Hall 2021

Advanced Structured Materials
Wayne Hall
Zia Javanbakht
Design and
Manufacture
of Fibre-Reinforced
Composites
Volume 158
Series Editors
Andreas Öchsner, Faculty of Mechanical Engineering, Esslingen University of
Applied Sciences, Esslingen, Germany
Lucas F. M. da Silva, Department of Mechanical Engineering, Faculty of
Engineering, University of Porto, Porto, Portugal
Holm Altenbach , Faculty of Mechanical Engineering, Otto von Guericke
University Magdeburg, Magdeburg, Sachsen-Anhalt, Germany
Common engineering materials reach in many applications their limits and new
developments are required to fulfil increasing demands on engineering materials.
The performance of materials can be increased by combining different materials to
achieve better properties than a single constituent or by shaping the material or
constituents in a specific structure. The interaction between material and structure
may arise on different length scales, such as micro-, meso- or macroscale, and offers
possible applications in quite diverse fields.
This book series addresses the fundamental relationship between materials and their
structure on the overall properties (e.g. mechanical, thermal, chemical or magnetic
etc.) and applications.
The topics of Advanced Structured Materials include but are not limited to
• classical fibre-reinforced composites (e.g. glass, carbon or Aramid reinforced
plastics)
• metal matrix composites (MMCs)
• micro porous composites
• micro channel materials
• multilayered materials
• cellular materials (e.g., metallic or polymer foams, sponges, hollow sphere
structures)
• porous materials
• truss structures
• nanocomposite materials
• biomaterials
• nanoporous metals
• concrete
• coated materials
• smart materials
Advanced Structured Materials is indexed in Google Scholar and Scopus.
More information about this series at http://www.springer.com/series/8611

Wayne Hall · Zia Javanbakht
Design and Manufacture

of Fibre-Reinforced
Composites
Wayne Hall Zia Javanbakht
School of Engineering and Built School of Engineering and Built
Environment Environment
Griffith University (Gold Coast Campus) Griffith University (Gold Coast Campus)
Southport, QLD, Australia Southport, QLD, Australia
ISSN 1869-8433 ISSN 1869-8441 (electronic)

ISBN 978-3-030-78806-3 ISBN 978-3-030-78807-0 (eBook)
https://doi.org/10.1007/978-3-030-78807-0
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Switzerland AG 2021
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse
of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
To my wife, Angela
and my children, Patrick and Edward
Wayne Hall
To my parents and my brother

Zia Javanbakht
Preface
There are many excellent composite design and/or manufacturing books in the liter-
ature, including those that have a mechanics of composites focus. Indeed, many of
these books are cited by this text. There are also textbooks that focus on step-by-
step practical fabrication methods for making Fibre-Reinforced Composite (FRC)
structures—for example, the outstanding ‘Composite Materials Fabrication Hand-
books 1–3’ by Wanberg. So, why write another textbook that focuses on composite
materials?
Well... this textbook has resulted from the course ‘3801ENG Design and Manu-
facture of Composites’ at Griffith University, Australia. This course offers an intro-
duction to FRCs with an emphasis on practical exploration. It aims to span (and
link) the content of the aforementioned specialist types of composite textbooks. The
purpose of this text is therefore to complement these specialist texts, and to bridge
the gap between those with a more design (mechanics) emphasis and those that offer
practical fabrication methods.
The course 3801ENG is core in the ‘Bachelor of Industrial Design’ degree
programme, and a free-choice elective for students enrolled in the ‘Bachelor of
Engineering with Honours in Mechanical Engineering’. The course comprises a 1
hour lecture and 3–5 hours of tutorial activities per week (depending on student
requirements), plus individual project-based assessment items that run in parallel
with the tutorial activities. Students are assumed to have a basic understanding of
mechanics of materials concepts (including tension, compression, flexure, transverse
shear and torsion), and some previous experience on design and make projects using
traditional materials. These skills are included in earlier courses in the Industrial
Design programme.
This textbook covers the lecture and tutorial content of 3801ENG. The tuto-
rial classes are used to reinforce the theoretical content covered in the lectures and
to establish practical, skill-based competencies in composite fabrication. Practical
fabrication methods include wet layup; vacuum bagging; and prepreg moulding. The
design and manufacturing examples in this text, including the step-by-step practical
fabrication activities, are primarily taken from the tutorial content. Project-based
vii
viii Preface
assessment tasks are not included in this textbook—projects are carefully selected,
each course offering to link with, as well as build on, the lecture and tutorial content.
Gold Coast, Australia Wayne Hall

August 2020 Zia Javanbakht
Acknowledgements
The authors would like to acknowledge the help and support provided by our friends
and colleagues. A special mention is given to Prof. Andreas Oechsner for his help
and encouragement, and to Dr. Ian Underhill for technical support provided in the
composite fabrication aspects of this book. The contribution and effort from Dr. Nick
Emerson in the intial stages of the textbook is gratefully appreciated. We would also
like to acknowledge Prof. John Summerscales for his advice on earlier versions of
the manuscript. Finally, we would like to extend our most grateful appreciation to our
families. Their continuous support and encouragement have made this book possible.
ix
Contents
1 Fibre-Reinforced Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Fibre Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4 Fibre Forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.6 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.7 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2 Mechanics of Composite Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2 Axial (Longitudinal) Modulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.3 Transverse Modulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.4 Shear Modulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.5 Effect of Fibre Orientation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.6 Strength of Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.7 Woven and Random Fabrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.8 Volume Fraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.9 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.10 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.11 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3 How to Make a Composite—Wet Layup . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.2 Fibre and Matrix Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.3 Mould Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.4 Layup and Consolidation (Hand Lamination) . . . . . . . . . . . . . . . . . . 39
3.5 Curing and Post-Curing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.6 Crosslinking: Chemical Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.7 Demoulding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.8 Post-Processing (Finishing) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
xi
xii Contents
3.9 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.10 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.11 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4 Advanced Methods—Vacuum Bagging and Prepreg Moulding . . . . . . 55
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.2 Vacuum Bagging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.3 Prepreg Moulding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.5 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.6 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5 Composite Testing—How Accurate Are Design Estimates? . . . . . . . . . 69
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.2 The Tensile Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
5.3 Fibre-Reinforced Composite Specimens . . . . . . . . . . . . . . . . . . . . . . 72
5.4 Tensile Properties: Elastic Moduli and Strength . . . . . . . . . . . . . . . . 73
5.5 Factor of Safety (FoS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.7 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.8 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
6 Moulding Composite Parts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
6.2 Composite Part Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
6.3 Mould Design and Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
6.4 Practical Task: Making a Frisbee . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
6.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
6.6 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
6.7 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
7 Hollow Sections—How to Make Composite Tubes . . . . . . . . . . . . . . . . . 105
7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
7.2 Mandrel Lamination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
7.3 Bladder Moulding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
7.4 Practical Task: a Helical Spring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
7.5 Extension Task: a Bicycle Handlebar . . . . . . . . . . . . . . . . . . . . . . . . . 122
7.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
7.7 Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
7.8 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Symbols and Acronyms
Latin Symbols (Capital Letters)
A Area
E Elastic modulus
F Force
G Shear modulus
I Second moment of area
J Polar second moment of area
L Length
M Moment (bending)
Q Shear force
T Torsional moment (torque)
V Volume fraction
W Weight fraction
Latin Symbols (Small Letters)
b Breadth
d Diameter
h Height
n Number of plies/layers
r Radius
t Thickness
y Distance from the neutral axis
xiii
xiv Symbols and Acronyms
Greek Symbols (Capital Letters)
Denotes the change in a variable.
Greek Symbols (Small Letters)
α Coefficient of thermal expansion

δ Deflection
Strain
η Fibre efficiency factor
ν Poisson’s ratio
ρ Density
σ Tensile or Compressive Stress
τ Shear Stress
θ Angle
Indices (Superscripts)
. . .∗ Strength (or failure)
Indices (Subscripts)
. . .avg Average
. . .b Bending
. . .c Composite
. . .cl Composite (longitudinal)
. . .comp Compression
. . .ct Composite (transverse)
. . .eq Equivalent
. . .f Fibre
. . .i Inner
. . .m Matrix
. . .man Mandrel
. . .max Maximum
. . .O Orientation
. . .o Outer
. . .s Shear
Symbols and Acronyms xv
Abbreviations
e.g. for example (from Latin ‘exempli gratia’)

et al. and others (from Latin ‘et alii’)
etc. and others (from Latin ‘et cetera’)
i.e. that is (from Latin ‘id est’)
viz. namely, precisely (from Latin ‘videlicet’)
vs. against (from Latin ‘versus’)
Chapter 1
Fibre-Reinforced Composites
Abstract The concept of fibre-reinforced polymer composites is introduced in this

chapter. Most fibre-reinforced composites (FRCs) comprise two constituent materi-
als (known as phases): a fibre phase and a matrix phase. It is noted that thermosetting
matrices are far more frequently used for FRCs than their thermoplastic counterparts.
An overview of the three most commonly used thermosetting matrices (i.e. unsatu-
rated polyester, vinylester and epoxy) is provided. Available fibre types and forms
are considered. The representative mechanical properties for the most common fibre
types (glass, carbon and aramid) are presented and compared to traditional materials.
1.1 Introduction
A composite is a material comprising two or more constituent materials. The com-

posite offers a combination of properties that are different from each of the con-
stituents [1, 2]. Many composites comprise only two materials (or phases) [3]: a
matrix that surrounds a dispersed phase. In a fibre-reinforced composite (FRC),
fibres act as the dispersed phase. Fibres may be long or short [4].
In recent years, there has been a shift towards the use of FRCs. Glass, carbon and
aramid fibres are frequently used as reinforcements since they provide high stiffness
and strength when used in structural applications. The matrix transfers the load to the
fibres, defines the shape and protects the fibres from the environment [1]. Polyester,
vinylester and epoxy are commonly used as matrices [5].
Fibre-reinforced composites are now widely used in the transport and energy
sectors [6] as well as for numerous other industrial and commercial applications,
including storage tanks and pressure vessels [7]. In addition, composites are finding
favour in the sports and leisure industry [8] and in the medical sector. Thus, FRCs are
now an essential element in industry and technology-based innovation, and therefore
an important part of the education of the next generation of mechanical engineering
and design graduates [9].
This chapter introduces the fundamentals of fibre-reinforced polymer compos-
ites, specifically focusing on the most commonly available fibre reinforcements and
matrices.
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 1
W. Hall and Z. Javanbakht, Design and Manufacture of Fibre-Reinforced Composites,
Advanced Structured Materials 158,
https://doi.org/10.1007/978-3-030-78807-0_1
2 1 Fibre-Reinforced Composites
1.2 Matrices
The polymer matrix of a FRC can be classified as thermoplastic or thermoset [3].

Thermoplastics soften (and melt) when heated moderately but return to a solid form
upon cooling. Thermosets are by far the most commonly used for FRC structures [10];
they are supplied as liquids and tend to offer an easier processing route. They are
formed from a chemical crosslinking reaction that is often initiated at ambient tem-
perature. The mixing of the liquid resin and a catalyst (or hardener) causes a non-
reversible chemical reaction and a solid polymer is produced. In contrast to thermo-
plastics, thermosets will not melt upon heating but their mechanical properties may
diminish at elevated temperatures. The most common thermosets used for composite
applications are polyester, vinylester and epoxy [5].
Polyester. Unsaturated polyester1 is the most widely used thermosetting resin sys-
tem (followed by epoxy) [8]. The term unsaturated arises from the carbon double
bonds (C=C) in the backbone of the molecular chain [3]. These carbon atoms are not
attached to the maximum number of atoms and hence other atoms are able to react at
the unsaturated bonds (reactive sites). The unsaturated bonds are therefore the focus
of the crosslinking reaction in polyester (as well as in vinylester) [10]. Most polyester
resins have a styrene reactive diluent of up to 50% styrene [11] to reduce the resin
viscosity (for ease of handling). The styrene molecules also play a significant role
in the crosslinking reaction, i.e. they form the links between the polyester chains
as polymerisation progresses [10]. Polyesters resins cure in the presence of a cata-
lyst. The catalyst is mixed in small quantities (1–2% is typical [12]) and is usually
peroxide-based—for example, methyl ethyl ketone peroxide or MEKP. Polyesters
are relatively low-cost and easy to use but they tend to be brittle and prone to high
cure shrinkage; a shrinkage of about 7% is typical, but this value could be as high as
12% [13].
Vinylester. Like polyester, vinylester is diluted with styrene and cured with a catalyst
[11]. The reactive carbon double bonds (reactive sites) are located at the ends of the
molecular chains [8] providing a tougher and more resilient matrix [11] with typically
lower shrinkage than polyester [14]—polyester typically shrinks 5–12%, whilst the
shrinkage of vinylester is only 5–10% [13]. In addition, the backbone is similar to that
of epoxy and this combined with fewer ester groups improves the water and chemical
resistance in comparison to its polyester counterpart—ester molecules are susceptible
to hydrolysis [11]. Vinylester resins are more expensive than polyester resins and are
often harder to source. Moreover, vinylesters tend to be out-performed by their more
expensive (and more widely available) epoxy counterparts, particularly in terms of
water and chemical resistance and, as a result, they are sometimes overlooked [5].
Epoxy. Epoxy resins are more expensive than either polyester or vinylester resins [3].
They offer higher performance in terms of mechanical and thermal properties, and as
1 There are two main polyester types: orthophthalic and isophthalic resins. Orthophthalic resins are
the standard economic (or basic) resins, whilst isophthalic are general purpose resins with a higher
cost but superior performance, especially in marine environments [11].
1.2 Matrices 3
Table 1.1 Advantages, disadvantages and typical mechanical properties of polyester, vinylester
and epoxy [3, 19, 20]
Material ρ E σ∗ ν α Advantages Disadvantages
g
( 3) (GPa) (MPa) (×10-6 ◦1C )
cm
Polyester 1.04–1.46 2–4.5 34.5–103.5 0.37–0.39 55–200 Lowest cost Moderate
resin. stiffness and
Low viscosity. strength.
Room- High styrene
temperature content.
curing. High cure
shrinkage.
Vinylester 1.12–1.32 3.0–3.5 73–81 53 Better More expensive
mechanical than polyester.
properties than High styrene
polyester. content.
Improved water Inferior to
and chemical epoxy.
resistance.
Lower cost than
epoxy.
Epoxy 1.1–1.4 2.41–4.10 27.6–130 0.38–0.40 45–117 Excellent Highest cost.
strength/stiffness. Poor UV
Good thermal resistance.
stability. Post-cure
Good solvent requirements.
resistance.
Lowest
shrinkage.
previously mentioned, improved water and chemical resistance [10]. The improved
water and chemical resistance stems from the lack of ester groups that are evident
in polyester and vinylester [11]. They also exhibit lower shrinkage than polyester
or vinylester (1.2–4% [4]). The main weakness of epoxies, however, is that they
discolour with exposure to UV light. Evidence of UV discolouration is a yellow
tinge [15] for clear epoxies. Epoxies are mixed with a hardener to initiate the
crosslinking reaction. Typically, an amine hardener is used (co-reaction) to open
the epoxy rings (two carbon atoms bonded to an oxygen atom) at either end of the
molecular chain [10].
Polyester, Vinylester or Epoxy? There are advantages and limitations of each ther-
mosetting resin type, so the choice of matrix depends largely on the suitability of
a resin for a given application. A comparison of various material properties (mass
density ρ, modulus of elasticity E, tensile strength σ ∗ , Poisson’s ratio ν, and the
coefficient of thermal expansion α along with some of the main advantages and dis-
advantages of unsaturated polyester, vinylester and epoxy resin are listed in Table 1.1.
1.3 Fibre Types
The most common fibres used as structural reinforcements are glass, carbon and
aramid [3, 16]. Other specialist fibres are also commercially available, which include
natural fibres such as flax, hemp, jute and kenaf [17, 18], but these are not considered
here.
Glass fibres. Glass fibres are usually E-glass (electrical), but other types of glass
fibres (for example, S- or structural glass) are also available [19]. The low-cost and
balanced mechanical properties of E-glass fibre make it the most common fibre
choice [4]. S-glass is stiffer (by approx. 20%) and stronger (30–40% stronger) than
E-glass (see Table 1.2), which provides some justification for why it is significantly
more expensive.
Carbon fibres. Carbon fibres are identifiably black in colour and are more expen-
sive than either glass or aramid fibres, but tend to offer superior elastic modulus and
(at least) comparable strength. Carbon is available in several grades such as stan-
dard modulus and high strength carbon (HS), or high modulus carbon (HM). High
strength (standard modulus) carbon fibres are the most readily available and tend to
be preferred, except for unique or specialist applications.
Table 1.2 Typical mechanical properties for selected fibre types [3, 4, 10, 19, 21]
Material ρ E σ∗ ν α Elongation
(g/cm3 ) (GPa) (MPa) α (%)
(×10-6 ◦1C )
Glass 2.50–2.62 69–81 2000–3450 0.22 4.9–5.2 2.6–4.9
(E-glass)
Glass 2.46–2.50 83–89 4585–4800 5.3–5.7 5.7
(S-Glass)
Carbon (HS∗ ) 1.75–1.80 228–300 3400–7100 0.2 −0.6 (axial), 1.1–2.4
10 (radial)
Carbon 1.80–1.95 350–550 1900–4500 0.2 −0.7 (axial), 0.4–0.7
(HM† ) 10 (radial)
Aramid 1.40–1.44 83–85 3000–3606 4.0
(HT‡ )
Aramid (HM) 1.40–1.44 130-131 3600–4100 −6 (axial), 60 2.8
(radial)
Aluminium 2.70 69 124–310 0.33 23.6 17-30
6061
Steel A36 7.85 207 400–500 0.30 11.7 23
(mild)
Titanium 4.43 110–114 900–1172 0.34 8.6 10–14
Ti-6Al-4V
∗ High strength
† High modulus
‡ High toughness
1.3 Fibre Types 5
Aramid fibres. Aramid fibres (from the term aromatic polyamide) are yellow or
golden fibres with high tensile strengths comparable to carbon fibres [19, 21], but
relatively low compressive strength [22]. The term Kevlar is sometimes incorrectly
used in place of aramid; Kevlar® is the DuPont trade-name. Like glass and carbon,
there are also different types of aramids [10]. Aramid fibres have low densities (even
lower than carbon) and due to their toughness, they tend to perform very well in
impact applications. As a result, aramid fibres are often used in laminates for ballistic
applications [4].
Glass, Carbon or Aramid? Similar to the matrix selection, the choice of a fibre type
depends on the composite application and requisite material properties. A comparison
of the physical, mechanical and thermal properties for a selection of fibre types is
provided in Table 1.2. To provide context, some common metals are shown for
comparison.
1.4 Fibre Forms
Fibre Bundles. Reinforcement fibres are sometimes supplied as fibre bundles and
come wound on a reel. The terms strand (or sometimes end), tow or roving may also
be used [4]. Strand is used as a generic term and may be applied to glass, carbon or
aramid fibres [5]. Tow is often used for high performance carbon and aramid fibres,
whilst roving is normally reserved to describe thicker, low-cost glass fibre bundles.
The use of strands is normally associated with manufacturing methods such as spray
up [10] and filament winding [8, 10], which are beyond the scope of this textbook.
Reinforcement Fabrics. The term fabric is used here to describe a fibre assembly
that produces a flat sheet in either woven or nonwoven format. Fabrics are usually
sold on a roll by the linear metre and are described in terms of fabric weight (areal
weight) in grams per square metre (gsm or g/m2 ). Narrow fabric strips are often referred
to as tapes.
Fabrics are characterised here into four main types: unidirectional, woven, braided
and random mat. Other fabrics that fall outside these classifications are beyond the
scope of this textbook. Fabrics tend to be based on long fibres, but random mats may
have a long, continuous swirl of fibres or chopped short fibres (for example, chopped
strand mat) [4].
Unidirectional Fibres. A unidirectional (UD) fabric is a nonwoven fabric with fibres
mainly orientated in one direction [8]—see Fig. 1.1. Only a small amount of fibres
(or a binder material) are orientated perpendicular to the main fibre reinforcements to
maintain the integrity of the fabric structure. A unidirectional fabric is easy to handle
and can be laid flat and straight in a mould as there are no gaps between fibres. This
tends to result in higher fibre volume fractions (i.e. a higher volume of fibres in the
composite) and hence, unidirectional fibres are used to produce composites with the
maximum stiffness and strength, but this performance is only in the direction of the
fibres [11]. Fabrics can be overlapped at multiple orientations to resist the requisite
Fig. 1.1 Unidirectional

