltage output using Ohm’s law. The current output (per base acceleration) frequency response maps (calculated using jv(t)/  Rlω2W0ejωt j) are also generated for the STO cantilever configurations of each geometric scale using the previously mentioned fixed volumetric aspect ratio as displayed in Fig. 4.6. The electric current

Flexoelectric-piezoelectric vibration energy harvesting

123

Fig. 4.5 Voltage output frequency response vs load resistance maps (in magnitude form and per base acceleration) for cantilevered STO harvesters with a fixed aspect ratio of 100/5/1 (L/b/h) for three different geometric scales with the following thickness (h) values: (A) 1 mm, (B) 1 μm, and (C) 1 nm.

124

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 4.6 Current output frequency response vs load resistance maps (in magnitude form and per base acceleration) for cantilevered STO harvesters with a fixed aspect ratio of 100/5/1 (L/b/h) for three different geometric scales with the following thickness (h) values: (A) 1 mm, (B) 1 μm, and (C) 1 nm.

Flexoelectric-piezoelectric vibration energy harvesting

125

decreases monotonically with increased electrical load resistance at all frequencies, which is the opposite trend as compared to the voltage output. At all frequencies, the maximum current is achieved under short-circuit conditions of the surface electrodes. As with the voltage output frequency response maps, similar trends are observed for each case study in terms of the flexoelectric coupling coefficient. The thickness levels of 1 mm and 1 μm show no noticeable shift in resonance frequency (Fig. 4.6A and B), indicating negligible electromechanical coupling, whereas the 1 nm thickness case results in significant frequency shift (Fig. 4.6C), revealing strong electromechanical coupling as discussed previously for the voltage output. As a product of two quantities which have opposite trends with changing load resistance, the electrical power output exhibits more interesting trends, such as the presence of an optimal electrical load resulting in the maximum power output at a given frequency. The electrical power output is calculated using jv(t)/  ω2W0ejωt j2/Rl (which is nothing but the product of the voltage and current figures) for each of the three geometric scales and the fixed aspect ratio discussed previously. The resulting graphs are shown in Fig. 4.7. Note that, since the output voltage and current are individually proportional to the base acceleration, the power output is proportional to base acceleration squared (hence normalized by g2), i.e., doubling the base acceleration increases the power output by a factor of 4 under the linear system assumption. The optimal load for peak power output can be determined for each case from the power output frequency response maps. For instance, the cases of both 1 mm and 1 μm-thick harvesters result in a peak power output around 1 MΩ (Fig. 4.7A and B). As with the previous frequency response maps, the 1 mm and 1 μm power output frequency response maps show the resonance frequency to be insensitive to the resistive load due to very low electromechanical coupling. Consequently, a single optimal load is observed in the power map for the fundamental vibration mode. On the other hand, the 1 nm case study exhibits two peak values for two distinct optimal electrical loads, 100 kΩ and 10 MΩ, respectively, at the short-circuit and open-circuit resonance frequencies, yielding the same power output. The existence of two peaks in the power output is also the case in strongly coupled and lightly damped piezoelectric energy harvesters [54, 63, 64]. The same power output can be extracted at the short-circuit resonance frequency (
ωsc 1 ) for a lower electrical load resistance or at the open-circuit resonance frequency (
ωoc 1 ) for a larger electrical load resistance. As a result, the former optimal condition results in larger current and lower voltage, while the latter gives larger voltage and lower current, which can also be confirmed with Figs. 4.5 and 4.6. On the practical side of energy harvesting implementation, in some cases higher voltage is preferred, such as in AC–DC conversion using a rectifier, in order to overcome forward bias voltage of diodes, when charging a storage component. In other scenarios (if the voltage output is not an issue), higher current may be preferred, e.g., to charge a storage component faster. Note that, with reduced thickness from mm to nm in Fig. 4.7, the resonant electrical power magnitude for a fixed base acceleration intensity decreases substantially (by 13 orders of magnitude); however, the volume also decreases (by 18 orders of magnitude), and hence, the power density increases

126

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 4.7 Power output frequency response vs load resistance maps (in magnitude form and per base acceleration squared) for cantilevered STO harvesters with a fixed aspect ratio of 100/5/1 (L/b/h) for three different geometric scales with the following thickness (h) values: (A) 1 mm, (B) 1 μm, and (C) 1 nm.

Flexoelectric-piezoelectric vibration energy harvesting

127

by a factor of about 105 if the mechanical base acceleration magnitude is kept the same. Finally, it is of interest to explore what happens to the structural response of the STO cantilever while generating electricity from strain gradient fluctuations in response to mechanical base excitation. The motion of the cantilever can be evaluated at any position (x1) using Eq. (4.33), while the focus is typically placed on the tip (x1 ¼ L). Fig. 4.8 shows the tip displacement map (per base acceleration input via jwrel(L, t)/  ω2W0ejωt j) for the cantilevers of all three geometric scales for the same load resistance and normalized excitation frequency ranges discussed previously. For the cases of 1 mm and 1 μm thickness levels, as another manifestation of very weak electromechanical coupling at these geometric scales, the vibration response of the cantilever is insensitive to changing electrical load resistance (Fig. 4.8A and B). That is, although some power output is delivered to the electrical load according to Fig. 4.7A and B, the level of this electrical output is so small that it is negligible as compared to mechanical (vibrational) energy of the harvester (confirmed by the coupling coefficient values), and this tiny level of electricity production does not alter the vibration response, though the converse effect is taken into account in the model (i.e., the converse flexoelectric effect is negligible at these geometric scales). Therefore, as a result of weak electromechanical coupling, Joule heating in the resistive load does not create any significant dissipation in the vibration response of the STO cantilever. However, for the cantilever with 1 nm thickness, as we know from the previous electrical output graphs (Figs. 4.5C, 4.6C, and 4.7C), the electromechanical coupling is relatively strong, and therefore, mechanical to electrical energy conversion is rather significant. As a consequence, the response of the harvester is sensitive to changing electrical load in Fig. 4.8C in the vicinity of the resonance. Certain load resistance values result in significant shunt damping, analogous to piezoelectric shunt damping [65], confirming thermodynamic consistency of the fully coupled electroelastodynamic model.

4.4

Size effects in piezoelectric energy harvesting due to flexoelectricity

We consider the problem of a bimorph piezoelectric cantilever under base excitation as shown in Fig. 4.9 for linear vibrations (i.e., linear-elastic material behavior and geometrically small oscillations). The sample geometry justifies beam assumptions, such that the width (b) and thickness (h) of the rectangular cross section are much shorter than the overhang length (L). We further assume that the beam dimensions are such that the continuum theory is applicable (the beam length for the smallest case is orders of magnitude larger than the lattice parameter of the respective material). “Static” flexoelectricity [25] is applicable since the beam thickness (smallest dimension) is much smaller than the wavelength at vibration frequencies of interest.

128

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 4.8 Tip displacement frequency response vs load resistance maps (in magnitude form and per base acceleration) for cantilevered STO harvesters with a fixed aspect ratio of 100/5/1 (L/b/h) for three different geometric scales with the following thickness (h) values: (A) 1 mm, (B) 1 μm, and (C) 1 nm.

Flexoelectric-piezoelectric vibration energy harvesting

129

Fig. 4.9 Bimorph piezoelectric cantilever undergoing bending vibrations (exhibiting combined piezoelectric and flexoelectric effects at very small thickness levels) for energy harvesting/ sensing in response to mechanical excitation. The piezoelectric layers are oppositely poled in the thickness direction (series connection) and the respective lateral faces have perfectly conductive and thin electrode layers.

4.4.1 Flexoelectrically and piezoelectrically coupled mechanical equation and modal analysis The partial differential equation governing the forced vibration of a bimorph cantilevered piezoelectric thin beam under base excitation (Fig. 4.9) is ∂2 Mðx1 , tÞ ∂5 wrel ðx1 , tÞ ∂wrel ðx1 , tÞ ∂2 wrel ðx1 , tÞ + m + c + c I s a ∂x1 2 ∂x1 4 ∂t ∂t ∂t2 d2 wb ðtÞ ¼ m dt2

(4.42)

where wrel(x1, t) is the transverse displacement of the beam (neutral axis) relative to its base and M(x1, t) is the internal bending moment at position x1 and time t, ca is the viscous air damping coefficient (mass-proportional damping), cs is the strain-rate damping coefficient (stiffness proportional damping), I is the second moment of area of the rectangular cross section, and m is the mass per unit length of the beam (m ¼ ρbh ¼ 2ρbhp, where b is the width of the beam, ρ is the mass density of the material, hp is the thickness of each piezoelectric layer), and h ¼ 2hp is the total beam thickness. The linear damping coefficients employed in Eq. (4.42) satisfy the proportional damping condition [66] so that the corresponding undamped system’s mode shapes can be used in modal analysis. The internal bending moment in Eq. (4.42) is the first moment of the axial stress field over the cross section of each layer: Z Mðx1 , tÞ ¼ b

Z

0

hp

!

hp

T11 x3 dx3 +

T11 x3 dx3

(4.43)

0

The axial strain component is due to bending only, and at a certain level (x3) from the neutral axis, is proportional to the curvature of the beam: S11 ðx1 , x3 , tÞ ¼ x3

∂2 wrel ðx1 , tÞ ∂x21

(4.44)

130

Ferroelectric Materials for Energy Harvesting and Storage

Substituting Eqs. (4.2) and (4.44) into the internal bending moment in Eq. (4.43) gives   ∂P3 cE1111 S11  e311 E3 + f1133 x3 dx3 ∂x3 hp

Z M ðx 1 , t Þ ¼ b

0

hp 

Z + 0

cE1111 S11  e311 E3

!  ∂P3 + f1133 x3 dx3 ∂x3

(4.45)

For a finite sample (in which the polarization varies continuously from its bulk value to zero at the electrode boundaries [49]), the flexoelectric term can be evaluated using integration by parts to identify the role of this term in the bending moment equation: ! Z hp ∂P3 ∂P3 bf 1133 x3 dx3 + x3 dx3 hp ∂x3 0 ∂x3 ! Z 0 Z hp ¼ bf 1133 P3 dx3 + P3 dx3 0 hp  ¼  bf 1133 hp hP3 i + bf 1133 hp hP3 i Z

0

(4.46)

where hP3i is the average polarization induced by the electric field in the beam. The spatial scale of the polarization variation at the interface is much smaller than the beam thickness, therefore hP3i  P, where the polarization in the bulk [49] can be given by P ¼ χ 33 E3

(4.47)

The electric field component E3 should be expressed in terms of the respective voltage term for the series configuration shown in Fig. 4.9. For the series connection of two oppositely poled identical piezoelectric layers, the voltage resultant is v(t). It is important to note that, for the series connection case, e311 has opposite signs for the top and bottom layers (due to opposite poling); therefore, the instantaneous electric fields are in the same direction (i.e., E3 ¼  v(t)/2hp in both layers) [54]. The polarization in Eq. (4.47) can then be substituted into Eq. (4.46) along with the appropriate electric field equations for the series connection case. This gives the following contribution from the flexoelectric effect:   bf 1133 hp hP3 i + bf 1133 hp hP3 i ¼ bμ1133 vðtÞ

(4.48)

The flexoelectric and piezoelectric coupling terms resulting from Eq. (4.45) are only a function of time, and therefore, it must be multiplied by [H(x1)  H(x1  L)] (where H(x1) is the Heaviside function) to ensure its survival when the bending moment is substituted into Eq. (4.42). The internal bending moment is then

Flexoelectric-piezoelectric vibration energy harvesting

Mðx1 , tÞ ¼ YI

∂2 wrel ðx1 , tÞ + ϑvðtÞ½H ðx1 Þ  H ðx1  LÞ ∂x21

131

(4.49)

where the coefficient of the backward coupling term (ϑ) for the series connection case is 1 ϑ ¼ e311 bhp + μ1133 b 2

(4.50)

and the bending stiffness term YI of the composite cross section (in short-circuit) is YI ¼

2b c1111 h3p 3

(4.51)

The coupled beam equation can then be obtained from Eq. (4.42) as ∂4 wrel ðx1 , tÞ ∂5 wrel ðx1 , tÞ ∂wrel ðx1 , tÞ ∂2 wrel ðx1 , tÞ + m + c I + c s a ∂t ∂t2 ∂x41 ∂x41 ∂t   dδðx1 Þ dδðx1  LÞ d2 wb ðtÞ ϑvðtÞ  ¼ m dx1 dx1 dt2 YI

(4.52)

where δ(x1) is the Dirac delta function that satisfies Eq. (4.14) for a smooth test function γ(x1). The vibration response relative to the moving base can be expressed as wrel ðx1 , tÞ ¼

∞ X

ϕr ðx1 Þηr ðtÞ

(4.53)

r¼1

Here, ηr(t) is the modal mechanical coordinate and ϕr(x1) is the mass normalized eigenfunction (obtained from the short-circuit problem) for the rth vibration mode for the series connection of the piezoelectric layers given by Eq. (4.16). The mechanical equation in modal coordinates can be obtained after substituting Eq. (4.53) into Eq. (4.52) (then multiplying the latter by the mode shape, integrating over the beam length, and applying orthogonality conditions) as d2 ηr ðtÞ dη ðtÞ + 2ζ r ωr r + ω2r ηr ðtÞ  θr vðtÞ ¼ fr ðtÞ 2 dt dt

(4.54)

where the modal piezoelectric-flexoelectric electromechanical coupling term is θr ¼ ϑ

    dϕðx1 Þ 1 dϕðx1 Þ e ¼ bh + μ b 311 p 1133 dx1 x1 ¼L 2 dx1 x1 ¼L

(4.55)

132

Ferroelectric Materials for Energy Harvesting and Storage

and the modal mechanical forcing function can be expressed as d2 wb ðtÞ fr ðtÞ ¼ m dt2

Z

L

ϕr ðx1 Þdx1

(4.56)

0

4.4.2 Flexoelectrically and piezoelectrically coupled electrical circuit equation and modal analysis To derive the governing electrical circuit equations of the bimorph series configuration, we first examine a single layer under bending vibrations. The only source of mechanical strain is assumed the axial strain due to bending, yielding the following electric displacement D3: D3 ¼ ε33 E3 + e311 S11 + μ1133

∂S11 ∂x3

(4.57)

where ε33 is the dielectric permittivity of the material, ε33 ¼ ε0 + χ 33 ¼ ð1 + χ 33 Þε0 (note that for high-K materials, χ 33 ≫1 and ε33  χ 33 ). The piezoelectrically and flexoelectrically coupled electrical circuit equation can be obtained from Eq. (4.25) resulting in the following circuit equation  Z L 3 ε33 bL dvðtÞ vðtÞ 1 ∂ wrel ðx1 , tÞ + dx1 ¼ 0 + e311 hp b + μ1133 b hp dt Rl 2 ∂x1 2 ∂t 0

(4.58)

which can be extended to the resultant of two layers in series connection (as in the case of a purely piezoelectric bimorph [54]): C

∞ dvðtÞ vðtÞ X dη ðtÞ + + θr r ¼ 0 dt Rl dt r¼1

(4.59)

where the modal electromechanical coupling is the same as Eq. (4.55), and the equivalent capacitance of two layers combined in series is C¼

ε33 bL 2hp

(4.60)

Eqs. (4.54) and (4.59) are the governing electromechanical piezoelectric-flexoelectric bimorph cantilever equations in modal coordinates.

Flexoelectric-piezoelectric vibration energy harvesting

133

4.4.3 Closed-form voltage response and vibration response at steady state For harmonic base excitation with wb(t) ¼ W0ejωt, the modal forcing function given by Eq. (4.56) can be expressed as fr(t) ¼ Frejωt, where the amplitude Fr is Z Fr ¼ ω mW

L

ϕr ðx1 Þdx1

2

(4.61)

0

Then, the steady state modal mechanical response of the beam and the steady state voltage response across the resistive load are also harmonic at the same frequency as ηr(t) ¼ Hrejωt and v(t) ¼ Vejωt, respectively, where the amplitudes Hr and V are complexed valued. Therefore, Eqs. (4.54) and (4.59) yield 

ω2r  ω2 + j2ζ r ωr ω Hr  θr V ¼ Fr



(4.62)

 ∞ X 1 + jωC V + jω θr Hr ¼ 0 Rl r¼1

(4.63)

where ζ r is the modal mechanical damping ratio. The steady state voltage response is obtained as ∞ X

vðtÞ ¼

jωθr Fr + j2ζ r ωr ω

ω 2  ω2 r¼1 r ∞ X

1 + jωC + Rl

jωθ2r 2 ω  ω2 + j2ζ r ωr ω r¼1 r

ejωt

(4.64)

Once the voltage across the electrical load is obtained, the current and power output can be calculated easily. For the case of a real-valued electrical load (i.e., a resistive load), the current delivered to the load is i(t) ¼ v(t)/Rl and the instantaneous power output is P(t) ¼ v2(t)/Rl as discussed previously. The steady state modal mechanical response of the beam (that accounts for the converse piezoelectric-flexoelectric effect) can be obtained as 3 1 jωθr Fr 7 C B ∞ 6 X ω2  ω2 + j2ζ r ωr ω 7 C 6B ϕr ðx1 Þejωt r¼1 r 7 C 6BFr  θr wrel ðx1 , tÞ ¼ ∞ 7 C 6B 2 2 2 X ω  ω + j2ζ ω ω 1 jωθ r r r 5 A 4 @ r r¼1 + jωC + 2 2 ω  ω + j2ζ ω ω Rl r r r¼1 r 20

∞ X

(4.65)

134

Ferroelectric Materials for Energy Harvesting and Storage

4.4.4 Flexoelectric-piezoelectric electromechanical coupling coefficient and size effects Using the dynamic (resonant) definition of the modal electromechanical coupling coefficient based on the difference between the open-circuit and short-circuit natural frequencies given by Eq. (4.34), the expression for the flexoelectric-piezoelectric coupling coefficient for the r-th vibration mode is found to be k2 ¼ 1+

1 4cE1111 ε33

(4.66)

3α2r ðe311 + 4μ1133 =hÞ2

where αr ¼ sin λr sinh λr + σ r(cosλr  cosh λr) and recall that h ¼ 2hp. Eq. (4.66) clearly captures the thickness dependence of the flexoelectric effect and shows that, with decreased thickness (h), the coupling coefficient (k) increases. This equation also shows the effect of material properties on the coupling coefficient and gives insight into the sign of the flexoelectric constant (μ1133), which has been reported with different signs in the literature as pointed out by Zubko et al. [24]. According to Eq. (4.66), the flexoelectric and piezoelectric constants should have the same sign to prevent nonmonotonic dependence of the coupling coefficient on the thickness (and its vanishing at a certain thickness value)—i.e., a negative e311 should be accompanied with a negative μ1133. It is worth mentioning that the coupling coefficient depends on the vibration mode, electrode coverage, etc. Typically, the first bending mode is of interest (r ¼ 1) for which full electrode coverage yields no charge cancellation.

4.4.5 Cases studies and results In this section, simulations are performed to show the effect of thickness on the electromechanical coupling and the frequency response behavior of a uniform bimorph cantilevered beam for energy harvesting. The simulations in this section are for Barium Titanate (BTO) using the atomistic value presented by Maranganti and Sharma [28] of μ1133 ¼  5.463  109 C/m along with the necessary material properties [67]: e311 ¼  4.4 C/m2, cE1111 ¼ 166 GPa, ρ ¼ 5720 kg/m3, and εs33 ¼ 12.56 nF/m.

4.4.5.1 Electromechanical coupling coefficient and size effects The electromechanical coupling coefficient due to combined piezoelectric and flexoelectric energy conversion is plotted for a range of cantilever thicknesses in Fig. 4.10. The focus is placed on the fundamental bending vibration mode (r ¼ 1), and the beam thickness in the simulations ranges from 1 mm to 1 nm. As stated previously based on Eq. (4.66), the coupling coefficient increases with decreased thickness and is illustrated graphically in Fig. 4.10. The isolated piezoelectric and flexoelectric coupling coefficients are also shown in Fig. 4.10 and it is seen that only for thickness levels below 100 nm does the flexoelectric effect become appreciable,

Flexoelectric-piezoelectric vibration energy harvesting

135

0.4 m1133 = –5.463 nC/m, e311 = –4.4 C/m2, BTO

0.35

m1133 = 0 mC/m, e311 = –4.4 C/m2 m1133 = –5.463 nC/m, e311 = 0

0.3

k

0.25 0.2 0.15 0.1 0.05

10–3

10–2

10–1

100

101

102

103

h [mm]

Fig. 4.10 Transverse mode coupling coefficient (k) vs bimorph thickness (h) of a BTO cantilever for combined piezoelectric and flexoelectric, piezoelectric only, and flexoelectric only effects (for the first bending mode).

and it strongly enhances the overall electromechanical coupling. For micron thickness and above, the overall electromechanical coupling is merely due to bulk piezoelectricity; however, the electromechanical coupling is dramatically enhanced due to flexoelectricity for thickness levels approaching the nanoscale.

4.4.5.2 Resonant energy harvesting: Electromechanical frequency response and size effects The electromechanical frequency response behavior of a bimorph cantilevered piezoelectric and flexoelectric energy harvester under base excitation is simulated with a focus on the first bending mode for a range of electrical load resistance values. Three different geometric scales are explored ranging from mm- to nm-scale. The bimorph is made of BTO and has perfectly conductive surface electrodes on the faces that are perpendicular to the transverse base excitation (Fig. 4.9). A mechanical quality factor (Q) of
50 is assumed, yielding an approximate modal mechanical damping ratio of 1% of the critical damping for resonant vibrations. Three cases with total thicknesses of 1 mm, 1 μm, and 1 nm (thickness of one layer of the bimorph is h ¼ 2hp) are analyzed while keeping a constant aspect ratio of L/b/h fixed at 100/5/1. The mechanical excitation is harmonic base acceleration, wb(t) ¼ W0ejωt. Therefore, the results are presented as frequency response magnitude maps normalized by the base acceleration quantified in terms of gravitational acceleration. To capture optimal load in power generation and respective trends with changing load, a range of

136

Ferroelectric Materials for Energy Harvesting and Storage

electrical resistive load values spanning from short- to open-circuit conditions (100 Ω to 1 GΩ) are simulated for each case. The voltage output (per base acceleration) frequency response maps for all geometric scales of the BTO bimorph is shown in Fig. 4.11. With increased electrical load resistance, the voltage increases monotonically at all frequencies, as a typical trend in energy harvesting [54]. The voltage output frequency response map for the 1 mm-thick (100 mm  5 mm  1 mm) case is shown in Fig. 4.11A. This case shows no noticeable shift in the fundamental resonance frequency with changing load resistance. The combined piezoelectric and flexoelectric coupling coefficient for the 1 mm thickness level and BTO material property combination is obtained from Eq. (4.66) or Fig. 4.10 as k ¼ 0.0652 (which is roughly the bulk piezoelectric value), confirming negligible contribution from flexoelectricity. The beam thickness is then decreased to 1 μm while keeping the same aspect ratio (i.e., the dimensions are now 100 μm  5 μm  1 μm). The voltage output frequency response map for this case is shown in Fig. 4.11B. As with the 1 mm thickness case, the 1 μm-thick BTO bimorph shows no noticeable shift in the fundamental resonance frequency with changing load resistance. The combined piezoelectric and flexoelectric coupling coefficient for this case is k ¼ 0.0655, which, again, indicates negligible flexoelectric contribution. The beam thickness is further decreased to 1 nm (beam dimensions of 100 nm  5 nm  1 nm) and the analysis is repeated. The nm-thick bimorph exhibits a shift in resonance from short- to open-circuit conditions as shown in Fig. 4.11C. This shows significant electromechanical coupling as confirmed by the coupling coefficient of k ¼ 0.365 and Fig. 4.10. The electric current flowing to the resistive load is simply obtained from the voltage output using Ohm’s law. The current output (per base acceleration) frequency response maps are also generated for the BTO bimorph for each geometric scale as shown in Fig. 4.12. The electric current output decreases with increased electrical load resistance, which is the opposite trend as compared to voltage output. At all frequencies, the maximum current is achieved under short-circuit conditions of the surface electrodes. As with the voltage output frequency response maps, similar trends are observed for each case study in terms of the coupling coefficient. The thickness levels of 1 mm and 1 μm show no noticeable shift in resonance frequency (Fig. 4.12A and B), indicating low electromechanical coupling. The 1 nm thickness case shows significant frequency shift (Fig. 4.12C), revealing strong electromechanical coupling as discussed previously for the voltage output, as a result of flexoelectric contribution. The electrical power output is calculated for each of the three geometric scales with the fixed aspect ratio. The resulting graphs are shown in Fig. 4.13. The optimal load for maximum power output can be determined for each case from the power output frequency response maps. Both the 1 mm and 1 μm-thick harvesters result in a peak power output around 100 kΩ (Fig. 4.13A and B). As with the previous frequency response maps, the 1 mm and 1 μm power output frequency response maps show the resonance frequency to be insensitive to the resistive load due to low electromechanical coupling. Consequently, a single optimal load is observed in the power map for the fundamental vibration mode for each case in Fig. 4.13A and B. However, the 1 nm-thick harvester exhibits two peak values for two distinct optimal electrical loads,

Flexoelectric-piezoelectric vibration energy harvesting

137

Fig. 4.11 Voltage output frequency response vs load resistance maps (in magnitude form and per base acceleration) for cantilevered BTO harvesters with a fixed aspect ratio of 100/5/1 (L/b/h) for three different geometric scales with the following thickness (h) values: (A) 1 mm, (B) 1 μm, and (C) 1 nm.

138

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 4.12 Current output frequency response vs load resistance maps (in magnitude form and per base acceleration) for cantilevered BTO harvesters with a fixed aspect ratio of 100/5/1 (L/b/h) for three different geometric scales with the following thickness (h) values: (A) 1 mm, (B) 1 μm, and (C) 1 nm.

Flexoelectric-piezoelectric vibration energy harvesting

139

Fig. 4.13 Power output frequency response vs load resistance maps (in magnitude form and per base acceleration) for cantilevered BTO harvesters with a fixed aspect ratio of 100/5/1 (L/b/h) for three different geometric scales with the following thickness (h) values: (A) 1 mm, (B) 1 μm, and (C) 1 nm.

140

Ferroelectric Materials for Energy Harvesting and Storage

100 kΩ and 1 MΩ, respectively, at the short-circuit and open-circuit resonance frequencies, yielding the same power output. This is an indication of a relatively strongly coupled harvester configuration, as a result of the electromechanical coupling enhancement due to the flexoelectric effect. Finally, we explore the structural response of the BTO bimorph while generating electricity from strain (piezoelectric effect) and strain gradient (flexoelectric effect) fluctuations in response to mechanical base excitation. The motion of the cantilever is evaluated at the tip using Eq. (4.65). Fig. 4.14 shows the tip displacement maps for all three geometric scales of the bimorph using the same load resistances and normalized excitation frequency range. For the 1 mm and 1 μm bimorphs, the vibration response of the cantilever is insensitive to change in electrical load resistance, again, showing negligible electromechanical coupling at these thickness levels (Fig. 4.14A and B). Therefore, as a result of weak electromechanical coupling, Joule heating in the resistive load does not create any significant dissipation in the vibration response of the BTO cantilever. However, for the bimorph with 1 nm thickness, the electromechanical coupling is relatively strong, as seen from previous electrical output graphs (Figs. 4.11C, 4.12C, and 4.13C), and therefore, mechanical to electrical energy conversion is rather significant. Consequently, the structural response of the bimorph is sensitive to changing load resistance near the resonant frequency (Fig. 4.14C). Certain load resistance values result in significant shunt damping, confirming thermodynamic consistency of the fully coupled model.

4.5

Conclusions

This chapter explores flexoelectric and combined flexoelectric-piezoelectric energy harvesting by leveraging size effects. First, an electroelastodynamic framework is developed and analyzed for flexoelectric energy harvesting from strain gradient fluctuations in centrosymmetric dielectrics, by accounting for the presence of a finite electrical load across the surface electrodes as well as two-way electromechanical coupling. The flexoelectric energy harvester model presented in this work is based on the Euler-Bernoulli beam theory and it assumes the main source of polarization to be static bulk flexoelectricity. Following recent efforts on the converse flexoelectric effect in finite samples, the proposed model properly accounts for thermodynamically consistent, symmetric, direct, and converse coupling terms, and it captures the size effect on the coupling coefficient. Based on a modal analysis procedure, closed-form solutions of the electromechanical frequency response functions (voltage across the electrical load and coupled vibration response) are given. Results of an extensive analysis are presented at different geometric scales (mm, μm, and nm thickness levels with a fixed aspect ratio) for a Strontium Titanate (STO) cantilever that is shunted to a resistive electrical load for quantifying the electrical power output and its feedback on the vibration response due to the converse effect. The transverse mode flexoelectric coupling coefficient, k, (as a direct and compact measure of energy conversion) is analytically extracted from the short- and

Flexoelectric-piezoelectric vibration energy harvesting

141

Fig. 4.14 Tip displacement frequency response vs load resistance maps (in magnitude form and per base acceleration) for bimorph cantilevered BTO harvesters with a fixed aspect ratio of 100/ 5/1 (L/b/h) for three different geometric scales with the following thickness (h) values: (A) 1 mm, (B) 1 μm, and (C) 1 nm.

142

Ferroelectric Materials for Energy Harvesting and Storage

open-circuit natural frequencies. Dependence of the coupling coefficient on the thickness and material parameters (figure of merit) is discussed in detail. The flexoelectric energy conversion and harvesting become significant only at nm thickness levels for typical flexoelectric coefficients obtained from atomistic simulations (with order of magnitude 109 C/m). For instance, the negligible flexoelectric coupling of an STO cantilever at the mm thickness level increases by six orders of magnitude when the thickness is reduced to nm-level. It is no surprise that the flexoelectric power output of an individual nanoscale centrosymmetric dielectric cantilever is very low. However, the quantitative understanding of this size dependence provided by the presented framework could help the designer tailor the individual beam dimensions (cross section for bending vibrations) and fabricate a cluster of flexoelectric energy harvesters to maximize energy conversion under the constraints of a fixed material volume, target frequency range, among other parameters. Substantially high values of flexoelectric coefficients (104 C/m) reported in the literature based on experiments conducted with mm-thick samples result in extremely high values of the coupling coefficient (k), yielding values nearly unity for all submicron thickness levels, and therefore, suggesting very high energy conversion even at μm thickness level, which is not the case in reality. This observation confirms that the identified constants for certain mm-thick samples probably do not represent bulk flexoelectricity and are not valid at other scales (and cannot be used in the proposed model). Therefore, the need for rigorous experiments at smaller scales is pointed out (at submicron scales, and preferably for less than 10 nm sample thickness, by eliminating or controlling the effects other than bulk flexoelectricity). While this work has considered static bulk flexoelectricity alone, future electroelastodynamic modeling efforts may consider incorporating surface piezoelectricity and flexoelectricity [24, 25], as well as size dependence of the mechanical quality factor (Q) due to intrinsic dissipation mechanisms [58] (since k2Q is a more complete figure of merit for resonant energy harvesting). The modeling framework is then extended to the combined transverse mode flexoelectric and piezoelectric energy harvesting for the bending vibration of a piezoelectric cantilever by accounting for two-way electromechanical coupling. The modeling framework properly accounts for thermodynamically consistent, symmetric, direct, and converse coupling terms, and it captures the size effect on the combined flexoelectric-piezoelectric coupling coefficient. Based on a modal analysis procedure, closed-form solutions of the electromechanical frequency response functions are presented along with various case studies for a broad range of geometric parameters. Thickness dependence of the electromechanical coupling (which is a measure of energy conversion) is analytically extracted and its size dependence is observed also in simulations of the electromechanical frequency response functions. The flexoelectric-piezoelectric coupling increases from the bulk piezoelectric value of k  0.0652 at the mm-scale to k  0.365 at the nm-scale owing to flexoelectric contribution. Overall, since the coupling coefficient is thickness-dependent, the energy conversion dramatically increases in submicron thickness levels due to the flexoelectric effect. The proposed model can be used for parameter identification

Flexoelectric-piezoelectric vibration energy harvesting

143

as well as performance quantification and optimization in combined flexoelectric and piezoelectric energy harvesting.

Acknowledgment This work was supported by the National Science Foundation grant CMMI-1463339, which is gratefully acknowledged.

References [1] K. Cook-Chennault, N. Thambi, A. Sastry, Powering MEMS portable devices—A review of non-regenerative and regenerative power supply systems with special emphasis on piezoelectric energy harvesting systems, Smart Mater. Struct. 17 (4) (2008) 043001. [2] J.Q. Liu, et al., A MEMS-based piezoelectric power generator array for vibration energy harvesting, Microelectron. J. 39 (5) (2008) 802–806. [3] W.J. Choi, et al., Energy harvesting MEMS device based on thin film piezoelectric cantilevers, J. Electroceram. 17 (2–4) (2006) 543–548. [4] Y. Qin, X.D. Wang, Z.L. Wang, Microfibre-nanowire hybrid structure for energy scavenging, Nature 451 (7180) (2008) 809–U5. [5] U. Aridogan, I. Basdogan, A. Erturk, Multiple patch–based broadband piezoelectric energy harvesting on plate-based structures, J. Intell. Mater. Syst. Struct. 25 (14) (2014) 1664–1680. [6] M. Carrara, et al., Metamaterial-inspired structures and concepts for elastoacoustic wave energy harvesting, Smart Mater. Struct. 22 (6) (2013) 065004. [7] S. Tol, F. Degertekin, A. Erturk, Gradient-index phononic crystal lens-based enhancement of elastic wave energy harvesting, Appl. Phys. Lett. 109 (6) (2016) 063902. [8] L. Moro, D. Benasciutti, Harvested power and sensitivity analysis of vibrating shoemounted piezoelectric cantilevers, Smart Mater. Struct. 19 (11) (2010) 115011. [9] N. Elvin, A. Erturk, Advances in Energy Harvesting Methods, Springer Science & Business Media, 2013. [10] K.A. Cunefare, et al., Energy harvesting from hydraulic pressure fluctuations, Smart Mater. Struct. 22 (2) (2013) 025036. [11] X.-B. Cui, C.-P. Huang, J.-H. Hu, Sound energy harvesting using an acoustic grating, J. Appl. Phys. 117 (10) (2015) 104502. [12] S. Shahab, M. Gray, A. Erturk, Ultrasonic power transfer from a spherical acoustic wave source to a free-free piezoelectric receiver: Modeling and experiment, J. Appl. Phys. 117 (10) (2015) 104903. [13] A. Erturk, et al., On the energy harvesting potential of piezoaeroelastic systems, Appl. Phys. Lett. 96 (18) (2010) 184103. [14] C. De Marqui Jr, A. Erturk, Electroaeroelastic analysis of airfoil-based wind energy harvesting using piezoelectric transduction and electromagnetic induction, J. Intell. Mater. Syst. Struct. 24 (7) (2013) 846–854. [15] J. Allen, A. Smits, Energy harvesting eel, J. Fluids Struct. 15 (3–4) (2001) 629–640. [16] A. Erturk, G. Delporte, Underwater thrust and power generation using flexible piezoelectric composites: an experimental investigation toward self-powered swimmer-sensor platforms, Smart Mater. Struct. 20 (12) (2011) 125013. [17] Standards Committee of the IEEE Ultrasonics, Ferroelectrics, and Frequency Control Society, IEEE Standard on Piezoelectricity. ANSI/IEEE Std 176–1987, (1984)p. 1–66.

144

Ferroelectric Materials for Energy Harvesting and Storage

[18] S. Trolier-McKinstry, P. Muralt, Thin film piezoelectrics for MEMS, J. Electroceram. 12 (1–2) (2004) 7–17. [19] E. Fukada, History and recent progress in piezoelectric polymers, IEEE Trans Ultrason Ferroelectr Freq Control. 47 (6) (2000) 1277–1290. [20] M.D. Maeder, D. Damjanovic, N. Setter, Lead free piezoelectric materials, J. Electroceram. 13 (1–3) (2004) 385–392. [21] M. Majdoub, P. Sharma, T. Cagin, Enhanced size-dependent piezoelectricity and elasticity in nanostructures due to the flexoelectric effect, Phys. Rev. B 77 (12) (2008) 125424. [22] A. Tagantsev, Piezoelectricity and flexoelectricity in crystalline dielectrics, Phys. Rev. B 34 (8) (1986) 5883. [23] A.K. Tagantsev, V. Meunier, P. Sharma, Novel electromechanical phenomena at the nanoscale: Phenomenological theory and atomistic modeling, MRS Bull. 34 (09) (2009) 643–647. [24] P. Zubko, G. Catalan, A.K. Tagantsev, Flexoelectric effect in solids, Annu. Rev. Mater. Res. 43 (2013) 387–421. [25] P. Yudin, A. Tagantsev, Fundamentals of flexoelectricity in solids, Nanotechnology 24 (43) (2013) 432001. [26] T.D. Nguyen, et al., Nanoscale flexoelectricity, Adv. Mater. 25 (7) (2013) 946–974. [27] J. Zhang, C. Wang, C. Bowen, Piezoelectric effects and electromechanical theories at the nanoscale, Nanoscale 6 (22) (2014) 13314–13327. [28] R. Maranganti, P. Sharma, Atomistic determination of flexoelectric properties of crystalline dielectrics, Phys. Rev. B 80 (5) (2009) 054109. [29] V. Mashkevich, K. Tolpygo, Electrical, optical and elastic properties of diamond type crystals, Sov. Phys. JETP 5 (3) (1957) 435–439. [30] S.M. Kogan, Piezoelectric effect during inhomogeneous deformation and acoustic scattering of carriers in crystals, Sov. Phys. Solid State 5 (10) (1964) 2069–2070. [31] V. Indenbom, E. Loginov, M. Osipov, Flexoelectric effect and crystal-structure, Kristallografiya 26 (6) (1981) 1157–1162.  [32] A. Buka, N. Eber, Flexoelectricity in Liquid Crystals: Theory, Experiments and Applications, World Scientific, 2012. [33] A.G. Petrov, Flexoelectricity of model and living membranes, Biochim. Biophys. Acta Biomembr. 1561 (1) (2002) 1–25. [34] W. Ma, L.E. Cross, Large flexoelectric polarization in ceramic lead magnesium niobate, Appl. Phys. Lett. 79 (26) (2001) 4420–4422. [35] W.H. Ma, L.E. Cross, Observation of the flexoelectric effect in relaxor Pb(Mg1/3Nb2/3) O-3 ceramics, Appl. Phys. Lett. 78 (19) (2001) 2920–2921. [36] W. Ma, L.E. Cross, Flexoelectric polarization of barium strontium titanate in the paraelectric state, Appl. Phys. Lett. 81 (18) (2002) 3440–3442. [37] W.H. Ma, L.E. Cross, Strain-gradient-induced electric polarization in lead zirconate titanate ceramics, Appl. Phys. Lett. 82 (19) (2003) 3293–3295. [38] W.H. Ma, L.E. Cross, Flexoelectric effect in ceramic lead zirconate titanate, Appl. Phys. Lett. 86 (7) (2005) 072905. [39] W.H. Ma, L.E. Cross, Flexoelectricity of barium titanate, Appl. Phys. Lett. 88 (23) (2006) 232902. [40] L.E. Cross, Flexoelectric effects: Charge separation in insulating solids subjected to elastic strain gradients, J. Mater. Sci. 41 (1) (2006) 53–63. [41] P. Zubko, et al., Strain-gradient-induced polarization in SrTiO 3 single crystals, Phys. Rev. Lett. 99 (16) (2007) 167601.

Flexoelectric-piezoelectric vibration energy harvesting

145

[42] P. Zubko, et al., Erratum: Strain-Gradient-Induced Polarization in SrTiO 3 Single Crystals [Phys. Rev. Lett. 99, 167601 (2007)], Phys. Rev. Lett. 100 (19) (2008) 199906. [43] W. Huang, et al., Scaling effect of flexoelectric (Ba, Sr) TiO3 microcantilevers, Phys. Status Solidi RRL 5 (9) (2011) 350–352. [44] M. Majdoub, P. Sharma, T. C¸ag˘in, Dramatic enhancement in energy harvesting for a narrow range of dimensions in piezoelectric nanostructures, Phys. Rev. B 78 (12) (2008) 121407. [45] Q. Deng, et al., Nanoscale flexoelectric energy harvesting, Int. J. Solids Struct. 51 (18) (2014) 3218–3225. [46] P. Harris, Mechanism for the shock polarization of dielectrics, J. Appl. Phys. 36 (3) (1965) 739–741. [47] E.V. Bursian, O.I. Zaikovskii, Changes in the curvature of a ferroelectric film due to polarization, Sov. Phys. Solid State 10 (5) (1968) 1121–1124. [48] R.D. Mindlin, Polarization gradient in elastic dielectrics, Int. J. Solids Struct. 4 (6) (1968) 637–642. [49] A.K. Tagantsev, A.S. Yurkov, Flexoelectric effect in finite samples, J. Appl. Phys. 112 (4) (2012) 044103. [50] A.G. Moura, A. Erturk, Electroelastodynamics of flexoelectric energy conversion and harvesting in elastic dielectrics, J. Appl. Phys. 121 (6) (2017) 064110. [51] A.G. Moura, A. Erturk, Combined piezoelectric and flexoelectric effects in resonant dynamics of nanocantilevers, J. Intell. Mater. Syst. Struct. 29 (20) (2018) 3949–3959. [52] P. Yudin, R. Ahluwalia, A. Tagantsev, Upper bounds for flexoelectric coefficients in ferroelectrics, Appl. Phys. Lett. 104 (8) (2014) 082913. [53] A. Erturk, D.J. Inman, A distributed parameter electromechanical model for cantilevered piezoelectric energy harvesters, J. Vib. Acoust. 130 (4) (2008) 041002. [54] A. Erturk, D.J. Inman, Piezoelectric Energy Harvesting, Wiley, Chichester, UK, 2011. [55] S. Leadenham, A. Erturk, Unified nonlinear electroelastic dynamics of a bimorph piezoelectric cantilever for energy harvesting, sensing, and actuation, Nonlinear Dyn. 79 (3) (2015) 1727–1743. [56] S. Leadenham, A. Erturk, M-shaped asymmetric nonlinear oscillator for broadband vibration energy harvesting: Harmonic balance analysis and experimental validation, J. Sound Vib. 333 (23) (2014) 6209–6223. [57] G.A. Lesieutre, C.L. Davis, Can a coupling coefficient of a piezoelectric device be higher than those of its active material? in: Proc. SPIE 3041, Smart Structures and Materials 1997: Smart Structures and Integrated Systems, San Diego, CA, USA, 1997, pp. 281–292. [58] K. Kunal, N. Aluru, Intrinsic dissipation in a nano-mechanical resonator, J. Appl. Phys. 116 (9) (2014) 094304. [59] R. Bell, G. Rupprecht, Elastic constants of strontium titanate, Phys. Rev. 129 (1) (1963) 90. [60] A. Kityk, et al., Low-frequency superelasticity and nonlinear elastic behavior of SrTiO 3 crystals, Phys. Rev. B 61 (2) (2000) 946. [61] K. Abe, S. Komatsu, Epitaxial growth of SrTiO3 films on Pt electrodes and their electrical properties, Jpn. J. Appl. Phys. 31 (9S) (1992) 2985. [62] J.Y. Fu, et al., Experimental studies of the converse flexoelectric effect induced by inhomogeneous electric field in a barium strontium titanate composition, J. Appl. Phys. 100 (2) (2006) 024112. [63] Y. Shu, I. Lien, Analysis of power output for piezoelectric energy harvesting systems, Smart Mater. Struct. 15 (6) (2006) 1499.

146

Ferroelectric Materials for Energy Harvesting and Storage

[64] Y. Shu, I. Lien, Efficiency of energy conversion for a piezoelectric power harvesting system, J. Micromech. Microeng. 16 (11) (2006) 2429. [65] G.A. Lesieutre, G.K. Ottman, H.F. Hofmann, Damping as a result of piezoelectric energy harvesting, J. Sound Vib. 269 (3) (2004) 991–1001. [66] L. Meirovitch, Fundamentals of Vibrations, McGraw Hill, 2001. [67] R. Bechmann, Elastic, piezoelectric, and dielectric constants of polarized barium titanate ceramics and some applications of the piezoelectric equations, J. Acoust. Soc. Am. 28 (3) (1956) 347–350.

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

5

Vamsi C. Meesalaa and Muhammad R. Hajjb a Department of Biomedical Engineering and Mechanics, Virginia Tech, Blacksburg, VA, United States, bDepartment of Civil, Environmental and Ocean Engineering, Davidson Laboratory, Stevens Institute of Technology, Hoboken, NJ, United States

5.1

Introduction

Piezoelectric materials are receiving increased attention in research and development bearing to their multifunctional behavior which allows researchers to employ them to various applications. Piezoelectricity is the phenomenon exhibited by materials with asymmetrical crystal lattices that generate electricity for an applied stress/strain. Conversely, they can also be used to create a mechanical motion or strain in the material due to an applied electric field, referred as inverse piezoelectric behavior. Due to these complementary behaviors, the piezoelectric materials have been proposed for developing self-powered microsystems and sensors [1–6], energy harvesters from fluid flows [7–9], broadband energy harvesters [10–13], and impedance-based structural health monitoring devices [5, 14, 15] and active-passive vibration control and damping systems [16–21]. In more recent studies, piezoelectric/ceramic materials have been proposed in biomedical applications as sensors for knee joints [22, 23] and as receivers to oxygenate tumors and charge pacemakers [24–26]. Structural designs that exploit bistable configurations or those that inherently induce large strains are proposed to augment the power generated [27–29]. It has to be noted that the large strains can potentially cause the material to behave in a nonlinear manner and exhibit amplitude-dependent resonant frequencies, super-harmonics in the response, saturation, and hysteresis behaviors. Various sources contribute to the nonlinear response of piezoelectric systems. These include large deformation of the substrate, large amplitude forcing, or nonlinear material properties. In general, the impact of nonlinear material properties has not been considered, mostly because of the lack of consistent empirical or even manufacturer’s data. However, few investigators have investigated the nonlinear response of unimorph and bimorph piezoelectric beam systems by considering the nonlinear material properties of the piezoelectric material [30–37]. Employing a nonlinear electromechanical constitutive relation, Aurelle et al. [32] and Guyomar et al. [33] studied the saturation and hysteresis behavior that they observed in their experiments on ultrasonic transducers. In the context of piezoelectric beam systems, Von Wagner and Hagedorn [34] derived an equation for the electric enthalpy density that accounts Ferroelectric Materials for Energy Harvesting and Storage. https://doi.org/10.1016/B978-0-08-102802-5.00005-4 © 2021 Elsevier Ltd. All rights reserved.

148

Ferroelectric Materials for Energy Harvesting and Storage

for nonlinearities in piezoelectric response. They noted a saturation and hysteresis behaviors of the response and also demonstrated that a bi-morph cannot be used to study the quadratic nonlinearities as the symmetrical nature of the bi-morph nullifies any contribution of quadratic nonlinearities. An extensive discussion about modeling of nonlinear behavior of piezoelectric materials has been mentioned by Leadenham and Erturk [37]. In order to design, analyze, optimize, and/or exploit the nonlinear response, there is a need to develop a parameter identification scheme to identify the nonlinear material parameters. The vibration behavior of any structure is governed by the intrinsic material characteristics of the structure and for the same reason it is probably the most direct way for parameter identification. For instance, free vibrational response is often used for estimating damping ratio and fundamental natural frequency. In the context of piezoelectric materials, the transfer functions are used to identify characteristic capacitance and electromechanical coupling of the harvester [10]. Since, the nonlinear material parameters are amplitude dependent and exhibit several nontrivial phenomena, they cannot be uniquely identified using simple transfer functions. Nayfeh [38, 39] presented a parameter identification procedure that exploits the jump and hysteresis phenomena in subcritical bifurcation exhibited by a generic coupled nonlinear oscillator to estimate the nonlinear coefficients in governing equations. The identification procedure is based on the approximate solution obtained by perturbation techniques. Zavodney [40] discussed a generalized theory to identify the nonlinearity in structural systems using random and impulse excitations and also the forcing and frequency sweeps. Hajj et al. [41] utilized approximate solution in conjunction with higher order spectral methods to identify nonlinear damping of a three beam and two mass system. Kerschen et al. [42] presented a concise review of nonlinear system identification in structures. As discussed earlier, piezoelectric materials exhibit material nonlinearity and there has been a limited research so far into developing a parameter identification procedure [35, 36]. Mahmoodi et al. [35] studied a piezoelectrically actuated microcantilever beam by considering only the nonlinear stiffness that results in quadratic nonlinearity. They assumed that the microsensors function at small electric fields and that the nonlinearity in electric field can be neglected. As suggested by Nayfeh [39], they implemented the method of multiple scales [43] to the coupled nonlinear problem and identified the nonlinear stiffness from the frequency response. Stanton et al. [36] considered the study by Von Wagner and Hagedorn [34] on a bimorph and demonstrated a parameter identification scheme to characterize the nonlinear material parameters by implementing the harmonic balance method. Their strategy involves fitting the deflection and voltage generated by the bimorph as a 10th-order polynomial which can be mathematically intensive. We noted that there is a need to develop a holistic parameter identification scheme that utilizes the vibration behavior to characterize most of the quadratic and cubic material nonlinearities in a general nonlinear constitutive relation. In this chapter, we present two ways of identifying the material parameters that govern the nonlinear behavior of piezoelectric systems by using a thin beam as a substrate and a thin piezoelectric layer with specified length, also referred to as unimorph. In particular, we detail the parameter identification procedure of the nonlinear piezoelectric material parameters proposed by Meesala and Hajj [44] on a unimorph that is

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

149

subjected to direct resonant excitation and extend the ideology to a parametric excitation. To accomplish this, we first introduce the constitutive relations and electromechanical enthalpy used for modeling the nonlinear behavior and then derive the governing equations of the structure by using (i) stress, strain, electric displacement, and electric field constitutive relations and (ii) electromechanical enthalpy. We then solve the governing equations for approximate solution using the method of multiple scales and present a parameter identification strategy. Finally, we assume values to the nonlinear material parameters and show the prediction capability of the proposed parameter identification strategy.

5.2

Representation and implementation of constitutive relations

Different approaches have been used to develop the constitutive relations that include the nonlinear material parameters required to model the nonlinear response of piezoelectric beam systems. For a thin beam undergoing bending the stress, σ is related to the strain, ε σ ¼ Yε

(5.1)

where σ and ε are considered along the curvilinear axis only, and Y represents Young’s modulus. The constitutive relations that govern the electromechanical response of the piezoelectric material are either (i) expressed by a relation between the stress (σ ij) and electric displacement (Dij) or (ii) derived from the electromechanical enthalpy density 

(Δ ). These quantities are interrelated to each other in that 



∂Δ ∂Δ and Di ¼  σ ij ¼ ∂Ei ∂εij

(5.2a)

where Eij is the strain tensor and Ei is the electric field strength vector. Eq. (5.2a) yields the compatibility condition 

∂σ ij ∂Di ∂2 Δ ¼ ¼ ∂Ei ∂εij ∂εij ∂Ei

(5.2b)

The constitutive relation for a transversely isotropic piezoelectric beam under bending strain when the response is limited to the linear regime is written as p σ 11 ¼ Y11 ε11  e31 E3

(5.3a)

D3 ¼ e31 ε11 + ES33 E3

(5.3b)

and

150

Ferroelectric Materials for Energy Harvesting and Storage

p where Y11 , e31, and ES33 are, respectively, Young’s modulus, coupling coefficient, and dielectric constant at constant strain of piezoelectric material. The linear constitutive relations for a thick beam (Timoshenko beam) that accounts for a transverse shear strain/stress and for a thin plate that takes Poisson’s effect into consideration are given by Erturk and Inman [45]. The earlier relations can alternatively be derived from the enthalpy density written as

1 p 2 1 S 2  Δ ¼ 2 Y11 ε11  e31 ε11 E3  2 E33 E3

(5.4)

which yields Eqs. (5.3a), (5.3b) when differentiated with respect to εij and Ei in accordance with Eq. (5.2a). Extending the representation of Eq. (5.4), nonlinearities in the material behavior can be considered by representing them in the form of higher-order terms in the enthalpy density and subsequently in the stress and electrical displacement relations. For instance, a quadratic stiffness coefficient, say 16 α1 , which is a material nonlinearity, is accounted for by adding 16 α1 ε311 to the right-hand side of Eq. (5.4), which yields 1 p 2 1 S 2 1 3  Δ ¼ 2 Y11 ε11  e31 ε11 E3  2 E33 E3 + 6 α1 ε11 and

(5.5)

1 p ε11  e31 E3 + α1 ε211 σ 11 ¼ Y11 2

(5.6)

when differentiated with respect to ε11 in accordance with Eq. (5.2a). The electric displacement constitutive relation of Eq. (5.3b) remains unaltered because the incorporated material nonlinearity is independent of the electrical coupling. A general enthalpy density that accounts for quadratic and cubic nonlinearities of the piezoelectric material can then be written as an extension of Eq. (5.6) as 1 p 2 1 1 S 2 1 3 1 3 1 4  2 Δ ¼ 2 Y11 ε11  e31 ε11 E3  2 α3 E3 ε11  2 E33 E3 + 6 α1 ε11  6 α2 ε11 E3 + 8 α4 ε11

(5.7)

The enthalpy density, as defined earlier, assumes that the level of the applied electric field, E3, is low and hence higher orders in E3 are neglected in the formulation. Following Eq. (5.2a), we differentiate to obtain the constitutive relations as 1 1 1 p ε11  e31 E3  α3 E3 ε11 + α1 ε211  α2 ε211 E3 + α4 ε311 σ 11 ¼ Y11 2 2 2

(5.8a)

1 1 D3 ¼ e31 ε11 + α3 ε211 + ES33 E3 + α2 ε311 2 6

(5.8b)

Next, we develop the mathematical framework to derive the governing equations of the coupled electromechanical problem by using the generalized Hamilton’s principle.

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

5.3

151

Direct excitation

As a first example, we consider the case of direct excitation of a cantilever beam. We assume a thin beam as a substrate and a thin piezoelectric layer as shown in Fig. 5.1. The beam, which has a length l, width b, mass per unit length ρs, and a thickness hs is bonded to a piezoelectric layer of width b, thickness hp, and mass per unit length ρp over the complete length and is harmonically excited with a frequency of Ω that is close to its natural frequency. The amplitude of the excitation at the base of the beam is denoted by ub. Furthermore, the piezoelectric layer is connected to an electric load having a resistance R to harvest energy.

5.3.1 Modeling using nonlinear stress and electric displacement constitutive relations Considering the system’s kinetic energy, T, potential energy, U, electrical energy, We, and work done by nonconservative forces, Wnc, the generalized Hamilton’s principle is written as Z

t2

ðδ½T  U + We  + δWnc Þ dt ¼ 0

(5.9)

t1

The kinetic energy is given by T¼

1 2

Z

l 0

  _ , tÞ + u_b ðtÞ2 ds _ , tÞ2 + ½wðs ρeq uðs

(5.10)

where u(s, t) and w(s, t) are, respectively, the axial contraction and transverse displacement, as shown in Fig. 5.1, and ρeq ¼ ρp + ρs. Accounting for the structural damping, the variational nonconservative work done is written as Z l (5.11) _ ds  QδVðtÞ δWnc ¼  c1 wðs,tÞδwðs,tÞ 0

Piezoelectric sheet Ub COS(Wt)

u(s,t) w(s,t)

hs+hp

hs/2

Fig. 5.1 Schematic representation of the beam with a piezoelectric patch subjected to direct excitation.

152

Ferroelectric Materials for Energy Harvesting and Storage

where Q is the net generated electric charge. Assuming a linear elastic restoring force, the elastic potential energy is defined as Z 1 σ ij εij dV U¼ (5.12) 2 V which yields Z  1  δU ¼ εij δσ ij + σ ij δεij dV 2 V

(5.13)

where σ ij is a linear function of εij. We should note here that the earlier definition is based on assuming a linear elastic restoring force. On the other hand, without making any assumptions about the nature of the restoring force, the work done by the restoring force (stress), σ ij, to cause a variational strain, δεij, can be defined as variation in the potential energy and written as [46, 47] Z δU ¼

V

σ ij δεij dV

(5.14)

To emphasize the impact of using the variation in elastic potential energy as represented in Eq. (5.13), instead of Eq. (5.14), we consider a general relation for the normal stress of a thin beam and the corresponding strain as σ 11 ¼ kεn11

(5.15a)

which yields δσ 11 ¼ n

σ 11 δε11 ε11

(5.15b)

where n is positive integer. We then derive the variation in the potential energy by considering definitions given by Eqs. (5.13), (5.14) and compare them. To this end, we consider the elastic potential energy given by Eq. (5.13), and write Z 1 ðε11 δσ 11 + σ 11 δε11 Þ dV δUa ¼ (5.16) 2 V Considering the variation of elastic potential energy given by Eq. (5.14), we write Z δUb ¼

V

σ 11 δε11 dV

(5.17)

Considering the general stress and strain relation defined in Eq. (5.15b), the variation in potential energy based on the definition in Eq. (5.13) is given by δUa ¼

n+1 2

Z V

σ 11 δε11 dV

(5.18a)

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

and the variational based on Eq. (5.14) is given by Z δUb ¼ σ 11 δε11 dV V

153

(5.18b)

We note that Ua6¼Ub for n > 1. In other words, the variationals do not yield the same expressions for a nonlinear relationship between the normal stress and strain. As explained earlier, this is because the potential energy defined in Eq. (5.12) is derived by assuming a linear elastic force and is thus limited to the linear regime. Similarly, the variational work done by electrical forces is expressed as Z δWe ¼

D3 δE3 dV

(5.19)

v

From Eq. (5.14) and assumed constitutive relations of Eqs. (5.8a), (5.8b), we write the variation in the potential energy as  Z  1 2 1 2 1 3 p Y11 ε11  e31 E3  α3 E3 ε11 + α1 ε11  α2 ε11 E3 + α4 ε11 δε11 dV δU ¼ 2 2 2 V (5.20a) and the variational of the electric forces as Z  δWe ¼

V

 1 1 e31 ε11 + α3 ε211 + ES33 E3 + α2 ε311 δE3 dV 2 6

(5.20b)

We then determine δU  δWe as Z  1 p Y11 ε11 δε11  e31 ½E3 δε11 + ε11 δE3   ES33 E3 δE3 + α1 ε211 δε11 δU  δWe ¼ 2 V    1 2 1 2 1 3 1 3  α3 E3 ε11 δε11 + ε11 δE3  α2 ε11 E3 δε11 + α2 ε11 δE3 + α4 ε11 δε11 dV 2 2 6 2 (5.21) In the following section, we compare this hdifference with the variational of the R  i electromechanical enthalpy given by δΔ ¼ δ V Δ dV .

5.3.2 Modeling using electromechanical enthalpy Using the electromechanical enthalpy density, the generalized Hamilton’s principle is written as Z t2 ðδT  δΔ + δWnc Þ dt ¼ 0 (5.22) t1

where Δ ¼

RlR 0 A



Δ dA ds.

154

Ferroelectric Materials for Energy Harvesting and Storage

The kinetic energy of the system remains unchanged. So, to this end, by considering the assumed electromechanical enthalpy density of Eq. (5.7), we write the variation of the electromechanical enthalpy as  1 p 2 1 1 1 1 1 ε11  e31 ε11 E3  α3 E3 ε211  ES33 E23 + α1 ε311  α2 ε311 E3 + α4 ε411 Δ ¼ Y11 2 2 2 6 6 8

(5.23a)

which yields Z

Z 



1 p Y11 ε11 δε11  e31 ½E3 δε11 + ε11 δE3   ES33 E3 δE3 + α1 ε211 δε11 2 V V    1 2 1 2 1 3 1 3  α3 E3 ε11 δε11 + ε11 δE3  α2 ε11 E3 δε11 + α2 ε11 δE3 + α4 ε11 δε11 dV 2 2 6 2 (5.23b)

δΔ ¼

δ Δ dV ¼

Comparing Eqs. (5.21), (5.23b), we note that δU  δWe ¼ δΔ. Inspecting the generalized Hamilton principle descriptions in Eqs. (5.9), (5.22), for a given system, the integrands are identical if δU  δWe ¼ δΔ, which is the case. By assuming that the beam is inextensible, the axial contraction and the lateral displacement are related by [47] Z

s

uðs,tÞ ¼ 0

i2 1 h ð1,0Þ w ðy, tÞ dy 2

(5.24a)

The axial strain is written as

  1 ε11 ¼ ½y  y0 κ ¼ ½y  y0 w00 1 + w02 2

(5.24b)

where y0 is the neutral axis defined by Y p hp ðhs + hp Þ  y0 ¼  11s p 2 Y11 hs + Y11 hp

(5.24c)

Using Eq. (5.24a), the kinetic energy defined in Eq. (5.10) is rewritten as ! Z s 2 Z 1 l 1 ∂  02  2 T¼ ρ w dy + ½w_ + u_b  ds 2 0 eq 0 2 ∂t

(5.25)

By including the elastic potential energy of the beam substrate and using the strain relation of Eq. (5.24b), the electromechanical enthalpy is written as  Z 1 02 VH 1 l VH e31 Λ1 w 1 + w ds + α3 Ip w002 ds h hp 2 2 p 0 0 0 Z Z Z Z 1 l S VH2 b 1 l 1 l VH 1 l 003 003 E ds  α1 Λ2 w ds  α2 Λ2 w ds + α4 Λ3 w004 ds  hp 2 0 33 hp 6 0 6 0 8 0 (5.26)

1 Δ¼ 2

Z

l

 YIeq w 1 + w02 ds  002

Z

l

00



Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

155

p where YIeq ¼ ðYI Þs + Y11 Ip ,

" 3  3 # Z hs hs hs 2 s s b 2  y0 + + y0 ðYI Þs ¼ h Y11 bðy  y0 Þ dy ¼ Y11 s 2 2 3  2

(5.27a)

" 3  3 # Z h s + hp b hs hs 2 2 (5.27b) + hp  y 0   y0 Ip ¼ h bðy  y0 Þ dy ¼ s 2 2 3 2 " " 2  2 # 4  4 # b hs hs b hs hs + h p  y0   y0 + hp  y 0   y0 Λ1 ¼ , Λ2 ¼ 2 4 2 2 2 2 (5.27c) Λ3

E3

b ¼ 5

"

hs + hp  y 0 2

8 VH > > > <  hp ¼ > >0 > :

5

 5 # hs b  y0  and Is ¼ h3s 2 12

hs hs  y  + hp 2 2 hs hs  y< 2 2

(5.27d)

(5.27e)

and VH ðs, tÞ ¼ VðtÞ½HðsÞ  Hðs  l1 Þ ¼ VðtÞHf ðsÞ, VH ðs, tÞn ¼ VðtÞn ½HðsÞ  Hðs  l1 Þ ¼ VðtÞn Hf ðsÞ and H(x) is Heaviside step function. The virtual work done by the nonconservative forces is represented by Z l (5.28) δWnc ¼  c1 wð0,1Þ ðs, tÞδwðs,tÞ ds  QδVðtÞ 0

where c1 is structural damping coefficient and Q is the generated electrical charge. Substituting Eqs. (5.25), (5.26), (5.28) into the generalized Hamilton’s principle given in Eq. (5.22), we obtain the distributed parameter governing equation of motion as ρeq w€  c1 w_  YIeq ðw0000 + ½w0 ðw0 w00 Þ0 0 Þ  

 Z Z 0 ρeq 0 s θ ∂2  02  w w dy dθ 2 2 l 0 ∂t

w02 VHf 00 w00 w0 VHf 0 VHf 00 + α1 Λ2 ½w0002 + w00 w0000  + e31 Λ1 + + 2hp hp hp  VHf 00 w002 2w00 w000 VHf 0 w0002 VHf w0000 w00 VHf + + + + α2 Λ2 hp hp 2hp hp   00 00 0 000 0000 VHf w 2Vhf w VHfw 3 002 0000 00 0002 + α4 Λ3 3w w  w w  ρeq u€b ¼ 0 + +  α3 Ip hp hp hp 2 (5.29)

156

Ferroelectric Materials for Energy Harvesting and Storage

and the corresponding electrical equation as Z

Z l 003 w02 w00 w ds + α2 Λ2 ds e31 Λ1 0 0 2hp 0 6hp Z l 002 Z l ε33 Vb w ds  α3 Ip ds ¼ QðtÞ + hp 0 0 2hp l

w00 ds + hp

Z

l

(5.30)

Since the beam is clamped at s ¼ 0, the essential boundary conditions at s ¼ 0 are given by wð0, tÞ ¼ 0 wð1, 0Þ ð0, tÞ ¼ 0

(5.31)

and the natural nonlinear boundary conditions at s ¼ l, which are moment and shear force balance equations, obtained from the generalized Hamilton principle, are written as 

w02 VHf VHf 1 1 1 + α1 Λ2 w002 + + α2 Λ2 HfVw002  α4 Λ3 w003 2hp hp 2 2hp 2 00   w VHf α3 Ip  YIeq w00 w02 + w00 ¼ 0, and s¼l hp02  w VHf 0 VHf 0 e31 Λ1 + YIeq ðw000 w02 + w002 w0 + w000 Þ + 2hp hp  002   00  w VHf 0 VHfw00 w000 w VHf 0 w000 VHf 00 000  α1 Λ2 w w +α3 Ip + +  α2 Λ2 2hp hp hp hp 3 + α4 Λ3 w002 w000 ¼ 0 2 s¼l

e31 Λ1

(5.32)

(5.33)

5.3.3 Reduced-order model: Galerkin discretization To simplify the representative distributed parameter governing equation and to study the response to a principal mode resonance excitation, we employ a Galerkin discretization and perform weighted residual method to determine the governing equation of the first mode. To this end, we determine the linear mode shapes of the system by considering the undamped, unforced, short-circuited ½VðtÞ ¼ 0, and linear free vibration description of motion as ρeq w€ + YIeq w0000 ¼ 0

(5.34)

Writing w(s, t) as the sum of discretized M modes having a shape ϕi(s) and temporal amplitude qi(t) wðs, tÞ ¼

M X i¼1

ϕi ðsÞqi ðtÞ

(5.35)

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

157

and substituting into Eq. (5.34), while realizing that q€ ¼ ω2n q, where ωn is the natural frequency of the nth mode, we obtain ϕ0000 i ¼ λ i ϕi 0 < s < l

(5.36)

whose general solution is of the form ϕi ðsÞ ¼ A1 cos ðλi sÞ + B1 sin ðλi sÞ + C1 cosh ðλi sÞ + D1 sinh ðλi sÞ 0 < s < l where λi ¼

(5.37)

 2 1=4 ωi ρeq . The linear short-circuited boundary conditions are given by YIeq

ϕi ð0Þ ¼ ϕ0i ð0Þ ¼ ϕ00i ðlÞ ¼ ϕ000 i ðlÞ ¼ 0

(5.38)

Moreover, for any two distinct modes, p and q, we write the orthogonality relations as [48] Z l ρeq ϕp ðsÞϕq ðsÞ ds ¼ δpq (5.39) Z l0 00 00 2 YIeq ϕp ðsÞϕq ðsÞ ds ¼ δpq ωq 0

where δpq is the Kronecker delta, defined as unity when p is equal to q and 0 otherwise. For any given problem, the natural frequency ωi and mass normalized mode shapes can be uniquely determined by solving for nontrivial solution of the linear shortcircuited boundary conditions given by Eq. (5.38) in conjunction with the orthogonality relations represented by Eq. (5.39). Using the description of w(s, t) from Eq. (5.36), and the Galerkin-weighted residual method, the equation of motion is written as 0 1 M M M M X X X X ρeq ϕi q€i  c1 ϕi q_ i  YIeq @ ϕ0000 qi qj qk ½ϕ0i ðϕ0j ϕ00k Þ0 0 A i qi + i¼1 i¼1 i¼1 i, j, k¼1 0 Z sZ θ  M  ρeq X ∂2  ϕ0i qi ϕ0j ϕ0k 2 qj qk dy dθ  2 ∂t l 0 i, j, k¼1 2 3 + e31 Λ1 4

M X

qi qj ϕ0i ϕ0j

M VHf 00 X VHf 0 VHf 00 5 + qi qj ϕ00i ϕ0j + 2hp hp hp i, j¼1

2 i, j¼1 3 M M M M 00 0 X X X X VHf 2VHf VHf VHf 00 00 00 000 000 000 0000 00 5 qi qj ϕi ϕj + qi qj ϕi ϕj + qi qj ϕi ϕj + qi qj ϕi ϕj + α2 Λ2 4 hp hp 2hp hp i, j¼1 i, j¼1 i, j¼1 i, j¼1 " # h i M M M M 00 X 0 X X X 000 + ϕ00 ϕ0000  α I 00 q VHf + 000 2VHf + 0000 VHf qi qj ϕ000 ϕ ϕ q ϕ q ϕ + α1 Λ2 i i i i 3 p i j i j i i h hp hp p i, j¼1 i¼1 i¼1 i¼1  M X 000 3 00 00 0000 €b ¼ 0 + α4 Λ3 qi qj qk 3ϕ00i ϕ000 j ϕk  2 ϕi ϕj ϕk  ρeq u i, j, k¼1 (5.40)

158

Ferroelectric Materials for Energy Harvesting and Storage

and the corresponding electrical equation is written as "

# Z Z l Z l M M l X X ε33 Vb 00 1 0 0 00 1 qi ϕi ds + qi qj qk ϕi ϕj ϕk ds + ds e31 Λ1 h 2h hp p p 0 0 0 i¼1 i, j, k¼1 Z l Z l M M X X 1 1 qi qj qk ϕ00i ϕ00j ϕ00k ds  α3 Ip qi qj ϕ00i ϕ00j ds ¼ QðtÞ + α2 Λ2 6hp 2hp 0 0 i, j¼1 i, j, k¼1 (5.41) Because we are interested in the response of the beam at primary resonance and assuming that the vibration modes of a cantilever beam are distant, we neglect the contribution of the higher modes. By considering only one mode (M ¼ 1) and using the orthogonal properties of the linear modes, the individual terms in Eq. (5.40) are simplified as shown in Appendix A. From the nonlinear boundary conditions along with orthogonality conditions, we write the equation of motion of the first mode as Z

l

q€1 + c1 q_ 1 0

Z ϕ21

ds + ω21 q1

+ 2YIeq q31

l 0

002 ϕ02 1 ϕ1

V ds  e31 Λ1 hp

Z 0

l

ϕ001 ds

Z Z l Z l 3V l 02 00 V 1 ϕ1 ϕ1 ds  α2 Λ2 q21 ϕ1 0003 ds  α1 Λ2 q21 ϕ001 3 ds 2hp 0 2hp 0 2 0 Z l Z l V 1 2 3 €1 Þ ϕ002 ϕ004 + α3 Ip q1 1 ds + α4 Λ3 q1 1 ds  ρeq q1 ðq_ 1 + q1 q hp 2 0 0  Z l Z l Z s Z θ 02 € ϕ02 ϕ dy dθ ds ¼ ρ ϕ1 ds u eq b 1 1  e31 Λ1 q21

l

0

0

(5.42)

0

and the corresponding electrical equation as 

e31

Z Z Z q1 l 00 q31 l 0 2 00 q31 l 00 3 Λ1 ϕ ds + ϕ ϕ ds + α2 Λ2 ϕ ds hp 0 1 2hp 0 1 1 6hp 0 1 Z Z l ε33 Vb q2 l 00 2 ds  α3 Ip 1 ϕ ds ¼ QðtÞ + 2hp 0 1 hp 0

(5.43)

Using the electrical damping due to dissipation by the load resistance [49], Q_ ¼  VR, we modify Eq. (5.43) to write 

e31

Z Z Z 3q21 q_ 1 l 0 2 00 q21 q_ 1 l 00 3 q_ 1 l 00 Λ1 ϕ ds + ϕ ϕ ds + α2 Λ2 ϕ ds hp 0 1 2hp 0 1 1 2hp 0 1 Z l Z ε33 V b q1 q_ 1 l 00 2 V + ds  α3 Ip ϕ1 ds + ¼ 0 hp 0 hp R 0 

(5.44)

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

159

5.3.4 Approximate solution: Method of multiple scales The earlier governing equations constitute a set of nonlinear coupled equations and the exact solution is unknown. So, we employ the method of multiple scales to determine an approximate solution. The first step in implementing the method of multiple scales is to scale the different terms in the governing equation on the basis of their contributions to the response. For structural components like metals, composites, and alloys, the effect of damping is manifested at a very slow rate in comparison to the time rate of the elastic response. This can be discerned by inspecting the perturbed response of a free cantilevered beam. A perturbed free cantilevered beam takes longer time to attain its unperturbed position than the time period of the damped oscillations, which suggests that the effect of damping is much lower than that of the elastic restoring forces. The material nonlinearity manifests itself only at higher amplitudes of the response and its effect increases as this amplitude increases. We also consider a weak forcing, which produces a response that is just high enough so as to ensure that the effects of material nonlinearities are lower than the linear elastic restoring forces. Moreover, it is reasonable to assume that the effects of quadratic nonlinear terms are higher than those of the cubic nonlinearities. Piezoelectric materials are weakly coupled electromechanical systems as they require considerable strains to generate electrical power and in-line with earlier discussion, the nonlinear coupling terms are considered to be smaller than the linear coupling terms. As mentioned earlier, the piezoelectric coupling is usually small and one can scale the q2V terms to O(E3) in Eq. (5.45) and to O(E2) in Eq. (5.46). Considering that their effects are not significant and can often be in the range of electrical and mechanical noise during experiments, we neglect them in the following analysis. Using the earlier insights, we scale the different terms appropriately by the powers of a small nondimensional bookkeeping parameter, E and rewrite Eq. (5.42) as ^ + Eδ1 q2 + E2 δ2 qV + E2 δ3 q3 q€ + ω2 q + 2E2 μ1 q_ + EθV € ¼ Eηf cos ðΩt + τe Þ + E2 δ4 qðq_ 2 + qqÞ

(5.45)

and Eq. (5.44) as V θ^v q_ + Cp V Eδ2v qq_ + E ¼ 0 R 

(5.46)

where 1 E2 μ 1 ¼ c 1 2 Cp

Z 0

l

ϕ21 ds ≡ ζω, Eθ^ ≡ θ^v ¼ e31 Λ1

ε33 bl 1 ¼ , Eδ1 ¼  α1 Λ2 hp 2

Z

l 0

ϕ1003

1 hp

Z

l 0

ϕ001 ds

1 ds, E δ2 ≡ Eδ2v ¼ α3 Ip hp

(5.47a) Z

2

l

ϕ1002 ds

0

(5.47b)

160

Ferroelectric Materials for Energy Harvesting and Storage

Z

l

E2 δ3 ¼ 2YIeq 0

Z E2 δ4 ¼ ρeq

l

0

Z Eη ¼ ρeq Ω

ϕ02 1 l

2

1 002 ϕ02 1 ϕ1 ds + α4 Λ3 2 Z l Z s

0

θ

Z 0

l

ϕ004 1 ds

 ϕ02 dy dθ ds 1

ϕ1 ds, f ¼ ub , and Ω ¼ ω + Eσ

(5.47c)

(5.47d)

(5.47e)

0

The subscripts representing mode numbers were removed from the equations for the sake of convenience. To solve Eqs. (5.45), and (5.46), we introduce three independent time scales T0, T1, and T2 defined by Tn ¼ En t, where n ¼ 0,1, 2 Using chain rule, we expand the derivatives up to O(E2) and write D ∂ ∂ ∂ ¼ +E + E2 ≡ D0 + ED1 + E2 D2 Dt ∂T0 ∂T1 ∂T2

(5.48a)

and     D2 ∂2 ∂2 ∂2 ∂2 2 ≡ D20 + 2ED0 D1 + E2 D21 + 2D0 D2 ¼ + 2E +E 2 + Dt2 ∂T02 ∂T1 ∂T0 ∂T0 ∂T2 ∂T12 (5.48b) The approximate solutions of q(t) and V (t) that are accurate up to E2 order are expressed as qðt, EÞ ¼ q0 ðT0 , T1 , T2 Þ + Eq1 ðT0 ,T1 ,T2 Þ + E2 q2 ðT0 ,T1 , T2 Þ

(5.49a)

Vðt, EÞ ¼ V0 ðT0 , T1 , T2 Þ + EV1 ðT0 , T1 , T2 Þ + E2 V2 ðT0 , T1 ,T2 Þ

(5.49b)

Substituting Eqs. (5.48a), (5.48b), and (5.49a) into Eqs. (5.45), and (5.46), followed by retaining terms up to O(E2), we obtain E0-order equation: D20 q0 + q0 ω2 ¼ 0

(5.50a)

Cp D0 V0  θ^v D0 q0 ¼ 0

(5.50b)

E1-order equation: ^ 0 + 1 f ηejτe + jT0 Ω + 1 f ηejτe jT0 Ω D20 q1 + ω2 q1 ¼ 2ðD0 D1 q0 Þ  δ1 q20  θV 2 2

(5.51a)

RCp D0 V1 +RCp D1 V0  δ2v Rq0 D0 q0  Rθ^v D0 q1  Rθ^v D1 q0 + V0 ¼ 0

(5.51b)

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

161

E2-order equation: D20 q2 + ω2 q2 ¼ δ4 q20 D20 q0  δ4 q0 ðD0 q0 Þ2  2μ1 D0 q0  2D0 D1 q1  D21 q0 ^ 1 and  2D0 D2 q0  δ3 q30  2δ1 q1 q0  δ2 V0 q0  θV RCp D0 V2 + RCp D1 V1 + RCp D2 V0  δ2v Rq0 D0 q1  δ2v Rq0 D1 q0  δ2v Rq1 D0 q0 Rθ^v D0 q2  Rθ^v D1 q1  Rθ^v D2 q0 + V1 ¼ 0

(5.52a)

(5.52b)

From Eqs. (5.50a), (5.50b), we write the E0-order solution as q0 ðT0 ,T1 , T2 Þ ¼ AðT1 , T2 ÞejωT0 + c:c and

(5.53a)

θ^v AðT1 , T2 ÞejT0 ω + c:c Cp

(5.53b)

V0 ðT0 ,T1 , T2 Þ ¼

Substituting q0 and V0 in E1-order equation and eliminating the terms that cause the solution to grow indefinitely, also referred as secular terms, we obtain D1 A ¼

 j  2A1 θ^θ^v  f ηCp ejτe + jσT1 4ωCp

(5.54)

Solving for q1 and V1 yields q1 ¼  V1 ¼

Aδ1 A A2 δ1 e2jT0 ω + + c:c and 3ω2 ω2

  1

3δ2v Rω2 Cp A21 e2jT0 ω + 2θ^v A21 δ1 RCp e2jT0 ω + 3jA1 ωejT0 ω + c:c 6Rω2 C2p

(5.55a)

(5.55b)

Next, substituting q1 and V1 with the right-hand side of Eqs. (5.55a), (5.55b), in the E2-order equation, and eliminating the secular terms, we obtain the relation for D2A. Using the relations for D1A and D2A, we obtain the amplitude and phase modulation equations as A_ ¼ ED1 A + E2 D2 A

(5.56)

Writing the complex amplitude, A, in polar coordinates, AðT1 ,T2 Þ ¼ 12 aðT0 ,T2 ÞejβðT1 , T2 Þ and defining γ ¼ Eσt + τe  β to remove the explicit dependence on time in the amplitude and phase modulation equations, we obtain ! " # θ^θ^v f ηðΩ  3ωÞ f ηθ^θ^v 2 1 a_ ¼ E sinγ + E a  2 2  2μ1  3 sin γ and Rω Cp 8ω Cp 4ω2 2 # θ^θ^v f ηðΩ  3ωÞ + cosγ 2ωCp 4aω2 "  #  2 a2 6δ4 ω4  9δ3 ω2 + 10δ21 f ηθ^θ^v θ^2 θ^v 2 + 3 2 cos γ +E 8ω Cp 8aω3 Cp 24ω3

(5.57)

"

γ_ ¼ E σ +

(5.58)

162

Ferroelectric Materials for Energy Harvesting and Storage

To determine the steady-state amplitude a0, for a given excitation, we set a_ ¼ 0 and γ_ ¼ 0 and eliminate γ in Eqs. (5.57), (5.58), which yields the following steady-state amplitude response relation  2 16a20 ω2 E2 2μ1 Rω2 C2p + θ^θ^v 1  2 +   2 2 2 2 2 2 2 ^ ^ ^ θ^v 2 f η R Cp 2ωCp ð3ω  ΩÞ  θEθ v 9f η Cp 2ωCp ð3ω  ΩÞ  θE  

  2 ^ 2 Cp θ^v ; 3θ^2 Eθ^2  C2 a2 E 6δ4 ω4  9δ3 ω2 + 10δ2 + 24σω3 a20 12θω ¼1 p v 1

(5.59)

The approximate solutions up to E1 order are written as  a2 δ1 a2 δ1 wðs,tÞ ¼ ϕ1 ðsÞ a0 cos ðΩt + τe  γ 0 Þ + E 0 2 cos ð2½Ωt + τe  γ 0 Þ  E 0 2 6ω 2ω

(5.60a)

a0 θ^v a0 θ^v cos ðΩt + τe  γ 0 Þ  E sin ðΩt + τe  γ 0 Þ Cp RωC2p a2 δ1 θ^v a2 δ2v + E 0 2 cos ð2½Ωt + τe  γ 0 Þ + E 0 cos ð2½Ωt + τe  γ 0 Þ 6ω Cp 4Cp

(5.60b)

VðtÞ ¼

these solutions can alternatively be written in the frequency domain as  a2 δ1 a2 δ1 wðs,νÞ ¼ ϕ1 ðsÞ a0 δðν  ΩÞ  E 0 2 δðν  0Þ + E 0 2 δðν  2ΩÞ 2ω 6ω a0 θ^v VðνÞ ¼ Cp

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ! E2 a20 δ1 θ^v δ2v δðν  2ΩÞ + 1 + 2 2 2 δðν  ΩÞ + E R ω Cp 2Cp 3ω2 2

(5.61a)

(5.61b)

where δ is Dirac delta function.

5.3.5 Parameter identification strategy Nayfeh [38, 39] detailed how specific responses can be used to perform parametric identification. It is well established that by examining the free response of a system one can determine the linear structural damping and damped natural frequency of the mechanical system. Similarly, one could design experiments that exploit different nonlinear resonances to provide accurate estimate of nonlinear parameters. For instance, examining Eqs. (5.60a), (5.61a), we note that for a unimorph under direct excitation, the quadratic nonlinear stiffness coefficient, δ1, which is related to the nonlinear material property α1, results in a nonzero mean transverse displacement. As such, the value of a0 from the component of the transverse displacement at forcing frequency can be used to uniquely determine the nonlinear material parameter α1. Further, we note that the amplitude of the frequency component in the response at twice the forced frequency is determined by δ1 and Eδ2v ð¼ E2 δ2 Þ. Having determined δ1 in the previous step, we can determine δ2, which is directly related to α3. We should note here if the phase relation j2χ V(Ω)  χ V(2Ω)j 0 or 2π, then the component of

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

163

voltage at twice the forced frequency is of the same sign as the component of voltage at the forced frequency or vice versa. To estimate α4, we determine δ3 from the steadystate amplitude response relation using the identified values of δ1 and δ2. It is to be noted that the steady-state amplitude response relation is essentially a quadratic equation in δ3 and hence will give two roots for the same. One of the root can be eliminated by performing the identification procedure for a different detuning value. Based on Eq. (5.47c), the cubic stiffness coefficient, δ3, is the sum of geometric nonlinearity and the material nonlinearity which depends on α4. Accounting for the contribution of geometric nonlinearity by evaluating the integral, α4 can be identified. The step-by-step parameter identification strategy to identify the material nonlinear parameters α1, α3, and α4 discussed earlier is summarized in the flowchart in Fig. 5.2.

5.3.6 Validation of parameter identification strategy In this section, we implement the proposed parameter identification strategy for a unimorph with geometric and linear material properties as presented in Table 5.1 by analyzing a numerically simulated response for assumed values of the nonlinear piezoelectric material parameters presented in Table 5.2. For the material and geometric properties listed in Tables 5.1 and 5.2, the linear mass normalized mode shape is determined as ϕ1  16:61½ cos ð31:25sÞ  cosh ð31:25sÞ  12:19½ sin ð31:25sÞ  sinh ð31:25sÞ (5.62) and the coefficients in the governing equations are presented in Table 5.3. The numerically simulated transverse displacement at 5 cm from the clamp and the voltage generated for an excitation amplitude of 4 m s2 at a detuning frequency Eσ ¼ 15 rad s1 is presented in Fig. 5.3A–D. Particularly, Fig. 5.3A and C shows the time

w s,t

w sv V v

f

a02d1 2w 2

ϵδ1 =

=

w s,t V t

Steady-s

1 αΛ 2 1 2

a20

l 0

δ1θˆν

+

2Cp 3ω2

ϵδ2ν = α3Ip

1 hp

ds

δ2ν

l 0

3

φ1

φ1

2

2

ds

Fig. 5.2 Flowchart detailing the parameter identification strategy for a unimorph subjected to direct excitation and hw(s, t)i denotes the mean of w(s, t).

164

Ferroelectric Materials for Energy Harvesting and Storage

Table 5.1 The physical and linear material properties of the unimorph subjected to direct excitation presented in Fig. 5.1. Parameters

Value

Length of the substrate, l (mm) Length of the piezoelectric layer, l (mm) Width of the structure, b (mm) Thickness of the substrate, hs (mm) Thickness of the piezoelectric layer, hp (mm) Density of the substrate (kg m3) Density of the piezo (kg m3) s Young’s modulus of the substrate, Y11 (G Pa) p Young’s modulus of the piezo, Y11 (G Pa) Piezoelectric coupling coefficient, e31 (C m2) Piezoelectric permittivity, ε33 (F m1) Damping ratio, ζ Load resistance, R (M Ohm)

60 60 10 0.5 0.26 8027 7800 210 66 12.54 1.328  108 0.004 10

Table 5.2 Assumed nonlinear material parameters. Parameters

Assumed value

α1 (T Pa) α3 (k C m2) α4 (P Pa)

100 5 50

Table 5.3 Coefficients in the governing equations corresponding to the parameters presented in Tables 5.1 and 5.2. Parameters (units)

Value

ω (rad) E2μ1 (s1) Eθ^ ≡ θ^v (C kg1/2 m1) Eδ1 (kg1/2 m1 s2) E2δ2 ≡ Eδ2v (C kg1 m2) E2δ3 (kg1 m2 s2) E2δ4 (kg1 m2) Eη (kg1/2 s2) Cp (nF)

834.82 3.33 31.2  103 19.86  108 88.67 9.37  1012 3.52  105 0.047Ω2 30.65

Time (s)

(C)

(B)

Frequency (Hz)

Voltage (V)

Voltage (V)

(A)

165

w (mm)

w (mm)

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

Time (s)

(D)

Frequency (Hz)

Fig. 5.3 Principal mode resonant response represented by the displacement at 5 cm away from the fixed end and the corresponding harvested voltage for a detuning value Eσ ¼ 15 rad s1 and excitation amplitude of ubΩ2 ¼ 4 m s2 in time domain (A and C) and their amplitudes in frequency domain (B and D).

series of the transverse displacement of the structure and the voltage generated. Fig. 5.3B and D shows the corresponding amplitudes in frequency domain. From Fig. 5.3B, the component of transverse displacement at forcing frequency, w(s, Ω)js¼5 cm ¼ 0.72 mm. Noting that w(s, Ω) ¼ ϕ1(s)a0, we determine that a0* ¼ 2.8  105 kg1/2 m. Also, the mean value of the times series is evaluated as hw(s, t)js¼5 cmi ¼ 0.0308 mm. Based on the approximate solution, which shows hwðs,tÞi ¼ ϕ1 ðsÞ

a∗2 0 Eδ1 2ω2

(5.63)

we identify Eδ1* ¼ 2.11  109 kg1/2 m1 s2. Using the value of Eδ1*, a0*, and Eq. (5.47b), the nonlinear material parameter, α1, is defined as α∗1 ¼ 106:32 TPa. By examining the phase of voltage at forcing and twice the forcing frequency, it is determined that j2χ V(Ω)  χ V(2Ω)j 0 and by noting that θv is negative, we conclude that the component of voltage at twice the forcing frequency is also negative. As such, the

166

Ferroelectric Materials for Energy Harvesting and Storage

Table 5.4 Summary of the results and efficiency of the parameter identification scheme proposed. Parameters

Value

Parameters

Value

% Error

α1 (T Pa) α3 (k C m2) α4 (P Pa)

100 5 50

α1* (T Pa) α3* (k C m2) α4* (P Pa)

106.32 5.13 47.4

6.3 2.5 5.18

amplitude of the first harmonic is written as V (2Ω) ¼ 0.97 V, and from the approximate solution of voltage we write a2 δ∗1 θ^v δ2v + Vð2ΩÞ ¼ E 0 2Cp 3ω2 2

! (5.64)

and determine that Eδ∗2v ¼ 90:9 C kg1 m2. Using Eq. (5.47b), α3 is identified as  5126.2 C m2. Using the steady-state amplitude response relation of Eq. (5.59) and the identified values of a∗0 , Eδ1*, Eδ2v, the value of δ3 is numerically evaluated as δ3* ¼ 8.87  1012 kg1 m2 s2. Using Eq. (5.47c) and by accounting for the geometric nonlinearity, the value of α4 is identified as α4 ¼ 47.41 P Pa. The comparison of the assumed to identified nonlinear material parameters is presented in Table 5.4. Based on the percentage errors in Table 5.4 which are less than 7%, we conclude that the proposed parameter identification approach can yield the nonlinear material parameters with a reasonable accuracy.

5.4

Parametric excitation

Next, we identify the nonlinear piezoelectric coefficients using a unimorph subjected to parametric excitation. Parametric excitation is essentially a nonlinear excitation in which the direction of the base displacement of the cantilever beam is in the axial direction and the response is transverse bending. We use the same dimensions of the piezoelectric layer used in Section 5.3. The parametric excitation requires a relatively high acceleration to exhibit nontrivial response. For this reason, we choose a unimorph with a substrate length larger than the piezoelectric layer as shown in Fig. 5.4. The beam, Piezoelectric sheet

l1

u(s,t) w(s,t)

hs+hp y0 Cross section

hs/2

Fig. 5.4 Schematic representation of the beam and tip mass system with a piezoelectric patch considered.

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

167

which has a length l, width b, mass per unit length ρs, with a thickness hs, is bonded with a piezoelectric layer of length l1(< l), width b, thickness hp, and mass per unit length ρp. The piezoelectric layer is connected to a load resistance, R and is harmonically excited with an amplitude of ub and frequency of Ω.