(UD) fabric Fibre direction
applied loads [19], making UD fabrics an extremely versatile option when designing
FRC parts.
Woven Fabrics. Woven fabrics are created by interlacing fibre bundles at right angles
to each other. The two orientations are referred to as the warp (0◦ ) and weft (90◦ ) as
shown in Fig. 1.2.
Three common woven fabrics are [4]:

• Plain weave.
• Twill weave.
• Satin weave.
In plain weave, each warp strand passes over and then under each weft strand—see
Fig. 1.3a. The result is a symmetrical and stable fabric, but a fabric that tends to
experience poor drape and is more prone to surface porosity (wet out issues) than its
twill or satin weave counterparts [23]. Thus, plain weave fabrics tend to be preferred
for flat, planar laminates or parts with only slight curves [5].
Twill weave fabrics have a balance of stability and drapeabilty so tend to find a
wide range of applications. They have lower crimp (each fibre bundle follows a less
wavy path) than plain weave fabrics and therefore a better drape [10]. They also tend
to not experience the same porosity issues sometimes associated with plain weaves.
A common twill weave arrangement is the 2/2 pattern; the warp fibre passes over two
and then under two weft strands as shown in Fig. 1.3b. This 2/2 twill fabric pattern
produces the commonly sought-after carbon fibre aesthetic [11]—see Fig. 1.3c.
Satin or harness weave [10] is the least stable but has good drape and the lowest
crimp. Thus, it tends to be mostly used for moulded parts with complex curvature [5]
as the lack of stability makes it more difficult to maintain accurate fibre orientations
than with the other two woven fabrics. A satin weave is asymmetrical with fibres
running predominately in the warp direction on one fabric face as shown in Fig. 1.3d
and in the weft on the other face [11]. This asymmetry should be considered when
designing composite parts to ensure coupling stresses [19] do not cause warping
under load.
1.4 Fibre Forms 7
Warp
Weft
Fig. 1.2 Warp and weft of a plain weave fabric
Braided Fabrics. Braiding can be used to produce flat tapes [8] and this style is
therefore considered to be a fabric [10]. It has some similarity with woven fabrics.
Most commonly, however, braiding is used to produce braided sleeves, a specialist
fibre preform. Braided sleeves (sometimes referred to as either socks or tubes [5]) are
supplied as a continuous tube of fibres—see Fig. 1.4. Multiple strands are interwoven
to create a tubular fabric [8]. Typically, braids are supplied with fibres at ±45◦ to
the tube’s longitudinal axis with orientation based on a set diameter. Expanding or
compressing the tube’s length reduces or increases the diameter with a corresponding
change in fibre orientation. Fibre orientations can change between 25 and 75◦ [11]
to accommodate tubular structures with variable diameter [5].
Random Fibres. Random fibres are supplied as either mat or veil—see Fig. 1.5.
Random mats may use long fibres (continuous strand mat) or short chopped fibres [4].
They provide a near uniform distribution of fibre orientations and therefore produce
near isotropic mechanical properties (i.e. elastic modulus and strength) [24]. Glass
fibres are extensively used in random mats. The random arrangement of the glass
fibres results in large gaps between strands and hence excess resin in the finished
laminate [5]. Nonetheless, random mats are cheap and relatively easy to handle, even
though (as is the case for chopped strand mat) they may not be as rugged as woven
fabrics [10]. They are used to quickly build up the FRC thickness but are rarely
used in high performance applications due to the low fibre volume fractions that
are usually achieved [11]. They therefore tend to be the chosen reinforcement for
composite mould tools (composite tooling), lower performance structures [25] and
for composite repair.
(a) (b)
(c) (d)
Fig. 1.3 Various fabric types: a Plain weave; b and c twill weave; and d satin weave
Fig. 1.4 Braided ±45◦

carbon sleeve
1.4 Fibre Forms 9
(a) (b)
Fig. 1.5 Random fabrics: a chopped strand mat; and b veil
A tissue or veil is similar to a random mat except that the fibres are much finer.
Their primary role is not as a reinforcement, but rather as a surface layer [5]. A veil
is often used as a supportive backing for a gelcoat [10]; a gelcoat is a resin with
pigment and fillers added to provide a high-quality finish that is wear and weather
resistant [26].
1.5 Summary
The most common thermosetting resins are:

• Polyester.
• Vinylester.
• Epoxy.
Typical fibre types used as structural reinforcement are:
• Glass.
• Carbon.
• Aramid.
Most fibres are sold as a flat sheet fibre assembly, referred to as a fabric. Fabrics are
usually supplied as a roll but narrow fabric strips referred to as tapes can also be
purchased. Fabric types can be categories as:
• Unidirectional.
• Woven (common types are plain, twill and satin).
• Braided (including sleeves).
• Random fibres (random mat and veil).
Material properties for some common fibres and matrices are shown in Table 1.3.
Table 1.3 Some common fibre properties

Material Density Modulus of elasticity Tensile strength
(g/cm3 ) (GPa) (MPa)
Glass (E-Glass) 2.50–2.62 69–81 2000–3450
Carbon (HS) 1.75–1.80 228–300 3000–7000
Carbon (HM) 1.80–1.95 350–550 1900–4500
Polyester 1.04–1.5 2.0–4.5 34.5–103.5
Vinylester 1.12–1.32 3.0–3.5 73–81
Epoxy 1.1–1.4 2.41–4.1 27.6–130
1.6 Questions
Question 1.1 Define the term composite. How does the term composite differ from
fibre-reinforced composite?
Question 1.2 What are the main advantages of FRCs?
Question 1.3 In a FRC, define the roles of the matrix and reinforcement fibres.
Question 1.4 Explain the difference between thermoplastic and thermosetting poly-
mers?
Question 1.5 Identify the three most common thermoset matrices used in composite
applications. List the advantages and limitations of each matrix.
Question 1.6 What are the three most common fibre types used to make composites?
Question 1.7 Define the following terms:
a. Strand.
b. Tow.
c. Roving.
Question 1.8 Describe (with the aid of sketches, if necessary) the following fabrics:
a. Unidirectional.
b. Woven.
c. Braided.
d. Random fibres.
1.7 Problems
Problem 1.1 Typical stress-strain curves for E-glass, carbon (HS) and aramid (HT)
fibre are shown in Fig. 1.6.
1.7 Problems 11
Tensile stress (MPa) 4000
3000
2000
1000
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
Strain (%)
Fig. 1.6 Tensile stress-strain behaviour of a HS carbon-epoxy composite
a. Identify the stress-strain curve for each fibre type.

b. Calculate the elastic moduli of the three fibres from their stress-strain curves.
Answer. a. Carbon, aramid and E-glass (left to right); b. 237 GPa, 85 GPa and
69.7 GPa.
Problem 1.2 From the stress-strain data in Problem 1.1, identify

a. The stiffest fibre.
b. The strongest fibre.
c. The most ductile fibre.
d. The toughest fibre.
Answer. a. Carbon; b. Carbon; c. E-glass; d. Aramid.
References
1. Askeland DR, Fulay PP (2009) Essentials of materials science and engineering, 2nd edn.
Cengage Learning, Australia
2. Shackelford JF (2015) Introduction to materials science for engineers, 8th edn. Pearson, Boston
3. Callister WD, Rethwisch DG (2018) Materials science and engineering: an introduction, 10th
edn. Wiley, Hoboken NJ
4. Barbero EJ (2017) Introduction to composite materials design, 3rd edn. Composite materials.
CRC Press, Boca Raton
5. Wanberg J (2009) Composite materials: fabrication handbook #1, vol 1. Composite garage
series. Wolfgang Publications, Stillwater, Minnesota
6. Jones RM (1999) Mechanics of composite materials, 2nd edn. Taylor & Francis, Philadelphia,
Pa. and London
7. Vasiliev VV, Morozov EV (2013) Advanced mechanics of composite materials and structural
elements, 3rd edn. Elsevier, Amsterdam
8. Astrom BT (2018) Manufacturing of polymer composites, 2nd edn. Routledge, Boca Raton
9. Hall W, Palmer S (2015) Student opportunities in materials design and manufacture: introducing
a new manufacturing with composites course. J Mater Educ 37(3–4):155–168
10. Strong AB (2008) Fundamentals of composites manufacturing: materials, methods and appli-
cations, 2nd edn. Society of Manufacturing Engineers, Dearborn, Mich
11. Gurit (2019) Guide to composites. https://www.gurit.com/Our-Business/Composite-Materials
12. Fibreglass & Resin Sales (2011) Directions for using MEKP catalyst. https://www.fibreglass-
resin-sales.com.au/wp-content/uploads/MEKP-sheet1.pdf
13. Fiore V, Valenza A (2013) Epoxy resins as a matrix material in advanced fiber-reinforced poly-
mer (frp) composites. In: Advanced fibre-reinforced polymer (FRP) composites for structural
applications. Elsevier, pp 88–121. https://doi.org/10.1533/9780857098641.1.88
14. Lubin G (ed) (2013) Handbook of Composites. Springer, New York, NY
15. Hollaway L (1990) Polymers and polymer composites in construction. Civil engineering design.
Thomas Telford Publishing
16. Smith WF, Hashemi J, Presuel-Moreno F (2019) Foundations of materials science and engi-
neering, 6th edn. McGraw-Hill Education, New York, NY
17. Summerscales J, Dissanayake NP, Virk AS, Hall W (2010) A review of bast fibres and their
composites. part 1 – fibres as reinforcements. Compos Part A: Appl Sci Manuf 41(10):1329–
1335. https://doi.org/10.1016/j.compositesa.2010.06.001
18. Summerscales J, Virk AS, Hall W (2013) A review of bast fibres and their composites:
part 3 - modelling. Compos Part A: Appl Sci Manuf 44:132–139. https://doi.org/10.1016/j.
compositesa.2012.08.018
19. Hull D, Clyne TW (1996) An introduction to composite materials, 2nd edn. Cambridge solid
state science series. Cambridge University Press, Cambridge
20. Agarwal BD, Broutman LJ, Chandrashekhara K (2006) Analysis and performance of fiber
composites, 3rd edn. Wiley and Chichester, Hoboken, N.J
21. Akovali G (2001) Handbook of composite fabrication. Woodhead Publishing, Shawbury
22. Virk AS, Hall W, Summerscales J (2009a) Multiple data set (mds) weak-link scaling analysis
of jute fibres. Compos Part A: Appl Sci Manuf 40(11):1764–1771. https://doi.org/10.1016/j.
compositesa.2009.08.022
23. Campbell FC (2010) Structural composite materials. ASM International, Materials Park, Ohio
24. Lokensgard E (2010) Industrial plastics: theory and application, 5th edn. Delmar Cengage
Learning, Clifton Park NY
25. Fibreglast (2019) The fundamentals of fiberglass. https://www.fibreglast.com/product/the-
fundamentals-of-fiberglass/Learning_Center
26. Vaitses AH (2008) The fiberglass boat repair manual. International Marine, Camden, Me
Chapter 2
Mechanics of Composite Structures
Abstract This chapter addresses the mechanical properties of a fibre-reinforced

composite (FRC). The focus is on calculation of the elastic modulus and strength for
unidirectional FRCs using the rule of mixtures expressions, but woven and random
fibres are also considered. A unidirectional FRC exhibits anisotropic behaviour. It
is stiffest and strongest in the fibre direction but is relatively compliant and weak in
the transverse orientation. Woven structures provide similar mechanical properties in
their axial (longitudinal) and transverse orientations, whilst random fibre composites
simulate in-plane isotropic behaviour. The effect of the fibre and matrix properties
on the structural behaviour of a composite is investigated in the context of fibre
volume fractions (and weight fractions), assuming no voids and perfect fibre-matrix
adhesion.
2.1 Introduction
The design of a FRC requires:

• The selection of materials (fibres and matrix).
• An appropriate choice of a manufacturing method.
In this chapter, fibre and matrix selection is considered in the context of structural
performance. The selection of an appropriate manufacturing method is covered in
Chaps. 3 and 4.
Material selection is a complex task as the mechanical properties of FRCs are
usually anisotropic, i.e. their mechanical properties are direction-dependent [1]. A
fibre-reinforced laminate is made up of several laminae (or plies) stacked on top of
each other. Often unidirectional or woven plies are used, but simple low-cost random
fibre mats also offer some design and manufacturing opportunities for composite
materials. Random mats tend to exhibit near-isotropic (referred to as quasi-isotropic)
behaviour [2], and are often used in the marine and automotive sectors for body
repair [3].
Herein, an introduction to the mechanics of FRCs is provided to assist mate-
rial selection. The calculations include many simplifications and assumptions, and
https://doi.org/10.1007/978-3-030-78807-0_2
14 2 Mechanics of Composite Structures
should not be considered more accurate than an estimate. Design calculations are an
important step in any product design process but, on their own, are not a substitute
for physical testing.
2.2 Axial (Longitudinal) Modulus
A fibre-reinforced composite subjected to an axial (tensile or compressive) load will

increase or decrease in length. A tensile force will cause an increase in length of the
composite, whilst a compressive one will cause a decrease in its length. This change
in length depends on several factors including:
• The magnitude of the load.
• Geometrical dimensions of the composite.
• Mechanical properties of the FRC material.
The mechanical properties of a fibre-reinforced composite are influenced by the
properties of both the fibres and matrix; the amount (usually referred to as the volume
fraction) of the fibres; the fibre length and the orientation of the fibres; and the fibre-
matrix adhesion characteristics [2, 4, 5].
Now, consider a fibre composite made up of a number of parallel unidirectional
plies stacked on top of each other to create a solid, homogenous composite material.
This FRC is shown in Fig. 2.1 under tension and compression. The plies consist
of long, continuous fibres in a matrix. The fibres are all orientated in the same
direction, referred to as the axial [5] or longitudinal [2, 4] direction. When all fibres
are orientated in the same direction, the composite is said to be an aligned fibre
composite, a unidirectional composite or simply a UD composite.
The mechanical properties of the composite material are anisotropic; a unidi-
rectional composite has the highest elastic modulus (stiffness1 ) and strength in the
fibre (longitudinal) direction. The stiffness and strength of a unidirectional compos-
ite perpendicular (often termed transverse) to the fibre is much lower [5]. The fibre
and matrix contributions govern the mechanical properties of the longitudinal and
transverse directions, respectively.
A typical tensile stress-strain curve for a high-strength (HS) carbon-epoxy unidi-
rectional composite loaded parallel to the fibres is shown in Fig. 2.2. The stress-strain
curves for the stiff and strong carbon fibres, and the weaker but more ductile epoxy
matrix are overlaid for comparison. The composite performance lies somewhere
between that of the fibres and matrix behaviour. Tensile stress is proportional to
strain, and the elastic modulus (or Young’s modulus of elasticity) is the constant of
proportionality between the stress and strain [1]. The tensile strength of the FRC is
1 Note, the terms stiffness and elastic modulus are sometimes used interchangeably in this text.
In fact, stiffness is not the same as elastic modulus but they are related. This stiffness and elastic
modulus relationship is dependent on specimen dimensions and the load application. For instance,
in axial members, E is actually related to stiffness (measured along the member’s length) via E A/L .
2.2 Axial (Longitudinal) Modulus 15
Fibre orientation Fibre orientation
Longitudinal Longitudinal
Transverse Transverse
F
(a) (b)
Fig. 2.1 Fibre composite members subjected to axial load: a tension; and b compression
4,000 HS carbon fibre

Tensile strength HS carbon-epoxy composite
3,500 Epoxy
Tensile stress (MPa)
3,000
2,500
2,000
1,500
1,000 Failure strain
500
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
Fig. 2.2 Tensile stress-strain behaviour of a HS carbon-epoxy composite
the maximum stress measurement, whilst the failure strain is the strain at composite
fracture, corresponding here to the strain at the maximum stress value.
The longitudinal modulus of the composite, E cl , is calculated by the rule of mixtures

(RoM) [1], viz.
E cl = E m (1 − Vf ) + E f Vf (2.1)
where E f and E m are the elastic moduli of the fibres and matrix, respectively, and Vf
is the fibre volume fraction.
Note that tensile and compressive stiffness measurements of the composite are found
to be similar, but some minor discrepancies exist, e.g. a difference in tensile modulus
of only about 10% is observed for carbon-epoxy composites [6]. Thus, Eq. (2.1) is
relevant under both tension and compression load cases.
The elastic moduli data for common fibres and matrices are readily available in
the literature and representative values have been presented in Tables 1.1 and 1.2.
These material properties are provided to enable the calculation of design estimates
(i.e. mechanical properties) for FRCs, and ultimately to size the composite. Note,
the term size is used here to mean ‘identify the requisite number of plies’.
Volume fractions are used to describe the proportion of fibres in the composite
structure. The upper theoretical or threshold fibre volume fractions are considered
later (see Sect. 2.8), whilst sensible (practically attainable) fibre volume fraction
estimates for each manufacturing process are considered in Chaps. 3 and 4.
Example 2.1
A unidirectional composite consists of glass fibres (E f = 69 GPa) embedded
in a polyester matrix (E m = 2.8 GPa). Assume Vf = 0.4. Calculate the longi-
tudinal elastic modulus for the composite.
Solution
The longitudinal modulus is computed from
E cl = E m (1 − Vf ) + E f Vf
= 3 × 109 (1 − 0.4) + 69 × 109 × 0.4
= 29.3 × 109 N/m2 = 29.3 GPa
Note. E cl is fibre-dominated; the fibres support nearly all of the applied load
with only a very minor matrix contribution [1]. Thus, we can simplify the RoM
calculation further by using only the fibre characteristics to offer a sensible
(albeit, a slightly lower) estimate of E cl for the unidirectional laminate.
Thus
E cl ≈ E f Vf
≈ 69 × 109 × 0.4
≈ 27.6 × 109 N/m2 = 27.6 GPa
The obtained value is similar but slightly lower (<10%) than the previous
calculation of 29.3 GPa.
2.3 Transverse Modulus 17
Fig. 2.3 Composite member F

with fibres orientated
perpendicular to the tensile
load
Fibre orientation
Longitudinal
Transverse
F
2.3 Transverse Modulus
So far, we have only considered composite members with fibres orientated parallel
(along the length of the member) to the applied load. Now consider the situation
where the fibres are orientated perpendicular to the loading direction in the transverse
direction—see Fig. 2.3.
In this situation, it is the matrix (rather than the fibres) that dominates the stiffness
response, and the transverse modulus E ct is given by [1]2
1 1 − Vf Vf
= + (2.2)
E ct Em Ef
The transverse modulus expression in Eq. (2.2) is sometimes referred to as the inverse
rule of mixture (IRoM) [8, 9]. Whilst the RoM expression in Eq. (2.1) assumes
the internal strains in the fibres and matrix are equal, referred to as iso-strain (or
sometimes the Voigt model [10]), this IRoM (Eq. (2.2)) assumes the internal stresses
are the same in the fibres and matrix, termed iso-stress or sometimes the Reuss
model [11].
Example 2.2
A unidirectional carbon fibre-reinforced epoxy composite has a fibre volume
fraction of 0.5. Calculate the elastic modulus of the composite in the longitu-
dinal and transverse directions. Assume the moduli of elasticity of the carbon
fibre reinforcements and the epoxy matrix are 230 GPa and 4 GPa, respectively.
2 A significant underestimate is known to result from this simple expression but it is considered here
(in this introductory text) to offer a sensible design estimate. If necessary, a more accurate estimate
can be calculated using the more complex, semi-empirical Halpin-Tsai expression in [7].
Solution
The longitudinal modulus is calculated from
E cl = E m (1 − Vf ) + E f Vf
= 4 × 109 (1 − 0.5) + 230 × 109 × 0.5
= 117 × 109 N/m2 = 117 GPa
The transverse modulus is calculated from

1 (1 − Vf ) Vf
= +
E ct Em Ef
1 (1 − 0.5) 0.5
= +
E ct 4 × 109 230 × 109
1
= 0.127 × 10−9
E ct
E ct = 7.9 × 109 N/m2 = 7.9 GPa
2.4 Shear Modulus
Figure 2.4 shows a FRC subjected to in-plane shear. The shear modulus is the ratio
of the shear stress, τ , to the shear strain, γ [1]. Shear stress and shear strain are
proportional, and the shear modulus (like the elastic modulus for the tensile and
compressive load cases) is the constant of proportionality [12]. The shear modulus
of the composite, G c , can be estimated using a similar expression to the transverse
modulus [5], via
1 1 − Vf Vf
= + (2.3)
Gc Gm Gf
where G f and G m are the shear moduli of the reinforcing fibres and matrix, respec-
tively.
Fig. 2.4 A unidirectional

element subjected to in-plane
shear stress: a undeformed;
and b deformed
(a) (b)
2.4 Shear Modulus 19
Example 2.3
A unidirectional E-glass fibre-reinforced epoxy composite is required to have
an axial modulus of more than 35 GPa, and a shear modulus of at least 2.5 GPa.
Table 2.1 provides the mechanical properties for the E-glass fibres and epoxy,
respectively. What is the minimum fibre volume fraction that is needed?
Table 2.1 Fibre and matrix properties
Material Elastic modulus Shear modulus
(GPa) (GPa)
E-glass 70 28.7
Epoxy 3.2 1.2
Solution
First, consider the composite’s longitudinal modulus
E cl = E m (1 − Vf ) + E f Vf
35 ≤ 3.0(1 − Vf ) + 70Vf
∴ Vf ≥ 0.48
A volume fraction greater than 0.48 is acceptable to satisfy the axial stiffness
requirement.
Now, let’s consider the shear modulus
1 1 − Vf Vf
= +
Gc Gm Gf
1 1 − Vf Vf
≥ +
2.5 1.2 28.7
∴ Vf ≥ 0.55
The minimum volume fraction to satisfy the shear stiffness requirement is

0.55.
Thus, to satisfy both design conditions, a fibre volume fraction, Vf , of at least
0.55 is needed (i.e. a minimum Vf of 0.48 and 0.55 for axial and shear stiffness,
respectively).
2.5 Effect of Fibre Orientation
Of course, fibre-reinforced composites are not always loaded either parallel or per-
pendicular to the fibres, for example, see Fig. 2.5. It is therefore necessary to briefly
consider the effect of loading angle on the elastic and shear moduli of the composite.
The effect of fibre orientation angle θ on the elastic and shear moduli of a FRC
is shown in Fig. 2.6. The longitudinal (0◦ ) and transverse (90◦ ) elastic moduli were
previously calculated in Example 2.2, whilst the in-plane shear modulus has been
calculated from Eq. (2.3) assuming typical fibres and matrix values. The off-axis
elastic and shear moduli values are calculated based on reference [5]; note that the
calculation of off-axis elastic and shear moduli values requires an understanding of
the Poisson effects. In Fig. 2.6, νc is Poisson’s ratio for the FRC material defined for
a strain applied in the longitudinal direction [5].
The off-axis elastic modulus E cθ (see Fig. 2.6) initially diminishes rapidly from the
longitudinal E cl (E cθ , θ = 0◦ ) value as the fibre orientation angle θ increases. At an
angle of 10◦ , the off-axis elastic modulus has reduced to almost half the longitudinal
modulus. The rate of decay of the elastic modulus reduces as the fibre angle increases
further but at 45◦ , the magnitude approaches the same modulus as that for transverse
loading (E cθ , θ = 90◦ ); a magnitude of about 10% of the maximum is typical for
carbon fibre composites [13]. In contrast, the shear modulus G cθ is a minimum at
an angle of 0 and 90◦ , and maximum at 45◦ . So, put simply, in designing optimised
composite structures or parts, aligned fibres (E cθ , θ = 0◦ ) are recommended to resist
tensile or compressive loads whilst fibres should be orientated at ±45◦ for shear loads.
The maximum shear modulus can be calculated from stress and strain transfor-
mations [5] to be
Fibre orientation Fibre orientation
Longitudinal Longitudinal
Transverse Transverse
(a) (b)
Fig. 2.5 Fibre-reinforced composite member with fibres orientated at an angle θ: a tensile loading;
and b in-plane shear
2.5 Effect of Fibre Orientation 21
120 c
Elastic or shear modulus (GPa)

c
100
80
60
40
20
0
0 10 20 30 40 50 60 70 80 90
Angle ( )
Fig. 2.6 Effect of loading angle on elastic and shear moduli of a unidirectional carbon-epoxy
composite (E cl = 117 GPa, E ct = 7.9 GPa, G c = 2.95 GPa and νc = 0.265)
1 1 + 2νc 1
◦
= + (2.4)
G c 45 E cl E ct
where νc can be calculated from a modified version of the RoM equation (see Eq. 2.1).
2.6 Strength of Composites
A composite will fail when the normal or shear stress exceed the strength values.
There are three obvious failure modes:
• Axial (longitudinal) failure.
• Transverse failure (of the matrix or fibre-matrix interface).
• Shear failure (of the matrix due to parallel shear forces).
These failure modes are shown schematically in Fig. 2.7. In the figure, the axial
tensile and transverse failures are shown for tensile loading only.
Axial (Longitudinal) Failure.
The longitudinal strength of a unidirectional composite depends on whether the
applied load is tensile or compressive. If the load is tensile, the Kelly-Tyson (KT)
model [14] (which is based on the RoM expression) can be used to predict composite
failure. Measured compression longitudinal strengths, however, are usually lower
than those noted in tension [15]—often equating to 50–60% [16] (or less [4]) of
Shear
Transverse failure
Axial failure
failure
(a) (b) (c)
Fig. 2.7 Fibre-reinforced composite failure modes: a axial tensile failure; b transverse tensile
failure and c in-plane shear failure
the tensile strength. The failure mechanism is the main reason for this behaviour;
microbuckling of fibres govern the longitudinal strength under compression [4].
In axial tension, if we assume the fibres fail first (prior to the matrix), the longitudinal
strength of the FRC σcl∗ is defined by the Kelly-Tyson (KT) model [14] as
σcl∗ = σm(f∗ ) (1 − Vf ) + σf∗ Vf (2.5)
where σf∗ is the failure strength of the fibres and σm(f∗ ) is the stress in the matrix at
fibre failure.
If the matrix fails first, it can be assumed that the fibres continue to resist the applied
tensile load after the matrix has failed; the matrix therefore provides no contribution
and the failure strength is [5]
σcl∗ = σf∗ Vf (2.6)
The matrix contribution to the tensile strength is small in comparison to the fibre
contribution and therefore it follows that the axial (longitudinal) strength of the com-
posite, irrespective of whether the fibres or matrix fail first, will be fibre-dominated.
Example 2.4
The unidirectional carbon fibre-epoxy composite in Example 2.2 is axially
loaded in tension until failure. The tensile strength of the HS carbon fibres and
the epoxy matrix is 4000 MPa and 65 MPa, respectively. Calculate the failure
strength of the composite assuming the stress in the epoxy matrix is 30 MPa
when fibre failure occurs.
2.6 Strength of Composites 23
Solution
The longitudinal tensile strength is calculated from
σcl∗ = σm(f∗ ) (1 − Vf ) + σf∗ Vf

= 30 × 106 (1 − 0.5) + 4000 × 106 × 0.5
= 2015 × 106 N/m2 = 2015 MPa
If the matrix stress at fibre failure was unknown, we could conservatively

neglect the matrix contribution and still provide a reasonable estimate of the
failure strength, via
σcl∗ = σf∗ Vf
= 4000 × 106 × 0.5 = 2000 × 106 N/m2 = 2000 MPa
This value is close to the 2015 MPa previously calculated when including the
matrix stress. Neglecting the matrix contribution provides a result that is akin
to a matrix-fail-first scenario. Put simply, the matrix contribution is minimal.
Transverse Failure. Transverse to the fibres, the internal stresses in the fibre-
reinforced composite will be the same in the fibres and matrix, and hence transverse
failure is likely to occur in the weaker matrix (or often at the fibre-matrix interface).
In contrast to the axial strength, similar (relatively low) strength values are noted
transverse to the fibres, irrespective of whether the load is tensile or compressive [5].
Most importantly, the strength of the composite will not exceed the matrix strength.
Often, however, the measured strength will be significantly less than that of the
matrix, viz.
σct∗ σm∗ (2.7)
Shear Failure. There is no simple analytical expression for estimating the shear
strength of a unidirectional composite based on the fibre volume fraction [5]. In the
absence of empirical data, the (in-plane) shear strength of a unidirectional composite
can (for Vf values below 0.7) conservatively be considered to be approximately equal
to that of the matrix, i.e. τc∗ ≈ τm∗ [17]. Moreover, using the Tresca criteria, we can
relate the shear strength of the matrix to its tensile strength via [18, 19]
τc∗ ≈ τm∗ = 0.5 × σm∗ (2.8)
Failure Under Off-Axis Loading. Evaluating the strength of a UD composite sub-

jected to off-axis loading (0◦ < θ < 90◦ ) is more complex than for the case where
fibres are orientated either parallel or perpendicular to the applied loads; a detailed
description is therefore beyond the scope of this text. The complexity arises from
an interaction between the normal and shear components [20]. In general, it is rea-
sonable to assume that fibre orientations influence normal and shear strengths in a
similar way to the elastic and shear moduli. The maximum tensile or compressive
strengths occurs when the fibres are aligned parallel to the normal load (i.e. at 0◦ ),
whilst the maximum shear performance occurs with fibre orientations at ±45◦ to
the shear forces. So, whilst fibres are most often aligned in a composite structure at
0◦ and 90◦ to resist longitudinal and transverse loads, respectively, some fibres are
orientated at ±45◦ . These ±45◦ fibres offer minimal resistance to longitudinal and
transverse loads, but are much better at resisting shearing and/or torsional loads [13].
Example 2.5
E-glass, S-glass and carbon fibre are available for the design of a unidirectional
reinforced epoxy composite. The composite must have a longitudinal modulus
of at least 50 GPa and a strength of more than 1000 MPa. Mechanical properties
for each of the fibre types and the epoxy are given in Table 2.2. Assuming a
maximum fibre volume fraction of 0.5 can be achieved, which of the fibres are
suitable?
Table 2.2 Mechanical properties of the constituents (fibres and matrix)
Material Elastic modulus Tensile strength
(GPa) (MPa)
Carbon fibre 230 4900
E-glass 70 2500
S-glass 95 4600
Epoxy matrix 3.2 60
Solution
Based on the values calculated in Table 2.3, only carbon fibre satisfies both
criteria.
Table 2.3 Mechanical properties of the fibre-reinforced composites
Material Elastic modulus check Tensile strength check
Ecl =Em (1 − Vf) + Ef V f σ ∗cl = σ m(f*) (1 − V f ) + σ ∗f V f ≈
σ ∗f V f
Carbon fibre 116.6 GPa ✓ 2450 MPa ✓
E-glass 36.6 GPa ✗ 1250 GPa ✓
S-glass 49.1 GPa ✗ 2300 GPa ✓
2.7 Woven and Random Fabrics 25
2.7 Woven and Random Fabrics
So far, the focus of this chapter has been on unidirectional fibre reinforcements
but many laminates also incorporate woven fabric into their structure (e.g. modern
skis [1]). Some composite structures even incorporate randomly orientated fibres,
especially for lower stress applications [21]. So, how can we consider these other
fabric forms?
Woven Fabrics. A woven ply contains fibres orientated in both the warp and weft
directions (i.e. perpendicular to each other) [3]—see Fig. 1.2. If half of the fibres
are orientated in the warp direction and the other half in the weft orientation, Vf is
half the total fibre volume fraction in each orientation; this is sometimes referred to
as a balanced fabric. To provide a simple approximation of the elastic moduli for
balanced woven cloths, similar stiffnesses can therefore be assumed in both the warp
and weft directions. The elastic modulus values are then calculated using the rule of
mixtures approximation from Eq. (2.1) with a fibre orientation efficiency factor, ηo ,
of 1/2 included, via
1
E c = E m (1 − Vf ) + E f Vf (2.9)
2
Similarly, the tensile strength can be approximated from Eq. 2.5 using the fibre
orientation efficiency factor. The shear modulus and in-plane shear strength of a
woven composite will be somewhat comparable to a UD laminate [18] since the
shear performance of a UD composite at either 0 or 90◦ orientations is the same—
see Fig. 2.6.
Random Mats. Since random fabrics have fibres orientated in all directions, their
in-plane elastic modulus can be considered to be the same in all orientations (quasi-
isotropic [22]). In other words, random fibre composites can be considered to exhibit
isotropic elastic properties. The limitation with this fibre form is that the elastic
modulus is usually lower than for the other aligned fibre forms. Moreover, fibre
volume fraction tends to be lower due to packing issues. An orientation efficiency of
3/8 [23] is frequently used, yielding
3
E c = E m (1 − Vf ) + E f Vf (2.10)
8
Example 2.6
A glass fibre-reinforced laminate comprises 40 vol% woven fibres in an epoxy
matrix. The glass fibres and epoxy matrix have elastic moduli of 69 GPa and
3.5 GPa, respectively. The tensile strength of the glass fibres is 3450 MPa. Cal-
culate the elastic modulus and tensile strength in the warp and weft directions.
Solution
Assuming a balanced fabric the warp and weft moduli are calculated from
1
E c = E m (1 − Vf ) + E f Vf
2
1
= 3.5 × 109 (1 − 0.4) + × 69 × 109 × 0.4
2
= 15.9 × 109 N/m2 = 15.9 GPa
and the warp/weft strengths are estimated from
1
σc∗ = σm(f∗ ) (1 − Vf ) + σf∗ Vf
2
Either the fibres or the matrix may fail first. This will depend upon the failure
strain of the constituents. Since the failure mode is not known, we can con-
servatively neglect the matrix contribution. Remember, even if the fibre fails
first, the matrix contribution is likely to be relatively small and will result in a
slightly lower failure strength. An underestimate of the composite’s strength
is better than an overestimate. Thus, conservatively we calculate the warp and
weft strengths as
1
σc∗ = × 3450 × 106 × 0.4
2
σc∗ = 690 × 106 N/m2 = 690 MPa
2.8 Volume Fraction
The calculation of elastic properties for a unidirectional fibre composite requires

knowledge of the fibre and matrix properties, as well as fibre volume fraction data.
In contrast, manufacturers usually supply dry fibres based on their weight (referred
to as areal weight; mass per unit area of a single ply), whilst prepreg suppliers often
specify an additional weight fraction of the matrix (or resin weight fraction [24]). It is
therefore essential to be able to convert these resin weight fractions into the requisite
fibre volume fractions needed for the above design estimates (i.e. for mechanical
property calculations).
An understanding of the mass density (mass per unit volume) of the fibres, ρf , and
matrix, ρm , is needed to be able to calculate fibre volume fractions from corresponding
weight fractions, Wf and Wm , and vice versa. Typical density values for matrices and
2.8 Volume Fraction 27
fibres were presented in Tables 1.1 and 1.2, respectively. The fibre volume fraction
can be calculated from the weight fraction of the fibres, via [5]
Wf
ρf
Vf = (2.11)
Wf
ρf
+ 1−Wf
ρm
In a similar manner, the fibre weight fraction can be calculated from the volume
fraction as [2, 5]
Vf ρf
Wf = (2.12)
Vf ρf + (1 − Vf )ρm
Example 2.7
A randomly orientated glass fibre-reinforced polyester composite has a fibre
weight fraction of 0.4 (i.e. 40 wt%). The density of the glass fibres and the
matrix are 2550 kg/m3 and 1130 kg/m3 , respectively. Calculate the fibre volume
fraction.
Solution
The fibre volume fraction is readily calculated from
Wf 0.4
ρf
Vf = = 2550
= 0.23
Wf
ρf
+ 1−Wf
ρm
0.4
2550
+ 1−0.4
1130
Example 2.8
A unidirectional HS carbon fibre-reinforced epoxy laminate has a fibre weight
fraction of 0.65. Estimate the mechanical properties (i.e. elastic modulus and
strength) of the carbon-epoxy composite under longitudinal tensile loading.
The fibre and matrix properties are shown in Table 2.4.
Table 2.4 Material properties of the carbon fibres and epoxy matrix
Material Density, ρ (g/cm3 ) Elastic modulus, E Tensile strength,
(GPa) σ ∗ (MPa)
Carbon fibre 1.76 230 4100
Epoxy matrix 1.18 3.2 55
Solution
The fibre volume fraction is calculated from
Wf 0.65
ρf
Vf = = 1760
= 0.55
Wf
ρf
+ 1−W
ρm
f 0.65
1760
+ 1−0.65
1180
The longitudinal elastic modulus is calculated from
E cl = E m (1 − Vf ) + E f Vf
= 3.2 × 109 (1 − 0.55) + 230 × 109 × 0.55
= 128.0 × 109 N/m2 = 128.0 GPa
Assuming the fibres fail first, the longitudinal strength is calculated from
σcl∗ = σm(f∗ ) (1 − Vf ) + σf∗ Vf
Since the stress in the matrix at fibre failure is not known, we can conservatively
neglect the matrix contribution, hence
σcl∗ = 4100 × 106 × 0.55 = 2255 × 106 N/m2 = 2255 MPa
Packing Arrangement. So far, we have used a relatively narrow range of typical

fibre volume fractions. It should be evident that there is a limit to the volume of fibres
that can be embedded into a composite. This limit is due to geometrical constraints
that result from the size and shape of the fibres, the packing arrangement and the
need for a surrounding matrix to transfer the load to the stiff and strong fibres. So,
what is this threshold (upper) limit?
Let us start by looking at two packing arrangements: a hexagonal and a simple
square arrangement of fibres—see Fig. 2.8. In the figure, the fibre cross-sections are
represented by circles and are assumed to have a uniform fibre diameter along their
length. All fibres have the same diameter and therefore the distance between the fibre
centres (s) in the arrangement must be separated by at least twice the fibre radius.
If we consider the square and hexagonal arrangements, we can show that the
square unit cell has a maximum fibre volume fraction of 0.785 whilst that of the
hexagonal arrangement is optimised at 0.907. The hexagonal unit cell offers the
maximum possible fibre volume fraction (and hence, the upper threshold limit). In
practice, however, several factors can influence fibre volume fractions, making these
theoretical values unlikely to be attained when manufacturing a FRC. The most
important factors include fibre form and the manufacturing method. Table 2.5 offers
sensible fibre volume fraction target ranges for unidirectional, woven and random
fibre mats.
2.9 Summary 29
r
r
s
60 s
(a) (b)
Fig. 2.8 Fibre packing arrays: a square arrangement; and b hexagonal arrangement
Table 2.5 Practical fibre volume fractions for random, woven and unidirectional composites
Fibre form Volume fraction
(Vf )
Random 0.1–0.3
Woven 0.2–0.55
Unidirectional 0.3–0.7
2.9 Summary
Table 2.6 provides a summary of formulae for estimating the mechanical properties
of FRCs.
Table 2.6 Design formulae for fibre-reinforced composites