5.4.1 Mathematical modeling We use the generalized Hamilton’s principle for deriving the distributed parameter governing equation. Similar to the case of direct excitation, we write the kinetic energy of the system as 1 T¼ 2

Z

l1 0

1 + 2

Z ρeq

Z

l l1

s

0

Z ρs

0

1 ∂  02  w dy  u_b 2 ∂t s

2

1 ∂  02  w dy  u_b 2 ∂t

1 ds + 2

2

Z

1 ds + 2

l1 0

Z

l l1

ρeq ½w_ 2 ds (5.65) ρs ½w_ 2 ds

where ρeq ¼ ρs + ρp. Using the assumed electromechanical enthalpy density, the electromechanical enthalpy of the beam and piezoelectric layer considered is written as Δ¼

  Z 1 VH 1 l1 VH e31 Λ1 w00 1 + w02 ds + α3 Ip w002 ds hp hp 2 2 0 0 0 Z Z Z 1 l1 S VH2 b 1 l1 1 l VH E33 ds  α1 Λ2 w003 ds  α2 Λ2 w003 ds  hp 2 0 hp 6 0 6 0 Z l1 Z l

 1 1 s α4 Λ3 w004 ds + Y11 Is w002 1 + w02 ds + 8 0 2 l1

1 2

Z

l1

 YIeq w002 1 + w02 ds 

Z

l1

(5.66) where the definitions of Λ1, Λ2, and Λ3 remain identical to that defined for the case of direct excitation. The only difference is that the neutral axis in this case is defined by 8 p < Y11 hp ðhs + hp Þ  s  0  s  l1 p y0 ¼ 2 Y11 hs + Y11 hp : 0 l1 < s  l

(5.67)

The virtual work done by the nonconservative forces is represented by Z

l

δWnc ¼ 

c1 wð0,1Þ ðs, tÞδwðs,tÞ ds  QδVðtÞ

(5.68)

0

Substituting the kinetic energy, electromechanical enthalpy, and the work done by nonconservative forces in Eqs. (5.65), (5.66), and (5.68) into the generalized Hamilton’s principle, we obtain the piecewise distributed governing equation as

168

Ferroelectric Materials for Energy Harvesting and Storage

ρeq w€  c1 w_  YIeq ðw0000 + ½w0 ðw0 w00 Þ0 0 Þ  

 Z Z 0 ρeq 0 s θ ∂2  02  w dy dθ w 2 2 l1 0 ∂t

w02 VHf 00 w00 w0 VHf 0 VHf 00 + e31 Λ1 + + 2hp hp hp  00 002 00 000 0 VHf w 2w w VHf w0002 VHf w0000 w00 VHf + α1 Λ2 ½w0002 + w00 w0000  + + + + α2 Λ2 2hp hp hp hp   VHf 00 w00 2Vhf 0 w000 VHfw0000 3 002 0000 00 0002 + α4 Λ3 3w w  w w + +  α3 Ip hp hp hp 2 Z Z

 ρs 00 l θ ∂2  02  w dy dθ + u€b ρeq w0 + ρeq ðs  l1 Þw00 + ρs ðl1  lÞw00 ¼ 0 + w 2 2 l1 0 ∂t (5.69) for 0 < s < l1 s ρs w€  c1 w_  Y11 Is ðw0000 + ½w0 ðw0 w00 Þ0 0 Þ 

+ u€b ðρs w0 + ρs ðs  lÞw00 Þ ¼ 0

 Z Z 0 ρs 0 s θ ∂2  02  w dy dθ w 2 2 l 0 ∂t (5.70)

for l1 < s < l and the corresponding electrical equation as Z

e31

Z l1 003 w02 w00 w Λ1 ds + α2 Λ2 ds 2h p 0 0 0 6hp Z l1 002 Z l1 ε33 Vb w ds  α3 Ip ds ¼ QðtÞ + hp 0 0 2hp l1

w00 ds + hp

Z

l1

(5.71)

Since the beam is clamped at s ¼ 0, the essential boundary conditions at s ¼ 0 are given by wð0, tÞ ¼ 0 wð1, 0Þ ð0, tÞ ¼ 0

(5.72)

and the natural nonlinear boundary conditions at s ¼ l, which are moment and shear force balance equations, obtained from the generalized Hamilton principle are written as s s Is w00 + Y11 Is w02 w00 s¼l ¼ 0 Y11

(5.73)

s s s Is w000 + Y11 Is w0 w002 + Y11 Is w02 w000 s¼l ¼ 0 Y11

(5.74)

the generalized Hamilton’s principle also provides the shear force and bending moment equilibrium at s ¼ l1 as compatibility conditions which are written as

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

169



w02 VHf VHf 1 1 1 + α1 Λ2 w002 + + α2 Λ2 HfVw002  α4 Λ3 w003 2hp hp 2 2hp 2 (5.75)  00 02    w00 VHf s α3 Ip  YIeq w w + w00 + Y11 Is w00 w02 + w00 ¼0 hp s¼l1  02  0 0 w VHf VHf s + YIeq ðw000 w02 + w002 w0 + w000 Þ  Y11 + Is ðw000 w02 + w002 w0 + w000 Þ e31 Λ1 2hp hp  002   00  w VHf 0 VHfw00 w000 w VHf 0 w000 VHf α2 Λ2  α1 Λ2 w00 w000 +α3 Ip + + 2hp hp hp hp 3 + α4 Λ3 w002 w000 ¼ 0 2 s¼l (5.76) e31 Λ1

5.4.2 Reduced-order model: Galerkin discretization The direct procedure for finding the response of the beam with piezoelectric layer involves solving the distributed governing equation, boundary conditions, and compatibility conditions, which constitute complicated integro-partial differential equations. In order to study the principal parametric resonance, we employ the Galerkin weighted residual method and find the governing equation of first mode. To this end, we write the linear, undamped, unforced, and short-circuited Eqs. (5.69)– (5.70) as 8 < ρeq w€ + YIeq w0000 ¼ 0 s ρ w€ + Y11 Is w0000 ¼ 0 : s

0 < s < l1 l1 < s < l

(5.77)

Assuming a discretization

wðs,tÞ ¼

8 M X > > > ϕi ðsÞqi ðtÞ >
> > > :

l1 < s < l

i¼1 M X

ψ i ðsÞqi ðtÞ

(5.78)

i¼1

and substituting into Eq. (5.77), while realizing that q€ ¼ ω2n q, where ωn is the natural frequency of the nth mode, we obtain ϕi 0000 ¼ λ1i ϕi 0 < s < l1

(5.79a)

ψ i 0000 ¼ λ42i ψ i l1 < s < l

(5.79b)

170

Ferroelectric Materials for Energy Harvesting and Storage

whose general solution is of the form ϕi ðsÞ ¼ A1 cos ðλ1i sÞ + B1 sin ðλ1i sÞ + C1 cosh ðλ1i sÞ + D1 sinh ðλ1i sÞ 0 < s < l1 (5.80a) ψ i ðsÞ ¼ A2 cos ðλ2i sÞ + B2 sin ðλ2i sÞ + C2 cosh ðλ2i sÞ + D2 sinh ðλ2i sÞ l1 < s < l (5.80b)  2 1=4  2 1=4 ωi ρeq ωi ρ s where λ1i ¼ , λ2i ¼ and the linear short-circuited boundary s YIeq Y11 Is conditions are given by ϕð0Þ ¼ ϕ0 ð0Þ ¼ 0, and ψ 00 ðlÞ ¼ ψ 000 ðlÞ ¼ 0

(5.81)

The continuity relations and linear compatibility conditions are written as ϕðl1 Þ ¼ ψðl1 Þ, ϕ0 ðl1 Þ ¼ ψ 0 ðl1 Þ

(5.82a)

s Is ψ 00 ðl1 Þ and YIeq ϕ00 ðl1 Þ ¼ Y11

(5.82b)

s YIeq ϕ000 ðl1 Þ ¼ Y11 Is ψ 000 ðl1 Þ

(5.82c)

Moreover, for any two distinct modes, p and q, we write Z ω2p

l1

s¼l1 s¼l1 Z 00 0 ρeq ϕp ϕq ds ¼ YIeq ϕ000 ϕ  YI ϕ ϕ + eq p q p q s¼0

0

s¼0

l1

0

YIeq ϕ00p ϕ00q ds (5.83a)

Z ω2p

l

l1

s¼l

s ρs ψ p ψ q ds ¼ Y11 Is ψ 000 p ψ q

s¼l1

s¼l

s  Y11 Is ψ 00p ψ 0q

s¼l1

Z

l

+ l1

s Y11 Is ψ 00p ψ 00q ds

(5.83b) Adding Eqs. (5.83a), (5.83b) and using the compatibility conditions, the orthogonality relations are written as Z

l1 0

Z

l1 0

Z ρeq ϕp ðsÞϕq ðsÞ ds + YIeq ϕ00p ðsÞϕ00q ðsÞ ds +

l

l1

ρs ψ p ðsÞψ q ðsÞ ds ¼ δpq

Z l1

l

s Y11 Is ψ 00p ðsÞψ 00q ðsÞ ds ¼ δpq ω2q

(5.84a)

(5.84b)

where δpq is the Kronecker delta, defined as unity when p is equal to q and 0 otherwise. Using the earlier discretization, the solution is approximated as the sum of a finite number of modes, that is,

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

wðs,tÞ ¼

8 M X > > > ϕi ðsÞqi ðtÞ > < i¼1 M >X

> > > :

171

0 < s < l1 (5.85)

ψ i ðsÞqi ðtÞ

l1 < s < l

i¼1

where the basis functions ϕi(s) and ψ i(s) represent the mode shape, qi(t) represents the modal coordinate (temporal amplitude), and M is the number of modes under consideration. For any given problem, the natural frequency ωi and mass normalized mode shapes can be uniquely determined by solving eight equations mentioned in Eqs. (5.81)–(5.82c) in conjunction with the orthogonality relations given by Eq. (5.84). Using the piecewise description of w(s, t) from Eq. (5.85), and the Galerkinweighted residual method, the equations of motion are rewritten as 0 1 M M M M X X X X ρeq ϕi q€i  c1 ϕi q_ i  YIeq @ ϕ0000 qi qj qk ½ϕ0i ðϕ0j ϕ00k Þ0 0 A i qi + i¼1 i¼1 i¼1 i, j, k¼1 " #0 Z sZ θ M  2 ρeq X 0 0 0 ∂ ϕ qi ϕϕ qj qk dy dθ  2 i, j, k¼1 i l1 0 j k ∂t2

M Z l Z θ X ρs 00 ∂2  ϕi qi 2 qj qk ψ 0j ψ 0k dy dθ 2 ∂t l1 0 i, j, k¼1 2 3 M M 00 0 VHf 00 X X VHf VHf 5 qi qj ϕ0i ϕ0j + qi qj ϕ00i ϕ0j + +e31 Λ1 4 2hp i, j¼1 hp hp i, j¼1 2 3 M M M M 00 0 X X X X VHf 2VHf VHf VHf 00 00 00 000 000 000 0000 00 5 +α2 Λ2 4 qi qj ϕi ϕj + qi qj ϕi ϕj + qi qj ϕi ϕj + qi qj ϕi ϕj hp i, j¼1 hp 2hp i, j¼1 hp i, j¼1 i, j¼1  M h i M X X 000 00 0000 00 000 000 3 00 00 0000 +α1 Λ2 ϕ qi qj ϕ000 ϕ + ϕ ϕ Λ q q q 3ϕ ϕ ϕ  ϕ ϕ + α i j k 4 3 i j i j i j k 2 i j k i, j¼1 i, j, k¼1 " # M M M X VHf 00 X 2VHf 0 X VHf ϕ00i qi + qi ϕ000 + qi ϕ0000  α 3 Ip i i h h h p p p i¼1 i¼1 i¼1

+

+u…b

M h i X qi ρe qϕ0i + ρeq ðs  l1 Þϕ001 + ρs ðl1  lÞϕ00i ¼ 0 i¼1

(5.86) 1 M M X X 0 0A 0000 0 0 00 ψ i qi + qi qj qk ½ψ i ðψ j ψ k Þ  ρs ψ i q€i  c1 i¼1 i¼1 i¼1 i, j, k¼1 0 M  Z s Z θ M

X  ρ X ∂2  ψ 0i qi 2 qj qk ψ 0j ψ 0k dy dθ + u€b qi ρs ψ 0i + ρs ðs  lÞψ 00i ¼ 0  s 2 i, j, k¼1 ∂t l 0 i¼1 M X

M X

0

s ψ i q_ i  Y11 Is @

(5.87)

172

Ferroelectric Materials for Energy Harvesting and Storage

and the corresponding electrical circuit equation is rewritten as Z

" # Z l1 Z l1 M M X X ε33 Vb 00 1 0 0 00 1 ds + e31 Λ1 qi ϕi ds + qi qj qk ϕi ϕj ϕk ds hp hp 2hp 0 0 0 i¼1 i, j, k¼1 Z l1 Z l1 M M X X 1 00 00 00 1 + α2 Λ2 qi qj q k ϕ i ϕj ϕk ds  α3 Ip qi qj ϕ00i ϕ00j ds ¼ QðtÞ 6hp 2hp 0 0 i, j¼1 i, j, k¼1 (5.88) l1

The individual terms in the weighted residual statement of Eqs. (5.86), (5.87) are simplified as shown in Appendix B. By considering only one mode (M ¼ 1) and using the continuity, compatibility, and nonlinear boundary conditions along with normalization conditions, we write the equation of motion of the first mode as Z 0

 + 2q31

Z

l1

q€1 + c1 q_ 1

ϕ21

Z

l1

YIeq 0

l

ds + l1

002 ϕ02 1 ϕ1

 ψ 21

ds + ω21 q1 Z

s ds + Y11 Is

l

l1

002 ψ 02 1 ψ1

 ds

Z Z Z l1 3V l1 02 00 V l1 00 2 V ϕ ϕ ds  e31 Λ1 ϕ ds  α2 Λ2 q1 ϕ 0003 ds 2hp 0 1 1 hp 0 1 2hp 0 1 Z l1 Z l1 Z l1 1 V 1 002 2 00 3 3 ϕ1 ds + α3 Ip q1 ϕ1 ds + α4 Λ3 q1 ϕ004  α 1 Λ 2 q1 1 ds 2 hp 2 0 0 0 Z l1 Z l Z θ  2 02 02 + ρs ðq_ 1 + q1 q€1 Þ ϕ1 ds ψ 1 dy dθ  e31 Λ1 q21

l1

0

Z

 ρeq q1 ðq_ 21

l1

+ q1 q€1 Þ 0

Z  ρs q1 ðq_ 21 + q1 q€1 Þ  Z ¼ u€b q1 ρeq

l1

0

l l1

ϕ02 1

ψ 02 1

0

Z s Z l1

0

Z s Z l

ðl1  sÞϕ01 2 ds + ρs

θ

θ

0

Z 0

ϕ02 1

(5.89)

 dy dθ ds

 ψ 02 dy dθ ds 1 l1

ðl  l1 Þϕ01 2 ds + ρs

Z l1

l

ðl  sÞψ 01 2 ds



The corresponding electrical equation after differentiating with respect to time on both sides of Eq. (5.88) and accounting for electrical damping is written as 

e31

Z Z 3q21 q_ 1 l1 0 2 00 q21 q_ 1 l1 00 3 ds + ϕ ϕ ds + α2 Λ2 ϕ ds 2hp 0 1 1 2hp 0 1 0 Z ε33 V bl1 q1 q_ 1 l1 00 2 V  α 3 Ip ϕ ds + ¼ 0 + hp hp 0 1 R

q_ Λ1 1 hp

Z 

l1

ϕ001

(5.90)

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

173

5.4.3 Approximate solution: Method of multiple scales Similar to the previous example with direct excitation, we employ the method of multiple scales to determine the approximate solution of Eqs. (5.89), (5.90) and rewrite them by scaling the nonlinearities appropriately as ^ + Eδ1 q2 + E2 δ2 qV + E2 δ3 q3 + E2 δ4 qðq_ 2 + qqÞ € ¼ Eqηub cos ðΩt + τe Þ q€ + ω2 q + 2E2 μ1 q_ + EθV (5.91) V  θ^v q_ + Cp V Eδ2v qq_ + E ¼ 0 R 

(5.92)

where 1 E μ1 ¼ c1 2

Z

l1

2

0

Z ϕ21

l

ds + l1



ψ 21

1 ds ≡ ζω; Eθ^ ≡ θ^v ¼ e31 Λ1 hp

Z

l1

0

ϕ001 ds (5.93a)

ε33 bl1 1 ; Eδ1 ¼  α1 Λ2 Cp ¼ 2 hp

Z

l1

ϕ1

003

0

1 ds; E δ2 ≡ Eδ2v ¼ α3 Ip hp

Z

2

l1

ϕ1 002 ds

0

(5.93b) Z

l

E2 δ3 ¼ 2YIeq

002 s ϕ02 1 ϕ1 ds + 2Y11 Is

0

Z E2 δ4 ¼ ρs ðq_ 21 + q1 q€1 Þ Z

Z

l1

l1

ϕ1 0 ds 2

Z l1

l

1 002 ψ 02 1 ψ 1 ds + α4 Λ3 2

Z l Z l1

0 l1 Z θ



θ 0

ψ 02 1 dy dθ

Z

l1

0

ϕ004 1 ds



ϕ02 ϕ02 1 1 dy dθ ds 0 0 s  Z l1 Z l Z θ 02 02 + ρs ψ1 ψ 1 dy dθ ds + ρeq

s

l1

 Eη ¼ Ω ρeq

Z

2

0

l1

(5.93c)

(5.93d)

0

ðl1  sÞϕ01 2

Z ds + ρs

0

l1

ðl  l1 Þϕ01 2

Z ds + ρs

l1

l

ðl  sÞψ 01 2

ds (5.93e)

f ¼ ub , and Ω ¼ 2ω + Eσ

(5.93f)

Following a similar approach to that in Section 5.3.4, we obtain the steady-state amplitude response relation as

174

Ferroelectric Materials for Energy Harvesting and Storage

  ^ 2 Cp θ^v + 3θ^2 Eθ^2 a2 EC2p 6δ4 ω4  9δ3 ω2 + 10δ21  12θω v 12ω3 C2p f 2 η2 ðΩ  4ωÞ2  2   2 ^ θ^v 8μ1 Rω2 EC2p + 4θE f 2 η2 E 4ω2 + 2ωΩ + Ω2 + 2σ +   2 ¼ 1 16ω3 Ωð2ω + ΩÞ 4f ηRωC2p  f ηRΩC2p 16ω4

(5.94) and the approximate solutions that are accurate up to E1 order as     1 1 a0 δ1 wðs, tÞ ¼ ϕ1 ðsÞ a0 cos ðγ  τe  tΩÞ + a0 E ½ cos ðγ  τe  ΩtÞ  3 2 6 ω2   3f η 1 cos ½γ  3τe  3tΩ for 0  s  l1  Ωð2ω + ΩÞ 2 (5.95a)     1 1 a0 δ1 ½ cos ðγ  τe  ΩtÞ  3 wðs, tÞ ¼ ψ 1 ðsÞ a0 cos ðγ  τe  tΩÞ + a0 E 2 6 ω2   3f η 1 cos ½γ  3τe  3tΩ for l1  s  l  Ωð2ω + ΩÞ 2 (5.95b) for the displacement, and VðtÞ ¼

   a0 θ^v 1 a 0 E  cos ðγ  τe  tΩÞ + 3aδ2v ω2 Cp + 2aδ1 Cp θ^v cos ðγ  τe  tΩÞ 2 2 2 12ω Cp Cp    # 2 6f ηω Cp θ^v 1 12ωθ^v 1 cos ðγ  3τe  3tΩÞ + sin ðγ  τe  tΩÞ  2 2 Ωð2ω + ΩÞ R (5.95c)

for the voltage. Alternatively, the approximate solutions can be written in the frequency domain as   2 a20 δ1 Ω a δ1 + 0 2 δðν  ΩÞ wðs, νÞ ¼ ϕ1 ðsÞ E 2 δðν  0Þ + a0 δ ν  2ω 6ω 2   f ηa0 3 δ ν  Ω for 0  s  l1 E 2Ωð2ω + ΩÞ 2

(5.96a)

   a20 δ1 Ω a2 δ 1 wðs, νÞ ¼ ψ 1 ðsÞ E 2 δðν  0Þ + a0 δ ν  + 0 2 δðν  ΩÞ 2ω 6ω 2   f ηa0 3 δ ν  Ω for l1  s  l E 2Ωð2ω + ΩÞ 2

(5.96b)



Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

175

and sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  2    E Ω a20 E  3δ2v ω2 + 2δ1 θ^v δðν  ΩÞ + 1+ δ ν 2 RωCp 2 12ω Cp   ^ Ea0 f ηθ 3 δ ν Ω + 2Cp Ωð2ω + ΩÞ 2 (5.96c)

a0 θ^v VðνÞ ¼ Cp

5.4.4 Parameter identification strategy We follow the same approach that was implemented earlier for the case of direct excitation to determine the nonlinear material parameters α1, 2, 3 from the approximate solutions of the parametric mechanical and electrical responses. The major difference between the approximate solution in the case of direct and parametric excitations is that the forcing frequency in the former is close to natural frequency of the first mode, whereas it is close to twice the natural frequency of the first mode in the latter. It is interesting to note that the mean value of the transverse displacement is identical to that in the case of direct excitation. Also, the components of approximate solution of transverse displacement and the voltage generated at the half of the forcing frequencies in the case of parametric excitation are identical to those at the forcing frequency and its first harmonic under direct excitation. So we implement the same parameter identification approach that we proposed in the case of direct excitation in Section 5.3.5 by appropriately accounting for frequency. The step-by-step parameter identification strategy to identify the material nonlinear parameters α1, α3, and α4 discussed earlier is summarized in the flowchart in Fig. 5.5.

Fig. 5.5 Flowchart detailing the parameter identification strategy for a unimorph subjected to parametric excitation and hw(s, t)i denotes the mean of w(s, t).

176

Ferroelectric Materials for Energy Harvesting and Storage

5.4.5 Validation of parameter identification strategy We choose a unimorph with geometric and linear material properties presented in Table 5.5. To implement the proposed parameter identification strategy we assume the same values for the nonlinear piezoelectric material parameters as presented in Table 5.2. For the material and geometric properties listed in Tables 5.2 and 5.5, the piecewise linear mass normalized mode shapes are determined as ϕ1  6:2½ cos ð6:35sÞ  cosh ð6:35sÞ  4:4½ sin ð6:35sÞ  sinh ð6:35sÞ for 0 < s < l1 (5.97a) ψ 1  10:43 cos ð6:84sÞ  11:19 cosh ð6:84sÞ  5:22 sin ð6:84sÞ

(5.97b)

+ 9:09 sinh ð6:84sÞ for l1 < s < l

and the coefficients in the governing equations are presented in Table 5.6. Time series and spectra of the numerically simulated response for the beam with piezoelectric layer subjected to parametric excitation with an excitation amplitude of 14 m s2 and a detuning frequency of Eσ ¼ 0.1 rad s1 are presented in Fig. 5.6A–D. In particular, the times series of the transverse displacement of the beam at 5 cm from the base and voltage generated are presented in Fig. 5.6A and C, the corresponding spectra are in Fig. 5.6B  and  D. We follow the procedure mentioned  in flowchart in Fig. 5.5. Noting that w s, Ω2 js¼5 cm  3:04 mm and given that w s, Ω2 ¼ ϕ1 ðsÞa0 , a0 is determined to be a0* ¼ 5.2  103 kg1/2 m. The mean value of the transverse displacement is then determined as hw(s, t)js¼5 cmi ¼ 0.19 mm. Noting hwðs, tÞi ¼ ϕ1 ðsÞ

a∗0 Eδ1 2ω2

(5.98)

Table 5.5 The physical and linear material properties of the unimorph subjected to direct excitation presented in Fig. 5.4. Parameters

Value

Length of the substrate, l (mm) Length of the piezoelectric layer, l1 (mm) Width of the structure, b (mm) Thickness of the substrate, hs (mm) Thickness of the piezoelectric layer, hp (mm) Density of the substrate (kg m3) Density of the piezo (kg m3) s Young’s modulus of the substrate, Y11 (G Pa) p Young’s modulus of the piezo, Y11 (G Pa) Piezoelectric coupling coefficient, e31 (C m2) Piezoelectric permittivity, ε33 (F m1) Damping ratio, ζ Load resistance, R (M Ohm)

300 60 10 0.5 0.26 8027 7800 210 66 12.54 1.328  108 0.004 10

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

177

Table 5.6 Coefficients in the governing equations corresponding to the parameters presented in Tables 5.2 and 5.5.

ω (rad) E2μ1 (s1) Eθ^ ≡ θ^v (C kg1/2 m1) Eδ1 (kg1/2 m1 s2) E2δ2 ≡ Eδ2v (C kg1 m2) E2δ3 (kg1 m2 s2) E2δ4 (kg1 m2) Eη (kg1/2 s2) Cp (nF)

34.50 0.14 1.06  103 26.22  103 0.06 2.18  106 6.13  103 5.48Ω2 30.65

(C)

(B)

Frequency (Hz)

Voltage (V)

Time (s)

Voltage (V)

(A)

w (mm)

Value

w (mm)

Parameters (units)

Time (s)

(D)

Frequency (Hz)

Fig. 5.6 Principal mode resonant response represented by the displacement at 5 cm away from the fixed end and the corresponding harvested voltage for a detuning value Eσ ¼ 0.1 rad s1 and excitation amplitude of ubΩ2 ¼ 14 m s2 in time domain (A and C) and their amplitudes in frequency domain (B and D).

178

Ferroelectric Materials for Energy Harvesting and Storage

Table 5.7 Summary of the results and efficiency of the parameter identification scheme proposed. Parameters

Value

Parameters

Value

% Error

α1 (T Pa) α3 (k C m2) α4 (P Pa)

100 5 50

α1* (T Pa) α3* (k C m2) α4* (P Pa)

107.94 5.06 44.99

7.94 1.23 10.01

we evaluate Eδ∗1 ¼ 28:3  104 kg1/2 m1 s2. Using Eq. (5.93b), Eδ∗1 , and a0*, we identify α1 as α1* ¼ 107.94 TPa. Next, by noting the amplitude of component of the voltage and phases of the two components, we determine V (Ω) ¼ 17.98 V. Given that VðΩÞ ¼

  a∗2 0 3Eδ2v ω2 + 2Eδ∗1 θ^v 12ω2 Cp

(5.99)

and using the identified values of a0*, Eδ1*, we evaluate that Eδ2v ¼ 6.45  102 C kg1 m2. Further, using Eq. (5.93b), we identify α3 as α3* ¼ 5061.8 C m2. We note all the parameters in the steady-state amplitude response relation mentioned in Eq. (5.94), except for δ3. Hence, we use the equation to numerically solve δ3 as δ3* ¼ 2.4  106 kg1 m2 s2. Based on Eq. (5.47c), and accounting for the geometric nonlinearity by evaluating the integrals of the mode shapes, we identify α4 as α4* ¼ 44.99 P Pa. The assumed and identified nonlinear material parameters are presented in Table 5.7. Based on percentage errors in Table 5.7, we conclude that the proposed parameter identification strategy can estimate the nonlinear material parameters with a reasonable accuracy.

5.5

Conclusions

In this chapter, we l

l

l

l

l

Presented general constitutive relations to model the nonlinear behavior exhibited by piezoelectric materials. Detailed the modeling of governing nonlinear equations using both stress-electric displacement constitutive relations and electromechanical enthalpy. Presented Galerkin weighted residual method to discretize the complicated distributed parameter coupled governing equation, boundary conditions, and compatibility conditions (if any). Presented the parameter identification strategies of the a unimorph under (i) direct and (ii) parametric excitation which involved analyzing the approximate solution of the governing nonlinear equations. The approximate solutions were determined using the method of multiple scales. The idea is to show two different types of excitations that can be utilized to identify the nonlinear material parameters for the same piezoelectric layer.

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting l

179

By assuming values for the nonlinear material parameters and from the simulated response, we showed that the both parameter identification strategies can be used to identify the assumed values with reasonable accuracy.

Appendices Appendix A Simplification of weighted residual statement: Direct excitation Z

l

YIeq q1 0

ϕ1 ϕ0000 1 dx ¼ YIeq q1

Z 0

l

00 0 ϕ001 ϕ001 ds + YIeq q1 ϕ000 1 ðlÞϕ1 ðlÞ  YIeq q1 ϕ1 ðlÞϕ1 ðlÞ

(A.1) Z YIeq q31

l 0

000 0 002 3 ϕ1 ðϕ01 ðϕ01 ϕ001 Þ0 Þ0 ds ¼ ðYIeq q31 ϕ1 ϕ02 1 ϕ1 + YIeq q1 ϕ1 ϕ1 ϕ1 Þjs¼l Z l 002 00 3  ðYIeq q31 ϕ03 ϕ Þj + 2YI q ϕ02 eq 1 1 s¼l 1 ϕ1 ds 1 0

(A.2) Z ρeq q1 ðq_ 21

l

+ q1 q€1 Þ



ϕ01

Z sZ

θ

ϕ02 1

0

ϕ1 dy dθ ds l Z0 Z  Z l  s θ 02 ϕ dy dθ ds ¼  ρeq q1 ðq_ 21 + q1 q€1 Þ ϕ02 1 1 0

0

l

(A.3)

0

Z

l V V 02 ϕ ϕ02 Hf 00 ds ¼ e31 Λ1 q21 ϕ ϕ Hf 0 js¼l 2hp 0 1 1 2hp 1 1 Z V l +e31 Λ1 q21 ϕ ϕ0 ϕ00 Hf 0 ds hp 0 1 1 1 Z l V + e31 Λ1 q21 ϕ03 Hf 0 ds 2hp 0 1 Z l 02 0 2 V 2V ¼ e31 Λ1 q1 ϕ ϕ Hf js¼l + e31 Λ1 q1 ϕ ϕ0 ϕ00 Hf 0 ds 2hp 1 1 hp 0 1 1 1 Z V 03 3V l 02 00 +e31 Λ1 q21 ϕ1 Hf js¼l  e31 Λ1 q21 ϕ ϕ Hf ds 2hp 2hp 0 1 1

e31 Λ1 q21

e31 Λ1

V hp

V e31 Λ1 hp

Z

l

ϕ1 Hf 00 ds ¼ e31 Λ1

0

Z

0

l

ϕ001 Hf

ds

V V ϕ1 Hf 0 + e31 Λ1 ϕ01 Hf hp hp s¼l s¼l

(A.4)

(A.5)

180

Ferroelectric Materials for Energy Harvesting and Storage

Z l V V ϕ ϕ00 2 Hf 00 ds ¼ α2 Λ2 q21 ϕ ϕ00 2 Hf 0 js¼l 2hp 0 1 1 2hp 1 1 Z l 2V +α2 Λ2 q1 ϕ ϕ00 ϕ000 Hf 0 ds hp 0 1 1 1 Z l  0003  V 0 002 V ϕ1 ϕ1 Hf  α2 Λ2 q21 ϕ1 + 2ϕ01 ϕ001 ϕ000 + α2 Λ2 q21 1 Hf ds 2hp 2hp 0 s¼l

α2 Λ2 q21

(A.6)

Z V l 00 000 0 2V 00 000 ϕ1 ϕ1 ϕ1 Hf ds ¼ α2 Λ2 q1 ϕ1 ϕ1 ϕ1 Hf hp 0 hp s¼l Z l   V 000 2 00 0000 + α2 Λ2 q21 ϕ01 ϕ001 ϕ000 1 + ϕ1 ϕ1 + ϕ1 ϕ1 ϕ1 Hf ds hp 0

α2 Λ2 q21

Z

(A.7)

Z l 2 00 000 2 2 ϕ1 ϕ001 ϕ0000 ds ¼ α Λ q ϕ ϕ ϕ + α Λ q ϕ1 ϕ000 1 2 1 2 1 1 1 1 1 s¼l 1 1 ds 0 0 Z l  00 2  00 + α1 Λ2 q21 ϕ01 ϕ001 ϕ001 s¼l  α1 Λ2 q21 ϕ1 + ϕ01 ϕ000 1 ϕ1 ds 0 Z l Z l 1 1 0 00 000 2 0 00 2 2 2 ϕ1 ϕ1 ϕ1 ds ¼  α1 Λ2 q1 ϕ1 ϕ1 js¼l + α1 Λ2 q1 ϕ001 3 ds since;  α1 Λ2 q1 2 2 0 0 Z l Z l 2 00 000 2 2  α1 Λ2 q21 ϕ1 ϕ001 ϕ0000 ds ¼ α Λ q ϕ ϕ ϕ + α Λ q ϕ1 ϕ000 1 2 1 1 1 1 s¼l 1 2 1 1 1 ds 0 0 Z l 1 1 2 0 00 2 2 2 0 00 00 ϕ001 3 ds + α1 Λ2 q1 ϕ1 ϕ1 ϕ1 s¼l  α1 Λ2 q1 ϕ1 ϕ1 js¼l  α1 Λ2 q1 2 2 0 α1 Λ2 q21

l

(A.8) Z l   V V 0 000 0 00 q1 ϕ1 ϕ001 Hf 00 + ϕ0000 Hf + 2Hf ϕ I q Hf ϕ ϕ ds ¼ α 3 p 1 1 1 1 1 hp h 0 s¼l Z l p  V  0 V 002 00 000 + α3 Ip q1 Hf ϕ1 ϕ1 + Hf ϕ1 ϕ1 + α3 Ip q1 Hf ϕ1 ds hp hp 0 s¼l

α 3 Ip

(A.9)  3 3 002 0000 002 000 3 + ϕ1 ϕ1 ϕ1 ds ¼ α4 Λ3 q1 ϕ1 ϕ1 ϕ1 2 2 0 s¼l Z l 1 1 3 0 003 3 004  α4 Λ3 q1 ϕ1 ϕ1 + α4 Λ3 q1 ϕ1 ds 2 0 s¼l 2

α4 Λ3 q31

Z l

3ϕ1 ϕ001 ϕ0002 1

(A.10)

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

Appendix B Z YIeq q1 0

l1

181

Simplification of weighted residual statement: Parametric excitation

ϕ1 ϕ0000 1

Z dx ¼ YIeq q1 0

l1

00 0 ϕ001 ϕ001 ds + YIeq q1 ϕ000 1 ðl1 Þϕ1 ðl1 Þ  YIeq q1 ϕ1 ðl1 Þϕ1 ðl1 Þ

(B.1) Z

l1

000 3 0 002 ϕ1 ðϕ01 ðϕ01 ϕ001 Þ0 Þ0 ds ¼ ðYIeq q31 ϕ1 ϕ02 1 ϕ1 + YIeq q1 ϕ1 ϕ1 ϕ1 Þjs¼l1 0 Z l1 002 3 03 00 3  ðYIeq q1 ϕ1 ϕ1 Þjs¼l1 + 2YIeq q1 ϕ02 1 ϕ1 ds

YIeq q31

0

(B.2)  Z sZ θ 0 0 02 + q1 q€1 Þ ϕ1 ϕ 1 ϕ1 dy dθ ds l1 0 0  Z sZ θ Z l1  02 ¼  ρeq q1 ðq_ 21 + q1 q€1 Þ ϕ02 ϕ dy dθ ds 1 1 Z

ρeq q1 ðq_ 21

l1

0

l1

(B.3)

0

Z l1 V V 02 ϕ ϕ02 Hf 00 ds ¼ e31 Λ1 q21 ϕ ϕ Hf 0 js¼l1 2hp 0 1 1 2hp 1 1 Z V l1 + e31 Λ1 q21 ϕ ϕ0 ϕ00 Hf 0 ds hp 0 1 1 1 Z l1 V ϕ03 Hf 0 ds + e31 Λ1 q21 2hp 0 1 Z V 02 V l1 ¼ e31 Λ1 q21 ϕ1 ϕ1 Hf 0 js¼l1 + e31 Λ1 q21 ϕ ϕ0 ϕ00 Hf 0 ds 2hp hp 0 1 1 1 Z V 03 3V l1 02 00 ϕ1 Hf js¼l1  e31 Λ1 q21 ϕ ϕ Hf ds + e31 Λ1 q21 2hp 2hp 0 1 1

(B.4)

Z V l1 V V ϕ1 Hf 00 ds ¼ e31 Λ1 ϕ1 Hf 0 + e31 Λ1 ϕ01 Hf : hp 0 hp hp s¼l1 s¼l1 Z l1 V 00 e31 Λ1 ϕ Hf ds hp 0 1

(B.5)

Z l1 V V ϕ1 ϕ001 2 Hf 00 ds ¼ α2 Λ2 q21 ϕ ϕ00 2 Hf 0 js¼l1 2hp 0 2hp 1 1 Z V l1 + α2 Λ2 q21 ϕ ϕ00 ϕ000 Hf 0 ds hp 0 1 1 1 Z l1  0003  V 002 0 2 V ϕ1 ϕ1 Hf  α2 Λ2 q21 ϕ1 + 2ϕ01 ϕ001 ϕ000 + α2 Λ2 q1 1 Hf ds 2hp 2hp 0 s¼l1

(B.6)

e31 Λ1 q21

e31 Λ1

α2 Λ2 q21

182

Ferroelectric Materials for Energy Harvesting and Storage

Z V l1 0 2V 00 000 ϕ1 ϕ001 ϕ000 Hf ds ¼ α Λ q ϕ ϕ ϕ Hf 2 2 1 1 1 1 1 hp 0 hp s¼l1 Z l1  0 00 000  000 2 00 0000 2V + α 2 Λ 2 q1 ϕ1 ϕ1 ϕ1 + ϕ1 ϕ1 + ϕ1 ϕ1 ϕ1 Hf ds hp 0

α2 Λ2 q21

Z α1 Λ2 q21

l1

0

ϕ1 ϕ001 ϕ0000 1



ds ¼ α1 Λ2 q21 ϕ1 ϕ001 ϕ000 1 s¼l1

+ α1 Λ2 q21 ϕ01 ϕ001 ϕ001 s¼l1  α1 Λ2 q21 Z since;  α1 Λ2 q21

l1

0

Z

Z

l1 

0

(B.7)

Z + α1 Λ2 q21

l1 0

 00 ϕ001 2 + ϕ01 ϕ000 1 ϕ1 ds

1 1 2 0 00 2 2 ϕ01 ϕ001 ϕ000 1 ds ¼  α1 Λ2 q1 ϕ1 ϕ1 js¼l1 + α1 Λ2 q1 2 2

Z

2 ϕ1 ϕ000 1 ds

l1

0

ϕ001 3 ds

Z l1 2 00 000 2 2 ϕ1 ϕ001 ϕ0000 ds ¼ α Λ q ϕ ϕ ϕ + α Λ q ϕ1 ϕ000 1 2 1 1 1 1 s¼l1 1 2 1 1 1 ds 0 Z l1 0 1 1 2 0 00 2 2 2 0 00 00 ϕ001 3 ds + α1 Λ2 q1 ϕ1 ϕ1 ϕ1 s¼l1  α1 Λ2 q1 ϕ1 ϕ1 js¼l1  α1 Λ2 q1 2 2 0

α1 Λ2 q21

l1

(B.8) Z l1   V V 0 000 0 00 q1 ϕ1 ϕ001 Hf 00 + ϕ0000 Hf + 2Hf ϕ I q Hf ϕ ϕ ds ¼ α 3 p 1 1 1 1 1 hp hp 0 s¼l1 Z l1   V V 002 0 00 000 + α3 Ip q1 Hf ϕ1 ϕ1 + Hf ϕ1 ϕ1 + α3 Ip q1 Hf ϕ1 ds hp hp 0 s¼l1

α 3 Ip

(B.9)  3 3 002 0000 002 000 3 ds ¼ 3ϕ1 ϕ001 ϕ0002 + ϕ ϕ Λ q ϕ ϕ ϕ ϕ α 4 3 1 1 1 1 1 1 1 1 2 2 0 s¼l1 Z l1 1 1 3 0 003 3 004  α4 Λ3 q1 ϕ1 ϕ1 + α4 Λ3 q1 ϕ1 ds 2 2 0 s¼l1 Z

α4 Λ3 q31

l1 

Z ρs ðq_ 21

l1

+ q1 q€1 Þ

ϕ1 ϕ001 q1

Z l Z

θ

ψ 02 1



dy dθ ds ¼ Zl1 l Z0 θ  2 02 0 ρs ðq_ 1 + q1 q€1 Þϕ1 ðl1 Þϕ1 ðl1 Þ ψ 1 dy dθ l1 0 Z l1   Z lZ θ 2 + ρs ðq_ 21 + q1 q€1 Þ ϕ1 0 ds ψ 02 dy dθ 1 0

0

Z

l1

(B.11)

0

 ϕ1 ρeq ϕ01 + ρeq ðs  l1 Þϕ001 + ρs ðl1  lÞϕ001 ds ¼ u€b q1 € ub q 1 0 Z l1 ρs ðl1  lÞϕ02 ds  u€b q1 ρs ðl1  lÞϕ1 ϕ01 s¼l1 + u€b q1 l1

(B.10)

Z 0

l1

ρeq ðs  l1 Þϕ02 ds

0

(B.12)

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

Z s Y11 Is q1

l

l1

s ψ 1 ψ 0000 1 ds ¼ Y11 Is q1

Z l1

l

183

s¼l s s 00 0 s¼l ψ 001 ψ 001 ds + Y11 Is q1 ψ 000 1 ψ 1 s¼l1  Y11 Is q1 ψ 1 ψ 1 s¼l1 (B.13)

Z

l

s 000 s 3 0 002 s¼l ψ 1 ðψ 01 ðψ 01 ψ 001 Þ0 Þ0 ds ¼ ðY11 Is q31 ψ 1 ψ 02 1 ψ 1 + Y11 Is q1 ψ 1 ψ 1 ψ 1 Þjs¼l1 Z l 002 s 3 03 00 s¼l s 3  ðY11 Is q1 ψ 1 ψ 1 Þjs¼l1 + 2Y11 Is q1 ψ 02 1 ψ 1 ds

s Y11 Is q31

l1

l1

(B.14)  Z sZ θ 0 0 02 + q1 q€1 Þ ψ 1 ψ 1 ψ 1 dy dθ ds ¼ l1  Z l Z l s Z 0θ 02 ψ dy dθ ds ρs q1 ðq_ 21 + q1 q€1 Þ ψ 02 1 1 l1 l 0 Z l Z θ  + ρs q1 ðq_ 21 + q1 q€1 Þψ 1 ðl1 Þψ 01 ðl1 Þ ψ 02 dy dθ 1 Z

ρs q1 ðq_ 21

l

l1

Z

0

 ψ 1 ρs ψ 01 + ρs ðs  lÞψ 001 ds ¼ u€b q1 € u b q1 l1 + u€b q1 ρs ðl1  lÞψ 1 ψ 01 s¼l l

(B.15)

Z l1

l

ρs ðs  lÞψ 02 ds

(B.16)

1

References [1] N.S. Shenck, J.A. Paradiso, Energy scavenging with shoe-mounted piezoelectrics, IEEE Micro 21 (3) (2001) 30–42. [2] P. Glynne-Jones, S.P. Beeby, E.P. James, N.M. White, The modelling of a piezoelectric vibration powered generator for microsystems, Transducers’ 01 Eurosensors XV, Springer, 2001, pp. 46–49. [3] G.K. Ottman, H.F. Hofmann, G.A. Lesieutre, Optimized piezoelectric energy harvesting circuit using step-down converter in discontinuous conduction mode, IEEE Trans. Power Electron. 18 (2) (2003) 696–703. [4] S. Roundy, P.K. Wright, A piezoelectric vibration based generator for wireless electronics, Smart Mater. Struct. 13 (5) (2004) 1131. [5] D.J. Inman, B.L. Grisso, Towards autonomous sensing, in: Smart Structures and Materials 2006: Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems, 6174, International Society for Optics and Photonics, 2006, p. 61740T vol. [6] N. Jalili, Piezoelectric-Based Vibration Control: From Macro to Micro/Nano Scale Systems, Springer Science+Business Media, LLC, New York, NY, 2009. [7] G.W. Taylor, J.R. Burns, S. Kammann, W.B. Powers, T.R. Welsh, The energy harvesting eel: a small subsurface ocean/river power generator, IEEE J. Ocean Eng. 26 (4) (2001) 539–547. [8] H.D. Akaydin, Piezoelectric Energy Harvesting From Fluid Flow (Dissertation), City University of New York, 2012.

184

Ferroelectric Materials for Energy Harvesting and Storage

[9] A. Abdelkefi, Aeroelastic energy harvesting: a review, Int. J. Eng. Sci. 100 (2016) 112–135. [10] A. Erturk, J.M. Renno, D.J. Inman, Modeling of piezoelectric energy harvesting from an L-shaped beam-mass structure with an application to UAVs, J. Intell. Mater. Syst. Struct. 20 (5) (2009) 529–544. [11] S.C. Stanton, C.C. McGehee, B.P. Mann, Nonlinear dynamics for broadband energy harvesting: investigation of a bistable piezoelectric inertial generator, Phys. D. Nonlinear Phenomena 239 (10) (2010) 640–653. [12] S. Adhikari, M. Friswell, D. Inman, Piezoelectric energy harvesting from broadband random vibrations, Smart Mater. Struct. 18 (11) (2009) 115005. [13] A.F. Arrieta, P. Hagedorn, A. Erturk, D.J. Inman, A piezoelectric bistable plate for nonlinear broadband energy harvesting, Appl. Phys. Lett. 97 (10) (2010) 104102. [14] M. Abe, G. Park, D.J. Inman, Impedance-based monitoring of stress in thin structural members, UNIT 45002 APO AP 96337-5002, 2001, p. 285. [15] G. Park, H. Sohn, C.R. Farrar, D.J. Inman, Overview of piezoelectric impedance-based health monitoring and path forward, Shock Vib. Digest 35 (6) (2003) 451–464. [16] T. Bailey, J.E. Ubbard, Distributed piezoelectric-polymer active vibration control of a cantilever beam, J. Guid. Control Dynam. 8 (5) (1985) 605–611. [17] N.W. Hagood, A. von Flotow, Damping of structural vibrations with piezoelectric materials and passive electrical networks, J. Sound Vib. 146 (2) (1991) 243–268. [18] W.W. Clark, Vibration control with state-switched piezoelectric materials, J. Intell. Mater. Syst. Struct. 11 (4) (2000) 263–271. [19] V. Giurgiutiu, Review of smart-materials actuation solutions for aeroelastic and vibration control, J. Intell. Mater. Syst. Struct. 11 (7) (2000) 525–544. [20] L. Librescu, S. Na, Active vibration control of doubly tapered thin-walled beams using piezoelectric actuation, Thin-Walled Struct. 39 (1) (2001) 65–82. [21] S.O.R. Moheimani, A survey of recent innovations in vibration damping and control using shunted piezoelectric transducers, IEEE Trans. Control Syst. Technol. 11 (4) (2003) 482–494. [22] M. Safaei, S.R. Anton, Sensing and energy harvesting performance, and fatigue life of embedded piezoelectric transducer in total knee arthroplasty, ASME 2016 Conference on Smart Materials, Adaptive Structures and Intelligent Systems, American Society of Mechanical Engineers, 2016. V002T07A010. [23] E. Hoummadi, M. Safaei, S.R. Anton, Design, analysis, and fabrication of a piezoelectric force plate, in: Health Monitoring of Structural and Biological Systems 2017, vol. 10170, International Society for Optics and Photonics, 2017, p. 101700W. [24] T. Maleki, N. Cao, S.H. Song, C. Kao, S.-C. Ko, B. Ziaie, An ultrasonically powered implantable micro-oxygen generator (IMOG), IEEE Trans. Biomed. Eng. 58 (11) (2011) 3104–3111. [25] M.G.L. Roes, J.L. Duarte, M.A.M. Hendrix, E.A. Lomonova, Acoustic energy transfer: a review, IEEE Trans. Ind. Electron. 60 (1) (2013) 242–248. [26] S. Shahab, M. Gray, A. Erturk, Ultrasonic power transfer from a spherical acoustic wave source to a free-free piezoelectric receiver: modeling and experiment, J. Appl. Phys. 117 (10) (2015) 104903. [27] M.F. Daqaq, R. Masana, A. Erturk, D.D. Quinn, On the role of nonlinearities in vibratory energy harvesting: a critical review and discussion, Appl. Mech. Rev. 66 (4) (2014) 040801. [28] S. Leadenham, A. Erturk, M-shaped asymmetric nonlinear oscillator for broadband vibration energy harvesting: harmonic balance analysis and experimental validation, J. Sound Vib. 333 (23) (2014) 6209–6223.

Modeling and identification of nonlinear piezoelectric material properties for energy harvesting

185

[29] R.L. Harne, K.W. Wang, A review of the recent research on vibration energy harvesting via bistable systems, Smart Mater. Struct. 22 (2) (2013) 023001. [30] G.A. Maugin, Nonlinear Electromechanical Effects and Applications, vol. 1, World Scientific Publishing Company, Singapore, 1986. [31] H.F. Tiersten, Electroelastic equations for electroded thin plates subject to large driving voltages, J. Appl. Phys. 74 (5) (1993) 3389–3393. [32] N. Aurelle, D. Guyomar, C. Richard, P. Gonnard, L. Eyraud, Nonlinear behavior of an ultrasonic transducer, Ultrasonics 34 (2–5) (1996) 187–191. [33] D. Guyomar, N. Aurelle, L. Eyraud, Piezoelectric ceramics nonlinear behavior. Application to Langevin transducer, J. Phys. III 7 (6) (1997) 1197–1208. [34] U. Von Wagner, P. Hagedorn, Piezo-beam systems subjected to weak electric field: experiments and modelling of non-linearities, J. Sound Vib. 256 (5) (2002) 861–872. [35] S.N. Mahmoodi, N. Jalili, M.F. Daqaq, Modeling, nonlinear dynamics, and identification of a piezoelectrically actuated microcantilever sensor, IEEE/ASME Trans. Mechatron. 13 (1) (2008) 58–65. [36] S.C. Stanton, A. Erturk, B.P. Mann, D.J. Inman, Nonlinear piezoelectricity in electroelastic energy harvesters: modeling and experimental identification, J. Appl. Phys. 108 (7) (2010) 074903. [37] S. Leadenham, A. Erturk, Unified nonlinear electroelastic dynamics of a bimorph piezoelectric cantilever for energy harvesting, sensing, and actuation, Nonlinear Dyn. 79 (3) (2015) 1727–1743. [38] A.H. Nayfeh, Parametric identification of nonlinear dynamic systems, in: Advances and Trends in Structures and Dynamics, Elsevier, 1985, pp. 487–493. [39] A.H. Nayfeh, Parametric identification of nonlinear dynamic systems, Comput. Struct. 20 (1–3) (1985) 487–493. [40] L.D. Zavodney, Identification of nonlinearity in structural systems: theory, simulation, and experiment, Appl. Mech. Rev. 44 (11S) (1991) S295–S303. [41] M.R. Hajj, J. Fung, A.H. Nayfeh, S. Fahey, Damping identification using perturbation techniques and higher-order spectra, Nonlinear Dyn. 23 (2) (2000) 189–203. [42] G. Kerschen, K. Worden, A.F. Vakakis, J.-C. Golinval, Past, present and future of nonlinear system identification in structural dynamics, Mech. Syst. Signal Process. 20 (3) (2006) 505–592. [43] A.H. Nayfeh, D.T. Mook, Nonlinear Oscillations, John Wiley & Sons, Toronto, ON, 2008. [44] V.C. Meesala, M.R. Hajj, Identification of nonlinear piezoelectric coefficients, J. Appl. Phys. 124 (6) (2018) 065112. [45] A. Erturk, D.J. Inman, Piezoelectric Energy Harvesting, John Wiley & Sons, Chichester, 2011. [46] V. Birman, Physically nonlinear behavior of piezoelectric actuators subject to high electric fields, MISSOURI UNIV-ROLLA, 2005 Tech. Rep. [47] A.H. Nayfeh, P.F. Pai, Linear and Nonlinear Structural Mechanics, John Wiley & Sons, Toronto, ON, 2008. [48] L. Meirovitch, Fundamentals of Vibrations (Master’s thesis), Waveland Press, 2010. [49] H.A. Sodano, Macro-Fiber Composites for Sensing, Actuation and Power Generation (Ph.D. thesis), Virginia Tech, 2003.

Woodhead Publishing Series in Composites Science and Engineering

Sustainable Composites for Lightweight Applications Hom Nath Dhakal Sikiru Oluwarotimi Ismail

Woodhead Publishing is an imprint of Elsevier The Officers’ Mess Business Centre, Royston Road, Duxford, CB22 4QH, United Kingdom 50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States The Boulevard, Langford Lane, Kidlington, OX5 1GB, United Kingdom Copyright © 2021 Elsevier Ltd. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library ISBN: 978-0-12-818316-8 For information on all Woodhead Publishing publications visit our website at https://www.elsevier.com/books-and-journals Publisher: Matthew Deans Acquisitions Editor: Gwen Jones Editorial Project Manager: Andrea Gallego Ortiz Production Project Manager: Anitha Sivaraj Cover Designer: Alan Studholme Typeset by TNQ Technologies

Preface

Environmental damage caused by using and disposing of plastics and composites is worrisome. If this trend continues, it could lead to unprecedented damage to our natural resources. For example, with new environmental legislations, the European Union (EU) environmental regulation aims to reduce CO2 emissions in automotive components by using new technology, improved design and overall weight reduction. Composite materials reinforced with carbon and glass fibres are extensively used as structural materials, owing to their excellent strengths and stiffness-to-weight ratios. These attributes make these materials very attractive options for critical industrial sectors, such as aerospace, marine and automotive. It is logical to think that being able to reduce the overall mass results in a significant volume reduction, which consequently leads to the use of less raw materials and overall CO2 reduction. However, composite reinforcements such as glass and carbon fibres, currently being used in the aforementioned transportation sectors, have low recyclability after the end-of-life and high overall energy consumption during their production. Due to these environmental and energy utilisation concerns, a new class of materials, including natural plant fibre-reinforced composites, are being introduced in these critical sectors. This book entitled Sustainable composites for lightweight applications is a reflection of several years of experience gained by the authors in the field of advanced and sustainable composite materials and manufacturing. In this instance, Professor Hom Nath Dhakal has been involved in the design and development of sustainable lightweight composite materials for over 20 years. He has over 30 years of teaching and research experience in the field of Mechanical Engineering, Materials and Manufacturing. Professor Dhakal has published in over 150 international peerreviewed journals, book chapters and conference proceedings. Professor Dhakal is a Chartered Engineer (CEng), a member of the American Society for Composites (MASC) and a Fellow of the Higher Education Academy (FHEA), the Institution of Engineering and Technology (FIET) and the Institute of Materials, Minerals, and Mining (IOM3) (FIMMM). Similarly, Dr Sikiru Oluwarotimi Ismail is currently a Senior Lecturer in Manufacturing Engineering and Materials. His research specialisation focuses on Advanced and Sustainable Materials and Manufacturing (Mechanical) Engineering: Design, development, testing, damage characterisation and optimisation of materials

x

Preface

(especially composites), innovative manufacturing processes/monitoring and optimisation of manufacturing. He has over 10 years of teaching and research experience in the aforementioned field. Dr Ismail has published in over 60 international peerreviewed journals, book chapters and conference proceedings. In addition, Dr Ismail is a Member of many national/local and international professional bodies: American Society of Mechanical Engineers (MASME), Institution of Mechanical Engineers (MIMechE), American Society for Composites (MASC), Chartered Engineer of Institution of Engineering and Technology (CEng MIET), a Fellow of Higher Education Academy (FHEA) and Royal Society for the encouragement of Arts, Manufactures and Commerce (FRSA).

Key features of this book The inception of the book was made when both authors were teaching undergraduate and postgraduate engineering modules at the University of Portsmouth, such as Sustainable Development and Environmental Management, Strategies for Resource and Environmental management, as well as Materials and Manufacture. The planning of this book was immersed and started in 2018, when the authors strongly felt that there was a knowledge gap in sustainable composites, especially applying them in lightweight applications. The main purpose of bringing up this book is to elucidate the importance of the development of environmentally friendly materials to the industrial sectors, with the required mechanical properties and functionality. This book introduces the exciting field of biobased composite materials. Some real-world examples have been provided to explain the relevant topics. For each individual chapter, an abstract, keywords and conclusions are provided.

Target audiences of this book The readers can explore important aspects of biobased composites, their properties, damage analysis and use in critical application areas. Future perspectives of sustainable composites have been systematically presented. The contents of this book will help to appreciate the sustainable materials of the future. Also, teachers, researchers, scientists and industries can appreciate the benefits and apply the knowledge for various applications.

Chapter highlights of this book This book has seven chapters. Chapter 1 introduces the general introduction of composite materials and their key features focusing on biobased composites. In Chapter 2, the structure and the morphological aspects of sustainable natural fibre reinforcements are highlighted. Interestingly, Chapter 3 discusses important aspects of lightweight composites, their properties and various applications. In Chapter 4, design, manufacturing processes and their effects on biocomposites properties are analysed

Preface

xi

and discussed. Moving forward, Chapter 5 focuses on the testing and damage characterisation of biobased composite materials. Chapter 6 considers the different improvement techniques and their influences on the properties enhancement. Finally, Chapter 7 summarises the various aspects and puts forward the future perspectives and the challenges of sustainable composite materials for various applications while moving forward in this thriving field. Hom Nath DHAKAL Sikiru Oluwarotimi ISMAIL

Introduction to composite materials 1.1

1

Background and context

A composite material is composed of at least two visually distinct materials, which combine to give properties superior to those of the individual constituents while retaining their respective chemical, physical and mechanical properties while contributing desirable properties to the whole (Hull and Clyne, 1996; Matthews and Rawlings, 1994). The primary reason why composite materials used for engineering applications are due to their high performance relating to improved specific strength and stiffness (strength to-weight-ratio). This attribute helps in reducing the overall weight of components. If one considers an automotive part, for example, the reduction of overall weight leads to the reduction of fuel consumption, increased performance and eventually leads to the reduction of CO2 emissions (Erbach, 2014). There are several factors that influence the overall performance of composite materials. Important properties such as tensile strength and modulus, impact resistance and fracture toughness behaviours (mode I, mode II and mixed-mode), vibration behaviour related to damping, thermal properties such as thermal decomposition, coefficient of thermal expansion (CTE), and thermal conductivity are directly related to the reinforcements types, their volume and geometry and how they were processed and prepared. Therefore, understanding these key parameters is very important in the process of design, manufacturing and service life of composites (Faruk et al., 2012; Monteiro et al., 2010; Paturel and Dhakal, 2020). Despite many research work directed in these composites, their structure-property relationships, reinforcement types and their influence on the various properties, their environmental impacts and end-of-life (EoL) aspects are still not fully understood. This chapter attempts to provide a basic introduction of composite materials and highlights the important aspects in terms of understanding their structure, properties, applications and EoL options while considering sustainable composites for lightweight applications. There are four main types of composites, namely, metal-matrix composites (MMCs), ceramic-matrix composites (CMCs), fibre-reinforced polymer matrix composites (FRPCs). Metal matrix composites consist of fibres reinforced with metal alloys. Such composites can withstand high temperatures, unlike other composites, but are heavy due to the presence of metal. MMCs are used in the automotive industry where metals and alloys are used as matrix material and reinforced with fibres or particulates. Under this, aluminium matrix composites are the most commonly used as matrix material. CMCs use ceramics as matrix material and reinforce with short fibres such as silicon carbide and boron nitride. CMCs have excellent resistance to high temperatures owing to the presence of ceramic. This is one of the major

Sustainable Composites for Lightweight Applications. https://doi.org/10.1016/B978-0-12-818316-8.00001-3 Copyright © 2021 Elsevier Ltd. All rights reserved.

2

Sustainable Composites for Lightweight Applications

Table 1.1 Advantages and some drawbacks of FRPCs (Faruk et al., 2012; Dhakal et al., 2007; Bhardwaj, 2017). Advantages

Some drawbacks

• low density, high specific strength and stiffness (high strength-to-weight ratio) • less fatigue sensitive • economical (cost-effective) • corrosion resistance • good surface finish can be obtained • properties can be tailored in the fibre direction • low coefficient of thermal expansion (LTE) • part count reduction

• susceptible to chemical and solvent attacks • low reusability or recyclability in terms of carbon and glass fibre-reinforced composites • fluctuating cost • damage modes difficult to detect

advantages of the ceramic matrix. Therefore, these composites are used in hightemperature applications where compressive strength is more demanding than tensile and impact properties (Silvestre et al., 2015). Carbon and glass fibre-reinforced composites are very important materials in many engineering applications due to their lightweight, commercially available in continuous form and their corrosion resistance behaviour. Due to their attractive attributes, composites materials are extensively used in many critical applications such as aerospace, automotive, marine, construction and sports equipment. In the last 3 decades, polymeric composites have been one of the most attractive materials due to their versatile properties. One of the most commonly utilized composite types is Polymer Matrix Composites (PMCs), also known as fibre-reinforced plastics (FRPs). These composite materials are reinforced with many different types of fibres such as synthetic fibres (carbon, glass and aramids), natural biofibres (hemp, flax, jute, date palm kenaf among others) (Bhardwaj, 2017; Dhakal and Sain, 2019). Due to the several outstanding properties of composite materials, different key industry sectors are increasingly using composites instead of metallic materials in many structural and semi-structural applications. Some key advantages and drawbacks of synthetic fibre-reinforced polymer composites (FRPCs) against their metal counterparts are highlighted in Table 1.1. Growing environmental concern, high rate of depletion of petroleum-based materials; new stricter environmental regulations have resulted in research for alternative fibre-reinforced composites that are compatible with the environment. Natural fibrereinforced composites (NFRCs) have received much attention in recent years due to their many attractive properties such as high specific tensile strength and modulus compared to conventional glass fibres. Natural fibres represent an environmentfriendly alternative to conventional fibre reinforcements. Natural fibres are emerging as low cost, lightweight and biodegradable alternatives to glass fibres in composites (Dhakal et al., 2013; Dhakal et al., 2014). Moreover, the environmental impact of natural fibre versus glass and carbon fibres as reinforcements in composite fabrication

Introduction to composite materials

3

Table 1.2 Total energy required for production and cost of different for natural fibres versus glass and carbon fibres (Huda et al., 2008). Fibre types

Cost (US$/ton)

Energy (GJ/ton)

Natural fibres

200e1000

4

Glass fibres

1200e1800

30

Carbon fibres

12,500

130

NFRPCs fully / partially biodegradable

Applications

Composite development

NFRPCs

Amalagamation of natural fibre and polymer

Life cycle of NFRPCs

Scrap disposal

Fibre extractuion Landfills

CO2+ sun light Composting

Figure 1.1 Life cycle stages of natural fibre composites (Khan et al., 2018).

have been reported through the use of life cycle assessment (LCA). These various reports suggest that natural fibres exhibit significantly lower cost and lower energy consumption compared to glass and carbon fibres, which is shown in Table 1.2. The life cycle stages of natural fibre composites are illustrated in Fig. 1.1. However, due to their biochemical composition, these fibres are hydrophilic and need some kind of treatments to enhance the compatibility with hydrophobic thermoplastic and thermosets matrices. This shortcoming of natural fibre composites restricts the use of these composites in many non-structural and structural applications. The moisture absorption can lead to swelling of the fibres creating voids and microcracks at the fibre-matrix interface region resulting in a significant reduction of load transfer capability from the matrix to reinforcing fibres leading to reduction of mechanical properties. Many reported works on natural fibre composites have revealed that these shortcomings of natural fibre reinforcements have been minimised by modifying

4

Sustainable Composites for Lightweight Applications

fibre surfaces using various treatments achieving improvements in physical, mechanical and thermal properties by making compatible with different polymer matrices (Mohanty et al., 2003). The properties of fibre-reinforced composites depend on many factors, for example, types of matrices and reinforcement used, fibre volume fraction, fibre aspect ratio (length divided by the diameter of the fibre, L/d), fibre geometry and interfacial adhesion between reinforcement and the matrix. It is generally assumed that the higher the aspect ratio, the better the properties, up to a certain threshold. With longer thinner fibres the stress is able to travel along the fibres distributing the load more evenly than if the fibres are shorter; this, in turn, increases the mechanical properties. However, only the case up until a threshold aspect ratio is met, if the fibre aspect ratio is too high, i.e., the fibre size is increased, then fibre bunching can occur resulting in poor cohesion at the interface and a larger number of voids. At a low aspect ratio, the addition of reinforcement into the composite can create the phase of discontinuity, leading to heterogeneity structure and can result in poor mechanical performance. At a higher aspect ratio up to its threshold, the mechanical properties are expected to increase, because of good interfacial interaction between the matrix and the reinforcement (Hull and Clyne, 1996). In composite fabrication, the matrix wets the fibres, then the matrix and the fibres are bonded together and the resulting material, i.e. the composite, becomes far superior compared to the individual constituents. It is worth noting that these reinforcing fibres possess very high strength on their own, but when the load is applied, they break easily due to various defects. With this, outstanding mechanical properties can be achieved. When the polymeric matrix and fibres are combined, they provide significantly higher strength of resultant composite compared to individual fibre and matrix on their own. As can be seen in Fig. 1.2, the tensile properties of neat polypropylene (PP) and polylactic acid (PLA) have been significantly improved with the reinforcement of flax fibre onto PP and PLA. The tensile stress and modulus for both PP and flax/PP and flax PLA have been significantly improved. It is evident that the fibre reinforcement significantly contributes to the overall properties improvement of the resulting composites.

(a)

(b) 70

10

60

8

40

GPa

MPa

50 30

6 4

20 2

10 0

0 PP 0%

PLA 30%

40%

PP 0%

PLA 30%

40%

Figure 1.2 Comparison of tensile stress and modulus of (a) tensile stress of PP and PLA/flax composites in comparison to (b) modulus of PP and PLA/flax fibre-reinforced composites (Oksman et al., 2003).

Introduction to composite materials

1.2

5

Matrices and their types

The word “matrix” is a multipurpose word with wide applications and interpretations. In the case of material science, matrix refers to ceramic, metal or polymer matrices of a composite material. In the field of composite science, polymeric matrices are the most abundantly used materials. Poly comes from the Greek word for “many” and “mer” comes from the Greek word for “parts”. Therefore, a polymer is a long chain of molecules made up of from a series of repeating basic units called “mer”. Polymeric materials are most commonly used in every aspect of our lives. These materials are usually known as plastic and are the most widely used in composite manufacturing. These matrices are sub-divided into two types; thermoplastics and thermosets polymers. Fig. 1.3 shows a simple classification of different polymer matrices used in composites. The main differences between these two types of matrices are that thermoplastics can be re-heated and reformed to alter the shape or allow the shape of a component to be reformed, i.e., they are re-workable at temperature. Thermoplastic matrices are with high molecular weight, long chain molecules that can either be amorphous or partially crystalline (Stewart, 2011). Thermoplastic-reinforced composite materials are advantageous when strength, as well as improved toughness behaviour are important. The ability to reheat thermoplastic matrices offers re-workability and being able to formulate pellets, which allows greater freedom in manufacturing (Mallick, 2008; Stewart, 2011). Thermoplastics, as a matrix in material composites, have been growing gradually due to their recycling ability and rapid production cycle (Armentia et al., 2019). Thermosets matrices are resins, which cross-link during curing (hardening), resulting in glassy brittle solids such as epoxy, vinyl ester and polyesters. The cross-linked structures prevent the polymer from flowing and melting, which can provide the thermal stability of the polymer. The used or waste thermosets cannot be reused or recycled. Comparison between these two polymers in terms of their key advantages and disadvantages are highlighted in Table 1.3. Composites Matrix

Metal

Polymer

Thermosetting

Reinforcement

Ceramic

Biodegradable (natural/manmade)

Non-biodegradable

Thermoplastics

Biodegradable/non-biodegradable

NFRPCs (fully/partially biodegradable)

Figure 1.3 Classification of matrices and reinforcements for polymeric composites (Khan et al., 2018).

6

Sustainable Composites for Lightweight Applications

Table 1.3 Key advantages and disadvantages of thermosets and thermoplastics polymers (Nordin et al., 2013). Advantages

Disadvantages

Thermosets

• low resin viscosity • good fibre wetting • excellent thermal stability once cured • chemically resistant • better creep resistant than thermoplastics

• prone to brittleness • non-recyclable via standard techniques • not post-formable

Thermoplastics

• recyclable • easy to repair by welding and solvent bonding • post-formable • good impact toughness

• poor melt flow • need to be heated above the melting point for processing purposes

1.2.1

Types and main functions and the properties of matrices

In order to get optimal composite properties, a proper cure of matrix (chemical reaction) is very important. The functions of a matrix are as follows: • • • • •

Binding of the fibres (reinforcements) and the transferring of the applied load to the fibres Isolates the fibres so that individual fibres can act separately (fibre-matrix interface) Provides good surface finish and machinability Provides protection to fibres from chemical attack and other mechanical wear Provides other secondary properties

In summary, the main functions of the matrix in reinforced composites are to support and transfer the stresses to the fibres (reinforcements), which carry most of the load and protect the fibres against physical damage and the environment. The commonly used thermosets and thermoplastics matrices and their physical and mechanical properties are illustrated in Table 1.4.

1.2.1.1

Epoxy resins

These resins are high-performance matrices in terms of their mechanical properties. Due to the cross-linked structure, the epoxy matrix has good dimensional stability. They have high strain to failure compared to other resins such as polyesters and vinyl esters. Epoxy resin has good chemical resistance, as well as less prone to moisture in comparison to polyesters and vinyl ester. However, epoxies are brittle polymers, and often the toughness of this resin is enhanced by using some additives. Generally, when the toughness is increased, the modulus and glass transition temperature (Tg) gets decreased.

1.2.1.2

Polyester resins

These are the most widely used resins, especially in the marine industry, in building yachts and boats. There are two types of polyester resins, namely saturated and

Introduction to composite materials

7

Table 1.4 Physical and mechanical properties of commonly used thermosets and thermoplastics matrices (Manaia et al., 2019). Density (g/cm3)

Young’s modulus (GPa)

Tensile strength (MPa)

Strain to failure (%)

Epoxy resins

1.1e1.4

3.0e6.0

35e100

1e6

Polyesters

1.2e1.5

2e4.5

40e90

4e7

Vinyl ester

1.2e1.4

3.1e3.8

40e90

2

HDPE

0.94e0.96

1.1e1.60

30e40

2e130

Polypropylene (PP)

0.89e0.92

1.0e1.4

0.02e0.04

20e400

PS

1.04e1.06

4e5

25e69

1e2.5

PLA

1.21e1.25

0.35e3.5

21e60

2.5e6

Matrix types

Thermosets

Thermoplastics

unsaturated. These resins are cost-effective resins, and the price is far lower than epoxy and vinyl ester. The mechanical properties, especially interlinear shear stress (ILSS) strain due to failure of these resins, are far lower than that of vinyl ester and epoxy. The unsaturated polyesters are the backbone of the composite industry, with 75% of the resin being used.

1.2.1.3

Vinyl ester resins

These resins have properties somewhere between epoxy and polyesters. They possess good toughness, as well as better repellence to water and other chemical attacks.

1.2.1.4

Phenolic resins

Phenolic resins are an attractive matrix, especially where fire resistance attributes are of importance. Phenolic resins show superior fire resistance behaviour compared to other thermosetting resins such as epoxy and vinyl ester.

1.2.1.5

Polyethylene

Polyethylene (PE) is a common yet extremely useful and cost-effective plastic polymer. PE is found nearly everywhere today, from plastic grocery bags, plastic wrap, drain pipes, milk cartons, to trash cans. PE is an easily processed thermoplastic, which can be made into a variety of shapes and forms, including tubing. An especially convenient quality of PE is its ability to be easily altered during processing to give a variety of forms that differ based on polymer chain length, density and crystallinity. These characteristics allow PE products to be tailored for a variety of uses. As an example, high-density PE (HDPE) has a comparatively more linear morphology and a higher degree of crystallinity than low-density PE (LDPE). HDPE is lightweight and

8

Sustainable Composites for Lightweight Applications

possesses good tensile strength, while LDPE exhibits good chemical resistance. PE can be further modified by resin manufacturers to increase its structural and functional properties. PE polymer chains can be extended to produce ultra-high molecular weight (UHMW) PE to give a very dense PE product. Linear LDPE (LLDPE) has a greater proportion of short branches resulting in greater flexibility.

1.2.1.6

Polypropylene

Polypropylene (PP) is a tough and rigid, semicrystalline thermoplastic produced from propene (or propylene) monomer. PP is among the most commonly used cheapest thermoplastics. PP has poor resistance to UV, impact and scratches with low service temperature. PP is widely used in various applications, including packaging (flexible and rigid), due to its good barrier properties, high strength and good surface finish. PP is also considered as one of the emerging polymers in the automotive industry. PP is used in medical applications due to high chemical and bacterial resistance, as well as good mechanical properties. The main applications in automotive include battery cases and trays, front and rear bumpers, fender liners, interior trim, instrumental panels and door trims. There are self-reinforced PP also available, which provides improved strength. The selfreinforced PP has attracted a great deal of attention in automotive applications recently due to their key advantages such as no skin irritation, easy to handle and good recyclability over glass fibre-reinforced composites. Additional advantages of self-reinforced PP are as follows: • Lightweight • Good toughness properties • Good strain to failure property.

1.2.1.7

Polystyrene

Polystyrene (PS) is one of the most commonly used thermoplastics. It is easy to process, and semi-finished products such as foams, sheets and films are produced from PS. Additional advantages of self-reinforced PS are as follows: • Lightweight • Optical clarity • Easy to process

Some disadvantages of PS are as follows: • Poor oxygen and UV resistance • Poor impact resistance.

1.2.1.8

Polylactic acid

One of the great attractions of using polylactic acid (PLA) is its green attributes. PLA is fully degradable, and PLA reinforced with natural fibres becomes 100% bio-based composites.

Introduction to composite materials

9

Additional advantages of PLA are as follows: • Lightweight, high glossy appearance • Good strength • Good transparency

Some disadvantages of PLA are as follows: • Poor toughness (highly brittle material) • Higher cost than other commonly used thermoplastic matrices.

1.3

Reinforcements and their types

In terms of reinforcements, commonly used in conventional composites, there are three main types: glass, carbon and ceramic fibres. These reinforcements provide required strength, durability and overall quality of the composite parts. Some of the factors governing fibre reinforcement contributions to the composites are; 1. 2. 3. 4. 5.

The mechanical properties of the fibre Adhesion (interaction) between the fibre and the matrix Volume or weight fraction of fibre in the composite Fibre orientation in the composites Manufacturing process used to make the fibre.

1.3.1 1.3.1.1

Conventional reinforcements and their types Glass fibres

Most of the glass fibre used in the composite is made from molten glass combined with silica. This gives glass fibre superior bulk properties such as strength, stiffness, flexibility and hardness. Due to these attributes, glass fibres are used in structural composites. The main types of glass fibre that are used in the production of reinforced composites are S-glass and E-glass. Glass fibres normally come in two types: • The low-cost general-purpose fibres (E-glass) and • The premium special-purpose fibres.

Among the key advantages for using glass fibres are the reduced costs, the inability to conduct heat and therefore increased potential for insulation, as well as its good mechanical properties and anti-corrosiveness. These fibres are extensively used in the automotive and marine sectors. However, nowadays, carbon fibres are used in these applications instead of glass fibres due to their high-strength-to weight ratio.

10

1.3.1.2

Sustainable Composites for Lightweight Applications

Carbon fibres

Due to their high strength-to-weight ratio, carbon fibres are one of the most attractive reinforcements in the manufacture of advanced structural composites compared to conventional load bearing materials such as steels. These fibres are light in weight (low density) and have excellent tensile strength and stiffness, high fatigue strength, low coefficient of linear thermal expansion and lower susceptibility to corrosion. These favourable attributes make them attractive an material in automotive, aerospace, sporting goods and many other applications. Carbon fibres are produced by using three different precursor feedstock, such as rayon, polyacrylonitrile and pitch. Due to their aforementioned attributes, these fibres are used in composites reinforcements where high strength and stiffness fatigue resistance, high-temperature applications, and chemical inertness are important. Despite their many attractive attributes, these fibres suffer from low impact resistance due to their brittleness behaviour. Carbon fibres have high electrical conductivity, which in many cases, can be disadvantageous. Additionally, the scale of their firmness and strength depends on the production processes. In many cases due to their high cost, carbon fibres are not used in may common commercial applications; however, they are extensively used in the aerospace industry where lightweight material contributes to the overall weight savings, which is a vital point compared to cost.

1.3.1.3

Ceramic fibres

Ceramic fibres as crystalline or amorphous synthetic minerals, which are characterized by their refractory nature, they are obtained mainly in the form of whiskers. They are normally composed of silica and alumina oxides. Most ceramic fibres are polycrystalline or polycrystalline oxides and are generally white in colour. These fibres are used where high temperature resistance is required. The synthetic fibres discussed above are created from unsustainable fossil-based materials through energy-intensive processes; hence, these fibres give a high carbon footprint as their production processes are high-energy intensive. Moreover, these synthetic fibres provide limited recyclability and non-biodegradability that have become a growing concern when disposing of waste end-of-life products. Glass fibre-reinforced composites are used in various applications. Glass fibrereinforced composites have proven to meet the structural and durability demands of automobile interior and exterior parts, various components used in the marine industry, among others. Good mechanical properties and well-developed manufacturing bases have aided in the insertion of fibreglass-reinforced plastics within the automotive and marine industry. However, glass fibre-reinforced composites exhibit few shortcomings, such as their relatively high fibre density (approximately 40% higher than natural fibre), difficult to machine, and poor recycling properties, not to mention the potential health hazards posed by glass fibre particulates (Tables 1.5 and 1.6). The energy consumption to produce a flax fibre mat obtained by using life cycle assessment (LCA) suggests (9.55 MJ/kg), including cultivation, harvesting and fibre separation, amounts to approximately 17% of the energy to produce a glass-fibre mat (54.7 MJ/kg) (Joshi et al., 2004).

Introduction to composite materials

11

Table 1.5 Advantages and drawbacks of natural fibres as composite reinforcements (Faruk et al., 2012; Dhakal et al., 2007; Prasad and Sain, 2003). Advantages

Some disadvantages

• low density and high specific strength and stiffness • fibres are obtained from renewable resources, for which production requires less energy, involves CO2 absorption, while returning oxygen to the environment • fibres can be produced at a lower cost than synthetic fibre • low hazardous manufacturing processes • low emission of toxic fumes when subjected to heat and during incineration at the end-of-life • less abrasive damage to processing equipment compared with that for synthetic composites • lower risk to human health (no skin irritation)

• lower durability compared to their synthetic fibre composites but can be improved considerably with treatment • high moisture absorption, which results in swelling • lower strength, particularly impact strength compared to synthetic fibres composites • greater variability of properties (depending on geographical location, local growing conditions and weather) • lower processing temperatures limiting matrix options (low decomposition temperature) • lower thermal resistance than synthetic fibre • heterogeneous size • lack of standard • prone to fire hazard

Table 1.6 Comparison of physical and mechanical properties of metallic, conventional reinforcements (carbon and glass) and natural hemp fibres and flax (bundles) with E-glass (Bledzki et al., 1996; Faruk et al., 2012; Gurunathan et al., 2015).

Young’s modulus (GPa)

Specific strength (MPa)

Specific modulus (GPa)   E r

Fibre types

Density (g/cm3)

Tensile strength (MPa)

Steel

7.8

1300

200

167

26

Aluminium

2.81

350

73

124

26

Carbon

1.51

2500

151

1656

100

E-glass

2.10

1100

75

524

28

Aramid

1.32

1400

45

1656

100

Hemp

1.4

690

30e70

453

21e50

Flax

1.5

345e1830

27e80

230e1220

18e53

12

Sustainable Composites for Lightweight Applications

1.3.2

Natural fibres and their types

These fibres are obtained from a natural source, such as plants, animals and minerals. There are certain benefits that have been attributed to the use of natural fibres as reinforcement. However, there are certain aspects of natural fibres, which make them disadvantageous to use. The following sections highlight the advantages and some drawbacks of natural fibres as reinforcements. The focus of the book being sustainable lightweight composites, natural fibre will be looked at in a more comprehensive way. • • • •

Natural fibre can be classified according to their origin in the following groups: Plant fibres (bast, leaf, fruit, seed, wood, grass) Animal fibres (wool, hair, silk) Mineral fibres (ceramics, metal).

1.3.2.1

Advantages and disadvantages of natural fibres

Natural fibres offer many advantages by virtue of being natural. They have a high strength to weight ratio due to their lower density. Moreover, they have many green attributes such as biodegradable, recyclable and renewable. Similarly, they have tremendous processing benefits in terms of tool wear and energy requirements. Despite their attractive attributes, natural fibres also have some drawbacks in comparison with their synthetic counterparts, such as glass and carbon fibres, especially in terms of their durability, their strength and the moisture absorption. Fibre property (physical and morphological) availability is another major factor in selecting a natural fibre to use as reinforcement, as this is often attributed to geographical location and local growing conditions (Dhakal et al., 2020). Table 1.5 summarises the main benefits and some drawbacks.

1.4

Main drivers of composite materials

Fibre-reinforced plastics are referred to as (FRP) composites, usually with carbon, glass, aramid or natural fibres embedded in a polymer matrix. Other matrix materials can be used, and composites may also contain fillers or Nano-materials such as graphene and layered silicates. The many component materials and different processes that can be used to make composites extremely versatile and efficient. They typically result in lighter, stronger, more durable solutions compared to traditional materials. However, a major driving force behind the development of composites has been the combination of the reinforcement and the matrix, which can be tailored to meet the required final properties of a component (Wambua et al., 2003). Composite materials are lightweight materials. For example, the density of carbon fibre is far lower than that of metallic materials. Industries such as aerospace, automotive, marine and construction are seeking strong and durable lightweight materials. Another key benefit of using composite materials is their directional properties, which can be tailored in order to achieve high strength and modulus.

Introduction to composite materials

13

The key drivers can be listed as follows: • High anisotropic properties • Sustainability of materials • High fatigue properties.

The key drivers of using composite materials in substations, compared to their metallic counterparts in structural and non-structural applications, lie in the specific strength and modulus of their reinforcements. The key mechanical properties (strength, stiffness and strain to failure) of composite materials are governed by the properties of reinforcements, their morphology, geometry, volume fraction and alignment. Table 1.6 illustrates the mechanical properties of metal and synthetic reinforcements. As can be seen, the specific properties of carbon and glass fibres are far superior to that of metals. Steel and aluminium are well-established materials, and they still top the list of materials used in the automotive sector due to their well-defined damage characterisation ease to manufacture and lower cost of manufacturing. A closer look at the comparative values of physical and mechanical properties of metal, synthetic reinforcements carbon and glass and hemp and flax fibres are presented in Table 1.6. The results demonstrate that carbon and glass fibres are superior to even metals in terms of their specific properties. While considering natural fibre composite reinforcements, hemp and flax fibres are the most commonly used materials. As expected, the ultimate tensile strength is particularly higher for carbon and glass fibre than that of hemp and flax fibres. However, if one considers the specific modulus of carbon, glass and aramid fibres (modulus/density), they are far superior to that of steel and aluminium. This signifies the greater potential for weight reduction. This is one of the key drivers for the substituting of metallic materials with lightweight and stronger composite materials (Gurunathan et al., 2015).

1.5

Application of sustainable composite materials

For the last decade, natural fibre-reinforced sustainable composites have been used in different industry sectors, including transport (automotive, marine, rail, aviation), sports, building/construction and medical applications. The detailed application areas have been extensively discussed in Chapter 3.

1.6

Summary

Due to the increase in environmental awareness amongst the industries, as well as consumer pressure and government legislations, the urgency to develop alternative lightweight materials to substitute heavy metallic materials, has already taken place. As there is a continual progress in replacing metallic materials with glass and carbon fibre-reinforced composites, a similar approach needs to be adapted to replace carbon and glass fibre composites with more sustainable natural fibre-reinforced composites with a continuous drive for property, durability and quality improvement with the

14

Sustainable Composites for Lightweight Applications

enhanced end-of life-option. This will help the industry to reduce energy consumption arising from its materials and processes. If this is linked to the transport sectors, for example, this will then help in creating more fuel-efficient and lower emission vehicles. In order to realise the full potential of composite materials, it is important that from the design stage to raw materials extraction, production, use and endof-life stages are fully understood. This chapter has provided an overview of composite materials, their main characteristics and case for the need for sustainable lightweight composite materials for engineering applications.

References Armentia, S.L., Enciso, B., Mokry, G., Abenojar, J., Martinez, M.A., 2019. Novel application of a thermoplastic composite with improved matrix-fibre interface. J. Mater. Res. Technol. 8 (96), 5536e5547. Bhardwaj, S., 2017. Natural fibre composites:an opportunity for farmers. Int. J. Pure Appl. Biosci. 5, 509e514. Bledzki, A.K., Reihmane, S., Gassan, J., 1996. Properties and modification methods for vegetable fibers for natural fiber composites. J. Appl. Polym. Sci. 59, 1329e1336. Dhakal, H.N.P., Méner, E.L., Feldner, M., Jiang, C., Zhang, Z., 2020. Falling weight impact damage characterisation of flax and flax basalt vinyl ester hybrid composites. Polymers 12, 806. Dhakal, H.N., Sain, M., 2019. Enhancement of mechanical properties of flax-epoxy composite with carbon fibre hybridisation for lightweight applications. Materials 13 (1), 109. Dhakal, H.N., Skrifvars, M., Adekunle, A., Zhang, Z.Y., 2014. Falling weight impact response of jute/methacrylated soybean oil bio-composites under low velocity impact loading. Compos. Sci. Technol. 92, 134e141. Dhakal, H.N., Zhang, Z.Y., Guthrie, R., MacMullen, J., Bennett, N., 2013. Development of flax/ carbon fibre hybrid composites for enhanced properties. Carbohydr. Polym. 96 (1), 1e8. Dhakal, H.N., Zhang, Z.Y., Richardson, M.W., 2007. Effect of water absorption on the mechanical properties of hemp fibre reinforced unsaturated polyester composites. Compos. Sci. Technol. 67 (7e8), 1674e1683. Erbach, G., 2014. Reducing CO2 Emissions from New Cars. European Parliamentary Research Service. Faruk, O., Bledzki, A.K., Fink, H.P., Sain, M., 2012. Biocomposites reinforced with natural fibres: 2000e2010. Prog. Polym. Sci. 37 (11), 1552e1596. Gurunathan, T., Mohanty, S., Nayak, S.K., 2015. A review of the recent developments in biocomposites based on natural fibres and their application perspectives. Compos. Appl. Sci. Manuf. 77, 1e25. Huda, M.S., Drzal, L.T., Ray, D., Mohanty, A.K., Mishra, M., 2008. Natural-fiber composites in the automotive sector. In: Properties and Performance of Natural-Fibre Composites. Woodhead Publishing, Oxford, UK, ISBN 9781845692674. Hull, D., Clyne, T.W., 1996. In an Introduction to Composite Materials, (Cambridge Solid State Science Series. Cambridge University Press, Cambridge, pp. IeVI. https://doi.org/10.1017/ CBO9781139170130. Joshi, S.V., Drzal, L.T., Mohanty, A.K., Arora, S., 2004. Are Natural Fiber Composites Environmentally Superior to Glass Fiber Reinforced Composites? Compos. Part A Appl. Sci. Manuf. 35, 371e376.

Introduction to composite materials

15

Khan, M.Z.R., Srivastava, S.K., MK Gupta, M.K., 2018. Tensile and flexural properties of natural fiber reinforced polymer composites: a review. J. Reinforc. Plast. Compos. 37, 1435e1455. Mallick, P.K., 2008. Fiber-Reinforced Composites. United States of America: Taylor & Francis Group. Manaia, J.P., Manaia, A.T., Rodriges, L., 2019. Industrial hemp fibres: an overview. Fibres 7, 106. Matthews, F.L., Rawlings, R.D., 1994. Composite Materials: Engineering and Science. Chapman and Hall, London. Mohanty, A.K., Misra, M., Hinrichsen, G., 2003. Biofibres, biodegradable polymers and biocomposites: an overview. Macromol. Mater. Eng. 276e277 (1), 1e24. Monteiro, S.N., Satyanarayana, K.G., Ferreira, A.S., 2010. Selection of high strength natural fibres. Matéria 15 (4), 488e505. Nordin, N.A., Yussof, F.M., Kasolang, S., Salleh, Z., Ahmed, A.M., 2013. Wear rate of natural fibre:long kenaf composite. Procedia Eng. 68, 145e151. Oksman, K., Skrifvars, M., Selin, J.F., 2003. Natural fibres as reinforcement in polylactic acid (PLA) composites. Compos. Sci. Technol. 63, 1317e1324. Paturel, A., Dhakal, H.N., 2020. Influence of water absorption on the low velocity falling weight impact damage behaviour of flax/glass reinforced vinyl ester hybrid composites. Molecules 25 (2), 1e16, 278. Prasad, B.M., Sain, M.M., 2003. Mechanical properties of thermally treated hemptreated hemp fibres in inert atmosphere for potential composite reinforcement. Materials Research and Innovation 7, 231e238. Silvestre, J., Silvestre, N., de Brito, J., 2015. Review article an overview on the improvement of mechanical properties of ceramics nanocomposites. Hindawi Publishing Corporation J. Nanomat. https://doi.org/10.1155/2015/106494. Article ID 106494. Stewart, R., 2011. Thermoplastic composites e recyclable and fast to process. Reinf. Plast. 55, 22e28. Wambua, P.W., Ivens, J., Verpoest, I., 2003. Natural fibres: can they replace glass in fibre reinforced plastics? Compos. Sci. Technol. 63, 1259e1264.

Further reading Agarwal, J., Sahoo, S., Mohanty, S., Nayak, S.K., 2019. Progress of novel techniques for lightweight automobile applications through innovative eco-friendly composite materials: a review. J. Thermoplast. Compos. Mater. 63, 1e36. Asdrubali, F., D’Alessandro, F., Schiavoni, S., 2015. A review of unconventional sustainable building insulation materials. Sustain. Mater. Technol. 4, 1e17. Bourmaud, A., Beaugrand, J., Shah, D.U., Placet, V., Baley, C., 2018. Towards the design of high-performance plant fibre composites. Prog. Mater. Sci. 97, 347e408. Dhakal, H.N., MacMullen, J., Zhang, Z.Y., 2016. Moisture measurement and effects on properties of marine composites. In: Marine Applications of Advanced Fibre-Reinforced Composites. Woodhead Publishing, Sawston, UK; Cambridge, UK, pp. 103e124. Dhakal, H.N., Zhang, Z.Y., Bennett, N., 2012. Influence of fibre treatment and glass fibre hybridisation on thermal degradation and surface energy characteristics of hemp/unsaturated polyester composites. Compos. B Eng. 43 (7), 2757e2761.

16

Sustainable Composites for Lightweight Applications

Feraboli, P., Gasco, F., Wade, B., Maier, S., Kwan, R., Masini, A., Oto, L.D., Reggiani, M., 2011. Lamborghini “forged composites” technology for the suspension arms of the Sesto Elemento. In: American Society of Composites. 2011: Montreal. Hitchen, S.A., Kemp, R.M.J., 1996. Development of novel cost effective hybrid ply carbon-fibre composites. Compos. Sci. Technol. 56 (9), 1047e1054. Li, Y.X., Choy, X.J., Guo, C.L., Zhang, Z., 1999. Compressive and flexural behavior of ultrahigh-modulus polyethylene fibre and carbon fibre hybrid composites. Compos. Sci. Technol. 59 (1), 13e18. Madurwar, M.V., Ralegaonkar, R.V., Mandavgane, S.A., 2013. Application of agro-waste for sustainable construction materials: a review. Construct. Build. Mater. 38, 872e878. Mussig, J. (Ed.), 2010. Industrial Applications of Natural Fibres: Structure, Properties and Technical Applications. John Wiley and Sons, Chichester. Mustafa, N.S., Omer, M.A.A., Garlnabi, M.E.M., Ismail, H.A., 2016. Reviewing of general polymer types, properties and application in medical field. Int. J. Sci. Res. 5 (8), 212e221. Pervaiz, M., Panthapulakkal, S., Birat, K.C., Sain, M., Tjong, J., 2016. Emerging trends in automotive lightweighting through novel composite materials. Mater. Sci. Appl. 7 (1), 26e38. Peças, P., Carvalho, H., Hafiz Salman, H., Leite, M., 2018. Natural fibre composites and their applications: a review. J. Compos. Sci. 2, 66. https://doi.org/10.3390/jcs2040066. Sarbu, A., Dima, S.O., Dobre, T., Udrea, I., Bradu, C., Avramescu, S., Mihalache, N., Radu, A.L., Nicolescu, T.V., Lungu, A., 2009. Polystyrene wastes recycling by lightweight concrete production. Rev. Chem. 60, 1350e1356. Yan, L., Chouw, N., Jayaraman, K., 2014. Flax fibre and its compositesda review. Compos. B Eng. 56, 296e317. Yan, L., Kasal, B., Huang, L., 2016. A review of recent research on the use of cellulosic fibres, their fibre fabric reinforced cementitious, geo-polymer and polymer composites in civil engineering. Compos. B Eng. 92, 94e132.

Sustainable natural fibre reinforcements and their morphological structures

2

2.1 Commonly used sustainable materials (plant-based natural fibres reinforcements in composites) Due to the environmental concerns, legislation for low emission materials, and consumer awareness towards the sustainable development aspirations, lightweight composite materials reinforced with plant fibres have been considered as an alternative to conventional fibre reinforcing materials (Bledzki and Gassan, 1999). The classification of natural fibres is illustrated in Fig. 2.1. For making composite materials sustainable and environmentally friendly, both matrices and reinforcements are expected to be originated from renewable resources. Natural fibres available for composite reinforcement can be classified into three main types (Gurunathan et al., 2015): 1. Lignocellulosic plant-based (hemp, flax, jute sisal, palm fibre, kenaf, date palm, etc.) 2. Animal-based (silk, wool and hair) 3. Mineral-derived (asbestos, wollastonite).

Lignocellulosic plant-based natural fibres can be further classified into: Bast fibres include flax, hemp, jute, kenaf and ramie. Leaf fibres comprising sisal, abaca, pineapple and henequen. Seed fibres, which are obtained from seeds, and a good example of such fibre is cotton and kapok. Just like plant-based fibres, natural fibres are further grouped as to their source. Stalk fibre include fibres from plant stalks such as rice, wheat, barley, typically extracted from plants. Grass and other fibre residue includes bamboo, bagasse, corn, widely available from tall grasses. Animal fibres include fibres that are obtained from hairy mammals such as sheep, which is the source of wool. Silk fibres, on the other hand, are obtained from insects such as spiders, which produce silk when making their cobwebs (Faruk et al., 2012). These reinforcements have outstanding properties such as abundant, biodegradable and recoverable after their end-of-life and in many cases, better specific properties (strength and stiffness) compared to their synthetic counterparts (Satyanarayana et al., 2009). Due to these several benefits, many European countries have already using biobased composites in automotive applications. Nonetheless, despite these, most of the polymers used to use in these composites are non-biodegradable. Especially when higher mechanical properties are required, matrices such as unsaturated polyester, epoxy, vinyl ester, polypropylene are used as they provide higher mechanical properties in comparison to biodegradable polymers, but these polymers are not fully biobased (Nayak et al., 2000).

Sustainable Composites for Lightweight Applications. https://doi.org/10.1016/B978-0-12-818316-8.00004-9 Copyright © 2021 Elsevier Ltd. All rights reserved.

18

Sustainable Composites for Lightweight Applications

Natural fibers

Natural

Animal

Synthetic

Mineral Asbestos

Silk

Inorg. fiber

Org. fiber

Glass

Aramid/kavlar

Wool

Aromatic polyester

Hair

Polyethylene

Carbon Boron Silicacarbide

Cellulose/lignocellulose

Leaf

Seed

Fruit

Wood

Stalk

Flax

Sisal

Kapok

Coir

Bamboo

Hemp

Cotton

Oil palm

Soft wood

Rice

Banana

Wheat

Bagasse

Jute

Abaca

Loof ah

Barley

Corn

Ramire

PALF

Sabai

Henequen

Milk weed

Maize

Mesta

Oat

Rape

Rye

Esparto

Bast

Kenaf

Agave

Roselle

Raphia

Hard wood

Grass/reeds

Canary

Figure 2.1 Classification of natural fibres (Gurunathan et al., 2015).

The European Union’s directive requiring that all new vehicles needing to use 95% recyclable materials to achieve the end-of-life of vehicle by 2015 has led further motivation for the development of commercially viable lightweight composites. Under this directive, about 85% of these materials must be recoverable through re-use or mechanical recycling and about 10% through energy recovery or thermal recycling. Similar
GHGs regulation in North America, and directives are in place, for example, CAFE, in Europe, regulation 2020. For example, in Euro 6 regulation from 2020, a higher tax will be applied to vehicles exceeding 95 g/km of CO2 emission. In order to meet this target, original equipment manufacturers (OEMs) are undertaking various measures towards the utilisation of lightweight concepts as the weight reduction has direct benefits towards the reduction of CO2 emission (Pervaiz et al., 2016;Directive 2000/53/EC, n.d.). Despite these many attractive attributes, natural plant fibres are still not fully utilised to their full potential. The natural plant fibre reinforced composites have been used mainly for non-structural applications in automotive, construction and packaging industries due to their inherent moisture absorption behaviour resulting from their chemical compositions (cellulose, hemicelluloses, lignin and pectin) and natural variability (Dhakal et al., 2007). With this phenomenon, even with higher fibre contents, obtaining high or full potential strength and stiffness is difficult to achieve for natural

Sustainable natural fibre reinforcements and their morphological structures

19

fibre reinforced composites due to the fact that the moisture absorption behaviour of these fibres causes weak fibre matrix adhesion leading to ineffective stress transfer from matrix to fibres during various loading scenarios. To minimise this problem, many researchers have already undertaken various fibre surface modification processes, together with improved manufacturing processes (Nayak et al., 2000). It is worth noting that in addition to lack of compatibility between hydrophobic polymers and hydrophilic natural fibres leading to weak mechanical properties, fibre availability is a major factor in selecting natural fibres as reinforcement, often constrained geographically due to local growing conditions, as well as lack of established supply chain mechanisms. The following sub-sections will elaborate on the most commonly used and some emerging plant-based natural fibres and their main characteristics.

2.1.1 Hemp fibres Hemp (Cannabis Sativa L.) is one of the oldest, strongest and stiffest available natural fibre. The chemical constituents of mature hemp fibre are cellulose (74%), hemicellulose (18%), lignin (4%) and pectin (1%). Hemp is an annual plant, tall (2e5 m), robust, annual herbaceous plant, which is sown in spring and harvested in autumn. Hemp has been used as a basic raw material for the production of rope, traditional medicine, canvas and clothing for many years (Carus et al., 2013). Hemp is considered to be a multipurpose crop since it offers a source of reinforcing fibre in composites, oil and molecules (Andre et al., 2016). In many countries, the production of hemp fibre is limited due to the ban as this fibre contains narcotics. The flowering tops, as well as to some extent, the leaves of hemp produce resin secretions containing the narcotic 9-tetrahydrocannabinol (THC) for which marijuana and hashish are famous. However, the industrial hemp produces less than 0.3% THC, which cannot be used as a narcotic (Weiblen et al., 2015). For the last few decades, the industrial hemp has been one of the major sustainable reinforcements in sustainable lightweight composite materials. Just before 2000, hemp plants were common agriculture crops grown in the moderate climates for the production of ropes and shipping sails. Hemp and the flax are only commercial sources of long natural fibres grown in the UK. Originally native to Central Asia, it has since spread to every inhabited continent, region and country and widely cultivated in Europe (Richardson et al., 1998; Hepworth et al., 2000). For the last decade, hemp fibre has been very popular due to its versatile features: rage of food, strong fibre, and attractive agriculture features resistance to drought and pests, prevention of soil erosion, lower fertiliser, herbicides and pesticides and water requirement to grow in comparison to other plant fibres (Andre et al., 2016). The production and use of fertilisers and herbicides consume a significant amount of energy, and is hence, a very big burden to the environment. From the economic and environmental point of view, this natural fibre offers a significant opportunity to be used as a sustainable resource. Due to their high stiffness and strength, a primary requirement for the reinforcement of composite materials, these fibres are increasingly being used in composites reinforcements. The specific strength and stiffness that are comparable to those of glass fibres make these fibres attractive to the automotive and construction industry for the production of non-structural components.

20

Sustainable Composites for Lightweight Applications

From both environmental and performance point of view, hemp offers excellent attractions as reinforcing materials in composites, especially for automotive applications (Karus and Kaup, 2002). The major disadvantage, however, is its low impact strength when compared to glass fibre reinforced plastic composites (Dhakal et al., 2012). Fig. 2.2 depicts a hemp plant, non-woven hemp mat and scanning electron microscopy (SEM) macrograph of hemp fibre.

2.1.2 Flax fibres Flax (Linum usitatissimum L.) is an annual plant, which grows from 0.5 to 1.5 m tall. Flax fibre is obtained from the stem of the flax plant. As most bast fibres, flax fibres are constituted of an outer wall of pectin, cellulose, hemicellulose and lignin. The cellulose content in flax fibres is approximately 71%, hemicellulose 18.6%e20.6%, lignin 2.2% and 1.2% wax. Flax is bast fibre extensively used in the textile industry, which possesses a smooth and straight texture (Dhakal et al., 2013). The mechanical properties such as strength and stiffness of flax fibres are comparatively higher than other natural plant fibres due to their high molecular weight and crystallinity of cellulose. The properties and the quality of flax fibres depend on the growing and harvesting conditions (retting and decorticating for specific uses), soil quality, use of fertiliser and the part of the stem where the fibres are extracted (position within the stem). Physical characteristics such as diameter, microfibrillar angle (as this fibril angle decreases, the mechanical properties increase), chemical compositions, as well as how the fibres were processed as it is well accepted that during the process, the fibres are damaged at varying degrees (Faruk et al., 2012; Zini and Scandola, 2011). Flax technical fibres (consisting of many elementary fibres bonded together) are seen as competitors to glass fibres for many applications because of their following key principal advantages: (1) (2) (3) (4)

Cheaper than glass fibres, Non-abrasive, biodegradable and less toxic than glass fibres, Higher specific strength and modulus than glass fibres, Less energy required to produce compared to glass fibres.

Due to their high strength and stiffness together with outstanding damping properties, this fibre has been used in many applications, including the automotive, marine and textile industries (Bos et al., 2002).

Figure 2.2 Hemp fibre (a) hemp plant, (b) non-woven hemp mat and (c) SEM of hemp fibre.

Sustainable natural fibre reinforcements and their morphological structures

21

Flax fibres, as most of the plant fibres, are heterogeneous in their physical structure and chemical compositions, which poses a challenge when using these fibres for reinforcing composites for structural uses. Fig. 2.3 illustrates flax plant, non-woven flax mat and SEM images of flax fibre.

2.1.3 Jute fibres Jute (Genus Corchorus oliotorius), one of the most common bast fibres, acquired from the bark of jute plant is produced worldwide, which grows from 2 to 3.5 m tall, and fibres are extracted after harvesting. Bangladesh is one of the largest jute-producing nations, where jute is also known as the golden fibre of Bangladesh. Other countries like India, China, Myanmar and Nepal also produce high quantities of jute fibres. The main chemical constituents of jute fibre include cellulose (58%e65%), hemicellulose (20%e24%) and lignin (12%e15%). The high lignin contents make the fibre just brittle with lower extension to break in comparison to other bast fibres such as hemp. The low cost and non-abrasive nature of the jute fibres allow one of the most commonly used reinforcing alternatives, thereby resulting in significant cost savings in composite manufacture. The tabular (cellular) structure of the natural fibre (jute) provides good insulation against heat and noise. Jute, like many other natural plant fibres exhibit good mechanical properties (Faruk et al., 2014). As a result, they are attractive reinforcements in the manufacture of composites in the construction and automotive industries (Dhakal et al., 2014; Bourmaud et al., 2018; Sinha and Panigrahi, 2009). As expected from other natural plant fibres, jute fibre reinforced composites can provide low production energy and cost, good specific mechanical properties compared to conventional glass fibre reinforced composites. Fig. 2.4 illustrates the Jute plant and SEM image of jute fibre.

2.1.4 Kenaf fibres Kenaf belongs to the Hibiscus cannabinus L. family and is a herbaceous annual plant that can be grown under a wide range of weather conditions. This plant is easy to grow without much attention. Kenaf is a bast fibre and environmentally friendly biodegradable crop (Ramesh et al., 2018; Nishino et al., 2003). It is reported that this fibre grows

Figure 2.3 Flax fibre (a) mature flax plant, (b) non-woven flax mat and (c) SEM of flax fibre showing non-uniform diameter.

22

Sustainable Composites for Lightweight Applications

Figure 2.4 Jute fibre, depicting (a) unidirectional jute fibre (b) jute yarn (c) SEM of untreated jute fibre (Aziz and Ansell, 2004).

to more than 3 m within 3 months, even in moderate ambient conditions. When natural fibres are used in structural/non-structural applications, the kenaf fibres reinforced composites have been very popular ones, especially in the automotive industry. However, due to their hydrophilic nature, they require prior treatment (Ramesh et al., 2018). These fibres also have been used for many decades as a rope, sacking and canvas. Fig. 2.5 depicts the kenaf elementary fibre and SEM images of treated and untreated kenaf fibre.

2.1.4.1 Advantages of kenaf fibres Kenaf fibre absorbs nitrogen and phosphorus included in the soil, and it accumulates carbon dioxide at a significantly high rate (Nishino et al., 2003). Like other natural fibres, kenaf fibre has many usual advantages that commonly used natural fibres have, which are: • • • •

They form a cellulosic source with ecological and economic benefits Low cost, low density, high specific properties (strength and stiffness) Kenaf fibres are fully biodegradable and are used in making sustainable biobased composites Extensively used as woven and non-woven mats in automotive industries

2.1.5 Date palm fibres Date palm fibres are derived from the date palm tree (Fig. 2.6). The date palm tree (Phoenix dactylifera L.) is a member of the palm tree family (Arecaceae), from this family derive more than 200 types of palm trees. Currently, it is estimated that there are more than 100 million date palm trees worldwide. This is one of the most important crops in North Africa and the Middle East. They are predominantly found in countries such as Saudi Arabia, Egypt, Iraq and Iran (Alawar et al., 2009). These fibres contribute significantly to the economy of these regions. The date palm fibres are emerging natural plant fibres. Although this fibre is cultivated worldwide for many years, these fibres are getting significant priorities in composite reinforcements in recent years due to their better water repellent properties and good thermal stability compared to other natural plant fibres. The diameter ranges from

Sustainable natural fibre reinforcements and their morphological structures

23

(a)

Nano scale microfibril Ø 5–10 nm Micro scale elementry fibre Ø 10–30 Pm

Cellulose

Micro scale fibre bundle Ø 50–100 Pm Kenaf pith Kenaf core Kenaf bast

Kenaf stem Ø 20–40 mm Kenaf tree

(b)

(c)

Figure 2.5 Kenaf fibre (a) microscale elementary fibre bundles (b) SEM image of un-treated kenaf fibre (c) SEM of treated kenaf fibre (Khalil et al., 2012; Yousif et al., 2012).

24

Sustainable Composites for Lightweight Applications

Figure 2.6 Date palm fibre (a) date palm fibre location of the sheath, (b) sheath (c) processed raw date palm fibre (d) SEM micrographs of date palm single fibre.

100 to 1000 mm and a measured density of 0.917
0.127 g/cm3. The surface of date palm fibres is rough compared to other natural plant fibres. As can be noticed (Fig. 2.6), the outer surface shows some impurities and residues. The main features of the date palm fibres are: • Lower density (0.92 g/cm3) compared to other natural fibres. • Low moisture absorption (5%) compared to other common natural fibres: jute (12%), flax (10%) and sisal (11%).

Date palm fibres are extensively used for manufacturing rugs, huts and shades. Due to its high cellulose (46%) contents, these fibres possess good mechanical properties and low water absorption. Consequently, these fibres are weaker compared to flax fibres. After the harvesting of the date palm, large quantities of residues in the form of leaves and fronds are wasted each year. The by-products made from these wastes are mainly low-value products. Often they are burned in the agricultural land causing environmental and health hazards. The part of the date palm tree, which is often used as fibres is the leaf sheath. The sheath is the part of the tree, which surrounds the trunk of the plant attached to its lateral edges near the top of the trunk Fig. 2.6. The sheath is also known under the name of leaf and is often torn lose when pruning the leaves. Due to lower moisture absorption, these fibres have been successfully used as reinforcements in polymeric composites in recent years (Jawaid and Abdul Khalil, 2011).

Sustainable natural fibre reinforcements and their morphological structures

25

2.1.6 Sisal fibres As common attributes of other commonly used natural fibres, sisal fibres (Agave sisalana), extracted from the leaves of the sisal plant, have the promising reinforcing potential for use in composite materials due to its low cost, low density, high specific strength and modulus, no health risk, easy availability in some countries and renewability. Sisal fibre has a high percentage of the lumen, which plays an important role in the overall properties of resultant composites. The main chemical constituents of sisal fibre include cellulose (70%), hemicellulose (12%), lignin (9%), Pectin (10%) and waxes (2%). The microstructure of sisal fibre is illustrated in Fig. 2.7 (Cheung et al., 2009; Nishino et al., 2003). Due to the large hollow lumen and porous microstructure, sisal fibres can be used for acoustic and thermal insulation applications. However, the large hollow section on sisal fibres can weaken the mechanical properties as the new cross-section area decreases, and the overall stress concentration increases.

Figure 2.7 (a) Sisal plant (b) sisal fibre (c) SEM image of sisal fibre showing hollow structure (Cheung et al., 2009; Nishino et al., 2003).