Composite Type Property Formula
Unidirectional Longitudinal elastic modulus E cl = E m (1 − Vf ) + E f Vf
Transverse elastic modulus 1
E ct = 1−V
Em + Ef
f Vf
Longitudinal failure strength σcl∗ = σm(f∗ ) (1 − Vf ) + σf∗ Vf ≈ σf∗ Vf

Transverse failure strength σc∗ = σm∗
In-plane shear strength τc∗ = τm∗
Woven Elastic modulus Ec = E m (1 − Vf ) + 21 E f Vf
Random mat Elastic modulus Ec = E m (1 − Vf ) + 38 E f Vf
2.10 Questions
Question 2.1 Define the term anisotropic.
Question 2.2 How should the fibre reinforcements be orientated to produce the
stiffest and strongest composite in tension and compression?
Question 2.3 The tensile strength of a unidirectional composite is higher than the
compressive strength. The compressive strength is often what proportion of the tensile
strength?
Question 2.4 How should the fibre reinforcements be orientated to produce the
stiffest and strongest composite in shear?
Question 2.5 A fibre orientation efficiency factor is used to approximate the elastic
moduli of woven and randomly orientated fibres. Why?
Question 2.6 Practically, what fibre volume fraction ranges can be achieved for
unidirectional, woven and random fibre composites?
Question 2.7 Fibre volume fractions for unidirectional composites are normally
higher than their random or woven counterparts. Why?
Question 2.8 Why is it important to be able to convert weight fractions into fibre
volume fractions and vice versa?
2.11 Problems
Problem 2.1 An aligned FRC consists of E-glass fibres in an epoxy matrix. The
volume fraction of E-glass fibres is 0.5. The fibres have a modulus of 69 GPa whilst
the epoxy has a modulus of 3.8 GPa. Calculate the composite’s axial (longitudinal)
and transverse elastic moduli.
Answer: 36.4 GPa, 7.2 GPa.
Problem 2.2 Determine the elastic moduli (longitudinal and transverse) of a unidi-
rectional carbon fibre-reinforced epoxy composite with 50 vol% fibres. Assume that
the fibres and epoxy have an elastic moduli of 230 GPa and 3.8 GPa, respectively.
Answer. 116.9 GPa, 7.5 GPa.
Problem 2.3 A random FRC consists of 30 vol% glass fibres in a polyester matrix.
Again, the E-glass fibres can be assumed to have an elastic modulus of 69 GPa
whilst the polyester has an elastic modulus of 3.0 GPa. Calculate the in-plane elastic
modulus of the composite.
Answer. 9.9 GPa.
2.11 Problems 31
Table 2.7 Material properties for the glass-polyester composite

Material Elastic modulus Tensile strength
(GPa) (MPa)
E-glass 70 2500
Polyester matrix 3.3 50
Problem 2.4 A tensile test is conducted on an aramid-epoxy sample 20 mm wide

and 1 mm thick. The aramid composite has 50 vol% of unidirectional fibres in an
epoxy matrix. If the tensile strength of the aramid fibres is 3600 MPa and the epoxy
matrix is 50 GPa, estimate the axial failure load. Assume the stress in the matrix at
fibre failure is 35 MPa.
Answer. 36.4 kN.
Problem 2.5 A UD glass-polyester composite has been selected for a structural tie in
a truss. The mechanical properties of the E-glass and polyester are given in Table 2.7.
It is believed that a fibre volume fraction of 0.35 can be consistently achieved during
the manufacturing process. The tie is 50 mm wide and approximately 1 m long. If
the composite tie is loaded to 70 kN and each ply is 0.34 mm thick, how many plies
are needed to ensure
a. tensile failure does not occur?
b. the extension of the tie does not exceed 30 mm?
Note. Axial stiffness F/δ = AE/l where F is the tensile force, δ is the extension, A is
the cross-sectional area, E is the elastic modulus and l is the length.
Answer. a. 5 plies; b. 6 plies.
Problem 2.6 A short rectangular-section beam is to be designed with the fixed width
of 15 mm and a length of 0.2 m. The beam will be manufactured from a unidirectional
carbon fibre-epoxy composite. The tensile modulus of the composite is 110 GPa, and
the tensile and compressive strengths are 1750 MPa and 1000 MPa, respectively. If the
beam is subjected to three-point bending test with a midspan load of 1 kN, determine
the minimum number of 0.25 mm thick plies, to ensure
a. Longitudinal (tensile or compressive) failure does not occur.
b. The beam does not exhibit a midspan deflection of more than 3 mm.
3
Note. The second moment of area for a rectangular cross-section is I = bh 12
where
b is the width and h is the height of the cross-section. Maximum deflection occurs
Fl 3
at the midspan of the three-point bending test, and it is equal to δmax = 48E I
where
l is the length of the beam, and F is the point load.
Answer. a. 3 plies; b. 30 plies.
References
2. Mallick PK (2007). Fiber-reinforced composites: materials, manufacturing, and design, 3rd
edn. Mechanical engineering. CRC/Taylor & Francis, Boca Raton
6. Adams D (2019) Optimum unidirectional compression testing of composites. www.
compositesworld.com/articles/optimum-unidirectional-compression-testing-of-composites
7. Halpin JC, Tsai SW (1967) Environmental factors in composite design. air force materials
laboratory
8. Javanbakht Z, Hall W, Virk AS, Summerscales J, Öchsner A (2020b) Finite element analysis
of natural fiber composites using a self-updating model. J Compos Mater 54(23):3275–3286.
https://doi.org/10.1177/0021998320912822
9. Javanbakht Z, Hall W, Öchsner A (2020a) An element-wise scheme to analyse local mechanical
anisotropy in fibre-reinforced composites. Mater Sci Technol 36(11):1178–1190. https://doi.
org/10.1080/02670836.2020.1762296
10. Voigt W (1889) Ueber die beziehung zwischen den beiden elasticitätsconstanten isotroper
körper. Annalen der Physik 274(12):573–587. https://doi.org/10.1002/andp.18892741206
11. Reuss A (1929) Berechnung der fließgrenze von mischkristallen auf grund der plastizitäts-
bedingung für einkristalle. ZAMM - Zeitschrift für Angewandte Mathematik und Mechanik
9(1):49–58. https://doi.org/10.1002/zamm.19290090104
12. Hibbeler RC (2014) Statics and mechanics of materials, 4th edn. Pearson, Upper Saddle River
N.J
13. Hart-Smith LJ (1992) The ten-percent rule for preliminary sizing of fibrous composite struc-
tures. Weight Eng 52:29–45
14. Kelly A, Tyson WR (1965) Tensile properties of fibre-reinforced metals: copper/tungsten
and copper/molybdenum. J Mech Phys Solids 13(6):329–350. https://doi.org/10.1016/0022-
5096(65)90035-9
15. Adams D (2017) Can flexure testing provide estimates of composite strength properties? www.
compositesworld.com/articles/can-flexure-testing-provide-estimates-of-composite-strength-
properties
16. Sun W, Guan Z, Li Z, Zhang M, Huang Y (2017) Compressive failure analysis of unidirectional
carbon/epoxy composite based on micro-mechanical models. Chin J Aeronaut 30(6):1907–
1918. https://doi.org/10.1016/j.cja.2017.10.002
17. DoITPoMS (2019) Strength of long fibre composites. www.doitpoms.ac.uk/tlplib/fibre_
composites/strength.php
18. Medina C, Canales C, Arango C, Flores P (2014) The influence of carbon fabric weave on the
in-plane shear mechanical performance of epoxy fiber-reinforced laminates. J Compos Mater
48(23):2871–2878. https://doi.org/10.1177/0021998313503026
19. Öchsner A (2016) Continuum damage and fracture mechanics, 1st edn. Springer, Singapore,
Imprint: Springer Singapore
20. Clyne (2019) An introduction to composite materials. Cambridge University Press, Cambridge
22. Lokensgard E (2010) Industrial plastics: theory and application, 5th edn. Delmar Cengage
Learning, Clifton Park NY
23. Krenchel H (1964) Fibre reinforcement; theoretical and practical investigations of the elasticity
and strength of fibre-reinforced materials. Akademisk forlag
24. Lavender Composites (2017) Prepeg stock list. http://www.lavender-ce.com/wp-content/
uploads/prepreg-ex-stock-101017.pdf
Chapter 3
How to Make a Composite—Wet Layup
Abstract This chapter introduces the six basic steps needed to design and make a
fibre-reinforced composite (FRC): (1) fibre and matrix selection; (2) mould prepara-
tion; (3) layup and consolidation; (4) curing (and post-curing); (5) demoulding; and
(6) post-processing (finishing). These six steps are considered using a simple wet
layup process for a flat unidirectional (UD) composite. The mechanical performance
of the composite is estimated based on the rule of mixtures (RoM), inverse rule of
mixtures (IRoM) and Kelly-Tyson (KT) models from Chap. 2. A description of the
wet layup process is offered and discussed in the context of the options available to a
composite fabricator. The chemical crosslinking processes for polyester, vinylester
and epoxy are presented. Consideration is given to demoulding the FRC and to the
post-processing options. The use of gelcoat, flow coat and paint are mentioned in
the context of a broader discussion on surface finishing of FRC. The outcome of the
chapter is a step-by-step design and manufacturing method that is easily replicated.
Caution!
The manufacture of FRC structures demands the use of fibres and resins. The
abrasive fibres, when cut or sanded, will result in fine dust particles that can
cause skin irritations. Resins start as liquids and are mixed with either an initiator
(catalyst) or hardener which then cures to create the solid polymer matrix that
surrounds the fibres. The chemicals used in the curing process of thermosetting
polymers can cause harm. Many other chemicals, including cleaning solvents
and release agents, may also be used in the fabrication process [1]. For the
composite manufacturing tasks covered in this text, it is the fabricator’s sole
responsibility to familiarise themselves with the associated risks for each of the
activities.

https://doi.org/10.1007/978-3-030-78807-0_3
34 3 How to Make a Composite—Wet Layup
Fig. 3.1 The six basic steps 1. Fibre and matrix selection
to make a FRC
• Glass, carbon, or aramid?
• Unidirectional, woven or random fabric?
• Number of plies and their fibre orientations?
• Polyester, vinylester or epoxy?
2. Mould preparation
• What release agent?
• How should it be applied?
• How many coats of agent?
3. Layup and consolidation

• What layup process will be used?
• How will the composite be consolidated?
• What shape/size of fibres should be cut?
• How much resin and hardener/catalyst mix?
4. Curing (and post-curing)

• What resin curing conditions should be used?
• How long will the curing process take?
• Is a post-cure needed? If so, what conditions?
5. Demoulding the composite

• How will demoulding occur?
• What potential issues might arise?
• How might these issues be resolved
6. Post-processing (finishing)
• Does the part need trimming?
• Is the surface texture/finish suitable?
• Is there a need for a protective surface?
3.1 Introduction
There are six basic steps to making a fibre-reinforced composite (FRC), as shown
in Fig. 3.1. This step-by-step process is adapted from the six stages reported by
Wanberg [2] and is considered in this chapter in the same logical and sequential
manner. The sequence in Fig. 3.1 assumes that a mould for forming the composite is
already available—design and construction of moulds for making composite parts is
considered in Chap. 6. The six steps are common for basic manufacturing methods
such as wet layup [2, 3] (also called hand lamination [4] or layup moulding [5]) as
well as for more complex moulding techniques such as vacuum bagging [3, 5–7]
and prepreg layup [3, 5] (referred to here as prepreg moulding [7, 8]). Wet layup, as
the most basic method, is introduced in this chapter to provide context to the six-step
process, whilst vacuum bagging and prepreg moulding methods are discussed and
linked back to the same six steps in Chap. 4.
3.1 Introduction 35
Herein, the six steps are described for a flat UD carbon FRC panel. The manufac-
turing process for the FRC is used to facilitate a broader discussion of the manufac-
turing options available to the composite fabricator.
3.2 Fibre and Matrix Selection
To design a composite structure (or part), the applied loads and in-service conditions
must be known. The reinforcing fibres and matrices are chosen to meet these structural
(and environmental or in-service) requirements. The fibre and matrix selection step
therefore relates to the ‘Mechanics of Composite Structures’ topic, covered earlier
in Chap. 2.
To meet the structural requirements, the following considerations need to be
addressed [9]:
• Fibre and matrix selections.
• Fibre volume fraction (linked to the manufacturing process).
• Fibre orientation in each layer or ply (and the stacking sequence).
• The number of plies in each orientation (and hence, the laminate thickness).
Here, to illustrate the wet layup process, UD carbon fibres (200 g/m2 ) [10] and an
epoxy matrix [11] are chosen. The UD fabric is cut to approximately 0.3 m squared.
Since wet layup is used, a relatively low fibre volume fraction is likely in comparison
to other more advanced moulding methods since minimal consolidation pressure
is provided during the wet layup process. An increase in consolidation pressure
during moulding will reduce laminate thickness and hence, increase fibre volume
fraction [12].
Typical fibre volume (and weight) fractions for wet layup (based on the fibre type
and form) are shown in Table 3.1. Here, we assume Vf ≈ 0.4: a mid-range value
typical for wet layup of unidirectional composites. Four layers (plies) of carbon fibre
are stacked and orientated in the same direction to create a UD FRC of approximately
1 mm thickness—see Eq. (5.5). These fibre and matrix selections are made to produce
a composite that is stiff (has a high elastic modulus) and strong in the longitudinal
or fibre direction, but is significantly more compliant in the transverse direction.
Similar fibre and matrix selections will be used later in Chap. 4 for vacuum bagging
and prepreg moulding to illustrate the comparative performance of wet layup against
these other more advanced moulding methodologies (see Chap. 5).
Example 3.1
The wet layup method is used in this chapter to create a UD carbon fibre-
reinforced epoxy laminate with a target fibre volume fraction Vf ≈ 0.4. Use
Eqs. (2.1)–(2.6) to estimate the elastic moduli and the tensile strengths (lon-
gitudinal and transverse) of the FRC. The fibres and resin are identified in
Refs. [10, 11].
Solution
Wet layup is a simple manufacturing process but is likely to produce relatively
low fibre volume fractions (0.3–0.5) and hence, we assume that Vf ≈ 0.4 (see
Table 3.1).
The elastic moduli and strength of the fibres are reported to be 250 GPa and
5516 MPa, respectively [10]. The tensile modulus and strength of the epoxy
resin are not readily available; typical values have been estimated, based on
mechanical properties for similar commercial resin systems. The elastic mod-
ulus and tensile strength are assumed here to be 3.0 GPa and 50 MPa, respec-
tively.
Considering E m = 3.0 GPa and E f = 250 GPa, the longitudinal modulus is
calculated from
E cl = E m (1 − Vf ) + E f Vf ,
N
= 101.8 × 109 2 = 101.8 GPa
m
The transverse modulus is calculated from
1 (1 − Vf ) Vf
= + ,
E ct Em Ef
N
E ct = 5.0 × 109 2 = 5.0 GPa
m
Assuming σf = 5516 MPa, the longitudinal strength is calculated from
σcl∗ = σm(f∗ ) (1 − Vf ) + σf∗ Vf
or conservatively, neglecting the matrix contribution
σcl∗ = σf∗ Vf
N
= 2206.4 × 106 = 2206.4 MPa
m2
The transverse strength is calculated from
σct∗ σm∗
σct∗ 50 MPa
Note. Later (in Chap. 5), we find the measured Vf for the panel is actually 0.34,
rather than the targetted value of 0.4.
3.3 Mould Preparation 37
Table 3.1 Fibre fractions for wet layup (hand lamination) based on fibre type and form
Random mat Woven Unidirectional
Vf Wf Vf Wf Vf Wf
Glass 0.1–0.3 0.19–0.48 0.2–0.4 0.35–0.59 0.3–0.5 0.48–0.68
Carbon 0.1–0.3 0.14–0.38 0.2–0.4 0.27–0.49 0.3–0.5 0.38–0.59
Aramid 0.1–0.3 0.12–0.34 0.2–0.4 0.23–0.44 0.3–0.5 0.34–0.55
(a) (b) (c)
Fig. 3.2 Mould shapes and terminologies: a female; b male; and c matched-die mould
3.3 Mould Preparation
A FRC is usually formed in a mould, often referred to as tooling [3, 5, 13]. A

mould can be manufactured from a wide variety of materials. The design of the
mould depends on the composite structure to be created and on the number of parts
to be produced by the mould tool (production run). Further details about mould
design, tooling material options, how to create mould tools and other basic moulding
considerations are provided in Chap. 6.
As a starting point, let’s assume our mould tool is a flat plate. A flat plate will
produce a composite laminate with a single smooth surface on the mould side. The
fibres and surrounding matrix will be pressed against the flat mould surface. On the
reverse side of the composite, a less appealing surface finish will be produced as the
fibres are not pressed firmly against a surface. More complex mould tool shapes can
produce concave or convex surfaces, but still with a single smooth surface [13, 14].
A female mould is identified by its concave form and is used to create a (convex)
laminate with a smooth outer surface, whilst a male mould has a convex shape and
is used to create a smooth inner (concave) surface on the composite [3, 7, 14]. To
create a smooth surface on both sides normally requires a matched mould [7] or
matched-die moulding system [3, 14], i.e. a combination of both male and female
mould tools. However, it is also possible to simulate a two-sided mould with a simple
caul plate [14, 15]. A caul plate is a smooth metal or plastic plate (or sometimes a
rubber sheet) used to provide a uniform normal pressure to the laminate during
moulding and hence, a smooth surface on the ‘reverse’ side of the cured composite
part. Female, male and matched-die moulds are shown in Fig. 3.2. The aesthetic
requirements of the composite therefore play a significant role in the selection of a
suitable moulding system.
An appropriate mould preparation process is essential for the successful creation