26

Sustainable Composites for Lightweight Applications

2.1.7 Oil palm fibres The oil palm is one of the fibre materials belonging to the species Elaeis guineensis under the family Palmacea, and originated in the tropical forests of West Africa. Oil palm empty fruit bunch (OPEFB) and oil palm mesocarp fibres are two main fibrous materials obtained from palm oil mill (Jawaid et al., 2013). Empty fruit bunch fibre is one of the largest agricultural wastes biomass generated in countries like Nigeria, Malaysia and India. These fibres are the by-product of factories that process oil palm. The chemical constituents of OPEFB fibre are cellulose (65%), hemicellulose (24.2%), lignin (19%) and other extractives (8.9%) (Suksong et al., 2016; Sreekala and Thomas, 2003). These fibres have a lower density (0.7e1.55 g/cm3) than other natural fibres, which makes them attractive in lightweight composites as reinforcements. A large amount of lignocellulose materials such as oil palm fronds, trunks and empty fruit bunches generation leads to an enormous amount of empty fruit bunch (EFB) fibres. These fibres if not used properly, pose a serious environmental threat. The empty fruit bunches are mainly incinerated to produce bunch ash to be distributed back to the field as fertiliser. Approximately 15 million tons of this agriculture waste is generated by oil palm milling operation annually and part of it is burned in incinerators. The conventional method of burning these residues often creates environmental problems, in that it generates severe air pollution and is prohibited by the environment protection act. In abiding by these regulations, these residues are becoming expensive to dispose off (Izani et al., 2013; Rahman et al., 2007). Fig. 2.8 depicts oil palm empty fruit bunch and non-woven oil palm fibre mat for composite reinforcement. However, converting this waste into a useful product could save the environment from hazardous pollution. Currently, an extensive research provides an alternative way of optimising the usage of fibre obtained from oil palm into value added products. OPEFB is a form of fibrous lignocellulosic residue has significant potential as reinforcing materials in composites, Reinforcements in composite materials could be a way forward to utilise these waste agriculture biomasses. The EFB fibres possess some attractive properties such as good tensile strength and stiffness, good water repellence

Figure 2.8 Oil palm fibre from empty fruit bunch (a) fibre showing non-uniform diameter, (b) selected finer fibres and (c) non-woven oil palm fibre mat.

Sustainable natural fibre reinforcements and their morphological structures

27

behaviour after the chemical treatment of fibres. The work carried out by (Sreekala and Thomas, 2003) reported that the tensile strength and Young’s modulus of salinetreated OPEFB fibres were recorded at 273 MPa and 5.25 GPa, respectively. With these attractive mechanical properties, these fibrous wastes have been significantly utilised as reinforcements to produce biobased composites for various engineering applications in recent years (Sreekala and Thomas, 2003).

2.1.8 Banana fibres The banana fibres are extracted from the stem of the banana plant, which is the after product or often termed as a waste product of banana cultivation. These fibres can be normally obtained without incurring any additional costs of production for industrial uses (Joseph et al., 2002). The chemical constituents of banana fibre are cellulose (63%e64%), hemicellulose (19%), lignin (5%) and the moisture absorption is approximately 10%e11% (Seena et al., 2002). These fibres have already been used as reinforcements in composites and have shown comparable mechanical properties against other natural fibres. Fig. 2.9 shows raw banana fibres, SEM images of banana fibre reinforced composites and a cylindrical component made from banana-based epoxy composites.

2.2 Influence of processing and chemical composition on the properties Natural plant fibres such as hemp, flax, jute and kenaf need to be separated from their barks. The fibres from these plants are typically extracted by retting, followed by mechanical processing such as scotching and hackling (Sultana, 1992). The processes used to manufacture conventional reinforcements such as carbon and glass are inherently different from those used in the production of natural fibre as composite reinforcements. As far as mechanical properties of commonly used natural fibres such as hemp, flax, kenaf and jute, for example, are concerned, their properties are relatively high but transforming these high-end mechanical properties to the resulting composites

Figure 2.9 (a) Raw banana fibres (b) SEM images of banana fibre reinforced epoxy composites fibres (c) banana fibre reinforced epoxy composite cylinder (Mohan and Kanny, 2019).

28

Sustainable Composites for Lightweight Applications

have been the challenge as fibre processing steps influences or affects the final properties of fibres. The fibre processing techniques introduce defects onto fibres at both micro and meso-levels. These defects created on the fibre surface, eventually degrade the mechanical properties of natural fibre reinforced composites (Gager et al., 2019). In the process of separation, different processes such as retting and scotching are utilised. These processes have significant effects on the quality of fibres, as well as in the chemical composition of these fibres (Merotte et al., 2016). Defects on the fibre surfaces as depicted in Fig. 2.10 can easily be introduced during these processes. These defects can significantly affect the overall mechanical and thermal properties of these fibres. The overall properties, therefore, depend on natural variability, as well as damage sustained during their processing.

2.2.1 Importance of fibre processing parameters The plants are taken out directly from the ground to harvest fibres such as hemp, flax and jute; this is done so as to retain the longest fibre length. The flower heads of the flax plant, for example, are then removed by rippling. After this, the plants are spread on the ground for retting. During this stage, the pectin layer, which binds the fibres to the bast tissue and stem, is removed. Retting is carried out by normally laying the fibres on the ground for around 3e7 weeks (dew retting). The retting process can weaken the middle lamella by the action of micro-organisms. The retting process is followed by breaking, scorching and hackling. In the breaking process, the fibre stems are passed between the rollers whereas the scorching separates the fibre bundles from the xylem. Hackling thins the fibre bundles by passing them through a series of combs (Zhang et al., 2005; Mohanty et al., 2001). The coarse fibre bundles are then combed into the hackling process. During this process, the ribbon-shaped fibre is refined towards a circular fibre structure, although recently, a few more fibre isolation methods have emerged. The dew retting process described is still the traditional and most common method used (Rowell et al., 2000).

Figure 2.10 SEM images of the surface of flax fibre showing complex fibre layers and defects (a) untreated (b) treated.

Sustainable natural fibre reinforcements and their morphological structures

29

Retting is controlled degradation of plant stems to allow the fibre to be separated from the woody core. The traditional method to separate the fibres from the plant is to cut and leave the stems on the field, where they are soaked during the night by the dew allowing natural bacterial degradation to take place (Kessler et al., 1998). Under these conditions, microorganisms grow and produce enzymes, which degrade the pectic substances, and the cortex fibres are progressively disassociated into fibre bundles and sub-bundles. This method is known as dew retting, is currently in practice, but the quality of fibres may vary due to variations in the climatic conditions. Alternatively, the plant stems are retted in water tanks to make the retting process more controlled. Other approaches include drying of stems artificially subjected to multiple passes in a system of rollers (Hobson et al., 2001). This can be an expensive process and can cause too much fibre damage (Zafeiropoulos et al., 2007; Zafeiropoulos and Baillie, 2007). The processing affects the final fibre aspect ratio, and thus, the mechanical properties of the product. Processing techniques such as internal mixing and injection moulding can cause high-fibre attrition. Fibres can be broken into smaller and shorter pieces due to various mechanisms: • Fibre-fibre interaction can also cause fibres to break; • Fibre-matrix interaction owing to the shear stresses acting in the viscous polymer; • Fibre wall interaction

Natural fibres have weak links as natural and artificial flaws, as well as kink bands. These weak links are the most probable rupture points along the fibre length when the fibre is mechanically stressed. When the natural plant fibres are obtained by using the above-mentioned process, the fibres go through some kind of disruption and damage. This phenomenon is described as micro-compression and kind bands. In fact, natural plant fibres experience extreme weather conditions such as wind, rain and these conditions exert some sort of forces. These conditions create some sort of damage. Moreover, the damages are worsened by various fibre-processing techniques used. Additionally, the manufacturing processes used to make the composites will further damage the fibres and their lengths, reducing the overall mechanical properties. The damage created during these processes, as well as inherent structural defects such as kink bands, lead to an overall reduction of properties (Oksman, 2001). A modern method such as steam-explosion and ultrasonic treatments to separate the main components of lignocellulosic biomass (cellulose, hemicelluloses, pectins and lignin), has been introduced as an alternative to the conventional process. However, there is an argument that these methods separate the material into its component fibre cells, and hence, destroying the structure of fibre bundles that the plant has provided. Also, these methods can be expensive. It seems that the only economical method currently available to separate fibres from stems is field retting despite the process having several drawbacks (Satyanarayana et al., 2007; Joffe et al., 2003). The fibre bundles extend continuously from bottom to top of the hemp plant; however, the single fibres are smaller units with lengths in the range of 5e55 mm (Vincent, 2000; Ranalli and Venturi, 2004). In taking composite materials as a whole, there are

30

Sustainable Composites for Lightweight Applications

many different material options to choose from in the areas of resins, fibres and cores, all with their own unique set of properties such as strength, stiffness, toughness, heat resistance, cost, production rate, etc. However, the end properties of a composite part produced from these different materials are not only a function of the individual properties of the resin matrix and fibre (and in sandwich structures, the core as well), but is also a function of the way in which the materials themselves are designed into the part and also the way in which they are processed. Natural plant fibres have high moisture absorption, and the low processing temperatures permissible. The processing temperature of the lignocellulose plant fibres is limited due to the potential fibre degradation at high temperatures. The polymer matrices that can be used are limited to low melting temperature plastics.

2.2.2 Chemical composition and their influences on the properties The major constituents (chemical compositions) of natural plant fibres include cellulose, hemicellulose, lignin, pectin, fat, waxes and water-soluble substances. Chemical composition plays an important role in the properties of natural fibres. The average chemical composition of commonly used natural fibres is illustrated in Table 2.1. When there is higher cellulose content, higher mechanical properties are achieved. Natural fibres themselves are considered as composites. As can be observed in Table 2.1 that coir fibres have the lowest cellulose contents and show the lowest tensile strength among all other natural fibres. Lower cellulose content has contributed to the lower strength and modulus. Natural fibres are mostly constituted of cellulose, a biopolymer of the plant sugar glucose. Other constituents present in natural fibres Table 2.1 Chemical composition and equilibrium moisture contents of selected natural plant fibres (Li et al., 2007; Rosa et al., 2010; Sukumaran et al., 2001; Sreekala and Thomas, 2003; Suksong et al., 2016). Fibre

Cellulose

Hemicellulose

Lignin

Wax (wt.%)

Moisture (wt.%)

Date palm fibre

46

e

20

e

e

Flax

71

18.6e20.6

2.2

1.5

7

Hemp

68

15

10

0.8

9

Jute

61e71

14e20

12e13

0.5

12

Bamboo

26e43

30

21e31

e

Kenaf

72

20.3

9

e

e

Cotton

82.7

5.7

e

e

e

Sisal

65

12

9.9

2

11

Oil palm fibre

65

24.2

19

e

e

8.9

Sustainable natural fibre reinforcements and their morphological structures

31

are hemicelluloses, pectin, lignin and waxes; flax fibres when considered for example, exhibit highest strength and modulus, which is attributed to the higher cellulose contents. Similarly, hemicellulose, pectin and lignin contents (percentages and interaction between them) play a significant role on the mechanical and thermal properties of natural plant fibres (Rouison et al., 2006). Chemical structures of cellulose, hemicellulose and lignin are depicted in Fig. 2.11. Similarly, due to their major chemical constituents, natural plant fibres such as flax, hemp, jute and bamboo absorb water. The moisture absorption is influenced by the humidity and duration they are exposed to the hygrothermal environments (temperature and humidity). The water absorption causes thickness swelling of fibres and causes a significant strength reduction compared to the dry specimens. Moreover, in order to attain higher mechanical strength of composites, reinforcing with higher fibre volume fraction is one of the ways that is generally employed. However, in the case of natural fibre reinforced composites, normally, higher the fibre volume fraction, greater the moisture absorption at wet environments, which induces significant weakness at the interface and reduces the mechanical properties significantly. Additionally, the chemical composition of natural plant fibres limits the temperatures at which they can be processed. Many authors have reported that this limits the use of these fibres as reinforcement in thermoplastic composites as these fibres start degrading at a rapid rate at approximately above 200e240 C with the degradation of hemicellulose (Bledzki et al., 2002).

(a)

(c)

O HO

OCH3

OH

OH

HO O

O

O

O H2C C OH

O OH

OH

CH2 n OCH3 O

(b)

H H

O H

H

H

O

O O

O H HO

H

OH H

CH3 C O

H

OH

H

H

O H

Figure 2.11 Chemical structure of (a) cellulose (b) hemicelluloses and (c) lignin.

32

Sustainable Composites for Lightweight Applications

2.2.3 Cellulose structure Cellulose are made up of glucose polymeric structures which are repeated on the lattice. Usually, a single cellulose molecule may comprise of a few thousand glucose repetitions. The way glucose structures are connected together has a profound effect on the properties of the cellulose. Therefore, a wide range of properties can be expected from different cellulose-based fibres. One of the biggest limitations of cellulose-based fibres is that they carry a large density of voids due to the particular glucose structure. A large number of voids become water or moisture storing sites when the composite is immersed in liquid. Therefore, the moisture absorption is a great concern for cellulose-based composites. Various types of surface treatments may help in improving this feature.

2.2.3.1 Cellulose Cellulose is the basic structural component of all plant fibres. The molecular structure of cellulose is shown in Fig. 2.12. It is the most important organic compound produced by plants and the most abundant in the biosphere. The primary or outer wall of a cell consists of a thin network of cellulose microfibrils, irregularly and loosely arranged and incrusted with hemicellulose, lignin and other compounds. Cellulose is the most abundant natural polymer in the world and the most essential component of all plant fibres and found to be at a higher percentage than other constituents in natural plant fibres. It is an isotactic B-1, 4-polyacetal of cellulose. The basic unit, cellulose, is composed of two molecules of glucose. As a result, cellulose is often called a polyacetal of glucose (Eichhorn and Young, 2004). Cellulose is a semi-crystalline polysaccharide with a large amount of hydroxyl group present making natural fibres hydrophilic, and as a result, poor adhesion between fibre and matrices. This is one of the concerns of natural fibres when reinforced with hydrophobic polymers, making less compatible between them.

2.2.3.2 Hemicellulose Hemicelluloses are also found in all plant fibres. Hemicelluloses are polysaccharides bonded together in relatively short, branching chains. The hemicellulose fraction of the plant consists of a collection of polysaccharide polymers. Hemicelluloses usually consist of more than one type of sugar unit (Bledzki et al., 2002). Hemicellulosic polymers are branched, fully amorphous and have a lower molecular weight than cellulose. For example, in flax fibres, their open structure and availability of polar groups make hemicellulose and pectin susceptible to chemical degradation.

Figure 2.12 Molecular structure of cellulose.

Sustainable natural fibre reinforcements and their morphological structures

33

2.2.3.3 Lignin Lignin is a Latin word for wood. Lignin is the compound, which gives rigidity to the plant. Without lignin, plants could not attain great heights (as in trees) or rigidity found in some annual crops (straw). Lignin is a three-dimensional, highly complex polymer with an amorphous structure and high molecular weight. Of the three main constituents in fibres, it is expected that lignin would be the one with the least affinity for water (least water sorption). Another important feature of lignin is that it is thermoplastic, that is at temperatures around 90 C, it starts to soften, and at temperatures around 170 C, it starts to flow (Bledzki et al., 2002; Eichhorn and Young, 2004).

2.3 Mechanical, physical and morphological characteristics of plant fibres Due to their nature, plant fibres vary in their mechanical properties. This is because fibre morphologies vary due to different factors: plant fibres vary in their anatomy, morphology. The following sub-sections further discusses the morphological structures of different plant fibres. Moreover, the diameter also plays an important role; if the diameter decreases, the mechanical properties increase (Dittenber and Ganga Rao, 2012).

2.3.1 Morphological structure of natural fibres All plant species are built up of cells. When a cell is very long in relation to its width, it is called a fibre. For example, wood fibres are mostly 50 to100 times as long as they are wide. The length and width of some common natural fibres are illustrated in Table 2.2. Knowledge about fibre length and width is important for comparing different kinds of natural fibres. A high aspect ratio (length/width) in crucial in cellulose-based fibre composites as it gives an indication of possible strength properties. Hemp single fibre has an aspect ratio of 1000, which is good for mechanical properties. The fibre is like a microscopic tube (i.e., wall surrounding a central void referred to as the lumen). Moreover, when the cell wall is made up mainly (85% or more) of cellulose, hemicellulose and lignin, we talk about lignocellulose fibres, and this includes woody species, scrubs and most agricultural crops. Typical lignocellulose fibres from agriculture are found, for example, in straws, flax, hemp, jute and sisal. Non-lignocellulose fibres are fibres that do not contain lignin and are found in potatoes, beets and cotton, among other crops (Arbelaiz et al., 2005). Many natural fibres have a hollow space, so-called lumen, as shown in Fig. 2.12. In irregular distances, there are nodes dividing the fibre into individual cells. The surface of the natural fibres is rough and uneven which can give good adhesion to the matrix in a composite structure. Natural plant fibres are lignocellulose in nature, and they mainly contain cellulose, hemicellulose and lignin at varying concentrations. The reliability and long-term durability of natural fibres are influenced by the structure (microfibrillar angle, the fibre

34

Sustainable Composites for Lightweight Applications

Table 2.2 Mechanical and physical properties of selected natural plant fibres (Cheung et al., 2009; Pickering et al., 2016). Tensile strength (MPa)

Young’s modulus (GPa)

Elongation (%)

Density (g/cm3)

Diameter (micro meter)

Length (mm)

Date palm fibre

58e230

0.3e7.5

5e50

0.9e1.2

100e1000

20e250

Flax

345e1035

27.6

2.7e3.2

1.5

10e25

10e65

Hemp

690

70

1.6

1.4

25e35

5e55

Jute

393e773

26.5

1.5e1.8

1.3

25e200

0.8e6

Bamboo

140e230

11e17

e

0.6e1.1

14

2.7

Kenaf

930

53

1.6

e

1.14e11

12e36

Cotton

287e800

5.5e12.6

7e8

1.25

10e34

2.7

Sisal

511e635

9.4e22

2e2.5

1.15

7e47

0.8e8

Oil palm (empty fruit)

130e248

3.58

9.7e14

0.7e1.55

191e250

0.8e0.9

S-glass

4570

86

2.8

2.5

e

e

Fibre

diameter, fibre surface characteristics) and chemical compositions (cellulose, hemicellulose and lignin) content. The strength and stiffness of plant fibres mainly depend on the percentage of cellulose content and microfibrillar angle. Similarly, the performance of natural fibres as reinforcements also largely depends on operating environments (temperature and humidity) and the presence of surface defects and the hydrophilic nature of fibres itself. Additionally, the performance also largely depends on the source of origin, length and diameter, and the retting process used. These are the key concerns for these reinforcements to be used fully in structural composites as reinforcements. The cell walls of natural fibres are predominantly made up of a number of layers including a primary wall (the first layer deposited during cell development) and the secondary wall (S), which comprises of three sub-layers (S1, S2 and S3) as depicted in Fig. 2.13. A hole located in the centre of the elementary fibre is called a lumen. Such a hierarchical organisation produces multi-interphase regions with different morphological characterisations. The interphase transition regions are usually small, which induces challenges to achieve an accurate evaluation of the nanoscopic interfacial properties. Therefore, it is of necessity to develop a reliable method to characterise the interfacial properties of plant fibres and their reinforcing composites at the nanoscopic level (Fig. 2.14).

(b) Secondary wall S3

Lumen

rTs1

nS1=4 – 6 Lamina

X lumen

Helically arranged crystalline microfibrils of cellulose

Secondary wall S2

Ts2

nS2=32 – 150 Lamina

Spiral angle

Z nS3=0 – 6 Lamina

Secondary wall S1

rTs3 Ts2

Primary wall Amorphous region mainly consisting of lignin and hemicellulose

Y Cell

Disorderly arranged crystalline cellulose microfibrils networks

rTs1

Primary wall

Figure 2.13 The structure of natural fibre (a) typical cell wall representation (Bourmaud et al., 2018; Bledzki and Gassan, 1999).

Sustainable natural fibre reinforcements and their morphological structures

(a)

35

36

Sustainable Composites for Lightweight Applications

(a)

(b)

(c)

(d) S1

Middle lamella

S2 S3

Elementary fibres

P Stem

Bundle

(e)

Secondary cell wall

Primary cell wall

Elementary fibre

S2 layer

Fibre bundle

Lumen Middle lamella Tricellular junctions

Stem transverse section Fibre bundle section

Wood or shives

Figure 2.14 Multi-scale structure of flax bast fibre (a) stem of flax plant (b) bundle of flax fibre (c) representation of elementary fibre (d) S2 layer of elementary fibres (Bourmaud et al., 2015; Goudenhooft et al., 2019).

Sisal leaf fibre and jute bast fibre, for example, have multi-wall structures. A schematic diagram of the microstructures of sisal leaf fibre and jute bast fibres are also illustrated in Fig. 2.15(a) and (b), respectively. As can be seen from the diagram that the microstructure is very complex, comprising of many elements. These parameters, such as average diameter of the primary wall, spiral angle, significantly influence the overall mechanical properties of sisal fibre (Li et al., 2017).

2.3.1.1 Primary and secondary cell walls Generally, there are two walls inside the natural fibres, and they are primary and secondary walls. The natural fibres are made up of cells, and the wall that encompasses the cell is called the primary cell wall. Then, the secondary cell wall is placed between the plasma membrane and the primary cell wall, as shown in Figs 2.14 and 2.15. The cells are protected by the secondary cell walls. Also, those walls are built of layered sheaths of cellulose microfibrils, and fibres positioned in parallel inside each layer. The lignin makes the secondary cellular wall much less leaky and less flexible to water than the primary cell wall (Eichorn et al., 2001). In most cases, the secondary cell wall involves cellulose with lignin, other polysaccharides, and glycoprotein. Also, it comprises layers (S1, S2 and S3), as shown in Fig. 2.14.

Secondary wall Pectin Primary wall

(b)

Lumen Elementary fibre

Stick

Fibre bundle

After retting

Microfibril

P

S2 Lumen S1 S3

Jute reed section (with meshy structure)

Technical fibre Lumen S3

Amorphous polymers (Amorphous region) MFA, θ

Secondary S2 wall

Cellulose microfibrils (Crystalline region)

Orientation angle 7–9°

Micro fibrils

Secondary wall Primary wall

Middle lamella

Cell wall

Sustainable natural fibre reinforcements and their morphological structures

(a)

S1

Primary P wall

Amorphous cellulose

Fibre Crystaline cellulose

Single elemetary fibre 37

Figure 2.15 Microstructures of (a) sisal fibre and (b) jute fibre (Li et al., 2017).

38

Sustainable Composites for Lightweight Applications

2.3.1.2 Lumen The inside structure of the tubular shape of natural fibre is called the lumen. The structure of lumen is a factor in determining the properties of natural plant fibres as shown in Figs 2.14 and 2.15. The structure and shape of the lumen play an important role in determining the properties of the natural fibre. In addition, an increased lumen space results in a lower density of the fibre. Besides, the shape of the lumen also dictates the load-bearing properties and stiffness of the fibre. A hollow tubular structure generally offers better specific strength and stiffness. Therefore, the lumen in the natural fibres provides a natural way of reducing the density of the fibre, and therefore, the total mass of the composite. Such a hierarchical organisation produces multi-interphase regions with different morphological characterisations. Without loss of generality, the interphase transition regions are usually small, which induces challenges to achieve an accurate evaluation of the nanoscopic interfacial properties. Therefore, it is of necessity to develop a reliable method to characterise the interfacial properties of plant fibres and their reinforcing composites at the nanoscopic level (Li et al., 2017; Eichorn et al., 2001).

2.3.2 Effects of variable morphological structure and mechanical properties From Table 2.2, it can be observed that the variation in fibre dimensions is significantly high in both length and diameter for many natural plant fibres. This variation can influence the properties as well as modelling parameters. When numerical results are desired, the input parameters are important, and when there is a large dimensional variation, the accuracy of the modelling prediction could be compromised. In such a situation, experimental work is required to complement the numerical results. Nonetheless, the tensile strength and modulus of plant-based fibres are reasonably high; given the other attributes such as low density, lower cost and higher specific strength and modulus, the aforementioned shortcomings can be compensated, and these fibres have significant potential to be used as sustainable reinforcing materials in composites. Fibre length is a critical factor in obtaining maximum strength potential. It is well accepted that as the fibre volume fraction is increases up to its threshold value, the mechanical properties also are increased, but the fibre length must be greater than the critical value. Although the great variation in dimensions is not desired, at the same time, this gives opportunities for selecting fibres of different dimensions for different purposes. Another important factor that significantly affects the properties of natural plant fibre is fibre defects, including kink bands and crack running along the fibre bundles. Kink bands are folds or bends in the fibre walls. These are areas of low strength, and when the fibre is loaded, kink starts to extend, leading to the failure of the fibre. Under the tensile strength, for example, the kink bands and crack running along the fibre act as stress concentration factors, and as a result, these points can be sites for the initiation of delamination and fibre matrix deboning. SEM of hemp fibre showing kink band and fibre split along its longitudinal axis are shown in Fig. 2.16.

Sustainable natural fibre reinforcements and their morphological structures

39

Figure 2.16 SEM micrographs of an example of (a) “kink bands,” and (b) fibre split/crack observed in hemp fibres.

The structural performance and reliability of biocomposites depend on many factors, including fibre architecture, aspect ratio, orientation of fibres (uni-directional or transverse), fibre volume fraction and processing technique used, among others. These factors eventually influence the fibre-matrix interface. It is obvious that the reinforcement must be stronger if it is to give good mechanical properties. The bond between the matrix and the reinforcement (fibres) is critical since good interfacial adhesion between the matrix and the fibres transfers the stress from the matrix to the fibres improving the mechanical properties of the composites (Bisanda and Ansell, 1991). Fig. 2.16(a) illustrates the variation in thickness along the span of a single fibre strand of hemp fibre. Materials heterogeneity and diameter variation is one of the issues of natural plant fibres. The thickness variation affects the mechanical strength of the composite material since the strength at the thinner section could be very much lower compared to the thicker section. When the diameter is larger, the loadbearing ability could be higher than the small diameter section. When fracture behaviour is required to investigate, fibre properties are important parameters. Therefore, it is an important aspect to investigate to understand the characteristics of the fibre. It is evident that there is a direct correlation between the mechanical properties of natural plant fibres and surface defects and kink bands. Kink bands are present on the surface of flax, hemp and jute fibres for example. The kink bands are not desired as these act as stress concentration points, whereupon, loading can initiate debonding, leading to matrix microcracking.

2.4 Effects of variable morphology on properties The quality and reliability of natural fibres are influenced by their structure (microfibrillar angle, fibre surface characteristics, the fibre diameter) and chemical composition (vis. cellulose, hemicellulose and lignin content). The properties of natural fibres depend on their lignocellulosic contents. Higher non-cellulose contents such as

40

Sustainable Composites for Lightweight Applications

hemicellulose, lignin, pectin and wax contents influence the properties and minimises the interfacial interaction between fibres and the matrix. The key for the property enhancement of the fibres lies in removing non-cellulosic components using various surface treatments without damaging the cellulose. Additionally, non-cellulosic contents such as hemicellulose and lignin contribute to hydrophilic nature and promote early thermal degradation during various processing. Moreover, the performance of natural fibres as reinforcements are significantly influenced by operating environments (temperature and humidity) and the presence of surface defects and the hydrophilic nature of fibres itself (Faruk et al., 2012). Notwithstanding, other parameters such as harvesting techniques, agronomic practices, genotype significantly influence the overall fibre quality (fineness, aspect ratios) of natural fibres (Placet et al., 2017; M€ussig and Amaducci, 2018). The structural shape and the morphology of reinforcing fibres influence the overall properties of resultant composites. Hemp and flax fibres, for example, exist in various morphological structures: technical fibres consisting of a number of elementary fibres bonded together by pectineus gums. Lignin, located in the middle lamella, provides the rigidity to the cell wall (Peji
c et al., 2020). For example, the hollow structure of fibre morphology can provide improved vibration and energy absorption properties. Morphologies of some fibres are unidirectional, polygonal and non-uniform. This significantly influences the overall mechanical properties. Moreover, the overall performance and various properties of composites depend on properties of constituents, their size, shape, structure and overall morphological characteristics. Amongst them, just the orientation of fibre and how the load is applied (either parallel to the fibre orientation or perpendicular to the fibre orientation) plays a significant role in the overall mechanical performance (Fig. 2.17). Tensile properties of flax and jute fibre reinforced PP composites presented by Tanguy et al. (2018) clearly shows how the direction of loading influences the properties. Their results exhibited a significant difference in tensile properties between longitudinal and transverse directions (Fig. 2.18).

0°

45°

90°

E-modulus [MPa]

Composite

5000 4000 3000 2000 1000

45

0

22.5°

0°

22.5°

45°

67.5°

90°

PP

67.5°

Figure 2.17 Influence of fibre orientation on the mechanical properties of PP/lyocell composites (Cordin et al., 2018).

Sustainable natural fibre reinforcements and their morphological structures

(a) 250

(b) PP-Flax 0° PP-Jute 0°

A B

100 50 0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Strain (%)

Stress (MPa)

Stress (MPa)

200 150

41

10 9 8 7 6 5 4 3 2 1

PP-Flax 90° PP-Jute 90°

A

B

0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Strain (%)

Figure 2.18 Influence of fibre orientations (longitudinal and transverse) on the tensile properties of PP/flax (a) and PP/jute composites (b) (Tanguy et al., 2018).

They reported that the unidirectional properties of reinforced composites, for example, is influenced by different parameters including mechanical properties of reinforcements itself, the property of matrix, the fibre aspect ratio, fibre volume fraction, as well as their morphological structures. It is well accepted that the overall mechanical properties of composites depend on the adhesion between fibre and the matrix (Armentia et al., 2019; Dhakal et al., 2007). This topic is covered and extensively discussed in the next chapter, Chapter 3.

2.5 Physical and mechanical investigation of single fibres and fibre bundles 2.5.1 Importance of single fibre and fibre bundle properties It is well established that natural plant fibres bear rough surface topography due to some chemical materials present in their surfaces such as lignin, pectin and waxy substances. It is a well-accepted fact that natural fibres have high variability both physically and chemically. The properties of composites start with the properties of fibres. If fibres have good strength and stiffness, then it will be able to transfer the load from the matrix to fibres. It is, therefore, important to have a good understanding of individual fibre (elementary fibre) and fibre bundle (technical fibre) properties so that the overall properties of composites can be predicted well. Fig. 2.16 illustrates the non-uniform diameter and irregular surface properties of flax fibre. This property variation can be a challenge when calculating or modelling the properties of heterogeneous natural fibres. The understanding of fibre-matrix adhesion, key mechanisms and their influence on the overall properties of composite materials play a significant role. In conventional composite reinforcements such as glass and carbon fibre, in order to characterise the interlaminar shear strength (ISS), the stress analysis testing is normally used. However, due to property and morphology variation, the techniques used for conventional fibres

42

Sustainable Composites for Lightweight Applications

may not be fully applicable to natural fibre-reinforced composites. Therefore, single fibre testing has been one of the most popular techniques employed to measure the ISS of natural fibres instead of fracture mechanics analysis. Commonly used testing methods for fibre-matrix adhesion are subsequently elucidated, according to Zhou et al. (2016). Single fibre fragmentation test (SFFT): In single fibre fragmentation testing, a single fibre is entirely covered in a polymer matrix, which is then loaded in tension mode. The fibre covered inside the resin breaks into smaller debris at locations where the fibre’s axial stress reaches its tensile strength. From this test, a critical fibre length (lc) is determined. The average interfacial shear stress can be calculated using the formula: s ¼ sf d=2lc

(2.1)

where, s is average interfacial shear strength, s is fibre strength, d is fibre diameter and lc is critical fibre length. Single fibre pull-out test (SFPT): In SFPT testing, the fibre is implanted in a block of the matrix, where the free end of the fibre is gripped, and a load is applied continuously. The load-displacement is measured as the fibre is being pulled out. When the load required to pull the fibre from the block is determined, the resultant interfacial shear strength is then calculated using the following formula. F ¼ spdl

(2.2)

where, F is maximum load, s is fibre-matrix shear strength, pd is fibre circumstance and l is embedded fibre length. The main difference between SFFT and SFPT, according to Zafeiropoulos et al. (2007), Zafeiropoulos and Baillie (2007) is that SFPT has no stochastic data reduction system. It has been highlighted that measuring critical fibre length and critical fibre strength pose challenges for natural fibres. Micro-bond test (MT): In this, a small amount of resin is first applied on the surface of the fibre in the form of a droplet creating a shape of an ellipsoid. Then after that, a shear force is applied by restraining the bead by the opposing knife edges. The applied load and blade displacement is recorded. Then the average shear stress is calculated by using the formula: s¼

F pdl

(2.3)

where, F is maximum load, s is fibre-matrix shear strength, pd is fibre circumstance and l is embedded fibre length. Table 2.3 illustrates the mechanical and physical properties of important natural plant fibres both single and fibre bundles (Pickering et al., 2016; Bisanda and Ansell, 1991; David and Hota, 2012; de Farias et al., 2009; Faruk et al., 2012; Paiva et al., 2007).

Sustainable natural fibre reinforcements and their morphological structures

43

Table 2.3 Mechanical and physical properties of plant-based natural fibres (single fibres and their bundles).

Property Density (g/ cm3)

Flax Single/Bundle

Hemp Single/ Bundle

Jute Single/ Bundle

1.45

1.48

1.46

Bamboo Single/ Bundle 1.4

E-glass Single/Bundle 2.55

Tensile strength (MPa)

1500

800

900

550

800

400

950

750

2400

2000

Tensile modulus (GPa)

75

55

65

40

30

10

50

30

74

70

Specific strength (MPa/g/ cm3)

1030

550

600

370

550

275

680

535

940

780

Specific modulus (GPa/g/ cm3)

52

38

44

27

21

7

36

21

29

27

Strain (%)

2

1.5

1.6

1.8

1.9

3

It can be observed that the specific mechanical properties of natural plant fibres are similar or even higher than that of glass fibres. Moreover, it is reported by many literatures that the energy consumption for the manufacture of natural fibre non-woven mats, including cultivation, harvesting and fibre separation, only amounts to a third, respectively, a fifth of the energy necessary for the manufacture of glass-fibre mats. This reality provides a tremendous opportunity to use sustainable lightweight natural plant fibres in composite reinforcements. It is well appreciated that the strength and stiffness of the unidirectional composites, for example, is directly related to the strength of its single fibre. The overall properties of resultant composites, therefore, depend on the strength and stiffness of single fibres. It is evident from Table 2.3 that the single fibre provides higher mechanical properties than the fibre bundles for all the natural plant fibres. Similarly, the specific strength and modulus of flax fibre is comparable to the properties of glass fibre, even slightly higher. This is one of the key properties of natural plant fibres that needs to be exploited. It is, at the same time, important to consider the differences in properties between single and fibre bundles. As explained in the previous section, defects are created in natural plant fibres due to the growing conditions, harvesting, retting and the extraction of the fibres.

44

Sustainable Composites for Lightweight Applications

It is well established that the interface between fibres and matrix is an important phenomenon when considering the load transferability of matrices to fibres during mechanical loading. It is a well-accepted phenomenon that the final mechanical properties of composites depend on the effective fibre/matrix adhesion or load transfer capability of the matrix to the fibres. Moreover, it is also believed that the weak viscoelastic interphase between single fibres in the reinforcing fibre bundle is responsible for the stiffness reduction of natural fibre reinforced composites. The measurement of strength and modulus of single fibre and bundles become complicated by the inherent variability and due to errors encountered during these testing. Single fibre may contain cell-wall defects linking to local misalignments of cellulose microfibrils originating during growth and during the processing. These defects are also known as kink-bands, nodes and slip planes. In order to have a reliable set of data, a large number of samples need to be tested. Flax fibres, for example, are present in the outer part of flax stems in the form of single fibres assembled into bundles (10e40 single fibres). The microfibrillar angle, crystalline and amorphous phases and size of the lumen parameters are significantly influenced by the plant health, growing conditions, and hence, influence the mechanical properties of single and fibre bundles. It is established that the fibre bundle has less mechanical properties than the single fibre. Nonetheless, the fibre bundles are the main constituents of composite materials as they are composed of single fibres that are held together by pectin-rich interphase, also known as middle lamella. Jute fibre has microfibrillar angle of 8 degrees and high cellulose content. One of the unique features of jute fibres in comparison to other natural fibres such as flax is that their length is significantly shorter, 0.8e6 mm in comparison to that of flax, 10e65 mm. Due to their short fibre length, the mechanical properties of jute fibres are obtained not from a single fibre but from the fibre bundles. It is worth noting that fibre length influences the mechanical properties of the resultant composites. The aspect ratio (length divided fibre diameter) provides surface areas, which are important parameters for high mechanical strength and stiffness. The tensile strength and modulus of fibre bundle are reported for flax fibre in the range of 300e600 MPa and 30e37 GPa, respectively, as it can be seen that there is a large property gap between single fibre and fibre bundles. The differences have been attributed to viscoelastic shearing within the weak-rich interphase between elementary fibres. The reported work by suggested a large variation on the tensile strength of flax single fibre ranging from 600 to 2000 MPa. It is obvious that the variation is contributed to the different parameters such as growing conditions, extraction methods used, etc. The values reported for modulus and strain to failure lies in the range of 60e80 GPa and 1.5%e2.0%, respectively (Fig. 2.19). Fig. 2.20 depicts the tensile stress-strain curves for flax and hemp single fibres. As can be seen, flax fibre displays higher stress compared to hemp fibre (Marrotte et al., 2018; Marrot et al., 2013).

Sustainable natural fibre reinforcements and their morphological structures

45

1200 Flax fibre Flax bundle Jute bundle

Stress (MPa)

1000

800

600

400

200

0 0.0

1.0

0.5

2.0

1.5

2.5

3.0

3.5

Strain (%)

Figure 2.19 Tensile stress versus strain curves for flax and jute fibres (single and bundles) (Baley and Bourmaud, 2014).

1000

Stress (MPa)

800

600

400

200

0 0.0

Flax Hemp 0.5

1.0

1.5

2.0

2.5

3.0

Strain (%)

Figure 2.20 Tensile stress versus strain curves for flax and jute fibres (single and bundles) (Marrotte et al., 2018).

The difference in stiffness between elementary fibres and technical fibres has recently been attributed to a viscoelastic shearing within the weak pectin-rich interphase between elementary fibres. Fig. 2.21 illustrates the SEM images of jute and flax bundle cross-section and longitudinal views.

46

Sustainable Composites for Lightweight Applications

Figure 2.21 SEM images of fibres: (a) jute bundle cross-section, (b) flax bundle cross-section, (c) jute longitudinal bundle view, and (d) flax longitudinal bundle view (Tanguy et al., 2018).

2.6 Summary The use of plant-based natural fibres as reinforcement in composite materials compared to their synthetic counterparts such as glass and carbon fibres provide multiple benefits, including but not limited to low density, non-abrasive processing, abundant, recyclability, biodegradability, excellent specific strength and stiffness. From the environmental sustainability point of view, replacing synthetic fibres with sustainable natural plant fibres would attract significant attention and acceptance in lightweight composite applications. Although the use of natural plant fibres as a reinforcement in composites provides multiple benefits, some inherent drawbacks of these fibres place some key challenges that can limit the full application of these materials. Understanding of their interfacial shear strength using single fibre testing approaches is important to use and appreciate in determining the critical length and interlaminar shear strength. Therefore, there has to be continued concerted efforts by the industry and research communities to minimise and overcome those challenges so that these outstanding environmentally friendly materials are utilised to their full potential in semi-structural and structural lightweight applications.

Sustainable natural fibre reinforcements and their morphological structures

47

References Alawar, A., Hamed, A.M., Al-Kaabi, K., 2009. Characterization of treated date palm tree fiber as composite reinforcement. Compos. B Eng. 40 (7), 601e606. Andre, C.M., Hausman, J-F., Guerriero, G., 2016. Canabis sativa: the plant of the thousand and one molecules. Front. Plant Sci. 7, 19. Arbelaiz, A., Fern
andez, B., Ramos, J.A., Retegi, A., Llano-Ponte, R., Mondragon, I., 2005. Mechanical properties of short flax fibre bundle/polypropylene composites: influence of matrix/fibre modification, fibre content, water uptake and recycling. Compos. Sci. Technol. 65, 1582e1592. Armentia, S.L., Enciso, B., Mokry, G., Abenojar, J., Martinez, M.A., 2019. Novel application of a thermoplastic composite with improved matrix-fibre interface. J. Mater. Res. Technol. 8 (96), 5536e5547. Aziz, S.H., Ansell, M.P., 2004. The effect of alkalization and fibre alignment on the mechanical and thermal properties of kenaf and hemp bast fibre composites: part 1 e polyester resin matrix. Compos. Sci. Technol. 64, 1219e1230. Baley, C., Bourmaud, A., 2014. Average tensile properties of French elementary flax fibres. Mater. Lett. 122, 159e161. Bisanda, E.T.N., Ansell, M.P., 1991. The effect of silane treatment on the mechanical and physical-properties of sisal-epoxy composites. Compos. Sci. Technol. 41 (2), 165e178. Bledzki, A.K., Gassan, J., 1999. Composites reinforced with cellulose base fibres. Prog. Polym. Sci. 24, 221e274. Bledzki, A.K., Sperber, V.E., Faruk, O., 2002. Natural and Wood Fibre Reinforcement in Polymers, vol. 13. Rapra Technology Ltd, UK, pp. 1e135 (8). Bos, H.L., Van den Oever, M.J.A., Peters, O., 2002. Tensile and compressive properties of flax fibres for natural fibre reinforced composites. J. Mater. Sci. 37 (8), 1683e1692. Bourmaud, A., Beaugrand, J., Shah, D.U., Placet, V., Baley, C., 2018. Towards the design of high-performance plant fibre composites. Prog. Mater. Sci. 97, 347e408. Bourmaud, A., Gibaud, M., Lefeuvre, A., Morvan, C., Baley, C., 2015. Influence of the morphology characters of the stem on the lodging resistance of Marylin flax. Ind. Crop. Prod. 66, 27e37. Carus, M., Karst, S., Kauffmann, A., 2013. The European Hemp Industry: Cultivation, Processing and Applications for Fibres, Shivs and Seeds. EIHA, Cologne, pp. 1e9. Cheung, H.Y., Ho, M.P., Lau, K.T., Cardona, F., Hui, D., 2009. Natural fibre-reinforced composites for bioengineering and environmental applications. Compos. B Eng. 40, 655e663. Cordin, M., Bechtold, T., Pham, T., 2018. Effect of fibre orientation on the mechanical properties of polypropyleneelyocell composites. Cellulose 25, 7197e7210. David, B.D., Hota, V.S., 2012. Critical review of recent publications on use of natural composites in infrastructure. Compos. Appl. Sci. Manuf. 43, 1419e1429. de Farias, M.A., Farina, M.Z., Pezzin, A.P.T., Silva, D.A.K., 2009. Unsaturated polyester composites reinforced with fiber and powder of peach palm: mechanical characterization and water absorption profile. Mater. Sci. Eng. C 29, 510e513. Dhakal, H., Zhang, Z., Guthrie, R., MacMullen, J., Bennett, N., 2013. Development of flax/ carbon fibre hybrid composites for enhanced properties. Carbohydr. Polym. 96, 1e8. Dhakal, H., Zhang, Z., Richardson, M., 2007. Effect of water absorption on the mechanical properties of hemp fibre reinforced unsaturated polyester composites. Compos. Sci. Technol. 67, 1674e1683.

48

Sustainable Composites for Lightweight Applications

Dhakal, H.N., Skrifvars, M., Adekunle, A., Zhang, Z.Y., 2014. Falling weight impact response of jute/methacrylated soybean oil bio-composites under low velocity impact loading. Compos. Sci. Technol. 92, 134e141. Dhakal, H.N., Zhang, Z.Y., Bennett, N., Reis, P.N.B., 2012. Low-velocity impact response of non- woven hemp fibre reinforced unsaturated polyester composites: influence of impactor geometry and impact velocity. Compos. Struct. 94, 2756e2763. Directive 2000/53/EC, October 21, 2000. Directive of the European parliament and the council for end-of-lifevehicles. Off. J. Eur. Commun. ABl.EG Nr. (OJ L 269 , 21.10.2000, p. 34). Dittenber, D.B., Ganga Rao, H.V.S., 2012. Critical review of recent publications on use of natural composites in infrastructure. Compos. Appl. Sci. Manuf. 43, 1419e1429. Eichhorn, S.J., Young, R.J., 2004. Composite micromechanics of hemp fibres and epoxy resin microdroplets. Compos. Sci. Technol. 64, 767e772. Eichorn, S., Baillie, C., Zafeiropoulos, N., Mwaikambo, L., Ansell, M., Dufresne, A., et al., 2001. Current international research into cellulosic fibres and composites. J. Mater. Sci. 36, 2107e2131. Faruk, O., Bledzki, A.K., Fink, H.P., Sain, M., 2012. Biocomposites reinforced with natural fibres: 2000e2010. Prog. Polym. Sci. 37, 1552e1596. Faruk, O., Bledzki, A.K., Fink, H.P., Sain, M., 2014. Progress report on natural fibre reinforced composites. Macromol. Mater. Eng. 299 (1), 9e26. Gager, V., Le duigoum, A., Bourmaud, A., Pierre, F., Belhouli, B.C., 2019. Understanding the effect of moisture variation on the hygromechanical properties of porosity-controlled nonwoven biocomposites. Polym. Test. 78, 1e10. Goudenhooft, C., Bourmaud, A., Baley, C., 2019. Flax (Linum usitatissimum L.) fibres for composites reinforcement: exploring the link between plant growth, cell walls development, and fibre properties. Front. Plant Sci. 10 (411), 1e23. Gurunathan, T., Mohanty, S., Nayak, S.K., 2015. A review of the recent developments in biocomposites based on natural fibres and their application perspectives. Compos. Appl. Sci. Manuf. 77, 1e25. Hepworth, D.G., Hobson, R.N., Bruce, D.M., Farrent, J.W., 2000. The use of unretted hemp fibre in composite manufacture. Compos. Appl. Sci. Manuf. 31, 1279e1283. Hobson, R.N., Hepworth, D.G., Bruce, D., 2001. Quality of fibre separated from unretted hemp stems by decortification. J. Agric. Eng. Res. 78, 153e158. Izani, M.A.N., Tahir, P.M., Uyup, M.K.A., H’ng, P.S., 2013. Effects of fibre treatment on morphology, tensile and thermogravimetric analysis of oil palm empty fruit bunches fibres. Compos. B Eng. 45 (1), 1251e1257. Jawaid, M., Abdul Khalil, H.P.S., 2011. Cellulosic/synthetic fibre reinforced polymer hybrid composites: a review. Carbohydr. Polym. 86, 1e18. Jawaid, M., Khalil, H.P.S.A., Hassan, A., Dungani, R., Hadiyane, A., 2013. Effect of jute fibre loading on tensile and dynamic mechanical properties of oil palm epoxy composites. Compos. B Eng. 45, 619e624. Joffe, R., Andersons, J., Wallstr€om, L., 2003. Strength and adhesion characteristics of elementary flax fibres with different surface treatments. Compos. Appl. Sci. Manuf. 34 (7), 603e612. Joseph, P.V., Rabello, M.S., Maattoso, L.H.C., Joseph, K., Thomas, S., 2002. Environmental effects on the degradation behavior of sisal fiber reinforced polypropylene composites. Compos. Sci. Technol. 62, 1357e1372. Karus, M., Kaup, M., 2002. Natural fibres in the European automotive industry. J. Ind. Hemp. 7, 117e129.

Sustainable natural fibre reinforcements and their morphological structures

49

Li, X., Tabil, L.G., Panigrahi, S., 2007. Chemical treatments of natural fibre for use in natural fibre-reinforced composites: a review. J. Polym. Environ. 15, 25e33. Kessler, R.W., Becker, U., Kohler, R., Goth, B., 1998. Steam explosion of flax-a superior technique for upgrading fibre value. Biomass and Bioenerg 14, 237e249. Khalil, H.P.S.A., Tehrani, M.A., Tehrani, M.A., Davoudpour, Y., Bhat, A.H., Jawaid, M., Hassan, A., 2012. Natural fiber reinforced poly(vinylchloride) composites: a review. J. Reinf. Plast. Compos. 1e27. Li, Q., Li, Y., Zhou, L., 2017. Nanoscale evaluation of multi-layer interfacial mechanical properties of sisal fibre reinforced composites by nanoindentation technique. Compos. Sci. Technol. 152, 211e221. Marrot, L., Lefeuvre, A., Pontoire, B., Bourmaud, A., Baley, C., 2013. Analysis of the hemp fibre mechanical properties and their scattering (Fedora 17). Ind. Crop. Prod. 51, 317e327. Marrotte, J., Duigoua, A.L., Kervoelen, A., Bourmaud, A., Behlouli, K., Sire, O., Baley, C., 2018. Flax and hemp nonwoven composites: the contribution of interfacial bonding to improving tensile properties. Polym. Test. 66, 303e311. Merotte, J., Le Duigou, A., Bourmaud, A., Behlouli, K., Baley, C., 2016. Mechanical and acoustic behaviour of porosity controlled randomly dispersed flax/PP biocomposite. Polym. Test. 51, 174e180. Mohan, T.P., Kanny, K., 2019. Compressive characteristics of unmodified and nanoclay treated banana fibre reinforced epoxy composites cylinders. Compos. B Eng. 169, 118e125. Mohanty, A.K., Misra, M., Drzal, L.T., 2001. Surface modifications of natural fibres and performance of the resulting biocomposites: an overview. Compos. Interfac. 8, 313e343. M€ ussig, J., Amaducci, S., 2018. Scanner based image analysis to characterise the influence of agronomic factors on hemp (Cannabis sativa L.) fibre width. Ind. Crop. Prod. 113, 28e37. Nayak, S.K., Tripahy, S.S., Rout, J., Mohanty, A.K., 2000. Coirepolyester composites: effect on fibre surface treatment on mechanical properties of composites. Int. Plast. Eng. Technol. 4, 79e86. Nishino, T., Hirao, K., Kotera, M., Nakamae, K., Inagaki, H., 2003. Kenaf reinforced biodegradable composite. Compos. Sci. Technol. 63, 1281e1286. Oksman, K., 2001. High quality flax fibre composites manufactured by the resin transfer moulding process. J. Reinforc. Plast. Compos. 20, 621e627. Paiva, M.C., Ammar, I., Campos, A.R., Cheikh, R.B., Cunha, A.M., 2007. Alfa fibres: mechanical, morphological and interfacial characterization. Compos. Sci. Technol. 67, 1132e1138.  c, A.A., Kosti
c, M.M., 2020. Peji
c, B.M., Karmar, A.D., Obradovi
c, B.M., Kuraica, M.M., Zeki
Effect of plasma treatment on chemical composition, structure and sorption properties of lignocellulosic hemp fibers (Cannabis sativa L.). Carbohydr. Polym. 236, 1e9. Pervaiz, M., Panthapulakkal, S., Birat, K.C., Sain, M., Tjong, J., 2016. Emerging trends in automotive light- weighting through novel composite materials. Mater. Sci. Appl. 7, 26e38. Pickering, K.L., Efendy, M.A., Le, T.M., 2016. A review of recent developments in natural fibre composites and their mechanical performance. Compos. Appl. Sci. Manuf. 83, 98e112. Placet, V., Day, A., Beaugrand, J., 2017. The influence of unintended field retting on the physicochemical and mechanical properties of industrial hemp bast fibres. J. Mater. Sci. 52, 5759e5777. Rahman, S.H., Choudhury, J.P., Ahmad, A.L., Kamaruddin, A.H., 2007. Optimization studies on acid hydrolysis of oil palm empty fruit bunch fibre for production of xylose. Bioresour. Technol. 98, 554e559.

50

Sustainable Composites for Lightweight Applications

Ramesh, P., Prasad, B.D., Narayana, K.L., 2018. Characterisation of kenaf fibre and its composites: a review. J. Plast. 37, 731e737. Ranalli, P., Venturi, G., 2004. Hemp as a raw material for industrial applications. Euphytica 140, 1e6. Richardson, M., Santana, M.T.J., Hague, J., 1998. Natural fibre composites e the potential for the Asian markets. Prog. Rubber Plast. Recycl. Technol. 14, 174e188. Rosa, I.M.D., Santulli, C., Sarasini, F., 2010. Mechanical and thermal characterization of epoxy composites reinforced with random and quasi-unidirectional untreated Phormium tenax leaf fibres. Mater. Des. 31, 2397e2405. Rouison, D., Sain, M., Couturier, M., 2006. Resin transfer moulding of hemp fibre composites: optimization of the process and mechanical properties of the materials. Compos. Sci. Technol. 66 (7e8), 895e906. Rowell, R.M., Han, J.S., Rowell, J.S., 2000. Characterization and factors effecting fibreproperties. In: Frollini, E., Leao, A.L., Mattoso, L.H.C. (Eds.), Natural Polymers and Agrofibres Composites, pp. 115e134. San Carlos, Brésil. Satyanarayana, K.G., Arizaga, G.G.C., Wypych, F., 2009. Biogradable composites based on lignocellulosic fibres-an overview. Prog. Polym. Sci. 34, 982e1021. Satyanarayana, K.G., Guimar~aes, J.L., Wypych, F., 2007. Studies on lignocellulosic fibres of Brazil. Part I. Source, production, morphology, properties and applications. Compos. Appl. Sci. Manuf. 38 (7), 1694e1709. Seena, J., Sreekala, M.S., Oommen, Z., Koshy, P., Thomas, S., 2002. A comparison of the mechanical properties of phenol formaldehyde composites reinforced with banana fibres and glass fibres composites. Sci. Technol. 62, 1857e1868. Sinha, E., Panigrahi, S., 2009. Effect of plasma treatment on structure, wettability of jute fibre and flexural strength of its composite. J. Compos. Mater. 43 (17), 1791e1802. Sreekala, M., Thomas, S., 2003. Effect of fibre surface modification on water-sorption characteristics of oil palm fibres. Compos. Sci. Technol. 63, 861e869. Sukumaran, K., Satyanarayana, K.G., Pillai, S.G.K., Ravikumar, K.K., 2001. Structure, physical and mechanical properties of plant fibres of Kerala. Met. Mater. Process. 13 (2e4), 121e136. Suksong, W., Kongjan, P., Prasertsan, P., Imai, T., Sompong, O., 2016. Optimisation and microbial community analysis for production of biogas from solid waste residues of palm oil mill industry by solid-state anaerobic digestion. Bioresour. Technol. 214, 166e174. Sultana, C., 1992. Flax fibre. In: Sharma, H. (Ed.), The Biology and Processing of Flax Fibre, pp. 263e264. Tanguy, M., Bourmaud, A., Beaugrand, J., Gaudry, T., Baley, C., 2018. Polypropylene reinforcement with flax or jute fibre; Influence of microstructure and constituents’ properties on the performance of composite. Compos. B Eng. 139, 64e74. Vincent, J.F.V., 2000. A unified nomenclature for plant fibres for industrial use. Appl. Compos. Mater. 7, 269e271. Weiblen, G.D., Wenger, J.P., Craft, K.J., ElSohly, M.A., Mehmedic, Z., Treiber, E.L., Marks, M.D., 2015. Gene duplication and divergence affecting drug content in Cannabis sativa. New Phytol. 208, 1241e1250. https://doi.org/10.1111/nph.13562. Yousif, B.F., Shalwan, A., Chin, C.W., Ming, K.C., 2012. Flexural properties of treated and untreated kenaf/epoxy composites. Mater. Design 40, 378e385. Zafeiropoulos, N.E., Baillie, C., 2007. A study of the effect of surface treatments onthe tensile strength of flax fibres. Part II. Application of Weibull statistics. Compos. Appl. Sci. Manuf. 38, 629e638.

Sustainable natural fibre reinforcements and their morphological structures

51

Zafeiropoulos, N.E., Dijon, G.G., Baillie, C., 2007. A study of the effect of surface treatments on the tensile strength of flax fibres. Part I. Application of Gaussian statistics. Compos. Appl. Sci. Manuf. 38 (2), 621e628. Zhang, J., Henriksson, H., Szabo, I.J., Henriksson, G., Johansson, G., 2005. The active component in the flax-retting system of the zygomycete rhizopus oryzae sb is a family 28 polygalacturonase. J. Ind. Microbiol. Biotechnol. 32, 431-156. Zhou, Y., Fan, M., Chen, L., 2016. Interface and bonding mechanisms of plant fibre composites: an overview. Compos. B Eng. 101, 31e45. Zini, E., Scandola, M., 2011. Green composites: an overview. Polym. Compos. 32 (12), 1905e1915.

Further reading Ishikawa, T., 2014. Overview of CFRP (carbon fiber reinforced plastics) application to future automobiles. J. Soc. Automot. Eng. Jpn. 68. M€ ussig, J. (Ed.), 2010. Industrial Applications of Natural Fibres: Structure, Properties and Technical Applications. Wiley, Chichester, UK. Omar, F., Andrzej, K.B., Hans-Peter, F., Mohini, S., 2012. Biocomposites reinforced with natural fibers: 2000e2010. Prog. Polym. Sci. 37, 1552e1596. Reddy, N., Yang, Y., 2004. Structure of novel cellulosic fibres from cornhusks. Polymer preprints, American chemical society, division of polymer chemistry. In: American Chemical Society, Division of Environmental Chemistry; Papers Presented at the Philadelphia, PA Meeting, Division of Polymer Chemistry, vol. 45, p. 411 (2). Wambua, P., Ivens, J., Verpoest, I., 2003. Natural fibres: can they replace glass in fibre reinforced plastics? Compos. Sci. Technol. 63, 1259e1264.

Lightweight composites, important properties and applications 3.1 3.1.1

3

Lightweight composite materials: requirements and their key features Lightweight concept

Composites are defined as a macroscopic mixture of two or more distinct materials having a finite interface between them. The ability to tailor composites for a specific application is one of the biggest advantages of using these materials. Thus, the composite materials are used in a variety of fields ranging from, aerospace, sports, motorsport and the civil industry, among other applications. The lightweight concept, as far as composite materials are concerned, involves using improved design, which entails important performance requirements and puts a high priority on weight reduction. Equally, the lightweight concept emphasis on the end-of-life of a product at the design stage, the use of multifunctional and recyclable materials and efficient manufacturing processes. This involves, for example, improving the performance of composites by aligning fibres correctly, and reducing defects by using correct manufacturing techniques. Application of lightweighting principles is further linked to economics of parts and components. One of the examples of lightweight materials is carbon fibre-reinforced polymer composites (CFRP). Epoxy-based carbon fibre-reinforced composite contains a reinforcement structure of carbon fibres and a matrix phase of epoxy, which holds or binds all together. Carbon fibre composites have a high strength to weight ratio, high tensile modulus to weight ratio (depending on the anisotropy), high rigidity and a high fatigue strength. Also, due to it its lower density (1.5e2.0 g/cm3) compared to glass fibre-S2 (2.46 g/cm3), the specific strength and modulus of carbon fibre-reinforced composite are significantly higher (131 GPa cm3/g) than glass fibre-reinforced composites (35.3 GPa cm3/g). Carbon fibre is also corrosion resistant and has great heat resistant properties. Carbon fibre through the years is becoming more and more affordable, but due to the complexity of the manufacturing process, it can be hard to manufacture parts correctly and this results in high costs. One of the drawbacks of carbon fibre composites is an end-of-life option in which it is not recyclable and consumes a high amount of energy for production. In addition, despite being an excellent lightweight material, carbon fibre-reinforced thermosets composite possess low ductility. This means that when a load is applied, it breaks without giving any warning, manifesting a low toughness behaviour. The behaviour is not desirable in many critical applications (Faruk et al., 2014).

Sustainable Composites for Lightweight Applications. https://doi.org/10.1016/B978-0-12-818316-8.00006-2 Copyright © 2021 Elsevier Ltd. All rights reserved.

54

Sustainable Composites for Lightweight Applications

Transport was the largest source of CO2 emissions (27% in 2017 and 2018), the majority arising from road transport, according to the report produced by the UK Automotive Sustainability Report (2019). With new environmental legislations, for example, EU regulation is aiming to reduce CO2 emissions from 132.2 g/km by 2020 in automotive components (from EU report on reducing CO2 emissions from passenger cars). The automotive industry is currently moving towards a low carbon economy, lower fuel consumption and lower running costs. The weight of the vehicles, for example, contributes towards fuel consumption, which eventually leads to greenhouse gases (GHGs). It is clear that the lightweighting approach can contribute towards the reduction of GHGs in the transport sector. It is a general assumption that a 10% weight reduction is known to improve fuel efficiency by 5%e7% (Taub et al., 2007; Mohanty et al., 2018). Automobile original equipment manufacturers (OEMs) and related parts manufacturers in the supply chain are seeking to achieve lightweight by developing lightweight materials. In the last decade or so, OEMs are producing newer models with significant weight reduction. Recently, there has been a growing interest in the use of natural fibres for composites design and manufacturing with the aim to reducing the overall weight of the vehicles.

3.1.2

Lightweight drives

The main drivers for lightweight composite materials to replace metal parts are the challenges face by our society with regard to unprecedented environmental degradation due to the use of fossil-based raw materials. In order to reduce this trend, a viable alternative is required, and lightweight biobased materials can contribute significantly towards this (Mohanty et al., 2018). The key drives for lightweight materials include: • • • • • • •


and ELVs) Government legislation towards low carbon emissions (CAFE Consumers behaviour towards greener materials Companies striving to becoming good steward of the resources Improved environmental performance Cost saving from per Kg weight reduction Enhancements in fuel efficiency and range Avoiding the issues of corrosion especially related to metallic materials

EU proposals for CO2 reduction from new cars and vans after 2021 were agreed in 2018. According to these proposals, a 15% CO2 reduction will be required by 2025. Further, an ambitious plan was brought forward, and according to which, the further reduction target is 37.5% for cars and 31% for vans. Also, end- of-life vehicles legislation (ELVs), for example, expects about 80% of components by weight need to be recovered after their end-of-life. In order to achieve the ELVs goal, one of the approaches, which has been successfully employed in the UK, is the remanufacturing approach. In this approach, a high value used product is returned to its original performance with warranty that the product is equivalent to its original or even better than a newly manufactured product. It is claimed that remanufacturing usually uses 85% less energy than manufacturing and reduces raw materials consumption significantly.

Lightweight composites, important properties and applications

3.1.3

55

Achieving lightweighting potentials

Lightweighting can be achieved by using an optimised design, by choosing correct materials and manufacturing techniques. The fundamental expectation of this approach is using less and lighter material to support the applied load in structures or components. This requires delivering functional requirements of the products with the use of less or lighter materials possible. For example, using concepts such as design for manufacture and assemble, the components are decided at the design stage on how they are going to be assembled/disassembled and reused at the end of their life. Thermoplastic-based composite materials have better end-of-life options than thermoset-based composites as far as recycling is concerned. However, thermoset-based composites provide higher mechanical properties and low-cost manufacturing when compared to thermoplastic-based composites. Thermoplastic composites offer the following additional benefits compared to thermosets matrices. • Increased impact resistance • Fast processing of pre-impregnated materials • The ability to reshape the products

However, thermoplastic matrices come in solid-state, and due to this, it is difficult to impregnate the fibres while using certain manufacturing techniques. This can lead to a high cost of components. Another negative aspect of using thermoplastic-based composites is propensity for creep behaviour and a high internal tension caused by thermal expansion differences between the thermoplastic and reinforced fibres after cooling down typically from relatively high temperatures. As far as using thermoset-based composites for lightweight applications are concerned, there are several benefits attached including: • New optimised manufacturing techniques such as resin transfer moulding (RTM) and out-of-autoclave techniques • Good mechanical and corrosion properties • Low investment cost • Superior surface finish

The key requirements for the development of lightweight materials include: • • • •

Reduced weight Reduced the cost Long term durability under the harsh environments Improved design methods for integrated structures so that fewer parts are required to make.

3.1.4

Lightweighting benefits

Several potential benefits can be gained from the utilisation of lightweight thermosets and thermoplastics composite materials. These include design flexibility, the high strength-toweight ratio in comparison to their metal counterparts. The lightweight structures can provide increased fuel efficiency, which can reduce the overall carbon footprint.

56

Sustainable Composites for Lightweight Applications

In fact, lightweight aspires to improve product performance, quality improvement and cost reduction. In the context of environmental concerns, the lightweighting approach motivates towards using fewer materials, less energy for materials extraction, which in turn can help to solve some of the environmental and sustainability challenges. The concept of lightweighting through materials design and development is very relevant, relating to sustainability. Key benefits of lightweighting: • • • •

Reduction of the overall weight of the component Less use of raw materials (helping to realise sustainability aspirations) Reduced fuel consumption (Leading to improved environmental performance) Reduce costs

Fig. 3.1 illustrates the benefits of using natural fibre-reinforced composites in comparison with fossil-based composites. The key parameters used to compare are important factors towards the overall CO2 reduction in comparison to glass fibre composites. Cost-saving through the reduction of overall weight is a primary driver towards the adaption of less heavy parts, lightweight. The main underlining fact to this approach as far as transport sector is concerned is that the lighter the weight of the parts, the lesser the fuel consumption, leading to lesser environmental damage and higher cost-effect. In recent years, the electric car has been put forward as a way to minimise the carbon footprint. However, there is a challenge of reducing the overall weight of the batteries. If lightweight materials can be used, then it can compensate for the heavy battery weight in some way. Table 3.1 depicts the cost benefits, as suggested by Taub et al. (2019).

GHG emissions in %: fossil- and hemp-based composites compared

100%

Hemp-based composites; accounted for carbon storage Hemp-based composites; not accounted for carbon storage

80%

Fossil-based composites

60%

40%

20%

4

5

6

7

8

9

Hemp/PP vs GF/PP battery tray

Hemp fibre/PTP vs GF/PES bus exterior panel

Hemp fibre/epoxy vs ABS automotive door panel

Hemp fibre/PP vs PP composites

Hemp fibre/PP vs GF composites

Hemp fibre/PP vs GF/PP mat

0%

Figure 3.1 Greenhouse gas emission potential comparison between natural hemp fibre composites versus glass fibre-reinforced composites (Akampumuza et al., 2017).

Lightweight composites, important properties and applications

57

Table 3.1 Lightweighting benefits towards the cost. Vehicle types

Value of lightweighting ($/kg)

Light vehicle

$4.50/kg

Heavy vehicle

$5e11/kg dry van dedicated routes

Heavy vehicle

$13e24/kg bulk carriers

Despite several advantages of carbon and glass fibre composites, the ever-worsening environmental situations, new environmental legislations, research and development into more sustainable, environmentally friendly and cost-effective materials have been the focus in recent years. With this new scenario, there is a paradigm shift into the use of lightweight materials such as natural fibre-reinforced composites and biocomposites as alternative materials to carbon and glass fibres due to their positive ecological attributes and attractive specific strength and modulus. Moreover, environmental sustainability is one of the key aspects of the lightweighting approach. Towards this, natural fibres such as flax, hemp, jute and kenaf have been attractive alternative reinforcing materials due to their lower density (1.2e1.6 g/cm3) compared to glass fibre-reinforced composites (2.46 g/cm3), ensuring the production of lightweight composites together with sustainable and renewable attributes. Over the past decades, the application of natural fibre-reinforced polymer composites (NFRPCs) materials in some industry sectors has increased significantly. The applications of NFRPCs have been mainly non-structural or semi-structural components where the use in primary structural systems has been limited because of several limiting factors, including lower mechanical properties, lack of enough test data, damage mechanisms not fully understood and processes parameters are limited for composites design and manufacturing. This becomes more apparent for natural fibre-reinforced composites and biobased composites due to the fact that these materials are not as tested and established as conventional reinforcements, to some industry sectors for use on primary structural systems, an in-depth understanding of design, failure modes and their structure-property relationships is paramount. Further barriers include weak fibre matrix interface, environmental degradation and susceptibility to thermal and oxidative degradation (Dhakal et al., 2007a,b; Dhakal et al., 2012; Faruk et al., 2012; Bourmaud et al., 2018).

3.2

Important properties

For any new device or components to be used in semi-structural or structural applications, key properties required to investigate are mechanical (strength, stiffness and toughness), thermal (long-term durability and degradability), and environmental performances (response of materials under harsh operating conditions). Similarly, the parts should be able to be manufactured using a cost-effective process, which has influences on the various properties. While developing lightweight composites

58

Sustainable Composites for Lightweight Applications

using natural fibres as reinforcements, balancing these parameters becomes even more important than conventional composites as these are a relatively new class of materials, many parameters are still not fully understood, and they are under development. Many of the application areas require balanced properties. For the automotive frame, for example, it requires high strength, stiff materials with long-term durability attributes. When the components are required of high stiffness, they generally have lower impact toughness. In the case of composite materials, two constituents (reinforcements and matrices) play an important role in the overall properties of composites. The following sections discuss the important properties and various contributing factors for such properties for lightweight composites.

3.2.1

Mechanical properties of biobased composites

The overall performance of natural fibre-reinforced composites depend on the key mechanical properties. The mechanical properties of natural fibre-reinforced composites depend on a number of parameters such as volume fraction of the fibres, fibre aspect ratio (L/d), fibre matrix adhesion, stress transfer ability at the interface, the orientation of fibres, microstructure and morphology of fibres among others. Most of the studies on conventional fibre composites involve the study of mechanical properties as a function of fibre content, the effect of various treatments on the mechanical properties and prediction of modulus and strength using well-established models and comparison with experimental data. When new lightweight materials such as natural fibre composites and biocomposites are investigated, mechanical properties, mainly strength and stiffness, first need to be predicted and influencing factors are understood. It is well established that the mechanical properties of plant fibre-reinforced composites mainly depend on the strength and the stiffness of the reinforcements along with other parameters. For example, fibre morphology and geometrical aspects significantly influence the overall mechanical properties of reinforced composites. Surface modifications and chemical treatments are employed in order to improve the compatibility between fibre and matrices, especially in the case of natural fibre composites (Zafeiropoulos et al., 2002).

3.2.1.1

Tensile properties

A tensile test is conducted to determine the tensile strength, tensile modulus, elastic limit, proportional limit, elongation, and reduction in cross-sectional area. The tensile properties of composites are enhanced with the reinforcement of fibres as fibres possess higher mechanical properties (strength and stiffness) than that of matrices. The test results conducted for many natural fibre-reinforced composites indicate that the tensile strength decreases after reaching the threshold volume fraction of fibres, which is attributed due to the matrices not being able to wet the fibres and as results, the fibre matrix interface becomes weak (Dhakal et al., 2007a,b). The weak fibre matrix interface would not be able to transfer the applied load from the matrix to fibres. In such a scenario, fibre pull-out takes place, and load transfer between fibre and matrix becomes weak because of reduced fibre matrix interfacial adhesion. Another reason for lower tensile properties is related to incompatibility between hydrophilic natural fibres and hydrophobic

Lightweight composites, important properties and applications

59

polymer matrices. In this situation, adhesion between reinforcement and the matrix becomes not so strong, and upon the application of load, the delamination at the interface and interphase takes place instead of fibre breakages. When there is a strong interface between fibre and matrix, upon the application of load, the fibres are normally broken and there is less delamination (Dhakal and Sain, 2019). Moreover, most of the thermoset matrices are brittle in nature. When natural fibres are reinforced, the overall composites become more ductile as most of the natural fibres are ductile in nature. The work presented by (Dhakal et al., 2007a) on hemp fibre-reinforced unsaturated polyester composite suggests that tensile strength and modulus of composites increased up to the critical fibre volume fraction, also called threshold fibre volume fraction, above that, the strength and modulus were decreased. On the other hand, it has also been reported that voids contents in the composite significantly influences the tensile properties. Higher void contents reduce the tensile properties of the non-woven hemp/UP composites. Tensile properties are often considered one of the key properties where most research and development work has been focussed on the investigation of the mechanical properties of natural fibre-reinforced composites. Typical tensile properties of plant fibre-reinforced composites in longitudinal and transverse directions are presented in Table 3.2. The reported data presented in Table 3.2 reveals that the stiffness and strength for aligned untreated flax/epoxy composites are approximately 5 GPa and 68 MPa, respectively. The reported data reveals that the stiffness and tensile strength for randomly oriented untreated hemp in fibre composites is 2.7 GPa and 33 MPa, Table 3.2 Typical reported tensile properties of important plant fibre composites in longitudinal and transverse directions. Fibre volume fraction (Vf)

Flax/ epoxy

Aligned

(Wf ¼ 0.50)

4.67

68.12

Dhakal and Sain (2019)

Flax/PP

Random

0.14

3.4

36

0.14a

3.4

39

Hornsby et al., (1997)

Aligned

0.51b

28.7

288

Madsen (2004)

Random

0.20

5.4

77

Oksmann (2000)

Aligned

0.60

45.0

1020

Gamstedt et al. (1999)

(MA) was used as a compatibilising agent Silane treated For comparison.

b c

Ultimate stress (MPa)

Fibre orientation

Glassc/ PP

a

Stiffness (GPa)

Fibre/ matrix

References

60

Sustainable Composites for Lightweight Applications

respectively. Additionally, the stiffness and strength for the composites where hemp fibres are aligned in the direction of tension forces are 27.6 GPa and 277 MPa, respectively. It is evident that the greatest stiffness and strength in the fibre composites are obtained when the fibres are aligned in the direction of the tension force (load applied parallel to the fibre direction). In other words, to achieve the optimal tensile properties, fibre orientation plays a significant role. Similarly, the stiffness and strength for flax/PP composite in the transverse direction (load direction not parallel to the fibre) is 3.4 GPa and 39 MPa, respectively. Moreover, the stiffness and strength for flax fibre-reinforced in the longitudinal direction are 28.7 GPa and 288 MPa, respectively. As expected, the addition of the natural fibre results in a stiffness and strength value in the longitudinal or aligned direction, which is significantly higher than that in the random and transverse direction. Also shown in Table 3.2 are the tensile properties of glass fibre-reinforced composites. This demonstrates that glass fibre composites are superior to plant fibre composites irrespective of fibre orientation, and the ultimate tensile strength, in particular, is higher for glass fibre composites. These results certainly provide encouraging steps in replacing glass fibre-reinforced composites by natural fibre-reinforced lightweight composites for some structural and non-structural applications. The most important benefit of using natural fibre-reinforced composites instead of glass fibre composites is the environmental aspect. Natural fibre composites are renewable, biodegradable, has less processing energy and provides recycling possibilities. One important aspect to note in this is that natural fibre composites are susceptible to moisture absorption, and as a result, the mechanical properties such as tensile strength and modulus are significantly reduced due to the weak fibre matrix interface created as a result of moisture ingress. This aspect has been elaborated in the section. In order to withstand the applied load, the properties of both reinforcements and the matrices are important. As much as the properties of fibres, the properties of the matrices are crucial, as it is the role of the matrix to protect the fibres and maintain the fibre as straight columns to prevent them from buckling and transfer the load to fibres. The properties of reinforced composites depend on the type, shape, and orientation of the reinforcing agents, the length of the fibres, and the volume fraction (percentage) of the reinforcing material. Short fibres are less effective than long fibres, and their properties are strongly influenced by time and temperature. Long fibres transmit the load through the matrix better and thus are commonly used in critical applications. Along with the fibre volume fractions, fibre wetting in the matrix phase and high fibre aspect ratio is important in order to get improved tensile properties. Moreover, the manufacturing process (optimised manufacturing process) and void contents are equally important factors that contribute towards the optimised tensile properties of natural fibre composites. For achieving optimal strength and stiffness of natural fibre-reinforced biobased composites, various parameters as mentioned above, need to be optimised. Fig. 3.2 depicts the advantage of biobased composites in comparison to conventional composites and metallic materials. It can be seen the advantages of biobased composites. Their specific properties are only lower than 15%e20% in comparison to carbon fibrereinforced composites. Natural fibre-reinforced biobased composites, therefore, present unique mechanical properties.

Longitudinal stiffness [GPa]

225 200 175 150

Flax / PP

125

Glass / PP

100

Carbon / PP

75

Steel

50

Aluminium

25 0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Specific stiffness [GPa.cm3/kg]

(b)

Longitudinal stiffness E 250

Longitudinal specific stiffness E/ρ 140 120 100 Flax / PP

80

Glass / PP

60

Carbon / PP Steel

40

Aluminium

20 0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90%100%

Fibre volume fraction

Fibre volume fraction

Longitudinal specific stiffness in bending E1/3/ρ

(c) Specific bending stiffness [GPa1/3 .cm3/kg]

4,0 3,5

Lightweight composites, important properties and applications

(a)

3,0 2,5

Flax / PP

2,0

Glass / PP Carbon / PP Steel Aluminium

1,5 1,0 0,5 0,0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Fibre volume fraction

Figure 3.2 The absolute and longitudinal tensile properties comparison of composites against established steel and alloys (Pil et al., 2016). 61

62

Sustainable Composites for Lightweight Applications

To summarise the tensile properties of natural fibre-reinforced lightweight composites, it is important to remember that the structures, physical and chemical compositions of reinforcements significantly influence the final properties. Notwithstanding, the effects of processing parameters and variables, the critical length of the fibres, the orientation of the fibre, wettability of the fibres, the fibre/matrix adhesion and most importantly, the intrinsic qualities of the fibres.

3.2.1.2

Flexural properties

Flexural properties of natural fibre composites generally determined by using a threepoint bending test. The test is conducted by placing a specimen onto two rounded supports and is subjected to a load that is central to the supports. The test provides the flexural properties, including flexural strength, modulus and strain of the composite materials at failure. Flexural stress/strength refers to the amount of force being applied to the test specimen; flexural strain is the percentage of strain that has been exerted onto the outer surface of the specimen. The flexural modulus is the tendency of the composite to bend and is worked out by the ratio of stress to strain in flexural deformation. Flexural modulus is generally dependent on the fibre/matrix interface bonding (Aziz and Ansell, 2004). As applied to tensile properties, the higher volume fraction of fibre up to the threshold point increases the flexural properties. However, flexural strength may also decrease as the fibre loading is increased beyond its critical fibre volume fraction. As above this, the resin may not be able to wet, and stress transfer from matrix to fibres weakens as the result of a weak fibre matrix interface. Stress concentrations are experienced at weak points due to low adhesion forces that exist between the matrix and the fibres. Flexural stress is calculated using the following formula. Flexural Stress ¼ sf ¼

3FL 2bh2

(3.1)

Where, f ¼ Flexural stress, in megapascals (MPa) F ¼ Load in Newtons (N). L ¼ Span, in millimetres (mm). h ¼ Thickness of the specimen, in millimetres (mm). b ¼ Width of the specimen, in millimetres (mm). Whereas flexural modulus is calculated using the following formula:
 L3 DF Flexural Modulus ¼ Ef 4bh3 Ds Where, Ef is the flexural modulus of elasticity, expressed in megapascals (MPa). s is the difference in deflection between s00 and sʹ F is the difference in load F00 and load Fʹ at s00 and sʹ respectively.

(3.2)

Lightweight composites, important properties and applications

63

Flexural properties of natural fibre-reinforced polymer composites have been widely reported. Banana fibre-reinforced unsaturated polyester composites were prepared using a compression moulding technique, and various properties, including the flexural strength and modulus were investigated (Kenned et al., 2020). The study reports that flexural strength at break and flexural modulus showed an increasing trend with the introduction of banana fibre as reinforcement into the unsaturated polyester matrix. It was reported that the matrix on its own showed a brittle failure behaviour. As fibre reinforcement was increased from 10 to 40 wt.%, the flexural strength was increased. However, beyond the threshold value of 40 wt.% of reinforcements, the flexural strength was decreased. The reasons for the decrease at higher fibre volume fraction was attributed to less strong fibre matrix interface as it is difficult to for matrix to wet the fibres at higher concentrations (over the threshold value). As for the tensile, the flexural properties of natural fibre composites depend on several factors. The most important are critical fibre length, optimal fibre volume fractions, fibre orientation and fibre matrix compatibility. Fibre morphology, geometrical aspects also play important roles in the overall properties of fibre quality, and hence, the properties of resultant composite materials. The flexural properties of natural fibre-reinforced composites alter with varying fibre volume fractions. Hemp fibrereinforced unsaturated polyester composites, for example, increased in flexural strength and modulus with an increase of layered of hemp reinforcement, optimal properties achieved at four-layered hemp fibres (21% FVF) according to (Dhakal et al., 2007a). The void content was another factor mentioned in the many reports published in the field of natural fibre composites that significantly influences the mechanical properties.

3.2.1.3

Impact properties

Impact testing is used to measure the impact toughness of the material used. This is described as the toughness and the ability of the material to absorb energy due to sudden loading. Toughness takes into account the ductility and strength of the material being tested. Composite materials are very likely to undergo impact damage during their service life, which in turn may reduce its structural strength and load-carrying capacity, often leading to catastrophic failure. The impact strength of any composite is dependent on the number of fibres contained in a composite and also the type of impact testing being used. The impact strength of a composite increases with an increase in fibre content; however, this is limited once the critical volume fraction is attained. The reason for this is because the addition of fibres creates points of stress. There are four different types of impact testing classified (Andrew et al., 2019): • Low-velocity impacts are considered when impacts occur at a velocity of 0-11 m/sec, • High-velocity impacts are considered when impact velocity is above 11 m/sec, Ballistic impacts are considered when impact velocity is above 500 m/sec. • Hypervelocity impacts are considered when impact velocity is above 2000 m/sec

Impacts can be at different velocities and energy levels, an example of some of these impacts are tool dropping, hail strike (weather), ballistic (military), bird strike,

Sustainable Composites for Lightweight Applications

Compression strength after impact

Backside fibre failure from ice impact

BVID•

64

Not visible

Visible on the back surface

Visible on both faces Impact energy

* BVID = barely visible impact damage

Barely visible impact damage (BVID) No penetration

Type 1 Delamination

Type 2 Backside fibre failure with minor delamination

Visible impact damage (VID) Penetration Type 3 Through-thickness cracks in recurring diamond shape pattern

Type 4 Extensive throughthickness cracks

Type 5 Clean hole

Increasing velocity/energy

Figure 3.3 Effects of various impacts levels on the damage of composite structures (Kim et al., 2003).

lightning strike, runway debris and towing damage. These impacts range from low velocity to high velocity, respectively. Each of these impact velocity/types can result in different damage characteristics on composite components and can take place under different environmental conditions. Fig. 3.3 shows a rough guide to the response of composite materials for a range of impact velocity or incident impact energies. Higher energy levels yield severe damage (Fig. 3.3). Natural fibre composites are considered susceptible to impact loadings. Lowvelocity impact damage is one of the serious threats to composite parts, especially to natural fibre-reinforced biobased composites. How they respond to various impact loadings and the capacity of the composites to withstand various impact conditions during their service life is important to understand. Published scientific reports suggest that many natural fibre-reinforced composite materials are very sensitive to impact loading. There are several published papers highlight that assessing and predicting impact resistance of a composite material is always challenging since the damage consists of different forms and mechanisms such as delamination at the interface, fibre breakage, matrix cracking and fibre pull out (Richardson and Wisheart, 1996). Normally, under the low-velocity impact test, the composite is not damaged fully and still capable of performing its primary function, whereas, in a high-velocity impact test, the composite is completely ruptured or penetrated by the impact striker (Thanomslip and Hogg, 2003).

Lightweight composites, important properties and applications

65

The impact loading on a particular composite involves relatively high contact forces acting over a small area for a period of short duration. Local strains generated at the point of contact between the two solids (projectile and composite specimen) result in the absorption of energy. When a projectile strikes a laminated composite, fracture damage such as delamination, matrix cracking and fibre breakage frequently occur. When the low-velocity impact damage process is involved, the internal damage created in the structures is not visible to the naked eyes. Dhakal et al., (2007b) analysed and compared the effects of low-velocity impact damage behaviour of hemp fibre-reinforced unsaturated polyester composites in comparison with chopped strand matt E-glass fibre-reinforced unsaturated polyester composites. They reported that the time elapsed for damage initiation to penetration to be shorter for unreinforced polyester than for hemp fibre-reinforced samples. The results showed that there was a significant improvement in the impact properties of neat polyester when non-woven hemp matt was used as reinforcing fibre. Neat polyester matrix showed brittle failure behaviour (Fig. 3.4(a)). As can be expected in this study, the unreinforced polyester under impact loading succumbed to brittle fracture and failure. Upon reinforcing the polyester with four and then five layers of hemp fibre, the impact resistance improved significantly. The hemp fibre-reinforced samples, in this case, showed a more ductile fracture and failure, suggesting that with the fibre volume increased, the samples are able to absorb

(a)

Front face UPE

Back face UPE

(b)

Front face 5 layered hemp

Back face 5 layered hemp

Figure 3.4 Low-velocity impact damage patterns (a) neat unsaturated polyester sample showing brittle failure front and rear sides of impacted surfaces (b) front and rear faces of hemp fibrereinforced unsaturated polyester samples showing ductile failure pattern (Dhakal et al., 2007b).

66

Sustainable Composites for Lightweight Applications

more energy (Fig. 3.4(b)). Their report highlighted that natural fibres, generally are ductile and are key factors to improve the impact resistance of composites. Although the study found that strength and stiffness were found to be lower in hemp fibre-reinforced polyester than chopped strand E-glass fibre-reinforced composites, results show that total energy absorption by the hemp fibre-reinforced composites was comparable to that of chopped strand E-glass fibre-reinforced composite. It is worth noting that the hollow structure of hemp fibre positively influenced the impact energy as it increases the damping property, which helps in dissipating the energy better than glass fibre-reinforced composites. Dhakal et al. (2012) investigated the influence of impactor geometry and impact velocity on hemp fibre-reinforced unsaturated polyester composites. Three different types of impactor geometries used were: hemispherical, 30, and 90 at four varying impact velocity levels: 2.52, 2.71, 2.89 and 2.97 m/s, using a drop-weight impact test machine. The results indicated that the impacted specimen absorbed higher impact energy and withstood higher loads when a hemispherical shaped impactor was used compared to 30 and 90 impactor. As part of understanding the influence of incident energy levels and the repeated impact on the damage behaviour of flax/unsaturated polyester composites fabricated by using vacuum bagging technique, Dhakal et al., (2019) conducted low-velocity falling weight impact testing using three different energy levels: 25, 27 and 29 J at room temperature with repeated impact scenario. It is reported that the peak load increases almost linearly with the impact energy until perforation of the plates is observed, and afterwards the peak load reduces significantly. Their findings suggested that the absorption of impact energy significantly decreases with repeated impact damage. The energy absorption values for the post-impacted specimens were the lowest compared to all other specimens. This was mainly due to matrix cracking and fibre breakage resulting in low impact resistance. The influence of thickness on the impact damage behaviour of plain woven flax fibre-reinforced epoxy composites fabricated using hand lay-up technique were investigated by Wang et al. (2016). Three different types of specimens used in their work were of two-layered, four-layered, and six-layered thickness under the drop weight impact testing using different incident energies up to penetration using a hemispherical impactor diameter of 16.5 mm. The two-layered flax fibre specimen showed microcracks on the impact side at an energy level of 10 J. In addition, a dented area of diameter 40 mm was also observed on the specimen with an approximate depth of 5 mm. However, with an increase in impact energy level to 15 J, the cracks increased, and perforation of the specimen occurred. In the case of four-layered flax fibre specimen, visible damage such as microcracks were seen on the specimen at an energy level of 25 J. The perforation of the four-layered specimen occurred at an energy level of 30 J. For six-layered specimen, the appearance of cracks started to occur at an energy level of 50 J. This specimen had larger damage compared to the other two types of specimens, and therefore, higher absorption of energy. The perforation of a six-layered flax fibre specimen occurred at an energy level of 70 J. Moreover, dents were observed only for two-layered specimens due to plastic deformation caused by bending under impact loading. Their report concluded that the thickness of the specimen has a

Lightweight composites, important properties and applications

67

16000 14000

Force, N

12000 10000 8000 6000 4000

20 J 40 J 80 J

2000 0 0

1

2

3

4

5

6

7

8

9

10 11 12

Time, ms

Figure 3.5 Force vs time traces for impacted GFRP specimen (Khomenko et al., 2017).

significant influence on the damage behaviour of flax fibre-reinforced epoxy composite. Higher thickness of flax composites requiring greater penetration energy, an indication of higher impact resistance of composites. A similar thickness effect was reported by Dhakal et al. (2014a) on jute fibre-reinforced methacrylated soybean oil (MSO) biocomposites, using different fibre architectures. However, the incident impact energy used on those studied samples were far smaller than the reported work by Wang et al. (2016). Moving forward, Khomenko et al. (2017) reported on GFRP specimens impacted at energy levels of 20, 40 and 80 J that the peak load for all impacted specimens increased at a higher incident impact energy level. The recorded force at the maximum peak was approximately 7, 12 and 15 kN for specimens impacted at an energy level of 20, 40 and 80 J, respectively. Moreover, the impact force versus time response was similar (sinusoidal) for all impacted specimens. The first drop in the curve was seen due to the initial strike of the impactor causing delamination and matrix cracking. However, with further increase in the force caused a second drop near the peak of the curve. This was mainly due to the more extensive damage area on the impacted specimen, such as fibre breakage, delamination, and appearance of cracks. These observed behaviours are depicted in Fig. 3.5. Investigation onto the effect of temperatures on the low-velocity impact damage response of jute fibre-reinforced unsaturated polyester was carried out by Dhakal et al. (2014b). The composites were subjected under impact testing at a lowvelocity, indicating that the impact was a non-penetrating impact. The results indicated that temperature is a major influencing factor on the energy absorption and damage characteristics. The impact condition had affected the property of residual flexural strength. When the temperature was between 30 and 50 C, the composite specimens showed the greatest fraction of its preliminary strength in contrast to the same specimens, which were tested at higher temperatures around 75 C. Their work suggested

68

Sustainable Composites for Lightweight Applications

that as the temperature increases, there was a reduction in the maximum impact force, which is depicted in Fig. 3.6. The pre- and post-impact test outcomes predominantly depicted that the mode of failure of jute fibre-reinforced unsaturated polyester composites were mainly caused due to delamination despite the changes in the impact temperature. Furthermore, Paturel and Dhakal (2020) studied the low-velocity response of flax/ vinyl ester composites at different temperatures. The report proposed that glass/flax/ vinyl ester hybrid systems were able to withstand higher load at elevated temperatures. An interesting observation made in this study was that as far as the impact damage resistance is concerned, the water immersed flax specimens displayed improved energy absorption than dry specimens. This was attributed to a weak fibre matrix interface allowing higher impact energy to dissipate. Their report concluded that the hybrid system was a viable way to prevent moisture ingress effects on flax/ vinyl ester composites and a way forward strategy to achieve improved impact resistance structures.

Parameters influencing the impact damage characteristics of composites There are several important parameters that influence the impact damage behaviour of composites materials, including: • • • • • • •

Fibre geometry and volume fraction Resin types and their toughness Thickness variation Impactor tup geometry Fabric types and stacking sequence Hygrothermal environments (service conditions (temperature, humidity) Incident energy applied

3.2.1.4

Fatigue properties

Fatigue failure takes place in composite materials as a result of cumulative cyclic loading and random loading in their service lifespan. Fatigue failure resembles, for example, property deterioration during the service life of composite materials. Compared to a metallic material, the fatigue life measurement of composites is a complex process because it fails under various damage modes and involves various fibre matrix properties. Under the cyclic loadings, polymers fail well below their failure loading under monotonic loading. Therefore, fatigue failure prediction and evaluation plays an important role for measuring the long-term performance of composite materials. However, measuring the fatigue behaviour of composites is always challenging and complex (Szebényi et al., 2020). Lightweight composite materials often expose to the harsh environment and cyclic loading and inherently susceptible to fatigue failure. Fatigue failure under the repetitive loading accumulated over a period of time causes structural components to fail catastrophically, even below their ultimate strength (Qi et al., 2020). This scenario becomes a serious issue for natural fibre composites because inherently,

(b) 1500

1500

Ambient 50ºC 75ºC

Ambient 50ºC 75ºC

Contact force (N)

Contact force (N)

Lightweight composites, important properties and applications

(a)

1000

500

0 0

0.5

1

1.5

2

2.5

3

Deformation (mm)

3.5

4

4.5

5

1000

500

0 0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

Deformation (mm)

Figure 3.6 Force vs deformation traces for impacted jute fibre-reinforced unsaturated polyester composite specimen at different temperatures impacted at a velocity of (a) 1.5 and (b) 2 m/s (Dhakal et al., 2014b).

69

70

Sustainable Composites for Lightweight Applications

Stress

natural fibre-reinforced composites have structural variation (heterogeneity) and when these materials can fail in a catastrophic manner at levels much lower than ultimate strengths due to internal damage accumulation over a period of time (Mahboob and Bougherara, 2018). Therefore, predicting fatigue failure in fatigue critical components is important while designing a part or component. There is a considerable amount of reported work covering fatigue behaviour of natural fibre composites; however, not enough data to have confidence while analysing the failure mechanisms of natural fibre composites under fatigue loading (Shah et al., 2013). Despite many attractive attributes, natural plant fibre composites are inherently susceptible to long-term durability due to moisture absorption and prolonged repetitive loadings such as fatigue and vibration (Barbiere et al., 2020). In addition, the failure mechanism under accumulative loading is complex and not fully understood for natural fibre composites compared to conventional glass and carbon fibre composites. Due to lack of enough experience on how natural fibre composites fail, their failure mechanisms under such loading, these composites are not yet fully utilised in structural applications (Mahboob and Bougherara, 2018). In their comprehensive fatigue test evaluation of hemp fibre-reinforced epoxy composite (Barbiere et al., 2020). They reported the influence of moisture absorption (ambient, wet and dry conditions) on the fatigue life by applying three levels of fatigue stresses: 60%, 75% and 90% of the tensile strength. They concluded that moisture immersed samples exhibited lower tensile strength and lower fatigue sensitivity. Another interesting result of the fatigue behaviour of flax fibre-reinforced epoxy composites is by Jeannin et al. (2019). It is highlighted in their findings that the quasi-static strength and rigidity were significantly affected by the ageing conditions, but the fatigue strength was improved. There are various models applied to predict the cumulative failure of composites under fatigue loading. According to an extensive review carried out by Mahboob and Bougherara (2018), most of the fatigue testing has been undertaken using constant stress amplitude tests. Under the fatigue cyclic loading conditions, the material is subjected to stress over the time between smax and smin as a sinusoidal, as shown in Fig. 3.7, where the sm is the mean stress, and the range is defined as Ds, meaning smax  smin :

V max

'V

Vm

V min

Time

Figure 3.7 Constant stress amplitude fatigue failure.

Lightweight composites, important properties and applications

71

To measure the fatigue life, a commonly used method is a typical stress-life (SeN) diagram, using power-law, and regression equation: Smax ¼ S0 N b (Fig. 3.7). Where, Smax is the maximum stress applied, N is the number of cycles to failure, S0 is the single cycle (static) ultimate strength of the material and b is the material fatigue strength coefficient. The work undertaken by Shah et al. (2013) highlights that carbon fibre-reinforced epoxy composites outperforms the fatigue behaviour of natural fibre composites and glass fibre-reinforced composites. However, the fatigue behaviour of flax fibre-reinforced composites is comparable to glass fibre-reinforced composites. Moreover, natural fibre composites have displayed an improved damage accumulation rate (slower) in comparison to glass fibre composites. Although the causes of this behaviour are not so clear, the following could have attributed to such behaviour (Shah et al., 2013; Liang et al., 2012; Baley, 2002): 1. Due to strain hardening at cumulative cyclic loading exhibited by NFCs. 2. NFCs show slower stiffness degradation compared to glass fibre-reinforced composites during their fatigue life. 3. The progressive reorientation of microfibrils towards the loading direction

The above-highlighted observation is certainly encouraging when employing fatigue-loading situations on NFCs. However, glass fibre composites possess significantly higher static properties than NFCs, and therefore, overall fatigue performance of GFRCs is far higher than that of NFCs (de Vasconcellos et al., 2014). Owing to their natural variability and inherent moisture absorption characteristics, the fatigue damage rate of natural fibre composites is significantly higher than that of synthetic fibresreinforced composites. Key elements influencing the fatigue life of natural fibre composites (Awaja et al., 2016): 1. 2. 3. 4.

Damage accumulation at the fibre/matrix interface Material (fibre and matrix) variability Fibre types, treated or untreated and fillers or additives types Fatigue parameters (stress amplitude, intensity and frequency, temperature,

3.2.1.5

Creep behaviour

Creep is defined as the time-dependent deformation (deformation over time) or timedependent viscoelastic properties of a material subjected to continuous stress. Creep failure, also called permanent deformation, involves a combination of elastic deformation and viscoelastic behaviour of polymer composites. To measure this timedependent failure, knowledge on how materials respond the long-term loading is essential. The increased demand for sustainable lightweight materials such as natural fibre-reinforced composites, withstanding short- and long-term loading, depends on short and long-term creep behaviour at different stress and temperature levels. Creep failure generally occurs at high temperatures. For natural fibre composites to be used in structural and semi-structural applications, they need to withstand harsh environmental conditions, including high temperatures during their service life. In

72

Sustainable Composites for Lightweight Applications

Strain

Creep rupture

Initial elastic strain

First stage

Second stage

Third stage

Time

Figure 3.8 A typical strain time curve under creep testing.

such environments, polymeric composites lose their mechanical strength. It has been reported that the creep behaviour of natural fibre composites showed poor performance in certain operating conditions (Dhakal et al., 2009). Thus, it is important to have a comprehensive understanding of their long-term dimensional stability and durability relating to their viscoelastic behaviour at harsh service conditions. Creep behaviour is generally expressed as strain versus time (ε as a function of time, t) as depicted in Fig. 3.8. The creep behaviour of polymers normally consists of four stages. There are three important regions, stages under the creep testing. Instantaneous deformation, ε0 : In this stage, when an external load is employed, the material responds with an instantons strain ðε0 Þ due to the elastic/plastic deformation of the polymer. Stage I (primary creep): In this stage, strain increases at a rapid rate and slows with time. The rapid increase in strain is related to the non-uniformity of fibres in the composites, which contributes to the continuance of the fracture of fibres. If the applied stress is high, the strain time curve becomes non-linear because of damage initiation in materials. Stage II (secondary creep): In this stage, the strain rate is stable, has a slow uniform rate, which is an indication that materials are capable of sustaining the applied stress. Stage III (tertiary creep): In this stage, the strain rate accelerates and terminates when martial breaks or rupture (point of fracture). For example, in composites, this scenario is attributed to damage at fibre intersections, deboning at the fibre matrix interface and finally the material fails. Rupture of the composite at this stage can be influenced by several factors such as fibre matrix interface bonding, fibre defects, voids and stress level and temperature applied. This rupture stage is not desirable in the composite. Creep performance is measured using either tension, flexural and compressive testing modes. In such modes, the permissible or ultimate strength of the materials are first determined. Several factor contribute towards the creep failure of polymer and composites. The creep rupture behaviour, also called lifetime prediction, is therefore, important to understand when designing composites and structures.

Lightweight composites, important properties and applications

73

Dhakal et al. (2009) report on hemp fibre-reinforced unsaturated polyester composites containing varying fibre volume percentages subjected to short-term flexural creep tests. They also evaluated the influence of stress levels and temperatures. Their findings suggested that the creep behaviour is significantly influenced by both stress levels and temperature. Creep performance deteriorated as temperature increased above the glass transition temperature. Fig. 3.9 shows the strain/time curves for hemp/UP specimens, as reported by Dhakal et al. (2009), which are found to have typical strain-time curves showing an instantaneous elastic strain on loading followed by a period of slow linear deformation. At the start point, the sudden increase in strain reflects the existence of constant elasticity. Similarly, the report highlights that the recovery can be divided into instantaneous contraction and creep recovery. A constant deformation indicates that the instantaneous deformation consisting of both plastic and elastic deformation. Their work also investigates the creep behaviour at higher temperature (50 C). They

(a)

0.35 0.3

UPE 2L hemp 3L hemp 4L hemp 5L hemp CSM

Strain (%)

0.25 0.2 0.15 0.1 0.05 0 0

100

200

300

400

500

Time (min)

(b)

1.4 1.2

UPE 2L hemp 3L hemp 4L hemp 5L hemp CSM

Strain (%)

1 0.8 0.6 0.4 0.2 0 0

100

200

300

400

500

Time (min)

Figure 3.9 Creep responses of composite with different fibre concentrations at a constant stress of (a) five and (b) 15 MPa at 25 C (Dhakal et al., 2009).

74

Sustainable Composites for Lightweight Applications

suggested that at higher temperature, the matrix, UP, becomes softer, and the creep strain becomes higher. A similar influence (increase in creep deformation) at a higher temperature was reported for hemp-reinforced UP composites. In general, when predicting the creep rupture of composite materials at different stress levels and elevated temperatures, the following factors significantly influence the overall creep behaviour (Dhakal et al., 2009; Nunez et al., 2004; Morreale et al., 2017). • • • • • •

Viscoelastic behaviour of matrix materials Fibre alignment and flaws in the fibres Fibre volume fraction Void contents Fibre matrix interfacial strength Creep parameters (stress levels, temperature and time).

The stress and temperature play an important role in the creep failure of composite materials. The strain time relationship behaviour (creep rupture) of composites is influenced by the magnitude of stress levels. Therefore, creep performance (strength and modulus) decreases with increased stress and temperature. The creep compliance increases with the increase in temperature because of the increased free volume of the amorphous fraction of the polymer matrix. This is because, particularly at the high temperatures, the material approaches melting and experiences viscous flow (Park and Balatinez, 1998; Varela-Rizo et al., 2010). At high temperatures, the matrix loses its stiffness, and as a result, creep strain increases. Moreover, Greco et al. (2007) evaluated the flexural creep behaviour of woven composite. Their work explained that the first stage of rapid increase in strain was attributed to the non-uniformity of reinforcing fibres in the composites, which contributed to the continuance of the fracture of fibres. At the high-stress levels, the strain time curve becomes non-linear because of damage initiation. Their work suggests that the creep failure under the flexural mode can take place due to fibre alignment in tension and microbuckling in the compression side. The final damage can take place due to the deboning of fibre and matrix at their interface region.