of moulded composite structures. Mould tools should be clean and released prior
to commencement of the layup process (described in Sect. 3.4) [2, 16]. To remove
contaminants, a clean cloth (ideally, a lint-free cloth) and a solvent (for example,
acetone) are typically used. The mould tool should then be left to dry before the
release agent [17] (sometimes called parting agent [13]) is applied to the moulding
surface. The release agent is used to provide a thin uniform barrier between the mould
tool and the composite laminate, preventing permanent adhesion [18]. If a release
agent is not applied across the entire mould surface (or is not appropriately applied),
the laminate may stick to the mould, ruining the laminate and/or rendering the mould
tool unusable.
A laminate is more likely to stick to a new mould tool than a well- used one [19].
Thus, new moulds tend to require additional preparation whereas older (sometimes
referred to as seasoned [2]) moulds need fewer release coats. As a rule of thumb, 2–3
coats of release agent are usually needed on a seasoned mould, but a minimum of 4 is
suggested in [2] for new tooling. There are many types of release agents [20] but three
types of mould release that are readily available from most composite suppliers are:
• Wax.
• Polyvinyl alcohol (PVA).
• Semi-permanent release agents.
In this text, wax is exclusively used as the release agent. Standard wax mould release
is normally used for low-temperature curing processes (room-temperature cures)
but some high-temperature wax formulations are used later for elevated temperature
cures up to 120 ◦ C—see Chap. 4. In Fig. 3.3, standard mould release wax is applied
to the mould tool (flat glass plate). Masking tape is applied around the plate to define
the region of wax application. When applying the wax (or any other release agent),
careful consideration should be given to specific application instructions. Here, three
coats of wax are applied to the seasoned flat plate with circular motions. A small
quantity of wax is applied over the entire mould with a sponge. The wax is allowed
to dry (a haze is observed on the surface) and then the surface is buffed to a sheen
with a clean lint-free cloth. The surface is left for a further time period before the
next coat of wax is introduced—the requisite timings are specified in the product’s
instructions.
Polyvinyl alcohol and semi-permanent release agents are not used in this textbook,
but their widespread availability and popularity amongst many composite fabricators
means they (at the very least) warrant a brief comment—wax is used consistently
here solely as personal preference. Polyvinyl alcohol can be used on its own as a sole
release agent, or in some circumstances as a secondary mould release for composite
parts where demoulding is problematic. If needed, and depending on the PVA product
chosen, PVA can be wiped, brushed or sprayed on the mould surface [21].
Semi-permanent release agents are another alternative to wax that can help speed
up the release agent application process. Sometimes, these releases are referred to
as polymer release agents since they comprise one or more polymer resins dissolved
3.3 Mould Preparation 39
Fig. 3.3 Application of a

wax mould release
in a solvent [17]. Semi-permanent release agents can offer multiple moulding cycles
before reapplication of the release agent is necessary. If PVA or a semi-permanent
releases are applied to a mould tool, they should always be allowed to dry before the
composite layup process commences.
3.4 Layup and Consolidation (Hand Lamination)
The first surface layer of a laminate is an optional gelcoat (a layer of thickened

resin); typically, a gelcoat is 0.4–0.6 mm thick [22] but may be applied to a thickness
of up to 1 mm [23] or more. A gelcoat, whether polyester-, vinylester- or epoxy-
based, is used to provide a high-quality finish on the surface of the composite and to
protect the laminate from environmental factors. Most often, gelcoats are polyester-
or vinylester-based and tend to contain a pigment to provide a smooth coloured
aesthetic appearance [5]. The hard (often white) outer surface of a boat hull is an
example of a gelcoat finish on a glass fibre composite.
Gelcoats can be purchased directly from composite suppliers but if you wish to
formulate your own, resin thickeners (gelcoat additives) and pigments can in most
cases be added to standard lamination resins [2]. Where a gelcoat is used, a thin layer
is brushed or sprayed onto the mould tool surface. If a gelcoat is not applied, either
a clear flow (top) coat or paint [3, 24, 25] would normally be used to produce the
finished look and (perhaps most critically) provide environmental protection for the
composites. A flow coat or paint is applied as the last finishing process. Here, no
gelcoat or finish coats (neither flow coat or paint) are used, but finishing applications
on composite parts will be illustrated later in this textbook. Further information in
relation to surface preparation for flow coat or painted surface finishes is given in
Sect. 3.8.
(a) (b)
(c) (d)
Fig. 3.4 Lamination: a carbon fibre layup (first ply); b wetting out with a brush; c excess surface
resin; and d application of roller to remove air trapped between the plies
As no gelcoat is used here, the first carbon fibre ply is positioned on the flat
plate. The resin is mixed. During the mixing process, additional substances such as
pigments may be added [26] without an adverse effect on the resin characteristics (or
finished composite). The premixed resin (in our case, epoxy [11]) is introduced to
the carbon fibre; an excess of resin is sometimes added to the first ply to ensure resin
adequately impregnates the fabric. A brush is often used (a spreader is an alternative
option) to ensure a uniform distribution of resin across the surface of each of the
carbon fibre plies. The manual fibre positioning and the introduction of the resin are
shown in Fig. 3.4a and b, respectively. In some cases, an alternative wet-out process
is employed; a small quantity of resin is applied to the mould surface prior to the
placement of the first ply—this is an alternative method to ensure a smooth surface
finish on the mould side of the composite. The wetting-out process is repeated for
each ply. After each layer is wet out, a roller is used to assist resin impregnation and
expel trapped air —see Fig. 3.4d. A roller will have minimal effect on consolidation.
Wet layup is relatively easy to perform, but to achieve the best results, it does require
some practise. It is easy to use too much resin and add unnecessary mass to the
composite. So, how much resin is needed?
3.4 Layup and Consolidation (Hand Lamination) 41
The mass of the resin (and catalyst/hardener mix) can be calculated from the area of
the laminate ( A), the number of layers (n), the areal weight of the fibres ( Aw ) and
the intended weight fraction of the matrix (Wm ) [27]
A × n × A w × Wm
Resin and hardener mix (g) = (3.1)
(1 − Wm )
or in terms of fibre weight fraction
A × n × Aw × (1 − Wf )
Resin and hardener mix (g) = (3.2)
Wf
These equations offer simple estimates of the mass of the resin and hardener mix
needed to create the laminate, but they do not account for any resin wastage that
may occur during fabrication. A factor between 1 (assuming no waste) and 1.5 is
therefore usually introduced to take account of resin wastage [27]. The factor is
usually chosen with an understanding of the composite layup and is based on the
fabricator’s expertise.
Example 3.2
Now, using Eq. (3.2), calculate how much resin and how much hardener are
needed to fabricate the UD carbon fibre-reinforced epoxy laminate shown in
Fig. 3.4.
Solution
As mentioned, the fibres were cut to 300 mm long × 300 mm wide. The surface
area of the fibre-reinforced panel was therefore (0.3 m squared) = 0.09 m2 .
The fibre volume fraction was previously chosen based on the unidirectional
carbon fibre values for wet layup in Table 3.2, i.e. Vf = 0.4 which is at the
midpoint between the upper and lower threshold values.
Thus, Wf = 0.49 and hence
A × n × Aw × (1 − Wf )
Resin and hardener mix (g) =
Wf
0.09 × 4 × 200 × (1 − 0.49)
Resin and hardener mix (g) = = 75 g
0.49
To account for wastage, a factor of 1.25 (between 1 and 1.5) is used for the
resin and hardener mix.
Resin and hardener mix (g) = 75 × 1.25 = 94 g
The epoxy resin used here has a resin to hardener ratio of 5:1, i.e. 5/6 of resin
and 1/6 of hardener
Resin (g) = 94 × 5/6 = 78.3 g

Hardener (g) = 94 × 1/6 = 15.7 g
Note. An appropriate selection of the wastage factor is an important skill for a

composite fabricator but it is simply a sensible estimate (based on experience)
for the composite being fabricated. Adding extra resin will, of course, reduce
Vf whilst not enough will result in a dry composite with suboptimal properties.
3.5 Curing and Post-Curing
Once the layup process is complete, the panel is left to cure. There will be chemical
and structural changes during the curing process, and the resin will transform from
a low molecular weight liquid into a solid polymer [28]. The cure time depends on
many factors but most need a curing time of at least 24 h at room temperature [29]
to reach the solid state.
During curing, heat will be discharged (referred to as an exothermic reaction [30])
and the viscosity of the resin will increase. The resin starts to gel—this process
typically takes 30–40 min (sometimes more, sometimes less) and is dependent on
the type and volume of the resin and the ambient temperature. As a rule of thumb,
a 10 ◦ C temperature increase will half the working time [27]. Moreover, it should
be noted that a thick laminate can generate a considerable amount of heat and could
be a potential safety hazard, whilst minimal heat will usually result from a thinner
laminate; the peak exothermic temperature increases with thickness [5], so care
should be taken in the manufacture of thicker composite structures.
After the tacky, gelatinous stage, the composite will continue to cure and gradually
become firm (sometimes referred to as the green stage [2]) before finally reaching
the solid state. Once the solid state is attained, the laminate has sufficient stiffness
to be demoulded (see Sect. 3.7), but further crosslinking may still be needed. Epoxy
resins often benefit from a post-cure at elevated temperatures to reach their opti-
mum performance values [31]—see Ref. [27]. Here, a post-cure is performed after
demoulding. Post-curing can be performed before or after demoulding [32], but is
usually performed after removal from the mould [29]. A post-cure should always be
performed in accordance with the supplier’s instructions [5, 33].
3.6 Crosslinking: Chemical Steps 43
Fig. 3.5 The chemical structure of an isophthalic polyester and styrene (ester groups are high-
lighted). Adapted from [5]
3.6 Crosslinking: Chemical Steps
As mentioned, during curing, the liquid polymer is irreversibly transformed into

a solid, but what chemical changes occur during this process? The details of the
chemical crosslinking process for polyester, vinylester and epoxy are elaborated in
this section.1
Unsaturated Polyester. The production of polyester is usually the result of reacting
unsaturated and saturated acids and a glycol (an alcohol) to produce a short chain
polymer, which is usually a stable liquid. Adjusting the amount of unsaturated acid
allows for the creation of potential sites for crosslinking whereas the amount of
saturated acid and glycol determines the chemical and fire resistance [33].
The chemical structure of a typical unsaturated (isophthalic) polyester is shown in
Fig. 3.5. The intermolecular bonds of the polyester are represented as a line between
two atoms, i.e. a pair of shared electrons. These bonds are covalent and the term
unsaturated identifies the existence of carbon double bonds in the backbone of the
chain [5]. The letter ‘n’ identifies the part of the molecule (in square brackets) that
is repeated n-times (typically n = 3 to 6 [27]). The double bonds are the reactive
sites that provide a bridge between polyester chains during the crosslinking process.
A small molecule such as styrene (a volatile organic compound, VOC) is used as a
reactive diluent [5] to enhance the reactivity of the liquid polymer (and to reduce
cost) whilst also facilitating the crosslinking process without the evolution of by-
products. The styrene molecule shown in Fig. 3.5 acts as the bridge between the
polyester chains.
A simplified representation of the molecular chains during the curing process
is presented in Fig. 3.6. The polyester chains are shown as a simple 2D structure
but will (in reality) form a 3D arrangement. The curing process starts when all the
ingredients are available: the unsaturated polyester (containing the styrene diluent)
1 The term crosslinking should not be confused with polymerisation; the latter happens during
manufacturing of a polymer from monomers, whilst the former occurs during the moulding (or
curing) process when polymer chains join together [34].
styrene molecule
free radical R CH CH2
O O O O O O
Stage 1 HO C HC CH C O CH2 CH2 O C C O CH2 CH2 O C HC CH C O CH2 CH2 OH
O O O O O O
O R O O O O O
O R O O O O O
CH CH2
O R O O O O O
CH CH2
styrene group (only two molecules are shown)
CH2 CH
O O O O O O
HO C CH2 CH C O CH2 CH2 O C C O CH2 CH2 O C C C C O CH2 CH2 OH
Fig. 3.6 Simplified representation of crosslinking in an unsaturated polyester (ester groups are
highlighted). Adapted from [5]
and an initiator (typically referred to as a catalyst2 ). An accelerator is already included

in the unsaturated polyester to promote the crosslinking reactions.
The catalyst is usually peroxide-based and the most common is methyl ethyl ketone
peroxide (MEKP) [5]. The accelerator decomposes the initiator/catalyst into free
radicals (i.e. molecules with unpaired electrons) [33], which are denoted by Ṙ in
Fig. 3.6 (Stage 1). These free radicals seek to react with loosely bonded electrons in
the carbon-carbon double bond (Stage 2). When a free radical reacts with an elec-
tron from the double bond, the carbon-carbon double bond is broken and the free
radical forms a single bond with carbon. This leaves the carbon chain with a single
carbon-carbon bond and one unpaired electron (Stage 3). The unsaturated carbon-
carbon bond of a styrene molecule bonds with this unpaired electron (breaking its
double bond) and leaving an unpaired electron (Stage 4). As this process repeats,
2 The initiator is often referred to as the catalyst but, since its structure changes during the reaction,
this is strictly not accurate; the accelerators are technically the catalysts in the current context [33].
The term catalyst is employed in this text as this is the name typically used in the composite industry.
3.6 Crosslinking: Chemical Steps 45
Fig. 3.7 The chemical structure of a typical vinylester (with an epoxy backbone). Adapted from [9,
27]
Fig. 3.8 The chemical

structure of a typical
DGEBA epoxy. Adapted
from [5, 9, 27]
the crosslinking bridge continues to grow until another polyester change comes into
the reaction to terminate the process; see Stage 5 in Fig. 3.6. Note that in the same
figure, two styrene molecules are represented schematically as the bridge between the
polyester chains, but evidently this styrene group could contain only a few or many
styrene molecules. The final result is several polyester main chains with crosslinking
styrene groups between them [9].
Vinylester. A typical vinylester molecule is shown in Fig. 3.7. The vinylester
molecule has an epoxy backbone with ester linkages and double bonds (reactive
sites) at each end instead of epoxy rings (cf. Figs. 3.7 and 3.8) [5]. The crosslinking
process for vinylester is similar to that for polyester and commences with a catalyst,
but there are fewer reactive sites. This chemical structure helps to make vinylester
tougher (less brittle) than its polyester counterpart [27]. Increasing the number of
styrene molecules in the crosslinks can cause the cured polymer to exhibit the brittle
characteristics of polystyrene. Moreover, having an epoxy backbone along with a
reduction in ester groups makes vinylester less susceptible to hydrolysis compared
to unsaturated polyesters [5, 6].
Epoxy. The chemical structure for a typical epoxy is shown in Fig. 3.8. Epoxies are
characterised by three-member rings (reactive sites) containing two carbon atoms
bonded to an oxygen atom [9]. Whilst epoxy rings can be located either at the
extremities or in the middle of the polymer chain [14], for Diglycidyl Ether of
Bisphenol-A (DGEBA), these reactive sites are at either end of the epoxy molecule.
As mentioned earlier, unlike polyester and vinylester that cure by a catalyst, epoxies
cure with a hardener. The hardener is usually an amine curing agent containing
reactive NH2 groups to facilitate ‘opening’ of the epoxy ring [5]. There are usually
two reactive groups at the end of the curing agent to allow a reaction with two epoxy
molecules.
A simplified representation of the crosslinking process of an epoxy ring is given
in Fig. 3.9. The reaction commences when the epoxy and the hardener are mixed, i.e.
when the reactive NH2 group encounters the epoxy ring. Then, a C–O bond in the
Fig. 3.9 Simplified representation of epoxy crosslinking with an amine hardener. Adapted from [9]
epoxy and an N–H bond in the hardener are broken and alternative bonds between
the epoxy and hardener are created, i.e. a C–N and a C–OH bond. The crosslinking
process continues as the remaining reactive site, at the other end of the hardener
molecule, can react with another epoxy molecule, which results in connecting two
epoxy chains together [9]. The epoxy and amine molecules co-react, and hence
require a specific ratio. If they are not mixed in the correct proportions, unreacted
resin or hardener will result, and this will ultimately reduce the cured polymer prop-
erties [5]. Note that the lack of ester groups ensures epoxies are less susceptible to
moisture uptake compared to unsaturated polyester or vinylester [5, 6, 27].
3.7 Demoulding
After a laminate is cured to the solid state, it can be demoulded. Here, to remove
the FRC (panel) from the flat plate, a soft plastic wedge is used—see Fig. 3.10 . For
laminates with more complex shapes, some extra encouragement may be needed.
Assisted demoulding processes that involve mechanical [2], temperature [35] or
compressed air extraction methods [29] will be discussed in Chap. 7. The purpose
of the wedge in our case is to simply lift one corner of the panel from the flat plate
without causing damage to either the UD composite or mould. Demoulding should
be relatively easy to achieve for a smooth flat plate mould that has been appropriately
released. Once a small part of the FRC is freed from the mould, the rest of the panel
3.7 Demoulding 47
Fig. 3.10 Demoulding with a plastic wedge
should release with relative ease. Note that care should always be taken during the
demoulding process as it is easy to damage a laminate or, in some cases, the mould.
It is never advisable to use metal hand tools (or any other hard or abrasive materials)
for demoulding as these may scratch or score the mould tool.
3.8 Post-Processing (Finishing)
A FRC will rarely leave the mould in the finished state [3]. As a minimum, trim-
ming and sanding are usually required. Moreover, laminates are often assembled (or
joined) to other traditional materials (metals, ceramics and polymers) or other com-
posite parts to form larger assembled structures [9]. These joining processes often
involve drilling [36], cutting or machining operations [37]; even bonding compos-
ites to other materials requires some abrasion of the surface [5, 15]. The post- (or
secondary) processing activities should be tackled with caution as these actions can
cause damage to a laminate (as well as rapid tool wear in the case of machining
fibrous composites) [38].
Machining Processes. Trimming, cutting, drilling and other machining processes
can often be performed (with care) using basic hand tools or conventional metal work-
ing equipment, e.g. lathes and milling machines [39]. In general, the cutting speeds
for composite materials should be higher and the feed rates lower than those selected
for similar operations on metals [3, 5]. This minimises the forces and temperature
effects [40] that can cause damage. In the same context, the use of carbide (medium
cost), diamond or boron nitride (high cost) tooling is normally recommended for
FRCs as traditional cutting tools can blunt quickly since fibre reinforcements are
abrasive [41]. The use of blunt cutting tools increases tool friction and hence thermal
issues.
To finish the demoulded carbon fibre panel in Fig. 6.16, it is marked to the appro-
priate dimensions and then simply and carefully trimmed using a table saw, but a
hacksaw or bandsaw are suitable alternatives. If a more complex shape is needed,
a jigsaw can be used to cut the outline from the panel. The optional gelcoat is not
integrated into our composite panel, and no extra surface coating is added here at
this post-processing phase; a surface coating is normally applied to a composite to
provide environmental protection and a visual appearance, but neither are considered
important in this instance. The panel will be used later to benchmark the mechanical
properties attained from wet layup in comparison to other advanced manufacturing
methods, and to benchmark the performance of the RoM and KT equations—see
Chap. 2.
Surface Finish. If a surface finish is needed at the post-processing step, options
include the use of a flow coat or paint (clear and coloured paints). A flow coat
is a viscous coating of resin that covers the external surface of the composite. The
application of both a flow coat and painting requires similar initial surface preparation
methods.
In terms of preparation for a flow coat or paint, the surface is cleaned to remove
debris or contaminates (including residual mould release) and then sanded to provide
a smooth finish. A clean cloth and warm water, and/or an appropriate solvent are used
to clean the surface. Sanding helps remove tiny imperfections from the surface of
the laminate but should ideally only be used to roughen the surface and not expose
the fibres [3]; to prevent dust, wet sanding is usually recommended. Sometimes
surface filling is also necessary, but this is less common when a composite is made
using a mould. Sanding often commences at 120–220 grit sandpaper (<P120–P220)
to initially remove the high spots and then moves through a series of grades until
finishing is completed with a fine 400 grit paper (P600–P800) or more—Wanberg [2]
suggest 320–600 grit. Once sanding is complete, the surface should be cleaned and
left until completely dry.
The flow coat or paint layers are applied after drying. Paints can often allow
the addition of a primer for maximum adhesion between the laminate surface and
the top coat [42]. Primer fillers may also need to be used to remove any remaining
imperfection, prior to application of the finish coat. Following the supplier’s recom-
mendations, additional sanding at very fine grits (400 grit) is sometimes needed
between each coat.
Example 3.3
Consider the wet layup process. List what you think are the main advantages
and disadvantages.
3.8 Post-Processing (Finishing) 49
Solution
The main advantages and disadvantages of wet layup process are summarised
as follows [25, 27, 43]:
Advantages Disadvantages
Simple and versa- Fibre orientation and resin mixing. The fibre
tile method. The wet orientation is aligned by the fabricator and
layup process is sim- resin mixing is performed by hand. Thus, lam-
ple to understand and inate quality depends on the fabricator.
can be used in many
applications—from
small parts to large
composite structures.
Low-cost manufacturing Fibre volume fraction. Lower fibre volume
process. Minimal equip- fractions (and higher void contents) are
ment is needed to pro- attained in comparison to other advanced
duce a fibre-reinforced methods since minimal consolidation can be
composite and hence, it achieved via a roller—see Chap. 4.
is cost-effective for pro-
totyping or small produc-
tion runs.
Laminate consistency. Maintaining consis-
tency in a composite laminate and between
samples is difficult, and hence, a greater varia-
tion can often be seen in laminate performance.
Health and safety considerations. The resin
mixing process is a manual process, and fab-
ricators are exposed to volatile organic com-
pounds (VOCs).
3.9 Summary
There are six basic steps to make a FRC:

1. Fibre and matrix selection.
2. Mould preparation.
3. Layup and consolidation.
4. Curing (and post-curing).
5. Demoulding.
6. Post-processing (finishing).
The wet layup process is a simple, low-cost and versatile composite manufacturing
process. Fibres are impregnated with resin manually and allowed to cure, usually at
room temperature.
Table 3.2 Weight fractions for wet layup (hand lamination) based on fibre type and form
Fibre type Random mat W f Woven W f Unidirectional W f
Glass 0.19–0.48 0.35–0.59 0.48–0.68
Carbon 0.14–0.38 0.27–0.49 0.38–0.59
Aramid 0.12–0.34 0.23–0.44 0.34–0.55
Resin and hardener (or catalyst) mix can be calculated in terms of the weight fractions
from
A × n × Aw × (1 − Wf )
Resin and hardener mix [g] =
Wf
3.10 Questions
Question 3.1 State the six basic steps needed to design and manufacture a FRC.
Question 3.2 Why is it important to clean and release a mould prior to commencing
a composite layup process?
Question 3.3 What is a gelcoat and why would you use one?
Question 3.4 What is a flow coat and how does it differ from a gelcoat?
Question 3.5 Describe a simple wet layup and consolidation process.
Question 3.6 What fibre volume fractions are likely to be achieved for each of the
following fibre reinforcements when using wet layup as a manufacturing process?
a. Plain weave carbon.
b. Unidirectional aramid.
c. Random (chopped strand) glass mat.
Question 3.7 Describe what is meant by curing and post-curing.
Question 3.8 After mixing, how long does it typically take for a resin to start to
gel? How long should it take to transform from a low molecular weight liquid to a
solid polymer?
Question 3.9 Describe the series of chemical steps involved in the curing (or
crosslinking) reaction of an unsaturated polyester mixed with a small amount of
initiator/catalyst. In doing so, identify the most common catalyst and reactive dilu-
ent used in the crosslinking reaction.
Question 3.10 Describe the series of chemical steps involved in the curing (or
crosslinking) reaction of an epoxy mixed with an amine hardener.
3.10 Questions 51
Question 3.11 Why should you never use metal hand tools (for example, a scrapper)
when demoulding a composite part?
Question 3.12 A FRC laminate will rarely leave the mould in the finished state.
What post-processing operations are commonly required?
Question 3.13 What surface preparation is needed before a FRC is painted?
3.11 Problems
Problem 3.1 A FRC panel (0.2 m squared) is to be manufactured using wet layup.
The laminate will comprise six layers of woven glass fibres (areal weight = 300 g/m2 )
in a polyester matrix. The resin specifies 2% MEKP catalyst by weight:
a. What fibre volume and weight fractions would you suggest?
b. What mass of the resin and catalyst mix should be used?
c. What mass of polyester resin and MEKP catalyst are needed?
Answer. a. Vf = 0.3, Wf = 0.47; b. 101.5 g (waste factor = 1.25); c. 99.5 g resin, 2.0 g
catalyst.
Problem 3.2 How much polyester resin and catalyst mix are required to wet out a
single layer (1 m2 ) of E-glass random mat with an areal weight = 450 g/m2 ?
Answer. 1918.4 g (Wf = 0.19)–487.5 g (Wf = 0.48).
Problem 3.3 Wet layup is used to manufacture a random FRC consisting of 20 vol%
glass fibres in a polyester matrix. Assume E-glass fibres have a modulus of 70 GPa
and the polyester has a modulus of 2.9 GPa. Estimate the composite’s in-plane elastic
modulus.
Answer. E c = 16.3 GPa.
Problem 3.4 A unidirectional carbon fibre-reinforced epoxy composite is produced
by wet layup. The tensile strength of the carbon fibres and the epoxy matrix is
3800 MPa and 70 MPa, respectively. If the FRC is axially loaded in tension (parallel
to the fibres) until failure occurs, estimate its failure strength.
Hint. You will need to assume a fibre volume fraction appropriate for wet layup.
Answer. σcl∗ = 1520 MPa (assuming Vf = 0.4).
References
1. Composites Australia (2019) Mould release agents. https://www.compositesaustralia.com.au/

for-industry/health-and-safety/mould-release-agents/
4. Thomas S (2014) Polymer composites. Wiley-VCH, Weinheim
7. Wanberg J (2010) Composite materials: fabrication handbook #2. Composite garage series.
Wolfgang Publications, Stillwater, Minnesota
8. Mayer RM (1993) Design with reinforced plastics: a guide for engineers and designers.
Springer, Netherlands, Dordrecht. https://doi.org/10.1007/978-94-011-2210-8
9. Mallick PK (2007) Fiber-reinforced composites: materials, manufacturing, and design, 3rd
edn. Mechanical engineering. CRC/Taylor & Francis, Boca Raton
10. Hyosung Corporation (2017) Tansome carbon fiber. https://www.hyosungusa.com/files/
advanced/tansome_catalog_2017.pdf
11. Fiber Glass International (2021) Es180 epoxy data system. https://martglass.com.au/wp-
content/uploads/FGI-Epoxy-Resin.pdf
12. Quinn J (ed) (1990) Compliance of composite reinforcement materials
13. Lee SM (1989-) Reference book for composites technology. Technomic Pub. Co, Lancaster
Pa. U.S.A
14. Wang R, Zheng S, Zheng Y (2011) Polymer matrix composites and technology. Woodhead
publishing in materials, Woodhead Pub. and Science Press, Oxford and Philadelphia and Beijing
15. Campbell FC (2004) Manfacturing processes for advanced composites. Elsevier Advanced
Technology, Oxford
17. Composites Australia (2019) Health and safety. https://www.compositesaustralia.com.au/for-
industry/health-and-safety/
18. Clark SL (2013) Release agents. In: Lubin G (ed) Handbook of composites. Springer, New
York, NY, pp 633–638
19. Potter K (1997) An introduction to composite products: design, development and manufacture.
Chapman & Hall, London
20. Hyer MW (2009) Stress analysis of fiber-reinforced composite materials, updated. DEStech
Publications Inc, Lancaster, Pennsylvania
21. Steele GL (1962) Fiber glass: projects and procedures. McKnight & McKnight
22. Yuhazri MY, Haeryip S, Zaimi ZA et al (2015) A review on gelcoat used in laminated composite
structure. Int J Res Eng Technol 4:49–58
23. Bunsell AR, Renard J (2005) Fundamentals of fibre reinforced composite materials. Series in
materials science and engineering, Institute of Physics Publishing, Bristol
24. Owen MJ, Middleton V, Jones IA (2000) Integrated design and manufacture using fibre-
reinforced composites. Woodhead, Cambridge
25. Karlsson KF, TomasÅström B (1997) Manufacturing and applications of structural sandwich
components. Compos Part A: Appl Sci Manuf 28(2):97–111. https://doi.org/10.1016/S1359-
835X(96)00098-X
26. Fan M, Fu F (2016) Advanced high strength natural fibre composites in construction. Woodhead
Publishing, Oxford
28. Vergnaud JW, Bouzon J (1992) Cure of thermosetting resins: modelling and experiments.
Springer, London. https://doi.org/10.1007/978-1-4471-1915-9
29. Weatherhead RG (1980) FRP technology: fibre reinforced resin systems. Springer, Netherlands,
Dordrecht
30. Saunders N (2008) Chemical reactions, 1st edn. Exploring physical science. Rosen Cen-
tral/Rosen Publication, New York
31. Torgal FP, Labrincha J, Cabeza L, Goeran Granqvist C (2015) Eco-efficient materials for
mitigating building cooling needs: design, properties and applications. Woodhead Publishing,
Oxford
References 53
32. Bai J (ed) (2013) Advanced fibre-reinforced polymer (FRP) composites for structural appli-
cations, vol 46. Woodhead publishing series in civil and structural engineering. Woodhead
Publishing, Oxford
33. Hollaway L (1994) Handbook of polymer composites for engineers. Woodhead Publishing Ltd,
Cambridge
35. Silcock MD, Garschke C, Hall W, Fox BL (2007) Rapid composite tube manufacture uti-
lizing the quicksteptm process. J Compos Mater 41(8):965–978. https://doi.org/10.1177/
0021998306067261
36. Eneyew ED, Ramulu M (2014) Experimental study of surface quality and damage when drilling
unidirectional cfrp composites. J Mater Res Technol 3(4):354–362. https://doi.org/10.1016/j.
jmrt.2014.10.003
37. Karpat Y, Bahtiyar O, Değer B (2012) Milling force modelling of multidirectional carbon
fiber reinforced polymer laminates. Procedia CIRP 1:460–465. https://doi.org/10.1016/j.procir.
2012.04.082
38. Gaugel S, Sripathy P, Haeger A, Meinhard D, Bernthaler T, Lissek F, Kaufeld M, Knoblauch
V, Schneider G (2016) A comparative study on tool wear and laminate damage in drilling of
carbon-fiber reinforced polymers (cfrp). Compos Struct 155:173–183. https://doi.org/10.1016/
j.compstruct.2016.08.004
39. Abrate S, Walton DA (1992) Machining of composite materials. part i: traditional methods.
Compos Manuf 3(2):75–83
40. El-Hofy MH, Soo SL, Aspinwall DK, Sim WM, Pearson D, M’Saoubi R, Harden P (2017) Tool
temperature in slotting of cfrp composites: 45th sme north american manufacturing research
conference, namrc 45, la, usa. Proc Manuf 10:371–381. https://doi.org/10.1016/j.promfg.2017.
07.007
41. Caggiano A (2018) Machining of fibre reinforced plastic composite materials. Materials (Basel,
Switzerland) 11(3). https://doi.org/10.3390/ma11030442
42. Biron M (2014) Thermosets and composites: material selection, applications, manufacturing,
and cost analysis, 2nd edn. PDL handbook series. Andrew Elsevier, Oxford
43. Lee SM (1992) Handbook of composite reinforcements. VCH, New York
Chapter 4
Advanced Methods—Vacuum Bagging
and Prepreg Moulding
Abstract Vacuum bagging and prepreg moulding methods are introduced in this
chapter. These advanced moulding techniques address some of the shortcomings
of wet layup (hand lamination) but require additional equipment and consumables,
as well as higher fabricator skill levels. The requisite equipment and consumables
are described in detail. A step-by-step guide to vacuum bagging is offered and the
storage and handling requirements of prepregs are considered. The advantages and
limitations of these advanced methods are presented alongside their typical fibre
volume (and weight) fractions. Finally, fibre volume fractions for wet layup are
compared to the higher fractions typically obtained by vacuum bagging and prepreg
moulding.
4.1 Introduction
The wet layup process is a simple and effective method to create FRCs. It is a versatile
and low-cost fabrication method that requires minimal expertise to create a laminate.
The main problems [1–3] associated with wet layup include:
• Accuracy of fibre orientation.
• Control of the fibre and resin ratio.
• Void content (entrapped air).
• Consistency of structure or parts.
• Health and safety concerns with resins.
To overcome some of the performance issues of wet layup, vacuum bagging and
prepreg moulding methods have been developed [3–7]. These advanced methods are
considered here.

https://doi.org/10.1007/978-3-030-78807-0_4
56 4 Advanced Methods—Vacuum Bagging and Prepreg Moulding
4.2 Vacuum Bagging
The vacuum bagging process is an advanced moulding technique that can be used
alongside wet layup to remove entrapped air and provide consolidation pressure
to a laminate [4, 7, 8]. A vacuum bag is made to cover the uncured laminate, air
is evacuated, and atmospheric pressure is used to improve consolidation. The air
evacuation process is sometimes referred to as debulking [2, 3].
This technique can be applied once the wet layup is complete, as a consolidation
phase. Recall the six-step process in Chap. 3 (see Fig. 3.1). All steps for vacuum
bagging are the same as for wet layup, except for Step 3: ‘Layup and consolidation’.
In step 3, consolidation is introduced via a vacuum bag. Namely, after the laminate is
rolled to expel entrapped air, an additional vacuum consolidation method is applied
to the wet layup.
A schematic of the vacuum bagging setup is shown in Fig. 4.1. To facilitate the
application of a vacuum, extra equipment and consumables are needed. In terms of
equipment, a breach unit, a vacuum pump, vacuum pot and a vacuum gauge are
used. Extra consumables include a vacuum hose (tube), peel ply (optional), release
film, bleeder and breather fabrics, vacuum bagging film and sealant tape (sometimes
referred to as tacky tape). The role of the vacuum equipment and consumables is
clearly outlined in Tables 4.1 and 4.2, respectively.
In this chapter, only the steps of the vacuum bagging process following completion
of the wet layup stage are explained. For consistency between the techniques, the
same fibre and matrix selections (described in Chap. 3) are repeated, i.e. four layers
of (200 g/m2 ) carbon fibre [9] and epoxy resin [10] are used. Figure 4.2a shows the
fibres on the mould (flat plate), prior to the vacuum bagging assembly process. The
fibres have been wet out using the wet layup process—the resin is still in the liquid
(uncured) state.
To commence the vacuum bagging process, a release film is placed directly on
top of the wet layup to prevent the consumables from adhering to the curing resin;
see Fig. 4.2b. Sometimes an optional peel ply is used prior to the release film, but not
in the current instance. The purpose of the peel ply is to provide a textured surface
for further lamination or to facilitate secondary bonding [4, 5, 11].
Breather fabric
Vaccum gauge
Bleeder fabric To vaccum
Breach unit Release film pump
Vaccum bagging film
Sealant
Mould
Laminate
Fig. 4.1 Schematic of the vacuum bagging process

4.2 Vacuum Bagging 57
Table 4.1 Vacuum bagging equipment

Item Purpose Image
Breach unit The breach unit allows a vacuum

hose to be connected to the
vacuum bagging film. Note, an
alternative is to simply seal the
vacuum hose between the vacuum
bagging film and mould perimeter.
Vacuum hose Provides the link between the

breach unit and the vacuum pot,
as well as between the vacuum
pot and vacuum pump.
Vacuum pot Prevents excess resin from

(resin trap) entering (and hence damaging)
the vacuum pump.
Vacuum pump A high-volume pump for drawing

a vacuum.
Vacuum (pres- Attached to the vacuum bag via a

sure) gauge breach unit to monitor the vacuum
(consolidation) pressure.
Table 4.2 Vacuum bagging consumables

Item Purpose Image
An optional fabric used to produce a textured sur-

face on the cured laminate. The textured surface
Peel ply
facilitates further lamination, secondary bond-
ing or painting.
A perforated sheet placed on the peel ply or

directly on top of the laminate. The perforations
Release allow the resin to ‘bleed’ from the laminate dur-
film ing the curing process. The release film prevents
the other consumable from sticking to the lami-
nate.
Bleeder/breather fabric serves two main pur-

poses: (i) to absorb excess resin (bleed) and reac-
Bleeder/
tants from the laminate; and (ii) to provide a flow
breather fabric
path to permit the escape of air, reactants and
volatiles.
Sealant tape, sometimes referred to as tacky tape,

Sealant
is used to ensure a seal between the mould and
tape
the vacuum bagging film.
The vacuum bagging film encapsulates the lam-