3.3

Thermal stability of biobased composites

In recent years, the use of fibre-reinforced polymer (FRP) composites, including carbon fibre-reinforced polymer (CFRP), glass fibre-reinforced polymer (GFRP) and natural fibre-reinforced polymer composites (NFRPCs) have become an important part of many applications. It is well-accepted that various environmental conditions affect the physical, mechanical and thermal properties. When these composites are exposed to outdoor applications, it causes degradation, including solar Ultra Violet (UV) rays, hot/cool cycling, humidity and other environmental pollutants. It is important that the environmental effects of the weathering process on composites are fully understood to prevent the damage of the parts and components, as well as health and safety issues. With a focus on the understanding of the changes to the thermal properties of the composites due to different environmental conditions, the life span of the composites can be optimised.

Lightweight composites, important properties and applications

75

Composites, both synthetic and natural fibre-reinforced, are prone to weakening and degradation of their mechanical and thermal properties, particularly when exposed to high temperatures. In addition, natural fibres are more vulnerable to the effects of oxidative degradation (Wielage et al., 1999; Azwa et al., 2013). The chemical composition of natural fibres as reinforcements in composites influences the overall properties of composites. Natural fibres such as flax, hemp, jute and kenaf absorb moisture at different conditions. The chemical compositions not only influence the moisture absorption behaviours, but they also limit the temperature such that they can processed. In addition, Dorez et al. (2013) reported on the thermal and fire properties of flax, hemp, and bamboo-reinforced Polybutylene succinate (PBS)-based biocomposites. They used ammonium polyphosphate (APP) as a fire-retardant agent. The influence of fibre type and fire retardant on the thermal stability and fire properties were investigated. It was reported that the degradation process of lignocellulosic fibres occurred in four key steps, including: 1. Release of water vapour by the fibres covering first mass loss between 50 and 150 C. 2. The second step corresponding to the depolymerisation of the hemicellulose and cleavage of glycosidic linkages of cellulose between 250 and 370 C. 3. The degradation of cellulose at the main peak between 340 and 370 C.

Often, degradation behaviour obtained from TGA is compared with the cone calorimeter test results. The thermal decomposition and thermal stability of the composites are assessed by thermogravimetric analysis (TGA). TGA is extensively used to measure the rate and quantity of weight change (physical and chemical) in composite samples as a function of temperature. The key parameters measured in TGA include weight loss at onset temperature (Tonset), degradation temperature at different percentages of mass loss, final decomposition and remaining of residue (char yield), also known as char. In TGA testing, the materials are exposed to increasing temperature and controlled atmospheres (oxygen and nitrogen). More also, Dhakal et al. (2012) investigated the degradation of hemp fibre using TGA and suggested that the onset of decomposition temperature started around 240 C with the degradation of hemicellulose. This temperature indicates that up this temperature, the material is thermally stable. This critical temperature needs to be taken care of when natural fibres are used in composites manufacturing processes such as injection and extrusion processes. When natural fibres, especially reinforced in thermoplastics matrices, processing parameters such as temperature and time become very critical. TGA test results can indicate the thermal stability of different materials. As far as thermal degradation of natural fibre-reinforced composites are concerned, their physical and chemical compositions play a significant role. Thermal degradation behaviour of natural fibres (lignocellulose fibres) occurs at different stages (Azwa et al., 2013; Dhakal et al., 2012; Methacanon et al., 2010). The key stages are as follows: Stage 1: the mass loss region between 50 and 120 C is attributed to the release of moisture absorbed by the fibres (evaporation, desorption of residual moisture trapped). Stage 2: The mass loss between 250 and 370 C corresponds to the depolymerisation or decomposition of hemicellulose.

76

Sustainable Composites for Lightweight Applications

Stage 3: The degradation of lignin. The degradation temperature range is larger for lignin (from 200 to 500 C). Lignin helps in char formation. Low lignin content, for example, for flax, can contribute to higher decomposition temperature.

3.3.1

Thermal degradation and stability of biobased composites

Thermal stability and decomposition behaviour of composites are measured using various techniques. One of the popular techniques is thermogravimetric analysis (TGA). Natural fibre as reinforcements in composites has several limitations. One of the main drawbacks of natural fibre composites being the degradation of natural fibre at low temperatures compared to their synthetic counterparts. The main degradation of natural fibres takes place at around 200e240 C. In such a condition, these reinforcements are mainly used for commodity plastics such as PP, PE and PVC, which have a melting temperature of about 185 C. Besides, the work of Bhattacharyya and Kim (2017) on different natural fibre composites and compared with glass fibre composites highlighted the thermal stability of polymer and natural fibre composites. Fig. 3.10 illustrates the TGA traces of natural fibre composites with reference flax-glass epoxy composites covering both thermoplastic and thermosets matrices. Acceptable processing temperature of natural fibre composites depends on other constituents such as chemical composition, fibril angle, fibre morphologies and others. It is worth noting that degradation temperature directly influences the processing temperatures. For example, flax fibre degradation takes place between 270 and 400 C due to the thermal decomposition of cellulosic structures of the fibre, whereas glass fibre does not show any sign of degradation below 800 C. Thus, with a similar matrix, the flax fibre shows a poor resistance to high temperature, with a degradation of 80% at 400 C. This makes flax fibre-reinforced composites venerable at high temperature processing and applications in comparison to glass fibre. Thermal stability of polymer and composites can be enhanced by various methods, including grafting, surface treatments and use of various surface coatings. Fig. 3.11 depicts TG and DTG curves of uncoated and coated flax fibres.

3.3.2

Flammability behaviour

NFRPCs are gaining gradual acceptance for use in many critical industry sectors due to their lightweight, low cost and better specific properties than glass fibre composites. However, their susceptibility to environmental degradation and poor flame retardancy characteristics, sensitivity to their processing temperature and flammability (fire behaviour) issue pose challenges for use in these critical industry sectors in various fireproof related demanding applications despite many additives and treatment processes have been developed to improve long-term durability and flammability issues (Norouzi et al., 2015). Polymer normally does not possess high flammability resistance characteristics. Similarly, when natural plant fibres are reinforced in different polymers, the overall flammability properties of resultant composites are influenced. The lignocellulosic fibres such as flax, jute, hemp and kenaf fibres are easily ignited when they are put into

(b) 100

80

80

Relative weight (%)

100

Relative weight (%)

Lightweight composites, important properties and applications

(a)

PP 60

Flax Wool

40

Flax-PP Wool-PP

20

Epoxy 60 Flax E-glass 40 Flax-epoxy Glass-epoxy 20

Flax-PP-FR

Flax-epoxy-FR

Wool-PP-FR

Glass-epoxy-FR

0

0 0

100

200

300

400

500

Temperature (ºC)

600

700

0

100

200

300

400

500

600

700

Temperature (ºC)

Figure 3.10 Thermogravimetric traces (a) neat PP, natural fibres and various types of composites (b) neat epoxy, flax, E-glass and various types of composites (Bhattacharyya and Kim, 2017).

77

78

(a)

(b) Uncoated flax fabric 100

EPS(granules) coated flax fabric

60

EPS(flocs) coated flax fabric

1.2

EPS(granules) coated flax fabric 0.8

0.4

20

0

0 0

200

400

600

Temperature (ºC)

800

0

200

400

600

800

Temperature (ºC)

Figure 3.11 TG and DTG curves of uncoated flax fabric, EPSflocs and EPSgranules coated flax fabrics; (a) TG and (b) DTG (Kim et al., 2020).

Sustainable Composites for Lightweight Applications

40

Mass loss rate (%/ºC)

EPS(flocs) coated flax fabric

80

Mass (%)

Uncoated flax fabric

1.6

Lightweight composites, important properties and applications

79

contact with fire. These fibres burn quite fast, leaving a small amount of char, depending on lignin contents. For improvement to fire resistance behaviour various fire different flame-retardants (FRs) additives and chemical treatment methods are employed. In this way, the char yield properties are improved. There are several techniques available to measure the fire properties of composites. Cone calorimeter is a widely employed technology to investigate fire-retardant behaviour of polymer and composites in a small scale. Flammability characteristics of fire properties of polymer and composites can be determined by measuring different parameters such as heat release rate, time taken for ignition, smoke generation, etc. The main testing techniques to measure the fire properties of composites include (Norouzi et al., 2015): 1. Cone calorimeter testing: This technology measures and monitors important fire properties in accordance with the oxygen consumption principle by deriving oxygen consumption rate into heat release rate (HRR) during the combustion of specimens. This is one of the most used and reliable testing method, which enables the analysis of various fire response parameters as listed below (Schartel et al., 2003; Norouzi et al., 2015): • Time to ignition (TTI, s), • Time to peak • Total heat release rate (THR, kW/m2), • Total heat release (THR(t)) • Peak of the total heat release rate (PHRR, kW/m2) • Mass loss rate (MLR) and • Effective heat of combustion (EHC) • Total smoke produced (TPS) 2. Limiting oxygen index (LOI): This is a parameter that characterises in which the lowest percentage of oxygen in the mixture with nitrogen at which the test specimen ignites and burns on its own. LOI is a useful, small-scale test, which correlates to the ignitability in polymers. According to Gou and Tang (2011), any material with an LOI less than 21% can easily burn in air, and they are flammable; a material with an LOI greater than 21% can reduce the flame after removal of the igniting source. It is a well-established phenomenon that materials with an LOI greater than 28% are generally flame retardants, and materials between the thresholds of 21% < LOI < 28% can be referred to as slow-burning materials (Norouzi et al., 2015). It is clear that LOI provides rough guidance to indicate the relative flammability of polymer and composites. Limiting oxygen index (LOI) is used to express the relative flammability of polymers and composites. This provides rough guidance, such as composites with higher LOI will show lower flammability compared to low LOI. The formation of char is a good indicator to suggest the flammability property of polymer and composites.

UL-94: This technique can help to classify the flammability behaviour of polymers and composites according to their burning behaviour as V-0, V-1 and V-2. As reported, in the V-2 category, the burning of sample stops within 30 s on a vertical configuration and flaming drips are permitted in this category. In V-1 category, burning discontinues within 30 s on a vertical sample and flaming drips are permissible as long as they are not burned. Whereas, in the V-0 category, burning stops within 10 s on a vertical sample and flaming drips are allowable as long as they are not ignited (He et al., 2007).

80

Sustainable Composites for Lightweight Applications

(a)

(b) Ignitablility Flammablility Flame spread Heat release Fire penetration

Spark ignition Temperature and differential pressure measurements

Fully developed fire

Gas analyzers: -FTIR -HORIBA PG250 Heated sampling line (170ºC)

Temperature

Soot filter

Cone calorimeter hood

Developing fire Ignition

Time

Cone calorimeter (electrical heater) Sample holder

Weighing device

(c)

Figure 3.12 Cone calorimeter technology, showing (a) schema of a cone calorimeter, (b) stages and fire, fire properties and (c) cone calorimeter set-up detail (Fateh et al., 2016; Schartel and Hull, 2007).

In order to evaluate the fire properties of polymer and composites, the development of the HRR over time, in particular the value of its peak/maximum (pHRR or HRRmax), is usually measured. The total heat release (TRR, kJ/m2) is obtained by integrating the HRR vs time curve. Furthermore, the cone calorimeter test also enables the analysis of fire response parameters such as the time of ignition (TOI), time of combustion or extinction (TOF), mass loss during combustion, quantities of CO and CO2 and total smoke released (TSR) (Schartel and Hull, 2007). A schema of cone calorimeter is illustrated in Fig. 3.12. Furthermore, Kim et al. (2015) investigated the effects of wool, fire-retardant additive and polypropylene viscosity (MFI) on fire retardant and mechanical performance of the PP-short wool fibre composites. Cone calorimeter technology was used to measure the fire properties. Their findings demonstrated significant enhancement of fire-retardant properties of PP/wool 30 wt.% composite with APP 20 wt.%. A decrease in PHRR (w82%) and an increase in time to PHRR (w170%), compared to those of neat PP, were reported as evidenced of improved fire properties, as depicted in Fig. 3.13. Similarly, Bhattacharyya and Kim (2017) reported that HRR traces of composite laminates showed more fluctuations than that of short fibres as depicted in Fig. 3.14(a). Compared to flax/epoxy composites, glass/epoxy composites exhibited significantly slower heat and smoke and ignited slower than natural flax/epoxy composites during the combustion process. This is attributed to the inert nature of glass fibres.

PP

(b)

PP

0.1

1200 Flax-PP

Wool-PP

Heat release rate (kW/m2)

1000

Flax-PP-FR

800

Wool-PP-FI

600

400

Smoke production rate (m2/s)

Flax-PP

Wool-PP

0.08

Flax-PP-FR Wool-PP-FR

0.06

0.04

Lightweight composites, important properties and applications

(a)

0.02

200

0

0 0

100

200

300

Time (s)

400

500

0

100

200

300

400

500

Time (s)

Figure 3.13 (a) Heat release rate and (b) smoke production rate curves of neat PP, natural fibres-PP and natural fibres-PP-FR composites (horizontal sample orientation) (vertical sample orientation) (Bhattacharyya and Kim, 2017).

81

82

(a)

700

(b)

Flax-epoxy

Flax-epoxy

Glass-epoxy

500

Flax-epoxy-FR Glass-epoxy-FR

400 300 200

Smoke production rate (m2/s)

600

Glass-epoxy Flax-epoxy-FR

0.15

Glass-epoxy-FR

0.1

0.05

100 0 0

100

200

Time (s)

300

400

0 0

100

200

300

400

Time (s)

Figure 3.14 (a) Heat release rate and (b) smoke production rate curves of natural fibres-epoxy resin and natural fibres-epoxy resin-FR composites (vertical sample orientation) (Bhattacharyya and Kim, 2017).

Sustainable Composites for Lightweight Applications

Heat release rate (kW/m2)

0.2

Lightweight composites, important properties and applications

3.3.2.1

83

Parameters influencing cone calorimeter performance

There are several parameters that influence the fire properties investigated by the cone calorimeter. Some of them are listed below, as per the work conducted by Schartel and Hull (2007). Physical observation: This is a crucial parameter needing to take extra care while running calorimeter testing in order to understand the burning pattern of specimens (surface rise, char formation and cracking, bubbling and sparking). Visual observation becomes an important part of the test and should be reported and recorded. Sample thickness: The important fire properties also largely depends on the sample size and especially the thickness. While presenting and discussing the results, this parameter should be correlated. It is reported that thermally thin samples exhibit decreased ignition time compared to thick samples. Type of fibres: When considering the fibres for reinforcements, their structural morphologies, chemical composition, as well as their fibre weaving and orientation, also play an important role in fibre properties. Distance between sample and cone heater: this parameter also influences the fire properties.

3.3.2.2

Ways for improvement of fire properties of natural fibre reinforcements and composites

There are several methods employed to improve the fire performance of natural fibrereinforced sustainable composites. For the last decade, nanoclay has been extensively used as an additive to improve the thermal stability and fire-retardant properties of composites. Nanoclay being cross-linked thermally and physically, helps to improve the fire-retardant behaviour. The addition of graphene weakens the reaction of flame retardant PP to a small flame. Lower loading of graphene is observed to improve the swelling of char, resulting in better insulation of the char and decrease in heat and smoke release. It was reported that flammability (fire behaviour) of wool/PP composites was significantly improved by adding wool and ammonium polyphosphate (APP). The cone calorimeter and vertical burn tests exhibited a significant decrease in HRR and a direct flame self-extinguishment of composites, respectively (Kim et al., 2015). Expanded graphite (EG) has also been used to improve the flammability of polymers. This flame retardant expands to its original volume up to 300 times upon heating at the temperature of 900 C. The expanded graphite acts as an insulating layer, and hence, reduces dripping (Schartel et al., 2003). Several reported work suggests that both EG and APP contribute to improving fire properties (for example, PHRR and THR). However, it is mentioned that APP can produce smoke and carbon monoxide production per unit mass loss. Their work highlighted that EG was a more attractive fire retardant as this reduced both fire risk and fire hazard. However, APP improved fire risk but increased fire hazard.

84

Sustainable Composites for Lightweight Applications

Another way to improve fire properties is by blending of two polymers, which involves blending the polymer with lower fire properties into the polymer with higher fire properties. For example, if phenolic, which has higher fire resistance is mixed with unsaturated polyester resin (UPE), then the good fire properties of phenolic resin can help in improving the fire properties of UPE resin (Kandola and Nazare, 2007; Chiu et al., 2000). The use of nanoparticles such as nanoclays (layered silicate, montmorillonite (MMT) organically modified clays) have been extensively used to improve the fire properties of polymer and composites. Although clays are not fire retardant themselves, however, their incorporation into the polymer matrices can help to reduce PHRR. Nanoparticles help in the melt flow of polymers. Moreover, clay contributes to the formation of a carbonaceous-silicate char that helps in increasing thermal stability by acting as a heat barrier. Also, Kozlowski et al. (2007) reported an improvement in the fire properties of poly (lactic acid) with the incorporation of organically modified montmorillonite to fabricate a platelet structure of nanocomposites. Nanoparticles not only improve the thermal stability but also equally enhances the mechanical properties of composite materials. With a small amount, two to five wt.% nanoclay mixed into polymers, especially exfoliated structure, contributing to a significant of mechanical properties are achieved (Yang and Nelson, 2011; Norouzi et al., 2015). A novel work on using extracellular polymeric substances (EPS), such as EPSflocs and EPSgranules, were extracted from activated and aerobic granular sludge, respectively, and tested as bio-based flame retardant materials have been researched (Kim et al., 2020). From these substances, flax fibre was coated, and flammability behaviour was investigated by employing a vertical burning test. Significant differences were observed during the burning test, where coated flax exhibited far superior flammability properties than uncoated, reference flax samples. The detail of the images is depicted in Fig. 3.15 (Kim et al., 2020). Similarly, fire resistance of natural fibre composites can also be improved by thermal coatings such as ceramic or mineral fibres. These coatings can be bonded to the composite, which reflects heat away from the composites, and as a result, this increases the ignition resistance. Treated fibres have display significantly reduced pHRR significantly (Fig. 3.16). The application of the flame-retardant matrix helped to reduce pHRR values greatly. Also, the modifications helped overall flame retardance behaviour. Moving forward, Maqsood et al. (2020) comprehensively investigated the influence of different flame retardants on the thermal stability and flame retardancy behaviour of polylactic acid (PLA) yarn. The acidic and carbonic sources, as well as a modified polyester-based plasticiser, were used as flame retardants. Their report achieved a significant decrease (59%) in the heat release rate with modified flame retardant PLA in comparison to neat PLA without any additives (Fig. 3.17). The flame retardant used this study were a phosphorous-nitrogen-based non-halogenated with commercial name (EXP PP/37), the kraft lignin (KL) powder, and a modified polyester-based plasticising agent (PES).

Lightweight composites, important properties and applications After coating

(a)

85

(f)

(b)

(g)

(h)

After removing the ignition flame After 5 second flame ignition

(c)

(d)

(i)

(e)

(j)

(k)

After 12 second flame ignition

Figure 3.15 The uncoated flax fabric, EPSflocs coated and EPSgranules coated flax fabrics before, during and after burning. (a) and (b) are EPSflocs and EPSgranules coated flax fabrics before burning, respectively. (c), (d) and (e) are uncoated flax fabric, EPSflocs and EPSgranules coated flax fabrics after burning for 5 s, respectively. (f), (g) and (h) are uncoated flax fabric, EPSflocs and EPSgranules coated flax fabrics after burning for 12 s, respectively, (i), (j) and (k) are uncoated flax fabric, EPSflocs and EPSgranules coated flax fabrics after the burning was completely finished (Kim et al., 2020).

THR Residue

1400 Referenec matrix

1200

(%)

84.91

1.8

THF 62.63 SiTHF 77.09

NHF

1000 HRR (kW/m2)

(MJ/m2)

800

THR Residue

FR matrix

(MJ/m2)

(%)

NHF

45.89

20.1

7.4

THF

45.62

20.4

9.6

SiTHF 49.52

23.2

NHF 600

THF SiTHF

400 200 0 0

50

100

150

200

250

300 0 Time (sec)

50

100

150

200

250

300

Figure 3.16 Cone calorimeter curves of the reference and flame retarded composites (Szolnoki et al., 2015).

86

450 400

PLA

350 300

PLA/EXP10/PES10

250

PLA/EXP15/PES10

200

PLA/EXP15/PES10/KL5

150 PLA/EXP15/PES10/KL7

100 50 0

Total heat release/MJm–2

(b)

500

70 60

PLA

50

PLA/EXP10/PES10

40

PLA/EXP15/PES10

30 PLA/EXP15/PES10/KL5

20

PLA/EXP15/PES10/KL7

10 0

0

50 100 150 200 250 300 350 400

Time/s

0

50 100 150 200 250 300 350 400

Time/s

Figure 3.17 (a) Heat release rate of net PLA, PLA/EXP/PES and PLA/EXP/PES/KL fabrics (b) total heat release curves of neat PLA/EXP/PES and PLA/EXP/PES/KL (Maqsood et al., 2020).

Sustainable Composites for Lightweight Applications

Heat release rate/kW m–2

(a)

Lightweight composites, important properties and applications

87

Figure 3.18 Thermal conductivity mechanisms comparison between (a) crystalline materials (metal) and (b) polymers (Burger et al., 2016).

3.3.3

Thermal conductivity measurements

The thermal conductivity of composites is related to the material’s capacity to transport or conduct heat. Therefore, specific heat capacity is always linked to thermal conductivity. The thermal conductivity of materials depends on how their crystal structures are arranged. Crystalline materials are different from amorphous materials in terms of their capacity to transfer heat. The crystalline and semi crystalline polymers for example will give greater thermal conductivity compared to amorphous polymers. Overall, polymers are far less conductive than metals. Fig. 3.18 shows the thermal conductivity mechanisms crystalline and amorphous polymers by illustrating as a Newton pendulum, as described by Burger et al. (2016). It is clear that crystallinity is an important factor when measuring the thermal conductivity of materials. It is important to know the thermal conductivity properties of natural fibre composites, especially when they are required for use in heat-related applications. The information on thermal conductivity plays an important role while selecting parts for various applications. Thermal conductivity of natural fibre-reinforced composites have different behaviour than that of glass and carbon fibre composites due to their hollow morphological structures. The thermal conductivity of natural fibrereinforced composites also greatly influenced by the conductivity and the type of matrices. Matrices are insulating materials, so they have limited thermal conductivity. As mentioned above, generally, the thermal conductivity of polymer depends on their types, amorphous or crystalline. For example, the conductivity of amorphous polymers increases with temperature up to its glass transition temperature (Tg). Depending on the applications, they can be made more electrically and thermally conductive by adding conductive fillers such as graphite, metallic particles, carbon black and carbon nanotubes (Han and Fina, 2011). Moreover, thermal conductivity also depends on the filler contents level. The specific heat capacity of the composites decreases with increasing particle contents.

88

Sustainable Composites for Lightweight Applications

Average thermal conductivity (W/m.K)

0.30

0.29

0.28

0.27

0.26

0.25

0.24 Neat (AR0) PCL

Aspect ratio 19

Aspect ratio 26

Aspect ratio 30

Aspect ratio 38

Figure 3.19 Thermal conductivity showing the influence of aspect ratio of hemp/PCL biocomposites (Dhakal et al., 2020).

There are two types of thermal conductivities measured for fibre-reinforced composites: 1. The in-plane thermal conductivity (to thermal conductivity in the direction parallel to fibre axis) and 2. Through-thickness thermal conductivity (refers to thermal conductivity in the direction perpendicular to fibre axis).

Due to the structure and morphology of natural fibres, natural fibre-reinforced composites have low through-thickness conductivities. It also becomes a challenge to measure the thermal conductivity of natural fibres and resultant composites. Fig. 3.19 shows the influence of fibre aspect ratio on the in-plane thermal conductivity of hemp fibre-reinforced poly(e-caprolactone) biocomposites. Their report highlighted that the thermal conductivity depended on the aspect ratio where it increased up to the fibre threshold values, as well as the interaction between fibre matrix interface. The thermal conductivity on composites reported that the through-thickness conductivity is far lower than in-plane conductivity. Thermal conductivity, for example, depends on various parameters such as morphology, density and homogeneity of materials. Measuring conductivity in the through-thickness direction is always more challenging than in the in-plane (Ming et al., 2015). The thermal conductivity of composites decreases as fibre volume fraction increase; however, with increase in fibre angle, thermal conductivity increases. Mounika et al. (2012) reported on the thermal conductivity of bamboo fibre-reinforced polyester composites in comparison to glass/polyester composites. Fibre volume fraction plays an important role in thermal conductivity. The thermal conductivity of polyester was highest compared to different bamboo composites (Fig. 3.20).

Lightweight composites, important properties and applications

89

Thermal conductivity (Wm–1k–1)

0.3 0.25 0.2

Vf = 0.304

0.15 0.1 0.05 0 Glass FRPC

Bamboo FRPC

Bamboo short FRPC

Polyester resin

Figure 3.20 Thermal conductivity of bamboo/polyester composites (Mounika et al., 2012).

a - LDPE b - SRP (20% sisal) c - GSRP (50/50 SRP/GRP) d - GRP (20% glass)

Thermal conductivity (W/mK)

0.58

d

0.50

c

0.42 0.34

b a

0.26 0.18 0.10 120

160

200

240

280

320

360

Temperature (K)

Figure 3.21 Thermal conductivity of bamboo/polyester composites (Kalaprasad et al., 2000).

Similarly, a study carried out by Kalaprasad et al. (2000), highlighted that thermal conductivity and thermal diffusivity for sisal-reinforced polyethylene (SRP), glassreinforced polyethylene (GRP) and sisal/glass hybrid fibre-reinforced polyethylene (GSRP), reported that GSRP showed increased thermal conductivity than nonhybridised sisal-reinforced composite (SRP), due to the addition of the glass fibre, as depicted in Fig. 3.21.

90

Sustainable Composites for Lightweight Applications

Thermal conductivity (W/mK)

0.8

0.6

0.4

0.2

0.0 0 wt% GnPs

0.1 wt% GnPs

0.3 wt% GnPs

0.5 wt% GnPs

Figure 3.22 Through-plane thermal conductivity of the GNPs/carbon fibre-reinforced epoxy composites (Wang and Cai, 2019).

3.3.3.1

Ways improving the thermal conductivity of polymer matrix composites

Several reported works have highlighted ways to improve the thermal conductivity of polymers and composites. Incorporation of thermally conductive fillers into less conductive polymers are common practice. For example, the work reported by Wang et al. (2015) recommends that thermal composites of epoxy composites were significantly enhanced with the incorporation of expanded graphite (EG) particles. The key to achieving such improvement was attributed to the proper dispersion of EG in the epoxy matrix. Similarly, the reported work by Wang and Cai (2019) achieved a significant improvement in through-plane thermal conductivity of carbon fibre laminated composites by incorporation of 0.3 wt% of graphene nanoplatelets (GNPs) as illustrated in Fig. 3.22. The thermal diffusivity of the composite plates in their work was measured using the laser flash method. The conductivity was calculated using Eq. (3.3). K ¼ a r Cp

(3.3)

Where, K ¼ thermal conductivity (W/m K), a ¼ thermal diffusivity (mm2/s), r ¼ densities of the samples (g/cm3), Cp ¼ specific heat capacity (J/g K). Idumah and Hassan (2016) added exfoliated graphene nanoplatelets (GNP) to kenaf fibre-reinforced PP composites. Their study reported enhancement in thermal conductivity by 88% with 3 phr GNP loading. According to various published works in the thermal conductivity of polymeric composites, the key parameters that influence the thermal conductivity can be listed as follows: 1. Defects found in crystalline structures according to the comprehensive work presented by Burger et al. (2016). They reiterated that thermal conductive will always be decreased

Lightweight composites, important properties and applications

91

Point defect

Dislocation

Grain boundary

Figure 3.23 Phonon scattering in crystalline materials as a result of various defects (Burger et al., 2016). 2. 3. 4. 5. 6. 7. 8. 9.

Mechanisms involved and fillers alignment Filler chemical nature Filler content types and dispersion Aspect ratios and sizes Polymer physical structure Polymer particle interactions and interface Temperature Processing techniques used in the fabrication of composite laminates.

when defects are presented in the structure, as illustrated in Fig. 3.23. They suggested that any phonon scattering caused by defects leads scattered transfer of waves through the crystal.

3.4

Environmental effects (water absorption) and their influence in different properties

Moisture ingress in composites creates negative influences. Long-term exposure in harsh environments (high temperatures and humidity) during their service life, for example, can lead to a significant deterioration on the properties, especially for natural fibre-reinforced composites and biocomposites. Natural fibre-reinforced composites, due to inherent affinity to moisture, degrade and lose their structural integrity when they are exposed to humidity and extreme weather conditions. The extreme hygrothermal environments (conditions such as temperature, humidity), corrosion and UV radiation weaken the fibre matrix interface and severely influence the mechanical, physical and thermal properties. Moreover, at the elevated temperatures, the reduction in mechanical properties accelerates. In the process, moisture first attacks in the matrix and then reaches the fibres, which then leads to a weak fibre matrix interface and degrades the mechanical properties (Dhakal et al., 2007a; Akil et al., 2009). Enough points have been already made in the earlier sections that there has been a significant uptake in the application of natural fibre-reinforced composites for nonstructural applications due to their many attractive attributes. Their semi-structural

92

Sustainable Composites for Lightweight Applications

and structural supplication largely depends on how their mechanical properties and damage behaviour are accurately evaluated, understood and applied in the design process. Water absorption creates heterogeneous structures in composite materials, which not only influences the mechanical properties but effects glass transition temperature, thermal conductivity and storage modulus as a result of damaged fibre matrix interfaces. It is, therefore, the study of moisture absorption and its effects on various properties of natural fibre composites that have attracted significant attention from academia and industry, especially in the last decade. The following sections highlight the hygrothermal behaviour and its influences on the properties of natural fibre composites.

3.4.1

Moisture diffusion mechanisms in composites

Concerning the natural fibre-reinforced composites, the presence of hydroxyl group and complex chemical compositions and morphological structures, a high percentage of moisture absorption takes place when exposed to moisture and humid conditions. This phenomenon also restricts the compatibility or adhesion with non-polar polymer matrices. The absorption of moisture by a natural fibre composites leads to fibre swelling. The moisture ingress further leads to the degradation of fibre matrix interface. When the fibre matrix interface is weakened, the load transfer capability from the matrix to the reinforcing fibres is reduced, and the overall mechanical properties are significantly compromised. According to (Dhakal et al., 2007a) water is absorbed along the fibree matrix interface, which can lead to a swelling of the fibre or a hydrolytic breakdown of any chemical bonding between the fibre and the matrix. Moisture absorption tests are conducted following different standards. Normally, moisture absorption tests are conducted at room temperature and elevated temperatures with different humidity. It is generally accepted that accelerated moisture absorption tests can reduce the time required for moisture absorption test to reach the saturation moisture uptake. The moisture absorption is measured, using Eq. (3.4), which allows calculating the percentage of water absorption in the polymer composites by measuring the weight difference between the samples immersed in water and the sample in dry condition (Dhakal et al., 2015). Mt ¼

Wt  W0  100 W0

(3.4)

where, Mt is moisture uptake, and W0 and W are the mass of the specimen before and during ageing, respectively. The moisture content versus square root of time then is normally plotted. From the moisture uptake curve (moisture gain versus the square root of time), the rate at which water moves from the surface to the interior of the specimens is described by the water diffusion coefficient. Diffusion coefficient Dx, an important parameter in Fick’s law, is used to determine the rate of water absorption and calculated by using Eq. (3.5), from the initial stage of the diffusion process curve; pffi 2 Mt = t  h Dx ¼ p 4  Mt


(3.5)

Lightweight composites, important properties and applications

(a)

Fibre swells after moisture absorption

93

(b)

Capillary mechanismwater molecules flow along fibre-matrix interface

Matrix microcrack around swollen fibres Water diffusion through bulk matrix

(c)

(d)

Water soluble substances leach from fibres

Ultimate fibre-matrix debonding

Figure 3.24 Effects of water in the fibre matrix interface (Azwa et al., 2013).

Where, Mt the maximum water uptake of the sample (%), h the thickness of the pffi samples (m) and t the time square root (s). The main reason for natural fibre composites degrading after prolonged moisture absorption is due to their chemical compositions. The structural integrity of cellulose, hemicellulose and pectin, for example, is significantly altered as the result of moisture absorption, especially at the elevated temperatures. Generally, when natural fibre-reinforced composites are exposed to hygrothermal environments water molecules can penetrate the composites by three different mechanisms (Dhakal et al., 2007a; Akil et al., 2014; Espert et al., 2002; Robert et al., 2010): 1. Water molecules ingress through microgaps between polymer chains. 2. Capillary transport process taking place via gaps and flaws at the fibre-matrix interface due to the manufacturing defects or poor wettability resins into fibres. 3. Swelling of fibres (especially in the case of plant fibres such as flax, hemp, jute and kenaf) causing the expansion of the microcracks (stress cracking) formed in the matrix.

The three mechanisms highlighted above are illustrated in Fig. 3.24 where water acts on cellulosic fibre and polymer matrix interfacial bond according to the four following stages, as suggested by Azwa et al. (2013) and (Dhakal et al., 2007a). (a) Micro cracking of the brittle thermosetting resin occurs. (b) As the composite cracks and gets damaged, capillarity and transport via micro cracks become active. (c) The capillarity mechanism involves the flow of water molecules along the fibreematrix interfaces and a process of diffusion through the bulk matrix.

94

Sustainable Composites for Lightweight Applications

(d) The water molecules actively attack the interface, resulting in the debonding of the fibre and the matrix leading to premature failure.

It is clear from the above three paths that both the matrix and fibre/matrix interface, and moreover, manufacturing flaws/defects, as well as the morphology of reinforcements, can contribute to moisture ingress in the composite materials. In the case of natural fibre composites, fibres are more susceptible to moisture absorption. With the polar nature of fibres, reinforced in hydrophobic matrices, the fibre-matrix interface and interphase regions are always a weak point where hydrolysis reaction takes place and causes interfacial damage leading overall deterioration of various properties (Azwa et al., 2013; Dhakal et al., 2007a). Moisture absorption-related behaviours, considering the above-described mechanisms, are normally assigned into three main categories: 1. Linear Fickian behaviour: in this moisture absorption, after an initial increase due to weight gain resulting from moisture absorption, gradual equilibrium is reached. 2. Non-Fickian behaviour: after an initial increase due to weight gain resulting from moisture absorption, no equilibrium is achieved. 3. Pseudo-Fickain behaviour: If the moisture absorption behaviour lies in between Fickian and non-Fickian, then often it is termed as Pseudo-Fickain behaviour.

The above mentioned moisture absorption behaviours were described by (Dhakal et al., 2007a) where hemp fibre-reinforced unsaturated polyester composites. The samples were immersed at room temperature (25 C) and elevated temperature (100 C). The implication was that the composite moisture was directly proportional to the fibre volume fraction. The water absorption behaviours of the composites were illustrating Fickian behaviours at room temperature and displaying non-Fickian behaviours at elevated temperatures.

3.4.2

Effects of moisture diffusion the mechanical properties

The effects of moisture absorption on the mechanical properties of natural fibre composites have been extensively studied. There are many studies reporting the effects of moisture absorption on the various mechanical properties of natural fibre composites. The reported work by Dhakal et al., (2007a), for example, on the effects of hemp fibre reinforcement on the mechanical properties (tensile and flexural) of unsaturated polyester composites extensively presented the moisture absorption and its influence on the mechanical properties. Moisture absorption at room and elevated temperatures had a significant effect on the tensile and flexural properties. Their findings highlighted that the moisture absorption was directly proportional to fibre volume fraction. At a higher fibre volume fraction of hemp fibre, the percentage of moisture absorption was increased. It was correlated that a high percentage of moisture absorption led to a significant reduction in tensile and flexural properties, which was attributed to a weak fibre matrix interface due to moisture absorption. Their report further highlighted that the rate of moisture ingress (diffusion coefficient) increased gradually with higher hemp fibre contents, which was attributed to the cellulose contents of the hemp fibre (Bismarck et al., 2002; Dhakal et al., 2007a). In their work, it was reported that strain to failure (tensile and flexural) was increased with imposture absorption. Furthermore,

Lightweight composites, important properties and applications

95

15,0

12,5

90 80

Weight gain (%)

10,0 70 7,5

Weight gain measurements Tensile modulus measurements

60 50

5,0

40

Decrease of tensile modulus (%)

100

2,5 30 0,0 0,0

2,5

5,0

7,5

10,0

12,5

15,0

17,5

20 20,0

√t (h)

Figure 3.25 Effects of water absorption on the tensile modulus (Habibi et al., 2019).

their work highlighted that the temperature has a substantial influence on the loadbearing capability. At high temperature, the tensile and flexural stiffness and strength were reduced compared to room temperature. Similarly, the work by Habibi et al. (2019), as presented in Fig. 3.25, shows a significant decrease on the mechanical properties at three different temperatures (room, 50 and 75 ). Their report suggests that the tensile strength of room temperature immersed specimen was decreased by 15% in comparison to dry specimens. Similar to the tensile strength, the tensile modulus decreased significantly. The interesting observation is that the tensile strain to failure was significantly increased with moisture absorption. The decrease in strength and modulus and an increase in strain (approximately by 205%) is attributed to a weak fibre matrix interface due to moisture ingress and induced plasticisation. Compared to dry specimens, the moisture immersed specimens at 50 C deceased the tensile modulus and strength by 35% and 45%, respectively. An increase in temperature to 75 C, leads to a reduction of modulus and strength by 57% and 53%, respectively. Additionally, Akil et al. (2009) studied the moisture absorption behaviour of pultruded jute fibre-reinforced unsaturated polyester composites at three different environments: distilled water, seawater and acidic solution. It was observed that in all three environments, flexural properties were decreased with increased moisture absorption (Fig. 3.26). The reason attributed for such behaviour was moisture absorption contributing to weak fibre matrix interface. Additionally, when plant fibres such as jute are exposed to water, it swells, and as a result, microcracks are formed in brittle matrix such as UPE, which creates a path for transport of water through the fibre matrix interface. Their work further suggested that strain to failure increased with an increase in moisture content. This behaviour was attributed to the cellulose contents of the jute

96

Sustainable Composites for Lightweight Applications

(a) 400

Flexural strength, MPa

350 300 250

Standard

200

1st day

150

1st week 2nd week

100

3rd week

50 0 Distilled water

Sea water

Acidic solution

Environmental condition 100 90 80

Maximum flexural strain, ×10

–3

mm/mm

(b)

70

Standard

60 50

1st day

40

1st week

30

2nd week

20

3rd week

10 0 Distilled water

Sea water

Acidic solution

Environmental condition

(c)

4

Flexural modulus, MPa

3 3 Standard

2

1st day 2

1st week 2nd week

1

3rd week

1 0

Distilled water

Sea water

Acidic solution

Environmental condition

Figure 3.26 Effects of moisture absorption on (a) flexural strength, (b) maximum flexural strain and (c) flexural modulus for pultruded jute fibre-reinforced unsaturated polyester composite after exposure to (a) distilled water, (b) seawater and (c) acidic solution (Akil et al., 2009).

Lightweight composites, important properties and applications

Flexural strength (MPa)

(a) 140

121.1

120

80

98.1

89.9

100 61.4

64.8 51.0

60 40 20 0

(b) 120 Flexural modulus (GPa)

Unconditioned Conditioned

97

100

Type ,

Type ,,, 107.1

Unconditioned Conditioned

87.3 76.3

80 60

50.6 37.9

40

Type ,,

31.1

20 0

Type ,

Type ,,

Type ,,,

Figure 3.27 Effects of hygrothermal exposure on the flexural properties of flax/PP and flax/ carbon/PP hybrid composites. Type I represents flax/pp composites. Type II represents flax/ carbon hybrid composites with flax ply at the surface. Type III represents flax/carbon hybrid composites with carbon ply at the surface (Cheng et al., 2020).

fibre. With moisture, plasticisation gives rise to plastic deformation, as well as cellulose degradation, and makes jute fibre more flexible, which contributes to an increase in failure strain as a result of moisture absorption. The flexural properties of flax and flax/carbon hybrid composites reported by Cheng et al. (2020) suggested that as for the tensile properties, the moisture ingress influenced the flexural properties of natural flax fibre composites. Fig. 3.27 illustrates the comparison of flax and flax PP composites and their hybrids. It is evident from the results that flexural strength and modulus decreased significantly due to moisture ingress leading to plasticisation of flax fibre, as well as PP matrix. A similar observation was reported by Paturel and Dhakal (2020). In their work, the flax and flax/glass hybrid composites were investigated in terms of moisture absorption (Fig. 3.28) and its influence on the low-velocity impact damage characteristics. The flax/vinyl ester composites and their hybrids were immersed at two different temperatures (room and 70 C). It was observed that flax hybrid composites absorbed far less moisture than flax/VE composites (Fig. 3.28). The sample at elevated

98

Sustainable Composites for Lightweight Applications 5

G6-RT G6-HT F6-RT F6-HT GF4G-RT GF4G-HT

4.5

Weight gain (%)

4 3.5 3 2.5 2 1.5 1 0.5 0

0

5

10

15

20

25

30

Time1/2 (h)

Figure 3.28 Moisture absorption versus square root of time at room temperature and high temperature (Paturel and Dhakal, 2020).

(a)

(b)

(c)

(d)

5 days

10 days

30 days

Figure 3.29 Surfaces of single flax fibres: (a) unaged flax fibre, reference; (bed) aged flax fibres (Cheng et al., 2020).

Lightweight composites, important properties and applications

99

temperatures reached to saturation in a short period of time compared to room temperature moisture immersed samples. Natural fibre absorbs more moisture from the surrounding environment than glass fibres. In the short-term moisture absorption, some composites can recover their mechanical properties upon the drying of the composites. However, if the samples are immersed for a prolonged period, the properties are irreversible (Cheng et al., 2020). Fig. 3.29 depicts the SEM images of flax fibre at dry and water immersed conditions for various periods. The surfaces of flax fibres are damaged as the immersion period is increased. This further promotes moisture absorption and leads to deterioration in the mechanical properties of composites.

3.5

3.5.1

Numerical modelling of mechanical properties and damage behaviour of natural fibre-reinforced biobased composites Background

Composite structures and their related properties and damage mechanisms are traditionally analysed by using the macroscopic/microscopic approaches. The failure modes of composite materials can be measured by various techniques, the main being experimental testing, finite element analysis (FEA) and analytical methods. Numerical modelling is extensively used to analyse composite structural damages. Finite Element Method (FEM) is a popular technique used to predict the damage behaviour of composite materials. Moreover, in recent years, FEM has been used in modelling of natural fibres, as well as natural fibre-reinforced composites (Xiong et al., 2018). There are many reported works on the prediction of bulk properties of composite materials. Among the analytical models, Hashin and Shtrikman (1963), Voigt-Reuss type bounds Voigt (1889), Mori and Tanaka (1973) and Rule of mixtures (ROM) are reliable models used to predict the bulk properties (strength and modulus) of composite materials as an analytical method. Although these analytical methods are attractive, in recent years, due to the development of fast computers, the analytical methods are replaced with numerical modelling. The modelling of composite materials, especially natural fibre composites, is a difficult task as it involves heterogeneous components. Upon loading, composite materials undergo many possible damage mechanisms before the ultimate failure of the material.

3.5.2

Predicting mechanical and damage behaviour of natural fibres and composites

Over the last decade, sustainable composite materials such as natural fibre-reinforced biobased composites are becoming an increasingly attractive alternative for lightweight applications covering important industrial sectors such as automotive, marine and construction. The full exploitation of these composites largely depends on how accurately one can predict their bulk properties and damage behaviours.

100

Sustainable Composites for Lightweight Applications

Different theoretical models to predict the mechanical behaviour of natural fibre composites have gained significant attention in recent years. The benefits of using such models are included but not limited to: • Cost effective as experimental studies take times and require materials and equipment • It is also easy to use and can predict long term behaviour of materials

Numerical modelling can be carried out using various methods such as finite difference approach (FDA), boundary element approach (BEA) or finite element analysis (FEA). In general, FEA is widely used to study the mechanical behaviour of composite materials compared to FDA and BEA. In FDA, the body is separated in equal element sizes. The body is discretised into equal and square or rectangular elements grid, and hence, this approach is not applicable in curve surfaces and model with complex geometry. It is also not applicable in cases where stress concentrations are varied. The BEA consists of separating the only boundary into different numbers of elements. Line elements in 2D or surface elements in 3D modelling are used. The discretisation of boundary is carried out into a number of finite elements, where loading is applied, and shape functions are used to calculate the unknown for each element. Finite element analysis (FEA) is a numerical technique, which helps to find approximate solutions to mathematical models that involves partial differential equations. This is one of the most effective analytical tools in the context of dynamic loading. In FEA, the body is discretised into a finite number of elements. Different shapes of elements can be utilised based on the problem. Different shape functions are utilised to solve the problems, and results are re-assembled to achieve the solution for the whole problem (Becker and Karamanlidis, 2004; Liu et al., 2015). There are various commercial FEA software available. The most commonly used include, but are not limited to, ANSYS, ABAQUS and LS-DYNA and Multiscale Designer. These software provide FE codes with an interface to ensure that the user has an understanding with minimum efforts.

3.5.2.1

Finite element method

In this approach, the problem under consideration is divided into a number of small segments of a simple basic shape, known as “spatial discretisation”, with each of the simpler shapes being known as an “element”, and the whole collection of elements being known as a “mesh”. For each element, the relevant properties of the element are then predicted in a simple way. In the case of structural analysis, the relationship between forces, displacements, strains and finally, stresses are evaluated, following the theory of elasticity. In FEA, the body is discretised into a finite number of elements. Different shapes of elements can be utilised based on the problem. Different shape functions are utilised to solve the problems, and results are re-assembled to achieve the solution for the whole problem.

3.5.2.2

Boundary element method

In this approach, only the surface or boundary of the problem under consideration is considered, as the name implies. The boundary is divided into a number of small segments over which the transformed governing differential equations, in the form

Lightweight composites, important properties and applications

(a)

(b)

25

30

Experiment, long fiber

Experiment, long fiber

FEA, 1000Pm fiber

25

20 15 FEA, 167 Pm fiber

10

Experiment, short fiber

5 0 0.00

Stress (MPa)

30

Stress (MPa)

101

FEA, 1000Pm fiber

20 15 10

FEA, 167 Pm fiber Experiment, short fiber

5

0.01

0.02

0.03

Strain

0 0.00

0.01

0.02

0.03

Strain

Figure 3.30 Stress versus strain traces of non-aligned fibre composites (a) without and (b) with 10% microcells (Kern et al., 2016).

of integral identities, are numerically integrated. As in the finite element analysis, provided that the boundary conditions are satisfied, a system of linear algebraic equations emerges, for which a unique solution can be obtained.

3.5.2.3

Finite difference method

In this approach, the body is separated into equal element sizes. The body is discretised into equal and square or rectangular elements grid, and hence, this approach is not applicable in curve surfaces and models with complex geometry. It is also not applicable in cases where stress concentrations are varied. For composite materials to be used in semi-structural and structural applications, predicting their mechanical performance using numerical modelling becomes important. There are several models established for glass and carbon fibre composites. However, not many developed for natural fibre-reinforced biobased composites. The development of an accurate prediction model is difficult due to the following factors: • • • • •

Their natural variability (non-uniform fibre diameter) Their variable microfibril angles Their non-straight shape and non-straight alignment Their complex architectures (a large portion of the lumen) Some natural fibres such as bast fibres having a hollow structure (lumen) needing more complex mesh to analyse

While considering the above factors, the reported work by Kern et al., have presented aligned and non-aligned fibre models. Their work highlights that the net stress versus strain traces were analysed from the FEA models and then were compared with the experimental results. It has shown a good agreement between FEA and experimental results which is illustrated in Fig. 3.30. They also mentioned that FEA model underpredicted the effect of fibre length (Kern et al., 2016).

102

Sustainable Composites for Lightweight Applications

Moreover, natural fibre-reinforced composites and biocomposites are a relatively new class of materials. When new materials are developed, it is important that these material properties can be analysed using certain design criteria and materials parameters. Due to these variations, properties and damage prediction of natural fibre composites are largely carried out by using experimental studies rather than numerical modelling. The designers consider some of the drawbacks of natural fibre composites such as quality variations, susceptibility to moisture absorption, quality variations and microbial growth as weaknesses and disadvantages. These drawbacks are not within the boundaries of the most commonly available software, which were mainly designed for the testing and modelling of classical conventional composite materials. Some researchers have developed a numerical model based on the finite elements method to predict the micro and macro mechanical behaviours of composite materials. The experimental and numerical comparison made for banana fibre-reinforced PP composites by Monzon et al. (2019) suggests that FEA model can be used for natural fibre composites but requires software with the capability of full meshing covering fine filaments. They suggested Multiscale Designer software capable of single unit cell simulation with matrix so that micro-mechanic level analysis can be carried out. In their work, experimental results were compared with the numerical modelling with accepted error in comparison to experimental results. The use of numerical methods provides several advantages compared to experimental studies. The main advantages being time and cost. For experimental studies, very often, expensive equipment is required, which adds cost. In numerical modelling, mechanical behaviour can be predicted by using various software using material characteristics such as strength, modules and fibre diameters. However, to model and predict the accurate mechanical behaviour of composites, numerical results are important to correlate with experimental studies. In other words, the numerical models are important to validate with the experimental results. Due to the heterogeneous behaviour of composites, the modelling of failure behaviour and mechanisms is always challenging. Therefore, using some simple models, a realistic behaviour of composites can be predicted. These numerical behaviours and values are important while designing composite materials, which, in turn, will help to reduce fully relying on experimental results. Finite element method has been widely applied in modelling the mechanical behaviour of natural fibres, and natural fibre-reinforced composites. One of the popular techniques used for predicting the mechanical behaviour unidirectional composites is to produce a micromechanical model by considering representative volume (RVE), a constitutive description of (Sun and Vaidya, 1996): • the reinforcing fibres and their packing sequence and the polymer matrices, • fibre matrix interface characterisation, • and appropriate boundary conditions on the RVE.

As the performance of biobased composites is significantly influenced by these parameters, so they do for the mechanical (static and dynamic) behaviours prediction of biobased composites using a certain model.

Lightweight composites, important properties and applications

103

25 Flax/pp composites

Moisture (%)

20

15 Flax/carbon hybrid composites 10

64 hrs

Type I - experiments Type I - fick’s model Type II - experiments Type II - fick’s model Type III - experiments Type III - fick’s model

5

0 0

2

4

6

8

10

12

14

16

18

Square root (time/h)

Figure 3.31 Moisture absorption behaviour of flax/pp composites and flax carbon hybrid composites (Cheng et al., 2020).

3.5.3

The prediction of static mechanical properties of composites using FEA

The prediction of static mechanical properties of composites using different theoretical models was carried out by Dayo et al. (2017). Fig. 3.31 shows experimental and Series or inverse rule of mixtures (IROM), Halpin-Tsai and Nielsen models used to predict the Young’s modulus of treated hemp fibres-reinforced polybenzoxazine composites. Their work reported that some models were closer to experimental than others. They suggested from their work that the best estimation of Young’s modulus was Nielsen and Halpin-Tsai. More so, inputting the right material parameters is key to get an accurate FEA result. If there are wrong parameters introduced, the prediction can completely go wrong, and the system safety can be compromised. Therefore, it is normal practice to validate FE simulation via analytical and experimental results. Therefore, in order to validate the accuracy and efficiency of FEA, experimental tests are normally conducted. One of the approaches, which has been implemented in recent years to reduce the risk of moisture absorption on the degradation of mechanical properties is the hybridisation technique. Both experimental data and theoretical calculations have been used to predict the moisture absorption curves. Inputting the right material parameters is key to get an accurate FEA result. If there are wrong parameters introduced, the prediction can completely go wrong, and the system safety can be compromised. Therefore, it is normal practice to validate FE simulation via analytical and experimental results. Therefore, in order to validate the accuracy and efficiency of FEA, experimental tests are normally conducted. In addition, Bauroni and Dhakal (2019) investigated the mechanical behaviours of flax and flax fibre-hybridised glass fibre composites by using a model. Their results showed that mechanical properties, especially low-velocity impact properties largely

104

Sustainable Composites for Lightweight Applications

(a)

(b)

6

6 Experiment Simulation

5

4

Load, kN

Load, kN

4 3

3

2

2

1

1

0

Experiment Simulation

5

0 0

1

2

3

4

Displacement, mm

5

6

0

2

4

6

8

10

12

Time, msec

Figure 3.32 Experimental and numerical results from low-velocity impact (a) forcedisplacement histories at 25 J of impact energy (b) force-time histories at 25 J of impact energy (Bauroni and Dhakal, 2019).

depend on fibre volume fraction, fibre matrix interface, properties of fibres and matrix. Their work predicted the experimental and numerical results of the low-velocity impact on flax and its hybrid composites (Fig. 3.32). The experimental results were shown in good agreement with the numerical modelling for impacted composites.

3.6

Applications of lightweight natural fibre composites

The functional requirements of a component dictate the properties. The applications of biobased composites, for example, depends on several factors. For some applications, wear and moisture resistance behaviours are important. For other applications, thermal stability is more important. For example, under-hood automotive parts, impact resistance and heat deflection temperature with reduced flammability properties are important. Over the past decades, biobased composites have been extensively used in various industry sectors such as construction, transport, marine and sports due to their excellent specific properties, corrosion resistance and low processing energy requirements, among others. Among the various sectors, transport, specifically these composites are predominantly used in automotive industries mainly for reinforcement of door panels, passenger rear decks, pillars and boot linings. In addition to the automotive industries, the aerospace and construction and low-cost building industries are investigating the possibility of using natural fibres as a reinforcing material as alternatives to conventional glass fibres.

3.6.1

Automotive application (road vehicles and land transport)

Body panels, engine components- rocker covers, whole car bodies in some cases. Performance road cars utilise carbon fibre composite for both chassis, body and engine components such as rocker covers and air intakes. Due to their lightweight,

Lightweight composites, important properties and applications

105

Hemp fibre

Non-woven mat

Finished door

Pre-finished door

Figure 3.33 Production of door frame using hemp fibre as reinforcement (Peças et al., 2018).

renewability and positive environmental benefits, biocomposites have gain popularity in the automotive industry. According to a report published by ECCC (2016a; WHO), the transport sector contributes the highest greenhouse gases (GHGs) in the World. In the current scenario, there is a significant demand and expectation that the transport sector considers environmental aspects as a high priority along with other aspects such as cost and passenger comfort. The GHGs are related to the burning of fossil fuel or energy consumption in all stages of the vehicle parts: raw materials extraction, production, use and disposal/re-use). Despite several advantages outlined, carbon fibres are expensive materials. It takes a large amount of energy to produce carbon fibre composites. Moreover, the high cost of producing carbon fibre (in terms of cost and energy consumption) makes it difficult to be used in automotive applications when the OEMs in this sector are seeking to bring the overall cost down due to high competition. There are efforts directed in bringing the cost down and using the recycled carbon fibres as composite reinforcements. Nevertheless, these reinforcements take a massive amount of energy for production and manufacturing (Agarwal et al., 2019). If the overall vehicle weight can be reduced, then the fuel consumption can also be reduced. In this context, lightweight, sustainable biobased composites can contribute significantly. Natural fibre composites have been extensively used in interior parts of automotive applications, as depicted in Fig. 3.33. This component was produced using 50% PP and 50% hemp non-woven mat. This indeed provides technological, ecological and economic benefits. Moreover, non-woven hemp/PP-based composites can also have an important impact on damaged properties. The automotive industry in Europe extensively uses short natural fibre-reinforced (hemp, flax, kenaf) composites with thermoplastic and thermoset matrices. These composites are mainly used in non-structural interior applications using mainly non-woven mats, as these composites do not possess high mechanical strength required for semistructural or structural applications (Fig. 3.33). Moreover, many automotive OEMs use

106

Sustainable Composites for Lightweight Applications

F-150

Focus Ford

3-series

X5

5-series

OEM

BMW

Touareg

Golf 7 VM

Elantra

Santa Fe

Hyundai

0

50

100

150 200 250 Weight reduction (kg)

300

350

400

Figure 3.34 Auto lightweighting agenda of major global OEMs and their models from 2013 to 2020 (Pervaiz et al., 2016).

natural fibre-reinforced composites in trim parts such as door panels, seat cushions, cabin linings and parcel shelves. Reported works suggest that GHGs generated by automotive vehicles counts for more than a quarter of GHGs generated (Pervaiz et al., 2016). It makes every sense that the automotive sector is now at the forefront of using natural fibre-reinforced lightweight composites. Plant fibre-reinforced composites have been widely used by automotive industries in Europe and North America due to their economical, technological and green credentials. The adaption of lightweight materials in transport sector has been a key drive towards weight saving initiatives. The extensively used fibres and semi-products are raw fibres and non-woven mats, and the composites. These composite parts possess moderate mechanical properties, which makes them well qualified for interior parts (Fig. 3.33). "The most environmentally friendly thing you can do for a car that burns gasoline is to make lighter bodies" (Henry Ford).

When Henry Ford said this, he was trying to move using steel parts to biocomposites parts, but the aim was not achieved due to World War II. The weight reduction in the automotive industry is still an on-going process as with most areas of transport through all the parts of a vehicle design. Other areas of performance can be benefited from a lighter vehicle, and from the manufacturer’s side it can also reduce the production cost, which can then help to decrease the fuel consumption and lead to overall GHGs emissions when the lighter vehicle could be made. Also, after the end-of-life of a vehicle, the parts are expected to be re-used or recovered. For this, biobased composites possess very attractive attributes, which has attracted the attention of OEMs in automotive sector. These materials can be recycled as new reinforcements, which still can hold many important properties (Mohanty et al.,

Lightweight composites, important properties and applications

107

Figure 3.35 Natural plant fibre application in the current E-Class Mercedes-Benz (Akampumuza et al., 2017).

2018). Fig. 3.34 shows the emphasis given by the North American automotive OEMs of lighting drive. For example, hemp fibre-reinforced polyester composites were used in the “Lotus Eco Elise” concept car, where hand lay-up and vacuum bagging techniques were used to form the parts. It was reported that the use of hemp fibre replacing glass fibres helped to reduce the Eco Elise’s weight by 32 kg resulting in a higher fuel economy and a better performance in comparison to the standard Elise S. Majority of the car manufacturers in Germany use natural fibre composites in various parts. Fig. 3.35 depicts the use of plant fibre parts used in the current E-Class Mercedes-Benz. Daimler/Chrysler has also manufactured door trim panels using a biocomposite plastic comprising 25% hemp, 25% kenaf and polypropylene. Due to the high sensitivity of plant fibres towards thermal stability, low flammability characteristics and the susceptibility to moisture absorption, the natural fibrereinforced composites are mainly used in interior parts. Besides their use in trim parts, plant fibres are also used for thermo-acoustic insulation. Another well-established field of application is the use of coconut fibres bonded with natural latex for seat cushions. For this application, the ability of plant fibres to absorb a large amount of humidity leads to an increased comfort that can not be reached with synthetic materials. An important step towards higher performance applications was achieved with the door panels of the Mercedes-Benz E-Class. The wood fibre materials previously used for the door panels were replaced by a plant fibre-reinforced material consisting of a flax/sisal fibre mat embedded in an epoxy resin matrix. A remarkable weight reduction

108

Sustainable Composites for Lightweight Applications

Figure 3.36 Applications of natural fibre composites by various original equipment manufacturers (Akampumuza et al., 2017).

of about 20% was attained, and the mechanical properties, important for passenger protection in the event of an accident, were improved. Recently, plant fibres have also been used in exterior components such as the engine and the transmission covers (Adekomaya, 2020). As highlighted in the earlier sections, due to the high specific strength and stiffness and the excellent damping properties, have promoted the use of flax fibre composites in semi-structural, as well as structural components, by replacing synthetic E-glass fibres in many in automotive parts. Motorsports is one of the popular application areas of lightweight composites. Carbon fibre reinforced composites (CFRP) and glass-fibre reinforced composites (GFRP) as outstanding lightweight materials which are used in many components including wings, aerofoils, side panels, nose cones, air boxes, and chassis components. Ranging from formula students projects to Formula One (F1) standard components, race teams use composites for the bodywork and aero packages of the racing cars. Motorsport teams are now using composites for the structural elements of the cars, such as carbon fibre monocoques, tubular structures and suspension components for potential reduction of weight and CO2 emissions Pervaiz et al., (2016). Both thermoplastic- and thermoset-based composites are used in automotive applications. These are due to advanced manufacturing techniques being proven in the aerospace industry. Fig. 3.36 depicts the advantages of natural fibre composites for

Lightweight composites, important properties and applications

109

Figure 3.37 Vibration damping related properties of different natural non-woven fibre-reinforced PP composites (Hadiji et al., 2020).

automotive lightweight applications, which is significantly lighter than conventional chassis with improved stiffness. Fig. 3.37 shows the vibration damping properties of non-woven natural fibrereinforced PP composites in comparison with glass/PP composites. The key parameters used to compare are important factors towards the overall damping analysis of natural fibre composites. It is evident from Fig. 3.37 that non-woven natural fibre composites exhibited the highest damping ratio compared to glass fibre-reinforced PP composites. The key properties compared were: bending modulus, fibre volume fraction, composite density, loss factor and porosity content. Due to these unique damping parameters of important natural fibre-reinforced composites, these materials can serve as alternative lightweight composites in the automotive industry (Hadiji et al., 2020).

3.6.2

Aerospace and related application

Carbon fibre and glass fibre-reinforced composites were developed with the aim of its usage in the aerospace sector. Carbon fibre-reinforced composites with the combination of properties such as lower density, high strength to weight ration attracts for both structural and non-structural applications in civil and military aircraft (Soutis, 2005). The Dreamliner aircraft (Boeing 787), for example, used 50% by weight continuous carbon fibre composites in order to reduce the overall weight of the aircraft for enhanced fuel economy. These applications demand excellent performance in terms

110

Sustainable Composites for Lightweight Applications

of high strength and stiffness, high operating temperatures, high thermal conductivity, among others. For natural fibre-reinforced composites to be used in aerospace applications, it is a long way away. However, if the fire-resistance behaviours of these composites are brought to the acceptable level, then there is a possibility of using these composites in non-load bearing applications in interior parts. With the classic fibrereinforced polymer composites (FRP and GRP), however, there are often considerable problems with respect to re-use or recycling after the end-of-tlife. Moreover, when it comes to the important mechanical properties such as vibration damping and toughness, carbon fibre composites do not necessarily exhibit the highest performance. With excellent damping properties of flax fibres as reinforcements, using these fibres as hybrid materials together with carbon using high-performance manufacturing technology, natural fibre-reinforced hybrid composites have a significant potential to be used in the non-structural applications in the aerospace sector.

3.6.3

Marine applications

Carbon and glass fibre-reinforced thermosets (epoxy, vinyl ester and polyester) composites (GFRP) are extensively used in marine applications due to their costeffectiveness and higher mechanical performance, as well as ease of fabrication (formability), and room temperature curing ability. Due to the lightweight nature of composites, boats can use structures and skins made from composite materials, and they not only benefit from the lightweight properties of composites but the composites can be easily repaired if they become damaged. Due to the cost factors, glass fibrereinforced composites are extensively used in marine industry as the cost of carbon fibres is almost five times than that of glass fibres. Glass fibre-reinforced composite applications include yachts, sailboats, dinghies and lifeboats, due partly to its corrosion resistance, lightweight and resistance to degradation by water. Any materials that require to be applied in marine applications will need to go through harsh environments during their service life. Moisture absorption elated issues such as weak fibre matrix interface, blistering, degradation due to the marine environment are important criteria to be understood. Therefore, lightweight, long-term durability and high strength and stiffness are prerequisite design criteria for composites to be used in marine applications. In recent years, concerns over microplastics generated by marine components and the harm that has caused to the marine lives have been highlighted by many scientific research papers as a serious concern. Therefore, there is a great potential for the use of sustainable, environmentally friendly biodegradable composites. However, due to their vulnerability to harsh environments, inconsistent long-term durability has hindered the use of these composites in marine outdoor applications. However, significant efforts have been made to make these composites withstand harsh environments (glass/carbon/flax hybrid composites for example) as these composite have significant potential to be used as lightweight, sustainable materials in marine applications. In recent years, composite sandwich structures have been used in the marine industry with the main aim being the forming of lightweight structures with high strength and stiffness. In sandwich structures, two skins are placed topsides usually, glass,

Lightweight composites, important properties and applications

111

carbon and aramid fibres and thick foam cores (usually polyvinyl chloride (PVC), polystyrene and honeycomb) are placed in the middle. The key benefits of using sandwich structures in marine components are cost and weight saving. After many years of using non-renewable conventional reinforced composites, the marine sector is in the search for more sustainable composites with environment and end-of-life as a focussed agenda. Going forwards for using a lightweighting approach in the marine sector requires a new shift, using natural fibre-reinforced biobased materials. For this, the matrices need to go more towards biodegradable/renewable from conventional ones. Similarly, with regard to reinforcements, the need it to move from glass fibres to more sustainable such as flax, hemp, kenaf and date palm fibres, for example. With this new approach, sustainability aspirations can be realised; however, the shortcomings of sustainable renewables mentioned need to be overcome.

3.6.4

The building construction application

Carbon fibre and glass fibre-reinforced composites have also been used extensively in construction industries. The applications include composite cables as an anchor for an earthquake-resistant building, bridges, precast concrete. There are numerous other applications where carbon and glass fibres can be used. They include turbine blades, gearbox, flywheel energy storage, conductive reinforcements for fuel cell and solar panel supports. Building construction materials consume about 40% of the World’s global energy, 25% of the global water, 40% of the global resources this is due to the majority of materials used come from conventional materials such as steel, cement and bricks where these conventional materials consume a significant amount of electrical and thermal energy (Asdrubali et al., 2015). In recent years, there have been efforts to replace conventional building materials with lightweight composite materials. Moreover, efforts are being made even to replace non-renewable carbon and glass fibres by natural fibres such as hemp, jute, kenaf, coir, flax and date palm fibres (Mark and Fam, 2019; Yorseng et al., 2020). These materials offer tremendous benefits in building materials, including they being renewable, lower cost, biodegradable and good mechanical and acoustic properties. Additionally, there are several reported works that highlighted the use of solid wastes generated from agricultural and industrial production and are used as reinforcements in construction and building materials. These initiatives not only provide the required mechanical properties but also provide economic and environmental benefits. Building and construction industries are aiming to use low carbon materials. This concept aims to correlate the amount of energy used (embodied energy), raw materials and energy intensity (Yan et al., 2016). While considering low carbon materials in building and construction sector, many literatures indicate that amongst the more common construction materials considered, the lowest energy option or low carbon material is timber, while the highest or high carbon material is steel, with concrete in between. Furthermore, if sustainable low

112

Sustainable Composites for Lightweight Applications

carbon materials are used in buildings, it can significantly reduce the overall energy consumption. There is a strong view that the replacement of high-energy intensive processes and materials can be replaced by the use of low carbon materials, with better design and recyclable materials (Yan et al., 2016). As detailed by Faruk et al. (2012), with businesses recently having an increasing environmental awareness and with a view to the ability to recycle end-of-life products, the need for the use of natural fibres over synthetic fibres is becoming more prevalent. The reduced cost and increased performance of NFRPCs go hand in hand with the drive to increase a company’s green credentials in inciting more research into the uses and capabilities of natural fibre composites. A significant encouragement that can be taken from the above sections is that the specific properties of hemp and flax fibres are comparable to glass fibre especially. This indicates that the specific mechanical properties of hemp fibres are approaching the properties of glass fibres. Because of the density of hemp and flax fibres is lower than that of glass fibre, the reinforcement of hemp and flax to the polymeric matrix reduces the density of the composite as a whole. These properties make hemp fibres attractive environmentally friendly reinforcing lightweighting reinforcements in composites (Gurunathan et al., 2015). It is well established that the key factor in the development of today’s modern buildings has been the use of structural steels due to their unique properties. However, the development of structural composites has demonstrated that advanced composites can be viable alternatives to steel due to advanced manufacturing processes and understanding of their damage mechanisms. This is due in part to improved material performance, but more particularly, the development of new linear processing techniques. Continuous pultrusion processing methods produce low cost sections, which have high specific stiffness, constant linear properties and good environmental stability, making them suitable for primary construction components. However, a principal limitation in the use of composites in mainstream construction has been the high environmental cost in the manufacture of energy incentive synthetic fibres, and the problems associated with their subsequent disposal at the end of their life (Ahmad et al., 2014). Natural fibre-reinforced (rice husk/phenolic) composites have been successfully tested and used to make temporary shelters. The rice husk particleboards are manufactured in various densities, thickness, types and grades to suit a wide range of applications (Arjmandi et al., 2017). The important market for the flax and hemp short fibres is their use in ecological thermal insulation materials. In many countries, this market is growing faster than the total market for insulation materials (Hussain et al., 2019; Saheb and Jog, 1999).

3.6.5

Other applications

Carbon fibre-reinforced composites have been used in sporting goods due to their lightweight and strength-to-weight ratio. Fishing rods, tennis racquets, race bicycles and prosthetics. Due to their excellent strength and damping properties, flax fibres have been used to make bicycle frames by hybridisation with carbon fibres. As indicated in the earlier

Lightweight composites, important properties and applications

0.015

113

Titanium Aluminium

0.01 Steel 0.005

Unidirectional carbon Custom layup composite

0

Bidirectional flax

Figure 3.38 Vibration damping properties of different materials (Amiri et al., 2018).

sections, despite having high strength and stiffness, carbon fibres are expensive materials, and they lack impact toughness (energy absorbing capabilities when a sudden load is applied) required for a bicycle frame (Amiri et al., 2018). Their work reported that by hybridising flax fibres into carbon, their frame exhibited similar strength and stiffness as commercially available carbon, titanium and aluminium frames while exhibiting superior damping properties. These outstanding properties were achieved by maintaining 40 wt.% bio-content. Fig. 3.38 shows the vibration damping of different materials where bidirectional flax outperforms the other important materials. With these unique properties, natural fibre-reinforced flax composites can also be sued in making fishing rods and other sporting equipment.

3.7

Conclusions

This chapter overviewed the importance of lightweight composites reinforced with sustainable and renewable plant-based fibres, which meet the expectation of lightweighting aspiration and drives, as well as providing several benefits, including performance, environmental and cost efficiency. While considering important mechanical properties, it was evident that the mechanical properties of natural fibre composites were comparable to the properties of conventional glass fibre-reinforced composites, mainly, the specific properties. A careful consideration is required when using lightweight composites in harsh environments, such as at humid, high temperature applications as these composites are susceptible to these harsh service conditions. The numerical modelling of natural fibres composites requires special attention as the composites have inherent property variations. With various surface modifications and optimised manufacturing processes already developed, the drawbacks of the composites can be eradicated. Therefore, the key application sectors, such as automotive, construction, marine and aerospace, can use the lightweight composites for making various environmentally friendly components.

114

Sustainable Composites for Lightweight Applications

References Adekomaya, O., 2020. Adaption of green composite in automotive part replacements: discussions on material modification and future patronage. Environ. Sci. Pollut. Control Ser. 27, 8807e8813. Agarwal, J., Sahoo, S., Mohanty, S., Nayak, S.K., 2019. Progress of novel techniques for lightweight automobile applications through innovative eco-friendly composite materials: a review. J. Thermoplast. Compos. Mater. 63, 1e36. Ahmad, F., Choi, H.S., Park, M.K., 2014. A review: natural fiber composites selection in view of mechanical, light weight, and economic properties. Macromol. Mater. Eng. 300 (1), 10e24. Akampumuza, O., Wambua, P.M., Ahmed, A., Li, W., Qin, X.H., 2017. Review of the applications of biocomposites in the automotive industry. Polym. Compos. 38 (11), 2553e2569. Akil, H.M., Cheng, L.W., Ishak, Z.A.M., Bakar, A.A., Rahman, M.A.A., 2009. Water absorption study on pultruded jute fibre reinforced unsaturated polyester composites. Compos. Sci. Technol. 69 (11e12), 1942e1948. Akil, H.M., Santulli, C., Sarasini, F., Tirillo, J., Valente, T., 2014. Environmental effects on the mechanical behaviour of pultruded jute/glass fibre-reinforced polyester hybrid composites. Compos. Sci. Technol. 94, 62e70. Amiri, A., Krosbakken, T., Schoen, W., Theisen, D., Ulven, C.A., 2018. Design and manufacturing of a hybrid flax/carbon fiber composite bicycle frame. Proc. Inst. Mech. Eng. P J. Sports Eng. Technol. 232 (1), 28e38. Andrew, J.J., Srinivasan, S.M., Arockiarajan, A., Dhakal, H.N., 2019. Parameters influencing the impact response of fibre-reinforced polymer matrix composite materials: A critical review. Comp. Struct. 2019, 111007. Arjmandi, A., Ismail, A., Hassan, A., Bakar, A.A., 2017. Effects of ammonium polyphosphate content on mechanical, thermal and flammability properties of kenaf/polypropylene and rice husk/polypropylene composites. Construct. Build. Mater. 152, 484e493. Asdrubali, F., D’Alessandro, F., Schiavoni, S., 2015. A review of unconventional sustainable building insulation materials. Sustain. Mater. & Technol. 4, 1e17. Awaja, F., Zhang, S., Tripathi, M., Nikiforov, A., Pugno, N., 2016. Cracks, microcracks and fracture in polymer structures: formation, detection, autonomic repair. Prog. Mater. Sci. 83, 536e573. Aziz, S.H., Ansell, M.P., 2004. The effect of alkalisation and fibre alignment on the mechanical and thermal properties of kenaf and hemp bast fibre composites: part 2- cashew nut shell liquid matrix. Compos. Sci. Technol. 64 (9), 1231e1238. Azwa, Z.N., Yousif, B.F., Manalo, A.C., Karunasena, W., 2013. A review on the degradability of polymeric composites based on natural fibres. Mater. Des. 47, 424e442. Baley, C., 2002. Analysis of the flax fibres tensile behaviour and analysis of the tensile stiffness increase. Compos. Appl. Sci. Manuf. 33, 939e948. Barbiere, R., Touchard, F., Chocinski-Arnault, L., Mellier, D., 2020. Influence of moisture and drying on fatigue damage mechanisms in a woven hemp/epoxy composite: acoustic emission and micro-CT analysis. Int. J. Fatig. 136, 1e13. Bauroni, A., Dhakal, H.N., 2019. Damage investigation and assessment due to low-velocity impact on flax/glass hybrid composite plates. Compos. Struct. 226, 1e11. Becker, A.A., Karamanlidis, D., 2004. An introductory guide to finite element analysis. Appl. Mech. Rev. 57 (4), 19e20.

Lightweight composites, important properties and applications

115

Bhattacharyya, D., Kim, N.K., 2017. Flammability of natural fiber-reinforcement polymeric composites. In: 21st International Conference on Composite Materials, 20-25th August 2017, Xi’an: China. Bismarck, A., Aranberri-Askargorta, I., Springer, J., Lampke, T., Wielage, B., Stamboulis, A., et al., 2002. Surface characterization of flax, hemp and cellulose fibers; surface properties and the water uptake behavior. Polym. Compos. 23 (5), 872e894. Bourmaud, A., Beaugrand, J., Shah, D., Placet, V., Baley, C., 2018. Towards the design of highperformance plant fibre composites. Prog. Mater. Sci. 97, 347e408. Burger, N., Laachachi, A., Ferriol, M., Lutz, M., Toniazzo, V., Ruch, D., 2016. Review of thermal conductivity in composites: mechanisms, parameters and theory. Prog. Polym. Sci. 61, 1e28. Cheng, M., Zhong, Y., Kureemund, U., Cao, D., Hu, H., Lee, H.P., et al., 2020. Environmental durability of carbon/flax fiber hybrid composites. Compos. Struct. 234, 1e12. Chiu, H.T., Chiu, S.H., Jeng, R.E., Chung, J.S., 2000. A study of the combustion and fireretardance behaviour of unsaturated polyester/phenolic resin blends. Polym. Degrad. Stabil. 70, 505e514. Dayo, A.Q., Gao, B., Wang, J., Liu, W., Derradji, M., Shah, A.H., et al., 2017. Natural hemp fiber reinforced polybenzoxazine composites: curing behavior, mechanical and thermal properties. Compos. Sci. Technol. 144, 114e124. de Vasconcellos, D.S., Touchard, F., Chocinski-Arnault, L., 2014. Tensionetension fatigue behaviour of woven hemp fibre reinforced epoxy composite: a multi-instrumented damage analysis. Int. J. Fatig. 59, 159e169. Dhakal, H.N., Arumugam, V., Aswinraj, A., Santulli, C., Zhang, Z.Y., Lopez-Arraiza, A., 2014b. Influence of temperature and impact velocity on the impact response of jute/UP composites. Poly. Test. 35, 10e19. Dhakal, H.N., Ghasemnejad, H., Zhang, Z.Y., Ismail, S.O., Arumugam, V., 2019. The post impact response of flax/UP composite laminates under low velocity impact loading. Int. J. Dam. Mech. 28 (2), 183e199. Dhakal, H.N., Ismail, S.O., Beaugrand, J., Zhang, Z., Zekonyte, J., 2020. Characterization of nano-mechanical, surface and thermal properties of hemp fibre-reinforced polycaprolactone (HF/PCL) biocomposites. Appl. Sci. 10 (2636), 1e16. Dhakal, H.N., MacMullen, J., Zhang, Z., 2015. In: Graham-Jones, J., Summerscales, J. (Eds.), Marine applications of advanced fibre-reinforced composites, 5. Woodhead Publishing. Chapter, pp. 103e122. Dhakal, H.N., Sain, M., 2019. Enhancement of mechanical properties of flax-epoxy composite with carbon fibre hybridisation for lightweight applications. Materials 13 (1), 1e15. Dhakal, H.N., Skrifvars, M., Adekunle, A., Zhang, Z.Y., 2014a. Falling weight impact response of jute/methacrylated soybean oil bio-composites under low velocity impact loading. Comp. Sci. Tech. 92, 134e141. Dhakal, H.N., Zhang, Z.Y., Bennett, N., 2012. Influence of fibre treatment and glass fibre hybridisation on thermal degradation and surface energy characteristics of hemp/unsaturated polyester composites. Compos. B Eng. 43, 2757e2761. Dhakal, H.N., Zhang, Z.Y., Richardson, M.O.W., 2007a. Effect of water absorption on the mechanical properties of hemp fibre reinforced unsaturated polyester composites. Compos. Sci. Technol. 67, 1674e1683. Dhakal, H.N., Zhang, Z.Y., Richardson, M.O.W., 2009. Creep behaviour of hemp fibre reinforced unsaturated polyester composites. J. Biobased Mater. Bioenergy 3, 1e6.

116

Sustainable Composites for Lightweight Applications

Dhakal, H.N., Zhang, Z.Y., Richardson, M.O.W., Errajhi, O.A.Z., 2007b. The low velocity impact response of non-woven hemp reinforced unsaturated polyester composites. Compos. Struct. 81, 559e567. Dorez, G., Taguet, A., Ferry, L., Lopez-Cuesta, J.M., 2013. Thermal and fire behavior of natural fibers/PBS biocomposites. Polym. Degrad. Stabil. 98, 87e95. ECCC e Environment and Climate Change Canada, 2016a. Ecological science approach: ecological risk classification of organic substances. https://www.ec.gc.ca/ese-ees/default. asp?lang¼En&n¼A96E2E98- 1. Espert, A., Vilaplana, F., Karlsson, S., 2002. Comparison of water absorption in natural cellulosic fibres from wood and one-year crops in polypropylene composites and its influence on their mechanical properties. Compos. Appl. Sci. Manuf. 35 (11), 1267e1276. Faruk, O., Bledzki, A.K., Fink, H.P., Sain, M., 2014. Progress report on natural fiber reinforced composites. Macromol. Mater. Eng. 299 (1), 9e26. Faruk, O., Bledzki, A.K., Fink, H.P., Sain, M., 2012. Biocomposites reinforced with natural fibers: 2000e2010. Prog. Polym. Sci. 37, 1552e1596. Fateh, T., Richard, F., Batiot, B., Rogaume, T., Luche, J., Zaida, J., 2016. Characterization of the burning behavior and gaseous emissions of pine needles in a cone calorimeterdFTIR apparatus. Fire Saf. J. 82, 91e100. Gamstedt, E.K., Berglund, L.A., Peijs, T., 1999. Fatigue mechanisms in unidirectional glassfibre-reinforced polypropylene. Compos. Sci. Technol. 59 (5), 759e768. Gou, J.H., Tang, Y., 2011. Flame retardant polymer nanocomposites. In: Leng, J., Lau, A.K. (Eds.), Multifunctional Polymer Nanocomposites, vols. 309e336. Taylor and Francis Group: CRC Press. Greco, A., Musardo, C., Maffezzoli, A., 2007. Flexural creep behaviour of PP matrix woven composites. Compos. Sci. Technol. 67 (6), 1148e1158. Gurunathan, T., Mohanty, S., Nayak, S.K., 2015. A review of the recent developments in biocomposites based on natural fibres and their application perspectives. Compos. Appl. Sci. Manuf. 77, 1e25. Habibi, M., Laperriere, L., Hassanabadi, H.M., 2019. Effect of moisture absorption and temperature on quasi-static and fatigue behavior of nonwoven flax epoxy composite. Compos. B Eng. 166, 31e40. Hadiji, H., Assarar, M., Zouari, W., Pierre, F., Behlouli, K., Zouari, B., Ayad, R., 2020. Damping analysis of nonwoven natural fibre-reinforced polypropylene composites used in automotive interior parts. Polym. Test. 89, 1e9. Han, Z., Fina, A., 2011. Thermal conductivity of carbon nanotubes and their polymer nanocomposites: a review. Prog. Polym. Sci. 36 (7), 914e944. Hashin, Z., Shtrikman, S., 1963. A variational approach to the theory of elastic behaviour of multiphase materials. J. Mech. Phys. Solid. 11, 127e140. He, S., Hu, Y., Song, L., Tang, Y., 2007. Fire safety assessment of halogen-free flame retardant polypropylene based on cone calorimeter. J. Fire Sci. 25 (2), 109e118. Hornsby, P.R., Hinrichsen, E., Tarverdi, K., 1997. Preparation and properties of polypropylene composites reinforced with wheat and flax straw fibres, Part I. Fibre characterisation. J. Mat. Sci. 32, 443e449. Hussain, A., Calabria-Holley, J., Lawrence, M., Jiang, Y., 2019. Hygrothermal and mechanical characterisation of novel hemp shiv based thermal insulation composites. Construct. Build. Mater. 212, 561e568. Idumah, C.I., Hassan, A., 2016. Characterisationand preparation of conductive exfoliated graphene nanoplatelets kenaf fibre hybrid polypropylene composites. Synth. Met. 212, 91e104.

Lightweight composites, important properties and applications

117

Jeannin, T., Gabrion, X., Ramasso, E., Placet, V., 2019. About the fatigue endurance of unidirectional flax-epoxy composite laminates. Compos. B Eng. 165, 690e701. Kalaprasad, G., Pradeep, P., Mathew, G., Pavithran, C., Thomas, S., 2000. Thermal conductivity and thermal diffusivity analyses of low-density polyethylene composites reinforced with sisal, glass and intimately mixed sisal/glass fibres. Compos. Sci. Technol. 60 (16), 2967e2977. Kandola, B.K., Nazare, S., 2007. An overview of flammability and fire retardancy of fibre reinforced unsaturated polyester composites. In: Proceedings of Fibre Reinforced Composites Conference, Port Elizabeth, South Africae. Kenned, J.J., Sankaranarayanasamy, K., Binoj, J.S., Chelliah, S.K., 2020. Thermo-mechanical and morphological characterization of needle punched non-woven banana fiber reinforced polymer composites. Compos. Sci. Technol. 185, 1e13. Kern, W.T., Kim, W., Argento, A., Lee, E.C., Mielewski, D.F., 2016. Finite element analysis and microscopy of natural fiber composites containing microcellular voids. Mater. Des. 106, 285e294. Khomenko, A., Karpenko, O., Koricho, E., Haq, M., Cloud, G., Udpa, L., 2017. Quantitative comparison of optical transmission scanning with conventional techniques for NDE of impact damage in GFRP composites. Compos. B Eng. 123, 92e104. Kim, H., Welch, D.A., Kedward, K.T., 2003. Experimental investigation of high velocity ice impacts on woven carbon/epoxy composite panels. Compos. Appl. Sci. Manuf. 34 (1), 25e41. Kim, N.K., Mao, N., Lin, R., Bhattacharyya, D., van Loosdrecht, M.C.M., Lin, Y., 2020. Flame retardant property of flax fabrics coated by extracellular polymeric substances recovered from both activated sludge and aerobic granular sludge. Water Res. 170, 1e8. Kim, N.K., Lin, R.J.T., Bhattacharyya, D., 2015. Effects of wool fibres, ammonium polyphosphate and polymer viscosity on the flammability and mechanical performance of PP/ wool composites. Polym. Degrad. Stabil. 119, 167e177. Kozlowski, M., Iwanczuk, A., Frackowiak, S., Szczurek, T., 2007. Biocomposites of functional properties. In: International Sustainable Materials, Polymers & Composites Conference, Warwick. Liang, S., Gning, P.B., Guillaumat, L., 2012. A comparative study of fatigue behaviour of flax/ epoxy and glass/epoxy composites. Compos. Sci. Technol. 72 (5), 535e543. Liu, J., Zhang, Z.Q., Zhang, G.Y., 2015. A smoothed finite element method (S-FEM) for largedeformation elastoplastic analysis. Int. J. Comput. Methods 12 (4), 1e26. Madsen, B., 2004. Properties of Plant Fibre Yarn Polymer Composites. PhD Thesis. Department of Civil Engineering, Technical University of Denmark, Denmark. Series R: No 082. Mahboob, Z., Bougherara, H., 2018. Fatigue of flax-epoxy and other plant fibre composites: critical review and analysis. Compos. Appl. Sci. Manuf. 109, 440e462. Maqsood, M., Langensiepen, F., Seide, G., 2020. Investigation of melt spinnability of plasticized polylactic acid biocomposites-containing intumescent flame retardant. J. Therm. Anal. Calorim. 139, 305e318. Mark, K., Fam, A., 2019. Performance of flax-FRP sandwich panels exposed to different ambient temperatures. Construct. Build. Mater. 219, 121e130. Methacanon, P., Weerawatsophon, U., Sumransin, N., Prahsarn, C., Bergado, D.T., 2010. Properties and potential application of the selected natural fibers as limited life geotextiles. Carbohydr. Polym. 82, 1090e1096. Ming, Y., Chi, H., Blaser, R., Xu, C., Yang, J., Veenstra, M., et al., 2015. Anisotropic thermal transport in MOF-5 com-posites. Int. J. Heat Transf. 82, 250e258.

118

Sustainable Composites for Lightweight Applications

Mohanty, A.K., Vivekanandhan, S., Pin, J.M., Misra, M., 2018. Composites from renewable and sustainable resources: challenges and innovations. Science 362 (6414), 536e542. Monzon, M.D., Paz, R., Verdaguer, M., Suarez, L., Badallo, P., Ortega, Z., et al., 2019. Experimental analysis and simulation of novel technical textile reinforced composite of banana fibre. Materials 12 (7), 1e16. Mori, T., Tanaka, K., 1973. Average stress in matrix and average elastic energy of materials with misfitting inclusions. Acta Metall. 21 (5), 571e574. Morreale, M., Mistretta, M.C., Fiore, V., 2017. Creep behavior of poly(lactic acid) based biocomposites. Materials 10 (4), 1e12. Mounika, M., Ramaniah, K., Ratna Prasad, A.V., Rao, K.M., Reddy, K.H.C., 2012. Thermal conductivity characterization of bamboo fiber reinforced polyester composite. J. Mater. Environ. Sci. 3 (6), 1109e1116. Norouzi, M., Zare, Y., Kiany, P., 2015. Nanoparticles as effective flame retardants for natural and synthetic textile polymers: application, mechanism, and optimization. Polym. Rev. 55 (3), 531e560. Nunez, A.J., Marcovich, N.E., Aranguren, M.I., 2004. Analysis of the creep behavior of polypropylene e woodflour composites. Polym. Eng. Sci. 44 (8), 1594e1603. Oksmann, K., 2000. Mechanical properties of natural fibre mat reinforced thermoplastic. Appl. Compos. Mater. 7, 403e414. Park, B.D., Balatinez, J.J., 1998. Short term flexural creep behavior of wood-fiber/polypropylene composites. Polym. Compos. 19 (4), 377e382. Paturel, A., Dhakal, H.N., 2020. Influence of water absorption on the low velocity falling weight impact damage behaviour of flax/glass reinforced vinyl ester hybrid composites. Molecules 25 (2), 1e16. Peças, P., Carvalho, H., Salman, H., Leite, M., 2018. Natural fibre composites and their applications: a review. J. Compos. Sci. 2 (66), 1e22. Pervaiz, M., Panthapulakkal, S., Birat, K.C., Sain, M., Tjong, J., 2016. Emerging trends in automotive light- weighting through novel composite materials. Mater. Sci. Appl. 7 (1), 26e38. Pil, L., Bensadoun, F., Pariset, J., Verpoest, I., 2016. Why are designers fascinated by flax and hemp fibre composites? Compos. Appl. Sci. Manuf. 83, 193e205. Qi, W., Yao, W., Shen, H., 2020. A bi-directional damage model for matrix cracking evolution in composite laminates under fatigue loadings. Int. J. Fatig. 134, 1e7. Richardson, M.O.W., Wisheart, M.J., 1996. Review of the low velocity impact properties of composites materials. Compos. Appl. Sci. Manuf. 27 (12), 1123e1131. Robert, M., Roy, R., Benmokrane, B., 2010. Environmental effects on glass fibre reinforced polypropylene thermoplastic composites laminates for structural applications. Polym. Compos. 31 (4), 604e611. Saheb, D.N., Jog, J.P., 1999. Natural fiber polymer composites: a review. Adv. Polym. Technol. 18 (4), 351e363. Schartel, B., Braun, U., Schwarz, U., Reinemann, S., 2003. Fire retardancy of polypropylene/ flax blends. Polymer 44 (20), 6241e6250. Schartel, B., Hull, T.R., 2007. Development of fire-retarded materials e interpretation of cone calorimeter data. Fire Mater. 31, 327e354. Shah, D.U., Schubel, P.J., Clifford, M.J., Licence, P., 2013. Fatigue life evaluation of aligned plant fibre composites through SeN curves and constant-life diagrams. Compos. Sci. Technol. 74, 139e149. Soutis, C., 2005. Carbon fibre reinforced plastics in aircraft construction. Mater. Sci. Eng. 412 (1e2), 171e176.

Lightweight composites, important properties and applications

119

Sun, C.T., Vaidya, R.S., 1996. Prediction of composite properties from a representative volume element. Compos. Sci. Technol. 56 (2), 171e179. Szebenyi, G., Bloßl, Y., Heged€us, G., Tabi, T., Czigany, T., Schledjewski, R., 2020. Fatigue monitoring of flax fibre reinforced epoxy composites using integrated fibre-optical FBG sensors. Compos. Sci. Technol. 199, 1e8. Szolnoki, B., Bocz, K., Soti, P.L., Bodzay, B., Zimonyi, E., Toldy, A., et al., 2015. Development of naturalfibre reinforcedflame retarded epoxy resincomposites. Polym. Degrad. Stabil. 119, 68e76. Taub, A., Moor, E.D., Luo, A., Matlock, D.K., Speer, J.G., Vaidya, U., 2019. Materails for automotive lightweighting. Annu. Rev. Mater. Res. 49, 327e359. Taub, A.I., Krajewski, P.E., Luo, A.A., Owens, J.N., 2007. The evolution of technology for materials processing over the last 50 years: the automotive example. J. Miner. Met. Mater. Soc. 59 (2), 48e57. Thanomslip, C., Hogg, P.J., 2003. Penetration impact resistance of hybrid composites based on commingled yarn fabrics. Compos. Sci. Technol. 63 (3e4), 467e482. UK Automotive Sustainability Report, twentieth ed., 2019 2018 data. Varela-Rizo, H., Weisenberger, M., Bortz, D.R., Martin-Gullon, I., 2010. Fracture toughness and creep performance of PMMA composites containing micro and nanosized carbon filaments. Compos. Sci. Technol. 70 (7), 1189e1195. Voigt, W., 1889. About the relationship between the two elastic constants of isotropic bodies. Ann. Phys. 38, 573e587. Wang, F., Cai, X., 2019. Improvement of mechanical properties and thermal conductivity of carbon fibre laminated composites through depositing graphene nanoplatelets on fibres. J. Mater. Sci. 54, 3847e3862. Wang, W., Chouw, N., Jayaraman, K., 2016. Effect of thickness on the impact resistance of flax fibre-reinforced polymer. J. Reinforc. Plast. Compos. 35 (17), 1277e1289. Wang, Z., Qi, R., Wang, J., Qi, S., 2015. Thermal conductivity improvement of epoxy composite filled with expanded graphite. Ceram. Int. 41 (10), 13541e13546. Wielage, B., Lampke, T., Max, G., Nestler, K., Starke, D., 1999. Thermogravemetric and differential scanning calorimetric analysis of natural fibres and polypropylene. Thermomochimicia Acta 337 (1e2), 169e177. Xiong, X., Shen, S.Z., Hua, L., Liu, J.Z., Li, X., Wan, X., et al., 2018. Finite element models of natural fibers and their composites: a review. J. Reinforc. Plast. Compos. 37 (9), 617e635. Yan, L., Kasal, B., Huang, L., 2016. A review of recent research on the use of cellulosic fibres, their fibre fabric reinforced cementitious, geo-polymer and polymer composites in civil engineering. Compos. B Eng. 92, 94e132. Yang, F., Nelson, G.L., 2011. Combination effect of nanoparticles with flame retardants on the flammability of nanocomposites. Polym. Degrad. Stabil. 96 (3), 270e276. Yorseng, K., Rangappa, S.M., Pulikkalparambli, H., Seingchin, S., Parameswaranpillai, 2020. Accelerated weathering studies of kenaf/sisal fiber fabric reinforced fully biobased hybrid bioepoxy composites for semi-structural applications: morphology, thermo-mechanical, water absorption behaviour and surface hydrophobicity. Construct. Build. Mater. 235, 270e276. Zafeiropoulos, N.E., Baillie, C.A., Hodgkinson, J.M., 2002. Engineering and characterisation of the interface in flax fibre/polypropylene composite materials. Part II. The effect of surface treatments on the interface. Compos. Appl. Sci. Manuf. 33, 1185e1190.

Design, manufacturing processes and their effects on biocomposite properties

4

4.1 Introduction and context In an attempt to improve the inherent properties of the biocomposite materials, different designs and manufacturing processes have been adopted. These design and manufacturing processes involve the selection of natural fibres or filler (reinforcements) and matrices used, consideration of sizes/part design, processing: fibre surface treatments, the stacking sequences, fibre orientation, compaction methods and curing techniques, among others. Importantly, all these processes determine the properties (mechanical, physical, electrical and thermal) of a particular biocomposite. Therefore, the intended enhancement of biocomposite materials begins from the design stage to the manufacturing phase, as all these stages are carefully monitored to avoid defects and alteration in the expected behaviours. Also, the functionality of biocomposite materials in engineering applications depends on the products of design and manufacturing processes, whereby the biocomposites have undertaken. The design and manufacturing processes of biocomposites have some challenges today. These include, but are not limited to, increasing applications of biocomposites, unpredicted materials responses (mainly non-linear properties), and availability of limited design data, because there are numerous varieties or species of natural fibres and matrices. In addition, the interactions between hydrophilic natural fibres and hydrophobic matrices, as well as presence of variable lumens (the centred holes) in many natural fibres are critical challenges that are confronting design and manufacturing processes of several biocomposites (Rudin and Choi, 2013). These natural limitations have a tendency of creating voids within the biocomposites. The presence of void, probably during the manufacturing process, has undesirable implications on the total quality of the fabricated products. Hence, various design techniques and manufacturing processes will be extensively considered, as well as their influences on the several properties of some commonly used biocomposite materials within this chapter to capture this concept comprehensively.

4.2 Eco-design and sustainability (design for environment and design for manufacturing) 4.2.1 Eco-design There are numerous advantages from the products manufactured from composite materials. One of these benefits includes environmental friendliness due to their

Sustainable Composites for Lightweight Applications. https://doi.org/10.1016/B978-0-12-818316-8.00005-0 Copyright © 2021 Elsevier Ltd. All rights reserved.

122

Sustainable Composites for Lightweight Applications

sustainable sources, biodegradability, renewability and corrosion resistance, as characterised by the natural or biofibres. Also, composite materials have lower weight when compared with some metals, alloys and synthetic fibre. Therefore, the overall weight of the cars is reduced when natural fibre-reinforced polymer (FRP) composites are used in the car or aeroplane components. These components can be car/aeroplane engine cover, body, seat, dashboards and among other interior parts of the cars/aeroplanes. As the overall weight of the car and aeroplanes are reduced, the fuel consumption will be reduced. Hence, it offers fuel savings throughout the lifetime of cars and aeroplanes. This results in a reduction in the quantity of carbon dioxide (CO2) release to the atmosphere, among other toxic gases from fossil fuels. Consequently, a better and cleaner environment is maintained. Also, natural fibres provide good acoustic insulating properties due to their hollow structure. Therefore, noise pollution can be reduced. In recent years, the use of natural FRP composite has gained much interest as a low cost, environmentally friendly alternative for the more commonly used reinforcing materials, such as glass and carbon fibres. Cost is one of the essential factors that influence the sustainability of a material. Several natural fibres, especially jute, coir and bamboo, are much cheaper (unit price) when compared with some commonly used synthetic fibres, including carbon and E-glass (Fig. 4.1). This consequently reduces the total cost of producing natural FRP composites. There has been increasing pressure from both nationally and internationally users, designers and manufacturers to lessen the environmental damage caused by the use of non-renewable materials by using more sustainable and environmentally friendly materials as reinforcements in the composite manufacturing. This has resulted in many manufacturers searching and adopting preventive measures to seek and minimise the environmental impact of composites after their service life (Dhahak et al., 2013, 2014; Thakur et al., 2014; Rodi et al., 2016).

$12.00

Cost range

Cost per weight ($/Kg)

$10.00

$8.00

$6.00

$4.00

$2.00

$0.00 Carbon E-glass

Abaca Bamboo

Coir

Cotton

Flax

Hemp

Jute

Kenaf

Ramie

Sisal

Figure 4.1 Difference between natural and synthetic fibres in terms of cost per weight (Lotfi et al., 2019).

Design, manufacturing processes and their effects on bio-composite properties

123

Natural fibres, as a main constituent/reinforcement of biocomposites, are abundantly available, sustainable and biodegradable. Natural fibres include, but are not limited to, hemp, jute, sisal, bamboo, kenaf, rice husk and straw, wheat straw, oil palm, henequen, curaua, olive pit/husk, coir, pineapple leaf, ramie, choir, abaca, flax, date palm, banana, pineapple and bagasse (fibrous residue of sugarcane stalk). They have some outstanding mechanical properties. For example, hemp fibres have better tensile strength at break of 550e1110 MPa and tensile modulus of 58e70 GPa when compared with other naturally available plant fibres: date palm, jute, flax, to mention but a few (Bledzki and Gassan, 1998; Dhakal et al., 2007; Mohanty et al., 2002; Bourmaud et al., 2017). Some of the mechanical properties of natural fibres are quite comparable or better than that of synthetic fibres (such as E-glass), as shown in Fig. 4.2 and comprehensively reported by Pickering et al. (2016). These significant properties of natural FRP biocomposites have greatly increased their areas of engineering application, as designs for both environment and manufacturing are carefully considered starting from the initiation journey of design concept. More also, although it is well understood that the processing of natural FRP biocomposites is safer than that of synthetic FRP composites with respect to health and the environment, there are a lot of environmental and sustainability issues still associated with the design and manufacturing processes of biocomposite materials today. These issues are chronologically discussed hereafter.

4.2.2 Sustainability Sustainability is one of the key developmental concepts of the next generation of products (materials) and processes (manufacturing). This gave birth to the advent of sustainable, eco-efficient, biodegradable and environmentally friendly biocomposite materials as a suitable substitute to the synthetic, non-renewable fibre-reinforced polymer composites. Sustainability encompasses land use, energy efficiency, resource availability, soil conversation, biodiversity, environmental impact and impact on social community (Quarshie and Carruthers, 2014). Therefore, the design for biocomposite materials should improve biodiversity, land use and energy efficiency, as

100

E-modulus (GPa)

Specific modulus

80 60 40 20 0 E-glass

Hemp

Flax

Figure 4.2 Comparison of some natural fibres with synthetic (E-glass) fibre (Mohanty et al., 2002).

124

Sustainable Composites for Lightweight Applications

well as conserve soil and resources available, reduce impact on both environment and social community, as consumer demands and satisfaction are met through design and manufacture of biocomposite materials. Also, biomass is available annually in a large quantity, nearly 70% of these crops, for energy use (Wool and Sun, 2005). Some are used as energy crops, crop residues and biogas, while the remaining proportion ends up as wastes. Soil quality is maintained with some residues that returned to the land. Some animal manures or beddings are produced from some of these crop residues. Biocomposite materials, with bio-based constituent(s), have been considered as a solution to the fast reducing petroleum supply and growing environmental threat (Mohanty et al., 2002). These natural constituents are readily available all over the world. They are available at a very large quantity and probably, throughout a year in thousands of tones (Lotfi et al., 2019). They are renewable, recyclable and biodegradable. Biocomposites with both bio-fibre(s) and bio-based matrix, commonly called ‘green composites’ are biodegradable; prone to environmental and microbial degradation after disposal, without having an adverse environmental effect. Hence, they are environmentally acceptable and commercially viable for those who are engaging in planting/farming of these natural fibres, as shown in Fig. 4.3(a). Also, some of these bio-based materials (fibres and matrices) are originated from plants. The photosynthesis reaction enables plants to maintain carbon dioxide neutrality, as they respire (take in) carbon dioxide. Therefore, the quantity of carbon dioxide in the atmosphere is reduced. This process is further explained with the aid of Fig. 4.3(b). The solar energy required by the plants is a sustainable energy source.

4.2.3 Design for environment Design for the environment is an approach towards achieving design goals without jeopardising or affecting the environment and human health throughout the stages of the product or process life cycle. It aims at improving product quality and cost, as well as minimising or eliminating all environmental impacts of a product over its life cycle. It involves five different product (biocomposite) life cycle stages: materials, production, distribution, use and recovery, as depicted in Fig. 4.4. All the processes involved in each of these stages are properly designed and monitored to maintain or protect a healthy environment. An environment includes the earth, and both atmosphere and hydrosphere, where plants and animals exist. Also, the selection of non-renewable resources and the release of toxic or inorganic substances to the environment during the five stages of design for the environment must be avoided, especially after use and recovery stages, as illustrated in Fig. 4.4.

4.2.3.1 Materials Basically, composites are desirable products of the combination of reinforcements (fibres/fillers) and binders (matrices), designed and produced for a specific engineering application. Biocomposites are composite materials of either biofibre or bio-based matrix, or both. The main reinforcing elements or materials of biocomposites are from natural sources, as bio-fibres/fillers. Some fibres are originated from plants/vegetables, while others are from either animals or minerals. Examples of fibres under each of the

Design, manufacturing processes and their effects on bio-composite properties

125

(a) Renewable

Recyclable

Triggered biodegradable

Environmental acceptability & commercial viability

Sustainable

(b) Moulding

Plastic converters

Biopolymeric materials

Trays, spoons, etc

Bio Biowaste collection

Innovation: polymer production Composting Renewable resources (cellulose, corn, etc.)

2

CO Photosynthesis

Figure 4.3 (a) Sustainability of the bio-based materials for biocomposites and (b) carbon dioxide sequestration (Mohanty et al., 2002).

aforementioned sources have been discussed in the previous chapters. Additional examples of natural sources of biofibres have been reported (Sanjay et al., 2015, 2018; Akil et al., 2011). Furthermore, the matrix can be classified into petrochemicalbased and bio-based. Petrochemical-based matrices are thermosetting and thermoplastic, while examples of bio-based matrices include wheat, gluten, soybean, starch, gelatin, polyhydroxybutyrate (PHB) and polylactic acid (PLA), among others (Lotfi et al., 2019; Sanjay et al., 2015). It is very important to ensure that the depletion of natural resources and deforestation are avoided in searching for biomaterials for biocomposite manufacturing. Biofibres and bio-based matrices are renewable and abundant resources (Table 4.1), most of them are very compactable to produce an improved fully biocomposite material. Also, some are non-hazardous and environmentally friendly with respect to health issues. Some of their wastes are water-based and biodegradable. These factors are very germane when designing biocomposite materials for good environments. This aspect has been extensively discussed in the previous Chapter two.

126

Sustainable Composites for Lightweight Applications Non-renewable resources

Resources

Post-industrial recycling

Materials

Renewable resources

Post-consumer recycling Natural decay

Natural “biological” life cycle Toxics

Recovery

Organics

Deposit

Production Remanufacturing

Product “industrial” life cycle

Distribution

Reuse

Inorganics

Use

Figure 4.4 Biocomposite double life cycles (Ulrich and Eppinger, 2012).

In addition, toxic materials (fibre and matrix) must be eliminated or carefully handled. Care must be taken to avoid toxic raw materials; in process, use and after use (Fig. 4.4). There are some materials that are not originally toxic, but they become harmful after they have reacted with fluids (liquids and gases), heat and ultraviolet rays from sunlight. These potentially harmful materials must be carefully considered during the design of biocomposites. Also, the use of raw materials should be reduced. The design of biocomposite materials to be used as a compartment/part of a system should encourage not too many raw materials and quantity. During materials selection, the use of recovered and recycled materials should be welcomed. This is required to reduce discards, minimise waste, time and energy during the production stage.

4.2.3.2 Production In this stage, there must be a reduction in the use of process materials. Process materials that can be fully recovered and recycled must be specified and adopted during the production phase of the design for the environment. Similar to the material stage, the toxic process materials must be eliminated. Also, processes with high energy efficiency should be selected. The use of clean and renewable (solar, water and wind) sources of energy must be encouraged or preferred during the production stage, rather than fossil fuels. The purpose of these preferences is to keep the working and general environments safe and clean from toxic fumes, gases and hazardous noises. Air and water pollution from industrial emissions and discharges, as well as waste generated during the production of biocomposites must be either well controlled or avoided. For example, processing of natural (hemp) fibres have a very good environmental impact, especially when compared with a synthetic or conventional fibres, such as glass. A comparison of hemp fibre with glass fibre is presented in Table 4.2. During

Design, manufacturing processes and their effects on bio-composite properties

127

Table 4.1 Yearly production of various important natural fibres, their origins, species and largest producer countries (Lotfi et al., 2019). Fibre type

Origin

Species

Largest producer countries

World production (103 tons)

Coir

Fruit

Cocos nucífera

India, Vietnam, Sri Lanka

100

Kenaf

Stem

Hibtscus connabtnus

India, Bangladesh, United States

970

Flax

Stem

Linum usitatissimum

Canada, France, Belgium

830

Bamboo

Stem

(>1250 species)

China, India, Indonesia

30,000

Abaca

Leaf

Muso textilis

Philippines, Ecuador, Costa Rica

70

Jute

Stem

Corchorus capsularis

India, Bangladesh

2500

Sisal

Leaf

Agave sisolana

Tanzania, Brazil, Kenya

378

Ramie

Stem

Boehmeria nivea

China, Brazil, Philippines

100

Cotton

Seed

Gossypium sp.

China, India, United States

25,000

Banana

Leaf

Musa indica

Brazil, India

200

Silk

Animal

Silkworms, honeybee

China, India, Europe

202

Wool

Animal

Sheep, alpaca, camel

Australia, New Zealand, China

2000

Hemp

Stem

Cannabis sativa

China, France, Philippines

215

Pineapple

Leaf

Ananas comosus

Philippines, Thailand, Indonesia

74

Agave

Leaf

Agave fourcroydes

Columbia, Cuba, Mexico

56

Kapok

Fruit

Ceiba pentandra

China, India

316

Bagasse

Stem

d

Brazil, India, China

75,000

128

Sustainable Composites for Lightweight Applications

Table 4.2 Comparison of environmental impacts during the production of 1 kg of glass and hemp fibres (Shahzad, 2011). Fibres Parameters

Hemp

Glass

Power consumption (MJ)

3.4

48.3

CO2 emission (kg)

0.64

20.4

SOx emission (g)

1.2

8.8

NOx emission (g)

0.95

2.9

BOD (mg)

0.265

1.75

their processing stages, lesser energy is required when compared to the synthetic (carbon and glass) FRP composites. Additionally, the production of controllable scraps and wastes should be reduced. More also, the production technique must be applied to minimise the total volume of materials used. The production process must be well planned in order to avoid rejects and reduce material waste, using clean, highly efficient production processes and few manufacturing steps as much as possible.

4.2.3.3 Distribution After the first two stages (materials and production) of design for the environment have been successfully accomplished, there is need for the biocomposite materials to be transported to the product manufacturing companies, such as automobile, aerospace, telecommunication, power/energy, sports/games or recreation, marine/naval, to mention but a few, where they will be used or processed to products. In another way, biocomposite products are also required to be moved to the market places. Therefore, it is a part of the design for the environment to ensure that the most effective and energy-efficient shipping plan is embraced and emission from transport is reduced. Both air pollution from transportation emissions and waste generation from packaging for distribution must be well controlled or avoided. In addition, harmful and hazardous packaging materials must be eliminated. If it is possible, packing should be eliminated; if not, reuse packaging materials should be encouraged. This should be considered during biocomposite design. The concept of lightweight biocomposite is also a good factor to be considered, with ideas of folding, disassembly or nesting to distribute biocomposite materials for use.

4.2.3.4 Use As a part of the design for the environment, it is of great importance to ensure that the useful biocomposite life is extended. During the use of biocomposite products, discharges of harmful substances should be eliminated and energy consumption should be reduced. For instance, the use of lightweight biocomposite materials in transportation industries to manufacture vehicles and aircraft has significantly reduced the

Design, manufacturing processes and their effects on bio-composite properties

129

amount of fuel consumption. Therefore, it has saved our environment from being attacked with excessive carbon dioxide, sulphur dioxide and other toxic gases that are released from the combustion of fossil fuel and their effects on ozone layers and ecosystem at large. Biocomposite products must be used under intended conditions, under clean and efficient servicing operations. Also, minimal maintenance is required when biocomposite materials are in use. They do not require painting or coating to prevent rusting, because they possess good corrosion and wear resistance properties, unlike metals and some alloys. Biocomposite materials can be reused and recycled, and therefore, biocomposite materials with excellent recovery should be encouraged. This will lead to the final stage of the design for the environment, known as recovery, as subsequently discussed.

4.2.3.5 Recovery The design of biocomposite materials involves natural fibres or bio-based matrix or both that can be recycled continually, without compromise quality and effective performance. These natural materials or constituents of biocomposites can completely return to the earth’s natural cycles to create new natural materials for sustainability. This concept has been fully and earlier explained, using Fig. 4.4. During recovery, incompatible materials must be easily separated. Easy disassembly of biocomposite parts should be considered during the design stage, either by using hands or simple tools. The biocomposite components must be designed to support easy recovery and manufacturing. It is expected that the waste volume for landfill deposits, as well as incineration, is reduced. Fig. 4.5 further illustrates a complete life cycle of biocomposite materials, with emphasis on recovery.

4.2.4 Design for manufacture The possibility of using biocomposite materials in many areas of engineering application has emphasised the need for design for manufacture. Biocomposite materials are manufactured using different processes: hand lay-up, compression moulding and extrusion, to mention but a few. These processes and their effects on the properties of various biocomposite materials are extensively discussed in the next section. The expected properties and intended areas of application of a biocomposite are important determinant factors to be considered during design for its manufacture. Close to these factors is the geometry of the biocomposite products, among others. Importantly, Mohanty et al. (2002) proposed a tri-corner approach in designing for the manufacture of biocomposites with excellent or desirable strength: (i) effective, (ii) low cost and efficient biofibre treatment, effective blending and functional matrix modification, and (iii) choice of efficient biocomposite processing conditions, as illustrated schematically in Fig. 4.6. A deeper discussion on each of these approaches can be found in the subsequent sub-sections and other chapters. Moreover, it is expected that biocomposite design for manufacture supports a reduction in the costs of manufacturing, components, assembly, supporting production and considers the impact of design for manufacturing decisions on other unavoidable

130

Sustainable Composites for Lightweight Applications

Recycling

Oil production

Incineratrion Refinery

Eco-Cycle

Biorefinery

Biodegradation

Green chemistry

Biodeterioration Biofragmentation Assimilation Biomass Emissions of VOCs nanoparticle release degradation during service life

Renewable raw material

Figure 4.5 Detailed life cycle of sustainable biocomposites (Vilaplana et al., 2010).

Bio-composite processings

Efficient biofibre surface treatments

Synergism

Matrix polymer modification

High performance bio-composite formulation

Figure 4.6 Proposed tri-corner approach to manufacture high efficient biocomposites (Mohanty et al., 2002).

Design, manufacturing processes and their effects on bio-composite properties

131

factors. Recently, design for manufacture has been combined with a design for assembly, as a simultaneous effort. They are jointly referred to as a design for manufacture and assembly (DfMA). DfMA has numerous benefits over the traditional design approach, where many manufacturing processes and post-manufacturing operations are required. A single part of biocomposite materials can be manufactured, just like forged or cast materials (Fig. 4.7(b)), instead of Fig. 4.7(a). A typical example is shown in Fig. 4.8, where an interior carpet of a car’s door is made by hemp fibre-reinforced polyethylene biocomposites. Furthermore, there are numerous advantages of having effective DfMA (single biocomposite part), as shown in both Figs. 4.7(b) and 4.8 over the traditional design approach. These benefits include, but are not limited to, the following: ⁃ ⁃ ⁃ ⁃ ⁃ ⁃ ⁃ ⁃

Reduced costs (of materials, design, manufacturing and assembly). Ease of design, manufacturing and installation/assembly. Not susceptible to failure, hence longer durability is guaranteed. Lower time/effective time management from design to installation/repairs or recycle stages. Use of less equipment, labour and facilities. Effective waste management. Ease of maintenance and repairs. Better aesthetics and reduced weight.

4.3 Manufacturing processes and their influences on properties of bio-composites Manufacturing processes are the series or stages of techniques undergo by the biocomposites before they are manufactured into semi or finished products for several engineering applications. These processes vary from one biocomposite to another. They determine the properties of the biocomposites. Therefore, they are selected based on the inherent properties of the fibres/fillers bio-fibres/natural or synthetic and matrices/binders (thermoplastics or thermosets), as reinforcing and binding elements, respectively. Also, the following factors determine the type of suitable process for manufacturing of a particular biocomposite: fibre loading, quantity, process parameters (production speed, temperature and pressure), surface treatments, costs, properties (of raw materials and final products), size, the complexity of shape and applications

Figure 4.7 (a) Traditional design approach (11 parts) and (b) DfMA approach (single part).

132

Sustainable Composites for Lightweight Applications

Figure 4.8 Single components containing natural fibres (flax, hemp and sisal) reinforced composite parts used in Mercedez-Benz. (From Holbery, J., Houston, D. Natural-fiberreinforced polymer composites in automotive applications. JOM 58, 80e86 (2006). https://doi. org/10.1007/s11837-006-0234-2).

of the biocomposites, among others. The size of the biocomposite is a leading factor among all these factors (Ho et al., 2012). Both compression and injection moulding processes are suitable for a small to medium-sized components, because of their fast processing cycles and simplicity. Nonetheless, open moulding and autoclave processes are often used to manufacture large structures (Ho et al., 2012). Therefore, suitable manufacturing techniques must be utilised to enhance the interfacial bonding, properties and produce an optimal biocomposite material, without manufacturing-induced defects. There are several methods for manufacturing different biocomposite materials both in laboratory/small and commercial/large scales, but the most common techniques have been extensively and hereafter discussed. These processes include, but are limited to, hand and spray lay-ups, injection moulding, compression moulding, extrusion, resin transfer moulding, filament winding, automated fibre placement, autoclave and out-of-autoclave, as well as additive manufacturing.

4.3.1 Hand and spray lay-ups 4.3.1.1 Hand lay-up Hand lay-up is the most basic, simplest and oldest process of manufacturing biocomposites. This method has been widely used for many years due to its simple principles to operate and teach. However, resin mixing, laminate resin contents and laminate qualities are very dependent on the skills of laminators. This method involves the placement of plies or fabric layers and succeeding application of matrix in the

Design, manufacturing processes and their effects on bio-composite properties

133

mould manually, to produce a biocomposite laminate (Lotfi et al., 2019). In other words, in the hand lay-up method, resins are impregnated into fibres by hand, which are in the form of woven, knitted, stitched or bonded fabrics. This is usually accomplished by rollers or brushes, with the increased use of nip-roller type impregnators for forcing resin into the fabrics by means of rotating rollers and a bath of resin. Laminates are left to cure under standard atmospheric conditions. According to Knoeller (2018), hand lay-up is a process whereby resin material is rolled into the already placed reinforcing fibres in the mould, after a release film and gel coating. This manufacturing process can be divided into four sequential stages, as thus listed. i. ii. iii. iv.

Mould preparation Gel coating Lay-up Curing

A suitable mould must be prepared and used for the hand lay-up process, except the biocomposite will be joined directly to another structure. A mould can be very simple or with some shapes, such as curves and edges. Mould with several shapes must be assembled and dismantled in parts or sections before lay-up and after curing to remove the biocomposites. A certain amount of fibres is cut and placed in the mould in a specific direction(s) before a catalysed resin is added to the well-laid fibres. The mixing and contents of the resin, as well as the quality of the biocomposite laminate, greatly depend on the skills of the operator or designer, and hence, it is a labour-intensive manufacturing process. Thereafter, impregnation of the fibres with resin is done using a roller or brush, as depicted in Fig. 4.9(a), or in a slightly modified configuration (Fig. 4.9(b)). Furthermore, the advantages of this method include simple operation and low cost. It is suitable for creating a corrosion-resistant and structural portion. Hand lay-up process does not involve too much of fibre loading, unlike other processes. It is suitable for fibre-reinforced thermosetting polymer biocomposites and many material (fibre and matrix) types. It accommodates higher fibre contents and longer fibres, when compared with the spray lay-up. An increased natural fibre content significantly increases the stiffness property and improves the impact strength of biocomposite (Arbelaiz et al., 2005b; George et al., 2010). However, both water uptake and odour of biocomposite increase with an increase in the fibre content. It requires less capital and infrastructure when compared with other methods. It supports versatility, the use of longer fibres and low tooling cost. Also, it can be used for several years (durability). It is very easy to operate and has simple principles to teach. These advantages have made hand lay-up attractive application in the transportation sector, to manufacture marine, military and aircraft structures. For instance, it is used for the fabrication of bulletproof composite panels from ramie fibre-reinforced epoxy composites and commonly used to fabricate polymer matrix composite parts for the American aerospace industries (Lei et al., 2006; Knoeller, 2018), big tanks/containers, boat/ship hulls, deck, swimming pools, aircraft and automotive components in transportation industries. Other typical applications of this technique include the manufacturing of standard wind-turbine blades and boats. Several different hybrid fibre-reinforced

134

Sustainable Composites for Lightweight Applications

(a)

Resin

Hand roller Fibreglass reinforcements

Gel coat Mold Release film

(b)

Vacuum regulator

Vacuum pump Breather Vacuum port Non-porous teflon

Pipe

Caul plate

Vacuum bagging

Composite panel

Alumium base table Non-porous teflon for dam structure

Sealant for dam structure

Sealant for vacuum bagging

Figure 4.9 Hand lay-up process, showing (a) a simple illustration and (b) complete set-up with vacuum bagging process (Cucinotta et al., 2017; Hakim et al., 2017).

polymeric biocomposites have been successfully manufactured using hand lay-up technique, with various common matrices: polyester resin (Singh et al., 1995; Ahmed and Vijayarangan, 2008; Athjayamani et al., 2010; Khanam et al., 2010; Ramesh et al., 2013; Isa et al., 2013; Alavudeen et al., 2015), epoxy resin (Venkateshwaran and ElayaPerumal, 2010; Jawaid et al., 2010, 2013; Zhang et al., 2012; Ramnath et al., 2013, 2014; Boopalan et al., 2013; Santulli et al., 2013; Dhakal et al., 2013; Srinivasan et al., 2014; Guermazi et al., 2014; Boroujeni et al., 2014; Yahaya et al., 2014a,b,c, 2015; Dong and Davies, 2015; Gupta and Srivastava, 2016), among others. Nonetheless, the mixing of resin, resin content and quality of laminates are significantly depended on the skills of the laminators. Production rate and the possibility of achieving a high fibre volume fraction with hand lay-up process is less and difficult, respectively, especially in comparison with other automated processes. In addition, this method creates high void contents. It generally supports a low volume production of biocomposites; hence it is not economical. High viscosity resin is difficult to use in

Design, manufacturing processes and their effects on bio-composite properties

135

this technique. The lower viscosity of the resins implies that they have an increased propensity to penetrate clothing, among others. Resins need to be low in viscosity to be workable by hand. This generally compromises their mechanical/thermal properties due to the need for high diluent/styrene levels. To remove the entrapped air or reduce occurrence of voids and pores, and obtain an even distribution of the resin within the fibres, a better fibre-matrix interaction and a desired thickness, rolling, brushing or squeezing of the wet biocomposites must be carried out, using hand rollers.

4.3.1.2 Spray lay-up Spray lay-up method is very similar to the hand lay-up technique. It is an advancement in hand lay-up. The main difference is that a spray gun is used to spread pressurised resin and chopped fibres, as depicted in Fig. 4.10. The matrix material and fibre can be sprayed separately or simultaneously. Other stages of manufacturing are very similar to, if not exactly, the same as the hand lay-up. The spray lay-up process supports high fibre volume fraction in biocomposites and the production of various sizes of components. It is a continuous process. It accommodates several materials, as moulds. Re-spraying can be used to correct errors. Therefore, the spray lay-up method is good for fabricating lower load-carrying components, such as small boats, fairing of trucks, bathtubs, to mention but a few. However, it is sometimes slow, inconsistent and environmentally unfriendly, and it has no control of fibre orientation. Moving forward, natural reinforcements (plant fibres such as sisal hemp, coir, jute, banana, flax, cotton and nettle) in the form of chopped short fibres, particles, flakes and fillers, as well as matrices (polyester, unsaturated polyester, epoxy, phenolic resin, polyvinyl ester and polyurethane resin), among others, are suitable materials for spray lay-up manufacturing process. Fiber Resin catalyst pot Air pressurized resin

Chopper gun

Optional gel coat

Figure 4.10 A simple illustration of a spray lay-up process (Balasubramanian, Sultan, and Rajeswari, 2018).

136

Sustainable Composites for Lightweight Applications

4.3.2 Vacuum bagging moulding A vacuum is used in the vacuum bagging process, as the name implies. In this process, the vacuum is introduced to eliminate unwanted enclosed air, excess resin or gases, as earlier illustrated in Fig. 4.9(b). A complete woven mat or fabric form and matrix layup are covered or sealed up with a non-adhering film of polyvinyl alcohol or nonporous teflon or nylon to create a vacuum bag within the mould (Fig. 4.11). Often, the air under the vacuum bag is sucked by an atmospheric pressure to produce a desirable and sustainable biocomposite laminate after the curing stage at a particular or room temperature has been completed. Also, the advantages of vacuum bagging manufacturing process include, but are not limited to, the following: ⁃ Removal of any air or voids within the resin and reinforcement lay-up, which leads to greater surface finish and mechanical properties. ⁃ Any volatile organic compounds produced in the addition of a catalyst or during the curing process are removed and confined safely. This also helps the aforementioned benefit, because if these volatile organic compounds are removed, they have no effect on the finished biocomposite products. ⁃ Possibility of delamination defect in biocomposite parts is also reduced, as the resin is encouraged to move between the layers and into an absorbent part within the bagging system.

However, the setbacks of this process are as follows: ⁃ Stringent measures are to be taken to ensure that there are no leaks within the vacuum system. This can be tested before introducing resin by turning the vacuum on and clamping the system to seal the vacuum, and the pressure should be noted on the generator and the system left to 10e15 min. Therefore, any reduction in pressure indicates a leak within the system. ⁃ Ensure that a thermally stable sealant tape is used. The tape must remain stable and hold the seal around the edge under curing and exothermic conditions. ⁃ Bridging may occur when there is an intricate or complex part that does not allow the bag to completely press against the part. Pleats may be used to prevent this unwanted condition or the use of pressure pads made from a flexible material, such as rubber, is also encouraged.

Toughened glass mould

Alumium foil

Sealing tape

Composite laminates

Release agent

Peel ply

Perforated relase film

Breather

To vaccum Vacuum hose

Vacuum bag

Figure 4.11 A simple arrangement of the vacuum bagging system (Hang et al., 2015).

Design, manufacturing processes and their effects on bio-composite properties

137

4.3.3 Injection moulding Injection moulding is used to force the measured quantity of the mixture of molten polymer and fibre into a predetermined mould cavity (Fig. 4.12). Initially, the thermoplastic polymer used for an injection moulding process is modified to plastic pellets. Now, the already chopped fibres with pellets are fed separately through a funnelshaped feed hopper into a heated compression barrel to fabricate fibre-reinforced polymer biocomposites. The heated barrel accommodates a rotating screw or screws (in case of the twin-screw extruder). The solid pellets are converted to a viscous liquid through heat produced from interfacial friction between the barrel, pellets and screw (Nystr€ om, 2007). The viscous liquid is required for an easy movement through the sprue nozzle of the injection machine. Therefore, the screw helps to generate heat, develop shear force for mixing the fibres and polymer and functions as a piston to push the biocomposite (fibre-reinforced molten polymer) through sprue nozzle to the close mould cavities. The extruded product is compressed into the closed mould cavities to produce a desirable geometry, after solidification and freezing (cooling) of the fibre orientation and distribution. Finally, the anticipated biocomposite product is ejected from the mould cavity. The compounding process considerably affects the fibre shortening, thermal deterioration and fibrillation at initial stages and the final properties of the biocomposite properties. The following various compounding methods are recommended for injection moulding of natural fibre-reinforced thermoplastics: mixing (cascade mixing), pull-drill process (bast fibres), extruder, pultrusion (bast fibres), pelletising (with matrix) and hybrid fibre non-woven pre-consolidation and cut (Faruk et al., 2012). The modulus distribution of the biocomposites produced by injection moulding is critically affected by the fibre orientation and residual stresses. Residual stresses could cause an earlier fracture of pure thermoplastic polymers and their biocomposites, which significantly affect the mechanical properties and quality (when it causes

Carbon fibre roving

Hopper

Real - time monitoring system

Quantitative feeding system Screw

Mould

Heater

Vent hole

Injection moulding machine

Figure 4.12 Descriptive set-up of the direct fibre feeding injection moulding system (Yan and Cao, 2018).

138

Sustainable Composites for Lightweight Applications

shrinkage, stress cracking, undue deformation and warpage) of the final biocomposite products. Residual stress distribution in the injection moulded components also affect their dimensional accuracy and other structural properties (Kim et al., 2002). An occurrence of fibre attrition is another limitation of the injection moulding process because it decreases the fibre length during processing (Pickering et al., 2016). However, optimal properties of the injection moulded biocomposites are achievable by carefully selecting, monitoring and improving the following three principal parameters: ⁃ Process: Melt and mould temperatures, speeds and pressures of both screw and injection process. The thermal and rheological properties of the matrix depend on the selected temperature and pressure for the moulding process. Too high temperatures can cause fibre degradation. For example, natural fibres need process at lower temperatures, around 240
C (Monteiro et al., 2012), or otherwise, they will degrade. This limits the thermoplastic matrices used, which must possess melting points lower than the degradation temperature of the concerned fibres (Pickering et al., 2016). ⁃ Materials: Molten polymer rheology and reinforcement/fibre type and volume content. The viscosity of the matrix is very germane during injection moulding, and therefore, this process is generally limited to fibre content of less than 40 m% in a biocomposite (Pickering et al., 2016). For example, semi-crystalline polylactic acid exhibited a greater shear viscosity than that of amorphous polylactic acid (Fang and Hanna, 1999). ⁃ Geometric: Size and shape of the mould cavity, location of the injection gates and the vents.

Several studies have been carried out, moving forward, on the manufacturing of natural fibre-reinforced with renewable polymeric biocomposites using injection moulding technique (Huda et al., 2005a,b, 2006a,b; Serizawa et al., 2006; Nystr€om, 2007). Studies have further shown the suitability of injection moulding method to fabricate natural fibre-reinforced thermoplastic biocomposites (Arbelaiz et al., 2005a; Khan et al., 2009; AlMaadeed et al., 2012; Kumar et al., 2013; Asaithambi et al., 2014; PérezFonseca et al., 2014; Bledzki et al., 2015; Gupta and Srivastava, 2016).

4.3.4 Compression moulding As the name implies, the moulding of the biocomposites is carried out under the action of pre-determined heat, high compression force or pressure. The compression moulding process combines a hot-press method and an autoclave process. A hotpress method is performed with or without a close mould, but an autoclave process involves a closed operation, whereby thermoplastic prepregs are sequentially laid up on a mould, followed by the vacuum bagging of the whole laminate, heating process at a pre-determined heat-pressure operational cycle and curing, before the final biocomposite is manufactured (Mallick, 2008; Hu and Lim, 2007). Furthermore, it is the widest choice for high volume biocomposite components, used with thermoset and thermoplastic matrices and short or long fibres. For instance, close to 70 wt% fibre can be used, and between 1 and 10 mm thickness can be manufactured. Compression is very similar to the hand lay-up process. The main difference between both processes is the presence of a closed set of matching dies during curing

Design, manufacturing processes and their effects on bio-composite properties

139

after pressure and heat are applied during compression moulding. Compression and hand lay-up mouldings are suitable for the manufacturing of small and large composites parts (Lotfi et al., 2019). More also, this manufacturing method for biocomposites involves two halves of a mould (female and male), and it utilises pressure and heat to produce parts. The pre-cut and measured quantity of mat, chopped or switched fibres are stacked together. The pre-heated closed mould cavity accommodates the fibres, then pressurised before applying temperature in order to melt the compounds and the molten compounds conform to the shape of the mould cavity (Fig. 4.13). Next, the component is ejected after the mould is opened. For reducing the damage of the fibre, the fibre is gently placed inside the mould with no vigorous motion and shear stress. Noticeably, this manufacturing process of biocomposite with long fibres produces a higher fibre volume fracture. Short and long fibres, as well as fibre mats can be used, when they are pre-mixed with the compounds. They reinforce the biocomposites and significantly decrease shrinkage of the final biocomposite part. Also, an increased filler content acts as a heat sink within the material and reduces the total quantity of heat released (Mallick, 2008), and improves the anisotropy of the final compression moulded products (Dumont et al., 2003). The compression moulding process generally involves two traditional initial charges: bulk moulding compound (BMC) and sheet moulding compound (SMC), according to Ho et al. (2012). The surface of the female mould cavity is usually covered by the charges (Mallick, 2008). Plastic materials are mingled with short fillers and fillers before they are placed in the mould cavity in the case of BMC. Conversely, SMC involves cutting of long fibre sheets according to the mould cavity, before putting them in the surface of the mould. The desired thickness of the compounds is sequentially built up by adding a sheet layer upon layer with resin on the fibre sheet (van Voorn et al., 2001). The physicomechanical properties of a flax fibre-reinforced SMC product have been enhanced (van Voorn et al., 2001). Pressure

Movable mould Charge

Fixed mould Ejector pin

Moulded part

Figure 4.13 A simple illustration of a compression moulding (Rajak et al., 2019).

140

Sustainable Composites for Lightweight Applications

For example, the mechanical properties and morphology of flax fibre-reinforced melamine-formaldehyde composites manufactured by compression moulding method were studied by Hagstrand and Oksman (2001). The mechanical properties were compared with that of glass fibre. The result showed that flax fibre-reinforced biocomposites exhibited lower mechanical properties against glass fibre. However, considering the cost and the density, the properties of flax fibre biocomposites were more competitive than that of glass fibre composites. The study also highlighted that compression moulding method helped to increase fibre-matrix interfacial adhesion. Moreover, the flammability of flax fibre-reinforced polypropylene biocomposites has been fabricated by the compression moulding process and investigated (Helwig and Paukszta, 2006). The heat release rate and mass loss rate, along with the mechanisms of thermal decomposition, were discussed. The results obtained depicted that the heat release rate depended on the fibre volume in the biocomposites. The characteristic of the biocomposite resembled that of lignocellulosic materials when its fibre volume exceeded 20%. Admittedly, the benefits of using the compression moulding process include, but are not limited to, the following. ⁃ Good dimensional tolerances under applications of pressure and temperature. ⁃ Process is very repeatable, as being used in the automotive industry for the production of small to moderate-sized parts at high volumes. ⁃ Low void content is achieved due to the application of appropriate pressure. ⁃ Minimal scrap. ⁃ High surface quality and impact strength of products. ⁃ Low labour requirements. ⁃ Can be used alongside with the prepreg and preform composites. ⁃ Compression moulding process attracts few limitations, as subsequently stated. ⁃ Equipment cost is high. ⁃ Depends on the size of the component. ⁃ Inappropriate for structural components.

In addition, there are many reported studies on the manufacturing of natural fibrereinforced renewable polymeric biocomposites using a compression moulding technique (Huda et al., 2005a,b, 2006b; Serizawa et al., 2006). Studies have further shown the suitability of compression moulding technique to fabricate both fibre-reinforced thermoset and thermoplastic biocomposites (Xian-bao et al., 2006; Athijayamani et al., 2009; Graupner et al., 2009; J unior et al., 2013; Dhakal et al., 2013; Shanmugam and Thiruchitrambalam, 2013; Zhang et al., 2013; Arthanarieswaran et al., 2014; Gupta and Srivastava, 2016). Moreover, it has been reported that the tensile strength of fibres reduced by 10% in only 10 min between temperatures of 150 and 200
C (Herrmann et al., 1998). An optimum compression temperature of nearly 80
C has been used to manufacture jute yarn-reinforced bacterial co-polyester biopol biocomposites with an improved range of mechanical properties (Mohanty et al., 2000). Van de Velde and Kiekens (2003) recorded the highest strength for a unidirectional and multidirectional non-woven flax-reinforced polypropylene biocomposite at 200
C. Therefore, processing parameters such as curing temperature, pressure and cycle time must be correctly selected and controlled. They vary with various materials

Design, manufacturing processes and their effects on bio-composite properties

141

and thicknesses of the plies or sheets used. For instance, high compaction pressure in the compression moulding process can elongate the flax or hemp fibre used. Thus, the fibre can be crushed and eventually lower the overall mechanical properties of the FRP biocomposites.

4.3.5 Vacuum resin infusion The vacuum can also be applied to the mould cavity to assist resin in-flow, being drawn into the fabrics (Fig. 4.14). This is known as vacuum-assisted resin injection (VARI). Once all the fabric is wet out, the resin inlets are closed, and the laminate is allowed to cure. Both injection and curing processes can take place at either ambient or elevated temperature. Generally, curing time depends on the type of polymer used for the biocomposite manufacturing process. Benefits of resin infusion include, but are not limited to: ⁃ Good tolerance control, as tooling controls and dimensions, give high repeatability. ⁃ Prototype tooling costs are low. ⁃ Volatile emissions, such as styrene from unsaturated polyester, are controlled as a close mould tooling process is used. ⁃ Very little waste is produced. ⁃ Process can be automated, and hence, productivity and efficiency are both enhanced. ⁃ Integration of heating up of the mould for the matrix.

4.3.6 Pre-impregnated resin The pre-impregnated resin can be used. This technique may involve the use of prepreg or simply be a manual lay-up method. The location of where the prepreg layout is carried out should be an environmentally controlled room. This is necessary to control the temperature and minimise dust particles. Workers are often required to wear protective clothing in order to further reduce contamination. Problems may occur with prepreg lay-up if silicone comes in contact with the prepreg. An occurrence of this compromises Resin channel

Bag sealing tape

Bag Vacuum channel

Resin valve Fibreglass and reinforcements

Mixed resin

Figure 4.14 Schema of vacuum resin infusion process (Cucinotta et al., 2017).

142

Sustainable Composites for Lightweight Applications

the surface of the prepreg and its ability to bond with another material. Prepreg materials also need to be kept under refrigeration in order to preserve their shelf life, as curing can occur when the material is exposed to ambient temperature. Prepreg can be applied by manual hand lay-up or automation process for mass production. Similarly, the advantages of using prepreg material include the following: ⁃ As the resin is already impregnated, there is no need to add liquid resins, which can often be messy in the hand lay-up method. ⁃ Reduction in the chances of getting air bubbles within lay-up and the possible presence of voids is also reduced. ⁃ Shelf life of prepreg is comparatively high if kept under refrigeration and free from contaminants.

4.3.7 Extrusion Manufacturing of a fibre-reinforced biocomposite materials can be done through an extrusion process, by steadily forcing materials through a pre-designed and uniform cross-sectioned die, as depicted in Fig. 4.15. This is possible by mixing softened bead-like pellets or crude polymer (matrix material) with pre-treated or crude lignocellulosic fibres (fibre bundles/reinforcements) before they are continually fed into a single or twin-screw extruder. Fig. 4.15 shows an extrusion compounding process of spent coffee grounds-reinforced polypropylene (SCG/PP) composites (Sohn et al., 2019). Both reinforcement and polymer matrix can be concomitantly fed into the screw extruder (Gallos et al., 2017). Malkapuram et al. (2009) stated that twinscrew extruders exhibited better fibre dispersion, and hence, mechanical performance than the single screw system. The extrusion process is suitable for both natural fibres, renewable and cellulosebased polymers. Uniform fibre dispersion is one of the important factors to consider during this process because it determines the properties of the biocomposite. Hence, a twin-screw extruder produces a high shear process towards good fibre dispersion (Oksman et al., 2009; Arrakhiz et al., 2012).

+ SCG/PP composites Spent coffee grounds Air-cooled fans

Nozzle

Polypropylene

Hopper

Heater 1 Heater 2 Heater 3 Heater 4

Screw

Cutting machine

Conveyor belt

Extrusion machine

Figure 4.15 An extrusion process of biocomposite material (Sohn et al., 2019).

Design, manufacturing processes and their effects on bio-composite properties

143

More also, the extrusion process is a fast, dimensional repeatable and economic process. It produces high fibre volume fraction and high structural properties of biocomposite products. However, the extrusion process of biocomposite parts is limited to component size because of the quantity of fibre and the extruding force required. It is also limited to constant or near-constant cross-section components (Lotfi et al., 2019). Tungjitpornkull and Sombatsompop (2009) reported a better tensile modulus of compression moulded glass fibre-reinforced wood/polyvinyl chloride (WPVC) composites than that of twin-screw extrusion counterparts, as shown in Fig. 4.16. This improved property was attributed to the less fibre breakage and thermal degradation of the PVC molecules. Consequently, a longer fibre length, higher specific density due to lesser void contents (manufacturing defect), and stronger composite was produced with compression moulding in comparison with extrusion counterpart.

4.3.8 Resin transfer moulding Resin transfer moulding (RTM) has become very popular to fabricate fibre-reinforced polymer biocomposites, due to its capability to produce high volume laminates with cost-effectiveness and low void content. It bridges the gap between capital intensive compression moulding and labour intensive hand lay-up method (Sreekumar et al., 2007). This process is a closed moulding process, whereby resin is transferred (injected) under pressure over the already placed fibre preform (woven mat, fabrics or chopped strand mat), as schematically shown in Fig. 4.17. Generally, these mats or fabrics are made from natural plant fibres, such as banana, hemp, sisal, flax, nettle fibres to fabricated biocomposite laminates. Epoxy, polyester, vinyl ester, methyl methacrylate and phenolic resins are commonly used in RTM. The fabrics used are sometimes pre-pressed to the mould shape and held together by a binder. These ‘preforms’ are then more easily laid into the mould tool. A second mould tool is then clamped over the first, and resin is injected into the mould cavity until it is filled up, 16

Tensile modulus (GPa)

14 12 10 8 6 4 2 Compression moulding

Twin-screw extrusion

0 0

2

4

6

8

10

12

14

16

18

Fibre content (% wt)

Figure 4.16 Effect of compression moulding and twin-screw extrusion techniques on tensile modulus of composites (Ku et al., 2011).

144

Sustainable Composites for Lightweight Applications

Resin

Vent

Mould

Figure 4.17 Illustration of a resin transfer moulding process (Rajak, 2019).

using single or multiple inlet ports (multiple injection gates). The component is then removed from the mould after cooling (Sreekumar et al., 2007). A full chemical reaction between the resin and its catalyst (curing) is required through the post-curing process. Both fire retardancy and surface finish of the biocomposite can be enhanced using appropriate mineral fillers. The mould temperature, mould configuration, resin viscosity, vent control, resin injection pressure, fibre mat permeability, gate location and configuration, preform placement methods, preform permeability and architecture, are important process parameters in RTM technique (Ho et al., 2012). They require careful selection, monitoring or control and optimisation to avoid manufacturing defects. For instance, mould deformation and fibre preform wash-out defects can occur by excessive injection pressure. Also, pre-mature resin gelation and its resultant short shot flaws are induced by an excessive high mould temperature (Ho et al., 2012). RTM is a very useful technique to produce high volume and low-cost biocomposite parts. Therefore, RTM is characterised with the following benefits: ⁃ Increased productivity through automation, as well as good temperature and pressure control. ⁃ Improved quality, as a result of the process being consistent. ⁃ Good dimensional tolerances and surface finish due to the use of suitable pressure in the process. ⁃ Required lower temperature, and therefore, thermomechanical degradation is avoided. ⁃ Manual, semi-automated and highly automated processes are possible. ⁃ Possibility of a combination of a wide range of reinforcing materials (fibres and fillers) to achieve numerous desired orientations. ⁃ Production of near net shape components is possible; hence it decreases the material waste. ⁃ Nearly zero air enclosure or entrapment (voids). ⁃ Production of uniform component thickness, large and complex components with high strength to weight ratio. ⁃ Due to pressure being utilised in this process, higher fibre volumes can be obtained, but high injection pressure is not required.

Design, manufacturing processes and their effects on bio-composite properties

145

The RTM is widely used in the automobile industry to produce car body components, in addition to other products, such as bathtubs and big containers. However, the following setbacks characterised the RTM process: ⁃ ⁃ ⁃ ⁃

The size of the biocomposite laminate is limited by the mould cavity. It requires complex mould design. It does not accommodate all reinforcing materials content. Tooling cost is relatively high.

In addition, there are many studies on RTM of sustainable natural fibre-reinforced polymeric biocomposites for lightweight applications in various engineering sectors, especially in automotive and aerospace (Ferland et al., 1996; Ikegawa et al., 1996; Willians and Wool, 2000; Kim and Daniel, 2003; Warrior et al., 2003; Goutianos et al., 2006; Sreekumar et al., 2007; Rassmann et al., 2010; Salim et al., 2011; Francucci et al., 2012). Importantly, Sreekumar et al. (2007) reported that the mechanical (tensile and flexural strengths, Young’s and flexural moduli) properties of sisal fibrereinforced biocomposites manufactured by RTM were greater than that of compression moulded counterparts. These improved mechanical behaviours of RTM biocomposite samples were attributed to their lower water absorption and void content, as a resultant advantage of better fibre-matrix interfacial adhesion. Also, Sebe et al. (2000) used the RTM technique to manufacture a series of hemp fibre-reinforced polyester biocomposites. The results obtained depicted a proportional increase in mechanical (impact and flexural properties) with an increased quantity of fibres during formulation. Sreekumar et al. (2007) reported a maximum water absorption and void content with compression moulded sisal-leaf fibre-reinforced polyester biocomposites when compared with RTM counterparts.

4.3.9 Automated fibre placement Automated fibre placement, as depicted in Fig. 4.18, is an innovative technique of manufacturing large and complex composite structures. Some of the setbacks of the hand lay-up method, especially the low productivity, can be improved with the automation of a programmed robotic system, as described in Fig. 4.19. An uninterrupted layering and building of biocomposite structure is performed as a robot places the Incoming tape

Tape feed

Tape cut Consolidation force

HGT

Process heat

Direction of travel

Previous ply Tool

Figure 4.18 The Automated fibre placement process (Lotfi et al., 2019).

146

Sustainable Composites for Lightweight Applications

Fibre feeding system

Robotic arm Spool with fibre tape

Head

Figure 4.19 Sub-components of an automatic fibre placement system (Kozaczuk, 2016).

continuous fibre-reinforced composite tape. The incoming tape is heated either by laser or hot nitrogen, before pressing to the mould for good compaction. Each ply can be laid using diverse orientations and angles (Lotfi et al., 2019). The main benefits of using automated fibre placement process when compared with other manufacturing techniques include, but are not limited to, reduced costs of labour, manufacturing time and material scraps or wastes, it supports producibility and repeatability. Some large unique structures can be produced by this process. Consequently, this process is now popular and affordable for several aviation industries today, such as Spirit, Boeing and Airbus. Nevertheless, the limitations from mould shape, head geometry and roller diameter, as well as ply edges caused by cut tapes, have reduced the application of this advanced method (Kozaczuk, 2016; Lotfi et al., 2019). Also, it is relatively expensive. Sometimes, it results in a distortion in the process of thermoplastic biocomposites.

4.3.10 Filament winding There are other processing methods widely used in the fabrication of natural fibrereinforced polymer composite materials. One of them is filament winding. This process involves the continuous movement of unwinding fibre strands through a resin container, where complete impregnation occurs before the resin-impregnated strands are passed to a rotating mandrel in a controlled pattern to give a specific fibre orientation (NPTEL, 2019b). The simple schematic illustration of the process is depicted in Fig. 4.20. From Fig. 4.20, it is evident that fibre creel, resin-impregnated system, carriage and rotating mandrel are the main components of this process. Filament winding is primarily suitable for hollow, circular or oval sectioned parts, such as pipes and tanks, to mention but a few. Filament winding has been used by Li (2015) to manufacture fibre-reinforced polymer, as shown in Fig. 4.21. Similarly, the fabrication of natural fibre-reinforced polymer biocomposites has been done with this process. For instance, filament winding was used successfully to fabricate hemp-reinforced thermoplastic polypropylene

Design, manufacturing processes and their effects on bio-composite properties

147

Fiber spools

Resin bath Nip rollers

Rotating mandrel

Figure 4.20 The schematic illustration of the filament winding process (Wang et al., 2020).

composites (Madsen, 2004). In this experimental study, the hemp fibre yarn was aligned by filament winding with a custom-built winding machine. The film stacking method for fibre/matrix mixing in a compression moulding was also applied in this work. The study had concluded that the commingled filament-winding method was found to be much more optimal approach for fibre/matrix mixing, when compared with the film-stacking method with a combination of compression moulding. More also, filament winding was recently used to manufacture a basalt fibre-reinforced epoxy composite pipes of diameter, thickness and filament winding angle of 100 mm, 6 mm and 55 degrees, respectively (Prabhakar et al., 2019). The results obtained from the drop weight low-impact tests conducted showed a higher breaking strength at a maximum impact energy of 60.42 J at an impactor’s height of 1.25 m, when compared with other conventional material pipes.

Figure 4.21 Manufacturing of fibre-reinforced polymer with filament winding (Li, 2015).

148

Sustainable Composites for Lightweight Applications

Furthermore, the fibre tension produces the required compaction. This is the reason why fibre tension is a vital factor to be carefully considered in the filament winding process. The fibre tension depends on fibre type, geometry and winding pattern produced by the rotating mandrel. An optimum fibre tension must be determined and used to avoid fibre surface fracture, which eventually leads to final breakage. Heat, commonly from an oven, is used to cure the biocomposite before the biocomposite part is finally removed from the metallic, collapsible rubber or soluble plaster mandrel using appropriate technique. Benefits of filament winding include: ⁃ Possibility of achieving a high strength to weight ratio and high fibre volume ratio. ⁃ An automated process, therefore it involves minimal labour, high production volume (very fast process), efficiency and cost saving, because fibres are not necessarily converted to fabric prior to use. ⁃ Control or metering of the resin is possible. ⁃ A specific direction of fibre orientation is easily achieved, with good structural properties of the laminates. ⁃ Possibility of design flexibility in biocomposite parts due to change in materials, winding patterns and curing option. ⁃ Excellent fibre distribution, placement and orientation are obtainable with a high measure of uniformity or consistency. ⁃ Use to manufacture biocomposite parts that require accurate tolerances ⁃ Requires a less and low cost materials to fabricate high strength biocomposite part, when compared with other biocomposite manufacturing processes. ⁃ Supports the production of variable sizes of components (NPTEL, 2019b). ⁃ Suitable for commercial production because of its low cost, high flexibility and repeatability (Sofi et al., 2018).

Nevertheless, there are few shortcomings that are associated with the application of the filament winding process, as subsequently highlighted. ⁃ ⁃ ⁃ ⁃ ⁃ ⁃

Expensive mandrel for some applications and large components. Surfaces of some biocomposite parts may be occasionally unacceptable. Relatively high capital investment. Within a single layer winding, it is impossible to change the fibre direction. Production of a reverse curvature (female feature) is impossible. The mechanism requires accurate and skilful control to achieve uniform fibre distribution and orientation (NPTEL, 2019b).

Filament winding process can be used to manufacture numerous composite products for military and defence sector (missile and rocket motor cases), aerospace industry (aircraft fuselages), sports/game (golf shafts), oil and gas (storage tank, pipelines and gas cylinders), marine/naval (vessels and sail boat mast), building/construction (ducting and cement mixture), among others (fishing rods). As technology advances with the advent and application of sophisticated machine centre and engineering software, filament winding can be used to manufacture complex engineered non-spherical and noncylindrical composite products. Also, a conventional 2-axis lathe type filament winding machine has been improved to higher degrees of freedom (Minsch et al., 2017). Independent monitoring of all movements of the entire process is possible today, using a computer controlled machine. Also, a robotic filament winding technique has been

Design, manufacturing processes and their effects on bio-composite properties

149

adopted for industrial application, as recently reviewed by Quanjin et al. (2018a). Further, recent studies on this process have been reported in an attempt to advance the innovative equipment, re-design and manufacture different optimised axial filament winding machines and processes (Mateen et al., 2018; Quanjin et al., 2018b, 2019).

4.3.11 Autoclave moulding The autoclave moulding method is very similar to the vacuum bagging, with few changes. Heat and pressure that are required by the biocomposites during the curing stage are supplied by the autoclave machine (Fig. 4.22(a)). This process involves firmly stacking of prepregs in a mould following a specific sequence. A release gel is applied to the surface of the mould to avoid the sticking between the polymer and the mould surface. In addition, this process allows the use of cores and inserts. Then, vacuum bagging follows to remove all possible entrapped air between the layers, as earlier explained (Fig. 4.22(b)). Afterward, the whole assembly is moved to the

(a)

(b) Bleeder pack Membrane

Prepreg pack Cork dam Seal

Fan Mould

Heaters Pressure

Vacuum

Figure 4.22 (a) An autoclave machine, (b) its internal components and process schematic (Halley, 2012; Dixit et al., 2016).

150

Sustainable Composites for Lightweight Applications

autoclave machine, where both heat and pressure are applied to aid uniform and effective distribution of the matrix, as well as good fibre-matrix interfacial adhesion or bonding for a definite time interval. This stage is called curing. Later, the composite component is removed from the mould after cooling of the assembly and removal of the vacuum bag, sequentially. Autoclave moulding process embraces the following advantages: ⁃ ⁃ ⁃ ⁃ ⁃ ⁃ ⁃

Applied to both fibre-reinforced thermosetting and thermoplastic polymer composites. Better inter-layer adhesion. Good control of both fibre and resin. Proper and sufficient fibre wetting. Degree of uniformity in component solidification is high. Supports high fibre volume fraction in the composite component. Absence of void content in the final component due to the benefit of the vacuum bagging mechanism. ⁃ As a part of the benefits of the vacuum bagging process, a better interfacial bond with inserts and cores is often achieved. ⁃ Used to manufacture high strength to weight ratio parts.

For profiting from the aforementioned advantages of the autoclave moulding process of manufacturing fibre-reinforced polymer composites, this process has been widely used to manufacture numerous engineering parts mainly by aerospace, marine and military companies. These products include, but are not limited to, aircraft components, military, marine and space crafts, as well as missiles. Despite the aforementioned benefits of the autoclave moulding process, it has the following few drawbacks: ⁃ Low production rate. ⁃ Restriction on composite component size, which depends on the size of the autoclave machine. ⁃ Involvement of skilled labour. ⁃ Expensive technique for processing composite.

4.3.12 Out-of-autoclave moulding In a bid to reduce rigid manufacturing environment and significant acquisition, tooling and operation costs that are associated with traditional autoclave moulding techniques, an out-of-autoclave (OoA) moulding was invented to produce autoclave-quality components, using vacuum bag-only (VBO). These costs are much with large components. In addition, the OoA process has gained attention today, especially due to its lower cure pressure through VBO, which consequently eliminates autoclave-induced manufacturing defects. However, a lower cure pressure must be used with care; it may result in a relatively low mechanical performance (mostly toughness), out-time of only a week and high porosity or uneven resin bleed, especially within high fibre volume fraction (Centea et al., 2015). OoA composite manufacturing technique is performed in a closed mould, where vacuum, pressure and heat are applied by means other than an autoclave. Examples

Design, manufacturing processes and their effects on bio-composite properties

151

of methods under this technique include, but are not limited to, resin transfer moulding and vacuum-assisted resin transfer moulding. Comparatively, the energy consumption in the autoclave method is relatively higher than that of the OoA method, due to higher cure pressure and temperature required. These consequently increase the operational costs and reduce the sustainability of the conventional autoclave technology. Also, the size of the autoclaved composite material or part is limited by the capacity of the autoclave machine. Hence, it hardly supports flexibility in sizes, from small to big composite parts.

4.3.12.1 Autoclave and out-of-autoclave curing processes Along with different manufacturing processes employed in the composite fabrication process, the curing methods and parameters equally play an important role in the final properties of the composites. Because the curing process determines resin-rich areas, porosity formation, adhesion between reinforcement and the matrices, the two key curing processes employed in composite fabrication are subsequently elucidated. ⁃ Autoclave curing: This is the most commonly used curing process used in the fabrication of primary and secondary composite structures. It uses combined pressure and vacuum, which helps to achieve composites with a very small amount of void contents and a highly reliable performance. However, this process uses a significant amount of energy and requires high operating costs, as previously discussed. ⁃ Oven or out-of-autoclave curing: In this process, high-quality composites are obtained. However, the OoA process is not much used to fabricate primary structural components, unlike the autoclave process. In recent years, due to its low operating costs and environmental benefits, the OoA process had been employed in composite manufacturing.

4.3.13 Additive manufacturing Advancement in technology has produced additive manufacturing (AM) technique, commonly referred to as 3D printing. AM or 3D printing is a process whereby desired components are built from a three-dimensional (3D) computer-aided models in a successive layer-by-layer pattern, through the ejection of already prepared materials via a nozzle (Berman, 2012; Parandoush and Lin, 2017). Manufacturing of parts began many decades ago with traditional subtractive manufacturing (SM) technology. SM technology involves the removal of unwanted parts from a whole material, mainly using machining processes. Due to the drawbacks of SM and several benefits of AM technology, attention and/or attraction of many manufacturing industries to the use of AM technology have been widely increasing. Disadvantages of SM technology include, but are not limited to, longer production time, lesser accuracy, higher cost of manufacturing, a greater possibility of failure and accident, higher energy consumption, higher scrap/wastes and inability to produce complex geometry or product with many intricacies. With the effective application of AM technology, the listed challenges can be easily eliminated.

152

Sustainable Composites for Lightweight Applications

The introduction of AM processes to the manufacturing of FRP composite manufacturing can be traced back to a few years ago. Today, AM technology has been developed in many sectors, such as building and construction, digital art, tissue/biomedical engineering, architectural design and importantly, composite manufacturing, to mention but a few. Recently, the use of robots in manufacturing is promoting various innovative FRP composite manufacturing techniques. Advances in AM or 3D printing techniques of FRP composites include fused deposition modelling (FDM), extrusion, selective laser sintering (SLS), stereolithography (SLT), laminated object manufacturing (LOM) and recently, the four-dimensional (4D) printing of active FRP composites (Parandoush and Lin, 2017), as simply and briefly explained in Fig. 4.23(aef), respectively. The 4D printing is a process that accommodates the use of smart or active materials, which can respond to external stimuli (such as heat, chemical, cold, among others) and change to pre-programmed shapes or self-transformed structures. The fourth dimension denotes time (Mitchell et al., 2018). The choice of process to use depends on the nature of reinforcement and matrix of the anticipated composite product, cost, quality, properties, quantity, volume/size, time, among other factors. AM of FRP composites has improved the efficiency of AM with the capability to fabricate highly customised components with better mechanical properties when compared with unreinforced polymeric products. Nevertheless, with the increasing benefits of AM technology in the field of FRP composite manufacturing, there are still some concerns that require attention to further optimise it. They include fibre orientation, blockage of nozzle and printer heads due to filler entanglement, the formation of a void, poor fibre-matrix interfacial adhesion, increased curing time, agglomerate and non-homogeneous FRP composite formation, as well as predictive modelling and simulation. Although, a few studies have been carried out on 4D printing of active FRP composites (Ge et al., 2013; Li et al., 2017; Miao et al., 2017; Rajkumar and Shanmugam, 2018; Zhang et al., 2019; Piedade, 2019; Ahmed et al., 2020), finite element methods (Weinan et al., 2007; Pineda et al., 2013; Zhang and Xu, 2013), analytical and modelling techniques (Li et al., 2002; Modniks and Andersons, 2013; Melenka et al., 2015) of additively manufactured FRP composite materials, but there are still opportunities for cutting edge research in a bid to continuously improving the AM or 3D and 4D printing of FRP composite materials/products.

4.3.14 Brief comparison among manufacturing processes The choice of use of the above-mentioned manufacturing process depends on cost contribution (from materials, labour and tooling) and required energy intensity. Therefore, the attention of many biocomposite manufacturers has been shifted from the extrusion process to either resin transfer moulding or compression moulding technique because of the high energy intensity required for the extrusion process. This is the reason why the foremost manufacturing techniques for natural fibre-reinforced polymer biocomposites are compression, resin transfer and injection moulding. For instance, Lotfi et al., (2019) conducted a comparative study on energy consumptions of extrusion, hand lay-up, resin transfer moulding, compression moulding and

Design, manufacturing processes and their effects on bio-composite properties

153

(a) Liquifier head Extrusion nozzles Part Support Print bed Support material spool (if necessary)

Build platform

Build material spool

(b) Fused filament fabrication

Other approaches Prepreg composite filament

In–situ coating with liquid resin To extruder

Impregnated fibre

Dry fibre

In–situ fusion with molten thermoplastic Dry fibre

Molten thermoplastics

Liquid deposition modelling Randomly oreinted short fibres in paste or liquid

Cross-section view Fibre

Matrix

Aligned fibres

Inter–layer boundaries

Voids

Figure 4.23 Additive manufacturing technologies, showing (a) fused deposition modelling, (b) extrusion, (c) selective laser sintering, (d) stereolithography, (e) laminated object manufacturing and (f) 4D printing of active FRP composites (Ning et al., 2015; Goh et al., 2019; Shahzad et al., 2014; Pan and Patil, 2017; Ahn et al., 2012; Miao et al., 2017), respectively.

154

Sustainable Composites for Lightweight Applications

Laser Mirror scanner

(c)

XY deflection Roller / scraper

f-Tlens Protective atmosphere

Y

X Z

Feed container Build cylinder Overflow container

(d)

DMD based projection unit

Z-stage Electrostatic deposition unit Platform Particle collection plate

Z

Built physical resin model

X Particles dropped on the resin surface

Liquid resin

Mirror

(e)

Laser beam X-Y moving optic head

Laser Heated roller

Current laser Part layer contour Previous layer

Material sheet Material supply roll

Figure 4.23 cont’d.

Fabricated part and support material Platform

Waste take-up roll

Design, manufacturing processes and their effects on bio-composite properties

155

(f) Top down Heat

Cool

Side view

Swelling

t=0

t = 5 min

Swelling

t=0

t = 25 min

Figure 4.23 cont'd.

automated fibre placement and recorded approximately 19, 15, 13, 12 and 3 MJ/kg, respectively. It was observed that the automated fibre placement exhibited minimum energy intensity among the manufacturing methods that were considered. Also, the gross costs (including sum of tooling, materials and total labour cost/unit) of approximately $953, 619, 534, 463 and 446 were spent on resin transfer moulding, hand layup, automated fibre placement, compression moulding and extrusion, respectively. This implies that the extrusion method, with the lowest cost of $446, was the most economical process, among others (Fig. 4.24).

156

Sustainable Composites for Lightweight Applications

(a) 1000 Tooling cost/unit

900

Material cost/unit

800

Total labor cost/unit

700 Cost($)

600

685 177

500

213

400

146

300 200

326 64

185 296

100

204

HL

102

136

73 64

AFP

CM

EX

CM

EX

0 RTM

261

83

Manufacturing process

(b)

25

Energy (MJ/kg)

20

15

10

5

0 RTM

HL

AFP Manufacturing process

RTM: resin transfer moulding

HL: hand lay-up

AFP: automated fibre placement

CM: compression moulding

EX: extrussion

Figure 4.24 Comparison of (a) cost per unit and (b) energy intensity of selected composite manufacturing methods (Lotfi et al., 2019).

Design, manufacturing processes and their effects on bio-composite properties

157

Similarly, the ultimate tensile strengths (UTS) and elastic moduli (EM) of hand layup, vacuum infusion and vacuum bagging manufactured composite samples have been comparatively studied (Abdurohman et al., 2018). The results obtained show that samples fabricated through the vacuum infusion method recorded the highest UTS and EM, as presented in Fig. 4.25. In moving forward, both quality and mechanical behaviour performance of OoA prepreg and autoclaved composite materials have been studied and compared (Sutter et al., 2019). From the result obtained, it was evident

(a) 400 346.15 350 300

271.298 260.982

250 200 150 100 50 0 Hand lay up

Vacuum infussion

Vacuum bagging

(b) 12000 10000

10673.4 9221.9 8660.52

8000 6000 4000 2000 0 Hand lay up

Vacuum infussion

Vacuum bagging

Figure 4.25 Comparison among (a) ultimate tensile strengths and (b) elastic moduli (MPa) of differently manufactured FRP composite laminates (Abdurohman et al., 2018).

158

Sustainable Composites for Lightweight Applications

from the early results that OoA manufactured composite material recorded a decrease in both quality and mechanical property performance after a prolonged out-time. Notwithstanding, it exhibited a similar performance to the autoclaved materials when subjected to an ambient shop floor handling, within a few days of processing duration. As technology advances to improve the manufacturing process of sustainable and lightweight biocomposite materials and parts, a combination of different processes has been designed and developed. These include, but are not limited to, liquid compound moulding processes (such as RTM light and compression RTM), extrusion-injection and extrusion-compression moulding with different screw, mould and die designs (Faruk et al., 2012), in addition to the application of automated systems (robots) and efficient software packages.

4.4 Key drivers for cleaner production or green manufacturing The concept of cleaner production includes the development and application of new, reliable, efficient, simple and new technologies, as well as intense, innovative activities that are better than the existing conventional or traditional ones in terms of environmental protection and waste minimisation. Biocomposites are environmentally friendly products, unlike metallic products and their alloys. They possess the following properties to support cleaner production: biodegradability, renewability, sustainability and recyclability, especially when both reinforcements and matrices are bio-products (green composites). Both the life cycle of biocomposite materials and their manufacturing processes support the goals of cleaner production. Cleaner production involves the use of cleaner and renewable sources of energy, such as sun, wind and water, and the possibility of building the recycling process into the manufacturing process of biocomposite materials, which makes the entire process cleaner. Biocomposite waste can be recycled. Therefore, our environments (air, water and land) are protected from harmful substances, called pollutants. Also, there is an assurance of satisfactory end-of product disposal. In addition, composite materials are recently used to produce filters for removing particles from gas streams. Hence, air pollution is reduced. For instance, powerful and efficient filters are manufactured by combining woven and non-woven fabrics. These filters are used to clean the air of commercial plants from several particles. Also, filtration fabrics are produced for paper industries, whereby a high volume of water is removed for each tonne of paper produced. This is achieved through the following strategies: 1. Decrease the water pollution by chemicals, using a higher density of yarns with greater retention capability in newly developed woven fabrics. 2. Reduce the energy consumption per production unit by decreasing the water content of paper with newly developed ultra-thin fabrics. 3. Increase or promote the use of recycled fibres by developing new yarns and fabrics that possess both sticky and contaminant resistance properties.

Design, manufacturing processes and their effects on bio-composite properties

159

Figure 4.26 A CAD illustration of three dimensional woven based composites. (From Gereke, T., Cherif, C., 2019. A review of numerical models for 3D woven composite reinforcements. Composite Structures 209, 60e66).

Also, the application of three-dimensional woven-based composites (Fig. 4.26) in aviation industry helps to reduce the quantity of fuel consumption and resultant carbon emissions, noise patters, as a considerable weight of the aircraft and noise level are reduced. For example, there is a 20% decrease in fuel consumption of a Boeing 787 aeroplane compared with other similar aircraft. There is a decrease in the energy required to manufacture the aircraft, quantity of scrapped materials and use of spare parts due to better composite properties that yield a greater wear resistance and durability or life span (Mourad, 2012).

4.5 Manufacturing defects Biocomposite materials have attracted wide applications due to their better inherent properties: corrosion and wear resistance, high strength and stiffness, low cost and density, sustainability and renewability, as well as recyclability and biodegradability, when compared with some metals and their alloys. Nevertheless, these outstanding properties can be compromised and destroyed due to the wrong choice of manufacturing processes and poor management or control of process parameters, which consequently cause various types of manufacturing defects. These defects could be caused during materials processing by environmental, mechanical and/or human factors. Fig. 4.27 shows various manufacturing defect formations and their associated or dependent relative process design parameters. Defects alter the materials, as well as mechanical, thermal, optical, acoustic and electrical properties of the manufactured biocomposites. The functionality of both fibres and matrices, which are the main constituents of a reinforced biocomposite, as well as coupling agents and fillers, can be adversely affected with an incorrect selection of both processes. It has been reported that nearly 44% of failures in fibre-reinforced polymer composites is caused by the manufacturing process (Knoeller, 2018). The possible common defects associated with the manufacturing of biocomposites and their causes include microcracks and cracks, temperature effects, moisture absorption, inclusions or contamination, porosity (void or pores), among others, as simply depicted in Fig. 4.28. They are subsequently and extensively discussed.

160

Sustainable Composites for Lightweight Applications

Winding/draping/forming design parameters: • • • • •

Pre-load tension Holding force profile Pre-shear Direction Starting point

Curing design parameters: • Pulling speed • Temperature profile • Tool geometry

Misalignment Shear deformation

Curing stage

Wrinkles Thermal properties variation

Winding/draping/forming stage

Residual stresses Incomplete cure

Fibre volume fraction/thickness variation Permeability variation

Distortion Cure induced voids and cracking

Resin rich area Voids

Filling/consolidation stage

Dry spots

Filling design parameters: • • • • •

Injection rate/pressure Mould filling temperature profile Gates/vents locations Gates/vents numbers Pressure profile

Figure 4.27 Various manufacturing defect formations and their dependent process design parameters (Struzziero et al., 2019). Delamination

Broken fibre

Debonding

Wrinkle

Porosity

Void

Resin rich

Matrix crack

Foreign object

Blister

Figure 4.28 Common manufacturing-induced defects of fibre-reinforced composites (Bowkett and Thanapalan, 2017).

4.5.1 Microcracks and cracks Both matrix and reinforcements expand and contract at different temperatures because of their dissimilar coefficients of thermal expansion (CTE). A volumetric contraction of the matrix occurs, the CTE of matrix is often higher than that of reinforcements (Wisnom et al., 2006). This phenomenon occurs after thermoplastic polymeric

Design, manufacturing processes and their effects on bio-composite properties

161

1  in 100

Matrix micro-cracks within plies

Delamination (separation) at ply boundary

Figure 4.29 Formation of matrix microcracks within plies, showing their consequent delamination defect.

biocomposites are cooled to their service temperature subsequent to moulding process. Thermal stresses are developed in the fibres and around matrix during cooling and heating operations, due to the different contraction rates of fibres and matrix (Parlevliet et al., 2006). An interface de-bonding probably occurs when the magnitude of residual stresses is more than the yield strength of the biocomposite, later results in a transverse or microcracks (Timmerman et al., 2003; Parlevliet et al., 2006; Parlevliet et al., 2007). In addition, microcracks can occur from the stresses formed within thermoset polymer biocomposites during curing. These stresses increase whenever the temperature increases. This causes a rise in the size or shape and density of the cracks and possibly a delamination defect (Wisnom et al., 2006). Also, the stress concentration increases at ply interfaces. Therefore, the formation of microcracks within these interfaces often results in delamination (Fig. 4.29). Microcracks are tolerated at low densities in materials engineering practice. However, experimental results have confirmed that delamination occurs when a critical microcrack density is reached in the presence of multiple microcracks (Parlevliet et al., 2007; Knoeller, 2018). Therefore, it may be a regrettable decision to accept microcracking within biocomposite materials during the manufacturing process, as it damages the materials by reducing their mechanical (CTE, cyclic or fatigue and longitudinal stiffness) properties (Timmerman et al., 2003). The crack occurs when there is an actually visible separation in a material. It initiates as a microcrack and propagates when the total local energy exceeds that which the material can absorb (Strong, 2008). Cracking is a processing flaw frequently associated with the use of a gel coat; a protective layer applied to the mould before reinforcement is placed, of a biocomposite material. This gel coat must be capable of withstanding large stresses during moulding and de-moulding stages. There is an additional stress added to the biocomposite parts whenever there is sticking between the mould and the material, except when an alternative improved technique is used. A well prepared and treated gel coats with release agents could either prevents or minimises the sticking effect. Crack leads to delamination that reduces both strength and stiffness of the biocomposite, and eventually, causes a catastrophic fracture.

162

Sustainable Composites for Lightweight Applications

4.5.2 Temperature effects Manufacturing of biocomposite involves thermal process, whereby temperature changes or heat transfers between the fibres and the matrix. This results in thermal stresses and strains in the biocomposites. Heat energy causes a free expansion of the laminate layers in a biocomposite laminate type. This causes internal stresses and failure if this expansion is constrained by the adjacent laminate layers of a biocomposite with different oriented reinforcements or fibres. The thermoplastic matrices may not be able to continuously transfer loads efficiently, prior to untimely failure, if the temperature higher than the glass transition temperature is allowed (Knoeller, 2018). Additionally, the use of improper time and/or temperature during the curing process is one of the principal and common sources of temperature effects.

4.5.3 Moisture absorption Biocomposites have a tendency of absorbing moisture or water through fibre, matrix, fibre-matrix interface, areas that have been already affected by porosity, microcracking, cracking and delamination. The fibre absorbs less moisture or other liquid such as water than the matrix materials through capillary action before the moisture is absorbed by the matrix. Consequently, the chemical composition of the liquid resin is altered. There is also a decrease in the glass transition temperature and mechanical properties, such as strength and elasticity (Knoeller, 2018). When moisture is restricted within reinforcements, due to the barriers caused by the reinforcements against moisture absorption, there is always a swelling effect. If this effect is prolonged and later freeze, it causes fibre-matrix de-bonding or an inter-laminar delamination. Therefore, materials that are susceptible to moisture absorption must be kept in a low humidity environment and moisture must be removed from the thermoset polymer matrix biocomposites before they are subjected to a high-temperature curing process to preclude expansion and resultant delamination (Knoeller, 2018).

4.5.4 Inclusions or contamination The physical and mechanical properties of biocomposite materials are often affected by an enclosure of foreign bodies, especially solid materials. Obviously, an encapsulated strange material possesses different properties, when compared with the main constituents of the biocomposites. Energy fields and structural stresses can be transmitted through inclusions. For instance, foreign particles such as irrelevant or unrelated fibres and pieces of a plastic release film/peel ply that are not removed from the surface of prepreg can cause contamination defects (Knoeller, 2018).

4.5.5 Porosity (void or pores) Porosity can be described as the volume fraction of small microvoids (a space that is neither occupied by fibre nor matrix) in a biocomposite material (Fernlund et al., 2016). A space occupied by entrapped volatile gases or air (non-solid foreign material)

Design, manufacturing processes and their effects on bio-composite properties

163

is known as void. Trapped gas when mixing resin or bridging ply and wrinkling, insufficient use of adhesives/matrices, inappropriate pressure used during curing cycle, as well as dissolved and absorbed gases and water can result in voids or blisters. Conversely, pores result from insufficient fibre wetting by the matrix or insufficient infiltration of the fibre tows (Matzkanin and Yolken, 2007). Both voids and pores are sometimes used interchangeably. Precisely, voids and pores are formed under (or beneath) and on the surface of fibre-reinforced biocomposites, respectively. These two defects can be detected by using ultrasound, acoustic, acousto-ultrasonics, X-ray, to mention but a few techniques. Porosity has a very significant effect on the mechanical behaviours of biocomposites, among other important properties. It is not easy to avoid in biocomposites. Therefore, much concentration has been given to minimise it in a biocomposite material. It is caused due to the inclusion of air during the manufacturing process, especially when there is poor wettability of the natural fibres, presence of lumens in most plant fibres (such as hemp and flax), low compatibility of fibres or hollow feature within natural fibres and/or fibre bundles. Although these hollow features and lumens could become closer during the manufacturing process when subjected to high pressure (Madsen et al., 2009; Pickering et al., 2016). In moving forward, there are different types of voids and porosity. Depending on the mode and location of formation, voids could be inter-laminar (between plies), fibre tow within partially impregnated fibre tows or resin (fully enclosed by resin), as depicted in Fig. 4.30. Both inter-laminar and fibre tow voids are commonly named bulk voids because of their connection to the breathing network and they can be removed by de-bulking. Similarly, types of porosity depend on their various shapes and sizes. There are spherical, cylindrical and microporosity, as shown in Fig. 4.31. Also, there is surface porosity, which occurs on the biocomposite laminate surface, as depicted Fig. 4.32. The image was captured by light reflecting off the surface. It has no significant impact on the mechanical behaviours, but it affects the aesthetic appearance and aerodynamic properties of the biocomposites.

Epoxy (matrix)

Perpendicular carbon fibre tows Fiber tow void

Parallel carbon fibre tows

Interlaminar void

Resin void

Figure 4.30 Typical types of voids and their locations of formation (Fernlund et al., 2016; Farhang, 2014).

164

Sustainable Composites for Lightweight Applications

Spherical porosity

Micro porosity

Cylindrical porosity

Resin rich region

Fibre rich region

Figure 4.31 The micrographs of different types of porosity and their features: shapes, sizes and locations (Hakim et al., 2017).

Figure 4.32 Formation of surface porosity (black) on a small and flat MTM45-1/CF2426A prepreg laminate (Wells, 2015).

Admittedly, porosity has been identified as one of the crucial manufacturing defects of biocomposite materials. It is quite uneasy to obtain 0% porosity in a biocomposite; the presence of it above a certain limit can be greatly disadvantageous to the properties, especially mechanical (moduli, delamination resistance, inter-laminar shear, flexural/ bending, compressive, static and fatigue strengths), water/moisture absorption behaviours and structural reliability of the final biocomposite products. However, it can be managed or reduced through the void sinks mechanism. This mechanism involves an application of vacuum evacuation, bubble mobility and increased matrix pressure. It depends on the materials (fibre and matrix) and the manufacturing process used. Therefore, in order to monitor and reduce the formation of porosity during composite manufacturing, Fernlund et al. (2016) proposed a model by assuming that gas is an ideal gas and flows according to Darcy’s law. Hence, the governing differential Eq. (4.1) is applied to formulate Eq. (4.2).

Design, manufacturing processes and their effects on bio-composite properties

vp K v vp   $ p$ vt m vx vx

165

! ¼0

(4.1)

where ; p and t represents porosity, pressure (Pa) and time (s), respectively and K, m and x denote gas permeability (m2), gas dynamic viscosity (Pa.s) and distance (m), respectively. t ¼

   1 m  1 F pv 0:6 2 In  L  p0 K 0:9  p0

(4.2)

where t  ; p0 F , pv and L are minimum required de-bulk time, initial pressure, a specific or final porosity level, uniform pressure and length, respectively. In addition, porosity in natural fibre-reinforced polymer biocomposites increases with the fibre content. It increases at a high rate if the geometrical compaction limit is exceeded, which depends on the types of fibres used and the fibre orientation. Its levels could be critical because of the high effect on the mechanical behaviours of natural FRP composites. For instance, Madsen and Lilholt (2003) reported that porosity fractions increased from 0.04 to 0.08 when the fibre weight fractions increased from 0.56 to 0.72 in a non-treated flax fibre-reinforced polypropylene biocomposite. It is important to quantify for this defect in a model in order to have a better prediction of the properties (axial stiffness and tensile strength) of the concerned biocomposite. Porosity defects can occur in a biocomposites if the matrix flow path is increasingly complex during the manufacturing process stage (Baghaei et al., 2014). Further explanation was given on the effects of alkali treatment, hemp yarn and non-woven types on porosity of the hemp-reinforced polylactic acid (PLA) biocomposite samples. The maximum improvement in terms of mechanical properties was achieved with alkalitreated hemp/PLA yarn when compared with the untreated counterparts. This was attributed to the better compact and close packing of the PLA/hemp yarns within the biocomposite system, different from the PLA/hemp non-woven biocomposite sample. Moreover, an occurrence of porosity depends on the type of manufacturing process and degree of process parameters considered. During the extrusion process, relative high screw speed could lead to high porosity due to non-uniform dispersion of the fibres and shorter residence time, and subsequently, reduce the mechanical properties (such as tensile strength) of the biocomposites (Ku et al., 2011). In addition, a low curing pressure (Fernlund et al., 2016), high humidity (Grunenfelder and Nutt, 2010; Kay and Fernlund, 2012), insufficient de-bulk or gas removal (Kay and Fernlund, 2012) and deficient vacuum (Kay and Fernlund, 2012; Centea and Hubert, 2014) during OoA processing of prepregs could result to porous biocomposite materials. However, a low porosity in OoA prepreg processing is possible by keeping volatiles in solution, through vacuum evacuation of trapped air and resin infiltration, as a three-step approach recommended by Fernlund et al. (2016). Hakim et al. (2017) reported that the presence of porosity reduced the mechanical (both mode I cyclic strain energy release fatigue life and mode I static inter-laminar fracture toughness) of carbon

166

Sustainable Composites for Lightweight Applications

fibre-reinforced polymer composites. Matzkanin and Yolken (2007) stated that approximately 7% of inter-laminar strength can be reduced for every 1% of voids present in a fibre-reinforced polymer composites, up to a void content of 4%. In addition, a nearly 5% and 50% of strength and fatigue life of a fibre-reinforced polymer composite product are reduced for each 1% increase in voids. Therefore, these composites are more liable to degradation in harsh environments. Between 2% and 2.5% void limit is accepted in practice (Knoeller, 2018). Both voids can be excluded from the final fibre-reinforced polymer biocomposite product by drying prepreg in a room with a controlled humidity before lamination and pores should be eliminated during the OoA process sincerely pores are almost impossible to eliminate from the prepreg (Wood and Bader, 1994).

4.5.6 Other manufacturing defects Other manufacturing defects in biocomposite materials include fibre and/or resin starved and rich areas (Fig. 4.31), usually caused due to non-uniform or uncontrolled distribution of fibres and flow of resin during moulding, respectively. Also, there are fibre kinking or waviness and misalignment of fibres. During prepreg preparation, pultrusion and filament winding processes, and improper tension of fibres causes a very complex phenomenon, called fibre kinking. Fig. 4.33 depicts an optical microscopic cross-section of a well-examined squared panel specimen after curing, with both low and high random waviness or kinking defect, and their failure progressions (Fig. 4.34).

(a)

(b)

High waviness specimen

304 mm

Area with high waviness

(c)

Low waviness specimen

304 mm

Figure 4.33 A well examined (a) composite panel, showing both (b) high and (c) low fibre waviness manufacturing defects (Elhajjar et al., 2016).

Design, manufacturing processes and their effects on bio-composite properties

High waviness specimens

(b)

Low waviness specimens

Fibre kinking

Damage progression

(a)

167

Delamination onset

Fibre fracture

Progressive delamination

Figure 4.34 Failure evolution of (a) high and (b) low fibre waviness or kinking defect of a composite panel (Elhajjar et al., 2016).

Fibre misalignment is resulted from washing out of fibres by excessive flow of resin, non-conformity of the pre-selected lay-up and/or fibre filament winding arrangement and misoriented fibres (Fig. 4.35). Scratching or cutting of fibres often result to a broken filaments or fibre fracture, as another common biocomposite material manufacturing-induced defect.

1000 Pm

Figure 4.35 Fibre misalignment defect during the manufacturing process (Krishnamurthy, 2006).

168

Sustainable Composites for Lightweight Applications

Last, there is a need for all the aforementioned manufacturing defects to be properly detected and characterised. Therefore, the subsequent Chapter five discusses different techniques used to characterise these defects. Importantly, Chapter six provides some improvements in designs and processes (manufacturing techniques) that are required and relevant to minimise the defects and enhance the properties of FRP biocomposites.

4.6 Conclusions The natural sources, better production processes and inherent properties of several fibres and matrices used for the manufacturing of reinforced polymeric biocomposites support the concepts of eco-design and sustainability. The era has gone when the most dependable and used engineering materials were only metals and alloys. The renewability, recyclability and biodegradability of many biocomposite products have been enhancing the design for environment. Actually, product design for the environment involves material, production, distribution, use and recovery stages. Each of these stages is very important and cannot be compromised in order to protect environments from harmful processes and products. Evidently, biocomposite technology is a key driver for cleaner production in terms of a low amount of energy consumption and waste emission, as well as the use of renewable energy, during processing. Biocomposite materials are easy to manipulate into final useful or desirable products, through well-designed manufacturing procedures. These include correct choice of fibres and matrices, suitable matrix polymer modification, efficient bio-fibre surface treatment and optimal manufacturing process. With the present advancement in processing and production technologies for fibrereinforced polymeric biocomposite, compression moulding, resin transfer moulding, extrusion and injection moulding techniques are the major manufacturing processes. But, an innovative design and development of a single method and combination of different processes are increasing. For example, extrusion-injection and extrusioncompression moulding with different screw, mould and die designs, the advent of AM or 3D and 4D printing, in addition to the application of automated or robotic systems with the help of efficient engineering software packages. However, fibre-reinforced polymeric biocomposites are susceptible to some manufacturing defects, as discussed. These defects could be caused during materials processing by environmental, mechanical and/or human factors. Defects reduce the quality of materials, as well as mechanical, thermal, optical, acoustic and electrical properties of the manufactured biocomposites. Hence, manufacturing defects must be avoided at all costs to maximally benefit from the outstanding properties of sustainable and lightweight biocomposites during various applications.

References Abdurohman, K., Satrio, T., Muzayadah, N.L., Teten, et al., 2018. A comparison process between hand lay-up, vacuum infusion and vacuum bagging method toward e-glass EW 185/ lycal composites. J. Phys. Conf. Ser. 1130, 1e11.

Design, manufacturing processes and their effects on bio-composite properties

169

Ahmed, K., Shiblee, M.D.N.I., Khosla, A., Nagahara, L., Thundat, T., Furukawa, H., 2020. Reviewdrecent progresses in 4D Printing of gel materials. J. Electrochem. Soc. 167, 1e13. Ahmed, K.S., Vijayarangan, S., 2008. Tensile, flexural and interlaminar shear properties of woven jute and jute-glass fabric reinforced polyester composites. J. Mater. Process. Technol. 207 (1e3), 330e335. Ahn, D., Kweon, J.H., Choi, J., Lee, S., 2012. Quantification of surface roughness of parts processed by laminated object manufacturing. J. Mater. Process. Technol. 212, 339e346. Akil, H.M., Omar, M.F., Mazuki, A.A.M., Safiee, S., Ishak, Z.A.M., Abu Bakar, A., 2011. Kenaf fibre reinforced composites: a review. Mater. Des. 32, 4107e4121. Alavudeen, A., Rajini, N., Karthikeyan, S., Thiruchitrambalam, M., Venkateshwaren, N., 2015. Mechanical properties of banana/Kenaf fibre-reinforced hybrid polyester composites: effect of woven fabric and random orientation. Mater. Des. 66 (Part A), 246e257. AlMaadeed, M.A., Kahraman, R., Khanam, P.N., Madi, N., 2012. Date palm wood flour/glass fibre reinforced hybrid composites of recycled polypropylene: mechanical and thermal properties. Mater. Des. 42, 289e294. Arbelaiz, A., Fernandez, B., Cantero, G., Ponte, R.L., Valea, A., Mondragon, I., 2005a. Mechanical properties of flax fibre/polypropylene composites: influence of fibre/matrix modification and glass fibre hybridization. Compos. A Appl. Sci. Manuf. 36 (12), 1637e1644. Arbelaiz, A., Fernandez, B., Ramos, J.A., Retegi, R., Llano-Ponte, R., Mondragon, I., 2005b. Mechanical properties of short flax fibre bundle/polypropylene composites: influence of matrix/fibre modification, fibre content, water uptake and recycling. Compos. Sci. Technol. 65 (10), 1582e1592. Arrakhiz, F.Z., El Achaby, M., Kakou, A.C., Vaudreuil, S., Benmoussa, K., Bouhfid, R., FassiFehri, O., Qaiss, A., 2012. Mechanical properties of high density polyethylene reinforced with chemically modified coir fibres: impact of chemical treatments. Mater. Des. 37, 379e383. Arthanarieswaran, V.P., Kumaravel, A., Kathirselvam, M., 2014. Evaluation of mechanical properties of banana and sisal Fibre reinforced epoxy composites influence of glass fibre hybridisation. Mater. Des. 64, 194e202. Asaithambi, B., Ganesan, G., AnandaKumar, S., 2014. Bio-composites: development and mechanical characterization of banana/sisal fibre reinforced poly lactic acid (PLA) hybrid composites. Fibres Polym. 15 (4), 847e854. Athijayamani, A., Thiruchitrambalam, M., Natarajan, U., Pazhanivel, B., 2009. Effect of moisture absorption on the mechanical properties of randomly oriented natural fibres/ polyester hybrid composite. Mater. Sci. Eng. A 517 (1e2), 344e353. Athjayamani, A., Thiruchitrambalam, M., Manikandan, M., Pazhanivel, B., 2010. Mechanical properties of natural fibres reinforced polyester hybrid composite. Int. J. Plast. Technol. 14 (1), 104e116. Baghaei, B., Skrifvars, M., Salehi, M., Bashir, T., Rissanen, M., Nousiainen, P., 2014. Novel aligned hemp fibre reinforcement for structural bio-composites: porosity, water absorption, mechanical performances and viscoelastic behaviour. Compos. A Appl. Sci. Manuf. 61, 1e12. Balasubramanian, K., Sultan, M.T.H., Rajeswari, N., 2018. Manufacturing techniques of composites for aerospace applications. In: Jawaid, M., Thariq, M. (Eds.), Structure Composites for Aerospace Applications, 1st ed. Chapter 4, Woodhead Publishing Series in Composites Science and Engineering, Elsevier Woodhead Publishing Limited, Cambridge, pp. 55e67. Berman, B., 2012. 3D printing: the new industrial revolution. Bus. Horiz. 55, 155e162.

170

Sustainable Composites for Lightweight Applications

Bledzki, A.K., Franciszczak, P., Meljon, A., 2015. High performance hybrid PP and PLA biocomposites reinforced with short man-made cellulose fibres and softwood flour. Compos. A Appl. Sci. Manuf. 74, 132e139. Bledzki, A.K., Gassan, J., 1998. Composites reinforced with cellulose based fibres. Prog. Polym. Sci. 24, 221e274. Boopalan, M., Niranjana, M., Umapathy, M.J., 2013. Study on mechanical properties and thermal properties of jute and banana fibre reinforced epoxy hybrid composite. Compos. B Eng. 51, 54e57. Boroujeni, A.Y., Tehrani, M., Nelson, A.J., Al-Haik, M., 2014. Hybrid carbon nanotubee carbon fibre composites with improved in-plane mechanical properties. Compos. B Eng. 66, 475e483. Bourmaud, A., Dhakal, H., Habrant, A., Padovani, J., Siniscalco, D., Ramage, M.H., Beaugrand, J., Shah, D.U., 2017. Exploring the potential of waste leaf sheath date palm fibres for composite reinforcement through a structural and mechanical analysis. Compos. A Appl. Sci. Manuf. 103, 292e303. Bowkett, M., Thanapalan, K., 2017. Comparative analysis of failure detection methods of composites materials’ systems. Syst. Sci. Control Eng. 5 (1), 168e177. Centea, T., Grunenfelder, L.K., Nutt, G.S.R., 2015. A review of out-of-autoclave prepregs e material properties, process phenomena, and manufacturing considerations. Compos. A Appl. Sci. Manuf. 70, 132e154. Centea, T., Hubert, P., 2014. Out-of-autoclave prepreg consolidation under deficient pressure conditions. J. Compos. Mater. 48, 2033e2045. Cucinotta, F., Guglielmino, E., Sfravara, F., 2017. Life cycle assessment in yacht industry: a case study of comparison between hand lay-up and vacuum infusion. J. Clean. Prod. 142, 3822e3833. Dhakal, H.N., Skrifvars, V., Adekunle, K., Zhang, Z.Y., 2014. Falling weight impact response of jute/methacrylated soybean oil bio-composites under low-velocity impact loading. Compos. Sci. Technol. 92, 134e141. Dhakal, H.N., Zhang, Z.Y., Guthrie, R., MacMullen, J., Bennett, N., 2013. Development of flax/ carbon fibre hybrid composites for enhanced properties. Carbohydr. Polym. 96 (1), 1e8. Dhakal, H.N., Zhang, Z.Y., Richardson, M.O.W., 2007. Effect of water absorption on the mechanical properties of hemp fibre reinforced unsaturated polyester composites. Compos. Sci. Technol. 67, 1674e1683. Dixit, D., Pal, R., Kapoor, G., Stabenau, M., 2016. Lightweight composite materials processing. In: Bhatnagar, A. (Ed.), Lightweight Ballistic Composites: Military and Law e Enforcement Applications, second ed. Elsevier Woodhead Publishing Limited, Cambridge, pp. 157e216 (Chapter 6). Dong, C., Davies, I.J., 2015. Flexural strength of bidirectional hybrid epoxy composites reinforced by E glass and T700S carbon fibres. Compos. B Eng. 72, 65e71. Dumont, P., Orgéas, L., Corre, S.L., 2003. Anisotropic viscous behavior of sheet molding compounds (SMC) during compression moulding. Int. J. Plast. 19 (5), 625e646. Elhajjar, R.F., Shams, S.S., Kemeny, G.J., Stuessy, G., 2016. A hybrid numerical and imaging approach for characterizing defects in composite structures. Compos. A Appl. Sci. Manuf. 81, 98e104. Fang, Q., Hanna, M.A., 1999. Rheological properties of amorphous and semi crystalline poly lactic acid polymers. Ind. Crop. Prod. 10 (1), 47e53. Farhang, L., 2014. Void Evolution during Processing of Out-Of-Autoclave Prepreg Laminates. Published PhD Thesis. The University of British Columbia, Vancouver, Canada.

Design, manufacturing processes and their effects on bio-composite properties

171

Faruk, O., Bledzki, A.K., Fink, H., Sain, M., 2012. Bio-composites reinforced with natural fibres: 2000e2010. Prog. Polym. Sci. 37 (11), 1552e1596. Ferland, P., Guittard, D., Trochu, F., 1996. Concurrent methods for permeability measurement in resin transfer molding. Polym. Compos. 17 (1), 149e158. Fernlund, G., Wells, J., Fahrang, L., Kay, J., Poursartip, A., 2016. Causes and remedies for porosity in composite manufacturing. Mater. Sci. Eng. 139, 1e16. Francucci, G., Rodrıguez, E.S., Vazquez, A., 2012. Experimental study of the compaction response of jute fabrics in liquid composite molding processes. J. Compos. Mater. 46 (2), 155e167. Gallos, A., Paes, G., Allais, F., Beaugrand, J., 2017. Lignocellulosic fibres: a critical review of the extrusion process for enhancement of the properties of natural fibre composites. RSC Adv. 7, 34638e34654. Ge, Q., Qi, H.J., Dunn, M.L., 2013. Active materials by four-dimension printing. Appl. Phys. Lett. 103, 1e5. George, G., Joseph, K., Boudenne, A., Thomas, S., 2010. Recent advances in green composites. Key Eng. Mater. 425, 107e166. Goh, G.D., Yap, Y.L., Agarwala, S., Yeong, W.Y., 2019. Recent progress in additive manufacturing of fibre reinforced polymer composite. Adv. Mater. Technol. 4, 1e22. Goutianos, S., Peijs, T., Nystrom, B., Skrifvars, M., 2006. Development of flax fibre based textile reinforcements for composite applications. Appl. Compos. Mater. 13 (4), 199e215. Graupner, N., Herrmann, A.S., M€ussig, J., 2009. Natural and man-made cellulose fibrereinforced polylactic acid PLA composites: an overview about mechanical characteristics and application areas. Compos. A Appl. Sci. Manuf. 40 (6e7), 810e821. Grunenfelder, L.K., Nutt, S.R., 2010. Void formation in composite prepregs e effect of dissolved moisture. Compos. Sci. Technol. 70, 2304e2309. Guermazi, N., Haddar, N., Elleuch, K., Ayedi, H.F., 2014. Investigations on the fabrication and the characterization of glass/epoxy, carbon/epoxy and hybrid composites used in the reinforcement and the repair of aeronautic structures. Mater. Des. 56, 714e724. Gupta, M.K., Srivastava, R.K., 2016. Mechanical properties of hybrid fibres-reinforced polymer composite: a review. Polym. Plast. Technol. Eng. 55 (6), 626e642. Hagstrand, P.O., Oksman, P., 2001. Mechanical properties and morphology of flax fibre reinforced melamine-formaldehyde composites. Polym. Compos. 22 (4), 568e578. Hakim, I.A., Donaldson, S.L., Meyendorf, N.G., Browning, C.E., 2017. Porosity effects on inter-laminar fracture behavior in carbon fibre-reinforced polymer composites. Mater. Sci. Appl. 8, 170e187. Halley, P.J., 2012. Rheology of thermosets: the use of chemorheology to characterise and model thermoset flow behaviour. In: Guo, Q. (Ed.), Thermosets: Structure, Properties and Applications, first ed. Elsevier Woodhead Publishing Limited, Cambridge, pp. 92e117 (Chapter 4). Hang, X., Li, Y., Hao, X., Li, N., Wen, Y., 2015. Effects of temperature profiles of microwave curing processes on mechanical properties of carbon fibreereinforced composites. Proc. IME B J. Eng. Manufact. 231 (8), 1332e1340. Helwig, M., Paukszta, D., 2006. Flammability of composites based on polypropylene and flax fibres. Mol. Cryst. Liq. Cryst. Sci. Technol. Sect. A 354 (1), 373e380. Herrmann, A.S., Nickel, J., Riedel, U., 1998. Construction materials based upon biologically renewable resourcesdfrom components to finished parts. Polym. Degrad. Stabil. 59 (1e3), 251e261. Ho, M.P., Wang, H., Lee, J.H., Ho, C.K., Lau, K.T., Leng, J., Hui, D., 2012. Critical factors on manufacturing processes of natural fibre composites. Compos. B Eng. 43 (8), 3549e3562.

172

Sustainable Composites for Lightweight Applications

Hu, R., Lim, J.K., 2007. Fabrication and mechanical properties of completely biodegradable hemp fibre reinforced polylactic acid composites. J. Compos. Mater. 41 (13), 1e15. Huda, M.S., Drzal, L.T., Misra, M., Mohanty, A.K., 2006a. Wood-fibre-reinforced poly (lactic acid) composites: evaluation of the physicomechanical and morphological properties. J. Appl. Polym. Sci. 102, 4856e4869. Huda, M.S., Drzal, L.T., Misra, M., Mohanty, A.K., Williams, K., Mielewski, D.F., 2005b. A study on bio-composites from recycled newspaper fibre and poly (lactic acid). Ind. Eng. Chem. Res. 44 (15), 5593e5601. Huda, M.S., Drzal, L.T., Mohanty, A.K., Misra, M., 2006b. Chopped glass and recycled newspaper as reinforcement fibres in injection molded poly (lactic acid) (PLA) composites: a comparative study. Compos. Sci. Technol. 66 (11e12), 1813e1824. Huda, M.S., Mohanty, A.K., Drzal, L.T., Schut, E., MIsra, M., 2005a. “Green” composites from recycled cellulose and poly (lactic acid): physico-mechanical and morphological properties evaluation. J. Mater. Sci. 40 (16), 4221e4229. Ikegawa, N., Hamada, H., Maekawa, Z., 1996. Effect of compression process on void behavior in structural resin transfer molding. Polym. Eng. Sci. 36 (7), 953e962. Isa, M.T., Ahmed, A.S., Aderemi, B.O., Taib, R.M., Mohammed-Dabo, I.A., 2013. Effect of fibre type and combinations on the mechanical, physical and thermal stability properties of polyester hybrid composites. Compos. B Eng. 52, 217e223. Jawaid, M., Khalil, H.P.S.A., Bakar, A.A., 2010. Mechanical performance of oil palm empty fruit bunches/jute fibres reinforced epoxy hybrid composites. Mater. Sci. Eng. A 527 (29e30), 7944e7949. Jawaid, M., Khalil, H.P.S.A., Hassan, A., Dungani, R., Hadiyane, A., 2013. Effect of jute fibre loading on tensile and dynamic mechanical properties of oil palm epoxy composites. Compos. B Eng. 45, 619e624. J unior, J.H.S.A., Amico, S.C., Botelho, E.C., Amado, F.D.R., 2013. Hybridization effect on the mechanical properties of curaua/glass fibre composites. Compos. B Eng. 55, 492e497. Kay, J., Fernlund, G., 2012. Processing conditions and voids in out-of-autoclave prepregs. In: Proceedings of the Society for the Advancement of Material and Process Engineering (SAMPE) Conference, United States: Baltimore. Khan, M.A., Ganster, J., Fink, H.P., 2009. Hybrid composites of jute and man-made cellulose fibres polypropylene by injection moulding. Compos. A Appl. Sci. Manuf. 40 (6e7), 846e851. Khanam, P.N., Khalil, H.P.S.A., Jawaid, M., Reddy, G.R., Narayana, C.S., Naidu, S.V., 2010. Sisal/carbon fibre reinforced hybrid composites tensile, flexural and chemical resistance properties. J. Polym. Environ. 18 (4), 727e733. Kim, S.K., Daniel, I.M., 2003. Determination of three-dimensional permeability of fibre performs by the inverse parameter estimation technique. Compos. A Appl. Sci. Manuf. 34, 421e429. Kim, S.K., Lee, S.W., Youn, J.R., 2002. Measurement of residual stresses in injection molded short fibre composites considering anisotropy and modulus variation. Korea Aust. Rheol. J. 14 (3), 107e114. Knoeller, S.L., 2018. Polymer matrix composite manufacturing induced defects. Adv. Mater. Manuf. Test. Inf. Anal. Cent. Q. 5 (4), 7e10. Kozaczuk, K., 2016. Automated fibre placement systems overview. Trans. Instit. Aviat. 245 (4), 52e59. Krishnamurthy, S., 2006. Prestressed Advanced Fibre Reinforced Composites: Fabrication and Mechanical Performance. Published Ph.D. thesis. Engineering System Department, Defence College of Management and Technology, United Kingdom: Cranfield University.