Vacuum
inate and consumables permitting a vacuum to
bagging film
be drawn.
(a) (b)
(c) (d)
Fig. 4.2 Consolidation using a vacuum bagging process (over a wet layup): a final stage of the
wet layup; b application of the release film; c vacuum bagging fixed with the sealant tape; and d
extracting the excess resin under vacuum
Next, bleeder and breather fabrics are placed to cover the laminate and release
film. These are often the same material [12] and are therefore (with careful control
of resin absorption [3]) sometimes combined into one as illustrated in Fig. 4.2c.
The combination fabric is referred to here as bleeder/breather fabric [5, 7, 13]. The
vacuum bagging film is then sealed to the mould tool with sealant tape allowing
a vacuum to be drawn within the vacuum bag. In the figure, the vacuum is drawn
simply by sealing the hose between the vacuum bagging film and mould, near the
perimeter. A more common method preferred by many fabricators is to use a breach
unit (sometimes called a vacuum valve [3] or vacuum connector [5, 7]) instead of
using a hose as the air outlet; see Fig. 4.1. The outcome of either methods should be
identical, but the use of a hose method is preferred herein.1
As the vacuum is drawn, the air is evacuated and the excess resin is absorbed by
the bleeder/breather fabric. In Fig. 4.2d, the excess resin can be seen as it is drawn into
1If using the vacuum hose method, to ensure a uniform consolidation pressure across the lami-
nate and to prevent blockage, the hose can be carefully placed between two additional pieces of
bleeder/breather fabric placed on top of the original fabric but to one side of the wet layup. If the
hose is left on top of the layup, vacuum pressure may cause an indentation on the cured composite.
the fabric by the vacuum—without the bleeder/breather fabric, it is more difficult to

evacuate the air and draw out the excess resin. A vacuum (pressure) gauge monitors
the absolute pressure inside the vacuum bag and helps to identify the leak rate. There
is no agreed threshold for an acceptable leak rate; the aim is simply to maintain a
leak rate as close to zero as possible [14], but as a general guide less than 1 mbar/min is
recommended here as a sensible threshold target for the vacuum bagging process.
Vacuum Bagging Technique. There is more than one method to create a vacuum bag
that surrounds a laminate and offers a suitable mould seal. Possible options include
the following:
• The vacuum bagging film (normally nylon [2, 15]) is laid directly onto the sealant
tape, which is already adhered to the mould surface [7].
• The sealant tape is stuck to the vacuum bagging film, and then the film and sealant
tape are simultaneously applied to the mould tool as illustrated in Fig. 4.3.
If a mould tool is not flat but is instead either concave or convex (i.e. a female or
male mould tool, respectively), additional vacuum bagging film will be needed to
(a) (b)
(c) (d)
Fig. 4.3 Vacuum bagging: a application of the sealant tape; b, c and d steps in the vacuum bagging
process
conform to the tool surface. As a result, there will be a need to introduce pleats in
the bag to ensure that there is enough bagging film to conform to the contours of the
mould [2]. A bag that is too large is better than one that is too small!
In Fig. 4.3, the pleated vacuum bagging process is illustrated in more detail (albeit
for the flat plate mould, rather than a more complex male or female mould surface).
The sealant tape is applied to the vacuum bagging film as shown in Fig. 4.3a, and
then after introducing the requisite consumables, the film and sealant tape are stuck
to the mould surface. The sealant tape is initially stuck to the mould tool at the four
corners as shown in Fig. 4.3b, and then working from each of the corners, the pleats
are formed at one or two locations on each edge; a pleat formed on one edge of the
mould tool is shown in Fig. 4.3c. Pleats are normally aligned with the edges of the
composite sample or geometric features. In Fig. 4.3d, the vacuum hoses are wrapped
in sealant tape before being placed in their position on the perimeter of the mould
as an alternative to using a breach unit (or vacuum valve/connector [3, 5, 7]). This
wrapping method helps to lift the vacuum hose from the mould surface and facilitates
a tight seal around the vacuum hose.
Once the vacuum bag is sealed around the mould perimeter, a vacuum is drawn and
the air is extracted. The pressure inside the vacuum bag is monitored and any evident
leaks are rectified; once the leaks are minimised, a leak test should be conducted
for a minimum of 5 minutes [14]. The monitoring and rectification of leaks should
continue until an acceptable leak rate is achieved whilst remembering that the curing
process is underway. The vacuum is maintained for several hours until the composite
is fully cured.
After the FRC is cured to a solid state (after 24 hours), the composite panel is
demoulded. The vacuum consumables are detached, and disposed of, before the FRC
panel is removed from the mould tool with a plastic wedge. The UD composite is
then post-cured using the same procedure as for the wet layup panel (see Chap. 3).
Table 4.3 refines this fibre volume fraction range for the vacuum bagging process
as well as offering estimates for the corresponding weight fractions (based on fibre
type and form). The vacuum bagging process offers notable volume fraction improve-
ments over the wet layup process; typical wet layup values were previously reported
in Table 3.2.
The vacuum bagging process addresses some (but not all) of the issues with wet
layup. The main advantages and disadvantages of vacuum bagging are outlined in
Table 4.4.
Table 4.3 Typical fibre volume and weight fractions for the vacuum bagging process
Material Random mat Woven Unidirectional
Vf Wf Vf Wf Vf Wf
Glass 0.2–0.4 0.35–0.59 0.3–0.5 0.48–0.68 0.4–0.6 0.59–0.76
Carbon 0.2–0.4 0.27–0.49 0.3–0.5 0.39–0.59 0.4–0.6 0.49–0.69
Aramid 0.2–0.4 0.23–0.44 0.3–0.5 0.34–0.55 0.4–0.6 0.44–0.64
Table 4.4 Advantages and disadvantages of vacuum bagging. Adapted from [2, 11, 16]
High fibre volume fractions. Air and excess Fibre orientation. The fibre orientation is still
resin are drawn from the laminate resulting in manually aligned by the fabricator, and hence
higher fibre volume fractions and lower void accuracy and consistency is still a potential
content than for wet layup. issue.
Improved consolidation. Laminate Fabricator skill levels. Fabricators require
consolidation and wet out are improved due to higher skill levels and there can be difficulty in
pressure and resin flow, and hence a more bagging complex shapes.
uniform thickness and a better surface finish
can be attained.
Health and safety. The vacuum bag minimises Increased process costs. There are extra
exposure to volatile organic compounds during equipment and consumable costs, and vacuum
curing, but the resin mixing process is still a tight moulds are needed.
manual task.
4.3 Prepreg Moulding
Prepregs are pre-impregnated materials usually comprising carbon fibre and epoxy.
They are supplied on a plastic backed roll and are stored in a protective bag in a
freezer (at −18 ◦ C) [3]. The prepregs are in a stable intermediate cure stage (partially
crosslinked) that only requires heat to cure. Prepregs can be stored in the freezer for
3–12 months [15], but at room temperature the prepreg can remain uncured for a
shorter period of time (1–2 months is typical) [17]. The time a prepreg can be stored
in the freezer is referred to as the freezer-life [15, 18] or shelf-life [14], whilst the
term out-time [2, 3] or out-life [18] identifies the time a prepreg can remain at room
temperature before it is no longer useable.
Prior to heating, the prepregs are removed from the freezer and protective bag, cut
to size and left to thaw at room temperature for several hours to prevent condensation
issues [2]. The prepreg roll is resealed and returned to freezer storage as soon as
possible to maintain the shelf-life [3]. Cutting can be performed with a sharp knife
or scissors. Once the plastic backing is removed, the prepreg is ready for layup. The
freezer and out-life of prepregs need to be carefully monitored to ensure appropriate
storage and handling times are not exceeded.
At room temperature, prepregs are tacky and are easily handled with safety
gloves [7] as shown in Fig. 4.4. To help conformation to the mould and to increase
tack and drape during layup, prepregs can be gently heated [4, 5]. To optimise per-
formance, the fibre and resin ratios are predefined by the supplier with near-net fibre
volume fractions [2]. Moreover, they are easily cut in the uncured state [7] and hence,
precut patterns can simplify prepreg layup. The tackiness allows for easy positioning
in the mould and hence provides maximum accuracy in fibre orientation. The mate-
rial costs for prepregs, however, are significantly higher (perhaps 4 times higher [19])
than those for wet layup and small quantities cannot often be purchased from supplies
4.3 Prepreg Moulding 63
Fig. 4.4 Prepreg: woven

carbon fibres in a partially
cured epoxy
(a) (b)
Fig. 4.5 Prepreg moulding process: a prepreg layup (4 plies); and b vacuum bagging and debulking
who tend to prefer not to ‘split’ a roll. They are therefore often only readily used in
the highest performance industries (e.g. aerospace and sports equipment sectors [3]).
The prepreg layup and consolidation is shown in Fig. 4.5. During layup, the
prepreg [20] is laid on a pre-released mould, rolled after each ply and then consol-
idated in a vacuum bag (see method in Sect. 4.2). Prepregs cure at elevated tem-
peratures, and therefore the mould release and the vacuum bagging consumables
(release film, bleeder/breather fabric, vacuum bagging film and sealant tape) all need
to be suitable at the cure temperature. There are three main temperature ranges for
prepregs: low (60 ◦ C); medium (120 ◦ C); and high (180 ◦ C). However, some prepregs
(including the one used here) allow a variable cure temperature with modified curing
times—lower curing temperatures require longer curing times than higher tempera-
tures.
To cure prepregs, the temperature is typically ramped up from room temperature,
held at the cure temperature in an oven for the requisite time (sometimes referred to
as the dwell time) and then cooled down. The temperature ramp-up and cool-down
rates are provided by the prepreg supplier, but are normally only a few degrees per
Table 4.5 Typical fibre volume and weight fractions for prepreg processes
Material Random mat Woven Unidirectional
Vf Wf Vf Wf Vf Wf
Glass N/A N/A 0.4–0.55 0.58–0.72 0.5–0.7 0.68–0.83
Carbon N/A N/A 0.4–0.55 0.49–0.64 0.5–0.7 0.59–0.77
Aramid N/A N/A 0.4–0.55 0.44–0.59 0.5–0.7 0.55–0.74
Table 4.6 Advantages and disadvantages of prepreg moulding. Adapted from [2, 16]
High fibre volume fractions. Fibre and resin Fibre orientation. The fibre orientation is still
ratios are predefined ensuring high fibre aligned by the fabricator but there is potential
volume fraction and low void content. for automation.
Longer working times. The extended working Fabricator skill levels. Fabricators require
times mean complex layups can be achieved higher skill levels—similar to vacuum bagging.
without concerns about resin working (curing)
times.
Shorter curing times. Prepreg cure times can be Moulding costs. Prepreg costs are high and
much shorter than for wet layup or vacuum storage controls are needed to monitor freezer
bagging—see Table 4.5. and out-life. Elevated cure temperatures mean
tooling needs are increased.
Health and safety. Health and safety issues are
minimised using prepregs.
minute (for example, 2 − 3 ◦ C/min). The laminate is maintained under vacuum pressure
during the entire curing process until it is removed from the oven, demoulded and
post-processed. Here, a dwell time of 5 hours at a temperature of 80 ◦ C is employed.
Table 4.5 provides fibre volume and weight fraction estimates (based on fibre type
and form) for prepreg moulding. The premium cost (and performance) associated
with prepreg means that random mat carbon fibre prepregs are not readily available—
it is not sensible to pay a premium cost for a suboptimal solution. Volume fractions
for prepreg moulding are towards the highest levels that can be sensibly realised in
practice.
The main advantages and disadvantages of prepreg moulding are outlined in Table 4.6.
4.4 Summary
Advanced moulding methods require additional equipment and consumables but

address many of the shortcomings of wet layup. They offer higher fibre volume
fraction (and hence weight fraction) laminates with lower void content. The result is
therefore a stiffer and stronger laminate, but moulding costs for advanced methods are
4.4 Summary 65
Table 4.7 Typical fibre volume fractions for wet layup and advanced moulding methods
Method Fibre form
Random (V f ) Woven (V f ) Unidirectional (V f )
Wet layup∗ 0.1–0.3 0.2–0.4 0.3–0.5
Vacuum bagging 0.2–0.4 0.3–0.5 0.4–0.6
Prepreg moulding N/A 0.4–0.55 0.5–0.7
∗ Wet layup values are restated from Chap. 3 for convenience
higher than for wet layup and necessitate higher fabricator skill levels. Prepreg mould-
ing has the highest cost for the methods described. Typical fibre volume fractions for
vacuum bagging and prepreg moulding are compared to wet layup in Table 4.7.
4.5 Questions
Vacuum Bagging
Question 4.1 The main problems associated with wet layup were listed as:
• Accuracy of fibre orientation.
• Control of the fibre and resin ratio.
• Void content (entrapped air).
• Health and safety concerns with resins.
Which of these are addressed, partially addressed or not addressed by
a Vacuum bagging.
b Prepreg moulding.
Question 4.2 Describe the role of each item of equipment in the vacuum bagging
process:
a Breach unit.
b Vacuum hose.
c Vacuum pot.
d Vacuum pump.
e Vacuum gauge.
Question 4.3 Identify the five consumables in Fig. 4.6. Describe their role in the
vacuum bagging process.
(5)
Vaccum gauge
(4) To vaccum
Breach unit (3) pump
(2)
(1)
Mould
Laminate
Fig. 4.6 Consolidation using a vacuum bagging process (over a wet layup)
Question 4.4 Provide fibre volume fraction and fibre weight fraction estimates (i.e.
the range of values) for each fibre reinforcement using the vacuum bagging process:
a Unidirectional carbon fibre.
b Plain weave carbon fibre.
c Twill weave glass fibres.
Question 4.5 A vacuum bag is used to remove air trapped between the fibres and to
consolidate the laminate. What term is used to describe the process of air evacuation?
Prepreg Moulding
Question 4.6 The term prepreg is an abbreviation for what term?

Question 4.7 Provide fibre volume fraction and fibre weight fraction estimates (i.e.
the range of values) for a unidirectional carbon fibre prepreg.
Question 4.8 The room temperature is 24 ◦ C, and the ramp-up and cool-down rates
are scheduled to be 2 ◦ C per minute. What will the total manufacturing time be, if
the prepreg is cured for 5 hours at 80 ◦ C?
4.6 Problems
Problem 4.1 A unidirectional carbon fibre-reinforced epoxy composite was pro-

duced by wet layup in Problem 3.4. If the same constituents (fibres and resin) are
used to fabricate a composite using the vacuum bagging process, estimate the strength
of the FRC loaded in axial tension. Compare the tensile strength of the wet layup
and vacuum bagged laminates—is there a significant difference? If so, why? Recall,
the tensile strength of the carbon fibres and the epoxy matrix were assumed to be
3800 MPa and 70 MPa, respectively.
Answer. σcl∗ = 1520 MPa (wet layup), σcl∗ = 1900 MPa (vacuum bagging). Why? Vf .
4.6 Problems 67
Problem 4.2 Which of the following materials and manufacturing processes would
you expect to be suitable to achieve a tensile modulus of 80 GPa and a tensile strength
of 1350 MPa?
• Wet layup of unidirectional HS carbon fibres (40 vol%) in an epoxy resin.
• Vacuum bagging of unidirectional HM aramid fibres (55 vol%) in an epoxy resin.
• Prepreg moulding of unidirectional (an E-glass fibres (60 vol%) in an epoxy matrix.
The elastic modulus and strengths for E-glass, HS carbon and HM aramid fibres, and
epoxy are given in Table 4.8.
Answer. Wet layup ✓, Vacuum bagging ✗, prepreg moulding ✗.
Table 4.8 Fibre and resin properties

Material Density Elastic modulus Tensile strength
(g/cm 3 ) (GPa) (MPa)
Glass (E-glass) 2.50–2.62 69–81 2000–3450
Carbon (HS) 1.75–1.80 228–300 3400–7100
Aramid (HM) 1.40–1.44 130–131 3600–4100
Epoxy 1.1–1.4 2.41–4.1 27.6–130
References
1. Karlsson KF, TomasÅström B (1997) Manufacturing and applications of structural sandwich

components. Compos Part A: Appl Sci Manuf 28(2):97–111. https://doi.org/10.1016/S1359-
835X(96)00098-X
6. Mayer RM (1993) Design with reinforced plastics: a guide for engineers and designers.
Springer, Netherlands, Dordrecht. https://doi.org/10.1007/978-94-011-2210-8
8. Kutz M (2002) Handbook of materials selection. Wiley, New York and Chichester
10. Fiber Glass International (2021) Es180 epoxy data system. https://martglass.com.au/wp-
content/uploads/FGI-Epoxy-Resin.pdf
11. Campbell FC (2010) Structural composite materials. ASM International, Materials Park, Ohio
12. Sheikh-Ahmad JY (2009) Machining of polymer matrix composites. Springer, New York and
London
13. Lubin G (ed) (2013) Handbook of composites. Springer, New York, NY

14. Summerscales J, Graham-Jones J, Pemberton R (2018) Marine composites: design and per-
formance. Woodhead Publishing series in composites science and engineering. Woodhead
Publishing, Oxford
Technology, Oxford
17. Brain NG (1991) Composites: tooling and component processing. iSmithers Rapra Publishing
18. Baker AA, Dutton S, Kelly D (2004) Composite materials for aircraft structures, 2nd edn.
AIAA education series. American Institute of Aeronautics and Astronautics, Reston, VA
19. Adams PJ (2000) Revolutionary materials: technology and economics: 32nd International
SAMPE Technical Conference, Boston Park Plaza Hotel, Boston, Massachusetts, November
5–9 2000, International SAMPE Technical Conference series, vol 32. Society for the Advance-
ment of Material and Process Engineering, Covina, Calif
20. Solvay (2021) Technical data sheet: Vtm® 260 series (prepeg). https://catalogservice.solvay.
com/
Chapter 5
Composite Testing—How Accurate Are
Design Estimates?
Abstract A selection of simple composite design (mechanics) equations were

introduced in Chap. 2 to enable prediction of the mechanical properties of fibre-
reinforced composites (FRCs). The rule of mixtures (RoM) and inverse rule of mix-
tures (IRoM) expressions were introduced for elastic moduli estimation, whilst lon-
gitudinal strength was considered via the Kelly-Tyson (KT) model. Chapters 3 and 4
then introduced wet layup, and vacuum bagging and prepreg moulding methods for
composite manufacture. A carbon fibre-reinforced panels was fabricated using sim-
ilar constituents (i.e. 200 g/m2 carbon fibres in an epoxy matrix) for each fabrication
method. This chapter builds on these earlier chapters via a comparison of design
(mechanical property) estimates and experimental data. Tensile tests are performed
on multiple specimens cut from each of the panels. This chapter provides guidance
on the accuracy of design estimates and offers support for the typical (rule of thumb)
volume fractions suggested in Chaps. 3 and 4 for each manufacturing approach.
5.1 Introduction
In previous chapters, unidirectional (UD) carbon fibre-reinforced epoxy composites

were fabricated using basic wet layup (Chap. 3), and vacuum bagging and prepreg
moulding (Sects. 4.2 and 4.3). The same six steps were followed to create similar
carbon fibre panels. The laminates were made from similar constituents, i.e. the
fibre type and form, and the matrices were all similar. Nonetheless, the laminate
performance will differ in each case as a result of the variations in material properties
and manufacturing processes.
The fibre volume fraction (Vf ) is of critical importance to the mechanical perfor-
mance of FRCs [1] and is directly affected by the fabrication method. An understand-
ing of Vf (for each method) is therefore an important consideration for a composite
designer. Here, measured fibre volume fractions are used to inform design estimates
that are compared to tensile experiments for specimens cut from each of the carbon
fibre panels. In doing so, the chapter offers insight into the accuracy of the composite
design equations (see Chap. 2) and provides validation for the rule of thumb values

https://doi.org/10.1007/978-3-030-78807-0_5
70 5 Composite Testing—How Accurate Are Design Estimates?
used to estimate fibre volume fractions for each of the manufacturing methods—see
Table 4.7.
Design estimates are a fundamental step in a product development process as they
enable the designer to make structural predictions (within margins of error) prior to
prototype testing, but it should be noted that structural predictions are not a substitu-
tion for physical testing. After design calculations are complete, a prototype should
always be fabricated and tested to the relevant product or engineering standards.
5.2 The Tensile Test
The uniaxial tensile test is a common method used to determine the characteristic
mechanical properties of materials [2], including composites [3].1 Tensile test data
from token specimens is frequently used to provide elastic moduli, tensile strength
and fracture strain measurements that are subsequently used to inform structural
design decisions [5].
To perform a tensile test, the specimen is held in a uniaxial test machine and
monotonically loaded (see Fig. 5.1) until failure occurs [4]. During a tensile test,
force (load) and extension (deformation) are recorded. A load cell on the machine
records the force, whilst the deformation of a specimen is typically measured using
a clip-on extensometer (or sometimes bonded strain gauges) rather than via the
machine’s own crosshead displacement measurements; crosshead displacements are
Fig. 5.1 Tensile test of a

carbon fibre-reinforced
composite
1Tensile tests are usually performed in accordance with a recognised testing standard; for FRCs,
ASTM D3039 [4] is often used.
5.2 The Tensile Test 71
HS carbon-epoxy composite (longitudinal)