Design, manufacturing processes and their effects on bio-composite properties

173

Ku, H., Wang, H., Pattarachaiyakoop, N., Trada, M., 2011. A review on the tensile properties of natural fibre reinforced polymer composites. Compos. B Eng. 42 (4), 856e873. Kumar, N.R., Prasad, G.R., Rao, B.R., 2013. Investigation on mechanical properties of banana fibre glass reinforced hybrid thermoplastic composites. Int. J. Eng. Res. Technol. 2 (11), 3701e3706. Lei, W., Lei, W., Ren, C., 2006. Effect of volume fraction of ramie cloth on physical and mechanical properties of ramie cloth/UP resin composite. Trans. Nonferrous Metals Soc. China 16 (2), 474e477. Li, L., Sun, Q., Bellehumeur, C., Gu, P., 2002. Composite modelling and analysis for fabrication of FDM prototypes with locally controlled properties. J. Manuf. Process. 4 (2), 129e141. Li, X., Shang, J., Wang, Z., 2017. Intelligent materials: a review of applications in 4D printing. Assemb. Autom. 37 (2), 170e185. Li, Z., 2015. Tension control system design of a filament winding structure based on fuzzy neural network. Eng. Rev. 35 (1), 9e17. Lotfi, A., Li, H., Dao, D.V., Prusty, G., 2019. Natural fibreereinforced composites: a review on material, manufacturing, and machinability. J. Thermoplast. Compos. Mater. 20 (10), 1e47. Madsen, B., 2004. Properties of Plant Fibre Yan Polymer Composites: An Experimental Study. Technical University of Denmark, Lyngby, Denmark. Madsen, B., Lilholt, H., 2003. Physical and mechanical properties of unidirectional plant fibre composites e an evaluation of the influence of porosity. Compos. Sci. Technol. 63 (9), 1265e1272. Madsen, B., Thygesen, A., Lilholt, H., 2009. Plant fibre composites e porosity and stiffness. Compos. Sci. Technol. 69 (7e8), 1057e1069. Malkapuram, R., Kumar, V., Negi, Y.S., 2009. Recent development in natural fibre reinforced polypropylene composites. J. Reinforc. Plast. Compos. 28 (10), 1169e1189. Mallick, P.K., 2008. Fibre-Reinforced Composites: Materials, Manufacturing and Design, third ed. CRC Press, Taylor and Francis Group, Boca Raton, Florida, USA. Mateen, M., Shankar, D.R., Hussain, M.M., 2018. Design and development of low cost two axis filament winding machine. J. Adv. Manuf. Technol. 12 (1), 117e126. Matzkanin, G.A., Yolken, H.T., 2007. Techniques for the nondestructive evaluation of polymer matrix composites. Adv. Mater. Manuf. Test. Inf. Anal. Cent. Q. 2 (4), 3e7. Melenka, G.W., Schofield, J.S., Dawson, M.R., Carey, J.P., 2015. Evaluation of dimensional accuracy and material properties of the MakerBot 3D desktop printer. Rapid Prototyp. J. 21 (5), 618e627. Miao, S., Castro, N., Nowicki, M., Xia, L., Cui, H., Zhou, X., et al., 2017. 4D printing of polymeric materials for tissue and organ regeneration. Mater. Today 20 (10), 577e591. Minsch, N., Herrmann, F., Gereke, T., Nocke, A., Cherif, C., 2017. Analysis of filament winding processes and potential equipment technologies. Procedia CIRP 66, 125e130. Mitchell, A., Lafont, U., Holynska, M., Semprimoschnig, C., 2018. Additive manufacturing - a review of 4D printing and future applications. Addit. Manuf. 24, 606e626. Modniks, J., Andersons, J., 2013. Modelling the non-linear deformation of a short-flax fibrereinforced polymer composite by orientation averaging. Compos. B Eng. 54, 188e193. Mohanty, A.K., Khan, M.A., Sahoo, S., Hinrichsen, G., 2000. Effect of chemical modification on the performance of biodegradable jute yarn-Biopol® composites. J. Mater. Sci. 35 (10), 2589e2595. Mohanty, A.K., Misra, M., Drzal, L.T., 2002. Sustainable bio-composites from renewable resources: opportunities and challenges in the green materials world. J. Polym. Environ. 10, 19e26.

174

Sustainable Composites for Lightweight Applications

Monteiro, S.N., Calado, V., Rodriguez, R.J.S., Margem, F.M., 2012. Thermogravimetric stability of polymer composites reinforced with less common lignocellulosic fibres e an overview. J. Mater. Res. Technol. 1 (2), 117e126. Mourad, S., 2012. Innovation: A Key Factor for Cleaner Production e the Case of Technical Textiles and its Role in Reducing Pollution. A document presented at a United Nation workshop organised in Damascus in Nov 2008. Available at: https://www.slideshare.net/ mourasab/innovation-cleaner-production. Ning, F., Cong, W., Qiu, J., Wei, J., Wang, S., 2015. Additive manufacturing of carbon fibre reinforced thermoplastic composites using fused deposition modelling. Compos. B Eng. 80, 369e378. NPTEL, 2019a. Lectures 5.4: Hand Lay-Up and Spray Lay-Up. Available at: https://nptel.ac.in/ courses/112107085/module5/lecture4/lecture4.pdf. (Accessed 7 August 2019). NPTEL, 2019b. Lectures 5.7: Filament Winding. Available at: https://nptel.ac.in/courses/ 112107085/module5/lecture7/lecture7.pdf. (Accessed 9 August 2019). Nystr€ om, B., 2007. Natural Fibre Composites Optimization of Microstructure and Processing Parameters. Published Licentiate Thesis, Luleå, Sweden: Luleå University of Technology, Department of Applied Physics and Mechanical Engineering Division of Polymer Engineering. Oksman, K., Mathew, A.P., Långstr€om, R., Nystr€om, B., Joseph, K., 2009. The influence of fibre microstructure on fibre breakage and mechanical properties of natural fibre reinforced polypropylene. Compos. Sci. Technol. 69 (11e12), 1847e1853. Pan, Y., Patil, A., 2017. A novel projection based electro-stereolithography (PES) process for production of 3D polymer-particle composite objects. Rapid Prototyp. J. 23 (2), 236e245. Parandoush, P., Lin, D., 2017. A review on additive manufacturing of polymer-fibre composites. Compos. Struct. 182, 36e53. Parlevliet, P.P., Bersee, H.E.N., Beukers, A., 2006. Residual stresses in thermoplastic composites: a study of the literature e Part I: formation of residual stresses. Compos. A Appl. Sci. Manuf. 37 (11), 1847e1857. Parlevliet, P.P., Bersee, H.E.N., Beukers, A., 2007. Residual stresses in thermoplastic composites: a study of the literature e Part III: effects of thermal residual stresses. Compos. A Appl. Sci. Manuf. 38, 1581e1596. Patel, I., March 7, 2017. Lecture on Integrated Product Engineering. Studynet Module. Masters Taught Postgraduate Course 2017/18. University of Hertfordshire, United Kingdom, Hatfield. Pérez-Fonseca, A.A., Robledo-Ortíz, J.R., Ramirez-Arreola, D.E., Ortega-Gudi~ no, P., Rodrigue, D., Gonzalez-Nu~nez, R., 2014. Effect of hybridization on the physical and mechanical properties of high density polyethyleneepine/agave composites. Mater. Des. 64, 35e43. Pickering, K.L., Efendy, M.G.A., Le, T.M., 2016. A review of recent developments in natural fibre composites and their mechanical performance. Compos. A Appl. Sci. Manuf. 83, 98e112. Piedade, A.P., 2019. 4D Printing: the Shape-morphing in additive manufacturing. J. Funct. Biomater. 10 (9), 1e7. Pineda, E.J., Bednarcyk, B.A., Waas, A.M., Arnold, S.M., 2013. Progressive failure of a unidirectional fibre-reinforced composite using the method of cells: discretisation objective computational results. Int. J. Solid Struct. 50 (9), 1203e1216. Prabhakar, M.M., Rajini, N., Mayandi, K., Subash, K., Subramanian, S., Seelan, S.V.R.B., Ayrilmis, N., 2019. Impact response of basalt composite pipe using filament winding. J. Phys. Conf. Ser. 1240, 1e6.

Design, manufacturing processes and their effects on bio-composite properties

175

Quanjin, M., Rejab, M.R.M., Idris, M.S., Kumar, N.M., Merzuki, M.N.M., 2018a. Robotic filament winding technique (RFWT) in industrial application: a review of state of the art and future perspectives. Int. Res. J. Eng. Technol. 5 (12), 1668e1676. Quanjin, M., Rejab, M.R.M., Kumar, N.M., Idris, M.S., 2019. Experimental assessment of the 3axis filament winding machine performance. Results Eng. 2, 1e3. Quanjin, M., Rejab, M.R.M., Sahat, I.M., Amiruddin, M., Bachtiar, D., Siregar, J.P., Ibrahim, M.I., 2018b. Design of portable 3-axis filament winding machine with inexpensive control system. J. Mech. Eng. Sci. 12 (1), 3479e3493. Quarshie, R., Carruthers, J., 2014. Technological Overview: Bio-Composites. Net composites, pp. 1e40. Available at: https://netcomposites.com/media/1211/biocomposites-guide.pdf. (Accessed 5 August 2019). Rajak, D.K., Pagar, D.D., Menezes, P.L., Linul, E., 2019. Fibre-reinforced polymer composites: manufacturing, properties, and applications. Polymers 11, 1e37. Rajkumar, A.R., Shanmugam, K., 2018. Additive manufacturing-enabled shape transformations via FFF 4D printing. J. Mater. Res. 33 (24), 4362e4376. Ramesh, M., Palanikumar, K., Reddy, K.H., 2013. Mechanical property evaluation of sisale juteeglass fibre reinforced polyester composites. Compos. B Eng. 48, 1e9. Ramnath, B.V., Kokan, S.J., Raja, R.N., Sathyanarayanan, R., Elanchezhian, C., Prasad, A.R., Manickavasagam, V.M., 2013. Evaluation of mechanical properties of abacaejuteeglass fibre reinforced epoxy composite. Mater. Des. 51, 357e366. Ramnath, B.V., Manickavasagam, V.M., Elanchezhian, C., Krishna, C.V., Karthik, S., Saravanan, K., 2014. Determination of mechanical properties of intra-layer abacaejutee glass fibre reinforced composite. Mater. Des. 60, 643e652. Rassmann, S., Reid, R.G., Paskaramoorthy, R., 2010. Effects of processing conditions on the mechanical and water absorption properties of resin transfer moulded kenaf fibre reinforced polyester composite laminates. Compos. A Appl. Sci. Manuf. 41 (11), 1612e1619. Richard, M.G., 2013. Cannabis in Your Car Doors, but Not to Smuggle it. Available at: https:// www.treehugger.com/sustainable-agriculture/cannabis-your-car-doors-not-smuggle-it. html. (Accessed 8 June 2020). Rodi, E.G., Mangeon, C., Dessauw, E., Sansalone, V., Lemaire, T., Renard, E., Langlois, V., 2016. Functionalization of miscanthus by photoactivated thiolene addition to improve interfacial adhesion with polycaprolactone. ACS Sustain. Chem. Eng. 4, 5475e5482. Rudin, A., Choi, P., 2013. Biopolymers. The Elements of Polymer Science and Engineering, third ed. Elsevier Inc., USA. Salim, M.S., Ishak, Z.A.M., Hamid, S.A., 2011. Effect of stitching density of nonwoven fibre mat towards mechanical properties of kenaf reinforced epoxy composites produced by resin transfer moulding (RTM). Key Eng. Mater. 471e472, 987e992. Sanjay, M.R., Arpitha, G.R., Yogesha, B., 2015. Study on mechanical properties of natural e Glass Fibre reinforced polymer hybrid composites: a review. Mater. Today Proc. 2 (4e5), 2959e2967. Sanjay, M.R., Madhu, P., Jawaid, M., Senthamaraikannan, P., Senthil, S., Pradeep, S., 2018. Characterization and properties of natural fibre polymer composites: a comprehensive review. J. Clean. Prod. 172, 566e581. Santulli, C., Sarasini, F., Tirillo, J., Valente, T., Valente, M., Caruso, A.P., Infantino, M., Nisini, E., Minak, G., 2013. Mechanical behaviour of jute cloth/wool felts hybrid laminates. Mater. Des. 50, 309e321. Sebe, G., Cetin, N.S., Hill, C.A.S., Hughes, M., 2000. RTM hemp fibre-reinforced polyester composites. Appl. Compos. Mater. 7 (5e6), 341e349.

176

Sustainable Composites for Lightweight Applications

Serizawa, S., Inoue, K., Iji, M., 2006. Kenaf-fibre-reinforced poly (lactic acid) used for electronic products. J. Appl. Polym. Sci. 100 (1), 618e624. Shahzad, A., 2011. Hemp fibre and its composites e a review. J. Compos. Mater. 46, 973e986. Shahzad, K., Deckers, J., Zhang, Z., Kruth, J.P., Vleugels, J., 2014. Additive manufacturing of zirconia parts by indirect selective laser sintering. J. Eur. Ceram. Soc. 32, 81e89. Shanmugam, D., Thiruchitrambalam, M., 2013. Static and dynamic mechanical properties of alkali treated unidirectional continuous Palmyra palm leaf stalk fibre/jute fibre reinforced hybrid polyester composites. Mater. Des. 97, 533e542. Singh, B., Gupta, M., Verma, A., 1995. Mechanical behavior of particulate hybrid composite laminates as potential building materials. Construct. Build. Mater. 9 (1), 39e44. Sofi, T., Neunkirchen, S., Schledjewski, R., 2018. Path calculation, technology and opportunities in dry fibre winding: a review. Adv. Manuf. Polym. Compos. Sci. 4 (3), 57e72. Sohn, J.S., Ryu, Y., Yun, C.S., Zhu, K., Cha, S.W., 2019. Extrusion compounding process for the development of eco-friendly scg/pp composite pellets. Sustainability 11 (6), 1e12. Sreekumar, P.A., Joseph, K., Unnikrishnan, G., 2007. A comparative study on mechanical properties of sisal-leaf fibre-reinforced polyester composites prepared by resin transfer and compression moulding techniques. Compos. Sci. Technol. 67 (3e4), 453e461. Srinivasan, V.S., Boopathy, S.R., Sangeetha, D., Ramnath, B.V., 2014. Evaluation of mechanical and thermal properties of bananaeflax based natural fibre composite. Mater. Des. 60, 620e627. Strong, B.A., 2008. Fundamentals of Composites Manufacturing: Materials, Methods, and Applications, second ed. Society of Manufacturing Engineers, USA. Struzziero, G., Teuwen, J.J.E., Skordos, A.A., 2019. Numerical optimisation of thermoset composites manufacturing processes: a review. Compos. A Appl. Sci. Manuf. 124, 1e13. Sutter, J.K., Kenner, W.S., Pelham, L., Miller, S.G., Polis, D.L., Nailadi, C., et al., 2019. Comparison of Autoclave and Out-Of-Autoclave Composites. Available at: https://ntrs. nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20110016095.pdf. (Accessed 22 June 2020). Thakur, V.K., Vennerberg, D., Kessler, M.R., 2014. Green aqueous surface modification of polypropylene for novel polymer nanocomposites. ACS Appl. Mater. Interfaces 6, 9349e9356. Timmerman, J.F., Hayes, B.S., Seferis, J.C., 2003. Cure temperature effects on cryogenic micro cracking of polymer matrix materials. Polym. Compos. 24 (1), 132e139. Tungjitpornkull, S., Sombatsompop, N., 2009. Processing technique and fibre orientation angle affecting the mechanical properties of E-glass reinforced wood/PVC composites. J. Mater. Process. Technol. 209 (6), 3079e3088. Ulrich, K.T., Eppinger, S.D., 2012. Product Design and Development, fifth ed. Irwin McGrawHill, New York. Van de Velde, K., Kiekens, P., 2003. Effect of material and process parameters on the mechanical properties of unidirectional and multidirectional flax/polypropylene composites. Compos. Struct. 62 (3e4), 443e448. van Voorn, B., Smit, H.H.G., Sinke, R.J., de Klerk, B., 2001. Natural fibre reinforced sheet moulding compound. Compos. A Appl. Sci. Manuf. 32 (9), 1271e1279. Venkateshwaran, N., ElayaPerumal, A., 2010. Mechanical and water absorption properties of woven jute/banana hybrid composites. Fibres Polym. 13 (7), 907e914. Vilaplana, F., Str€omberg, E., Karlsson, S., 2010. Environmental and resource aspects of sustainable biocomposites. Polym. Degrad. Stabil. 95, 2147e2161. Wang, Q., Li, T., Wang, B., Liu, C., Huang, Q., Ren, M., 2020. Prediction of void growth and fiber volume fraction based on filament winding process mechanics. Compos. Struct. 246, 1e10.

Design, manufacturing processes and their effects on bio-composite properties

177

Warrior, N.A., Turner, T.A., Robitaille, F., Rudd, C.D., 2003. Effect of resin properties and processing parameters on crash energy absorbing composite structures made by RTM. Compos. A Appl. Sci. Manuf. 34 (6), 543e550. Weinan, E., Engquist, B., Li, X., Ren, W., Vanden-Eijnden, E., 2007. Heterogeneous multiscale methods: a review. Commun. Comput. Phys. 2 (3), 367e450. Wells, J., 2015. Behaviour of Resin Void in Out-Of-Autoclave Prepreg Processing. Published MSc Dissertation, Vancouver, Canada: The University of British Columbia. Willians, G.I., Wool, R.P., 2000. Composites from natural fibres and soy oil resins. Appl. Compos. Mater. 7 (5e6), 421e432. Wisnom, M.R., Gigliotti, M., Ersoy, N., Campbell, M., Potter, K.D., 2006. Mechanisms generating residual stresses and distortions during manufacture of polymer-matrix composite structures. Compos. A Appl. Sci. Manuf. 37 (4), 522e529. Wood, J.R., Bader, M.G., 1994. Void control for polymer-matrix composites (1): theoretical and experimental methods for determining the growth and collapse of gas bubbles. Compos. Manuf. 5 (3), 139e147. Wool, R.P., Sun, X.S., 2005. Bio-Based Polymers and Composites, first ed. Elsevier, Netherlands. Xian-bao, Y.U., Chun, W., Shao-rong, L.U., Jing-hong, Y.U., Chi-hua, X.U.D., 2006. Preparation and mechanical properties of TLCP/UP/GF in-situ hybrid composites. Trans. Nonferrous Metals Soc. China 16, 529e533. Yahaya, R., Sapuan, S.M., Jawaid, M., Leman, Z., Zainudin, E.S., 2014a. Mechanical performance of woven kenaf-Kevlar hybrid composites. J. Reinforc. Plast. Compos. 33 (24), 2242e2254. Yahaya, R., Sapuan, S.M., Jawaid, M., Leman, Z., Zainudin, E.S., 2015. Effect of layering sequence and chemical treatment on the mechanical properties of woven kenafearamid hybrid laminated composites. Mater. Des. 67, 173e179. Yahaya, R., Sapuan, S.M., Jawaid, M., Leman, Z., Zainudin, E.S., 2014b. Effects of kenaf contents and fibre orientation on physical, mechanical, and morphological properties of hybrid laminated composites for vehicle spall liners. Polym. Compos. 36, 1469e1476. Yahaya, R., Sapuan, S.M., Jawaid, M., Leman, Z., Zainudin, E.S., 2014c. Selection of natural fibre for hybrid laminated composites vehicle spall liners using analytical hierarchy process (AHP). Appl. Mech. Mater. 564, 400e405. Yan, X., Cao, S., 2018. Structure and interfacial shear strength of polypropylene-glass fibre/ carbon fibre hybrid composites fabricated by direct fibre feeding injection molding. Compos. Struct. 185, 362e372. Zhang, C., Xu, X., 2013. Finite element analysis of 3D braided composites based on three unitcells models. Compos. Struct. 98, 130e142. Zhang, J., Chaisombat, K., He, S., Wang, C.H., 2012. Hybrid composite laminates reinforced with glass/carbon woven fabrics for lightweight load bearing structures. Mater. Des. 36, 75e80. Zhang, Y., Li, Y., Ma, H., Yu, T., 2013. Tensile and interfacial properties of unidirectional flax/ glass fibre reinforced hybrid composites. Compos. Sci. Technol. 88, 172e177. Zhang, Z., Demir, K.G., Gu, G.X., 2019. Developments in 4D-printing: a review on current smart materials, technologies, and applications. Int. J. Soc. Netw. Min. 10 (3), 205e224.

Testing and damage characterisation of biocomposite materials 5.1

5

Introduction and context

The detrimental effects of synthetic materials to the environment in terms of pollution and the need for material sustainability have informed a recent and increased research on biocomposites worldwide, thus leading to noteworthy achievements in eco-friendly green technology (Faruk et al., 2012). Also, it has been reported that the mechanical behaviours of biocomposite materials are preferred when compared with their synthetic counterparts, having a comparative advantage in terms of cost and energy consumed in their production (Sanjay et al., 2018). These biocomposites are abundantly available by nature’s endowment and have found applications in such fields as electronics, aircraft, automobiles, sports (Shekar and Ramachandra, 2018), as well as construction, thus leading to the elimination of wastes in structural works (Sanjay et al., 2018). However, while biocomposites have gained commercial successes in these fields, care must be taken to ensure that the material matches the application of interest (Dicker et al., 2014). A number of variables affect the properties of biocomposites and these include fibre source environment, methods employed in their processing, type of fibre and fibre modification (Faruk et al., 2012).

5.2

Testing methods for damage characterisation and their importance

During their service life, composites materials undergo various loading conditions. Moreover, depending on loading methods employed, the damage modes and mechanisms involve complex scenarios where composites can lose its structural integrity fully or partially. In order to prevent such damages and failures, reliable damage detection and monitoring techniques are paramount important. Among these techniques, non-destructive evaluation (NDE) or also called non-destructive testing (NDT), is one of the most commonly used methods. This testing method requires that the damage on the surface of materials and their interiors are duly identified and characterised without cutting or modify the material in a sense that causes damage, i.e., materials are inspected, evaluated and characterised for defect assessment following established standards for testing and materials standards (ASTM E2533, 2017), without changing the fundamental features of the material nor causing any harm or damage to the material undergoing the test. Worthy of mention is the cost-effectiveness that these NDT

Sustainable Composites for Lightweight Applications. https://doi.org/10.1016/B978-0-12-818316-8.00007-4 Copyright © 2021 Elsevier Ltd. All rights reserved.

180

Sustainable Composites for Lightweight Applications

techniques provide to either a sample or whole material investigation (Gholizadeh, 2016). NDT methods in biocomposites have found applications in quite a lot of fields. These include, but are not limited to, nuclear industry, manufacturing, storage tanks, aerospace, nuclear industry, tube and pipe production industries, security, military and defence, as well as characterisation of manufacturing defects in biocomposites. There are numerous techniques that have been designed and utilised in the NDT of various biocomposites. These include the following methods: n n n n n n n n n n n

Visual inspection (VI)/testing (VT) Ultrasonic examination Thermographic testing Radiographic testing Electromagnetic testing Acoustic emission inspection Acousto-ultrasonic testing Shearography testing Computed tomography scanning X-ray micro-computed tomography examination Scanning electron microscopy.

Summarily, various categories of NDT and their evaluation techniques are illustrated in Fig. 5.1. The explanations, applications, benefits and drawbacks of the aforementioned NDT techniques are subsequently explained. Other commonly used, available and improved testing and evaluation techniques are listed as thus: optical testing, magnetic particle examination, liquid penetrant inspection, digital image correlation, vibration method, digital X-ray radiography and infrared thermography inspection, although, a few of these methods have been briefly discussed under some main NDT techniques.

Non-destructive testing & evaluation

Visual inspection

Acoustic wave-based

Optical techniques

Imaging techniques

Electromagnetic fields

Visual and optical testing Liquid penetrant Dye penetrant

Acoustic emission Nonlinear acoustics Ultrasonic testing Acoustoultrasonic

Infrared thermography Terahertz testing Shearography Digital image correlation

X-ray radiography

Eddy-current Remote field testing testing Magnetic Magnetic particle flux leakage testing inspection

Neutron radiography γ -ray radiography

Figure 5.1 Broad classification of NDT and their related evaluation techniques (Wang et al., 2020).

Testing and damage characterisation of biocomposite materials

5.2.1

181

Visual inspection or testing

Usually, visual inspection (VI) or testing (VT), as the utmost used and simple form of NDT, has the advantage of saving time, as well as money. This is because other tests may not be necessary or a reduction in the number of other types of testing. Most importantly, visual inspection (VI) has the merit of fast processing. The process is also relatively affordable because it does not require any equipment, although not without its intrinsic demerits. The main drawback of VI includes the ability to examine external defects and damage to bio/composite materials. It cannot detect internal and/ or microscopic damage, such as inter-laminar delamination, voids, particle inclusion, cracks, fibre fracture, wavering and kinking, to mentioning but a few. This limitation is ascribed to human sight capacity. For instance, Dhakal et al. (2014a) visually examined the effects of temperatures and low-impact velocities on jute FRP biocomposites, as shown in Fig. 5.2(a). The same visual method was used to examine the influence of low-velocity impact load on jute FRP/methacrylated soybean oil biocomposite samples (Dhakal et al., 2014b). Similarly, De Rose et al. (2012) conducted a visual inspection of hemp/unsaturated polyester biocomposite samples after post-impact static and cyclic flexural tests.

(a)

(b)

Front

Rear

Figure 5.2 Visual examination results of the impact damage responses on front and rear/back faces of (a) jute and (b) hemp FRP/unsaturated polyester biocomposite samples (Dhakal et al., 2014a; De Rose et al., 2012).

182

Sustainable Composites for Lightweight Applications

The results obtained are shown in Fig. 5.2(b). In addition, VI has been used to assess elastic deformation, crack initiation and propagation when investigating the influence of water absorption on hybrid flax/basalt FRP composites (Almansour et al., 2017).

5.2.2

Ultrasonic testing

In contrast to visual inspection, ultrasonic testing (UT) involves the use of some electronic components, such as a transmitter, receiver circuit, transducer and display device. With the UT testing method, the location of the crack, size of the flaw, orientation and other relevant features could be determined based on the signal information. Some of the high points of ultrasonic testing are its scanning speed, good flaw detection, good resolution and field deployment. However, it has disadvantages with respect to set up and skill requirements, i.e., ultrasonic testing is difficult to set up and requires a highly-skilled labour to ensure the accuracy of scanning. Assembly lines with the repetitive part design is an excellent application of ultrasonic testing. Generally, ultrasonic NDT has two approaches that are used in various fields. These are pulse-echo and through transmission approaches. In these approaches, the frequency of the sound waves used for detecting internal flaws in a material is quite high and lies between 1 and 50 MHz. Three modes which characterised ultrasonic testing are transmission, reflection and backscattering. All these modes depend on transducer range, coupling agents and frequencies. Defects present in homogeneous materials can be easily located using the pulseecho ultrasonic method. Here, waves scattering on flaws, wave energy loss transit time and energy loss as a result of attenuation are of utmost importance to the operator. This helps in locating material inconsistency, whether homogeneous or otherwise. Measurements of ultrasonic pulse velocity have been reported to be quite suitable for detecting, locating, imaging and quality control of large defects (Gholizadeh, 2016). The second approach, i.e., through transmission ultrasonic method ensures the transducer and receiver are placed at a fixed position away from the composite specimen and not on the sample surface. It is quite dissimilar from the common traditional ultrasonic techniques. Also, the transmission ultrasonic method has advantage of application to complex geometries where traditional transducer and receiver cannot make contact with part surface. Moreover, Fig. 5.3 depicts an experimental set-up. It can be observed that the ultrasonic wave passed through the defect-free Carbon fibre reinforced polymer (FRP) panel, but it rebounded by the internal defect present in the similar carbon FRP structure, as shown in Fig. 5.3(a) and (b), respectively. The UT results are compared with that of thermography testing subsequently.

5.2.3

Thermography testing

Thermography testing (TT) method is also called thermal imaging. Defects have the potential of changing the thermal properties of a material, such as thermal conductivity and amplitude, among others. When these defects move deeper into a part surface, they

Testing and damage characterisation of biocomposite materials

(a)

Water

Receiver

Emitter

183

(b)

CFRP panel

Tank

Figure 5.3 Experimental design of ultrasonic testing, showing the reaction of its wave to both (a) defect-free and (b) internal defect in the carbon FRP composite panels, respectively (Duan et al., 2019).

have a tendency of producing reduced or lower heat fluctuation when compared with material with a nearer defect on a part surface. Hence thermography examination is applicable for thin sample parts. A general rule is that thermography testing is not capable of picking defects with sample part depth greater than its diameter. When a part experiences impact damage or delamination, it causes a flaw which changes the thermal radiation of the impacted region of the material. This method has the advantage of inspecting large part surfaces. Also, it does not require a coupling, thus allowing for part inspection with only one side accessibility. Another point is that it does not have to couple distinguished thermography testing from many other types of inspections. However, the need for expensive and sensitive instruments, highly skilled inspection workers and lack of defect clarity in the part surface depth forms the demerits of this type of testing. A type of thermography testing, such as the Infrared thermography testing (IRT), records thermal radiation that is discharged from the surface of a specimen with the aid of an infrared camera, as illustrated in Fig. 5.4. The results of pulsed-TT are very similar to UT. For instance, Duan et al. (2019) compared the impact damage responses of 35 carbon FRP composite samples, with thickness and impact energy ranged from 2.3 to 4 mm and 10e36 J, respectively. The results obtained are shown in Fig. 5.5. As for the drawbacks of each of these non-destructive examinations, pulsed-TT and UT could not detect 45 and 0
oriented defects, respectively.

5.2.4

Radiography testing

Radiographic testing (RT) has been reported as the most widely applied inspection method. Delamination as the highest critical and commonly reported damage that FRP composite materials experience, often results in an air pocket damage. The delamination is clearly detected during RT only, provided the X-ray beam and its orientation is not perpendicular to each other. Many types of radiography exist with their various unique applications. When parts are not too thin or thick, then traditional RT is most appropriate. When parts are thin, in the range of 1e5 mm, the appropriate type of radiography is low voltage radiography. Gamma rays (g-rays) radiography applies to thick

184

Sustainable Composites for Lightweight Applications

Infrared sensor

Heating lamps

Infrared waves Infrared wave absorption Defect Periodic heating

Heat diffusion

N images Amplitude and phase images FFT

i -th pixel

Figure 5.4 Illustration of a lock-in thermography technique, depicting thermal amplitude and phase micrograph generation (de Oliveira et al., 2020).

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

Figure 5.5 Comparative examination results between (a-d) UT and (e-h) pulsed-TT images carbon FRP composite samples (Duan et al., 2019).

Testing and damage characterisation of biocomposite materials

185

b a

c

a

d p

a

P

d

Washer edge d = 6 mm [90
/45
/45
/0]s

Figure 5.6 (i) Damaged bolted joint sample, showing (ii) various X-ray radiographic images of drilling-induced damage at ultimate load: (a) transverse matrix cracks, (c) delamination, (b) 0
and (d) 45
axial splits, respectively (Atas and Soutis, 2013).

parts as they penetrate the composites and have shorter wavelengths. Another type of radiography used specifically for detecting delamination and minor FRP sample matrix cracks is the penetrant-enhanced radiography. These radiography types help to detect non-uniform fibre distribution, large voids, translaminar cracks, inclusions, as well as fibre disorientation. The defects/damage are associated with weld lines and fibre wrinkles. Additionally, there are several types of radiographic testing techniques, with particular applications of each. These methods include computed radiography, digital radiography, film radiography and computed tomography. The X-ray computed tomography (XCT) can be described as a non-destructive technique for examining the interior structures of solid materials/FRP composite samples and also to digitally obtain information about their three-dimensional (3D) features and properties. XCT has a comparative benefit over projection radiology, due to the 3D captured micrograph of the FRP composite samples that it provides, i.e., the projection radiology generates a 2-D image only. Hence the data from XCT is simple and quickly readable. Its results are also reliable because XCT alters the observation scale from macroscopic to microscopic scale. More discussion on XCT can be found in subsequent sub-chapters. For instance, local damage mechanisms and joint strengths of bolted joints in different oriented carbon FRP composite laminates were identified (Atas and Soutis, 2013), using penetrant enhanced X-ray radiography testing (Fig. 5.6). This was done in an attempt to reduce the drilling-induced delamination damage on drilled holes of the samples in question. Another comprehensive study has been reported, whereby three different X-ray imaging inspections (micro-computed tomography, computed and digital radiography) were concomitantly applied to examine the presence of voids in a glass FRP composite sample (Rique et al., 2015).

5.2.5

Electromagnetic testing

As the name suggests, the electromagnetic testing method employs electricity and magnetism in the detection and evaluation of faults, fractures, corrosion or other material conditions. The approach here is that both magnetic fields and electric currents (or either of them) are induced within a test FRP composite sample, while the electromagnetic signal is detected. Some identified electromagnetic (EM) techniques are Eddy current (EC) inspection, Microwave open-ended waveguide imaging,

186

Sustainable Composites for Lightweight Applications

(a)

Vertical linear axles

Manually adjustable linear axles Horizontal axle

Rotational axle Emitting coil

(b) Eddy current

Driver coil

Eddy current

Driver coil

In-plane waviness

Pickup coil

Pickup coil

Figure 5.7 Eddy current inspection: (a) measurement rig and (b) schematic induced EC in UD carbon FRP material without fibre waviness (left) and with in-plane waviness (right). (Berger, D., Lanza, G., 2017. Development and application of eddy current sensor arrays for process integrated inspection of carbon fibre preforms. Sensors 14 (4), 1e12.; Mizukami, K., Mizutani, Y., Todoroki, A., Suzuki, Y., 2016. Detection of in-plane and out-of-plane fibre waviness in unidirectional carbon fibre reinforced composites using eddy current testing. Composites Part B: Engineering 86, 84e94).

Couple spiral inductors, alternating current field measurement (ACFM), magnetic flux leakage (MFL) and remote field testing (RFT). They all have different physics as governed by their unique partial differential equations (PDEs). On moving forward, EC inspection depends on Faraday’s law, whereby the discontinuity in the conductivity distribution causes a change in the coil impedance, as illustrated in Fig. 5.7. The currents flow from one fibre to another through their contact points, along the fibres. This method is often used to detect damage in carbon FRP

Testing and damage characterisation of biocomposite materials

187

composites, provided defects are reflected in the change of the impedance recorded by its analyser. The EC inspection is commonly used for crack, impact damage, fibre breakage and corrosion detections, among other defects. For instance, Mook et al. (2001) employed special static, as well as rotary EC probes to identify some defects and characterised the shape of the damaged area after impact test. The local defects included impact damage, resin-rich zones, fibre breakage and delamination, as well as fibre orientation. Similarly, an EC theta probe was used to detect purposely induced low impact defects, produced at the energy of 0.25 J. Also, the correlation between the impact energy level and the signal phase was obtained (Koyama et al., 2011, 2013).

5.2.6

Acoustic emission inspection

Acoustic emission (AE) has proven to be an effective method for analysing imperfections. Material defects, such as fibre-matrix de-bonding, matrix microcracking, localised delamination or fibre pull-out and breakage, generate mechanical vibrations and produce stress waves at the origin, which concentrically disperse. Then, a group of piezoelectric detects the stress waves with high sensitivity. Moreover, two aspects uniquely distinguish the AE method from other widely used non-destructive testing methods. Firstly, the signal origin is considered, whereby the sound from the energy that is released in the FRP composite sample is captured in this method, as against the supply of energy to the object. The second distinguishing feature of the AE technique is its ability to address dynamism in a material, i.e., AE discerns between stagnant and developing defects significantly. Other merits of the acoustic emission method are its high sensitivity, assurance of process control with permanent sensor mounting without having to dismantle for specimen clean up and allowance of the use of multiple sensors which account for global and fast inspection. It is also worthy of note to mention that AE is useful in the detection of different types of damage due to fatigue loading, i.e., AE can detect fatigue damage types such as fibre fractures, fatigue cracks, fibre-matrix de-bonding, matrix microcracks and delamination. The disadvantage, however, of acoustic emission testing is that it requires a highly skilled inspector to map its data with the exact type of damage mechanism. Furthermore, an extensive literature review has been conducted by De Rosa et al. (2009) to report some applications of AE testing. It was used to monitor mechanical properties of natural FRP composites, such as interface investigations in single FRP composite tests, damage progression and detection of failure mechanisms, as well as crack propagation. For instance, AE has been used to understand the influence of fibre weaving on behaviours of three glass FRP composite materials (Mi et al., 2020). These weaved samples were a uniaxial, biaxial and triaxial direction, with fibre orientations/ ply sequences of [0
]8, [þ45
, 45
]4 and [0
, þ45
, 45
]4s, designated as UD, 2AX and 3AX, respectively. The signal amplitudes of the samples depicted the magnitude of their tensile strengths, especially after fast Fourier transformation (FFT), as shown in Fig. 5.8. AE also showed the damage patterns (matrix cracking, fibre pull-out, as well as fibre-matrix de-bonding), stress-strain curves and strain rates of the three samples. It was concluded that AE could effectively monitor the structural health of fibre-resin composite materials with complex fibre weavings.

188

Sustainable Composites for Lightweight Applications

(b)

(a)

2000

Amplitude (dB)

Amplitude (dB)

1 0.5 0

−0.5 −1 3AX

300 2AX UD

0

100

200

Time (s)

1500 1000 500 0 3AX

1000 800 600 400 200 Frequency (kHz)

2AX UD 0

Figure 5.8 The AE waveforms of the three glass FRP composite laminate samples and their FFT outcomes (Mi et al., 2020). 100

Matrix cracking Interface failure Fiber breakage

Partial power 1 (%)

80

60

40

20 0 0

100

200 300 400 500 600 Weighted peak frequency (kHz)

700

Figure 5.9 AE results of carbon fabric laminate depicting various failures under flexural testing (Ali et al., 2019).

Moving forward, Ali et al. (2019) employed AE to analyse microscopic matrix cracking, interface failure and fibre breakage in woven carbon FRP composite laminates, as shown in Fig. 5.9. AE root-mean-square was applied to analyse the drilling evolution and drilling-induced damage mechanism of 16-layers glass FRP composite laminates (Ravishankar and Murthy, 2000). Andrew et al. (2016) also characterised damage and residual strength of repaired glass FRP composite laminates with the AE method, among others reported studies on the application of the AE testing technique.

5.2.7

Acousto-ultrasonic testing

As its name implies, the acousto-ultrasonic testing (AUT) type of testing combines AE and UT types. It is precisely applied to evaluate the level of impact of internal failures, irregularities and heterogeneity in FRP composite structures. This testing method is quite useful in spotting and assessing non-critical flaws in composites. It also helps to indicate accumulated damage in a structure caused by impact or fatigue, as shown in Fig. 5.10. From Fig. 5.10, the current and baseline signals depicted a fatigue crack

Testing and damage characterisation of biocomposite materials

189

1.5

Baseline Current

Normalized amplitude

1.0 0.5 0.0 −0.5 −1.0 −1.5

0

30

60

90

120

150

Time (us)

Figure 5.10 Acousto-ultrasonic results, showing the presence of fatigue crack (current) and notched prior to the fatigue treatment (baseline) (Su et al., 2014).

and notched prior to the fatigue treatment, respectively. However, a low-point of this method is the compulsory pre-calculations and setup prior to testing. Another disadvantage of this method is that individual large flaws, such as voids or delamination, cannot be detected.

5.2.8

Shearography testing

This type of examination depends on the laser optical technique. Composite failure often occurs by stress concentrations. The strain concentration level that surrounds a specific defect determines the criticality of the defect. Shearography is often used on foam and honeycomb structures. It also has an advantage of less susceptibility to noise when compared with many other types of NDT techniques. This allows for the use of less-skilled workers without extensive training to inspect and determine whether a part is useable or not. However, one major drawback of shearography is the difficulty in characterising other defect types apart from delamination and dis-bonding. This makes pairing with other NDT techniques inevitable in order to identify certain defects in composites (Fig. 5.11). Therefore, there are laser digital shearography in addition to laser and digital types.

5.2.9

Computed tomography scanning

The computed tomography (CT) scanner is typically a large, box-like machine with a short tunnel in the middle. The narrow examination table slides in and out with the electronic X-ray and X-ray tube detectors in the gantry. A computer workstation is used for imaging information processes. A CT system contains an X-ray source, an X-ray detector, a rotary table and a data process unit for visualisation, computation

190

Sustainable Composites for Lightweight Applications

(a)

(b)

Shearography system

Image of disbonds

(c) Laser source

Loading frame

Display and processing

CCD camera

Polariser

Shearing device

Loading frame

Figure 5.11 (a) Testing for bonded insulation defect on the outer part of a rocket body, (b) dis-bond defect between the outer foam insulation and skin of the rocket, through (c) shearography system (Bossi and Giurgiutiu, 2014; Wang et al., 2020).

CT scanner

Detector

X-ray source

Rotary table

Figure 5.12 A CT scanning machine.

and data analysis of the results measured (Fig. 5.12). A CT system generates crosssection images through the projection of emitted photons beam and an object plane using angle positions that is angled while performing one revolution (Cantatore and Muller, 2011). When the emitted photons go through an object, some of them are scattered, absorbed with the rest transmitted. The attenuation represents the absorbed or

Testing and damage characterisation of biocomposite materials

(a)

(b) mm 3 2.5 2 1.5 1 0.5 0 −0.5 −1

191

μm 400 300 200 100 0 −100 −200 −300 −400 −500 −600 −700

Figure 5.13 CT scans of conventional and ultrasonically assisted drilled hemp FRP composite laminate samples (Wang et al., 2019).

scattered X-rays often result from interactions with an object (Hsieh, 2009). The attenuation prevents some X-ray from reaching the detector. The transmitted photons go through the object at an individual angle and are collected on the detector, which is then visualised by the computer. The visualisation creates a complete reconstruction of the object being scanned. Bartscher et al. (2007) stated that the 3D grey value data structure that is derived in this manner depicts the distribution density of the electron within the object measured. Unlike other visual scanning technologies that capture a point cloud, the CT scanning develops numerous X-ray images. They are combined to form a voxel data set. A voxel represents volume element or volumetric pixel, and it deploys X-ray attenuation effectively in distinguishing parts. For example, surface CT scans of conventional and ultrasonically assisted drilled hemp FRP composite laminate samples exhibit fibre push-out and burrs, as shown in Fig. 5.13 (Wang et al., 2019). These damages are not effectively observed, using a visual inspection and optical microscopy testing.

5.2.10 X-ray micro-computed tomography examination An example of model X-ray micro-computed tomography (mCT) scan XT H 225. XTH 225 (Fig. 5.14). It is ideal for a wide range of sample sizes and materials. It possesses three interchangeable sources: optional 225 kV rotating target, 180 kV transmission target and 225 kV reflection target. This model provides a microfocus X-ray source, high image resolution, a large inspection volume and readily available for ultrafast CT reconstruction. Also, its application area is wide, and it includes small castings, complex mechanisms and plastic parts together with natural specimens and several engineering materials. This instrument is advantageous because of its straightforward inspection automation, low cost-of-ownership, easy system operation, proprietary 225 kV microfocus X-ray source with a focal spot size of 3 mm and a high level of safety. It captures data very fast, together with images of high quality. From Fig. 5.14, part 1 shows the X-ray source (microfocus X-ray tube), part 2 indicates the sample in the rotating stage and part 3 depicts the phosphor detector.

192

(a)

(b) SDD

1

3

Fibre-optic taper

SOD

Micro-focus X-ray tube

1

2

2

Rotating stage

Controller

Phosphor CCD detector camera

Acquisition and display computer Trends in biotechnology

Figure 5.14 (a) XT H 225 X-ray mCT scanner, showing its set-up and (b) scanning/working mechanism.

Sustainable Composites for Lightweight Applications

3

Testing and damage characterisation of biocomposite materials

193

The microfocus X-ray source illuminates the object and a planar X-ray detector collects magnified projection images. Based on hundreds of angular views acquired while the object rotates, a computer synthesises a stack of virtual cross-section slices through the object. Then, scroll through the cross sections can be done, interpolating sections along different planes to inspect the internal structure and selecting simple or complex volumes of interest. Measurement of 3D morphometric parameters and the creation of realistic visual models for virtual travel within the object is achieved. This technology is of the order of >two to three micrometres. Moreover, an X-ray mCT scanner can be used to assess the barely visible impact damage failure, delamination of the layer of composite, matrix cracking, fibres breakage and fibres pull out the matrix, among other damage. It can assess FRP composite manufacturing defects. The X-ray mCT machine has asimilar working principle to CT scanner, but with better magnification. It is possible to observe an impacted FRP composite laminate sample with a scale/lamina of one failure to show the different mechanisms of the impact failure (Dhakal et al., 2018a,b). More also, Ismail et al. (2016) used X-ray CT micrographs to show peel-up and push-out inter-laminar drilling-induced delamination damage and fibre uncut on drilled holes of carbon and hemp FRP composite samples, as depicted in Fig. 5.15(a) and (b), respectively. Moreover, in recent years, this technique has been used to model the crack and damage progression in composite materials. In which, a high-resolution synchrotron X-ray CT is utilised to map the crack distribution, porosity formation and defects detection, among others, using 3D reconstruction for FE modelling of natural fibre and biodegradable matrix. For instance, Jiang et al. (2020) employed or established 3D finite element model to investigate the water diffusion response of jute/PLA composite based on the X-ray CT technique, as depicted in Fig. 5.16. Peel-up delamination

Feed direction Push-out delamination

(a)

Back

Front Uncut fibres

Uncut fibres

Small burrs

(b)

HFRP

Figure 5.15 X-ray CT micrographs, depicting (a) delamination damage and (b) fibre uncut and burrs on drilled holes of carbon and hemp FRP composite samples (Ismail et al., 2016).

194

Mould flow

direction

(a)

3D FE model

(b)

(e)

Z Threshold segmentation X

y Element model

Assembley, material parameters and boundary conditions assign

(d) (i)

(ii)

(ii)

Import Abaqus software PLA matrix

Jute fibres

PLA matrix

Jute fibres

Figure 5.16 3D model established for jute fibre and PLA matrix, depicting (a) 3D view, (b) 3D volume representation by introducing a threshold segmentation, (c) 3D rebuilding element model of jute fibre and PLA matrix after surface and mesh generation, (d) element model of jute fibre and PLA matrix using ABAQUS software and (e) Already built 3D finite element model with assembled jute fibre and PLA matrix, assigned material parameters and boundary conditions (Jiang et al., 2020).

Sustainable Composites for Lightweight Applications

(c) (i)

Surface and mesh generation

Testing and damage characterisation of biocomposite materials

195

5.2.11 Scanning electron microscopy A scanning electron microscope (SEM) focuses electron beam over a prepared FRP composite sample surface to create an image or micrograph. A comprehensive working principle of SEM is depicted in Fig. 5.17. The electrons in the beam interact with the sample to produce various signals that can be used to obtain information about the surface topography and composition. The trimmed and sputter-coated sample is attached to a mount with putty, as shown in Fig. 5.18(a). Then, it is loaded onto a platform and closed in the sample chamber (Fig. 5.18(b)) before the vacuum is turned on. When the right pressure is reached, the beam starts; first, at a low intensity to avoid bombardment damage to the sample and then switched to a higher one ready for imaging. Images of the sample can be taken at a low magnification to show an overview of the damage, followed by increasing levels of magnification at key points in order to better identify and view any specific failure mechanisms. Furthermore, SEM provides an excellent technique for examination of surface morphology of natural and treated biofibres. For example, the micrograph of untreated biofibres, as shown in Fig. 5.19(a), indicates the presence of impurities on the surface, mostly waxes and oil. Waxes and oils provide a protective layer to the surface of the

FE-gun

Condensor

EsB detector Filter grid

Beam booster

Inlens SE detector Gemini objective

Magnetic lens

Scan coils Electrostatic lens Sample

Figure 5.17 Detailed illustration of ZEISS EVO LS 10 scanning electron microscopy.

196

Sustainable Composites for Lightweight Applications

Figure 5.18 (a) Coated and mounted SEM sample and (b) SEM vacuum chamber loading platform.

(a)

0406 5.0KV

X500 10Pm WD35

(b)

0421 5.0KV

X500 10Pm WD35

Figure 5.19 SEM micrographs of hemp fibre (a) before and (b) after NaOH treatment.

fibres. However, a surface free of impurities is obtained after alkali treatment, as presented in Fig. 5.19(b). In addition, numerous studies have been conducted to show various applications of SEM. For instance, Dhakal et al. (2018c) were able to detect entanglement of some garnet abrasive particles within both peel-up and push-out types of drilling-induced inter-laminar delamination damage on drilled hole surfaces of hybrid carbon/flax FRP composite samples, as shown in Fig. 5.20. Similarly, various damage caused by drilling (Ismail et al., 2016; Niamat et al., 2019), wear, scratch (Parikh and Gohil, 2017; Akpan et al., 2018; Rajini et al., 2019; Chanda et al., 2019; Chegdani and Mansori, 2019; Belotti et al., 2019; Cheng et al., 2020), impact, flexural and tensile (Dhakal et al., 2019; Thiagamani et al., 2019; Wu et al., 2019; Nagaprasad et al., 2020) tests on different FRP composite samples have been observed and well analysed using SEM examination. These damages include, but are not limited to, fibre fracture and breakage, matrix melting and cracking, voids, crack, de-bonding, resin-rich region, filler accumulation, delamination, fibre pull-out and uncut.

Testing and damage characterisation of biocomposite materials

Mag =

102 X

EHT = 15.00 kV WD = 17.0 mm

Mag = 102 X

Signal A = SE1 I Probe =

50 pA

EHT = 15.00 kV Signal A = SE1 WD = 18.5 mm I Probe = 50 pA

Date : 28 Jul 2017

200 μm

Mag = 102 X

EHT =15.00 kV WD = 19.5 mm

Signal A =SE1 I Probe = 50 pA

Date :2 Aug 2017 Time :9:33:59

20 μm

Mag = 500 X

EHT = 15.00 kV WD = 19.0 mm

Signal A = SE1 I Probe = 50 pA

Date :26 Jul 2017 Time :9:58:53

Time :9:55:44

Date :26 Jul 2017 Time :10:03:16

197

Figure 5.20 SEM micrographs, showing entanglement of garnet abrasive particles within interlaminar delamination after drilling operation (Dhakal et al., 2018c).

5.3

Damage mechanisms and types (key factors for improving damage resistance)

The inherent properties of FRP composite materials make their failures or damage quite different from that of homogeneous materials, such as metals (Fig. 5.21). It has been reported that more than 50% of composite failure occurs within the first 20% of its life. This implies that composite structure can sustain cracks in its environment. Generally, damage in the form of crack initiation occurs in a metal after more than 75% of its fatigue life (Jollivet et al., 2013). Mechanisms of damage in metals are well understood more that of FRP composites. Defects and damage in composites are associated with their manufacturing process and in-service life stages, respectively. Damages of various composite structures are inevitable, either after a short- or longterm service. Thick-wall, large-scale composite, composite laminates, sandwich and smart structures are subjected to various kinds of loads/forces or stresses as structural, semi-structural and/or non-structural components. Composite constituents, types of loads and loading conditions determine the modes and mechanisms of damage. For instance, impact or flexural load, P induces combined (a) compressive, (b) shear and (c) tensile stresses, which they eventually orchestrate the final failure of the impacted composite sandwich structure (Fig. 5.22), after the decrease in mechanical properties

198

Sustainable Composites for Lightweight Applications

Damage

Delemination

Composite

Metal

Cracks & delamination Fiber failure

Damage in early cycles

Propagation & cracks

Nb cycle to failure

Nb cycles nb cycles for crack initiation

Figure 5.21 Comparison between fatigue damage progression in FRP composite and metallic materials (Jollivet et al., 2013). Impact location

Impact load, P

Surface deformation

Delamination

Skin (a) Compression (b) Shear

Core Skin

(c) Tension Support

Support

Matrix cracks due to bending

Fibre breakage

Matrix cracks due to shear

Figure 5.22 (a) Analysis of sandwich composite panel and (b) composite laminate (right) under impact loading (Greenhalgh, E.S., 2009. Defects and damage and their role in the failure of polymer composites. In: Greenhalgh, E.S. (Ed.). Failure analysis and fractography of polymer composites, Chapter 7, pp. 365e440. Woodhead Publishing: Cambridge, UK).

due to matrix cracking, fibre-matrix de-bonding and fibre breakage. These damages could similarly be caused by fatigue, lightning strikes and such alike. There are numerous damages associated with FRP composite materials. These include, but are not limited to, crack, de-bonding, delamination and waviness, through a particular evolution. Fig. 5.23 summarily depicts evolutions of some flaws/defects and damage associated with composite manufacturing and in-service/applications, respectively. The first aspect of composite manufacturing defects has been already and extensively discussed in the previous sub-chapters of this chapter. Therefore, the second part that deals with various damage or failures of FRP composite structures, their types, modes and mechanisms are hereby subsequently elucidated.

5.3.1

Damage types and mechanisms

Damage types and mechanisms depend on the types and locations of the failure. Some of the common types and mechanisms include interface damage or fibre-matrix de-bonding, as shown in Fig. 5.24(a). There is a matrix cracking or matrix damage.

Testing and damage characterisation of biocomposite materials

199

Scale dimension Nano-scale

Micro-scale

Meso-scale

Macro-scale

Manufacturing-induced flaws/defects

Matrix cracking

Nanoscale particles & imperfections

Fibre waviness

Voids, porosity, inclusions Fibre wrinkling

Fibre kinks/breakage

Misalignment of plies Misalignment of fibre orientation Laminate warping and buckling In-service damage evolution Yielding and failure of molecular, crystalline, or cross-linked networks

Matrix cracking

Delamination

Interface debonding Micro-buckling and waviness Fibre fracture and pull-out

Figure 5.23 Evolution of FRP composite manufacturing defects and in-service damage (Wang et al., 2020).

Stress

Stress

a a

b

C a

10 Pm

C

10 Pm

Figure 5.24 (a) fibre-matrix interfacial de-bonding and (b) matrix micro-cracks, caused by (c) locally strained area around the fibre (Arif et al., 2014).

In matrix cracking, stress, such as residual, could concentrate in an area to develop defects or flaws, which later promote crack propagation, as depicted in Fig. 5.24(b). Transversal de-bonding is followed by matrix failure, pull-out and longitudinal fibre failure. Others are compressive and shear damage. They can lead to delamination of the composite materials. Fibre fracture or breakage/fragmentation occurs when fibre ruptures through the through-thickness direction. Higher impact energy can cause fibre damage. Other types are described in Fig. 5.25. Detailed mechanisms of these and other damage initiations and propagations are elucidated later.

200

Sustainable Composites for Lightweight Applications

Type A

Type B Type C

Type F horizontal

Type D

Type E

Inspection surface

Opposite skin of inspection area

Figure 5.25 Various damage, showing Types A: Delamination, B and E: dis-bonding, C: crack, D: crush and F: fluid ingress in a composite sandwich structure (Olympus, 2020).

5.3.2

Failure or damage modes

Inter-laminar damage or delamination is one of the main failure modes in the impact damage. Also, trans-laminar damage occurs when the final damage stage tallies with the fibre failure. Matrix damage can occur in mode I as an intra- and inter-laminar damage (Fig. 5.26). Delamination means the separation of individual layers and often ensues from the weak interface between the fibre and matrix. It is an inter-laminar crack between resin-rich area and fibre plies. Delamination at the interface layers can cause final failure. Delamination damage is commonly reported when FRP composites are subjected to compression load. Another factor that could contribute to the susceptibility of composites to delamination is the brittle nature of the matrix resin, which binds the laminates. Delamination can occur due to buckling and compressive loadings, among other stresses. It occurs often due to weak fibre-matrix interfacial adhesion. Delamination modes include modes I, II and III, which are caused by opening or peeling, sliding shear and tearing shear, respectively (Fig. 5.27). In composite drilling, it could be either peel-up (entry) or push-out (exit) delamination (Fig. 5.28), when the drilling thrust force is higher than the threshold value.

(a)

(b)

20 μm

(c)

100 μm

10 μm

Figure 5.26 SEM micrographs of (a) inter-laminar, (b) intra-laminar and (c) trans-laminar damage in FRP composites (Jollivet et al., 2013).

Testing and damage characterisation of biocomposite materials

201

Figure 5.27 The basic fracture modes in delamination, depicting (a) mode I (Opening or peeling), (b) mode II (Sliding shear) and (c) mode III (tearing shear).

Mode I

Mode III

HSS drill bit Feed direction

a

Drill bit rotation

b

Peel-up Composite laminate

delamination Push-out delamination

Mode I

Mode II

Figure 5.28 Schematic illustration of (a) peel-up and (b) push-out types of delamination and their associated damage modes during drilling of FRP composite laminate (Ismail et al., 2016; Ojo et al., 2017).

5.3.3

Failure or damage mechanisms associated with FRP composites

The damage begins at a nano- or micro-scale level when it can be detected only by using some non-destructive techniques, as previously discussed. At this first stage, there is a change in either molecular, crystalline or cross-linked structures as a result of initial yielding and failure. Shortly after this phase, damage such as matrix cracking, interfacial de-bonding and buckling and waviness progressively occur as the composite material failures in-service from micro (mm) to meso (mm)-scale level, while carrying loads. Fibre fracture or cracking and pull-out are associated at the mesoscale stage. Without control, these failures continue and later develop into macro (m) scale leveled damage, such as delamination. It is necessary to note that the failure of the matrix precedes the fibre failure/crack. The damage progression is clearly illustrated in Fig. 5.29. Finally, the mechanical properties (strengths and stiffness), structural integrity, load-carrying capacity and service time of the composite component are decreased prior to the final rupture (Wang et al., 2020).

202

Sustainable Composites for Lightweight Applications

Matrix cracking Matrix Fi

be

r

Delamination Layer n–1 Matrix crack growth: Layer n Over thickness of ply Along fibre-matrix interface Layer n+1 matrix er

Failure

Damage index

fib

Stage II

Percent of life

Fi

Stage III

Stage I

Fiber cracking Matrix be

r

100 %

Figure 5.29 Damage mechanisms and evolution in unidirectional FRP composite structures (Romanowicz and Muc, 2018).

In other words, low-velocity impact modes of damage follow four distinctive stages: (a) matrix failure - where matrix crack occurs parallel to the fibres under combination effects of compression, shear and tensile (as previously discussed). (b) delamination e occurs by inter-laminar stresses. (c) fibre failure e orchestrates by the combination of tensile and compressive fibre buckling and (d) penetration e This is peculiar to high, ballistic and hypervelocity impact response types, where impactor or penetrator finally penetrates and completely perforates the FRP composite structures (Razali et al., 2014). Various damage modes (Fig. 5.30) can lead to perforation.

Due to the anisotropic and heterogeneous nature of FRP composite materials, they primarily and commonly exhibit brittle factures. Other uncommon fractures include ductile or in a combined form: ductile-brittle. For instance, Skalskyi et al. (2020) employed AE signals to estimate and rank various types and mechanisms of fracture in four different samples of Twaron 1000 and 1014 aramid fibre reinforced epoxy composites: untreated, silicone oil treated, plasma and 10% polystyrene-co-glycidyl methacrylate (PS-GMA) and 100% PS-GMA particle treated FRP composites, designated as samples 1, 2, 3 and 4, respectively. The optimum or best composite was sample 4, followed by samples 3 and 2, while sample 1 recorded the least performance. Furthermore, matrix cracking and shifting mechanisms in matrix recorded brittle and ductile types of fractures, while the interfacial de-bonding and fibre breakage of fracture mechanisms exhibited ductile-brittle, brittle and ductile, as well as ductile-brittle brittle types of fractures, as depicted in Fig. 5.31.

Testing and damage characterisation of biocomposite materials

Delamination

203

Impact location Surface buckling

Matrix cracks due to bending

Matrix cracks due to shear

Fiber breakage

Figure 5.30 Schematic illustration of various common impact damage modes of FRP composite laminate (Shyr, T-W., Pan, Y-H., 2003. Impact resistance and damage characteristics of composite laminates. Composite Structures 62, 193e203).

Type of fracture The number of events, %

80

Ductile (plastic deformation) Ductile-brittle (microcracking) Brittle (crack growth)

60

40

20

0

1

3

2

4 Type of sample

A, mV ΠT 0.16

1 0

0

0.08

60

−1

20

40

t,

μs

30 t, μs

A, mV

ΠT 0.08

0.6 0

0 400 0 800 f , kHz

0.04

30 20

−0.6 0

20

40

t, μs

t, μs

0

10 400 0 800 f , kHz

Figure 5.31 Various fracture types of aramid FRP composite samples and their recorded corresponding recorded number of AE events, as well as waveforms and continuous wavelet transform from AE signals of the tensile fractured samples 1 and 3. Adapted from Skalskyi, V., Stankevych, O., Zosel, T., Vynnytska, S., Thomas, H., Pich, A., 2020. Ranking of fibre composites by estimation of types and mechanisms of their fracture. Eng. Fract. Mech. 235, 1e13.

204

(a)

Sustainable Composites for Lightweight Applications

10:59:11

10:59:12

28.7°C

(b)

49.1°C

28 45 26

24

22 21.2°C

40 35 30 26.5°C

Figure 5.32 Crack pattern and temperature distribution of (a) plain and (b) notched samples at failure (Belmonte et al., 2017).

It was concluded that fracture resistance of FRP composites increased with their interfacial bond. Also, there was a decrease and an increase in the number of AE signals associated with brittle and ductile-brittle damage, showing macro-crack growth and microcracking fractures, respectively. The decrease was attributed to the fibre surface treatment. Moving forward, Belmonte et al. (2017) studied the damage mechanisms on a plain and notched short glass FRP composite samples subjected to fatigue loading at room temperature and humidity. Fractographic features, such as ductile-brittle matrix damage response, fibre failure or pull-out and degree of the glass-polyamide interfacial bond, were observed with the aid of infrared thermography, optical and electron microscopy. The crack pattern and temperature distribution of the two samples at failure are shown in Fig. 5.32. Further results showed that the plain sample failed due to unstable fatigue crack propagation (FCP), as illustrated in Fig. 5.33, while, crack initiation, stable and last unstable crack propagation, respectively, characterised the fracture steps or failure modes of the notched samples (Fig. 5.34). These failure modes determined the damage mechanisms of the samples, as matrix fracture responses were microductile and brittle on the fractured surfaces produced by stable and unstable FCP, respectively, as depicted in both Figs 5.33 and 5.34. Other damage mechanisms in FRP composite materials can be summarised in Fig. 5.35.

5.3.4

Damage detection in FRP composite structures

Several experimental methods of non-destructive examinations of damage in FRP composite structures have been previously explained in this chapter. Detection of the presence of damage and prediction of their locations are possible with many structural health monitoring techniques, which have been developed in the last few decades (Su et al., 2006; Diamanti and Soutis, 2010; Raghavan and Cesnik, 2007). The most famously used techniques are the guided wave base method. They are very popular because of their ability to detect small size damage, large detection zone, as well as their low attenuation. Lamb waves, the most famous among them, are applied to a localise damage, using piezoelectric as an actuator to transfer Lamb wave-imaging to the FRP composite structure, before the examination.

c

20 μm

(c)

b

Hochsp = 5.00 kV Arbeitsabstand = 10mm

Signal A = SE2 Vergroserung = 1.00 K X

Hochsp = 5.00 kV Arbeitsabstand = 11mm

Testing and damage characterisation of biocomposite materials

20 μm

(b)

(a)

Signal A = SE2 Vergroserung = 1.00 K X

Figure 5.33 (a) Tensile fractured plain sample and (b) its SEM micrographs, showing the morphology of the fractured surface at crack initiation and (c) caused by unstable crack propagation (Belmonte et al., 2017).

205

206

Sustainable Composites for Lightweight Applications

(a) d

60 μm

(b) c

b

EHT = 5.00 kV WD = 11.4 mm

u

u

S

S

60 μm

(c)

EHT = 5.00 KV WD = 11.3 mm

Signal A = SE2 Mag = 1.00 K X

60 μ m

(d)

Signal A = SE2 Mag = 1.00 K X

EHT = 5.00 KV WD = 11.0 mm

Signal A = SE2 Mag = 1.00 K X

Figure 5.34 (a) Tensile fractured notched sample, (b) its SEM micrographs, showing the morphology of the fractured surface at crack initiation, (c) the end of stable FCP and (d) caused by unstable FCP (Belmonte et al., 2017). Note: s and u represent stable and unstable FCP on the fractured surface. Damage mechanisms in composite materials

Fiber microdefects

Matrix microdefects

Fiber splitting (exfoilation) including organic fiber. Fiber surface crack especially in glass fibers. Fiber rupture (tension breakage). Fiber buckling in compression.

Volumetric pores especially in powdered-metal matrix. Transverse, axial and shear matrix microcracks. Idiabatic shear band (dynamic loading).

Matrix macrodefects Matrix transverse cracks in transverse layers. pseudo-matrix macrocracks.

Fiber-matrix microdefects Axial microcrack at ends ruptured fibers debond. Transverse micro crackings at ends of ruptured fibers, Shear microcracks directed at 45° from the applied loading’s direction.

Interlayer microdefects Interlaminar micro cracks delaminations. Intralaminar micro cracks which may also lead to loss of ply’s stability.

Figure 5.35 Summary of damage mechanisms in FRP composite structures (Lurie and Minhat, 2015).

Testing and damage characterisation of biocomposite materials

207

Moving forward, another structural health monitoring technique that is mostly suitable for complex FRP composite structures and has been extensively used is called wave-based imaging (Sohn et al., 2011; Rogge and Leckey, 2013; Radzienski et al., 2013; Kudela et al., 2015). Details of this method are available in their reported studies. Zhao et al. (2007) introduced a RAPID method. This method enables probabilistic inspection of damage in FRP composite materials through the reconstruction of the algorithm. A wind panel was tested and the presence and location of damage were obtained, implying the capability of the RAPID technique. Other structural health monitoring methods include, but are not limited to, delay-and-sum, cross-correlation and windowed energy arrival, as effectively used by Michaels and Michaels (2007), Veidt et al. (2008) and Sharif-Khodaei and Aliabadi (2014), respectively. Moreover, composites structures are often subjected to heat from fire, engine, electric wires/cables, exhaust wash and sun, due to their various areas of applications during service life. The effects of heat energy cause resin degradation and decrease in room-temperature mechanical behaviours. The affected FRP composites may appear undamaged to the visual examination, traditional, ultra-modern NDT methods and evaluation systems, below a particular exposure threshold, and without knowing that up to 60% of their initial strengths have been lost (Pereira et al., 2013; Razali et al., 2014). This phenomenon is referred to as incipient thermal damage (ITD). In a bid to solve the challenge of ITD in FRP composite components, many techniques have been adopted. These methods include, but are not limited to, the use of fluorescent thermal damage probes. They are embedded in the composite matrix during the manufacturing stage (Howie et al., 2012). The aviation industry also used the Fourier transform infrared spectroscopy technique to identify composite components that require repairs (Fu et al., 2014). Laser-induced fluorescence technique is another similar technique introduced for ITD detection (Fisher et al., 1997). Due to the necessary understanding of the failure throughout the whole thickness of FRP composite structure, the effect of composite thickness in terms of the decision on a part replacement and/or cost-effectiveness during repairs, Lindgren et al. (2006) employed sonicinfrared technique (also referred to as a sonic IR) to detect ITD through FRP composite structure thickness. The effectiveness and accuracy of the sonic IR technique depend on the standoff distance of the ultrasonic horn, which is used for the heating of the composite locally. An advanced NDT tool, known as thermography technique, has been introduced to evaluate ITD. This method includes the use of non-destructive heating equipment for sample excitation and infrared camera for collection of the full-time history of the surface heating. The thermography technique produces a similar sensitivity of ITD determination via the thickness of the composite material, excluding the requirement of a surface profile-following fixture (Razali et al., 2017).

5.3.5

Key factors for improving damage resistance

The service life of FRP composite materials depends on their damage resistance. Damages are undesirable responses in any material, but unfortunatelys they are inevitable. Several properties of composite reduce with time, under some certain conditions. However, these properties can be improved right from an effective design stage of

208

Sustainable Composites for Lightweight Applications

composite structure, because good damage resistance is a function of an improved property. This initial stage involves suitable selection of compactible reinforcements (fibres, particulates and fillers) and matrix, efficient process parameters, among others. These activities are very germane prior to the second stage of manufacturing. In addition, many damages in composites can be traced to poor manufacturing processes that produce defects, such as voids, porosity, the inclusion of foreign objects, to mention but a few. It also involves the use of suitable curing parameters (temperature and pressure) and techniques. In a simple word, the improvement in damage resistance of a composite material should involve a defect-free manufacturing process. For avoiding repetition, details on manufacturing defects associated with FRP composite materials can be obtained from the previous Chapter 4, under Section 4.5 of this book. Close to this factor is the use of efficient manufacturing methods: injection, compression, resin transfer moulding, automated fibre placement, filament winding, extrusion, additive manufacturing/3D printing, vacuum resin infusion, autoclave and out-of-autoclave moulding and effective techniques for enhancing properties of FRP composite structures. For instance, hybridisation technique improves some mechanical (tensile, impact and flexural) behaviours, z-pinning increases delamination resistance, fibre surface treatments (either chemical, physical, biological, among other methods) enhance the resistance to interfacial de-bonding and delamination by improving the fibrematrix interfacial adhesion, weaving, braiding, knitting and such alike can improve resistance to impact damage, to mention but a few. Section 4.3 of Chapter 4 extensively discussed all these methods/techniques with relevant and self-explanatory figures. Last, there is no multi-purpose FRP composite material; composite is designed and manufactured for a particular application. Wrong use of composite materials could reduce and destroy their properties, and hence, result in catastrophic failure and short service life. Therefore, all the composite structures must be used correctly and according to their stated conditions of optimum service in order to maintain their outstanding properties and performances.

5.4

5.4.1

Characterisation of damage modes using destructive and non-destructive damage analysis techniques (SEM, X-ray micro CT, AE, AFM, etc.) Categorisation of NDT methods for FRP composite materials

Composites literally suggest a material composition with more than one base material, such that the structure and properties of the individual base materials remain unchanged, i.e., they do not form an alloy. Sometimes, it may be challenging to choose an appropriate NDT method for a particular inspection of the composite structure for aerospace applications. But, this difficulty has been solved using standards (ASTM E2533, 2017) as a practical manual or guide. Testing situations and material applications are the determinants of the NDT method categorisation. The following are some of the categories.

Testing and damage characterisation of biocomposite materials

209

NDT

Through transm.ultrasonic

Visual

Shearography

Holography

Infrared

Thermography

Radiography

Liquid penetrant

methods

Penetrant

Non-Contact

Electromagnetic

Contact methods

Magnetic

Eddy current

Traditioal ultrasonic

techniques

Figure 5.36 Contact and non-contact NDT techniques.

5.4.2

Contact versus non-contact techniques

The fundamental NDT techniques are broadly divided into contact and non-contact techniques. Each of these methods has its definite applications in FRP composite material examination and evaluation. To obtain data or results from the contact method, reliability, good contact between the surface of the composite material being tested and the sensor is required. NDT testing methods that require good contact include UT, EC, penetrant, magnetic, as well as electromagnetic testing. The non-contact approach is performed without physical interaction or contact between the measuring sensor and the FRP composite material being inspected. Non-contact technique also helps to accelerate the process of collecting data. This category includes transmission ultrasonic, thermography, shearography, RT and VI methods. Optical methods, such as shearography and thermography, are mostly used methods within the non-contact category. Fig. 5.36 summarises various examples under the two categories or basic types of NDT techniques, as adapted (Gholizadeh, 2016).

5.4.3

Inspection type versus NDT methods

Many inspection types have been reported in the literature for the evaluation of composite materials. For instance, ultrasonic testing has been widely applied to identify the damage, assess defects and monitor health in various aerospace composite structures. Different researchers have also proposed other methods, which include thermographic inspection, vibration techniques, infrared thermography, X-ray computed tomography and shearography. Ultrasonic testing is the most applied method for monitoring the composite structure of aircraft wing-box. This method also has wide applications: identification of

210

Sustainable Composites for Lightweight Applications

Table 5.1 Inspection types and NDT methods (Gholizadeh, 2016). Inspection types

NDT methods

n

Identification of damage in composite parts of aircraft Assessment of composite components of aircraft Health monitoring of composite parts of aircraft.

Shearography Vibration techniques Ultrasonic inspection Infrared thermography Thermographic inspection XCT

n

Health monitoring of composite structure of aircraft wing-box

Ultrasonic examination

n

Structural health monitoring

Ultrasonic testing

n

Damage in GFRP

Thermographic testing Radiography

n

Auto-detection of impact damage in carbon FRP composite components Characterisation of damage in carbon FRP composite structures

Thermographic testing Radiography

n

Assessment of manufacturing defects and impact damage in glass FRP/epoxy composite components

Infrared thermography

n

Damage evaluation in composite sandwich components parameters that influence damping properties of composite structures The structure behaviour Dynamic behaviours for detection of damage in composite parts Statistical recognition/identification and restoration assessment of skin damage in composite components.

Vibration methods

n n

n

n n n n

Multiple cracks detection

Neutron radiography

damage, assessment of defects, as well as structural health monitoring (SHM) of composite components of aircraft. Table 5.1 shows different types of inspections and the methods used in each case.

5.4.4

Physical behaviours and structural integrity

There are several methods of testing composite structures. However, attention must be given to efficiency, safety and cost of the operation when determining best method to adopt. Composite failures are majorly caused as a result of material defects. These failures can manifest as fibre pull-out, fibre de-bonding, matrix cracking and fibre fracture. Structural integrity thus utilises advanced NDT approach to determine, detect and

Testing and damage characterisation of biocomposite materials

211

Table 5.2 Class of NDT techniques according to the detecting factors (Gholizadeh, 2016). Class

Use

Evaluation of both mechanical and physical behaviours, as well as detection of the material defects in various composite structures.

Measurement of: n n n n n n n n n n

To evaluate the structural integrity of the composite components, after manufacturing.

amount of fibre portion dynamic mechanical analysis mechanical behaviours: Stiffness and strength elastic constants delamination occurrence material content initiation of damage and succeeding damage progression construction connected with laminate resin cure condition fibre-matrix interfacial adhesion condition.

Detection of: n n n n

fibre breakage mechanical rubbing fibre pull-out micro-cracks and de-bonding

localise size of damage. Table 5.2 presents the NDT categories based on the factors that they evaluate. On moving forward, material mechanical properties are quite important as they determine whether or not the material can be manufactured. They also determine the performance and service life of the material. Therefore, the knowledge of the mechanical behaviours is of the essence for proper physical and mechanical characterisation of composite parts. Hence, this aspect of composite engineering has attracted a lot of studies.

5.5

Experimental and numerical modelling of damage modes and mechanisms

Detailed knowledge and understanding of FRP composite damage modes and mechanisms are very important in order to guide an effective composite design, development/fabrication and applications. Therefore, numerical formulation or modelling and simulation are carried out to mimic and evaluate the real damage response and evolution in composite structures, based on previously obtained relevant experimental results/data (Razali et al., 2017). Numerical analysis is relevant

212

Sustainable Composites for Lightweight Applications

to reduce the cost of materials and fabrication, production time and other challenges in the manufacturing of FRP composite materials. The use of the finite element method is very rampant in modelling and simulating material property degradation, various damage modes and mechanisms in FRP composites, using several simulation software, such as ANSYS, ABAQUS and LS-DYNA and Multiscale Designer, among others. Examples of the results obtained are given towards the end of this chapter. Some structural health monitoring or damage detecting and analysing methods that have been previously discussed, such as Lamb wave, wave filled imaging, RAPID, delay-and-sum, cross-correlation and windowed energy arrival, require a comparison between baseline and real-time data obtained from sensors. Experimental tests still remain the most dependable technique of obtaining the required baseline data. This is either impossible or can only be achieved with an enormous job because it requires testing a large FRP composite component in every single condition during its in-service life. Therefore, the finite element method is employed to obtain the required baseline data. However, drawbacks of the finite element method include expensive computation (Chakherlou and Yaghoobi, 2010) and its invalid equations in discontinuities, such as crack tips. Application of meshless methods, for example, peridynamics, is one of the ways out of these limitations. Therefore, Yaghoobi and Chorzepa (2015, 2017) used peridynamics and micropolar peridynamics (Chorzepa and Yaghoobi, 2016) to model fibre reinforcement in cementitious composites and solve the complex and unguided fracture response of FRP composite beams. In addition, the use of the spectral finite element method was another effective and alternative method of solving the limitations of the finite element method towards detecting damage in FRP composite structures. This method was first made prominent by Doyle (1997). Fourier-based spectral finite element method was employed and the results showed that it was very computationally efficient when compared with the finite element method. However, the main limitation of the spectral finite element method includes modelling of realistic composites components and complex features. Moreover, the wavelet spectral finite element was introduced with still some challenges. In an attempt to further improve on the wavelet spectral finite element method, wavelet spectral finite element-based user-defined element was formulated for one-dimensional composite beams to model complex structures (Khalili et al., 2014; Khalili et al., 2017). Wavelet spectral finite element-based user-defined element method was improved to effectively simulate delamination damage in composite beams (Khalili et al., 2015a). After much efforts, wavelet spectral finite element-based user-defined element method was further developed, and it was able to simulate two-dimensional composite laminate structures (Khalili et al., 2015b; Khalili et al., 2016), with computational efficiency and ability to model realistic structures and provide baseline data for the structural health monitoring benefits.

Testing and damage characterisation of biocomposite materials

213

Furthermore, FRP composite damage detection, among other heterogeneous and anisotropic complex materials, is very essential and hence a continuous possibility today, especially with machine learning methods coupled with the frequency response functions (Abbasi et al., 2015). Therefore, an Euler-Bernoulli model has been formulated to mimic the response of damaged FRP composite parts with possibility of inspecting different modes of delamination by embedding artificial immune-based method (Mohebbi et al., 2012; Mohebbi et al., 2013; Babaei et al., 2013, 2014).

5.5.1

Impact damage

Impact property of an FRP composite material is the ability of the material to resist the sudden release of load. It can be classified into four different categories, based on the velocity ranges: n

n n n

Low-velocity impact (LVI): Less than 20 or 40 m/s, such as dropping of hand tools on a material (usually less than 31 m/s), ship collision, vehicle impact and crash-worthiness, among other impact events. This often causes barely visible impact damage (BVID) response or effect on the FRP composite materials. High-velocity impact (HV): Between 40 and 240 m/s, such as bird striking and/or colliding with composite structures, such as aircraft. Ballistic velocity impact (BVI): Greater than 240 m/s, such as a projectile fired from a gun, free-falling bombs and missiles, among other military activities. Hypervelocity impact (HVI): Up to 15,000 m/s, such as orbit debris roving in outer-space, space vessels and meteoroid impact (Razali et al., 2017).

For instance, Petrucci et al. (2015) reported an experimental investigation on impact and post-impact damage behaviours of various hybrid FRP composite laminates. The impact responses showed that the glass-hemp-basalt (GHB) hybrid FRP composite laminate sample recorded a minimum performance, and the flax-hemp-basalt (FHB) sample slightly recorded better impact resistance than the glass-flax-basalt (GFB) combination. Importantly, the SEM micrographs obtained show their variable modes of fractured surfaces morphologically (Fig. 5.37). The possibility of delamination damage was observed with the FHB samples when AE was used to monitor the post-impact flexural tests. Conclusively, the FHB hybrids recorded better impact damage resistance that other hybrid samples. Similarly, Sarasini et al. (2016) experimental investigated into the low velocity falling weight impact and 4-point flexural damage tolerance of carbon/flax FRP hybrid composites, at different energy level ranged from 5 to 30 J. From the result obtained, it was evident that there was interfacial damage and cracks in the flaxbased laminates under impact and flexural loads, as depicted in Fig. 5.38(a) and (b), respectively. But, the presence of outer flax laminates (acting like skins) exhibited a greater impact damage resistance, preventing crack propagation within the laminates. Monitoring through digital image correlation supported preliminary

214

Sustainable Composites for Lightweight Applications

Figure 5.37 SEM micrographs of the FRP hybrid composites, showing brittle fractured hemp fibre surfaces of (a) GHB, (b) FHB and (c) glass-flax fibre layers of GFB laminates (Petrucci et al., 2015).

identification of associated failure modes and mechanisms of the hybrid FRP composite laminate structures. Summarily, the failure modes of impact loading include matrix cracking and breakage, fibre cracking and breakage, fibre pull-out and delamination, among others (Razali et al., 2017). The aforementioned modes of damage associated with FRP composite structures depend on materials (fibre/reinforcement and matrix) properties, target geometry, impact velocity, impactor/projectile nose shape and relative mass, support conditions, as well as the target geometry, among other factors.

5.5.2

Fatigue life model

Various approaches have been employed to model fatigue damage in FRP composite materials. Residual strength and stiffness degradation or commonly called damage

Testing and damage characterisation of biocomposite materials

215

Figure 5.38 (a) Interfacial damage and (b) crack paths in flax-based FRP composite laminates (Sarasini et al., 2016).

mechanism theories, are the most used fatigue life models (Zhang et al., 2015). These models accepted that the fatigue damage in composites is caused as a result of changes in their material properties (Romanowicz and Muc, 2018). Various loading parameters are considered for fatigue residual strength or stiffness models recently available and reported (D’Amore et al., 2016; D’Amore and Grassia, 2017; Mejlej et al., 2017; Stojkovic et al., 2017). One of the popular and effective model is Epaarachchi and Clausen model (Epaarachchi and Clausen, 2003), as stated in Eqs. (5.1)e(5.4) (Romanowicz and Muc, 2018).  
  s su 0:64 jsinðqÞj 1 u a Nfb  1 ¼ 1 fb 0:64 jsinðqÞj smax smax ð1  4Þ

(5.1)

216

Sustainable Composites for Lightweight Applications

Where,

4¼

8 > < R

9 for  N < R < 1  tension  tension and tension  compression > = 1=R for 1 < R < þN  compression  compression > ;

> :

(5.2) And R¼

smin smax

(5.3)

where Nf and su represent the numbers of cycles to failure and material ultimate stress in the loading direction, respectively. And, q, f and R are the smallest angles between the loading direction and fibre direction, loading frequency and stress ratio. While a and b denote the material constants obtained from experimental fatigue tests. smin and smax stand for the minimal and maximal applied fatigue stress in the loading direction, respectively. From Eq. (5.1) the number of cycles to failure is expressed as Eq. (5.4). "

1 Nf ¼ 1 þ f

5.5.3



su 1 smax



su smax

0:64 jsinðqÞj

1

ð1  4Þ

f 0:64 jsinðqÞj

#1 b

b

(5.4)

Thermal effects

Yang et al. (2015) used the finite element method to study the effects of temperature on low-velocity impact damage in woven carbon/epoxy composite sandwich structures with closed-cell polymeric foam core. The composite panels were subjected to energy levels of 10 and 50 J, under cold, room and high or elevated temperature dry of 45.6
C (50
F), 21
C (70
F) and 82.2
C (180
F), simply designated as CTD, RTD and ETD, respectively. The models depicted the face sheets, inter-laminar/delamination damage progression and temperature effect, with the aid of ultrasonic inspection and high-speed cameras. From the results obtained, it was observed that the larger damage areas for both impact energy levels were caused by the larger temperature exposure (Fig. 5.39a), resulting in visible damage with high penetration rate and indent or impact depth (Fig. 5.39(b)). However, the low energy level of 10 J produced a barely visible impact damage with no fibre fracture. Also, the simulation results for both energy levels are shown in Fig. 5.40(a) and (b).

Testing and damage characterisation of biocomposite materials

(a)

217

6 35 5

Damage area, (in2)

4

25

20

3

15 2

Damage area, (cm2)

30

10 1 5

0 ETD-50J

(b)

ETD-10J

RTD-50J

RTD-10J

CTD-50J

CTD-10J

0

1 24 22 0.8

20

16

0.6

14 12 0.4

10

Indent depth, (mm)

Indent depth, (in)

18

8 6 0.2 4 2 0

0 ETD-50J

ETD-10J

RTD-50J

RTD-10J

CTD-50J

CTD-10J

Figure 5.39 Comparison of effects of temperatures and energy levels on (a) impact damage zone and (b) indent depth, showing visible and barely visible damage responses at 50 and 10 J, respectively (Yang et al., 2015).

218

(a)

Sustainable Composites for Lightweight Applications

Shear damage view

Core damage Damage

Core crushing in limited zone under impactor

Cohesive layer damage

Delamination in the top facesheet

Figure 5.40 Low-velocity impact damage simulation results, exhibiting (a) barely visible and (b) visible impact failure, at 10 and 50 J, respectively (Yang et al., 2015).

In like manner, Buenrostro and Whisler (2018) employed both experimental and finite element analyse to investigate the impact, flexural and compressive behaviours of polyurethane foam (PUF) and glass fibre reinforced polyurethane form (GFRPUF). From the results obtained (Fig. 5.41), it was observed that GFRPUF composite specimens were more brittle than the PUF samples under flexural bending.

Testing and damage characterisation of biocomposite materials

(b)

219

Shear damage view

Continuum

Cohesive layer damage

damage in facesheets Fiber breakage at the top and bottom facesheets

Core damage

Delamination

Damage

Foam is completely crushed from red to green zone

Figure 5.40 cont'd.

Also, GFRPUF samples exhibited a better impact resistance property than PUF. This can be traced to the reinforcing effect of glass fibres on PUF material and better effective distribution of impact energy or force than the un-reinforced PUF samples. However, GFRPUF samples recorded a lower quasi-static compressive damage resistance when compared with the PUF counterparts, due to the manufacturing process and observed large voids in the GFRPUF composite samples. Both experimental and numerical simulation (finite element) results are depicted Fig. 5.41(a) and (b), respectively, and in a combined form to show their dynamic compressive damage responses or results (Fig. 5.41(c)).

220

Sustainable Composites for Lightweight Applications

PUF, front

GFRPUF, front

PUF, back

GFRPUF, back

(a)

(b)

S,Mises SNEG, (fraction = -1.0) (Avg: 75%) +8.0e+07 +7.3e+07 +6.7e+07 +6.0e+07 +5.3e+07 +4.7e+07 +4.0e+07 +3.3e+07 +2.7e+07 +2.0e+07 +1.3e+07 +6.7e+06 +0.0e+00

(a)

S,Mises SNEG, (fraction = -1.0) (Avg: 75%) +8.0e+07 +8.0e+07 +7.3e+07 +6.7e+07 +6.0e+07 +5.3e+07 +4.7e+07 +4.0e+07 +3.3e+07 +2.7e+07 +2.0e+07 +1.3e+07 +6.7e+06 +0.0e+00

PUF

GFRPUF

back

back

(b)

Figure 5.41 Impact damage responses of PUF and GFRPUF composite sandwich samples, showing their (a) experimental and (b) finite element results at a velocity of 220 m/s (1.2 kJ), and (c) both results of dynamic compression of GFRPUF and PUF cores. Adapted from Buenrostro, E., Whisler, D., 2018. Impact response of a low-cost randomly oriented fibre foam core sandwich panel. J. Compos. Mater. 52 (25), 3429e3444.

Testing and damage characterisation of biocomposite materials

221

4.5 PUF actual PUF FEA 3% fiber actual 3% fiber FEA

4 3.5

Force (kN)

3 2.5 2 1.5 1 0.5 0 0

0.005

0.01

0.015 Time (s)

0.02

0.025

(c)

Figure 5.41 cont'd.

5.6

Conclusions

Testing and damage characterisation of FRP bio/composite materials are more crucial and required in many manufacturing industries today. This is because FRP bio/composite structures are widely utilised in critical-safety applications, such as automobile, aircraft and marine primary constructions, where efficiency, cost and safety are important determinant factors. Therefore, several testing and damage characterisation techniques that are relevant to FRP bio/composite materials have been comprehensively elucidated in this chapter. Various destructive and non-destructive testing techniques, either contact or non-contact methods, can be used to detect and analyse manufacturing defects and test damage in FRP bio/composite structures internally (microscopically) and/or externally. Each of these methods has benefits and drawbacks. It is evident that two or more methods can be used to reduce their limitations, inspect and analyse a single damage, such as inter-laminar delamination, crack initiation and propagation, de-bonding, fibre fracture or kinking, to mention but a few. Therefore, the choice of their application depends on the nature of defect and damage, their penetrating or scanning capability, properties of the FRP composite materials and extent of analysis required, among others. More also, there is a possibility of an increase in the future concerning the number of examination techniques, efficiency, accuracy, fast analysis/processing of data and decision making with rapid advancement in science, technology and engineering (metrology): use of highly intelligent and automated/robotic inspection systems or diagnostic systems, machine and deep learning, artificial intelligence and robotics,

222

Sustainable Composites for Lightweight Applications

as well as artificial neural network coding/algorithms. The future is already here! The trend has started as a combination of two or more traditional or conventional techniques is fast increasing. These include, but are not limited to, acousto-ultrasonic testing (a combination of acoustic emission and ultrasonic inspections), digital X-ray radiography and infrared thermography inspection, to mention but a few.

References Abbasi, F., Mojtahedi, A., Ettefagh, M.M., 2015. Fault diagnosis using noise modelling and a new artificial immune system based algorithm. Earthq. Eng. Eng. Vib. 14, 725e741. Akpan, E.I., Wetzel, B., Friedrich, K., 2018. A fully biobased tribology material based on acrylic resin and short wood fibres. Tribol. Int. 120, 381e390. Ali, H.Q., Tabrizi, I.E., Khan, R.M.A., Tufani, A., Yildiz, M., 2019. Microscopic analysis of failure in woven carbon fabric laminates coupled with digital image correlation and acoustic emission. Compos. Struct. 230, 111515. Almansour, F.A., Dhakal, H.N., Zhang, Z.Y., 2017. Effect of water absorption on Mode I interlaminar fracture toughness of flax/basalt reinforced vinyl ester hybrid composites. Compos. Struct. 168, 813e825. Andrew, I.J., Arumugam, V., Bull, D.J., Dhakal, H.N., 2016. Residual strength and damage characterisation of repaired glass/epoxy composite laminates using A.E. and D.I.C. Compos. Struct. 152, 124e139. Arif, M.F., Meraghni, F., Chemisky, Y., Despringre, N., Robert, G., 2014. In situ damage mechanisms investigation of PA66/GF30 composite: effect of relative humidity. Compos. B Eng. 58, 487e495. ASTM E2533, 2017. Standard Guide for Nondestructive Testing of Polymer Matrix Composites Used in Aerospace Applications. Atas¸, A., Soutis, C., 2013. Subcritical damage mechanisms of bolted joints in CFRP composite laminates. Compos. B Eng. 54, 20e27. Babaei, M., Babaei, M., Niasati, M., 2013. Parametric analysis of over voltages caused by backflashover in “Siah-Bishe” 400 kV GIS substation. In: 2013 3rd International Conference on Electric Power and Energy Conversion Systems, Istanbul, 2-4 October 2013, pp. 1e6. Babaei, M., Babaei, M., Nourirad, G., 2014. Analysis of influential factors in determining very fast transient over voltages of GIS substations. In: 2014 IEEE 8th International Power Engineering and Optimisation Conference, Langkawi, 24-25 March 2014, pp. 79e84. Bartscher, M., Hilpert, U., Goebbels, J., Weidemann, G., 2007. Enhancement and proof of accuracy of industrial computed tomography (CT) measurements. CIRP Ann. Manuf. Technol. 56 (1), 495e498. Belmonte, E., De Monte, M., Hoffmann, C.J., Quaresimin, M., 2017. Damage mechanisms in a short glass fibre reinforced polyamide under fatigue loading. Int. J. Fatig. 94, 145e157. Belotti, L.P., Vadivel, H.S., Emami, N., 2019. Tribological performance of hygrothermally aged UHMWPE hybrid composites. Tribol. Int. 138, 150e156. Bossi, R.H., Giurgiutiu, V., 2014. Nondestructive testing of damage in aerospace composites. In: Irving, P.E., Soutis, C. (Eds.), Polymer Composites in the Aerospace Industry. Woodhead Publishing, Cambridge, UK, pp. 413e448. Buenrostro, E., Whisler, D., 2018. Impact response of a low-cost randomly oriented fibre foam core sandwich panel. J. Compos. Mater. 52 (25), 3429e3444.

Testing and damage characterisation of biocomposite materials

223

Cantatore, A., Muller, P., 2011. Introduction to Computed Tomography. Manufacturing Engineering. Department of Mechanical Engineering, Technical University of Denmark, Denmark, pp. 1e77. Chakherlou, T.N., Yaghoobi, A., 2010. Numerical simulation of residual stress relaxation around a cold-expanded fastener hole under longitudinal cyclic loading using different kinematic hardening models. Fatig. Fract. Eng. Mater. Struct. 33 (11), 740e751. Chanda, A., Sinha, S.K., Datla, N.V., 2019. Tribological studies of epoxy-carbon nanofibre compositeseEffect of nanofibre alignment using AC electric field. Tribol. Int. 138, 450e462. Chegdani, F., Mansori, M.E., 2019. Tribo-functional effects of double-crossed helix on surface finish, cutting friction and tool wear mechanisms during the milling process of natural fibre composites. Wear 426e427, 1507e1514. Cheng, Q., Jiang, H., Li, Y., 2020. Effect of fibre content and orientation on the scratch behaviour of short glass fibre reinforced PBT composites. Tribol. Int. 146, 1e10. Chorzepa, M.G., Yaghoobi, A., 2016. Innovative meshless computational method for the analysis of fibre-reinforced concrete (FRC) structures. Geotech. Struct. Eng. Congr. 1209e1221. Phoenix, 14-17 February 2016. D’Amore, A., Grassia, L., 2017. Phenomenological approach to the study of hierarchical damage mechanisms in composite materials subjected to fatigue loadings. Compos. Struct. 175, 1e6. D’Amore, A., Grassia, L., Ceparano, A., 2016. Correlations between damage accumulation and strength degradation of fibre reinforced composites subjected to cyclic loading. Procedia Eng. 167, 97e102. de Oliveira, B.C.F., Nienheysen, P., Baldo, C.R., Goncalves Jr., A.A., Schmitt, R.H., 2020. Improved impact damage characterisation in CFRP samples using the fusion of optical lock-in thermography and optical square-pulse shearography images. NDT E Int. 111, 1e15. De Rosa, I.M., Santulli, C., Sarasini, F., 2009. Acoustic emission for monitoring the mechanical behaviour of natural fibre composites: a literature review. Compos. A Appl. Sci. Manuf. 40, 1456e1469. De Rose, I.M., Dhakal, H.N., Santulli, C., Sarasini, F., Zhang, Z.Y., 2012. Post-impact static and cyclic flexural characterisation of hemp fibre reinforced laminates. Compos. B Eng. 43, 1382e1396. Dhakal, H.N., Arumugam, V., Aswinraj, A., Santulli, C., Zhang, Z.Y., 2014a. Influence of temperature and impact velocity on the impact response of jute/UP composites. Polym. Test. 35, 10e19. Dhakal, H.N., Ghasemnejad, H., Zhang, Z.Y., Ismail, S.O., Arumugam, V., 2019. The postimpact response of flax/UP composite laminates under low velocity impact loading. Int. J. Damage Mech. 28 (2), 183e199. Dhakal, H.N., Ismail, S.O., Jiang, C., Zhang, Z., Sweatman, T., 2018b. Influence of barely visible impact damage on post-impact residual flexural properties of hybrid Fibri Rock e Aero Eco-composites. Mater. Lett. 233, 233e237. Dhakal, H.N., Ismail, S.O., Ojo, S.O., Paggi, M., Smith, J.R., 2018c. Abrasive water jet drilling of advanced sustainable bio-fibre-reinforced polymer/hybrid composites: a comprehensive analysis of machining-induced damage responses. Int. J. Adv. Manuf. Technol. 99, 2833e2847. Dhakal, H.N., Ismail, S.O., Zhang, Z., Barber, A., Welsh, E., 2018a. Development of sustainable biodegradable lignocellulosic hemp fibre/polycaprolactone biocomposites for light weight applications. Compos. A Appl. Sci. Manuf. 113, 350e358.

224

Sustainable Composites for Lightweight Applications

Dhakal, H.N., Skrifvars, M., Adekunle, K., Zhang, Z.Y., 2014b. Falling weight impact response of jute/methacrylated soybean oil bio-composites under low velocity impact loading. Compos. Sci. Technol. 92, 134e141. Diamanti, K., Soutis, C., 2010. Structural health monitoring techniques for aircraft composite structures. Prog. Aerosp. Sci. 46, 342e352. Dicker, M.P.M., Duckworth, P.F., Baker, A.B., Francois, G., Hazzard, M.K., Weaver, P.M., 2014. Green composites: a review of material attributes and complementary applications. Compos. Appl. Sci. Manuf. 56, 280e289. Doyle, J.F., 1997. Wave Propagation in Structures: Spectral Analysis Using Fast Discrete Fourier Transforms. Springer, New York. Duan, Y., Zhang, H., Maldague, X.P.V., Ibarra-Castanedo, C., Servais, P., Genest, M., Sfarra, S., Meng, J., 2019. Reliability assessment of pulsed thermography and ultrasonic testing for impact damage of CFRP panels. NDT E Int. 102, 77e83. Epaarachchi, J.A., Clausen, P.D., 2003. An empirical model for fatigue behaviour prediction of glass fibre-reinforced plastic composites for various stress ratios and test frequencies. Compos. A Appl. Sci. Manuf. 34, 313e326. Faruk, O., Bledzki, A.K., Fink, H.P., Sain, M., 2012. Biocomposites reinforced with natural fibres: 2000-2010. Prog. Polym. Sci. 37 (11), 1552e1596. Fisher, W.G., Meyer, K.E., Wachter, E.A., Perl, D.R., 1997. Laser induced fluorescence imaging of thermal damage in polymer matrix composites. Mater. Eval. 55 (6), 726e729. Fu, T., Liu, Y., Lau, K., Leng, J., 2014. Impact source identification in a carbon fibre reinforced polymer plate by using embedded fibre optic acoustic emission sensors. Compos. B Eng. 66, 420e429. Gholizadeh, S., 2016. A review of non-destructive testing methods of composite materials. Procedia Struct. Integr. 1, 50e57. Howie, T., Shi, Z., Jen, A., Georgeso, G., Flinn, B., 2012. In: Detection of thermal damage of CFRP composites using fluorescent thermal damage probe. Proceedings of the 44th International SAMPLE Tech 2012 Conference and Exhibition: Navigating the Global Landscape for the New Composites, Charleston, South Carolina. Society for the Advancement of Material and Process Engineering. Hsieh, J., 2009. Computed Tomography: Principles, Design, Artifacts and Recent Advances. SPIE Press, Bellingham, WA, USA. Ismail, S.O., Dhakal, H.N., Popov, I., Beaugrand, J., 2016. Comprehensive study on machinability of sustainable and conventional fibre reinforced polymer composites. Eng. Sci. Technol. Int. J. 19, 2043e2052. Jiang, N., Li, Y., Li, D., Yu, T., Li, Y., Xu, J., et al., 2020. 3D finite element modelling of water diffusion behaviour of jute/PLA composite based on X-ray computed tomography. Compos. Sci. Technol. 199, 1e10. Jollivet, T., Peyrac, C., Lefebvre, F., 2013. Damage of composite materials. Procedia Eng. 66, 746e758. Khalili, A., Jha, R., Samaratunga, D., 2016. Spectrally formulated user-defined element in conventional finite element environment for wave motion analysis in 2D composite structures. Eur. J. Comput. Mech. 25, 446e474. Khalili, A., Jha, R., Samaratunga, D., 2017. The wavelet spectral finite element-based userdefined element in ABAQUS for wave propagation in one-dimensional composite structures. Simul. Trans. Soc. Model. Simul. Int. 93 (5), 397e408.

Testing and damage characterisation of biocomposite materials

225

Khalili, A., Samaratunga, D., Jha, R., Gopalakrishnan, S., 2014. Wavelet spectral finite element modelling of laminated composite beams with complex features. In: Proceedings of the American Society of Composites 2014-Twenty-Ninth Technical Conference on Composite Materials, San Diego, 8-10 September 2014. Khalili, A., Samaratunga, D., Jha, R., Lacy, T.E., Gopalakrishnan, S., 2015a. Wavelet spectral finite element based user-defined element in ABAQUS for modelling delamination in composite beams. In: 23rd AIAA/AHS Adaptive Structures Conference, Kissimmee, 5-9 January 2015, pp. 1e11. Khalili, A., Samaratunga, D., Jha, R., Lacy, T.E., Gopalakrishnan, S., 2015b. WSFE-based user-defined elements in ABAQUS for modelling 2D laminated composites with complex features. In: 30th ASC Technical Conference, East Lansing, 28-30 September 2015. Koyama, K., Hoshikawa, H., Hirano, T., 2011. Investigation of impact damage of carbon fibre reinforced plastic (CFRP) by eddy current non-destructive testing. In: Smart Materials, Structures and NDT in Aerospace: International Workshop, 2-4 November 2011, Montreal, Canada, pp. 582e594. Koyama, K., Hoshikawa, H., Kojima, G., 2013. Eddy current non-destructive testing for carbon fibre reinforced composites. J. Pressure Vessel Technol. 135 (4), 1e5. Kudela, P., Radzienski, M., Ostachowicz, W., 2015. Identification of Cracks in thin-walled structures by means of wavenumber filtering. Mech. Syst. Signal Process. 50e51, 456e466. Li, Z., Meng, Z., 2016. A review of the radio frequency non-destructive testing for carbon-fibre composites. Meas. Sci. Rev. 16 (2), 68e76. Lindgren, E., Welter, J., Sathish, S., Ripberger, E., 2006. In: Detection of incipient thermal damage in polymer matric composites. Air Force Research Laboratory, United States Air Force. Lurie, S., Minhat, M., 2015. Modelling of damage evaluation and failure of laminated composite materials across length scales. In: Beaumont, P.W.R., Soutis, C., Hodzic, A. (Eds.), Structural Integrity and Durability of Advanced Composites: Innovative Modelling Methods and Intelligent Design, Chapter 7. Elsevier Woodhead Publishing, Cambridge, UK, pp. 143e196. Mejlej, V.G., Osorio, D., Vietor, T., 2017. An improved fatigue failure model for multidirectional fibre-reinforced composite laminates under any stress ratios of cyclic loading. Procedia CIRP 66, 27e32. Mi, Y., Zhu, C., Li, X., Wu, D., 2020. Acoustic emission study of effect of fibre weaving on properties of fibre-resin composite materials. Compos. Struct. 237, 1e10. Michaels, J.E., Michaels, T.E., 2007. Guided wave signal processing and image fusion for in situ damage localization in plates. Wave Motion 44, 482e492. Mohebbi, B., Abbasidoust, F., Ettefagh, M.M., 2012. Delamination detection in CFRP composite beam using modified AIS algorithm. In: 2012 International Symposium on Innovations in Intelligent Systems and Applications, Trabzon, 2-4 July 2012, 1-5. Mohebbi, B., Abbasidoust, F., Ettefagh, M.M., Biglari, H., 2013. Delamination Diagnosis in Composite Beam Using Ais and Bga Algorithms Based on Vibration Characteristics. Springer International Publishing, New York, pp. 73e89. Mook, G., Lange, R., Koeser, O., 2001. Non-destructive characterisation of carbon-fibre-reinforced plastics by means of eddy-currents. Compos. Sci. Technol. 61 (6), 865e873.

226

Sustainable Composites for Lightweight Applications

Nagaprasad, N., Stalin, B., Vignesh, V., Ravichandran, M., Rajini, N., Ismail, S.O., 2020. Effect of cellulosic filler loading on mechanical and thermal properties of date palm seed/vinyl ester composites. Int. J. Biol. Macromol. 147, 53e66. Niamat, M., Sarfraz, S., Shehab, E., Ismail, S.O., Khalid, Q.S., 2019. Experimental characterisation of electrical discharge machining of aluminium 6061 T6 alloy using different dielectrics. Arabian J. Sci. Eng. 9, 1e10. Ojo, S.O., Ismail, S.O., Paggi, M., Dhakal, H.N., 2017. A new analytical critical thrust force model for delamination analysis of laminated composites during drilling operation. Compos. B Eng. 124, 207e217. Olympus, 2020. Non-destructive Bond Testing for Aircraft Composites. Available at: https:// www.olympus-ims.com/en/applications/non-destructive-bond-testing-aircraft-composites/. (Accessed 30 July 2020). Parikh, H.H., Gohil, P.P., 2017. Experimental investigation and prediction of wear behaviour of cotton fibre polyester composites. Friction 5 (2), 183e193. Pereira, G., Frias, C., Faria, H., Frazao, O., Marques, A.T., 2013. Study of strain-transfer of FBG sensors embedded in unidirectional composites. Polym. Test. 32 (6), 1006e1010. Petrucci, R., Santulli, C., Puglia, D., Nisini, E., Sarasini, F., Tirillo, J., et al., 2015. Impact and post-impact characterisation of hybrid composite laminates based on basalt fibres in combination with flax, hemp and glass fibres manufactured by vacuum infusion. Compos. B Eng. 69, 507e515. Radzienski, M., Dolinski, Ł., Krawczuk, M., Palacz, M., 2013. Damage localisation in a stiffened plate structure using a propagating wave. Mech. Syst. Signal Process. 39, 388e395. Raghavan, A., Cesnik, C.E.S., 2007. Review of guided-wave structural health monitoring. Shock Vib. Digest 39, 91e114. Rajini, N., Mayandi, K., Prabhakar, M.M., Siengchin, S., Ayrilmis, N., Bennet, C., et al., 2019. Tribological properties of Cyperus pangorei fibre-reinforced polyester composites (friction and wear behaviour of cyperus pangorei fibre/polyester Composites). J. Nat. Fibres 7 (4), 1e13. Ravishankar, S.R., Murthy, C.R.L., 2000. Application of acoustic emission in drilling of composite laminates. NDT E Int. 33, 429e435. Razali, N., Sultan, M.T.H., Jawaid, M., 2017. A review on detecting and characterising damage mechanisms of synthetic and natural fibre based composites. BioRes. 12 (4), 9502e9519. Razali, N., Sultan, M.T.H., Mustapha, F., Yidris, N., Ishak, M.R., 2014. Impact damage on composite structures e a review. Int. J. Eng. Sci. 3 (7), 8e20. Rique, A.M., Machado, A.C., Oliveira, D.F., Lopes, R.T., Lima, I., 2015. X-ray imaging inspection of fibreglass reinforced by epoxy composite. Nucl. Instrum. Method Phys. Res. B Beam Int. Mater. Atoms 349, 184e191. Rogge, M.D., Leckey, C.A.C., 2013. Characterisation of impact damage in composite laminates using guided wave field imaging and local wave number domain analysis. Ultrasonics 53, 1217e1226. Romanowicz, P., Muc, A., 2018. Estimation of notched composite plates fatigue life using residual strength model calibrated by step-wise tests. Materials 11, 1e18. Sanjay, M.R., Madhu, P., Jawaid, M., Senthamaraikannan, P., Senthil, S., Pradeep, S., 2018. Characterisation and properties of natural fibre polymer composites: a comprehensive review. J. Clean. Prod. 172, 566e581. Sarasini, F., Tirillo, J., A’Altilia, S., Valente, T., Santulli, C., Touchard, F., et al., 2016. Damage tolerance of carbon/flax hybrid composites subjected to low velocity impact. Compos. B Eng. 91, 144e153.

Testing and damage characterisation of biocomposite materials

227

Sharif-Khodaei, Z., Aliabadi, M.H., 2014. Assessment of delay-and-sum algorithms for damage detection in aluminium and composite plates. Smart Mater. Struct. 23, 1e20. Shekar, H.S.S., Ramachandra, M., 2018. Green composites: a review. Mater. Today Proc. 5 (1), 2518e2526. Skalskyi, V., Stankevych, O., Zosel, T., Vynnytska, S., Thomas, H., Pich, A., 2020. Ranking of fibre composites by estimation of types and mechanisms of their fracture. Eng. Fract. Mech. 235, 1e13. Sohn, H., Dutta, D., Yang, J.Y., Park, H.J., DeSimio, M., Olson, S., et al., 2011. Delamination detection in composites through guided wave field image processing. Compos. Sci. Technol. 71, 1250e1256. Stojkovic, N., Folic, R., Pasternak, H., 2017. Mathematical model for the prediction of strength degradation of composites subjected to constant amplitude fatigue. Int. J. Fatig. 103, 478e487. Su, Z., Ye, L., Lu, Y., 2006. Guided lamb waves for identification of damage in composite structures: a review. J. Sound Vib. 295, 753e780. Su, Z., Zhou, C., Hong, M., Cheng, L., Wang, Q., Qing, X., 2014. Acousto-ultrasonics-based fatigue damage characterisation: linear versus nonlinear signal features. Mech. Syst. Signal Process. 45, 225e239. Thiagamani, S.M.K., Krishnasamy, S., Muthukumar, C., Tengsuthiwat, J., Nagarajan, R., Siengchin, S., Ismail, S.O., 2019. Investigation into mechanical, absorption and swelling behaviour of hemp/sisal fibre reinforced bio epoxy hybrid composites: effects of stacking sequences. Int. J. Biol. Macromol. 140, 637e646. Veidt, M., Hames, S., Ng, C.T., Wattinger, T., 2008. Imaging laminar damage in plates using lamb wave beamforming. Adv. Mater. Res. 47e50, 666e669. Wang, B., Zhong, S., Lee, T.L., Fancey, K.S., Mi, J., 2020. Non-destructive testing and evaluation of composite materials/structures: a state-of-the-art review. Adv. Mech. Eng. 12 (4), 1e28. Wang, D., Onawumi, P.Y., Ismail, S.O., Dhakal, H.N., Popov, I., Silberschmidt, V.V., et al., 2019. Machinability of natural-fibre-reinforced polymer composites: conventional vs ultrasonically-assisted machining. Compos. A Appl. Sci. Manuf. 119, 188e195. Wu, C., Yang, K., Gu, Y., Xu, J., Ritchie, R.O., Guan, J., 2019. Mechanical properties and impact performance of silk-epoxy resin composites modulated by flax fibres. Compos. A Appl. Sci. Manuf. 117, 357e368. Yaghoobi, A., Chorzepa, M.G., 2017. Fracture analysis of fibre reinforced concrete structures in the micropolar peridynamic analysis framework. Eng. Fract. Mech. 169, 238e250. Yaghoobi, A., Chorzepa, M.G., 2015. Meshless modelling framework for fibre reinforced concrete structures. Comput. Struct. 161, 43e54. Yang, P., Shams, S.S., Slay, A., Brokate, B., Elhajjar, R., 2015. Evaluation of temperature effects on low velocity impact damage in composite sandwich panels with polymeric foam cores. Compos. Struct. 129, 213e223. Zhang, W., Zhou, Z., Zhang, B., Zhao, S., 2015. A phenomenological fatigue life prediction model of glass fibre reinforced polymer composites. Mater. Des. 66, 77e81. Zhao, X., Qian, T., Mei, G., Kwan, C., Zane, R., Walsh, C., et al., 2007. Active health monitoring of an aircraft wing with an embedded piezoelectric sensor/actuator network: II. Wireless approaches. Smart Mater. Struct. 16 (4), 1218e1225. Zukas, J.A., Nicholas, T., Swift, H.F., Greszczuk, L.B., Curran, D.R., 1982. Impact Dynamics. John Wiley and Sons, New York.

Sustainable composites and techniques for property enhancement 6.1

6

The context of sustainability in composites (comparison of sustainability of biocomposites versus conventional composites through a details life cycle analysis)

The concept of sustainability has become an object of serious focus of advanced technology, especially in the 21st century. This is necessary for continuous processes and products in the field of composite technology in order to meet the ever-increasing needs and satisfaction of the users and manufacturers. Sustainability cuts across many aspects of life. It involves social, ecological and economic relationships, as well as a balance among earth, life and environment to achieve equitable, bearable and viable inter-dependent benefits among its elements (Fig. 6.1). The natural sources of several bio-fillers and fibres, as well as matrices, have tremendously contributed to the sustainability of biocomposite materials for engineering applications. For instance, biofibres such as jute, pineapple, kenaf, choir, hemp, sisal, flax, cotton, rice husk, date palm, to mention but a few, are processed for the reinforcement of numerous biocomposite components. These natural fibres combined with either petroleum-based matrices to produce bio-based composites or natural matrices, such as polycaprolactone (PCL), polylactic acid (PLA), among others, to fabricate complete biocomposite parts. These components are used in transportation (aerospace, marine and automobile), telecommunication and other engineering sectors. Moreover, natural (plant/vegetable, animal and mineral) fibres and matrices are renewable, biodegradable and sustainable, due to their reliable sources when compared with the conventional composites. Complete synthetic composites are mainly products of the combination of synthetic fibres (carbon, glass, Kevlar, among others) and petroleum-based matrices (epoxy, vinyl ester, among other resins). Sustainability of biocomposites is enhanced through good productive farming or agricultural practices. Most of the natural plant fibres can be grown or cultivated in many countries/continents and harvested more than once in a year; after short periods to produce biocomposites. For example, rice, flax, maize/corn and hemp can be planted and harvested up to 2e4 times in a year, depends on a climatic/geographical condition and land suitability. Other natural perennial plant fibres, such as coconut, date palm, wood, oil palm and bamboo, can produce a huge quantity of fibres for continuous production of fibres and/or fillers for reinforcement of the biocomposite materials. These plants can exist for many years, similar to the stable availability of many natural or biobased matrices/resins.

Sustainable Composites for Lightweight Applications. https://doi.org/10.1016/B978-0-12-818316-8.00002-5 Copyright © 2021 Elsevier Ltd. All rights reserved.

230

Sustainable Composites for Lightweight Applications

Environment

Economic

Equitable

Social

Sustainable Viable

Bearable

Life

Earth Ecological

Figure 6.1 The key elements of sustainability (Yemm, 2013).

Conversely, the increasing rate of the depletion of the petroleum-based matrices and synthetic fibres, in addition to their tedious, harmful, high capital cost of processing and hazardous products made conventional composites to be relatively less sustainable, when compared with biocomposites. With the current trend in fossil fuel consumption rate by the geometrically increasing world population, there is a possibility that many wells of fossil fuel will be exhausted, if there are no new petroleum oil wells detected. Therefore, the manufacturing of many conventional composites will be directly and severely affected, if not totally stopped. With the recent government legislation on recommendation, interest and promotion of the use of renewable energies to prevent environments from pollution, it is obvious that the attention of many composite manufacturing industries will shift to the production of biocomposites. This is because the manufacturing of biocomposites consumes lesser energy and releases lesser CO2, among other toxic gases to the atmosphere, when compared with processing of conventional composites from synthetic carbon and glass fibres, as well as petroleum-based matrices. The process of biocomposite is almost harmless, considering the health of the workers. Environmental pollution is majorly caused by releasing toxic fumes and gases, such as CO2, SO2 and NOx, among other pollutants. These gaseous pollutants are very rampant with the manufacturing processes of synthetic composites (Pervaiz and Sain, 2003;Toffe et al., 2019). In addition, considering the life cycle analyses of both conventional composites and biocomposites. Although many conventional composite parts can be easily recyclable, but with higher release of aforementioned pollutants/gases. Significantly, complete product (composite) environmental life cycle analysis involves five main stages: raw material extraction, production, distribution, use and disposal/recovery. The other

Sustainable composites and techniques for property enhancement

231

first four stages have been directly and indirectly elucidated. Now, considering the last stage on product disposal/recovery after use, both partial and full biodegradability of biocomposite materials support a good disposal mechanism. Biocomposites are discomposed and easily returned to the soils, due to the nature of their compositions. Soil regains nutrients from the disposed biocomposite parts and supports the growth of the plants/vegetables, animals and production of minerals for the fabrication of more natural fibres and matrices. Well-grown plants respire by absorbing more CO2, and therefore, preventing global warming effect and depletion of ozone layers. However, complete synthetic composites are non-biodegradable. Hence, they are liable to cause environmental pollution, whereas, biocomposites are commonly referred to as environmentally friendly products because they thereby support environmental sustainability.

6.2

Inherent properties of natural fibres of biocomposite materials

The inherent properties of several natural fibres significantly depend on their structures and chemical compositions. These two factors are functions of growing/climatic condition, harvest time (Pickering et al., 2007), extraction technique (Bos et al., 2002), fibre treatment and storage procedures of the plants/fibres (Pickering et al., 2016). Based on the natural source and structure of plant fibres, flax fibres, for example, are produced in the stems of flax bast plants. Flax is a cellulose polymer with a more crystalline structure, which makes it stronger and stiffer to handle (Yan et al., 2014). The length of the flax plant ranges up to 900 mm, which possesses strong fibres all along the stem. These fibres have an average diameter of nearly 14 mm. At a macroscopic level, a stem consists of bark, phloem, xylem and a central void when described from outside to the inner region. Looking at a cross section of a fibre bundle at the mesoscopic level, it consists 10e40 fibres linked together mainly by pectin. The fibre has a complex hierarchical microstructure with different length scales and different materials in varying proportions. The outer cell wall, known as the primary cell wall, is 0.2 mm and coats the thicker outer secondary wall (S3). The S3 is responsible for the strength of the material and encloses the lumen, and S2 holds the bulk of the fibre and is the thickest cell wall, as shown in Fig. 6.2. Additionally, the S2 comprises molecules of crystalline cellulose microfibrils and amorphous hemicellulose, orientated at 10 degrees off the fibre axis. It gives flax high tensile strength. The S1 consists of concentric cylinders with a channel in the middle called the lumen. Lumen contributes to the low density, high water uptake (hydrophilic nature), and wettability of several natural/plant fibres. Each layer contains cellulose microfibrils, which arranges parallel to each other. The fibrils of the fibre form an angle with the fibre direction. This angle could affect the strength of the fibres, and usually, the fibre is more ductile, if the microfibrils have a spiral orientation. The microfibrils contribute approximately 70% of the weight of an individual flax fibre (Yan et al., 2014). The secondary cell wall (S1eS3) has a very crystalline structure, making it strong and stiff along the fibre length. The orientation

232

Sustainable Composites for Lightweight Applications

S1 =10°

S2

Secondary cell wall

S3

P

Primary cell wall

Figure 6.2 Schematic microstructure of a flax fibre cell (Yan et al., 2014).

of the structure and amorphous regions between it implies that these properties will differ greatly in the lateral direction (Fig. 6.3). Therefore, it significantly affects the properties of the biocomposites made with them. Furthermore, flax fibres are one of the strongest natural fibres. The tensile strength (mechanical property) of elementary fibres has been observed to be around 1500 MPa, whilst the technical fibres exhibited approximately 800 MPa at a 3 mm clamped length. The modulus of flax fibre varied, due to the diameter of the elementary fibres. It ranged from 39 to 78 GPa for 35e5 mm, respectively (Bos et al., 2006). Also, there are variations in the geometries or dimensions of the same fibre, due to the differences in their diameter, morphology and length. The dissimilarities have been attributed to the different climatic/environmental conditions, ages, soil types and geographical regions, where such agricultural produce (plant fibres) are grown before harvest and process stages. These variations contributed to the inhomogeneous quality and dimensional instability of natural fibre-based biocomposites. An aspect ratio, the ratio of fibre length to its diameter, is an important factor that determines the mechanical properties and damage responses of biocomposite materials. Due to the inconsistent diameters of many natural fibres, it is thus difficult to obtain a consistent fibre aspect ratio without approximation within a single biocomposite material. The strengths of biocomposites increase with fibre aspect ratios. For instance, the influence of drilling process parameters and fibre aspect ratios on surface roughness and delamination drilling-induced damage responses of hemp fibre-reinforced polycaprolactone (HFRP) complete biocomposite samples has been investigated (Ismail et al., 2016).

Elementary fibre

Cellulose microfibril

Lumen

1 – 1.5 m

Secondary walls

Cellulose macrofibril

S3 Crystallised regin

S2 Primary wall

Hemicellulose& lignin matrix

S1 Amorphous regin

Woody body (xylem)

Bast fibre bundle

2 – 4 mm

S2 walls with 30 - 150 lamina

Sustainable composites and techniques for property enhancement

Technical fibre

Cross-section of the plant’s stem

Hydrogen bonds between microfibrils

Pectin between technical fibres

100 – 200 Pm

10 – 40 Pm

ca. 50 nm

233

Figure 6.3 A detailed flax fibre structure, showing its macro to the nanoscale level. Adapted from Woigk, W., Fuentes, C.A., Rion, J., Hegemann, D., van Vuure, A.W., Dransfeld, C., Masania, K., 2019. Interface properties and their effect on the mechanical performance of flax fibre thermoplastic composites. Compos. Appl. Sci. Manuf. 122, 8e17.

234

Sustainable Composites for Lightweight Applications

The optimal HFRP biocomposite sample was recorded with the smallest aspect ratio of 19, implying that the machinability of biocomposites (materials) reduces with an increase in their fibre aspect ratios (strengths). In moving forward, natural fibres and fillers contain hydroxyl groups, which means they attract moisture. This causes weak or poor fibre/filler-matrix interfacial adhesion when embedded with hydrophobic polymer matrices. The weak fibre-matrix interfacial bonding prevents sufficient load transfer from matrices to fibres (reinforcements). Due to this phenomenon, biocomposites possess poor mechanical properties compared to their conventional counterparts. However, this limitation has been reduced by introducing several suitable fibre treatments (chemically - using compatibilising agents, physically, additively and biologically) and overall property enhancement techniques (hybridisation, stitching, pinning, weaving, knitting, among others). These treatments and techniques are later elucidated comprehensively. In addition, natural fibres and bio-resins possess biodegradation properties. They could not resist microbial attack. Therefore, they can completely disintegrate into natural ecosystems, including lakes, oceans, active sludge, marines and natural soils. Still on the advantage from the inherent properties of natural reinforcements (fibres/fillers) and binders (matrices/resins) of biocomposites, they can exhibit chemical transformation, under actions of microorganism and biological enzymes. The major structural component of all plant fibres is cellulose. Bast pineapple, ramie, curaua, kenaf, jute/ flax and hemp fibres with higher cellulose contents of approximately 81, 76, 74, 72, 71 and 68 wt%, respectively (Faruk et al., 2012) and microfibrils exhibit higher performance. Cellulosic fibre and/or bio-polymer materials present in biocomposites are easily attacked by microorganisms, thus enable their easy decomposition, unlike synthetic/conventional fibre and/or petroleum-based materials. Biodegradability, caused by activities of microorganisms, involves hydrolytic depolymerisation, whereby giant cellulose materials are broken down to lower molecular weight compounds to produce monomeric glucose elements. Microorganisms attack starch fast (Reddy et al., 2016). This property supports biocomposite materials to be eco-friendly or environmentally friendly products. Several studies on the biodegradability property of biocomposites have been conducted and reported within this book. Briefly, it has been studied that flax fibres are more biodegradable than polylactic acid (PLA) within the same biocomposite material. Various microorganisms and soil water attacked flax fibre faster than PLA, under the same burial time during soil burial test (Alimuzzaman et al., 2014).

6.3 6.3.1

Improvement of reinforcements and matrices through various treatments and fillers Fibre treatments

Due to some drawbacks of natural fibre-reinforced polymer (FRP) biocomposite materials, it is necessary to treat natural fibres to reduce challenges of weak interfacial bonding formed with polymers, hydrophilicity and poor thermal stability. The processes of mechanical interlocking, chemical, electrostatic and inter-diffusion bonding produce fibre-matrix interfacial bonding (Matthews and Rawlings, 1999; Pickering

Sustainable composites and techniques for property enhancement

235

et al., 2016). The mechanical interlocking depends on the surface roughness of the natural fibres, and electrostatic bonding is influenced by the metallic interfaces and chemical bonding forms in the presence of chemical groups in the matrix and on the surface of the fibre used. The interfacial strength obtained from the chemical bonding depends on the density and type of the bonds, with the application of coupling agents. The coupling agent acts as a bridge, and thus, reduces the incompatibility and provides a chemical bond between the fibre and matrix. An interfacial interaction between the molecules and atoms of the fibres and matrix produces an inter-diffusion bonding. Importantly, it is possible to have multiple types of bonding at the same interface and time (Pickering et al., 2016). Many studies have been conducted to improve on the properties of the natural fibres, and consequently, improve the compatibility between fibres and the matrices (Roy et al., 2020). One of the key motivations of various surface treatments of natural fibre is that it enhances the wettability of the fibres so that matrices wet fibre sufficiently, which leads to enhanced adhesion between the fibres and the matrices. These biocomposite properties include, but are not limited to, mechanical (toughness, tensile strength, modulus, elongation at break, impact, flexural, interfacial strength, among others), physical (wettability) and thermal stability. Mainly, several techniques under chemical, physical, additive and biological treatments have been used (Pommet et al., 2008; Quarshie and Carruthers, 2014). The detailed treatments are subsequently discussed.

6.3.2

Chemical treatments

It is well established that there is an incompatibility between the surface of hydrophilic cellulose fibres and hydrophobic polymeric matrices. Importantly, the chemical modification method contributes towards the compatibilisation of fibres and matrices in which the efficiency of wetting is significantly improved. Therefore, the surfaces of natural fibres are treated with some chemicals to improve the fibre-polymer interfacial bonding or adhesion, by reducing their incompatibility. There are numerous coupling mechanisms in biocomposite materials. These include, but are not limited to: n n n n n n

Acid-base effect Restrained layers Chemical bonding Weak boundary Wettability Deformable layers (Faruk et al., 2012).

Also, there are better improvements and numerous reported studies on chemical treatments than physical treatments. Chemical modifications of natural fibres are performed using alkali, zirconate, acryl, peroxide, titanate, isocyanate, permanganate, acrylonitrile, benzyl and acetyl treatments, as well as maleated anhydride grafted coupling agent (Singh et al., 1996; Faruk et al., 2014; Pickering et al., 2016). The commonest one is the alkaline treatment, using sodium hydroxide (NaOH), commonly known as a caustic soda (Quarshie and Carruthers, 2014), followed by silane, acetyl and maleated anhydride grafted coupling agents. This surface modification technique is considered one of the effective methods to improve interfacial

236

Sustainable Composites for Lightweight Applications

compatibility. In order to make natural fibres as reinforcements in composites, the requirement for improving the properties of natural fibres has been recognised as important steps established by a number of investigations. In alkaline treatment, the components such as hemicellulose and pectin present on the fibre surface are aimed to remove. Non-celluloses, including fats, wax and pectin were removed in the work carried out by Sawpan et al. (2011). The SEM images shown in Fig. 6.4 depict the difference between untreated and surface-treated hemp fibres. For instance, Beckermann and Pickering (2008) treated hemp fibres with a solution of 5 wt% NaOH/2 wt% Na2SO3 to improve thermal stability, crystallinity index, Young’s modulus, lignin reduction, tensile strength and fibre separation of hemp fibres/maleic anhydride grafted polypropylene biocomposites. Similarly, Bera et al. (2010) treated jute fibres with 4%, 8%, 10%, 14% and 28% NaOH with maleic anhydride grafted polypropylene (MA-PP) and vinyl trimethoxy silane (VTMO) coupling agents to improve or increase the tensile, interfacial strength and flexural properties of the resultant biocomposites. The optimum tensile properties of the jute fibres were obtained with 4% NaOH treatment. Furthermore, it was observed that the MA-PP treated fibres exhibited a higher tensile strength of 145% increase when compared with a 42% increase with VTMO counterpart, at a fibre volume fraction of 40%. Also, MA-PP treated Biocomposites recorded a better interfacial strength against fibre pull out. Bachtiar et al. (2008) improved the tensile modulus properties of alkali-treated sugar palm fibre-reinforced epoxy composite when compared with the untreated counterpart. However, this improvement depended on the concentration of the alkali and the soaking time, as depicted in Fig. 6.5. The fibre surface modification using alkaline treatment not only removes the hemicellulose and lignin but also removes the waxes, fatty acids and other organic residues. Similarly, Liu et al. (2004) used alkali treatment to improve the mechanical (tensile and flexural) and impact strength properties of Indian grass fibre-reinforced soy protein-based biocomposites. These enhancements were attributed to the reduction in the fibre inter-fibrillary region by eliminating hemicellulose and lignin, increased

Figure 6.4 Surface morphology of (a) untreated and (b) treated hemp fibre obtained from SEM characterisation (Sawpan et al., 2011).

Sustainable composites and techniques for property enhancement

237

(a) 4000

Tensile modulus (MPa)

3800

3600

3400 1 h. soaking time 3200

4 h. soaking time 8 h. soaking time

3000 0.25 M

Untreated

0.5 M

Concentration of alkali

Tensile modulus (MPa)

(b) 4000 3800 3600 3400 0.25 M NaOH 3200 0.5 M NaOH

h 8

h 4

h 1

Un

tre

at

ed

3000

Soaking time

Figure 6.5 The effects of (a) alkali/NaOH treatment and its concentrations and (b) soaking time on average tensile modulus of sugar palm-reinforced epoxy composite (Bachtiar et al., 2008).

homogeneous distribution of the bio-filler within the matrix and fibre aspect ratio in the biocomposite and consequently, improved efficiency of fibre reinforcement and fibrematrix interfacial bonding. Natural fibre surface impurities are removed while these various treatments are employed. It is important to bear in mind that optimal wt% of treatment must not be compromised. If there is an excessive amount of treatment employed, this will damage the fibre instead of improving the properties. Therefore, the percentage concentration of the alkali and the soaking time are important, as shown in Fig. 6.6(a) and (b).

238

Sustainable Composites for Lightweight Applications

(a) 16

3 Tensile strength

Modulus of elasticity

2 1.5

8

1 4

Modulus (GPa)

Tensile strength (MPa)

2.5 12

0.5 0

0 A

C

B

D

E 3

30 Bend strength

Modulus

2

20 15

1

10

Modulus (GPa)

Bend strength (MPa)

25

5 0

0 A

(b)

B

D

C

E

70

Impact strength (J/m)

60 50 40 30 20 10 0 A

B

C

D

E

Figure 6.6 The (a) tensile, flexural and (b) impact strength properties of grass fibre-reinforced soy-based biocomposites of (A) soya plastic, (B) raw fibre-reinforced plastic composites, (C) 5% alkali at 2 h, (D) 10% alkali at 2 h and (E) 10% alkali at 4 h FRP composites (Liu et al., 2004).

Sustainable composites and techniques for property enhancement

239

There are other reported studies on improvements of several mechanical, thermal stability and moisture resistance properties of natural fibres using alkali treatments (Goda et al., 2006; Gomes et al., 2007; Ibrahim and Hadithon, 2010; Islam et al., 2010; Kabir et al., 2011). The work carried out by Dhakal et al. (2012) used NaOH treatment on the hemp fibre and fabricated hemp fibre reinforced unsaturated composites. Their findings suggest that the surface treatment significantly improved thermal stability and surface characteristics because of 10% NaOH treatment. In the same study, they also highlighted that the wetting behaviour of treated hemp fibre was improved, as shown by the contact angle measurement. Furthermore, acetylation is another type of chemical treatment for the improvement of interfacial adhesion, mechanical (tensile, stiffness and flexural) properties, biological/fungal resistance, the thermal and dimensional stability of natural fibre-reinforced polymer composites/biocomposites (Khalil et al., 2001; Tserki et al., 2005; Bledzki et al., 2008). Sometimes, acetylation can be done after alkaline pretreatment for optimal performance. The natural fibres are soaked in acetic acid, treated with acetic anhydride before the final stage of washing (Quarshie and Carruthers, 2014; Wallenberger and Weston, 2004). Esterification occurs during this process due to the reaction between acetyl groups (CH3CH-) and hydroxyl groups (-OH), as illustrated in Fig. 6.7 (Hill et al., 1998; Pickering et al., 2016). The fibre surfaces are rendered more hydrophobic, thus increases the fibre-matrix interfacial compatibility, reduces the moisture uptake and biological or fungal attack by replacing the hydroxyl groups of a natural fibre cell wall with acetic anhydrides (Quarshie and Carruthers, 2014). This treatment is very common to woods for dimensional stability environmental protection, without reducing other important mechanical (strength and modulus) properties of the concerned natural fibres. Care must be taken to avoid overtreatment of fibres during the acetylation process because it causes fibre cellulose degradation and cracking with the catalyst used. Consequently, the mechanical properties (especially impact strength) of the overtreated fibres and resultant biocomposites are reduced (Bledzki et al., 2008; Pickering et al., 2016). Also, Wallace (2005) reported the use of a coupling agent, known as polymethylene-polyphenyl-isocyanate, to enhance the interfacial bonding (strength and stiffness) between wood fibre and polypropylene (PP) of the wood FRP composites. The synthetic isocyanate has adverse health effects. Therefore, Lee and Wang (2006) manufactured a bio-based lysine diisocyanate to improve the adhesive bond and hence, enhance the mechanical properties and water resistance of bamboo fibrepolylactide biocomposites.

O OH

+

H 3C H 3C

C

O O

CH3 + H3C

O C

C O

C OH

O

Figure 6.7 Esterification reaction between an acetic anhydride and a hydroxyl groups of a natural fibre cell wall (Pickering et al., 2016).

240

Sustainable Composites for Lightweight Applications

In addition, there is a silane treatment. Silanes are very effective molecules used as a surface modifier or a grafting mediator used in composite materials. Due to their silicon-oxygen structure and larger bond angles than hydrocarbon equivalents, they are more popular as an interfacial modifier (Tingju et al., 2013). Primarily, silane treatment involves hydrolysis of alkoxy groups on with water to produce silanol (Si-OH) group and later reacts with eOH groups presents on the surface of natural fibres, as shown in Fig. 6.8. This reaction could lead to the formation of either hydrogen of covalent bonding. Silanes increases the strength of natural fibre-reinforced polymer biocomposites by increasing the hydrophobicity of their natural fibres. A large increase in strength is associated with covalent bonding between silane and matrix used (Rachini et al., 2012; Pickering et al., 2003). The most rampantly used silanes are methacryl, alkyl, amino and glycidoxy silanes. An improved interfacial bonding of both conventional and natural FRP composite materials is possible with the use of an organo-silane coupling agents. For instance, Valadez-Gonzalez et al. (1999) achieved a stronger interaction between short henequen fibres with an organo-silane coupling agent. Consequently, the mechanical properties, such as tensile strength of the biocomposite, were improved, as depicted in Fig. 6.9. The reaction illustrated in Fig. 6.8 can be simply represented as Eqs (6.1) and (6.2), according to Rachini et al. (2012). R
SiðOR1 Þ3 þ 3H2 O/R
SiðOHÞ3 þ 3R1 OH

(6.1)

R
SiðOHÞ3 þ Fibre
OH/Fibre
OðXÞ2 Si
R þ H2 O

(6.2)

Where, R and R1 represent functional carbon and carbon chains, respectively and X denotes eOH or eOSi. Xu et al. (2009) modified the surface of kenaf fibre with a silane coupling agent and improved the interfacial adhesion between the natural fibres and the polystyrene matrix. This was performed after an occurrence of a condensation reaction between the alkoxysilane and hydroxyl groups of kenaf cellulosic fibres. Consequently, the resultant kenaf fibre-reinforced polystyrene biocomposites exhibited a lower tan d and greater storage modulus, when compared with the untreated counterparts. This implies a higher interfacial adhesion between the treated fibres and the matrix resin. Pothan and Thomas (2003) chemically treated the surface of banana fibres using two different silanes: A174 (gamma- methacryloxypropyltrimethoxysilane) and OH

OH OH

+

2 OH

Si

O

R’

OH H

OH

Si

R’

OH

OH O

OH

Si

R’ + H2O

OH

Figure 6.8 Silane reacts with natural fibre surface, where hydrogen bonding (Pickering et al., 2016).

R0

and . denote organic group and

Sustainable composites and techniques for property enhancement

241

30

Tensile strength (MPa)

25 20 15 10 5 0 FIB

FIBNA

FIBSIL

FIBNASIL

Fibre treatment

Figure 6.9 The effects of different fibre surface treatments on the tensile strength of a 20%v/v high-density polyethylene-henequen fibre composites. FIB, untreated henequen; FIBNA, henequen fibre treated with an aqueous alkaline solution; FIBSIL, henequen fibre treated with a silane coupling agent; FIBNASIL, henequen fibre treated first with alkaline solution and then with a silane coupling agent. (Valadez-Gonzalez et al. 1999).

A151 (vinyl triethoxy silane), as well as 0.5% NaOH. Fig. 6.10 depicts the structures of the types of silanes used. Silane A151 treated fibres recorded the highest overall polarity parameter when compared with other treatments. Also, the highest storage modulus was recorded from the same silane A151 treated banana fibre-reinforced polyester biocomposites. While, the lowest hydrogen-bond donating acidity and maximum shift in the loss modulus curve were recorded from the silane A174 treated fibres and its biocomposites, respectively after pre-treatment with 0.5% NaOH. However, banana fibres treated with silane A151 recorded the highest value of hydrogenbond donating acidity. In conclusion, it was evident that silane A174 was an ideal coupling agent for enhancing banana fibre-polyester matrix interfacial adhesion. Hence, an improvement in the storage modulus of banana fibre-reinforced polyester biocomposites has been achieved through chemical modification. Cantero et al. (2003) reported on the influence of different chemical treatments on flax fibre-polypropylene compatibility, considering the surface energy and mechanical properties of the resultant biocomposites. Both natural flax and flax pulp kinds of flax fibres were modified. Maleic anhydride-polypropylene copolymer (MA-PP), maleic anhydride (MA) and vinyl trimethoxy silane (VTMO) were the three applied CH3

OC2H5 CH2

CH

Si

OC2H5

CH2

C

OC2H5 A151

C O O

(CH2)3

OCH3 Si OCH3 OCH3

A174

Figure 6.10 The structures of A151 and A174 silanes used for modification (Pothan and Thomas, 2003).

242

Sustainable Composites for Lightweight Applications

treatments. Their results showed that the polar component of the flax fibre surface energy was reduced using the three treatments. Evidently, the 10% MA-PP treated fibrereinforced biocomposites exhibited the greatest mechanical (tensile and flexural strengths) and other properties (total surface energy), whereas both MA and VTMO treated fibre biocomposites exhibited similar values when compared with the untreated counterparts. Similarly, Ismail and Abdul Khalil (2000) experimentally investigated into the influence of silane coupling agent treatment and partial replacement of oil palm empty fruit bunch fibres (ground into oil palm wood flour) by silica on mechanical properties of the resultant reinforced biocomposites. Based on their report, the addition of a silane coupling agent increased the cure time, scorch time and the mechanical properties (hardness, tear strength, tensile strength, fatigue life and tensile modulus). Also, the addition of silane coupling agent and partial replacement of oil palm wood flour by silica improved the interfacial interaction between the rubber and the filler. In a similar manner, Abdul Khalil and Ismail (2000) treated oil palm empty fruit bunch (EFB) and coconut (Coir) mat fibres with titanate (neopentyl (diallyl) oxy tri (dioctyl) pyro-phosphate titanate) or silane (g-methacryloxypropyltrimethoxy silane), before fabricated into a resultant non-woven fibre mat-reinforced polyester biocomposites. The results obtained from the soil burial (biological) tests conducted were dependent on the silane and acetylation treatments. Both untreated and titanate treated fibres biocomposites recorded a high decrease in both tensile and impact strengths. The acetylated fibre biocomposite exhibited the highest bioresistance to decay or degradation after 12 months of exposure, followed by silane treated FRP biocomposites, cast resin and 4-ply glass FRP composite that were included in the experiment for the purpose of clear comparison. The high soil degradation exhibited by the fibre composites were related to high reductions in their mechanical properties (impact and tensile strengths). Moreover, maleated coupling is extensively utilised todays to improve the properties of the natural fibre-reinforced polymer biocomposites. Comparatively, maleated coupling has a fundamental difference when compared with other chemical treatments. Maleic anhydride is used to modify the fibre surfaces and the polymeric matrices to obtain better mechanical properties and fibre-matrix interfacial adhesion (Faruk et al., 2012). The maleic anhydride grafted polymers are coupling agents, extensively used to enhance the performance of biocomposite materials. Commonly, maleic anhydride is used with compatible polymer matrices. For instance, there is a maleic anhydride grafted polypropylene matrix (MA-PP). MA-PP type is widely used in many reported works, as subsequently discussed. It reacts with eOH groups on the natural fibre surfaces to produce either hydrogen or covalent bonding. There are many reported studies on the application of coupling agents, such as commonly used maleic anhydride grafted polylactic acid or polylactide and polypropylene (MA-PLA and MA-PP) towards enhancement of the properties of natural fibrereinforced polymer biocomposites. A 0.5% MA-PP (G-3015) concentration in toluene was used to modify the jute fibre surfaces (Mohanty et al., 2004). They reported a better fibre-matrix interfacial adhesion that resulted in an approximately 72.3% increase in the flexural strength and significantly reduced water absorption of the treated jute fibre-reinforced polypropylene biocomposite. Average fibre lengths, fibre loading

Sustainable composites and techniques for property enhancement

243

and impregnation time of 6 mm, 30% and 5 min were used, respectively. Similarly, the influence of maleic anhydride (MA) on the mechanical properties (impact strength, hardness, Young’s and flexural moduli) and swelling behaviour of banana, hemp and sisal fibre-reinforced novolac biocomposites have been studied (Mishra et al., 2000). They observed improvements in these properties with all the maleic anhydride treated fibre biocomposites than the untreated counterparts. Evidently, there was an increase in all the mechanical properties and a decrease in both steam and water absorptions, as the maleic anhydride acted as a compatibiliser. Yang et al. (2007) reported an increase in the mechanical (tensile strength and modulus) properties of rice-husk flourreinforced polypropylene biocomposites, using MA-PP coupling agents. This was attributed to the enhancing effect of compatibilising agents on the filler-matrix interfacial adhesion. However, the tensile strengths of the biocomposites decreased with an increased reinforcing filler (rice-husk flour) loading. Also, the addition of the compatibilising agents almost had no significant influence on both notched and unnotched Izod impact strengths of the biocomposites. Moving forward, Liu et al. (2009) used both maleic anhydride grafted polyethylene (PE-g-MA) and maleic anhydride grafted styrene/ethyleneebutylene/styrene triblock polymer (SEBS-g-MA) to improve the impact property and interfacial adhesion between banana fibre and high-density polyethylene (HDPE)/Nylon-6 blends. The HDPE/Nylon-6 based banana fibre-reinforced biocomposites exhibited better strengths and moduli with the addition of SEBS-g-MA when compared with counterpart HDPE based biocomposites. Additionally, both biocomposite systems exhibited slightly different thermal stability, apart from an additional decomposition peak caused as a result of minor Nylon-6 in the biocomposite with HDPE/Nylon-6 blends. Gassan and Bledzki (2000) treated the jute fibre surfaces with a maleic anhydride polypropylene copolymer (MAH-g-PP), as a coupling agent and consequently, improved the jute fibre-polypropylene interfacial bonding of the biocomposites. This better fibre-matrix interfacial adhesion decreased by nearly 30% loss-energy by nonpenetration recorded after the impact test on the concerned biocomposite samples. But, with greater loss energies at different temperatures: cold and warm of
30 and 70 C test conditions, respectively, when compared with a room temperature of 23 C. Furthermore, various improvements on properties and interfacial adhesion of different natural fibres and their fibre-reinforced polypropylene (FRPP) biocomposites with a widely used maleic anhydride grafted polypropylene (MAH-g-PP or PP-gMAH or simply, MA-PP), as a coupling agent have been reported. These include, but are not limited to, trans-crystallinity of jute FRPP biocomposites (Bledzki et al., 2001), surface properties and water uptake behaviour of different types of flax, hemp and cellulose fibres (Bismarck et al., 2002), thermal and crystallisation properties of sisal FRPP biocomposites (Joseph et al., 2003), dynamic-mechanical properties of flax and hemp FRPP biocomposites (Wielage et al., 2003), wetting behaviour of flax FRP biocomposites (Aranberri-Askargorta et al., 2003), characterisation of corn FRPP biocomposites: effects of micro-sized cellulose based corn fibers (Fuqua and Ulven, 2008) and tensile properties of Hildegardia FRPP biocomposites (Li et al., 2010), as reported by Faruk et al. (2012).

244

Sustainable Composites for Lightweight Applications

In addition, as earlier and broadly discussed, Beckermann and Pickering (2008), Bledzki et al. (2010), Bera et al. (2010), as well as Franco-Marqu
es et al. (2011) treated hemp, abaca, jute and newsprint/paper (natural) fibres, respectively, with different concentrations of MA-PP before they were fabricated into their respective biocomposites. Summarily, several mechanical (impact, tensile, flexural strengths and stiffness), thermal, physical and importantly interfacial adhesion or bonding of the natural fibres/ polypropylene biocomposites were improved. Evidently, among all these surface treatments, the use of the MA-PP coupling agent could be regarded to be the most successful method of enhancing the interfacial bonding. It provided almost twice the strength of the concerned biocomposite result obtained with silane treatment (Bera et al., 2010). This significant improvement with MA-PP, especially in mechanical performance, has been attributed to its wettability on natural fibres and improved dispersion (Kazayawoko and Balatinecz, 1997). In a like manner, maleic anhydride grafted polylactic acid or polylactide (MA-PLA) significantly increased the thermal stability, mechanical (tensile, flexural and impact strengths), interfacial adhesion and other important properties of PLA matrix natural (kenaf, bamboo and ramie) FRP composites (Avella et al., 2008; Kang et al., 2013; Yu et al., 2014), respectively. Also, enzyme treatment has been used successfully to improve the properties of biocomposite materials. The use of enzyme treatment on natural fibres is gaining increasing attention today because of its environmental benefits and eco-friendliness (Faruk et al., 2014); for example, Bledzki et al. (2010) modified abaca fibres by fungamix and natural enzymes, before being prepared into abaca fibre-reinforced polypropylene biocomposites. The fungamix modified abaca fibre biocomposites exhibited an increase of 85% in elongation at break, 45% in tensile strength, 35% in flexural strengths, 25% in impact strength and the lowest moisture absorption of 45%, and when compared with the unmodified fibre composites, and among others: maleic anhydride grafted polypropylene (MA-PP) and natural digestion system (NDS) modified abaca fibre biocomposites, as shown in Fig. 6.11(aed), respectively. Also, the modification supported a better resistance in both acid and base medium, as well as an improved decomposition temperature. A comprehensive and critical review on effects of different alkaline treatments (including a percentage of concentration) and coupling agents (such as MA-PP) on mechanical properties tensile strength, elongation at break, tensile modulus, Young’s various biocomposites (old newsprint-filled reinforced polypropylene, high-density polyethylene and linear low-density polyethylene/flax fibre-reinforced polypropylene, corn chaff-reinforced polypropylene, jute fibre-reinforced/poly (butylene succinate), coir fibre-reinforced polypropylene, hemp fibre-reinforced plylactic acid, kenaf fibre-reinforced polypropylene, jute fibre-reinforced polypropylene, bleached hemp fibre-reinforced 1-pentene/polypropylene copolymer, polypropylene/hemp fibres, micro winceyette fibre-reinforced thermoplastic corn starch, flax fibre-reinforced/highdensity polyethylene, high-density polyethylene-reinforced with hemp fibres, hardwood fibres and rice hulls fibres) have been reported by Ku et al. (2011). From this review, it was evident that both coupling agents and alkaline treatments improved the mechanical and other properties of biocomposites to a certain threshold value.

Sustainable composites and techniques for property enhancement

(a)

Abaca

Elongation at break [%]

8

Abaca_MA-PP

245

Abaca_fungamix

Abaca_NDS

6 4 2 0 Tensile

(b)

Abaca

Strength [MPa]

80

Flexural

Abaca_MA-PP

Abaca_fungamix

Abaca_NDS

60 40 20 0 Flexural

6

4

2

0 Abaca

Abaca_MA-PP

Abaca_NDS

Abaca_fungamix

16 Moisture absorption [%]

(d)

Charpy impact strength [mJ/mm2]

Tensile

(c)

12 8 4 Abaca

Abaca_MA-PP

Abaca_NDS

Abaca_fungamix

0 0

20

40

60

80

100

Conditioning period [days]

Figure 6.11 comparison of (a) elongation at break (b) tensile and flexural strengths, (c) impact strength and (d) moisture absorption of modified and unmodified abaca-PP composites (Bledzki et al., 2010).

246

Sustainable Composites for Lightweight Applications

The chemical surface treatment has been reported to have a long-term mechanical performance of natural FRP composites in both wet and humid conditions, with regards to enhancements of instantaneous mechanical behaviour of biocomposite materials (Singh et al., 1996).

6.3.3

Physical treatments

The main aim of using physical treatment methods is to improve the fibre-matrix adhesion through the increase of surface characteristics. With physical treatments, the functional surface and structural properties of natural fibres can be enhanced using plasma, corona discharge, ultraviolet (UV), fibre beating, heat treatment and electron radiation. The chemical compositions of the natural fibres are not extensively changed with physical treatments. Physical treatments improve the mechanical properties by enhancing the fibre-matrix bonding, therefore interfacial adhesion of the biocomposites. Another physical technique is called plasma treatment. The use of plasma treatment helps to roughen the fibre surfaces. This treatment is conducted in oxygen and using a vacuum chamber. It accommodates a variety of fibre surface modifications, depending on the nature and type of the gases used. Plasma treatments can be carried out in a lowtemperature condition to enhance the surface roughness of the natural fibres. It results in hydrophobicity at fibre surfaces, and thus, an increase in the surface roughness of the fibre. Therefore, the interfacial adhesion is increased (Sinha and Panigrahi, 2009; Pickering et al., 2016). For instance, Seki et al. (2009) modified the jute fibre surfaces by subjecting the fibres to a low-temperature oxygen plasma treatment. They investigated the effects of this treatment on the interlaminar shear strength (ILSS) and flexural strength (mechanical) properties of the jute fibres/high density polypropylene composites. They obtained the best interfacial bonding with the biocomposite, using an optimum plasma power of 60 W for 15 min, followed by 30 W with nearly 47% and 32% increase, respectively, compared with the untreated biocomposites. The mechanical properties, such as storage moduli of the plasma treated FRP composites depend on different strain frequencies, hybridisation of the reinforcements, plasma treatment speed and/or scan rate, as depicted in Fig. 6.12. More also, Marais et al. (2005) investigated the effect of cold helium plasma-treated flax fibre-reinforced unsaturated polyester biocomposites on their mechanical (stiffness) and water permeation properties. Their results showed that the plasma treatment enhanced the fibre-matrix adhesion and produced better stiffness. Also, Martin et al. (2000) improved the mechanical properties of a plasma-treated sisal-high density polyethylene biocomposites. Sinha and Panigrahi (2009) studied the effect of argon cold plasma treatment on the structural and wettability of jute fibres, as well as flexural strength of its reinforced unsaturated polyester biocomposite. Their result showed that the application of plasma treatment caused an occurrence of rough surface morphology and hydrophobicity in the jute fibres. This was attributed to a decrease in lignin, phenolic, secondary alcoholic groups or oxidation and hemicelluloses after plasma treatment, as shown in the Fourier transform infrared-spectroscopy results obtained. Consequently, these phenomena improved the interfacial jute fibre-thermosets adhesion, as depicted from micrographs of the plasma-treated fibres biocomposites obtained from the scanning electron microscope used. There was a nearly 14% increase in

35

at 0.1 Hz, 80°C PC/CF

40

PC/CNT/CF

35 Eꞌ (GPa)

Eꞌ (GPa)

30 25 20

15

5

5

(c)

50

at 10 Hz, 80°C PC/CF

PC/CNT/CF

PC/CNT/CF

20

10

40 30 20 10

0

0

6m/min 2m/min Plasma treatment speed

4.5

at 0.1 Hz, 80°C

(d) PC/CF PC/CNT/CF

3 2.5 2 1.5 1 0.5 0

0 Neat

6m/min 2m/min Plasma treatment speed

1m/min

4

at 1 Hz, 80°C

(e) PC/CF

3.5

6m/min 2m/min Plasma treatment speed

2m/min

6m/min Plasma treatment speed

PC/CNT/CF

3 2.5 2 1.5 1 0.5 0

Neat

Neat

4.5 Normalized storage modulus (Eꞌt/Eꞌn)

Neat

Normalized storage modulus (Eꞌt/Eꞌn)

PC/CF

25

10

3.5

at 1 Hz, 80°C

30

15

4

60

(b)

Eꞌ (GPa)

40

45

(a)

Normalized storage modulus (Eꞌt/Eꞌn)

45

4.5 4 3.5

(f)

at 10 Hz, 80°C PC/CF PC/CNT/CF

3 2.5 2

Sustainable composites and techniques for property enhancement

(a)

1.5 1 0.5 0

Neat

6m/min 2m/min Plasma treatment speed

1m/min

Neat

6m/min 2m/min Plasma treatment speed

Figure 6.12 Influences of different strain frequencies, hybridisation of the reinforcements, plasma treatment speeds and/or scan rates on plasma treated FRP composites. Cho, B-G., Hwang, S-H., Park, M., Park, J.K., Park, Y-B., Chae, H.G., 2019. The effects of plasma surface treatment on the mechanical properties of polycarbonate/carbon nanotube/carbon fibre composites. Composites Part B: Engineering 160, 436e445. 247

9

10

(a)

7

(b)

10

at 1 Hz, 170°C

9

PC/CF PC/CNT/CF

8

8

7 Eꞌ (GPa)

6 5 4

6 5

4 3

2

2

2

1

1

1

0

0

PC/CF PC/CNT/CF

2 1.5 1 0.5 0 Neat

Figure 6.12 Cont’d

6m/min 2m/min Plasma treatment speed

3

(e)

1m/min

Neat

6m/min

2m/min

Plasma treatment speed

at 1 Hz, 170°C PC/CF PC/CNT/CF

2.5 2 1.5 1 0.5 0

3.5

(f) 3

at 10 Hz, 170°C PC/CF PC/CNT/CF

2.5 2 1.5 1 0.5 0

Neat

6m/min 2m/min Plasma treatment speed

1m/min

Neat

6m/min Plasma treatment speed

2m/min

Sustainable Composites for Lightweight Applications

2.5

3.5

6m/min 2m/min Plasma treatment speed

Normalized storage modulus (Eꞌt/Eꞌn)

at 0.1 Hz, 170°C

0 Neat

Normalized storage modulus (Eꞌt/Eꞌn)

3

(d)

PC/CNT/CF

5

3

3.5

at 10 Hz, 170°C PC/CF

6

3

6m/min 2m/min Plasma treatment speed

(c)

7

4

Neat

Normalized storage modulus (Eꞌt/Eꞌn)

9

PC/CF PC/CNT/CF

8

Eꞌ (GPa)

at 0.1 Hz, 170°C

Eꞌ (GPa)

10

248

(b)

Sustainable composites and techniques for property enhancement

249

the mechanical (flexural) strength of the 10 min plasma-treated fibre biocomposites, among all the treated fibres, when compared with the raw/untreated fibre biocomposites. In addition, plasma treatment causes a crystallisation, etching, formation of free radicals, chemical implantation and polymerisation (Mohanty et al., 2005), among other effects. Many manufacturers prefer this treatment to others because it is cost effective, environmentally friendly and sustainable. However, it has few problems of scaling up for a commercial purpose and depends on atmospheric conditions, which vary from place to place. Seki et al. (2010) treated jute fibres using oxygen and different powered plasma treatments in radio frequency (RF) and low frequency (LF) plasma reactors. It was reported that mechanical (tensile, flexural and short beam shear) properties of the jute fibre-reinforced polyester biocomposites were enhanced. The LF and RF oxygen plasma jute treated fibre-reinforced polyester biocomposites exhibited a significant increase in interlaminar shear strength (ILSS) of 19.8 and 26.3 MPa, respectively, when compared with 11.5 MPa for the untreated jute fibre-reinforced polyester biocomposite. There were improvements in both tensile and flexural strengths of jute fibre-reinforced polyester biocomposites with oxygen plasma treatment for both LF and RF plasma systems. Evidently, a better improvement on the mechanical performance of the jute fibre-reinforced polyester was achieved with an RF oxygen plasma system in comparison with LF counterpart. Similarly, an increased level of corona treatment has been used to tremendously improve the wettability of some natural fibres, such as woods (Wallace, 2005) and surface energy of the cellulosic fibres. It causes both physical and slight chemical changes to the fibres, such as an increase in the surface polarity and fibre roughness. It increases the compatibility between hydrophilic fibres and hydrophobic matrix. For example, Ragoubi et al. (2010) reported on a significant increase in or improvements on mechanical (tensile strength, compressive stress-strain and 30% of Young’s modulus) properties of corona treated hemp fibre-reinforced polypropylene biocomposites. Likewise, Pizzi et al. (2009) enhanced the mechanical (surface contact angles, hardness, bending/flexural, tensile strength and modulus of elasticity) of corona treated nonwoven flax and hemp fibre mat-reinforced 50/50 natural resin matrices (mimosa flavonoid tannin with hexamine and mimosa tannin with both hexamine and glyoxalated organosolv lignin) lower density thicker and higher density thin biocomposites. However, the application of corona treatment to the three-dimensional (3-D) surfaces such as natural fibres is never found easy (Gassan and Gutowski, 2000; Ragoubi et al., 2010). Furthermore, Gassan and Gutowski (2000) improve the mechanical (flexural and tensile strengths, storage modulus, polarity and tenacity) of tossa jute fibres/epoxy biocomposite, using both corona discharge and ultraviolet treatments on the natural fibres. These treatments increased the fibre polarity. However, the corona treatment decreased the flexural strength of the jute fibres, and consequently, the flexural strength of the biocomposite, while UV treatment enhanced the flexural strength of the biocomposite to nearly 30% under optimum treatment conditions. Also, the jute yarn tenacity (tensile strength) decreased due to the oxidation reaction at the fibre surface. The fibre modulus of elasticity (stiffness) was not affected with various corona energy levels. But, there was nearly 15% improvement in storage modulus at optimal treatment condition of 60 C.

250

Sustainable Composites for Lightweight Applications

Also, fibre beating causes fibre defibrillation, and thus, increases the fibre surface area to promote better mechanical interlocking. For example, Beg and Pickering (2008) enhanced the tensile strength of kraft fibre-reinforced PP biocomposites with modest levels of fibre beating due to the improved interfacial bonding. The unbeaten fibre biocomposite tensile strength of 41 MPa was increased to a threshold value of 45 MPa of beaten fibre composites at an optimum beating time of 5.5 min. Noticeably, it was observed that the tensile strength of the beaten fibre composites reduced upon further beating, shortly after the optimum beating time. Close to fibre beating is the heat treatment. As its name implies, this is a process of heating the fibres near their degradation temperature. This process affects the fibre chemical (chain scission, the formation of carboxyl, carbonyl and peroxide, as well as free radical production groups), mechanical and physical properties (Pickering, 2008). Other changes include the degree of polymerisation, water content, strength, cellulose crystallinity and chemistry. The heat treatment is very similar to both plasma and corona treatment because it depends on the composition, time and temperature of the gases used during the treatment. For instance, the crystallinity index of kenaf fibres has been increased with heat treatment (Cao et al., 2007). Consequently, this caused an improvement of its mechanical (tensile strength and elongation at break) properties. A maximum or an optimal tensile strength of the treated fibre was obtained at 140 C, which later decreased upon further heating. Rong et al. (2001) improved the mechanical properties of unidirectional (UD) sisal-reinforced epoxy biocomposite after heating the sisal fibres at 150 C for 240 min in an air-conditioning oven. Additionally, electron radiation produces free radicals that assist crosslinking between the hydrophilic bio-fibres and the hydrophobic matrix. The interfacial bonding between natural fibres (single hemp, ramie, flax and cotton)/fibre bundles and polymer (polypropylene) has been enhanced with electron radiation of 10 MeV at intensities of 5, 15 and 33 kGy on the fibres (Huber et al., 2010). The interfacial shear strength of the biocomposites increased up to 50%. This was attributed to the formation of radicals with the cellulose chains of the natural fibres and the molecules of the polypropylene, causing a crosslinking, and consequently, and improved adhesion between the fibres and the matrix. Moreover, both hydrophilicity (rate of moisture uptake) and thermal degradation of natural fibres can be decreased by using hydrothermal treatments. For instance, Wallenberger and Weston (2004), as well as Stamboulis et al. (2000) used this method to reduce the thermal degradation and improve the moisture and thermal resistance of treated Duralin flax, as dew retting process was eliminated, when compared with the untreated counterpart.

6.3.4

Additive treatments

This method aims to enhance the properties of matrix materials, as well as fibres so that the overall properties of composites are improved. The natural fibre-polymer matrix interfacial bonding can be enhanced by impregnating the natural fibres with a suitable

Sustainable composites and techniques for property enhancement

251

matrix. This method of fibre treatment is not common, unlike the two aforementioned and extensively discussed methods. Additive treatments involve using either a solvent exchange method or by infusion method, whereby monomers are infused into the fibres followed by polymerisation with the aid of a catalyst, radiation or heat, as reported by Mohanty et al. (2005).

6.3.5

Biological treatments

Similar to chemical treatments, the biological process could be used to modify and improve the surface of natural fibres to produce enhanced biocomposites. For example, Pommet et al. (2008) deposited cellulose nanofibrils on the surfaces of both natural hemp and sisal fibres, using bacteria (Acetobacter xylinum), as shown in Figs 6.13 and 6.14, respectively. The nanofibrils acted as substrates during the fermentation stage of bacterial cellulose. After a single fibre pull-out/tensile test was carried out, a significant improvement in the biologically treated fibre biocomposites was observed. This was attributed to an improved interfacial bond between the 5%e6% bacterial cellulose treated fibres and the polymeric matrices (cellulose acetate butyrate and polylactic acid) used. Recently, Roy et al. (2020) investigated the effects of various chemical treatments on the performance of jute fibres filled natural rubber (NR) composites. They used three different types of surface treatments, namely; alkali treatment, combined alkali/stearic acid treatment and combined alkali/silane treatment. The influence of surface treatments on the jute fibres was characterised, using advanced techniques such as X-ray diffraction (XRD), Fourier transform infrared (FTIR) spectroscopy and field emission scanning electron microscopy (FESEM). The morphology from their report suggested that surface roughness of treated fibres were found rougher than the nontreated, as shown in Fig. 6.15. The rough surface acted as a mechanical interlocking, which contributed to improved properties. Their results exhibited that there was a significant improvement in the properties of jute fibre-reinforced NR biocomposites. The improvement in the properties was attributed to strong fibre-matrix interaction due to various surface treatments used.

Figure 6.13 SEM micrographs of (a) untreated natural and (b) bacterial cellulose modified hemp fibre surfaces (Pommet et al., 2008).

252

Sustainable Composites for Lightweight Applications

(a)

(b)

1 Pm

1 Pm

(c)

1 Pm

Figure 6.14 SEM micrographs of (a) untreated natural, (b) sisal fibre with the attachment of bacterial cellulose and (c) acetone treated and bacterial cellulose modified sisal fibre surfaces (Pommet et al., 2008).

JFun A–JF A-St-JF

65

A-Si-JF

60 55 50 Intensity (counts)

45 40 35 30 25 20 15 10 5 0 –5 0

10

20

30

40

50

60

70

80

2 theta (degree)

Figure 6.15 XRD patterns for treated and non-treated jute fibres (Roy et al., 2020).

Sustainable composites and techniques for property enhancement

6.4

253

Approaches towards overall property enhancement via hybridisation, pinning, stitching, among others

Several efficient techniques have been developed towards overall property enhancement of fibre-reinforced polymer biocomposite laminates. These approaches improve or increase damage tolerance and properties of biocomposites, such as the ultimate strength, delamination resistance and impact damage tolerance, among others. Although, as the application of biocomposite increases, there may be a need for further improvement of these approaches using new technology. The most widely used approaches are stitching, hybridisation, pinning, knitting, weaving, braiding and tufting, as subsequently expatiated.

6.4.1

Stitching

Stitching is a method adopted to enhance the interlaminar fracture toughness of biocomposite, among many textile techniques of inserting of a through-thickness yarn into traditional two-dimensional (2-D) fibre reinforcements or preforms. Other methods include, but are not limited to, knitting and weaving. Stitching is probably the most commonly used method due to its versatility and ease of handling (Beier et al., 2007; Velmurugan and Solaimurugan, 2007; Tan et al., 2010). It is considered to be one of the main methods for the automated and cost-reduced production of complex textile preforms, later used for liquid composite moulding of high performance fibre-reinforced polymeric biocomposites (Beier et al., 2007; Weimer and Mitschang, 2001). It enhances the impact damage resistance, interlaminar fracture toughness, lap joint strength, delamination crack growth resistance and tolerance of the final biocomposites with suitable and efficient thread and pattern (Mouritz et al., 1997). For instance, Yudhanto et al. (2015) enhanced the tensile strength of composites stitched a with thicker thread because of the effective hindrance of edge delamination. Also, Velmurugan and Solaimurugan (2007) reported that stitching increased the Mode I delamination toughness of fibre polymeric-reinforced composite up to 20 times greater than the unstitched counterpart sample. But, Jain et al. (1998) recorded an enhanced Mode I delamination resistance up to 4 times; lower than that of Velmurugan and Solaimurugan. They study crack propagation of the fibre-reinforced polymer composites, depending on stitch density, thread diameter and type. These results were similar to the results observed by Jain and Mai (1994), as well as Dransfield et al. (1998). The stitching pattern could be lock, modified lock, chain or zigzag type (Fig. 6.16). In detail, Fig. 6.17 depicts a clear description of a stitched pattern through a quasi-isotropic composite laminate. Due to the aforementioned benefits associated with the use of stitching techniques, its potential applications in automobile parts include floor panels, bumper bars and door frame members. And stitching is used as a part of techniques for manufacturing some aircraft structures, such as wing panels, fuselages and blade stiffened parts (Mouritz et al., 1999; Mouritz and Cox, 2000).

254

Sustainable Composites for Lightweight Applications

(a)

(b)

(c)

(e) Nominal 5 mm stitch pitch

(d) 5 mm

Nominal stitch row spacing

Figure 6.16 Illustration of the seam (a) lock, (b) modified lock, (c) chain, (d) plain and (e) zigzag patterns of stitching (Beier et al., 2007; Velmurugan and Solaimurugan, 2007; Mouritz and Cox, 2000). Needle thread

Laminate

Bobbin thread y

Spacing

Needle thread

z

Performs

Knot

Pitch Bobbin thread x

1 mm

Figure 6.17 A stitched quasi-isotropic composite laminate, showing its modified-lock pattern (Mouritz and Cox, 2010; Yudhanto et al., 2014).

Sustainable composites and techniques for property enhancement

(a)

(b)

Resin-rich zone Yarn

F

z

255

Misaligned yarn

F

Laminate

z

Stitching yarn tension

x

x

Through-thickness yarn

(d)

(c) Surface yarn

Laminate

Resin-rich zone

Ondulation

y

z x

x

Figure 6.18 Schematic illustration of various fibre distortions caused by stitching (Beier et al., 2007).

However, damaging effects of stitching on fibre preforms for biocomposites include fibre waviness or distortion, resin crack, crimping, induced undulations of the fibres, broken in-plane fibres, increased resin-rich zone and porosity, among others. Consequently, they lead to the degradation of the mechanical (flexural, tensile and compressive) properties of the composites (Khan and Mouritz, 1996; Velmurugan et al., 2004). In moving forward, some of the stitch-induced damages are illustrated in Fig. 6.18. In addition, the final extent of fibre damage depends on tension on stitch thread, the stitch type, alteration in effective fibre volume fraction, stitch span, consideration process during curing of the biocomposites, type of stitch yarn material, stitch density, the diameters of the stitching yarn and the needle (Mouritz et al., 1997; Yudhanto et al., 2015). The typical range of stitch diameter is 0.1e1.0 mm, above this range often results in detrimental distortion in the fibre architecture (Mouritz and Cox, 2010). Furthermore, the influences of through-thickness stitching of natural fibre on the mechanical (Mode I interlaminar fracture toughness and tensile) properties of flax fibre-reinforced epoxy biocomposite laminates have been reported (Ravandi et al., 2016). Preforms of woven fabric and unidirectional (UD) flax fibre laminates were stitched in various stitch densities, using flax yarn and cotton thread. A decrease in tensile properties (strength and stiffness) of the composites was similarly observed with both stitch materials (flax yarn and cotton thread), as shown in Fig. 6.19(a) and (b). This was attributed to the imperfection that resulted by the stitching. The interlaminar fracture toughness, GIC of the composite was not significantly enhanced with cotton thread stitch, but flax yarn stitch improved this property by at least 10% at the lowest stitch fibre area fraction. Evidently, the flax yarn stitch exhibited a higher fracture toughness of the biocomposite laminates than the cotton thread stitch, as depicted in Fig. 6.19(c).

256

Sustainable Composites for Lightweight Applications

(a)

95

Tensile strength (MPa)

90 85 80 75 70 65 60 Cotton thread stitch

55

Flax yarn stitch 50 0.000

0.005

0.010

0.015

0.020

0.025

Stitch areal fraction

(b)

9

Tensile stiffness (GPa)

8.5 8 7.5 7 6.5 6 Cotton thread stitch

5.5

Flax yarn stitch

5 0.000

0.005

0.010

0.015

0.020

0.025

Stitch areal fraction

(c)

4500 PWF4 4000

PWF2

PWF3

Gc(J/m2)

PWF1 3500 PW0 PWC2

3000

PWC4

PWC3 PWC1

2500

Cotton thread stitch Flax yarn stitch

2000 0.000

0.005

0.010

0.015

0.020

0.025

Stitch areal fraction

Figure 6.19 Effects of various stitch materials (cotton thread and flax yarn) and their areal stitch fractions on the tensile (a) strength, (b) stiffness and (c) Mode I interlaminar fracture toughness properties of natural flax fibre-reinforced epoxy biocomposite laminates (Ravandi et al., 2016).

Sustainable composites and techniques for property enhancement

257

In the following year, the influences of through-thickness stitching of natural fibre on the low-velocity impact response of continuous flax fibre-reinforced epoxy biocomposite laminates have been additionally studied (Ravandi et al., 2017). Similar impact behaviour of both flax and cotton stitched biocomposites was observed, regardless of differences in their stitch fibre type and thickness. Besides, the intra-laminar fracture toughness of the stitched woven flax fibre-reinforced biocomposite laminate reduced by nearly 16% with flax yarn stitch, while the intra-laminar fracture toughness of the cotton yarn stitched laminate was decreased by only 5%. Hence, it was evident that the flax yarn stitch reduced the impact resistance of the composite laminate under low-velocity impact, and the use of flax yarn stitch is unfavourable to the higher structural performance applications of the composite. More also, the effect of water absorption on the Mode I interlaminar fracture toughness of stitched and unstitched hybrid flax/basalt fibre-reinforced vinyl ester biocomposite laminates have been experimentally investigated (Almansour et al., 2017). Using three types of composite laminates: flax fibre-reinforced vinyl ester (FVE), unstitched hybrid (FBVEu) and stitched hybrid (FBVEs). The results obtained showed that the water immersed FVE laminate exhibited decreased Mode I interlaminar fracture toughness initiation and propagation by an average of 27% and 10%, respectively, when compared with the dry counterparts, whereas that of water immersed FBVEu and FBVEs biocomposite laminates were enhanced by nearly 15% and 17%, compared with the dry counterparts. These further established the fact that basalt fibre hybridisation improves both durability and water/moisture absorption resistance of natural fibre-reinforced polymeric biocomposites. It was observed that FVE laminate sample recorded the highest water absorption or weight gain of 5.38% before a plateau was reached around 5 weeks (900 h). Importantly, both FBVEs and FBVEu types of hybrid biocomposite laminates exhibited similar water absorption responses of 3.70% and 3.57%, respectively, because of the same balanced lay-up stacking sequence used. Comparatively, FBVEs recorded higher weight gain or water uptake than the FBVEu counterpart, as shown in Fig. 6.20. This can be attributed to the absorption property of the stitching thread, materials and more space created by needle and thread for water to lodge (water ingress). Hence, stitched hybrid natural fibrereinforced biocomposite attracts an increasingly small percentage of water uptake, and consequently, limits the stitching performance and some mechanical properties. These results were further evaluated and confirmed by using efficient microscopic examination techniques: scanning electron microscopy (SEM) and X-ray computed micro-tomography (mCT). In addition, in the following year, the same authors advanced the experimental investigation to cover Mode II interlaminar fracture toughness of the same hybrid stitched and unstitched flax/basalt-reinforced vinyl ester polymeric biocomposite laminates. They reported that the crack length of FBVEu dry biocomposites recorded high delamination resistance and an increased crack propagation resistance. The delamination phenomenon was effectively arrested by the stitch yarn used in FBVEs laminates and resultantly caused an increase in the initial interlaminar fracture energy, GIIC init. by 62%. However, the propagation of interlaminar fracture toughness, GIIC prop. was reduced by 36% (Almansour et al., 2018).

258

Sustainable Composites for Lightweight Applications 6

Weight gain (%)

5 4 3 2 1 0 0

5

10

15

20

25

30

35

Time (h)1/2 FVE

FBVEs

FBVEu

Figure 6.20 Comparative water absorption responses of stitched and unstitched hybrid natural flax/basalt fibre-reinforced biocomposite laminates (Almansour et al., 2017).

6.4.2

Hybridisation

The technique of combining more than one type of fibre-reinforced in the same matrix of a biocomposite system is called hybridisation. With the use of the hybrid system, the mechanical properties of natural fibre-reinforced polymeric biocomposites can be improved. In recent years, the hybrid technique has been used to overcome some of the drawbacks of natural fibre-reinforced bio-based composites, which tailors the important properties of two reinforcements in the same matrix. It is one of the widely used techniques in the contemporary world of composite technology. The hybridising material (sisal fibre) can be arranged in different stacking sequences (Fig. 6.21); either as skins (Fig. 6.21(a)) or core (Fig. 6.21(b)). The combined inherent properties of the two or more constituent fibres provided greater advantages over the use of conventional single fibre-reinforced polymeric biocomposites. Other attractive factors of employing hybridisation technique include, but are not limited to, the following benefits:

(a)

(b) Hemp (H)

Sisal (S)

(c)

(d)

(a) - SHHS (b) - HSSH (c) - HHSH (d) - HSHS

Figure 6.21 Stacking sequences of hybridising sisal fibre mats into hemp fibre mat-reinforced polymeric biocomposites (Kumar et al., 2019).

Sustainable composites and techniques for property enhancement n n n

n

259

Weight and cost reduction. Both weight of a hybrid biocomposite of a certain thickness and its cost of production are lower than that of synthetic fibre composite counterparts. Improvement to environmental effects since more natural fibres are used to obtain biocomposite laminates of higher thickness, density and strengths. Enhancement of mechanical, thermal and other properties of the natural fibre-reinforced polymeric biocomposites, compared with the synthetic fibres (glass, carbon and such like) reinforced polymeric composites. Application where high performance is required (Almansour et al., 2017).

However, hybrid natural fibre-reinforced polymeric biocomposites still have some reduced limitations initially associated with the single natural fibre-reinforced polymeric biocomposites, despite their aforementioned outstanding properties. Some of the major drawbacks of single natural fibre-reinforced polymeric biocomposites have been extensively discussed in the earlier chapters. In addition, natural fibrereinforced polymeric biocomposites have not been widely used as structural components, due to their lower strength and stiffness, as well as higher moisture absorption properties (Faruk et al., 2012). Therefore, some of the challenges of hybrid biocomposite systems are thus briefly highlighted. n

n n n n

Moisture absorption and swelling effect over a long duration, due to the combination of more hydrophilic natural plant fibres, probably the hybridising fibre has a higher hydrophilic nature. Higher cost of manufacturing occasionally. Requirement of special manufacturing process(s) or technique(s) in a few cases. Possibility of more or complex cracking, delamination effects, among other manufacturinginduced defects and post-manufacturing/test damage. Problem of incompatibility (weak interfacial adhesion) between two or more different natural fibres and a single matrix.

There are greater benefits derived from the properties and application of hybrid natural fibre-reinforced polymeric biocomposites over single fibre-reinforced polymeric biocomposites and much more over the conventional fibre-reinforced polymeric composite. Therefore, many studies have been conducted and reported on hybrid biocomposite systems. For instance, due to the excellent chemical, acoustic, mechanical and thermal properties of basalt fibres, they have been used as hybridising materials into several natural fibre-reinforced polymeric biocomposites (Dhakal et al., 2015; Fiore et al., 2016, 2017). The post-impact, quasi-static and cyclic flexural properties of hybrid hemp/basalt fibre-reinforced unsaturated polyester biocomposite were enhanced (Dhakal et al., 2015). The hybrid hemp/basalt (HB) reinforced biocomposite exhibit greater flexural stress-strain (modulus) of non-impacted biocomposite samples and normalised flexural strength of impacted samples under different impact energies of 3, 6 and 9 J, when compared with single natural hemp (H) fibre-reinforced unsaturated polyester biocomposite, as depicted in Fig. 6.22. Similar results on higher values of flexural strength, modulus and Mode II interlaminar fracture toughness of hybrid flax/basalt fibre-reinforced vinyl ester biocomposite laminates were obtained by Almansour et al. (2015), when compared with the single or non-hybrid flax fibrereinforced laminate counterparts.

260

Sustainable Composites for Lightweight Applications

(a) 180

H HB

160

Flexural stress (MPa)

140 120 100 80 60 40 20 0 0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Strain (%)

(b)

Normalized flexural strength

1.0 Hemp Hemp/basalt 0.9

0.8

0.7

0.6

0.5 0

2

4

6

8

10

Impact energy (J)

Figure 6.22 (a) The stress-strain curves of hemp and hemp/basalt biocomposites before impact test and (b) varied residual normalised flexural strength with impact energy (Dhakal et al., 2015).

Similarly, the effects of hybridisation technique with basalt fibre as double external layers on the mechanical (impact and flexural strength), water absorption and modulus properties of flax fibre-reinforced epoxy biocomposites under 0, 15, 30, 45 and 60 ageing days of salt-fog environment conditions have been investigated (Fiore et al., 2016). The basalt and flax fibres were arranged to form two basalt fibres (skin)-six flax fibres (core)-two basalt fibre (skin) stacking sequence of flax/basalt biocomposite laminates, using a balanced lay-up method. The hybridised flax/basalt biocomposite

Sustainable composites and techniques for property enhancement

(a)

261

120 V [MPa]

Flax

100

Flax–basalt

80 60 40 20 0 0.0

(b) 8000

H [%]

2.0

4.0

E [MPa]

6.0

8.0

Flax

6000

Flax–basalt

4000 2000 0 –2000 –4000 –6000 H [%]

–8000 0.0

2.0

4.0

6.0

8.0

Figure 6.23 The effects of hybridisation technique on mechanical properties of flax fibre biocomposites, showing (a) stress-strain and (b) modulus estrain curves of flax and flax/basalt fibre-reinforced unaged biocomposites (Fiore et al., 2016).

laminate exhibited greater impact, flexural modulus and strength by 28%, 49% and 71% without ageing day, respectively, when compared with non-hybridised flax fibre biocomposite counterpart (Fig. 6.23). In addition, the hybrid flax/basalt laminates possessed a greater water uptake resistance property than that of the non-hybrid flax fibre laminate, due to the presence of hybridising material (basalt fibre) skins. The hybrid flax/basalt and non-hybrid fibre laminates recorded maximum water uptake values of 9% and 23%, respectively, before a plateau was reached by both laminates after an approximately aging time of 700 h (Fig. 6.24). This result was very similar to that of Almansour et al. (2017) with the same hybrid flax/basalt fibre-reinforced biocomposite, but different polymeric matrix (vinyl ester) was used. However, the impact strengths (durability) of the hybrid biocomposite laminates were lower after 15, 30, 45 and 60 ageing days, when compared with the non-hybrid flax biocomposite counterparts, as depicted in Fig. 6.25. The letters A, B, C, D and E represented 0, 15, 30, 45 and 60 ageing days, respectively.

262

Sustainable Composites for Lightweight Applications 35

Flax Water uptake [%p/p]

30

Flax–basalt 25 20 15 10 5 0 200

0

600

400

800

1000

1200

1400

1600

Aging time [h]

Figure 6.24 Influence of basalt fibre, as a hybridising material on water uptake of flax fibrereinforced biocomposite laminate (Fiore et al., 2016).

Fiore et al. (2017) hybridised jute fibre-reinforced epoxy biocomposite laminates with basalt fibre, using intercalated and sandwich-like basalt-jute stacking sequences, as well as ply-substitution approach. The durability of the jute-basalt interply hybrid laminates exposed to salt-fog was improved. Evidently, the optimum durability results of mechanical properties (quasi-static flexural modulus and impact peak load) were obtained with sandwich-like configuration after 90 days of ageing, when compared with jute fibre-reinforced and intercalated configurations. Admittedly, the improvement in properties of natural fibre-reinforced polymeric biocomposite can be attributed to the source of hybridising basalt fibre. Basalt fibres are obtained from the natural mineral 55

Flax

50

Flax–basalt

Impact strength [KJ/m2]

45 40 35 30 25 20 15 10 5 0

A

B

C

D

E

Figure 6.25 Comparative effects of 0, 15, 30, 45 and 60 ageing days of salt-fog environment conditions on impact strengths of both hybrid flax/basalt fibre-reinforced epoxy polymeric biocomposite and non-hybrid flax fibre counterparts (Fiore et al., 2016).

Sustainable composites and techniques for property enhancement

263

produced when rocks are melted. It is a non-harmful resource, hence considered as an environmentally friendly fibre when compared with the synthetic glass and carbon fibres (Sarasini et al., 2013; Fiore et al., 2016). Consequently, basalt hybridisation into hydrophilic natural fibres enhances the mechanical properties and water absorption resistance of biocomposites by improving the interfacial adhesion between fibres and matrix. With these benefits, basalt fibres have potential of being used as hybridising materials of natural fibre-based biocomposites for structural applications (Sim et al., 2005). Moreover, Aji et al. (2012) studied the effects of hybridisation on mechanical (tensile, impact and flexural strengths) properties kenaf/pineapple leaf fibre-reinforced high-density polyethylene biocomposites. Both tensile and flexural of the biocomposites were increased due to the addition of the pineapple leaf fibre. The kenaf fibre enhanced both impact strength and water absorption resistance. In like manner, the influence of hybridisation on the mechanical properties of another natural banana and sisal fibres has been evaluated (Idicula et al., 2010). The addition of banana fibres improved the tensile properties of the hybrid short banana/sisal fibre-reinforced polyester biocomposites. The biocomposites with the ratio of banana to sisal (3:1) at 67 vol% fibre content recorded a maximum tensile strength of 58 MPa. This was attributed to the smaller diameter of banana fibres when compared with the sisal fibres, better stress transfer and improved banana-polyester interfacial bond present in the biocomposite system.

6.4.3

Pinning

Pinning is usually carried out on prepregs using a hand-held ultrasonic device (Tong et al., 2002; Mouritz, 2007), with a higher possibility of distortion compared with weaving. Also, the influence of the pinning technique on FRP composite damage and mechanical responses depends on the impact of energy, damage mechanisms (distribution and modes) and laminate layup (Francesconi and Aymerich, 2018). Zpinning (Fig. 6.26) is quite different from other techniques because it is the only method applicable to the reinforcement of the prepreg laminates, usually in the direction of through-thickness (Mouritz, 2007).

z-pin

Figure 6.26 A typical pinned composite (Mouritz and Cox, 2010).

264

Sustainable Composites for Lightweight Applications

An ultrasonically assisted method has been used to improve the insertion of z-pins into FRP composite samples (Hoffmann et al., 2019). From the results obtained, there were microstructural changes, absence of fibre waviness, as well as resin-rich regions that surrounded each steel pin. Additionally, lack of friction and poor pin-laminate interfacial bonding were observed. The volume content of z-pins in FRP composite materials improved their mode I interlaminar fracture properties. It enhanced the delamination, fatigue and compression resistance of composite materials, especially after impact damage, as investigated and reported by several studies (Cartié et al., 2006; Zhang et al., 2008; Cartié et al., 2009; Pingkarawat and Mouritz, 2014, 2016; Mouritz and Koh, 2014; Pegorin et al., 2015; Ladani et al., 2016). To reduce variance associated with the pinning technique and obtain better results necessitated the use of a robotic insertion device or automatic multi-pin insertion system (Qinghua et al., 2014). The automatic system was capable of inserting multiples of pins during the preparation of Z-pin FRP composite laminates has been developed. It has 5 times more efficiency than the extant machines. The innovative system enhances accuracy, productivity and supports commercialisation of zpin technology. Pinning resulted in high fibre breakage and waviness, crimping, resin pockets and inplane fibre distortion around a pin inserted into biocomposites (Fig. 6.27), if not properly handled by an experienced manufacturer. Resin-rich or pocket zones occur around the pin, where fibres have been displaced by the inserted pins in the biocomposite system.

(a) Load axis T

Pin

Resin pockets

(b)

Figure 6.27 Micrograph of a pinned biocomposite, showing the defects of (a) resin-rich, fibre distortion or waviness and (b) crimping defect (Chan et al., 2006; Mouritz, 2007; Mouritz and Cox, 2010).

Sustainable composites and techniques for property enhancement

265

Furthermore, cohesive modelling has been used to predict the behaviour of zpinned FRP composites. The simulation observations obtained were very close to the experimental results. It was evident that the fracture resistance curve must be discretised along the z-pin field in order to accurately capture the delamination dynamics in FRP composites (Ranatunga and Clay, 2012). In moving forward, a new finite element model has been developed to predict mode I interlaminar fracture toughness of z-pinned FRP composite laminates. Three different z-pinned materials were inserted into the composites used. These included high strength z-pins (stainless steel and titanium) and soft material (copper). It was observed that the model accurately predicted the crack growth resistance curves, as well as the interlaminar fracture behaviours of the FRP composites with various z-pins (Fig. 6.28). The lines and points denote the finite element prediction and experimental values obtained, respectively. The titanium and steel z-pinned composites failed by pull-out, while the copper z-pinned sample failed by both pull-out and rupture. Several methods have been used to evaluate and determine the through-thickness resistivity of FRP composites. For instance, the use of metal z-pins decreased the resistivity, and therefore, it depended on the electrical conductivity of the z-pin material. An increase in z-pin conductivity produced a decrease in the resistivity. The copper z-pin has a higher conductivity, compared with other materials, as reported (Pegorin et al., 2018).

6.4.4

Knitting

Interlaminar fracture toughness (kJ/mm2)

Fabric reinforcements are required for the biocomposite manufacturing processes, such as resin transfer moulding, vacuum infusion and hand or spray lay-up (as earlier and extensively discussed in Chapter 4). Natural fibres (such as flax, hemp, sisal,

12

Ti

Steel

8

Cu (pullout)

Cu (rupture)

4

Unpinned 0

0

20

40

60

Crack growth (mm) Unpinned

Copper

Titanium

Steel

Figure 6.28 Experimental and finite element results of the interlaminar fracture toughness of a z-pinned FRP composite material, showing the effects of different z-pin materials (Blacklock et al., 2016).

266

Sustainable Composites for Lightweight Applications

among others) can be warp knitted into either uniaxial or biaxial fabrics. Then, it can be woven into biaxial plain weaves. For instance, flax fibres were warp knitted by Goutianos et al. (2006). Warp knitting can be carried out using a warp knitting machine with weft insertion, whereby each warp thread is knitted to form interlaced loops, and fibre yarn straight inlays are introduced in the weft direction. Therefore, the fabric products are crimp-free in both directions. The fire yarns can be oriented in a uniaxial direction. Second, yarn can also be inserted in the warp direction to produce a 2-layer biaxial woven non-crimp fabric system. Numerous studies have been conducted on the properties of several knitted composite materials. Mechanical behaviours of 3D stitched woven-knitted basalt fabric biocomposites, to start with, were investigated (Hu et al., 2010). Four categories of knitted configurations were considered (Fig. 6.29). From the results obtained, it was observed that the mechanical properties of the 3D stitched woven-knitted composites were significantly influenced by the type of knitted structure. The interlock structure recorded the maximum energy absorption and tensile strength, while the plain knit structure exhibited the maximum Young’s modulus, when compared with other types. Figs 6.30e6.32 show tensile, bending and impact responses obtained by the four

(a)

(b)

Warp

(c)

Weft

(d)

Figure 6.29 Various knitted arrangements adopted: (a) varied plain knit, (b) 1 x 1 rib, (c) interlock and (d) Milano (Hu et al., 2010).

Sustainable composites and techniques for property enhancement

267

(a) 700 600

Tensile stress (MPa)

Varied plain knit 500 400

Milano Interlock

300 200 Rib 100 0 0.00

0.02

0.04

0.06

0.08

Strain

(b) 250 Interlock

Tensile stress (MPa)

200

Milano

150 Rib 100

50 Varied plain knit 0 0.00

0.02

0.04

0.06

0.08

0.10

Strain

Figure 6.30 Tensile properties of the four samples, showing their responses at both (a) warp and (b) weft directions (Hu et al., 2010).

samples, respectively, at both warp and weft directions. Similar results on mechanical performances of an interlock2 knitted composite samples have been reported (Demircan et al., 2015), with maximum tensile, 3-point bending, as well as 3-point bending impact responses, when tested at weft direction and compared with the tuck-s-miss, plain and tuck counterparts. Also, interlock knitted composite exhibited highest interlaminar fracture toughness when compared with rib, full Milano and uniweave types

268

Sustainable Composites for Lightweight Applications

(a) 500 Interlock

Bending stress (MPa)

400

300

Milano

200

Varied plain knit

100

Rib 0 0

5

10 15 Deflection (mm)

20

25

20

25

(b) 300 Interlock

Bending stress (MPa)

250 Milano 200 Varied plain knit 150

100

50

Rib

0 0

5

10

15

Deflection (mm)

Figure 6.31 Bending or flexural properties of the four samples, showing their responses at both (a) warp and (b) weft directions (Hu et al., 2010).

(Kim et al., 2005). The knitted composite samples recorded nearly 10e20 times interlaminar fracture toughness greater than the uniweave counterpart (Fig. 6.33). The mechanical performances (impact, tensile and flexural strength and modulus) of knitted composite materials can be traced to their laminate fibre lay-up systems, especially in sandwich laminates, as recently investigated on by Skrifvars et al. (2019), using uniaxial warp-knitted/non-woven viscose fabrics/unsaturated polyester sandwich biocomposites.

Sustainable composites and techniques for property enhancement

269

(a)

Load (KN)

8

6

4

Interlock Varied plain knit

2

Milano Rib

0 0

2

4

6

8 10 12 Displacement (mm)

14

16

18

20

(b) Spectific energy absorbed (KJ/m3)

8000

Interlock Varied plain knit Milano

7000 6000

Rib

5000 4000 3000 2000 1000 0

0

3

6

9

12

15

18

Displacement (mm)

Figure 6.32 Impact properties of the four samples, showing their (a) load-displacement and (b) specific energy absorbed-displacement plots, taken at the highest speed of 4.43 m/s (Hu et al., 2010).

6.4.5

Weaving

Weaving is a process of producing fabric by interlacing the fibre elements. Weaving could take various patters, as illustrated in Fig. 6.34. An innovative multi-axis 3D weaving technique for composites has been studied, where multi-axis 3D woven fabrics possessed several layers without delamination defect, as a result of the presence of both z-fibres and in-plane behaviours and Hand bias yarn layers (Bilisik, 2012). A typical illustrative 3D woven pattern is depicted in Fig. 6.35.

270

Sustainable Composites for Lightweight Applications

8 Interlock Rib Full milano Uniweave

GIC[kJ/m2]

6

4

2

0 0

25

50 75 Crack Length,a [mm]

100

Figure 6.33 Interlaminar fracture toughness responses of the four types of knitted composites (Kim et al., 2005).

There are various weaving machines available today to produce different plain weave fabrics of various yarns and their densities in the warp and weft directions. A 3D weaving (Fig. 6.36) is generally carried out on sewing machines and computercontrolled looms to achieve consistent properties to an extent. Weaving produces the least damage when compared with stitching and pinning, though it depends on

(a)

(b)

(c)

(d)

(e)

(f)

Figure 6.34 Various fabric weaving, showing (a), (b), (e) plain, (d) unidirectional, (c) and (f) twill 2/2 (Karahan and Karahan, 2014).

Sustainable composites and techniques for property enhancement

Weft yarns

271

Z yarns

Warp yarns

Figure 6.35 A typical illustrative 3D woven pattern (Yao et al., 2012).

the processing conditions and fibre arrangement (Mouritz and Cox, 2010). However, one of the frequent setbacks of weaving is the crimping defect, as illustrated in Fig. 6.37. Furthermore, the ineplain fibres constrained to tows oriented at 0 and 90 in 3D woven biocomposites are commonly referred to as warp and weft directions, respectively, as illustrated in Fig. 6.38(a). Additionally, yarn can be stitched and combined with another knitted yarn to form a biaxial weft knitted fabric for biocomposite materials of highly improved properties, as shown in Fig. 6.38.

(a)

Woven z-binder

(b)

Woven z-binder

Figure 6.36 A typical 3D (a) orthogonal and (b) angle interlocked woven composites (Mouritz and Cox, 2010).

272

Sustainable Composites for Lightweight Applications

Filler

Stuffers

Warp weaver

Figure 6.37 Crimping defect in interlock woven fabric (Mouritz and Cox, 2010).

(a)

(b)

(c)

Warp (wale) (0º) Weft (course) (90º) Stitch yarn

Weft yarn

Biaxial weft knitted fabric

Warp yarns

(d)

Stitch yarn

Weft yarn

Warp yarn

5 mm Biaxial weft knitted fabric

Figure 6.38 Representative illustration of the structure of a biaxial weft knitted fabric, showing its (a) warp and weft yarns, (b) stitch yarn with knitted fabric, (c) schema and (d) micrograph biaxial weft knitted fabrics (Demircan et al., 2015).

Sustainable composites and techniques for property enhancement

273

More also, the influence of fibre weight contents of kenaf/Kevlar fibre on the mechanical behaviours of woven hybrid laminated biocomposites has been reported (Yahaya et al., 2014). The mechanical performances of the hybrid composites were differently influenced by the fibre weight contents, with the optimum ratio of 78/22, when compared with other hybrid counterparts and tested under impact loading. This indicated the combined benefits of weaving and hybridisation techniques on biocomposite materials. In moving forward, the influence of hemp/PLA yarn laminate, 8harness satin hemp/PLA and basket hemp/PLA woven fabric structures on mechanical (tensile, flexural and impact strengths) and water absorption performance of hemp unidirectional woven fabric-reinforced polylactic acid biocomposites has been investigated (Baghaei et al., 2015). The optimal mechanical behaviours were obtained from the satin fabric pattern, with expressively minimum porosities and fibre misalignment. Both weave patters exhibited higher storage modulus and water absorption than the yarn laminate biocomposite and neat PLA matrix, respectively.

6.4.6

Braiding

Braiding is a smart manufacturing technique for tubular structures, including hollow shafts, tube and struts (Fig. 6.39). It is similar to hybrid filament winding and weaving methods of manufacturing biocomposite components. Braided fabric or reinforced composite components are usually characterised with high quality, good dimensional stability, low cost, better fatigue life, efficient fibre orientation, delamination and impact resistance. Process parameters such as braid angle is an important factor that determines the quality of braided biocomposite materials. For instance, it was reported that an increase in braid angle of a natural hemp fibre/bio-matrix (ecopoxy) tubular-braided biocomposites caused an increase in its void content (Bruni-Bossio et al., 2019). Therefore, smaller braid angles of 40 (with a void fraction of 1.5%) and 50 were preferred to 60 . However, the total void content was decreased with the application of vacuum-bagging. The presence of voids reduces the mechanical properties (elastic moduli and ultimate strengths) of braided cellulose-reinforced composites (Qamhia et al., 2015), as well as tensile strength and ductility of finite element modelled 3D braided composites (Dong and Gong, 2018). These results have been similarly reported (Dong and Huo, 2018). Conversely, highest bending stiffness and impact resistance has been observed with the highest braid angle of 45 circular braided composites when compared with 15 and 30 samples (Zhou et al., 2015). In addition, it has been observed that the braided thermoplastic matrix composites exhibited a greater impact resistance when compared with the braided thermosetting matrix based composite (Lei et al., 2017). This established the need for the right choice of matrix in a braided composite material, in addition to a smaller braided angle.

274

Sustainable Composites for Lightweight Applications

Figure 6.39 Braiding technology (simulation) in composite manufacturing. Sun, X., Kawashita, L.F., Wollmann, T., Spitzer, S., Langkamp, A., Gude, M., 2018. Experimental and numerical studies on the braiding of carbon fibres over structured end-fittings for the design and manufacture of high performance hybrid shafts. Production Engineering 12, 215e228.

6.4.7

Tufting

The process of tufting is schematically presented in Fig. 6.40. It is a one-side sewing method, whereby a thread binder is inserted through a ply stack, using a needle. This property enhanced technique can be totally automated for ease of change in parameters, using robots (Dell’Anno et al., 2016). It can probably be referred to as an advancement of frequently used stitching. Few studies on the application of tufting techniques are available. A reduced delamination and enhanced energy absorption have been achieved under impact loading, although they depended on the tufting pitch (Deconinck et al., 2014).

Needle Ply stack Tufting

Loop Figure 6.40 Schematic description of a tufting method (Deconinck et al., 2014).

Sustainable composites and techniques for property enhancement

6.5

275

Summary and further evaluation

Sustainability is an indispensable concept that must be considered at every stage of manufacturing processes through any technique. Both production processes and biocomposite products are expected to be sustainable in order to keep meeting the insatiable human needs. There is a high possibility that biocomposite sustainability will definitely increase with advancement in material engineering, especially in composite technology. The inherent properties of natural fibres and matrices determine their mechanical performances and other properties. Therefore, the need for fibre treatment is very germane. Natural or bio-fibres treatments are possible chemically, physically, additively and biologically. Among these treatment methods, both chemical and physical techniques are widely employed. In addition, several approaches have been adopted to enhance the overall property of several biocomposites, including hybrid types. These techniques include, but are not limited to, hybridisation, pinning, stitching, weaving, knitting, braiding and tufting. The main benefit of these approaches include improvement of the mechanical behaviours (tensile, flexural, impact, interlaminar fracture toughness), among other properties of biocomposite materials. The application of these technologies are prone to change and improvement as new biocomposite materials are designed and developed. Therefore, the quest to improve on these manufacturing techniques remains a continuous task in the field of composite engineering.

6.6

Conclusion

The sustainable composites and techniques for their properties’ enhancement have been extensively elucidated in this chapter. Comparatively, it was evident that biocomposites support the concept of sustainability more than conventional composites, considering the five stages of product environmental life-cycle analysis, especially material, production and disposal/recovery stages. Due to the inherent properties of natural fibres and matrices, there are needs for fibre treatments towards a better reinforcing performance within a particular biocomposite system. Among all the methods of fibre treatments, chemical (acetylation) treatment has been generally accepted for the treatment of natural fibres, because it supports a continuous process and reduces the moisture uptake greatly, not at the detriment of other important properties of the fibres (especially stiffness and strength). Other aforementioned property enhancement techniques could support the functionality and durability of biocomposite parts. However, there is no specific technique that could improve all the properties of all types of biocomposites; instead, a combination of methods may serve better. Summarily, it is evident that enhancement of performance behaviours of biocomposite materials depends on the nature of the reinforcements (fibres/fillers) and matrices (binders), manufacturing techniques/approaches, fibre treatments, process parameters used and areas of application.

276

Sustainable Composites for Lightweight Applications

References Abdul Khalil, H.P.S., Ismail, H., 2000. Effect of acetylation and coupling agent treatments upon biological degradation of plant fibre reinforced polyester composites. Polym. Test. 20, 65e75. Aji, I.S., Zainudin, E.S., Abdan, K., Sapuan, S.M., Khairul, M.D., 2012. Mechanical properties and water absorption behavior of hybridized kenaf/pineapple leaf fibrereinforced highdensity polyethylene composite. J. Compos. Mater. 47 (8), 979e990. Alimuzzaman, S., Gong, R.H., Akonda, M., 2014. Biodegradability of non-woven flax fiber reinforced poly lactic acid bio-composites. Polym. Compos. 35 (11), 2094e2102. Almansour, F.A., Dhakal, H.N., Zhang, Z.Y., Ghasemnejad, H., 2015. Effect of hybridization on the Mode II fracture toughness properties of flax/vinyl ester composites. Polym. Compos. 38 (8), 732e740. Almansour, F.A., Dhakal, H.N., Zhang, Z.Y., 2017. Effect of water absorption on Mode I interlaminar fracture toughness of flax/basalt reinforced vinyl ester hybrid composites. Compos. Struct. 168, 813e825. Almansour, F.A., Dhakal, H.N., Zhang, Z.Y., 2018. Investigation into Mode II interlaminar fracture toughness characteristics of flax/basalt reinforced vinyl ester hybrid composites. Compos. Sci. Technol. 154, 117e127. Aranberri-Askargorta, I., Lampke, T., Bismarck, A., 2003. Wetting behavior offlax fibers as reinforcement for polypropylene. J. Colloid Interface Sci. 263, 580e589. Avella, M., Bogoeva-Gaceva, G., Buzarovska, A., Errico, M.E., Gentile, G., Grozdanov, A., 2008. Poly (lactic acid)-based biocomposites reinforced with kenaf fibers. J. Appl. Polym. Sci. 108 (6), 3542e3551. Bachtiar, D., Sapuan, S.M., Hamdan, M.M., 2008. The effect of alkaline treatment on tensile properties of sugar palm fibre reinforced epoxy composites. Mater. Des. 29 (7), 1285e1290. Baghaei, B., Skrifvars, M., Berglin, L.M., 2015. Characterization of thermoplastic natural fibre composites made from woven hybrid yarn prepregs with different weave pattern. Compos. Appl. Sci. Manuf. 76, 154e161. Beckermann, G.W., Pickering, K.L., 2008. Engineering and evaluation of hemp fibre reinforced polypropylene composites: fibre treatment and matrix modification. Compos. Appl. Sci. Manuf. 39 (6), 979e988. Beg, M.D.H., Pickering, K.L., 2008. Mechanical performance of Kraft fibre reinforced polypropylene composites: influence of fibre length, fibre beating and hygrothermal ageing. Compos. Appl. Sci. Manuf. 39 (11), 1748e1755. Beier, U., Fischer, F., Sandler, J.K.W., Altst€adt, V., Weimer, C., Buchs, W., 2007. Mechanical performance of carbon fibre-reinforced composites based on stitched preforms. Compos. Appl. Sci. Manuf. 38 (7), 1655e1663. Bera, M., Alagirusamy, R., Das, A., 2010. A study on interfacial properties of jute-PP composites. J. Reinforc. Plast. Compos. 29 (20), 3155e3161. Bilisik, K., 2012. Multiaxis three-dimensional weaving for composites: a review. Textil. Res. J. 82 (7), 725e743. Bismarck, A., Aranberri-Askargorta, I., Springer, J., Lampke, T., Wielage, B., Stamboulis, A., Shenderovich, I., Limbach, H.H., 2002. Surface characterization of flax, hemp and cellulose fibers, surface properties and the water uptake behavior. Polym. Compos. 23, 872e894.

Sustainable composites and techniques for property enhancement

277

Blacklock, M., Joosten, M.W., Pingkarawat, K., Mouritz, A.P., 2016. Prediction of mode I delamination resistance of z-pinned laminates using the embedded finite element technique. Compos. Appl. Sci. Manuf. 91, 283e291. Bledzki, A.K., Gassan, J., Mildner, I., 2001. Transcrystallization of polypropylene on different modified jute fibers. Compos. Interfac. 8, 443e452. Bledzki, A.K., Mamun, A.A., Jaszkiewicz, A., Erdmann, K., 2010. Polypropylene composites with enzyme modified abaca fibre. Compos. Sci. Technol. 70 (5), 854e860. Bledzki, A.K., Mamun, A.A., Lucka, M., Gutowsk, V.S., 2008. The effects of acetylation on properties of flax fibre and its polypropylene composites. Express Polym. Lett. 2 (6), 413e422. Bos, H.L., M€ussig, J., van den Oever, M.J.A., 2006. Mechanical properties of short-flax-fibre reinforced compounds. Compos. Part A Appl. Sci. Manuf. 37, 1591e1604. Bos, H.L., Van den Oever, M.J.A., Peters, O., 2002. Tensile and compressive properties of flax fibres for natural fibre reinforced composites. J. Mater. Sci. 37 (8), 1683e1692. Bruni-Bossio, B.M., Melenka, G.W., Ayranci, C., Carey, J.P., 2019. Micro-computed tomography analysis of natural fiber and bio-matrix tubular-braided composites. J. Compos. Mater. 53 (28e30), 4003e4013. Cantero, G., Arbelaiz, A., Llano-Ponte, R., Mondragon, I., 2003. Effects of fibre treatment on wettability and mechanical behavior of flax/polypropylene composites. Compos. Sci. Technol. 63, 1247e1254. Cao, Y., Sakamoto, S., Goda, K., 2007. Effects of heat and alkali treatments on mechanical properties of kenaf fibers. In: Proc. 16th International Conference on Composite Materials, 8e13 July, 2007, Kyoto, Japan. Cartié, D.D.R., Laffaille, J.M., Partridge, I.K., Brunner, A.J., 2009. Fatigue delamination behaviour of unidirectional carbon fibre/epoxy laminates reinforced by Z-Fiber® pinning. Eng. Fract. Mech. 76 (18), 2834e2845. Cartié, D.D.R., Troulis, M., Partridge, I.K., 2006. Delamination of z-pinned carbon fibre reinforced laminates. Compos. Sci. Technol. 66 (6), 55e61. Chan, P., Mouritz, A.P., Cox, B.N., 2006. Properties and failure mechanisms of z-pinned laminates in monotonic and cyclic tension. Compos. Appl. Sci. Manuf. 37 (10), 1501e1513. Deconinck, P., Capelle, J., Bouchart, V., Chevrier, P., Ravailler, F., 2014. Delamination propagation analysis in tufted carbon fibre-reinforced plastic composites subjected to highvelocity impact. J. Reinforc. Plast. Compos. 33 (14), 1353e1363. Dell’Anno, G., Treiber, J.W.G., Partridge, I.K., 2016. Manufacturing of composite parts reinforced through-thickness by tufting. Robot. Comput. Integrated Manuf. 37, 262e272. Demircan, O., Ashibe, S., Kosui, T., Nakai, A., 2015. Effect of various knitting techniques on mechanical properties of biaxial weft-knitted thermoplastic composites. J. Thermoplast. Compos. Mater. 28 (6), 896e910. Dhakal, H.N., Zhang, Z.Y., Bennett, N., 2012. Influence of fibre treatment and glass fibre hybridisation on thermal degradation and surface energy characteristics of hemp/unsaturated polyester composites. Compos. Part B 43, 2757e2761. Dhakal, H.N., Sarasini, F., Santulli, C., Tirill
o, J., Zhang, Z., Arumugam, V., 2015. Effect of basalt fibre hybridisation on post-impact mechanical behaviour of hemp fibre reinforced composites. Compos. Appl. Sci. Manuf. 75, 54e67. Dong, J., Gong, Y., 2018. Influence of void defects on progressive tensile damage of threedimensional braided composites. J. Compos. Mater. 52 (15), 2033e2045.

278

Sustainable Composites for Lightweight Applications

Dong, J., Huo, N., 2018. Progressive tensile damage simulation and strength analysis of threedimensional braided composites based on three unit-cells models. J. Compos. Mater. 52 (15), 2017e2031. Dransfield, K.A., Jain, L.K., Mai, Y.W., 1998. On the effects of stitching in CFRPs-I. Mode I delamination toughness. Compos. Sci. Technol. 58 (6), 815e827. Faruk, O., Bledzki, A.K., Fink, H., Sain, M., 2012. Biocomposites reinforced with natural fibers: 2000e2010. Prog. Polym. Sci. 37 (11), 1552e1596. Faruk, O., Bledzki, A.K., Fink, H.P., Sain, M., 2014. Progress report on natural fiber reinforced composites. Macromol. Mater. Eng. 299 (1), 9e26. Fiore, V., Scalici, T., Calabrese, L., Valenza, A., Proverbio, E., 2016. Effect of external basalt layers on durability behaviour of flax reinforced composites. Compos. B Eng. 84, 258e265. Fiore, V., Scalici, T., Sarasini, F., Tirillo, J., Calabrese, L., 2017. Salt-fog spray aging of jutebasalt reinforced hybrid structures: flexural and low velocity impact response. Compos. B Eng. 116, 99e112. Francesconi, L., Aymerich, F., 2018. Effect of Z-pinning on the impact resistance of composite laminates with different layups. Compos. Appl. Sci. Manuf. 114, 136e148. Franco-Marqu
es, E., Méndez, J.A., P
elach, M.A., Vilaseca, F., Bayer, J., Mutjé, P., 2011. Influence of coupling agents in the preparation of polypropylene composites reinforced with recycled fibers. Chem. Eng. J. 166 (3), 1170e1178. Fuqua, M.A., Ulven, C.A., 2008. Characterization of polypropylene/corn fiber composites with maleic anhydride grafted polypropylene. J. Biobased Mater. Bioenergy 2 (3), 258e263. Gassan, J., Bledzki, A.K., 2000. Possibilities to improve the properties of natural fiber reinforced plastics by fiber modificationejute polypropylene composites. Appl. Compos. Mater. 7, 373e385. Gassan, J., Gutowski, V.S., 2000. Effects of corona discharge and UV treatment on the properties of jute-fibre epoxy composites. Compos. Sci. & Technol. 60 (15), 2857e2863. Goda, K., Sreekala, M., Gomes, A., Kaji, T., Ohgi, J., 2006. Improvement of plant based natural fibers for toughening green compositesd effect of load application during mercerization of ramie fibers. Compos. Appl. Sci. Manuf. 37 (12), 2213e2220. Gomes, A., Matsuo, T., Goda, K., Ohgi, J., 2007. Development and effect of alkali treatment on tensile properties of curaua fiber green composites. Compos. Appl. Sci. Manuf. 38 (8), 1811e1820. Goutianos, S., Peijs, T., Nystrom, B., Skrifvars, M., 2006. Development of flax fibre based textile reinforcements for composite applications. Appl. Compos. Mater. 13 (4), 199e215. Hill, C.A.S., Khalil, H.P.S., Hale, M.D., 1998. A study of the potential of acetylation to improve the properties of plant fibres. Ind. Crop. Prod. 8 (1), 53e63. Hoffmann, J., Brast, A., Scharr, G., 2019. Z-pin insertion process for through-thickness reinforced thermoplastic composites. J. Compos. Mater. 53 (2), 173e181. Hu, H., Zhang, M., Fangueiro, R., Araujo, M.D., 2010. Mechanical properties of composite materials made of 3D stitched woven-knitted preforms. J. Compos. Mater. 44 (14), 1753e1767. Huber, T., Biedermann, U., Muessig, J., 2010. Enhancing the fibre-matrix adhesion of natural fibre reinforced polypropylene by electron radiation analyzed with the single fibre fragmentation test. Compos. Interfac. 17 (4), 371e381. Ibrahim, N.A., Hadithon, K.A., 2010. Effect of fiber treatment on mechanical properties of kenaf fiber-ecoflex composites. J. Reinforc. Plast. Compos. 29 (14), 2192e2198. Idicula, M., Joseph, K., Thomas, S., 2010. Mechanical performance of short banana/sisa hybrid fiber reinforced polyester composites. J. Reinforc. Plast. Compos. 29 (1), 12e29.

Sustainable composites and techniques for property enhancement

279

Islam, M.S., Pickering, K.L., Foreman, N.J., 2010. Influence of alkali treatment on the interfacial and physico-mechanical properties of industrial hemp fibre reinforced polylactic acid composites. Compos. Appl. Sci. Manuf. 41 (5), 596e603. Ismail, H., Abdul Khalil, H.P.S., 2000. The effects of partial replacement of oil palm wood flour by silica and silane coupling agent on properties of natural rubber compounds. Polym. Test. 20, 33e41. Ismail, S.O., Dhakal, H.N., Dimla, E., Beaugrand, J., Popov, I., 2016. Effects of drilling parameters and aspect ratios on delamination and surface roughness of lignocellulosic HFRP composite laminates. J. Appl. Polym. Sci. 133 (7), 1e8. Jain, L.K., Dransfield, K.A., Ma, Y.W., 1998. On the effects of stitching in CFRPs-II. mode II delamination toughness. Compos. Sci. Technol. 58 (6), 829e837. Jain, L.K., Mai, Y.W., 1994. Analysis of stitched laminated ENF specimens for interlaminar mode II fracture toughness. Int. J. Fract. 68 (3), 219e244. Joseph, P.V., Joseph, K., Thomas, S., Pillai, C.K.S., Prasad, V.S., Groeninckx, G., Sarkissova, M., 2003. The thermal and crystallisation studies of short sisal fibre reinforced polypropylene composites. Compos. Appl. Sci. Manuf. 34, 253e266. Kabir, M.M., Wang, H., Lau, K.T., Cardona, F., Aravinthan, T., 2011. Mechanical properties of chemically-treated hemp fibre reinforced sandwich composites. Compos. B Eng. 43 (2), 159e169. Kang, J.T., Park, S.H., Kim, S.H., 2013. Improvement in the adhesion of bamboo fiber reinforced poly-lactide composites. J. Compos. Mater. 48 (21), 2567e2577. Karahan, M., Karahan, N., 2014. Influence of weaving structure and hybridization on the tensile properties of woven carbon-epoxy composites. J. Reinforc. Plast. Compos. 33 (2), 212e222. Kazayawoko, M., Balatinecz, J.J., 1997. Adhesion mechanisms in wood fiber polypropylene composites. In: Fourth International Conference on Wood Fiber-Plastic Composites. Forest Products Research Society, Madison, pp. 81e93. Khalil, H., Ismail, H., Rozman, H.D., Ahmad, M.N., 2001. The effect of acetylation on interfacial shear strength between plant fibres and various matrices. Eur. Polym. J. 37 (5), 1037e1045. Khan, M.Z.S., Mouritz, A.P., 1996. Fatigue behaviour of stitched grp laminates. Compos. Sci. Technol. 56 (6), 695e701. Kim, K.Y., Curiskis, J.I., Ye, L., Fu, S.Y., 2005. Mode-I interlaminar fracture behaviour of weftknitted fabric reinforced composites. Compos. Appl. Sci. Manuf. 36, 954e964. Ku, H., Wang, H., Pattarachaiyakoop, N., Trada, M., 2011. A review on the tensile properties of natural fiber reinforced polymer composites. Compos. B Eng. 42 (4), 856e873. Kulshrestha, P., 2020. Development of Composite Materials Using Braiding Technology. Available at: https://www.technicaltextile.net/articles/development-of-composite-materialsusing-braiding-technology-7848 (Accessed 04.03.2020). Kumar, T.S.M., Senthilkumar, K., Chandrasekar, M., Tengsuthiwat, J., Rajini, N., Siengchin, S., Ismail, S.O., 2019. Investigation into mechanical, absorption and swelling behaviour of hemp/sisal fibre reinforced bioepoxy hybrid composites: effects of stacking sequences. Int. J. Biol. Macromol. 140, 637e646. Ladani, R.B., Ravindran, A.R., Wu, S., Pingkarawat, K., Kinloch, A.J., Mouritz, A.P., Ritchie, R.O., Wang, C.H., 2016. Multi-scale toughening of fibre composites using carbon nanofibres and z-pins. Compos. Sci. Technol. 131, 98e109. Lee, S.H., Wang, S., 2006. Biodegradable polymers/bamboo fiber bio-composite with bio-based coupling agent. Compos. Appl. Sci. Manuf. 37 (1), 80e91.

280

Sustainable Composites for Lightweight Applications

Lei, X., Hayat, K., Ha, S.K., 2017. A multi-scale approach to predict the impact resistance of braided composites using progress failure. Adv. Compos. Lett. 26 (1), 24e29. Li, X., Zhang, J., He, J., Reddy, D.J.P., Rajulu, A.V., 2010. Tensile properties of hildegardia fibers reinforced polypropylene biocomposites. J. Compos. Mater. 44, 1681e1688. Liu, H., Wu, Q., Zhang, Q., 2009. Preparation and properties of banana fiber-reinforced composites based on high density polyethylene (HDPE)/Nylon-6 blends. Bioresour. Technol. 100, 6088e6097. Liu, W., Mohanty, A.K., Askeland, P., Drzal, L.T., Misra, M., 2004. Influence of fibre surface treatment on properties of Indian grass fibre reinforcement soy protein based biocomposites. Polymer 45 (22), 7589e7596. Marais, S., Gouanve, F., Bonnesoeur, A., Grenet, J., Poncin-Epaillard, F., Morvan, C., Metayer, M., 2005. Unsaturated polyester composites reinforced with flax fibers: effect of cold plasma and autoclave treatments on mechanical and permeation properties. Compos. Appl. Sci. Manuf. 36, 975e986. Martin, A.R., Manolache, S., Mattoso, L.H.C., Rowell, R.M., Dense, F., 2000. Plasma modification of sisal and high-density polyethylene composites: effect on mechanical properties. In: Natural Polymers and Composites Proceedings, pp. 431e436. Matthews, F.L., Rawlings, R.D., 1999. Composite Materials: Engineering and Science. Woodhead Publishing, Cambridge, England. Mishra, S., Naik, J.B., Patil, Y.P., 2000. The compatibilising effect of maleic anhydride on swelling and mechanical properties of plant fiber- reinforced novolac composites. Compos. Sci. Technol. 60, 1729e1735. Mohanty, A.K., Misra, M., Drzal, L.T. (Eds.), 2005. Natural Fibres, Biopolymers and BioComposites. CRC Press Taylor and Francis Group, New York. Mohanty, S., Nayak, S.K., Verma, S.K., Tripathy, S.S., 2004. Effect of MAPP as a coupling agent on the performance of juteePP composites. J. Reinforc. Plast. Compos. 23, 625e637. Mouritz, A.P., 2007. Review of z-pinned composite laminates. Compos. Appl. Sci. Manuf. 38 (12), 2383e2397. Mouritz, A.P., Bannister, M.K., Falzon, P.J., Leong, K.H., 1999. Review of application for advanced three-dimensional fibre textile composites. Compos. Appl. Sci. Manuf. 30 (12), 1445e1461. Mouritz, A.P., Cox, B.N., 2000. A mechanistic approach to the properties of stitched laminates. Compos. Appl. Sci. Manuf. 31 (1), 1e27. Mouritz, A.P., Cox, B.N., 2010. A mechanistic interpretation of the comparative in-plane mechanical properties of 3D woven, stitched and pinned composites. Compos. Appl. Sci. Manuf. 41 (6), 709e728. Mouritz, A.P., Koh, T.M., 2014. Re-evaluation of mode I bridging traction modelling for zpinned laminates based on experimental analysis. Compos. B Eng. 56, 797e807. Mouritz, A.P., Leong, K.H., Herszberg, I., 1997. A review of the effect of stitching on the inplane mechanical properties of fibre-reinforced polymer composites. Compos. Appl. Sci. Manuf. 28 (12), 979e991. Pegorin, F., Pingkarawat, K., Mouritz, A.O., 2018. Electrical-based delamination crack monitoring in composites using z-pins. Compos. Appl. Sci. Manuf. 104, 120e128. Pegorin, F., Pingkarawat, K., Mouritz, A.P., 2015. Comparative study of the mode I and mode II delamination fatigue properties of z-pinned aircraft composites. Mater. Des. 65, 139e146. Pervaiz, M., Sain, M.M., 2003. Carbon storage potential in natural fiber composite. Resour. Conserv. Recycl. 39, 325e340.

Sustainable composites and techniques for property enhancement

281

Pickering, K., 2008. Properties and Performance of Natural-Fibre Composites. Woodhead Publishing, Cambridge, England. Pickering, K., Beckermann, G., Alam, S., Foreman, N., 2007. Optimising industrial hemp fibre for composites. Compos. Appl. Sci. Manuf. 38 (2), 461e468. Pickering, K.L., Abdalla, A., Ji, C., McDonald, A.G., Franich, R.A., 2003. The effect of silane coupling agents on radiata pine fibre for use in thermoplastic matrix composites. Compos. Appl. Sci. Manuf. 34 (10), 915e926. Pickering, K.L., Efendy, M.G.A., Le, T.M., 2016. A review of recent developments in natural fibre composites and their mechanical performance. Compos. Appl. Sci. Manuf. 83, 98e112. Pingkarawat, K., Mouritz, A.P., 2014. Improving the mode I delamination fatigue resistance of composites using z-pins. Compos. Sci. Technol. 92, 70e76. Pingkarawat, K., Mouritz, A.P., 2016. Comparative study of metal and composite z-pins for delamination fracture and fatigue strengthening of composites. Eng. Fract. Mech. 154, 180e190. Pizzi, A., Kueny, R., Lecoanet, F., Massetau, B., Carpentier, D., Krebs, A., Loiseau, F., Molina, S., Ragoubi, M., 2009. High resin content natural matrixenatural fibre biocomposites. Ind. Crop. Prod. 30, 235e240. Pommet, M., Juntaro, J., Heng, J.Y.Y., Mantalaris, A., Lee, A.F., Wilson, K., Kalinka, G., Shaffer, M.S.P., Bismarck, A., 2008. Surface modification of natural fibers using bacteria: depositing bacterial cellulose onto natural fibers to create hierarchical fiber reinforced nanocomposites. Biomacromolecules 9, 1643e1651. Pothan, L.A., Thomas, S., 2003. Polarity parameters and dynamic mechanical behavior of chemically modified banana fiber reinforced polyester composites. Compos. Sci. Technol. 63, 1231e1240. Qamhia, I.I., Shams, S.S., El-Hajjar, R.F., 2015. Quasi-isotropic triaxially braided cellulosereinforced composites. Mech. Adv. Mater. Struct. 22 (12), 988e995. Qinghua, S., Yong, L., Junwei, Q., Liwei, W., Jun, X., Jinhui, N., 2014. Study on an automatic multi-pin insertion system for preparing Z-pin composite laminates. Chin. J. Aeronaut. 27 (2), 413e419. Quarshie, R., Carruthers, J., 2014. Technological Overview: Biocomposites. Net Composites, pp. 1e40. Available at: https://netcomposites.com/media/1211/biocomposites-guide.pdf (Accessed 05.08.2019). Rachini, A., Le Troedec, M., Peyratout, C., Smith, A., 2012. Chemical modification of hemp fibers by silane coupling agents. J. Appl. Polym. Sci. 123 (1), 601e610. Ragoubi, M., Bienaimé, D., Molina, S., George, B., Merlin, A., 2010. Impact of corona treated hemp fibres onto mechanical properties of polypropylene composites made thereof. Ind. Crop. Prod. 31 (2), 344e349. Ranatunga, V., Clay, S.B., 2012. Cohesive modeling of damage growth in z-pinned laminates under mode-I loading. J. Compos. Mater. 47 (26), 3269e3283. Ravandi, M., Teo, W.S., Tran, L.Q.N., Yong, M.S., Tay, T.E., 2017. Low velocity impact performance of stitched flax/epoxy composite laminates. Compos. B Eng. 117, 89e100. Ravandi, M., Teo, W.S., Tran, L.Q., Yong, M.S., Tay, T.E., 2016. The effects of through-thethickness stitching on the mode I interlaminar fracture toughness of flax/epoxy composite laminates. Mater. Des. 109, 659e669. Reddy, T.R.K., Kim, H.J., Park, J.W., 2016. Renewable biocomposite properties and their applications. In: Poletto, M. (Ed.), Composites from Renewable and Sustainable Materials, Chapter 10, pp. 177e197.

282

Sustainable Composites for Lightweight Applications

Rong, M.Z., Zhang, M.Q., Liu, Y., Yang, G.C., Zeng, H.M., 2001. The effect of fiber treatment on the mechanical properties of unidirectional sisal-reinforced epoxy composites. Compos. Sci. Technol. 61 (10), 1437e1447. Roy, K., Debnath, S.C., Tzounis, L., Pongwisuthiruchte, A., Potiyaraj, P., 2020. Effects of various surface treatments on the performance of jute fibres filled natural rubber (NR) composites. Polymers 12 (2), 1e15369. Sarasini, F., Tirill
o, J., Valente, M., Valente, T., Cioffi, S., Iannace, S., Sorrentino, L., 2013. Effect of basalt fiber hybridization on the impact behavior under low impact velocity of glass/basalt woven fabric/epoxy resin composites. Compos. Appl. Sci. Manuf. 47, 109e123. Sawpan, M.A., Pickering, K.L., Fernyhough, A., 2011. Effects of various chemical treatments on the fibre structure and tensile properties of industrial hemp fibres. Compos. Appl. Sci. Manuf. 42, 888e895. Seki, Y., Sarikanat, M., Sever, K., Erden, S., Gulec, H.A., 2010. Effect of the low and radio frequency oxygen plasma treatment of jute fiber on mechanical properties of jute fiber/ polyester composite. Fibers Polym. 11, 1159e1164. Seki, Y., Sever, K., Sarikanat, M., G€uleç, H.A., Tavman, I.H., 2009. The influence of oxygen plasma treatment of jute fibers on mechanical properties of jute fiber reinforced thermoplastic composites. In: 5th International Advanced Technologies Symposium (IATS’09), May 13e15, 2009, Karab€uk, Turkey, pp. 1007e1010. Sim, J., Park, C., Moon, D.Y., 2005. Characteristics of basalt fiber as a strengthening material for concrete structures. Compos. B Eng. 36 (6e7), 504e512. Singh, B., Gupta, M., Verma, A., 1996. Influence of fiber surface treatment on the properties of sisal-polyester composites. Polym. Compos. 17 (6), 910e918. Sinha, E., Panigrahi, S., 2009. Effect of plasma treatment on structure, wettability of jute fiber and flexural strength of its composite. J. Compos. Mater. 43 (17), 1791e1802. Skrifvars, M., Dhakal, H., Zhang, Z., Gentilcore, J., Åkesson, D., 2019. Study on the mechanical properties of unsaturated polyester sandwich biocomposites composed of uniaxial warpknitted and non-woven viscose fabrics. Compos. Appl. Sci. Manuf. 121, 196e206. Stamboulis, A., Baillie, C.A., Garkhail, S.K., van Melick, H.G.H., Peijs, T., 2000. Environmental durability of flax fibres and their composites based on polypropylene matrix. Appl. Compos. Mater. 7 (5e6), 273e294. Tan, K., Watanabe, N., Iwahori, Y., 2010. Stitch fiber comparison for improvement of interlaminar fracture toughness in stitched composites. J. Reinforc. Plast. Compos. 30 (2), 99e109. Tingju, L., Jiang, M., Jiang, Z., Hui, D., Wang, Z., Zhou, Z., 2013. Effect of surface modification of bamboo cellulose fibers on mechanical properties of cellulose/epoxy composites. Compos. B Eng. 51, 28e34. Toffe, G.T., Ismail, S.O., Montalv~ao, D., Knight, J., Ren, G., 2019. A scale-up of energy-cycle analysis on processing non-woven flax/pla tape and triaxial glass fibre fabric for composites. J. Manuf. & Mater. Process. 3 (92), 1e20. Tong, L., Mouritz, A.P., Bannister, M.K., 2002. 3D Fibre Reinforced Polymer Composites, first ed. Elsevier Inc., London. Tserki, V., Zafeiropoulos, N.E., Simon, F., Panayiotou, C., 2005. A study of the effect of acetylation and propionylation surface treatments on natural fibres. Compos. Appl. Sci. Manuf. 36 (8), 1110e1118. Valadez-Gonzalez, A., Cervantes-Uc, J.M., Olayo, R., Herrera-Franco, P.J., 1999. Chemical modification of henequen fibers with an organosilane coupling agent. Compos. B Eng. 30 (3), 321e331.

Sustainable composites and techniques for property enhancement

283

Velmurugan, R., Gupta, N.K., Solaimurugan, S., Elayaperumal, A., 2004. The effect of stitching on FRP cylindrical shells under axial compression. Int. J. Impact Eng. 30 (8e9), 923e938. Velmurugan, R., Solaimurugan, S., 2007. Improvements in mode I interlaminar fracture toughness and in-plane mechanical properties of stitched glass/polyester composites. Compos. Sci. Technol. 67 (1), 61e69. Wallace, S., September 2005. Review of Potential Surface Modifications and Fibre Treatments to Aid Compatibility of Hydrophilic Natural Fibres with Hydrophobic Matrix Polymers, Rapra Report, vol. 45170. Wallenberger, F.T., Weston, N. (Eds.), 2004. Natural Fibers, Plastics and Composites. Kluwer Academic Publishers, Netherlands. Weimer, C., Mitschang, P., 2001. Aspects of stitch formation process on the quality of sewn multi-textile-preforms. Compos. Appl. Sci. Manuf. 32 (10), 1477e1484. Wielage, B., Lampke, T., Utschick, H., Soergel, F., 2003. Processing of natural-fibre reinforced polymers and the resulting dynamic-mechanical properties. J. Mater. Process. Technol. 139, 140e146. Xu, Y., Kawata, S., Hosoi, K., Kawai, T., Kuroda, S., 2009. Thermomechanical properties of the silanized-kenaf/polystyrene composites. Express Polym. Lett. 3, 657e664. Yahaya, R., Sapuan, S.M., Jawaid, M., Leman, Z., Zainudin, E.S., 2014. Mechanical performance of woven kenaf-Kevlar hybrid composite. J. Reinforc. Plast. Compos. 33 (24), 2242e2254. Yan, L., Chouw, N., Jayaraman, K., 2014. Flax fibre and its composites: a review. Compos. B Eng. 56, 296e317. Yang, H.S., Kim, H.J., Park, H.J., Lee, B.J., Hwang, T.S., 2007. Effect of compatibilizing agents on rice-husk flour reinforced polypropylene composites. Compos. Struct. 77, 45e55. Yao, L., Rong, Q., Shan, Z., Qiu, Y., 2012. Static and bending fatigue properties of ultra-thick 3D orthogonal woven composites. J. Compos. Mater. 47 (5), 569e577. Yemm, D.H., 2013. What’s New in Eco-Materials. Available at: (Accessed 05.03.2020). https:// whatisnewinecomaterials.wordpress.com/2013/11/10/connectivity-sustainability/. Yu, T., Jiang, N., Li, Y., 2014. Study on short ramie fiber/poly (lactic acid) composites compatibilized by maleic anhydride. Compos. Appl. Sci. Manuf. 64, 139e146. Yudhanto, A., Lubineau, G., Ventura, I.A., Watanabe, N., Iwahori, Y., Hoshi, H., 2015. Damage characteristics in 3D stitched composites with various stitch parameters under in-plane tension. Compos. Part A Appl. Sci. Manuf. 71, 17e31. Yudhanto, A., Watanabe, N., Iwahori, Y., Hoshi, H., 2014. Effect of stitch density on fatigue characteristics and damage mechanisms of stitched carbon/epoxy composites. Compos. Appl. Sci. Manuf. 60, 52e65. Zhang, A.Y., Liu, H.Y., Mouritz, A.P., Mai, Y.W., 2008. Experimental study and computer simulation of z-pin reinforcement under cycle fatigue. Compos. Appl. Sci. Manuf. 39, 406e414. Zhou, H., Zhang, W., Liu, T., Gu, B., Sun, B., 2015. Finite element analyses on transverse impact behaviors of 3-D circular braided composite tubes with different braiding angles. Compos. Appl. Sci. Manuf. 79, 52e62.

Future outlooks and challenges of sustainable lightweight composites 7.1

7

Journey of composite materials towards sustainability

Reducing the weight of the components is one of the major drives of sustainable lightweight composites under which materials used are expected to provide both weight saving and required functional properties. This is an innovative new class of promising approach, as this provides efficient materials utilisation, the high strength-to-weight ratio in comparison to their metal counterparts leading to increased fuel consumption, benefits to the environment as this can reduce the carbon footprint significantly (Pervaiz et al., 2016; Bourmaud et al., 2018; Dhakal et al., 2013). For the last few decades, the use of advanced lightweight composite materials derived from renewable sources, especially in high-technology industry sectors: automotive, aerospace, marine and construction, has increased significantly in the quest for reducing reliance on heavy metal parts and non-renewable conventional composites moving towards more sustainable composite materials with less carbon footprint attributes. However, there are still some questions whether the lightweight composite materials and lightweighting approach can provide structural properties such as strength, modulus and long-term durability required, and still meet the environmental expectations (Holbery and Houston, 2006; Faruk et al., 2012; Mohanty et al., 2018). The environmental impact such as global warming potential (GWP), ecotoxicity, natural resource depletion and acidification are far less for sustainable lightweight composites compared to glass and carbon fibre-reinforced composites (Wu et al., 2018). If the inherent drawbacks of these composites can be addressed, their acceptance in the key industrial sectors is expected to grow significantly in the future. This can help the composite sector as a whole towards achieving sustainable development aspiration; an economically, environmentally and socially sustainable future. The following sections highlight and discusses the outlook and key challenges for lightweight materials and initiatives.

7.2

Market outlook and supply chain scenario

Public expectations for greener products, government legislation towards less dependence on fossil fuels, due to their attractive attributes such as high strength to weight

Sustainable Composites for Lightweight Applications. https://doi.org/10.1016/B978-0-12-818316-8.00003-7 Copyright © 2021 Elsevier Ltd. All rights reserved.

286

Sustainable Composites for Lightweight Applications

ratio, high resistance to chemical and corrosion together with compliance with the ‘lightweighting’ materials, composites have tremendous scope of application. The emergence of these lightweight composites has motivated high technology sectors, including automotive, marine, aerospace, sports leisure, construction and wind energy moving towards taking benefits from these new class of lightweight, sustainable materials. The composite materials hold a significant market share in today’s economy (Qu et al., 2019). One of the key sectors using lightweight composites is the automotive sector. Major original equipment manufacturers (OEMs), including Toyota, Mercedes, BMW, Audi and Ford, amongst others, use carbon fibre for reinforcements to reduce the overall mass of the vehicles where CO2 emission is expected to be reduced in the overall life span of the products. Similarly, biobased thermoplastic composites have been utilised in this sector (Shah, 2013; Almansour et al., 2017; Ishikawa et al., 2018). According to the report published by IDTechEX (Collins, 2017), the forecast for all synthetic fibres used in composite parts will exceed $9 bn by 2027. It can also be envisaged emerging reinforcements such as flax, hemp and kenaf are gaining demand and popularity, due to their green credentials. However, due to the lack of an established sustainable supply chain network, it is accepted that there is not enough supply to meet emerging demands.

7.3 7.3.1

Challenges of achieving properties for lightweight applications Materials and manufacturing process

Despite their attractive attributes, resolving their weaknesses, such as poor recyclability after their end-of-life, especially for carbon and glass fibres, low mechanical performance, especially for natural fibre-reinforced lightweight composites and improving their long-term durability under various harsh environments remain as challenges to make these materials and lightweight approach full commercially viable. Moreover, some of the materials, such as natural fibres used in lightweight applications have been considered difficult to maintain their supply chain (Dhakal et al., 2007; Rohan et al., 2015). In order to achieve high quality, with required mechanical properties at a low cost is a big challenge in achieving lightweight aspirations. Processing automation using latest technology is important. However, the sustainable lightweight reinforcements such as natural fibres require a different set of processing parameters than the glass and carbon fibres. Reducing the cycle time without reducing the impregnation ability and mechanical properties are challenging tasks. It is important that instead of creating waste materials at the end of life, recycling and re-using materials instead of using virgin materials. When new materials are developed, it is important that material properties can be analysed using certain design criteria and materials parameters. The designers consider

Future outlooks and challenges of sustainable lightweight composites

287

some of the drawbacks of natural fibre composites such as quality variations, susceptibility to moisture absorption, non-uniform diameter and microbial growth as weaknesses and disadvantages for modelling. Additionally, safety issues relating to matrices, processing techniques used and the costs are concerned challenges (Takahashi et al., 2012).

7.3.2

Recyclability and end-of- life option

When selecting appropriate matrices and reinforcements for lightweight composites, the recyclability aspect is of prime concern. Lightweighting materials such as glass and carbon fibres are attractive materials due to their many good attributes. However, the poor recyclability and recovery aspect poses a significant challenge. This is one of the biggest issues that OEMs in the transport sector face. This issue becomes even greater when using thermoset matrices. Several studies suggest that the demand for lightweight materials are expected to rise due to a rise in the demand for environmentally friendly, low emission vehicles. There is a strong correlation between vehicle weight and CO2 emission per km. Many companies in the transport sector have put their priorities to reduce CO2 emissions. There is a clear drive to reduce the vehicle weight, which can be achieved by using lighter materials, increase of the functionalities of the parts and improvement in the car system and powertrain design. Therefore, optimising performance with maintaining the environmentally sustainability is of a significant drive. More importantly, after the end-of-life consideration is crucial, as these materials cannot be landfilled as these materials leach toxic substances. Hence, end-of-life (EoL) management of lightweight composites, especially glass and carbon fibre-reinforced composites, is of a huge challenge. Additionally, an in-depth understanding of lightweight materials and their environmental impact arising from important stages of their life cycle such as raw materials extraction, processing, production, use and disposal is very important. It is vital that there is a credible life cycle analysis (LCA) technique used in order to determine the environmental impact of various products and processes.

7.3.3

Long-term durability

Understanding long-term durability in different harsh environments is still a challenge for natural fibre-reinforced lightweight composites and biobased composites (Dhakal et al., 2014). Overcoming some key obstacles such as inadequate impact toughness, reduced long-term stability in certain environmental conditions (temperature, humidity and UV radiation) is critical (Wang et al., 2006). Moreover, applications of these materials in automotive, marine, aerospace, wind turbine require a thorough understanding of how these materials behave under different loading conditions (Pickering et al., 2016). The relationship between microstructure and various mechanical properties by using suitable characterisation techniques is crucial. Achieving acceptable mechanical properties, low density, environmentally sustainable materials coupled with long-term structural durability, remain key challenges (Das et al., 2017; Dhakal et al., 2018).

288

7.4

Sustainable Composites for Lightweight Applications

Future outlook

The prospective of use and continuous development of lightweight and ultralightweight composite materials present an exciting potential of meeting sustainable materials and low carbon economy (LCE). To fully realise the benefits of these technical challenges, such as understanding design, selection of materials and costeffective manufacturing process, acceptable strength and modulus, failure mechanisms together with establishing material structure-property is very important. For achieving these properties, some advanced materials, manufacturing and characterisation technologies are emerging, which include: • Graphene incorporated natural fibre composites, the inclusion of nanoparticles into composites for enhanced thermal and functional properties, automated fibre placement technology, additive manufacturing (AM), non-destructive characterisation techniques such as MicroComputed Tomography (m-CT) capable with either in-situ experiments or numerical simulations to assess the physical properties of materials, as well as the measurement of volumetric variability (fibre distribution and matrix volume contents). • New materials and various manufacturing processes have emerged in the last few years, which offer significant opportunities in reducing the overall weight of the components. Hybrid composites such as fibre/metal laminates (FMLs), hybridising natural fibres with carbon and glass fibres could displace steel and alloys in many applications. However, the cost factor and their interlayer compatibility can be a challenging issue (Dhakal et al., 2015; Sarasini et al., 2013). • Utilisation of cellulose nanofiber-reinforced composites and nanotechnology concepts in natural fibre composites and biobased lightweight composites can play a significant role in improving water repellence, coating and barrier properties.

Admittedly, the current development of the manufacturing process, natural fibrebased lightweight new sustainable materials and their technological innovation in key technology sectors (automotive, marine, aerospace and construction) will provide answers to the shortcomings of new lightweight materials and processes. In this context, a paradigm shift from fossil fuel-based economy to circular economy and involvement of multidisciplinary research initiatives must receive high priority with the involvement of key stakeholders, including legislative bodies, OEMs, academic institutions and the consumers.

References Almansour, F.A., Dhakal, H.N., Zhang, Z.Y., 2017. Effect of water absorption on Mode I interlaminar fracture toughness of flax/basalt reinforced vinyl ester hybrid composites. Compos. Struct. 168, 813e825. https://doi.org/10.1016/j.compstruct.2017.02.081. Bourmaud, A., Beaugrand, J., Shah, D., Placet, V., Baley, C., 2018. Towards the design of highperformance plant fibre composites. Prog. Mater. Sci. 97, 347e408. Collins, C., 2017. IDTechEX Report. Composite Materials for the Automotive Sector - Who Needs Who More? IDTechEx Research Article.

Future outlooks and challenges of sustainable lightweight composites

289

Das, O., Kim, N.K., Sarmah, A.K., Bhattacharyya, D., 2017. Development of waste based biochar/wool hybrid biocomposites: flammability characteristics and mechanical properties. J. Clean. Prod. 144, 79e89. Dhakal, H.N., Zhang, Z.Y., Richardson, M.O.W., 2007. Effect of water absorption on the mechanical properties of hemp fibre reinforced unsaturated polyester composites. Compos. Sci. Technol. 67, 1674e1683. Dhakal, H.N., Zhang, Z.Y., Guthrie, R., MacMullen, J., Bennett, N., 2013. Development of flax/ carbon fibre hybrid composites for enhanced properties. Carbohydr. Polym. 96, 1e8. https://doi.org/10.1016/j.carbpol.2013.03.074. Dhakal, H.N., Zhang, Z.Y., Bennett, N., Lopez-Arraiza, A., Vallejo, F.J., 2014. Effects of water immersion ageing on the mechanical properties of flas and jute fobre biocomposites evaluated by nanoindentation and flexural testing. J. Compos. Mater. 48, 1399e1406. Dhakal, H.N., Sarasini, F., Santulli, C., Tirillo, J., Zhang, Z., Arumugam, V., 2015. Effect of basalt fibre hybridisation on post-impact mechanical behaviour of hemp fibre reinforced composites. Compos. Part A Appl. Sci. Manuf. 75, 54e67. Dhakal, H., Bourmaud, A., Berzin, F., Almansour, F., Zhang, Z., Shah, D.U., Beaugrand, J., 2018. Mechanical properties of leaf sheath date palm fibre waste biomass reinforced polycaprolactone (PCL) biocomposites. Ind. Crop. Prod. 126, 394e402. Faruk, O., Bledzki, A., Fink, H.-P., Sain, M., 2012. Biocomposites reinforced with natural fibers: 2000-2010. Prog. Polym. Sci. 37, 1552e1587. Holbery, J., Houston, D., 2006. Natural-fiber-reinforced polymer composites in automotive applications. JOM J. Min. Met. Mater. Soc. 58, 80e86. Ishikawa, T., Amaoka, Masubuchi, Y., Yamamoto, T., Yamanaka, A., Arai, M., Takahashi, J., 2018. Overview of automotive structural composites technology developments in Japan. Compos. Sci. Technol. 155, 221e246. Mohanty, A.K., Vivekanandhan, S., Pin, J.-M., Misra, M., 2018. Composites from renewable and sustainable resources: challenges and innovations. Science 362, 536e542. https:// doi.org/10.1126/science.aat9072. Pervaiz, M., Panthapulakkal, S., KC, B., Sain, M., Tjong, J., 2016. Emerging trends in automotive light- weighting through novel composite materials. Mater. Sci. Appl. 7, 26e38. Pickering, K.L., Efendy, M.G.A., Le, T.M., 2016. A review of recent developments in natural fibre composites and their mechanical performance. Compos. Part A Appl. Sci. Manuf. 83, 98e112. https://doi.org/10.1016/j.compositesa.2015.08.038, 2016. Qu, Z., Pan, X., Hu, X., Guo, Y., Shen, Y., 2019. Evaluation of nano-mechanical behaviour on flax fiber metal laminates using atomic force microscope. Materials 12, 3363. Rohan, K., McDonough, T.J., Ugrestic, V., Potyla, E., Henning, F., 2015. Mechanical study of direct long fiber thermoplastic carbon/polyamid 6 and its relation to processing parameters. In: Proceedings 14th Annual SPE Automotive Composites Conference & Exhibition, 1 June 2015. Sarasini, F., Tirillo, J., Valente, M., Valente, T., Ciof, S., Iannace, S., Sorrentino, L., 2013. Effect of basalt fibre hybridization on the impact behavior under low impact velocity of glass/ basalt woven fabric/epoxy resin composites. Compos. Part A Appl. Sci. Manuf. 47, 109e123. Shah, D.U., 2013. Developing plant fibre composites for structural applications by optimising composite parameters: a critical review. J. Mater. Sci. 48, 6083e6107. https://doi.org/ 10.1007/s10853-013-7458-7. Takahashi, J., Uzawa, K., Matsuo, T., Yamane, M., 2012. Technological Challenges for Realizing Ultra Lightweight Mass Production Automobile by Using CFRTP. ITHEC 2012, Bremen, Germany.

290

Sustainable Composites for Lightweight Applications

Wang, W., Sain, M., Cooper, P.A., 2006. Study of moisture absorption in natural fiber plastic composites. Compos. Sci. Technol. 66, 379e386. https://doi.org/10.1016/j.compscitech. 2005.07.027. Wu, Y., Xia, C., Cai, L., Garcia, A.C., Shi, S.Q., 2018. Development of natural fibre reinforced composite with comparable mechanical properties and reduced energy consumption and environmental impacts for replacing automotive glass-fibre sheet moulding compound. J. Clean. Prod. 184, 92e100.

Further reading Hughes, M., 2012. Defects in natural fibres: their origin, characteristics and implications for natural fibre-reinforced composites. J. Mater. Sci. 47, 599e609. https://doi.org/10.1007/ s10853-011-6025-3.

Index Note: ‘Page numbers followed by “t” indicate tables, “f ” indicates figures and “b” indicate boxes.’ A Acoustic emission (AE), 187e188, 188f Acousto-ultrasonic testing (AUT), 188e189, 189f Additive manufacturing (AM) technique, 151e152, 153f Agave sisalana, 25, 25f Alkaline treatment, 235e237, 236fe237f Animal fibres, 17 Autoclave curing, 151 Autoclave moulding, 149e150, 149f Automated fibre placement, 145e146, 145fe146f B Banana fibres, 27, 27f Bast fibres, 17 Biobased composites cone calorimeter performance, 83 cone calorimeter testing, 79, 80f environmental conditions, 74 fibre-reinforced polymer (FRP) composites, 74 flammability behaviour carbonaceous-silicate char, 84 cone calorimeter performance, 83 cone calorimeter testing, 79, 80f expanded graphite (EG), 83 extracellular polymeric substances (EPS), 84 flammability properties, 76e79 heat release rate (HRR), 80, 81f limiting oxygen index (LOI), 79 nanoclay, 83 natural fibre composites, 84, 85f peak of the total heat release rate (PHRR), 80 polylactic acid (PLA) yarn, 84, 86f smoke production rate curves, 80, 81fe82f

UL-94, 79 uncoated flax fabric, 84, 85f unsaturated polyester resin (UPE) resin, fire properties, 84 flammability properties, 76e79 heat release rate (HRR), 80, 81f lignocellulosic fibre degradation, 75 limiting oxygen index (LOI), 79 natural fibre-reinforced biobased composites boundary element approach (BEA), 100e101 failure modes, 99 finite difference approach (FDA), 100e102, 101f finite element analysis (FEA), 100 Rule of mixtures (ROM), 99 static mechanical properties prediction, 103e104, 103fe104f peak of the total heat release rate (PHRR), 80 smoke production rate curves, 80, 81fe82f thermal and fire properties, 75 thermal conductivity measurements aspect ratio, hemp/PCL biocomposites, 88, 88f bamboo fibre-reinforced polyester composites, 88, 89f crystalline and semi crystalline polymers, 87 fibre-reinforced composites, 88 natural fibre composites, 87 polymer matrix composites, 89e91, 90fe91f thermal degradation behaviour, natural fibres, 75 thermal stability and decomposition behaviour, 76, 77fe78f thermogravimetric analysis (TGA), 75

292

Biocomposites applications, 121, 179 bio-based constituent(s), 124 biomass, 124 cleaner production/green manufacturing, 158e159, 159f damage mechanisms and types composite manufacturing and in-service/ applications, 198, 199f delamination damage, 200, 201f fatigue life model, 214e216 fibre-matrix de-bonding, 198e199, 199f finite element method, 211e212 fibre reinforced polymer composite and metallic materials. See Fibre reinforced polymer composites (FRPCs) impact damage, 213e214, 214fe215f initiations and propagations, 198e199, 200f inter-laminar damage or delamination, 200, 200f intra-laminar damage, 200, 200f matrix micro-cracks, 198e199, 199f numerical analysis, 211e212 peridynamics and micropolar peridynamics, 212 thermal effects, 216e219, 217fe218f, 220f wavelet spectral finite element, 212 damage resistance, 207e208 design and manufacturing processes, 121 additive manufacturing (AM) technique, 151e152, 153f advantages, 131 autoclave moulding, 149e150, 149f automated fibre placement, 145e146, 145fe146f compression moulding, 129, 131e132, 138e141, 139f cost contribution, 152e155 cost reduction, 129e131 design for manufacture and assembly (DfMA), 129e131, 131f economical process, 152e155, 156f elastic moduli (EM), 157e158, 157f extrusion, 129, 142e143, 142fe143f, 152e155 filament winding, 146e149, 147f

Index

hand lay-up, 129, 132e135, 134f injection moulding processes, 131e132, 137e138, 137f natural fibre-reinforced polymer biocomposites, 152e155 open moulding and autoclave processes, 131e132 out-of-autoclave (OoA) moulding, 150e151 pre-impregnated resin, 141e142 resin transfer moulding (RTM), 143e145, 144f spray lay-up, 135, 135f traditional design approach, 129e131, 131f tri-corner approach, 129, 130f ultimate tensile strengths (UTS), 157e158, 157f vacuum bagging moulding, 136, 136f vacuum resin infusion, 141, 141f distribution, 128 eco-design, 121e123, 122fe123f functionality, 121 life cycle stages, 124, 126f loading methods, 179e180 manufacturing defects failure evolution, 166, 167f fibre kinking, 166, 166f fibre misalignment defect, 167, 167f fibre-reinforced composites, 159, 160f formations and process design parameters, 159, 160f inclusions/contamination, 162 microcracks and cracks, 160e161, 161f moisture absorption, 162 porosity (void or pores), 162e166, 163fe164f temperature effects, 162 materials bio-based matrices, 125, 127t biofibres, 125, 127t design, 126 petrochemical-based matrices, 124e125 toxic materials, 126 mechanical behaviours, 179 natural fibres, 123 natural limitations, 121 non-destructive testing (NDT)

Index

acoustic emission (AE), 187e188, 188f acousto-ultrasonic testing (AUT), 188e189, 189f computed tomography (CT) scanner, 189e191, 190fe191f contact vs. non-contact techniques, 209, 209f electromagnetic testing, 185e187, 186f inspection type vs. NDT methods, 209e210, 210t physical behaviours and structural integrity, 210e211, 211t radiographic testing (RT), 183e185, 185f scanning electron microscope (SEM), 195e196, 195fe197f shearography testing, 189, 190f thermography testing (TT) method, 182e183, 184f ultrasonic testing (UT), 182, 183f visual inspection (VI) or testing (VT), 181e182, 181f x-ray micro-computed tomography (mCT) scan, 191e193, 192fe194f production, 126e128, 128t recovery, 129, 130f reliable damage detection and monitoring techniques, 179e180 soil quality, 124 sustainability, 123e124, 125f Biomass, 124 Braiding, 273, 274f C Cannabis Sativa L., 19e20, 20f Carbon and glass fibre-reinforced composites, 2 Ceramic-matrix composites (CMCs), 1e2 Cleaner production/green manufacturing, 158e159, 159f Compression moulding, 131e132 advantages, 140 autoclave process, 138 bulk moulding compound (BMC), 139 fibre-matrix interfacial adhesion, 140 vs. hand lay-up process, 138e139 hot-press method, 138 illustration, 139, 139f processing parameters, 140e141 sheet moulding compound (SMC), 139

293

Computed tomography (CT) scanner, 189e191, 190fe191f Cone calorimeter testing, 79 D Date palm fibres, 22e24, 24f Design for manufacture and assembly (DfMA), 129e131 E Eddy current (EC) inspection, 186e187, 186f Elaeis guineensis, 26e27, 26f Epoxy resins, 6 Euro 6 regulation, 18 Extrusion compounding process, 142e143, 142fe143f F Fibre reinforced polymer composites (FRPCs) advantages and drawbacks, 2t failure/damage mechanisms crack pattern and temperature distribution, 204, 204f ductile-brittle, 202 fatigue crack propagation (FCP), 204, 205f fractographic features, 204 fracture resistance, 204 guided wave base method, 204 incipient thermal damage (ITD), 207 low-velocity impact modes, 202, 203f macro scale leveled damage, 201 matrix cracking and shifting mechanisms, 202, 203f nano- or micro-scale level, 201 progression, 197, 198f, 201, 202f RAPID method, 207 structural health monitoring techniques, 204, 207 tensile fractured notched sample, 204, 206f Twaron 1000 and 1014 aramid fibre reinforced epoxy composites, 202 wave filled imaging, 207 Fibre swelling, 92

294

Filament winding advantages, 148 applications, 148 basalt fibre-reinforced epoxy composite pipes, 146e147 commingled filament-winding method, 146e147 conventional 2-axis lathe type filament winding, 148e149 description, 146 fibre-reinforced polymer manufacturing, 146e147, 147f film stacking method, 146e147 hemp fibre yarn, 146e147 military and defence sector, 148e149 robotic filament winding technique, 148e149 schematic illustration, 146, 147f Flax fibres, 20e21, 21f scanning electron microscope (SEM) images, 45, 46f tensile stress-strain curves, 44, 45f 4D printing, 152 Fused deposition modelling (FDM), 152 G Genus Corchorus oliotorius, 21, 22f Glass fibre-reinforced composites, 10, 11t Glass fibres, 9 Grass fibre, 17 Guided wave base method, 204 H Hand lay-up advantages, 133e134 applications, 133e134 with common matrices, 133e134 curing, 133 description, 132e133 gel coating, 133 illustration diagram, 133, 134f lay-up, 133e134 low volume production, 134e135 mould preparation, 133 nip-roller type impregnators, 132e133 production rate, 134e135 set-up, 133, 134f Hemp fibres, 19e20, 20f Hibiscus cannabinus L., 21e22, 23f

Index

I Incipient thermal damage (ITD), 207 Injection moulding processes, 131e132, 137e138, 137f J Jute fibres, 21, 22f scanning electron microscope (SEM) images, 45, 46f tensile stress-strain curves, 44, 45f K Kenaf fibres, 21e22, 23f Knitting configurations, 266e268, 266f interlaminar fracture toughness responses, 266e268, 270f interlock structure, 266e268 tensile, bending and impact responses, 266e268, 267fe269f 3D stitched woven-knitted basalt fabric biocomposites, 266e268 warp knitting, 265e266 L Laminated object manufacturing (LOM), 152 Leaf fibres, 17 Lightweight composites advantages, 55e57, 56f, 57t aerospace application, 109e110 automotive application chassis, body and engine components, 104e105 E-Class Mercedes-Benz, 107, 107f Formula One (F1) standard components, 108 fuel consumption, 105 hemp fibre door frame, 105, 105f motorsports, 108 natural fibre-reinforced composites, 105e106, 108e109, 108f North American automotive OEMs, 106e107, 106f plant fibre-reinforced composites, 106 vibration damping, 109, 109f weight reduction, 106e107 biobased composites. See Biobased composites

Index

building construction application, 111e112 carbon fibre-reinforced polymer composites (CFRP), 53 cost benefits, 56, 57t creep behaviour creep compliance, 74 definition, 71 hemp fibre-reinforced unsaturated polyester composites, 73 instantaneous deformation, 72 lifetime prediction, 72 natural fibre composites, 71e72 stage I (primary creep), 72 stage II (secondary creep), 72 stage III (tertiary creep), 72 strain vs. time, 72e74, 72fe73f stress and temperature, 74 drivers, 54 environmental properties, 57e58 environmental sustainability, 57 epoxy-based carbon fibre-reinforced composite, 53 fatigue properties carbon fibre-reinforced epoxy composites, 71 constant stress amplitude tests, 70, 70f failure prediction and evaluation, 68 flax fibre-reinforced composites, 71 hemp fibre-reinforced epoxy composite, 70 heterogeneity, 68e70 natural fibre composites, 70e71 property deterioration, 68 stress-life (SeN) diagram, 71 flexural properties, 62e63 greenhouse gases (GHGs) reduction, 54 impact properties composite structure damage, 63e64, 64f energy absorption values, 66 flax fibre-reinforced epoxy composite, 66e67 flax/unsaturated polyester composites, 66 flax/vinyl ester composites, 68 four-layered flax fibre specimen, 66e67 glass fibre-reinforced polymer (GFRP) specimens, 67, 67f hemp fibre-reinforced samples, 65e66 high-velocity impacts, 63 hypervelocity impacts, 63

295

impact velocity levels, 66 jute fibre-reinforced methacrylated soybean oil (MSO) biocomposites, 66e67 jute fibre-reinforced unsaturated polyester, 67e68, 69f low-velocity impacts, 63e65 neat polyester matrix, 65, 65f parameters, 68 plain woven flax fibre-reinforced epoxy composites, 66e67 six-layered specimen, 66e67 toughness, 63 two-layered flax fibre specimen, 66e67 unreinforced polyester, 65e66 long-term durability, 287 marine application, 110e111 market outlook, 285e286 materials and manufacturing process, 286e287 recyclability and end-of-life option, 287 requirements, 55 sporting goods, 112 supply chain scenario, 285e286 surface modifications and chemical treatments, 58 sustainability, 285 tensile properties adhesion, 58e59 glass fibre-reinforced polymer (GFRP) composites, 60 lower tensile properties, 58e59 natural fibre-reinforced biobased composites, 60, 61f, 62, 64 plant fibre composites, 59e60, 59t reinforced composites, 60 stiffness and strength, 59e60 threshold fibre volume fraction, 59 thermal properties, 57e58 thermoplastic-based composite materials, 55 water absorption diffusion coefficient, 92e95 fibre matrix interface, 93e94, 93f fibre swelling, 92 flax/PP and flax/ carbon/PP hybrid composites, 97, 97f flax/vinyl ester composites, 97e99 hemp fibre reinforcement, 94e95

296

Lightweight composites (Continued) heterogeneous structures, 91e92 hygrothermal environments, 91 linear Fickian behaviour, 94 low-velocity impact damage characteristics, 97, 98f moisture absorption tests, 92e93 moisture uptake curve, 92e93 natural fibre composites, 93 non-Fickian behaviour, 94 pseudo-Fickain behaviour, 94 pultruded jute fibre-reinforced unsaturated polyester composites, 95, 96f single flax fibres, 98f, 99 tensile modulus, 95, 95f Lignocellulosic plant-based natural fibres, 17 Limiting oxygen index (LOI), 79 Linear Fickian behaviour, 94 Linum usitatissimum L., 20e21, 21f M Matrices advantages and disadvantages, 5, 5f classification, 5, 5f epoxy resins, 6 functions, 6 phenolic resins, 7 physical and mechanical properties, 6, 7t polyester resins, 6e7 polyethylene (PE), 7e8 polylactic acid (PLA), 8e9 polymeric matrices, 5 polypropylene (PP), 8 polystyrene (PS), 8 thermoplastics, 5, 6t thermosets polymers, 5, 6t vinyl ester resins, 7 Metal-matrix composites (MMCs), 1e2 N Nano-materials, 12 Natural fibre-reinforced composites (NFRCs), 2e3 environmental impact, 2e3 vs. glass and carbon fibres, 2e3, 3t life cycle stages, 2e3, 3f mechanical properties, 4 tensile properties, 4, 4f

Index

Natural fibre-reinforced polymer composites (NFRPCs), 57 Natural fibres vs. glass and carbon fibres, 2e3, 3t NDT. See Non-destructive testing (NDT) NFRCs. See Natural fibre-reinforced composites (NFRCs) Non-destructive testing (NDT) acoustic emission (AE), 187e188, 188f acousto-ultrasonic testing (AUT), 188e189, 189f computed tomography (CT) scanner, 189e191, 190fe191f contact vs. non-contact techniques, 209, 209f electromagnetic testing, 185e187, 186f inspection type vs. NDT methods, 209e210, 210t physical behaviours and structural integrity, 210e211, 211t radiographic testing (RT), 183e185, 185f scanning electron microscope (SEM), 195e196, 195fe197f shearography testing, 189, 190f thermography testing (TT) method, 182e183, 184f ultrasonic testing (UT), 182, 183f visual inspection (VI) or testing (VT), 181e182, 181f x-ray micro-computed tomography (mCT) scan, 191e193, 192fe194f Non-Fickian behaviour, 94 O Oil palm empty fruit bunch (OPEFB), 26 Oil palm fibres, 26e27, 26f Out-of-autoclave (OoA) curing, 151 moulding, 150e151 P Petrochemical-based matrices, 124e125 Petroleum-based materials, 2e3 Phenolic resins, 7 Phoenix dactylifera L., 22e24, 24f Plant-based natural fibres reinforcements banana fibres, 27, 27f cellulose structure hemicellulose, 32

Index

lignin, 33 moisture absorption, 32 molecular structure, 32, 32f chemical compositions, 30e31, 30t, 31f classifications, 17, 18f complex fibre layers and defects, 28, 28f composite reinforcement, 17 date palm fibres, 22e24, 24f European Union’s directive, 18 fibre bundles, 44 fibre processing techniques, 27e28 dew retting, 29 fibre isolation methods, 28 hackling process, 28 injection moulding, 29 internal mixing, 29 micro-compression, 29 retting process, 28e29 steam-explosion, 29 ultrasonic treatments, 29 flax fibres, 20e21, 21f scanning electron microscope (SEM) images, 45, 46f tensile stress-strain curves, 44, 45f hemp fibres, 19e20, 20f jute fibres, 21, 22f scanning electron microscope (SEM) images, 45, 46f tensile stress-strain curves, 44, 45f kenaf fibres, 21e22, 23f lignocellulosic plant-based natural fibres, 17 mechanical processing, 27 middle lamella, 44 moisture absorption behaviour, 18e19 morphological structure flax bast fibre, 34, 36f high aspect ratio, 33 length and width, 33, 34t lumen, 33, 38 and mechanical properties, 33, 34t, 38e39, 39f natural fibre, 33e34, 35f and physical properties, 33, 34t primary and secondary cell walls, 36 reliability and long-term durability, 33e34 sisal leaf fibre and jute bast fibre, 36, 37f strength and stiffness, 33e34 temperature and humidity, 33e34

297

non-cellulose contents, 39e40 non-structural applications, 18e19 oil palm fibres, 26e27, 26f physical and mechanical investigation, 42, 43t composite reinforcements, 41e42 fibre-matrix adhesion, 41e42 interlaminar shear strength (ISS), 41e42 micro-bond test (MT), 42 single fibre fragmentation test (SFFT), 42 single fibre pull-out test (SFPT), 42 PP/lyocell composites, mechanical properties, 40, 40f sisal fibres, 25, 25f strength and modulus, 44 strength and stiffness, 43 tensile properties, 40, 41f tensile strength and modulus, 44 Polyester resins, 6e7 Polyethylene (PE), 7e8 Polylactic acid (PLA), 8e9 Polymeric matrices, 5 Polymer matrix composites (PMCs), 2 Polypropylene (PP), 8 Polystyrene (PS), 8 Porosity (void or pores) defects, 165 description, 162e163 extrusion process, 165e166 fibre-reinforced biocomposites surface, 162e163 hemp-reinforced polylactic acid (PLA) biocomposite samples, 165 inter-laminar and fibre tow voids, 163 natural fibre-reinforced polymer biocomposites, 165 OoA prepreg, 165e166 small and flat MTM45-1/CF2426A prepreg laminate, 163, 164f types, 163, 163fe164f void sinks mechanism, 164 Pre-impregnated resin, 141e142 Pseudo-Fickain behaviour, 94 R Reinforcements carbon fibres, 10 ceramic fibres, 10 glass fibres, 9 natural fibres, 12

298

Resin transfer moulding (RTM) advantages, 144 automobile industry, 145 description, 143e144 hemp fibre-reinforced polyester biocomposites, 145 process parameters, 144 schematic diagram, 143e144, 144f sisal fibre-reinforced biocomposites, 145 sustainable natural fibre-reinforced polymeric biocomposites, 145 S Scanning electron microscope (SEM), 195e196, 195fe197f Seed fibres, 17 Selective laser sintering (SLS), 152 Shearography testing, 189, 190f Silk fibres, 17 Single fibre fragmentation test (SFFT), 42 Single fibre pull-out test (SFPT), 42 Sisal fibres, 25, 25f Spray lay-up, 135, 135f Stalk fibre, 17 Stereolithography (SLT), 152 Stitching advantages, 253 areal stitch fractions, 255, 256f basalt fibre hybridisation, 257 composite laminates, 257 damaging effects, 255 description, 253 FBVEu dry biocomposites, 257 fibre distortions, 255, 255f flax yarn stitch, 255 microscopic examination techniques, 257 mode I delamination toughness, 253 mode I interlaminar fracture toughness, 257 mode II interlaminar fracture toughness, 257 natural fibre, 255, 257 patterns, 253, 254f quasi-isotropic composite laminate, 253, 254f two-dimensional (2-D) fibre reinforcements, 253 water absorption responses, 257, 258f Sustainable composites, 13 additive treatments, 250e251

Index

biological treatments Acetobacter xylinum, 251, 251f field emission scanning electron microscopy (FESEM), 251 Fourier transform infrared (FTIR) spectroscopy, 251 jute fibres filled natural rubber (NR) composites, 251 x-ray diffraction (XRD), 251, 252f braiding, 273, 274f cellulosic fibre and bio-polymer materials, 234 chemical treatments acetylation, 239 alkaline treatment, 235e237, 236fe237f banana fibres, 240e241, 241f bio-based lysine diisocyanate, 239 biocomposite, tensile strength, 240, 241f coupling agents, 242e243 coupling mechanisms, 235 enzyme treatment, 244, 245f esterification, 239, 239f flax fibre-polypropylene compatibility, 241e242 HDPE/Nylon-6 based banana fibre-reinforced biocomposites, 243 Hildegardia FRPP biocomposites, 243 kenaf fibre-reinforced polystyrene biocomposites, 240 maleated coupling, 242 maleic anhydride (MA), 241e242 maleic anhydride grafted polyethylene (PE-g-MA), 243 maleic anhydride grafted styrene/ ethyleneebutylene/styrene triblock polymer (SEBS-g-MA), 243 maleic anhydride-polypropylene copolymer (MA-PP), 241e244 NaOH treatment, 236, 237f, 239 natural fibre surface impurities, 236e237 polymethylene-polyphenyl-isocyanate, 239 silane A151 treated fibres, 240e241 silane treatment, 240e242, 240f soil burial (biological) tests, 241e242 tensile, flexural and impact strength properties, 236e237, 238f vinyl trimethoxy silane (VTMO), 241e242

Index

damage tolerance and properties, 253 environmental life cycle analysis, 230e231 environmental pollution, 230 fibre treatments, 234e235 flax fibres bio-resins, 234 crystalline structure, 231 drilling-induced damage responses, 232e234 macro to nanoscale level, 231e232, 233f mesoscopic level, 231 primary cell wall, 231 schematic microstructure, 231, 232f secondary wall (S3), 231e232 tensile strength, 232 weak fibre-matrix interfacial bonding, 234 government legislation, 230 hybridisation advantages, 258e259 flax/basalt biocomposite, 260e261, 261f flax fibre-reinforced biocomposite laminate, 260e261, 262f hybrid hemp/basalt (HB) reinforced biocomposite, 259, 260f jute fibre-reinforced epoxy biocomposite, 262e263 kenaf/pineapple leaf fibre-reinforced high-density polyethylene biocomposites, 263 natural fibre-reinforced polymeric biocomposites, 259 salt-fog environment conditions, 260e261, 262f stacking sequences, 258e259, 258f inter-dependent benefits, 229 key elements, 229, 230f knitting configurations, 266e268, 266f interlaminar fracture toughness responses, 266e268, 270f interlock structure, 266e268 tensile, bending and impact responses, 266e268, 267fe269f 3D stitched woven-knitted basalt fabric biocomposites, 266e268 warp knitting, 265e266 natural sources, 229 physical treatments

299

cold helium plasma-treated flax fibre-reinforced unsaturated polyester biocomposites, 246e249 corona treatment, 249 electron radiation, 250 fibre beating, 250 Fourier transform infrared-spectroscopy, 246e249 functional surface and structural properties, 246 heat treatment, 250 hydrothermal treatments, 250 plasma treatment, 246, 247f, 249 ultraviolet treatments, 249 pinning, 263e265, 263fe265f productive farming/agricultural practices, 229 stitching advantages, 253 areal stitch fractions, 255, 256f basalt fibre hybridisation, 257 composite laminates, 257 damaging effects, 255 description, 253 FBVEu dry biocomposites, 257 fibre distortions, 255, 255f flax yarn stitch, 255 microscopic examination techniques, 257 mode I delamination toughness, 253 mode I interlaminar fracture toughness, 257 mode II interlaminar fracture toughness, 257 natural fibre, 255, 257 patterns, 253, 254f quasi-isotropic composite laminate, 253, 254f two-dimensional (2-D) fibre reinforcements, 253 water absorption responses, 257, 258f tufting, 274, 274f weaving advantages, 273 biaxial weft knitted fabric, 271, 272f crimping defect, 270e271, 272f definition, 269 patterns, 269, 270f 3D woven pattern, 269, 271f

300

T Thermography testing (TT) method, 182e183, 184f Thermoplastics, 5, 6t Thermosets polymers, 5, 6t 3D printing, 151 Tufting, 274, 274f U Ultrasonic testing (UT), 182, 183f

Index

V Vacuum bagging moulding, 136, 136f Vacuum resin infusion, 141, 141f Vinyl ester resins, 7 Visual inspection (VI), 181e182, 181f X X-ray micro-computed tomography (mCT) scan, 191e193, 192fe194f

Biomechanical energy harvesting with piezoelectric materials

7

Shashikant Guptaa, Bushara Fatmaa, Ritamay Bhuniaa, Prateekb, Raju Kumar Guptab,c, and Ashish Garga a Department of Materials Science and Engineering, Indian Institute of Technology Kanpur, Kanpur, India, bDepartment of Chemical Engineering, Indian Institute of Technology Kanpur, Kanpur, India, cCenter for Environmental Science and Engineering, Indian Institute of Technology Kanpur, Kanpur, India

7.1

Introduction

Energy harvesting implies extracting and storing energy from ambient sources, in the form of vibrations, thermal energy, or light that are present around us. The idea has become especially relevant for a variety of electronic device applications having low or modest power requirements and where one can devise methods either for charging the battery or eliminating the need for a battery altogether. With the growing need to harvest “green energy,” energy harvesting from abundant ambient sources to run low-power electronic devices has gained a lot of attention over the past few decades. William and Yates described three major mechanisms for the energy harvesting from vibrations, viz.,: (1) electromagnetic, (2) electrostatic, and (3) piezoelectric harvesting [1]. Among these, the power generation capacity from piezoelectric-harvesting mechanism is significantly higher than the electromagnetic and electrostatic mechanisms. Hence, the piezoelectric energy harvesting, typically direct conversion of vibration energy into electrical energy, has emerged as a major research area because of higher power densities and ease of implementation [1–3]. The term piezoelectric, derived from the combination of Greek words “piezo,” is synonymous with “to press” and “electric” which refers to the “electricity.” This implies that one can generate an electric field in a material upon application of mechanical stress. Thus, piezoelectricity is a coupling between mechanical force and electrical output energy. This phenomenon was first discovered by French brothers P. Curie and J. Curie in 1880 when external mechanical stress was applied to single crystals (SCs) of materials such as tourmaline, quartz, topaz, cane sugar, and Rochelle salt [4]. The first application of piezoelectric materials was demonstrated by Paul Langevin and his coworkers in France during World War I, in the form of an ultrasonic submarine detector (sonar) with a resonance frequency of 50 kHz, which used a composite system with thin quartz crystals sandwiched between two steel plates. This transducer was used to measure the depth of the water by transmitting a high-frequency chirp signal and measuring the time of the return echo. In 1921, Cady made an important invention of an electrical oscillator, stabilized by a quartz crystal, called quartz crystal-controlled Ferroelectric Materials for Energy Harvesting and Storage. https://doi.org/10.1016/B978-0-08-102802-5.00007-8 © 2021 Elsevier Ltd. All rights reserved.

210

Ferroelectric Materials for Energy Harvesting and Storage

oscillators. These two inventions led to the foundation behind development of new piezoelectric materials and exploring their applications in different fields after World War I. For example, ultrasonic transducers were used to measure fluid viscosity and detect the flaws inside solids. However, naturally occurring piezoelectric materials such as quartz have the main drawback of low piezoelectric constants. Thus, there was a need to develop high piezoelectric constant materials [3, 5]. Subsequently, development and commercialization of piezoelectric materials were revolutionized during and after World War II in United States, Japan, and the former Soviet Union due to their applications in microphones, ultrasonic transducers, accelerometers, and various other devices [3]. A major breakthrough was achieved in the 1950s with the discovery of polycrystalline ceramic materials with high dielectric constants (e.g., barium titanate (BaTiO3)) and high piezoelectric properties (e.g., lead zirconate titanate (PZT)). In 1946, it was found that BaTiO3 exhibits piezoelectric properties after poling. In 1954, it was discovered that PZT has strong piezoelectric coupling properties and can be controlled by doping. Subsequently, piezoelectric-based devices such as sensors, actuators, and transducers were developed for use in a variety of applications, e.g., defense equipment, medical devices as well as consumer electronics. Remarkable developments in the material manufacturing technologies after the 1960s, especially in vacuum technologies and cleaner environments, led to the development of thin and thick film-based piezoelectric devices which reduced the material requirements as well as device dimensions while offering excellent performance. These advances also led to the development of low power-generating mechanical energy-harvesting devices for powering portable/remote electronics subsequent to the development of highperformance materials. These modern piezoelectric material-based devices are attractive energy-harvesting devices due to their potential benefits of ease of installation, low maintenance, less space requirement, ease of fabrication at both macro-scale and micro-scale, long lifetime, possibility of replacing long wires and batteries, and not being restricted to specific climate conditions in contrast to conventional renewable energy sources such as solar and wind energy [2, 5, 6]. While many ceramics have good piezoelectric properties, their brittleness has led researchers to explore polymeric piezoelectric materials which have good intrinsic flexibility, lightweight, lower density (for example, piezoelectric polymers have densities less than 1/4th of the PZT ceramics), ability to sustain much higher strain and resilience to mechanical shock. These can be used in wearable electronic devices that are subjected to bending and conforming to a curved mounting surface [7]. The first use of piezoelectric materials in a biological environment was experimentally shown by H€asler and Stein (1984). Here, a PVDF film was used to transform the mechanical energy generated from respiration of a mongrel dog to electrical energy that was capable of producing a peak voltage of 18 V, which corresponded to a power output of
17 μW [8]. Another study by Ramsay and Clark [9] utilized piezoelectric transducer made of square thin plate driven by blood pressure for power supply in an in vivo micro-electromechanical system (MEMS) application. Since the late 1990s, a substantial amount of research has been conducted toward developing devices for harnessing energy from human body motions and activities as well as from various biological activities, such as cardiac/lung/diaphragm motions,

Biomechanical energy harvesting with piezoelectric materials

211

breathing, blood circulation, muscle stretching, eye blinking, etc. The devices based on such processes are known as biomechanical energy harvesters, as they transform the mechanical energy originated from biological stimuli into electrical energy. The harvested energy could be useful to power various healthcare and monitoring devices as well as to charge the batteries for powering low-power electronic goods. Over time, researchers have been able to use bioenergy via wearable piezoelectric devices that are not only energy conversion-efficient but are also flexible, stretchable, durable, processable, biocompatible as well as nontoxic [10]. Nowadays, given the importance of implantable devices for health monitoring, energy generators utilizing biomechanical energy could pave the way for self-powered in vivo biomedical devices [11]. This chapter presents a review of various aspects of the biomechanical energy harvesters. The chapter outlines the basic principles and summarizes the previous theoretical studies done by various researchers, which provide an understanding of the concepts behind building an energy harvester harnessing biomechanical energy. Further, the requirement for harnessing energy via human motion or any other biological activity should be known before fabricating an energy harvester. In addition, how much mechanical energy is available during a specific human activity should also be known. These aspects are discussed in the first and second sections of the chapter. Next, this chapter discusses various possible designs of a piezoelectric energy harvester, allowing one to appropriately choose the design based on given boundary conditions. Further, we discuss the useful figures of merit for device characterization and quantification of performance. Subsequently, the state-of-the-art section presents a brief account of the recent research works for the technological development of the biomechanical energy harvester and piezoelectric materials. Finally, a summary is provided along with future outlook.

7.2

Principles of biomechanical energy harvesting

The successful implementation of the energy-harvesting technology can eliminate the need for batteries, the requirement of their frequent replacement, and complex wiringrelated issues. Therefore, energy harvesting has evolved as a field of considerable interest to the researchers in their quest to provide a long-lasting and reliable power sources for portable and wearable devices such as mobiles, smart watches, biomedical implants, etc. [12, 13]. Harnessing energy from human activities mainly focusses on activities such as heel strikes, movements of the knee, ankle joint, hip and shoulder, and elbow joint motion during arm flap [14]. For the evaluation of energy produced from the above stated motions, one can use two basic work equations corresponding to linear and angular motions: ðS

!

!

W ¼ F :d S for linear displacement 0

(7.1)

212

Ferroelectric Materials for Energy Harvesting and Storage

and ðθ

!

!

τ :d θ for angular circular displacement

W¼

(7.2)

0 !

!

Here, F and τ are the force and torque vectors applied during linear and angular motions, respectively, while S and θ are total linear and angular displacements, respectively. We can analyze every motion of the body and can estimate the total work done with a sign (positive/negative) at the corresponding joints (knee, hip, ankle, elbow, etc.) of the body during walking. A more tangible way to comprehend the negative work is to imagine force acting in some direction, but the displacement is in the opposite direction. One example is that when something is thrown in the upward direction, the gravitational force acts in the downward direction. Here, the velocity is reduced with time, also leading to a reduction in the kinetic energy, and hence, one can say the force has done the negative work. On the contrary, in case of the same direction of applied force and displacement (e.g., downward fall of a mass), both velocity and kinetic energy increase resulting in positive work.

7.2.1 Theoretical background (analysis with few cases) 7.2.1.1 Heel strike Researchers have tried to model these motions with the help of collision principles using various models, namely elastic collision [15, 16] and perfect plastic collision [17]. Fundamentally, collision is a short-duration interaction between two or more bodies causing perturbation in their motion because of the internal forces acting among them. During an elastic collision, linear (for straight-line motion) or angular (for curved motion) momentum, as well as total kinetic energy, is conserved, while for plastic collision, only linear (for straight-line motion) or angular (for curved motion) momentum is conserved. Considering the linear displacement model, Shorten [16] evaluated the energy dissipated during running in a shoe and suggested that the total energy stored as elastic energy in the shoe sole can be utilized, while other parts of the energy are lost. It was predicted that the dissipated energy for a typical runner with speed 4.5 m/s could range between 1.72 and 10.32 J during a single step and most of the energy would be lost during heel strike [16]. In a simple model, where an external force can be assumed to act on the shoe sole over an entire stride, while the reactive ground force is maximum and equal to body weight, most of the heel compression occurs just after the heel strike. In such a situation, the power of the heel strike can be expressed as: Power ¼ mghn

(7.3)

where, m is mass of the person, g is gravitational acceleration, h is the vertical distance traveled by heel during human gait cycle, and n is a number of steps per second. As an

Biomechanical energy harvesting with piezoelectric materials

213

example, if the m is 70 kg, fall of heel is 5 cm during walking, and number of footsteps per second is 2 for an adult, then the power available for harvesting will be equal to (70 kg)*(9.8 m/s2)*(0.05)*(2 steps/s), i.e., 68.6 W. However, only 50–80% of this power is stored as elastic energy during heel strike, as the remaining power is dissipated through heat generation and sound generation during walking [16].

7.2.1.2 Lower body parts (ankle, knee, hip) motion Our muscles create torque at the knees, hip joints, and ankles during walking, and at the elbows during the motion of hands. This torque is a vector quantity and it has three components along X, Y, and Z axes. The dependency of these vector components on the gait cycle leads to a change in the torque at every moment of the motion. Fig. 7.1 shows the trajectory of joint angle θ (deg), and variation in torque, angular velocity during a gait cycle [14]. In this experiment which was done during a human (weight 58 kg) motion with 1.3 m/s walking speed and 0.9 Hz frequency, zero ankle angle, zero knee angle, and zero hip angle imply the perpendicular position of shank and foot, straight leg, and 90° angle between the thigh and ground, respectively. Essentially, simultaneous measure of 0° for all angles indicate person’s vertical fixed position. Winter et al. [18] defined normalized work as the total work performed at different joints of the body divided by the respective person’s mass during a single step. For this, 12 healthy young adults (7 men and 5 women) having age in the range of 21–28 years, and 15 elder persons (10 men and 5 women) having age in the range of 62–78 years, were considered. Different phases of motion were evaluated during the measurements. The calculated values are shown in Fig. 7.2 [18]. During a gait cycle, one can estimate the total work and the negative work (which cannot be utilized) per step at the ankle, hip, and knee using the data provided in Fig. 7.2. The total and net work done per step are expressed as: X Total work done ¼ Mass  |work phase| Step

(7.4)

Net work done Total work done Total negative work done ¼  Step Step Step

(7.5)

and

The net mechanical work produced by the body joints, as given by the above equation, can be transformed into electrical energy using piezoelectric devices.

7.2.1.3 Center of mass (CM) motion To quantitatively analyze the biomechanical energy from human motion, one needs to consider parameters such as mass, position of the center of mass, and moment of inertia. During motion, a person and his CM move like an inverted pendulum [19, 20]. Also, when one foot is put down and the body vaults over it, it is akin to the motion

Velocity (rad/s)

Angle (deg)

Ankle TO

0

60 40

–10

20

–20 0

0 20

40

60

80

100

0 –20 –40 –60 –80 0

TO

Hip

HC

20

0

20

40

60

80

60

80

100

0 40

20

40

60

80

40

60

80

0 40

60

% Gait cycle

80

20

40

60

80

100

40

60

80

100

80

100

0

20

20

0 2

Ext

Plantar

100 4

–5 40

HC

0

0

20

TO

10

5

2 0 –2 –4 0

Torque (nm)

Knee

HC

–20 Flex 100 0

20

% Gait cycle

100 –2 0 20 40 Ext 20 0 –20 –40 Flex 100 0 20

40 60 % Gait cycle

Fig. 7.1 Variation of bending angle, angular velocity, and torque during walking cycle of a man of mass 58 kg with speed and frequency 1.3 m/s and 0.9 Hz, respectively. Reproduced with permission from R. Riemer, A. Shapiro, Biomechanical energy harvesting from human motion: theory, state of the art, design guidelines, and future directions. J. Neuroeng. Rehabil. 8 (2011) 22.

Biomechanical energy harvesting with piezoelectric materials

215

Fig. 7.2 The value of work performed at different leg joints during walking steps of young and elderly.

h h h

q

m

q q

Inverted pendulum

Fig. 7.3 Schematic diagram of the movement of the center of mass of a man during walking.

of an inverted pendulum, as shown in Fig. 7.3. During walking, our hips, as well as the CM, move up and down by 4–6 cm during each gait cycle [21]. Besides, if the person carries an external load (e.g., a backpack) during walking, the motion of this load must follow the trajectory of CM. We can determine the energy required to move the load during each gait cycle using the following relation, assuming if there is no exchange

216

Ferroelectric Materials for Energy Harvesting and Storage

between potential and kinetic energy during up and down motion of CM. The total energy can be expressed as: E ¼ kinetic energy + potential energy ¼ 2mgh

(7.6)

For example, if the mass of the load is 35 kg and the average distance traveled by the load during up/down motion for each cycle is 5 cm, then E will be 34.3 J and power (P) will be 34.3 W (assuming two steps per second and two steps causing up and down motion of CM). Therefore, we can observe that applying this simple consideration for an average 5 cm movement of CM with 35 kg load, there is a potential of 34.3 W power to be harnessed during walking.

7.2.1.4 Arm motion This motion consists of the backward and forward swing of the arm during walking and running. The arm motion can be divided into two parts: (i) change in the elbowangle (due to relative motion between upper and forearm) and (ii) change of angle at the shoulder (due to relative motion between upper arm and trunk). These motion patterns can further be deconvoluted into vertical and horizontal components. To calculate the parameters such as velocity, torque, and energy, one can use the Newton-Euler equation of motion and Lagrange’s equation. Using recursive inverse dynamics, Riemer and Shapiro estimated that approximately 2.1 and 2.2 W of power would be available from elbow and shoulder motion [14].

7.2.2 Motion’s trajectory during human walking This was modeled by placing energy transducers made of piezoelectric materials that are attached to the elbow, knee, heel, ankle, and hip. The piezoelectric transducers devices can be used over the human body as a tight suit as shown in Fig. 7.4 [22] and data can be collected and modeled. Frequency dependence of the bending angle (θ) is analyzed by conducting Fourier transformation of the periodic signals, which results in discrete frequencies. To analyze a periodic motion, one can demonstrate the frequencies in the multiples of one fundamental frequency f0 with integer multiples of f0 known as corresponding integer harmonics. So, any periodic function can be split into a summation of f0 and its harmonics. Cha [23] demonstrated fitting this frequency-dependent bending angle (θ) to the function shown in Eq. (7.7), considering m ¼ 1, 2, 3, 4, 5, and 6. The values of fitted θ matched well with the experimental measurements collected from hip, knee, and ankle motion, as shown by Riemer and Shapiro [14] in case of higher harmonic (m ¼ 4, 5, and 6). The function used for fitting the data can be expressed as: n

f ðωtÞ ¼ a0 + å am sin ðmω0 t + φm Þ, where m is an integer m¼1

¼ a0 + a1 sin ðω0 t + φ1 Þ + a2 sin ð2ω0 t + φ2 Þ + ⋯ + an sin ðnω0 t + φn Þ

(7.7)

Biomechanical energy harvesting with piezoelectric materials

217

Fig. 7.4 The figure shows a person wearing the blue suit and the green pockets where the transducers were placed in. Reproduced with permission from A. Proto, M. Penhaker, D. Bibbo, D. Vala, S. Conforto, M. Schmid, Measurements of generated energy/electrical quantities from locomotion activities using piezoelectric wearable sensors for body motion energy harvesting. Sensors 16 (2016) 524.

Here, ω0 ¼ 2πf0 and φm is the phase angle of the mth harmonic, while f0 essentially implies the walking frequency. From walking frequency (f), we can estimate the walking speed U based on a relationship given by Bertram and Ruina [24], and Kuo [25]: f ¼ cU 0:58

(7.8)

where, c is a constant and depends on the walking speed. The magnitude of c can be determined by fitting the experimental data to obtain f and U. Cha [23] determined c ¼ 0.773 considering the values of walking speed U ¼ 1.3 m/s and frequency f ¼ 0.9 Hz from Riemer and Shapiro [14].

7.2.3 Electrical response of piezoelectric material: Modeling Generally, these devices have beam type structure which deflects along the curvature of the joint (see Fig. 7.5). Cha [23] modeled this deflection based on small deformations of the piezoelectric beam structure, assuming that the rotation angle is not large enough. For small angle approximation, one can write the beam deflection as: [23]

Dðx;tÞ ¼

8 x2 > > > > < 2r sin ðθðtÞÞ ;when 0  x < x1 ðtÞ > > > > :

x2 + ðx  x1 ðtÞÞ tan ðθðtÞÞ;when x1 ðtÞ  x < L 2r sin ðθðtÞÞ

(7.9)

where, D(x, t) is the beam deflection along length direction (x-axis), x1(t) is r sin (θ(t)) tan (θ(t)), and L is the total initial length of the piezoelectric material. The rotation

218

Ferroelectric Materials for Energy Harvesting and Storage

Piezoelectric device

q r

q D (x) x

q

Fig. 7.5 Schematic of a piezoelectric device attached to the body and its deflection.

of the beam can be related to the charge stored in the device using the following equation: [26–28] L ∂Dðx, tÞ QðtÞ ¼ CV ðtÞ + ϑ ∂x x¼0

(7.10)

where, Q(t) is the amount of stored charge, C and V are the capacitance and voltage generated between the two electrodes of a piezoelectric device, respectively, and ϑ is the electromechanical coupling coefficient. The coupling coefficient (ϑ) can be calculated from the following equation, considering that the piezoelectric material attached to a same-sized substrate acts like a composite beam made of piezoelectric material and substrate: [29] ϑ¼

2  Yp d31 bp  2  hc  hc  hp 2hp

    h2p + 2hp hs Ys =Yp + h2s Ys =Yp     hc ¼ 2 hs Ys =Yp

(7.11)

(7.12)

where, Yp and Ys are Young’s modulus of the piezoelectric material and the substrate material, respectively, d31 is the piezoelectric coefficient of the piezoelectric material, bp and hp are the width and thickness of the piezoelectric material, respectively, hs denotes the thickness of the substrate, and hc gives the information about the distance between the top of piezoelectric layer to the neutral axis before bending. Thus, considering all the above approximations, the open circuit voltage (Voc(t)) can be estimated by considering Q(t) ¼ 0 in Eq. (7.10) and can be expressed as: L ! ϑ ∂Dðx, tÞ VOC ðtÞ ¼ C ∂x x¼0

(7.13)

Biomechanical energy harvesting with piezoelectric materials

219

Now, applying Ohm’s law, we can calculate the current using a load resistance (RL) connected to the piezoelectric device from the following equation: IL ðt Þ ¼

dQðtÞ VL ðtÞ ¼ dt RL

(7.14)

where, VL(t) is the voltage across the load resistance. If the trajectory of the deflection of the beam is expressed by an expression, then VL(t) can be calculated using Eqs. (7.10) and (7.13). For periodic movements such as human walking, output power (PL) from the piezoelectric driven through RL can be predicted with the help of the equation: ðT 1 VL2 ðtÞ PL ¼ dt T RL

(7.15)

0

Here, we have assumed that the piezoelectric beams attached to the ankle, elbow, hip, and knee deform along with the angular motion of each body part during walking, as depicted in Fig. 7.4. According to this hypothesis, change in the deflection between L ðx, tÞ initial and final positions, i.e., ∂D∂x  , could be calculated from the tangent of the x¼0

angles. Cha and Seo [30] studied the comparison between theoretical prediction and experimental values of power transferred to the external load from piezoelectric devices attached to the ankle (during walking with a slipper). They considered three walking speeds of 2, 4, and 6 km/h and observed a good agreement of experimental values with the theoretically predicted values for walking speeds of 2 and 4 km/h. However, at 6 km/h speed, the bending angles were changed due to an increase in speed; therefore, theoretical values were higher than the experimental results.

7.3

Design considerations and performance criterion

Harvesting energy from human motions has its unique challenges primarily due to low frequency body movements which also possess large amplitude [31] and difficulties in achieving required energy density within the constrained size of the harvester. For harvesting biomechanical energy from the ankle, shoulder, etc., knowledge of certain governing parameters such as nature of the mechanical force (tension, compression, shear force, amount, and frequency of force) is very important, which can affect the design of any harvester [31], in turn affecting the device performance. Also, one should be able to synchronize the piezoelectric material energy harvesters of different sizes with the motion of stimuli [7]. The design aspects of a piezoelectric energy harvester can be classified based on geometry as follows: [7] l

l

Cantilever Discs (cymbals, diaphragms) a. Cymbal transducers b. Circular diaphragms c. Other configurations

220

Ferroelectric Materials for Energy Harvesting and Storage

7.3.1 Cantilevers Ease in the fabrication of cantilevers and resonance frequency matching with the biomechanical stimuli renders wide use of cantilevers (unimorphs and biomorphs) for building piezoelectric energy harvesters. Unimorphs, as shown in Fig. 7.6A, consist of single active piezoelectric ceramic or polymer thin film bonded with a metal contact (s). The piezoelectric element which could be a ceramic or a polymer thin film is bonded to a substrate. The piezoelectric material could be fixed at one end to obtain the flexural mode of the structure. Further, to enhance the output of the harvester, one can have two piezoelectric layers separated by a substrate in order to get cumulative output electrical signal during mechanical deformation, resulting in a bimorph structure, as shown in Fig. 7.6B. Bimorph’s output voltage will be enhanced in case of series connection of the films, while output current will be enhanced in case of parallel connection of the piezoelectric film. The resonant frequency of such a cantilever can be expressed as: ϑ2 1 fr ¼ n 2 2π L

rffiffiffiffiffiffiffi EI mw

(7.16)

where, ϑn ¼ 1.875 (Eigenvalue for the fundamental vibration mode), L is the length, m is the mass per unit length, w is the width of the cantilever, E is the modulus of elasticity, and I is the moment of inertia. Application of the proof mass, as shown in Fig. 7.6A, further enables one to control this resonance frequency which would result in enhanced output electrical signal of the cantilever harvester as suggested by Roundy and Wright [32]. In such a case, the resonant frequency would be governed by the following equation: ϑ2 1 fr ¼ n 2 2π L

rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi K me + Δm

(7.17)

(a)

Proof mass (b)

Piezoelectric element Substrate

Fig. 7.6 (A) Unimorph with proof mass at its end (B) Bimorph.

Biomechanical energy harvesting with piezoelectric materials

221

Oscilloscope

Electrode

(a)

Piezoelectric element Electrode PET

(b) Oscilloscope Electrode

Piezoelectric element Electrode PET

Fig. 7.7 Schematic configuration of (A) d33 mode with IDEs and (B) d31 mode with TBEs.

pffiffiffiffiffiffiffiffiffiffiffiffiffi 2 0:236 where, ϑ02 =3 , effective mass me ¼ 0.236mwL, Δ m is the proof mass, while n ¼ ϑn K is the effective spring constant of the cantilever structure. To utilize d31 or d33 mode of the piezoelectric harvester, the electrode configuration could be top-bottom electrodes (TBEs) and interdigitated electrodes (IDEs), as shown in Fig. 7.7. Piezoelectric coefficient d33 is larger than d31 because directions of the mechanical stress and the polarization are the same in the case of d33 and different in the case of d31. Thus, the d33 mode is preferred. Fig. 7.7A and B show a schematic configuration of d33 and d31 modes with IDEs and TBE connections of the electrode for a piezoelectric element-based flexible PET substrate. The application of bimorphs is reported by Pozzi et al. where they proposed an upgraded design of their bimorphs having a trapezoidal shape, which worked on plucking mechanism [33]. Adopting the trapezoidal shape for bimorph helped in reducing the volume of the harvester from 230 to 66 cm3 as well as the weight from 0.24 to 0.1 kg. The device performance was enhanced as the number of bimorphs increased from 4 to 16. This was made possible due to use of permanent adhesive fixing for the bimorphs, while earlier designs used demountable mechanical fixings. So, this harvester was able to produce 0.34 mW of energy at 32 kΩ of load during a running gait.

7.3.2 Disks: Cymbal Energy harvesters based on this design aspect are mostly suitable to harness large impact mechanical energy having high-frequency vibrations. The schematic design of such harvesters made of piezoelectric ceramic disk covered with steel metal cap at both the faces is shown in Fig. 7.8. There is a gap between the metal cap and

222

Ferroelectric Materials for Energy Harvesting and Storage

Metal end-cup fd dc

tp fc

PZT

f f29mm, fc: 17mm, fc: 5mm, dc : 1mm

Fig. 7.8 Cymbal energy harvester. Reproduced with the permission from H.W. Kim, S. Priya, K. Uchino, R.E. Newnham, Piezoelectric energy harvesting under high pre-stressed cyclic vibrations. J. Electroceram. 15 (2005) 27–34.

the piezoelectric element, which serves a very special purpose in enhancing the net piezoelectric coefficient dnet. Due to the gap, when axial stress is applied to the metal gap, a certain portion of that stress is applied radially along the surface. Thus, the net piezoelectric coefficient can be written as: [34] d net ¼ d33 + Ajd31 j

(7.18)

where, A is the amplification factor. Kim et al. fabricated a cymbal with a 29 mm diameter piezoelectric disk with maximum harvested power of 52 mW under bipolar stress of 70 N at a frequency of 100 Hz [34]. Palosaari et al. demonstrated a cymbal-based energy harvester, which can harness the impact energy at the shoe sole during human walking or running gait, as shown in Fig. 7.9 [35]. They utilized the evenly curved convex steel plates instead of traditional

Energy harvester 1500 Brass 150 500 PZT-5H

~750–1000

Bonding layer (Epoxy)

Piston d31

Electrodes (Ag) 20

Ball Dynamic bearing load cell

150, 200, 250, 300 Steels

CPU

35,000

(A)

(B)

Plastic cushions

CPU

Harvesting Electronics

Fig. 7.9 (A) Architecture of cymbal-based energy harvester. All dimensions are in microns. (B) Schematic of harvesting process via stress application by a computer-controlled positioning system. Reproduced with the permission from J. Palosaari, M. Leinonen, J. Hannu, J. Juuti, H. Jantunen, Energy harvesting with a cymbal type piezoelectric transducer from low frequency compression. J. Electroceram. 28 (2012) 214–219.

Biomechanical energy harvesting with piezoelectric materials

223

more angular-shaped design. This harvester was exposed to a cyclic mechanical load via a computer-controlled positioning drive system. This process can simulate the condition of the actual walking or running gait of a human. The optimized conditions had a frequency of 1.19 Hz similar to walking gait and a load of 24.8 N. Thus, the harvester was able to produce 0.66 mW of energy.

7.3.3 Disk: Diaphragms Most of the diaphragms based on piezoelectric materials were used for actuator and sensor applications [36, 37]. One of the diaphragms is shown in Fig. 7.10. Diaphragms generally consist of a circular ceramic disk, adhered on a metal shim followed by clamping at the edges. Unlike the cantilever, the clamping covers the whole edge, while cantilevers get clamped at only one end. Prestressed piezoelectric element can give better electrical output at lower frequencies, which could be achieved by placing a proof mass at the center of the diaphragms [38]. Similar to cantilever harvester, diaphragm-based harvesters can also be used in d31 mode. To achieve larger output by exploiting d33 mode, spiral electrode design was proposed where positive and negative polarity-based spirals were fabricated in an alternate order with spirals coinciding with the center of the piezoelectric disk [39]. These are also known as radial field diaphragm (RFD). Nowadays in the health care sector, various implanted devices, e.g., neural stimulators, pacemakers, and blood pressure monitors, are used by millions of patients. All these devices are based on conventional batteries for their power requirements. Although there are very efficient batteries in the market, they still have limitations due to the need of recharging and replacing after their life, and may also have some adverse effects on human health. So, the approach of having self-powered health monitoring devices will be extremely attractive. Shi et al. demonstrated an implanted energy harvester based on the piezoelectric diaphragm as shown in Fig. 7.11 [40]. Their harvester was meant to harvest vibration of the ultrasonic energy applied from outside the body. This approach aimed to reduce all the risks and disadvantages of conventional energy storing devices used in implanted energy harvesters. Fig. 7.11 shows the working and the architecture of micro-fabricated PZT-diaphragm arraybased broadband piezoelectric ultrasonic energy harvester (PUEH). Such energy Metal film (top electrode) Piezoelectric ceramic (Active element) Metal plate (bottom electrode)

Fig. 7.10 Piezoelectric diaphragm. Reproduced with the permission from R.G. Bryant, R.T. Effinger, I. Aranda, B.M. Copeland, E.W. Covington, J.M. Hogge, Radial field piezoelectric diaphragms. J. Intell. Mater. Syst. Struct. 15 (2004) 527–538.

224

Ferroelectric Materials for Energy Harvesting and Storage

Fig. 7.11 (A) Powering of implanted ultrasonic energy harvester attached to a pacemaker, (B) and (C) architecture of the PUEH, and (D) PUEH on the tip of a finger. Reproduced with the permission from Q. Shi, T. Wang, C. Lee, MEMS based broadband piezoelectric ultrasonic energy harvester (PUEH) for enabling self-powered implantable biomedical devices. Sci. Rep. 6 (2016) 24946.

harvester generated 77 nW of energy by application of 1 mW/cm2 ultrasonic energy at a frequency of 240 kHz.

7.3.4 Other configurations To overcome the challenge of achieving low resonance frequency harvesters based on ceramic piezoelectric materials that have large elastic moduli, a few MEMS-based designs were proposed such as ring MEMS (RMEMS) as shown in Fig. 7.12, which have additional torsional modes [41]. Those modes are normally absent in the planar cantilever counterparts. There is no requirement of any proof mass to achieve lower resonance frequency for biomechanical energy-harvesting purposes in such kinds of designs.

(A)

(B)

Rolled up layers

100 nm 0.5 mm Air gap 100 nm Mo layer 1 AIN Mo layer 2

Mo AIN Mo

(C)

x y z

Sacrificial layer SiO2

Fig. 7.12 (A) Device architecture Mo/AlN/Mo, (B) SEM image of the unimorph showing torsional moments of the tip, and (C) top view of the unimorph in SEM image. Reproduced with the permission from A. Massaro, S. De Guido, I. Ingrosso, R. Cingolani, M. De Vittorio, M. Cori, A. Bertacchini, L. Larcher, A. Passaseo, Freestanding piezoelectric rings for high efficiency energy harvesting at low frequency. Appl. Phys. Lett. 98 (2011) 053502.

Biomechanical energy harvesting with piezoelectric materials

225

Fig. 7.13 Schematic of the S-shaped cantilever energy harvester. Reproduced with the permission from H. Liu, C. Lee, T. Kobayashi, C.J. Tay, C.J.M.T. Quan, A new S-shaped MEMS PZT cantilever for energy harvesting from low frequency vibrations below 30 Hz. Microsyst. Technol. 18 (2012) 497–506.

An S-shaped PZT-based MEMS was micro-fabricated by Liu et al. to harvest the mechanical vibration at very low frequencies (below 30 Hz) and at low acceleration (100

600,000

>100

0.75

0.2

–

0.33–0.78

800–1200 14–15

600 4–13

27 –

300 2.0–3.1

58  10
8 380

>105 510

130 399

– 680

12

6–10

7.6

–

0.35 13.5

0.7 0.0005

– –

– –

25–35 1640–1940 28 700

100–200 >5000 20 100–200

Galfenol

Magneto-mechanical Relative permeability (annealed) Coupling factor Magnetostriction Strain (ppm) Energy density (kJ/m2) Electrical Resistivity (Ω-m) Curie temperature (°C)

Thermal Coefficient of thermal expansion (ppm/°C) Specific heat (kJ/kg-K) Thermal conductivity (W/m-K)

Mechanical Young’s modulus(GPA) Sound speed (m/s) Tensile strength (MPa) Compressive strength (MPa)

1–2