2,000
HS carbon-epoxy composite (transverse)

1,500
1,000
500
0
0 0.5 1 1.5
Strain (%)
Fig. 5.2 Tensile stress-strain curve for a UD carbon fibre composite in longitudinal and transverse
directions
not usually used as the measurements need to be calibrated based on the machine
compliance [6].
The load and deformation data is converted to engineering stress (σ ) and engineering
strain (ε), via [2]
F
σ = (5.1)
A
and
L
ε= (5.2)
L
where F is the force, A is the cross-sectional area of the specimen, L is the change
in length (extension) and L is the original length of the specimen or the extensometer
gauge length in the tensile test.
A typical tensile stress-strain curve for a unidirectional carbon FRC with fibres
aligned parallel (longitudinal) and transverse (perpendicular) to the loading direction
is shown in Fig. 5.2.
The tensile modulus (stiffness) of the material sample is determined via a corre-
lation of stress and strain. To calculate the stiffness, a submaximal load is correlated
to the deformation. As can be seen in Fig. 5.2, the tensile stress is proportional to
strain and hence, the elastic modulus of the sample is the constant of proportionality,
viz. [2]
σ
E= (5.3)
ε
where E cl could be used in place of E to denote axial (longitudinal) modulus for a

composite sample, and E ct could be used for the transverse modulus.
The tensile strength of the FRC (σ ∗ ) is the stress at failure (i.e. the maximum stress)
given by
Fmax
σ∗ = (5.4)
A
in which σcl∗ could replace σ ∗ for longitudinal strength, and σct∗ could be interchanged
for transverse strength.
5.3 Fibre-Reinforced Composite Specimens
The composite panels fabricated in Chaps. 3 and 4 were used to create the tensile test
specimens; longitudinal and transverse samples were created for each fabrication
method. Here, more longitudinal specimens were created from the plate than trans-
verse specimens as the mechanical properties of the former are usually of greater
significance to the designer—as previously mentioned, the transverse stiffness and
strength of unidirectional laminates are much lower than those in the longitudinal
direction.
The tensile test specimens were cut from the panel into rectangular strips approx-
imately 20 mm wide. At least five longitudinal and three transverse specimens were
cut from each panel.2 An example specimen is shown for each fabrication method
in Fig. 5.3.
The thickness of the FRC specimens is dependent upon the manufacturing method.
The wet layup specimens are the thickest, whilst the vacuum bagging and prepreg
specimens are thinner. A thicker specimen has a lower fibre volume fraction (and
hence, a higher matrix fraction) since the same volume of carbon fibres are used in
each fabrication method. The lowest fibre volume fractions are therefore evident in the
wet layup specimens, whilst the vacuum bagging and prepreg specimens have higher
volume fractions. A laminate’s thickness is related to its fibre volume fraction [7]
via
n Aw
t= (5.5)
ρf Vf
where t is the test specimen thickness, n is the number of layers, Aw is the areal
weight of the fabric, ρf is the mass density of the fibre used and Vf is the fibre volume
fraction.
The mean dimensions of the tensile test specimens and the fibre volume fractions
are given in Table 5.1. In the table, Vf from Eq. (5.5) is calculated based on the average
of five thickness measurements (measured prior to testing) from the successful test
specimens (see Sect. 5.4) using ρf = 1800 kg/m3 [8, 9]. The typical (rule of thumb)
fibre volume fractions previously reported (for unidirectional laminates in Table 4.7)
2 It is common practice to report mean measurements for at least five successful test specimen.
5.3 Fibre-Reinforced Composite Specimens 73
Fig. 5.3 Tensile

specimens—wet layup,
vacuum bagging and prepreg
moulding (left to right)
Table 5.1 Tensile test specimens: thickness and fibre volume fraction
Manufacturing Specimen thickness Vf Vf
method (mm) from Eq. (5.5) typical (rule of thumb)
Wet layup 1.32 0.34a 0.3–0.5
Vacuum bagging 0.89 0.50 0.4–0.6
Prepreg moulding 0.83 0.54 0.5–0.7
a Wet layup estimates in Example 3.1 assumed a slightly higher Vf (= 0.4) than the measured value
of 0.34
are also included in the table for comparison. Note the correlation between the typical
fibre volume fractions for each fabrication method and the corresponding calculated
value.
5.4 Tensile Properties: Elastic Moduli and Strength
Testing Method. Tensile tests data was produced with an Instron Universal Testing
System (UTS) using displacement control. The strain rate was maintained at 2 mm/min
(monotonic loading) until failure, in accordance with ASTM D3039 [4]. A clip-gauge
extensometer (with a 50 mm gauge length) was used to accurately measure gauge
length deformation. The stresses and strains observed during testing are derived from
the force-deformation data and the specimen dimensions using Eqs. (5.1) and (5.2),
respectively.
Results and Discussion. A typical plot of the stress-strain measurements for each
manufacturing method (wet layup, vacuum bagging and prepreg moulding) is shown
Fig. 5.4 Typical

stress-strain curves: 2,000 Wet layup
Vacuum bagging

a longitudinal; and
Prepeg
b transverse loading 1,500
1,000
500
0
0 0.5 1 1.5
Strain (%)
(a)
Wet layup
30 Vacuum bagging
Prepeg
20
10
0
0 0.1 0.2 0.3 0.4 0.5
Strain (%)
(b)
in Fig. 5.4. The measurements for the longitudinal and transverse fibre orientations
for each fabrication method are plotted on the same graph to clearly illustrate the
significant performance difference.
The mean elastic moduli and tensile strength measurements are compared to
design estimates (based on the RoM, IRoM and KT models) in Table 5.2. The
design estimates are calculated using carbon fibre properties taken from supplier
data sheets—for the fibres used in the wet layup and vacuum bagging processes, E f
and σf∗ are 250 GPa and 5516 MPa [8], whilst E f and σf∗ for the prepreg fibres are
230 GPa and 4900 MPa [10]. E m and σm∗ are assumed (in all estimates) to be 3.0 GPa
and 50 MPa, respectively. The fibre volume fractions Vf are taken from Table 5.1.
As a further comparison, Table 5.2 shows supplier data for the prepreg material,
normalised to 60%Vf [9].
In Table 5.2, the number of specimens successfully tested are identified in round
brackets. A successful test is identified by a specimen fracture in the gauge length
region. Thinner specimens tend to be more prone to premature fracture in the jaws
(even if tabs are used) and hence fewer prepreg samples were successfully recorded.
Transverse properties tend to be more difficult to accurately measure than longitudi-
Table 5.2 Tensile test specimens: estimated and measured mechanical properties
Manufacturing Orientation Elastic modulus Tensile strength
method (mean, GPa) (mean, MPa)
Experimental Estimated Prepreg Experimental Estimated Prepreg
60%Vf 60%Vf
Wet layup Longitudinal 80.0 (5)∗ 87.0 [8.8]† – 1310.4 (5) 1875.4.0‡ [43.1] –
Transverse 5.7 (3) 4.5 [−21.1] – 40.6 (3) 50.0 –
Vacuum bagging Longitudinal 117.4 (5) 126.5 [7.8] – 1855.0 (5) 2758.0 [48.7] –
5.4 Tensile Properties: Elastic Moduli and Strength
Transverse 7.5 (3) 5.9 [−21.3] – 25.3 (3) 50.0 –

Prepreg Longitudinal 126.2 (3) 125.6 [−0.5] 131.0 1967.5 (3) 2499.0 [27.0] 2575.0
moulding
Transverse 7.2 (1) 6.4 [−11.1] 9.1 13.6 (1) 50.0 40.0
a Number of specimens are printed in round brackets.
† Percentage relative errors are printed in square brackets.
‡ KT strength is calculated using fibre strength only, assuming σ ∗ = 5516 MPa (wet layup and vacuum bagging) or 4900 MPa (prepreg moulding).
f
75
nal specimens—particularly transverse strengths. During clamping, stresses can be

induced that cause the more fragile specimens to fracture.
In considering the results, we can conclude... Design estimates (based on RoM,
IRoM and KT models) offer the first approximation of the mechanical properties
of FRCs. In practice, design and modelling assumptions along with manufacturing
imperfections mean that estimates are likely (at best) to offer only an appraisal of
the real situation, rather than a perfect agreement.
Elastic Moduli. In Table 5.2, the RoM and IRoM estimates provide a reasonably
accurate prediction of the elastic moduli—as confirmed by others [1, 11]. In terms
of the longitudinal elastic moduli, the experimental measurements are in excess of
90% of the design estimates. Transverse moduli estimates are less accurately pre-
dicted than their longitudinal counterparts but still lie within a reasonable margin of
error. Moreover, the accuracy of transverse estimates can (as previously mentioned in
Sect. 2.3) be improved using the more complex, semi-empirical Halpin-Tsai expres-
sion in Ref. [12].
Tensile Strength. The longitudinal and transverse strengths of FRCs tend to be
more difficult to accurately predict than those for the elastic moduli since the failure
mechanisms are influenced by many factors [1]. The KT estimates in Table 5.2 are
calculated using mean fibre bundle strengths from supplier data [8, 10]. This method
is simple but tends to offer a significant overestimate of composite performance as
it does not consider the nonuniformity of fibre strengths. In reality, fibre strengths
are dependent upon the presence of flaws along their length [1]. These flaws cause a
considerable spread of fibre strengths about a mean value [11, 13]. Moreover, fibre
damage can occur during handling [11] and fabric processing [14]. In tension, the
weaker fibres will fail first resulting in a local stress concentration in the adjacent
unbroken fibres, which in turn increases the likelihood of their failure [15]. The use
of a mean fibre bundle strength is therefore likely to offer an upper bound strength
estimate whilst a lower bound estimate can be taken as the stress at the first fibre fail-
ure. To improve the longitudinal strength estimate therefore requires consideration of
the strength variability. Here, an arbitrary fibre strength factor of 0.7 in the KT model
yields estimates that closely correlate to our experiments: wet layup = 1312.8 MPa
(0.2%); vacuum bagging = 1930.6 MPa (4.1%); and prepreg = 1852.2 MPa (–5.9%).
In this chapter, tensile properties (only) have been reviewed. As an exercise for
the reader, it would be worthwhile to benchmark the design estimates suggested for
compression and shear performance (see Chap. 2). The measurement of compressive
properties for unidirectional composites (see ASTM D3410 [16] and D6641 [17])
is more difficult than for tensile properties [18], but a simple flexural test (ASTM
D7264 [19]) can be used to approximate the compressive strength (values can differ
by ≈10–20%) [20]. The in-plane shear properties (shear modulus and strength) can be
measured using a uniaxial tensile test with a ±45◦ laminate—see ASTM D3518 [21].
5.5 Factor of Safety (FoS) 77
5.5 Factor of Safety (FoS)
In a practical sense, design estimates from the RoM, IRoM and KT models offer
the designer a starting point for structural predictions. However, to ensure composite
failure does not occur in service and to offer a safety net, it is usual to make a structure
stronger than needed. This approach is justified by the fact that there is some level
of uncertainty in the material properties (e.g. fibre properties), design calculations,
manufacturing processes and the mechanical loads that a structure will experience
in service. To account for the level of uncertainty in the calculations, a FoS is often
specified.
The FoS is defined as the maximum allowable stress (or failure strength) of the
composite structure (σc∗ ) divided by the composite’s design (or working) stress (σc ),
viz.
σ∗
FoS = c (5.6)
σc
The FoS is selected as a positive real number. The chosen value depends on several
factors, but two of the most important are:
• The accuracy to which structural loads and material properties can be estimated
or characterised.
• Consequences that would result from failure of the composite structure or part
(severity of the application).
An excessively large FoS results in an overdesign, and hence, the chosen values are
usually between 1.2 and 4.0 [2]. A higher FoS is usually applied to FRCs than to
traditional (isotropic) materials and therefore values ≥ 2.0 tend to be considered the
norm.
5.6 Summary
The RoM, IRoM and KT models offer the first approximations of the mechanical
properties (elastic moduli and strength values) of FRCs.
Design estimates for elastic moduli are more accurately predicted than composite
strengths. Fibre strengths are dependent upon the presence of flaws and there can be
considerable variation in the mean value. Moreover, fibre processing and composite
fabrication methods can cause damage to pristine fibres.
The thickness of a FRC is related to fibre volume fraction (Vf ) via
n Aw
t=
ρf Vf
In practice, model assumptions and manufacturing imperfections mean design esti-

mates are unlikely to offer perfect agreement with experimental measurements.
Thus, the designer is responsible for the selection of an appropriate safety net (i.e.
factor of safety, FoS) to compensate for any potential errors, as well as for design
uncertainties and other in-service practical implications outside of the idealised test
cases used.
The FoS is defined as
σc∗
FoS =
σc
Factor of safety values are chosen based on the application and usually range from
1.2 to 4.0.
5.7 Questions
Questions 5.1 In a tensile test, a specimen is monotonically loaded until failure.

What measurements are usually recorded?
Questions 5.2 What is an extensometer and why is it used? What could be used
instead of an extensometer?
Questions 5.3 How is the tensile modulus determined from the measured data?
What about the tensile strength?
Questions 5.4 Design (RoM, IRoM and KT) estimates tend to offer an approxima-
tion of the tensile properties of FRCs. The estimations are not exact. Why?
Questions 5.5 What is a factor of safety (FoS)? Why is it important when designing
composite structures?
5.8 Problems
Problem 5.1 A tensile test is conducted on a unidirectional glass-polyester com-

posite. The stress-strain curve is shown in Fig. 5.5.
a. What is the tensile modulus of the specimen?
b. What are the longitudinal strength and failure strain?
Answer. E = 35 GPa, σcl∗ = 1000 MPa, cl∗ = 2.85%.
Problem 5.2 If the composite in Problem 5.1 is fabricated using 6 plies of 300 g/m2
glass fibres and the specimen is 1.41 mm thick, what is the fibre volume fraction?
Assume ρg = 2550 kg/m3 .
Answer. Vf = 0.5.
5.8 Problems 79
1,000

750
500
250
0
0 0.5 1 1.5 2 2.5 3
Fig. 5.5 Tensile stress-stain curve of the glass-polyester composite
Problem 5.3 A unidirectional carbon fibre-reinforced composite has a tensile

strength of 1800 MPa (experimentally measured). Assuming the member is 20 mm
wide:
a. What is the minimum thickness that is required to support a tensile load of 35 kN?
b. What thickness would be required if a FoS of 2.0 is selected for the application?
Answer. a. 0.97 mm, b. 1.94 mm.
Problem 5.4 Sketch the stress-strain curves for a HS carbon fibre-reinforced epoxy
composite with Vf = 0.45. Assume the carbon fibre and epoxy have elastic moduli of
245 GPa and 3.4 GPa, and tensile strengths of 4900 MPa and 60 MPa, respectively.
References
edn. Wiley, Hoboken
cations, 2nd edn. Society of manufacturing engineers, Dearborn, Mich
4. ASTM (2017) Standard test method for tensile properties of polymer matrix composite mate-
rials
5. Smith WF, Hashemi J, Presuel-Moreno F (2019) Foundations of materials science and engi-
neering, 6th edn. McGraw-Hill Education, New York
6. Huerta E, Corona JE, Oliva AI, Avilés F, González-Hernández J (2010) Universal testing
machine for mechanical properties of thin materials. Revista mexicana de física 56(4):317–
322
7. Curtis PT (1988) Crag test methods for the measurement of the engineering properties of fibre
reinforced plastics

9. Solvay (2021) Technical data sheet: Vtm(mohana)series (prepeg). https://catalogservice.
solvay.com/
10. Toray Inc (2021) Technical data sheet: T700s no. CFA-005. http://stg.toray.testcrafting.com
11. Agarwal BD, Broutman LJ, Chandrashekhara K (2006) Analysis and performance of fiber
composites, 3rd edn. Wiley and Chichester, Hoboken
12. Halpin JC, Tsai SW (1967) Environmental factors in composite design. Air force materials
laboratory
13. Virk AS, Summerscales J, Hall W, Grove SM, Miles ME (2009b) Design, manufacture, mechan-
ical testing and numerical modelling of an asymmetric composite crossbow limb. Compos B
Eng 40(3):249–257. https://doi.org/10.1016/j.compositesb.2008.10.004
14. Lee B, Leong KH, Herszberg I (2001) Effect of weaving on the tensile properties of carbon
fibre tows and woven composites. J Reinf Plast Compos 20(8):652–670. https://doi.org/10.
1177/073168401772679011
15. Zweben C (1968) Tensile failure analysis of composites. AIAA J 2:2325
16. ASTM (2008) Standard test method for compressive properties of polymer matrix composite
materials with unsupported gage section by shear loading
17. ASTM (2009) Standard test method for compressive properties of polymer matrix composite
materials using a combined loading compression (CLC) test fixture
18. Adams D (2019) Optimum unidirectional compression testing of composites. www.
compositesworld.com/articles/optimum-unidirectional-compression-testing-of-composites
19. ASTM (2015) Standard test method for flexural properties of polymer matrix composite mate-
rials
20. Adams D (2017) Can flexure testing provide estimates of composite strength properties? www.
compositesworld.com/articles/can-flexure-testing-provide-estimates-of-composite-strength-
properties
21. ASTM (2007) Standard test method for in-plane shear response of polymer matrix composite
materials by tensile test of a ±45◦ laminate
Chapter 6
Moulding Composite Parts
Abstract So far, this book has focused on flat, planar laminates in order to simply
and easily illustrate composite design and manufacture considerations. However, cre-
ating most composite structures (or parts) requires a mould. This chapter considers
composite design in the context of part mouldability (i.e. design for manufacture,
DFM), and offers the composite designer some guidance on mould design and con-
struction. In terms of part mouldability, special attention is given to draft angles (and
undercuts), surface textures and sharp corners. Mould design and construction are
examined in the context of mould durability; speed and ease of construction; and
the mould tool fabrication cost. A step-by-step mould-making process is presented
to enable the reader to better understand moulding considerations for laminates. A
simple composite part (a frisbee) is then moulded to illustrate the fabrication process.
6.1 Introduction
To make composite panels with a smooth surface on one side (on the mould side), a
flat plate was previously used as a mould surface. The same flat plate was used to make
composites using wet layup (Chap. 3), and vacuum bagging and prepreg moulding
methods (Chap. 4) [1–3]. Of course, not all composite laminates flat surfaces. To
create curved composite structures, or to reproduce a composite part numerous times,
a mould tool is usually needed [4]. A female mould can be used to create a laminate
with a smooth convex surface on the composite structure have, whilst a male mould
can be used to create a laminate with a smooth concave surface. Matched moulds can
be used to create a smooth surface on both the inner (concave) and outer (convex)
composite surfaces, i.e. a laminate with two finished surfaces [5, 6].
In this chapter, we focus on composite part design and how to make mouldable
fibre-reinforced composite (FRC) parts, including mould design and construction—
for further information on mould design, the reader is directed to references [5,
7], whilst mould- making examples are given in [8]. Of course, not all structures
lend themselves to composite manufacture, and it is therefore important to recog-
nise which shapes and features make a part a suitable candidate for moulding (and
which features do not!). The shape and features of a composite part can also have
https://doi.org/10.1007/978-3-030-78807-0_6
82 6 Moulding Composite Parts
an influence on the selection of mould-making materials which in turn can affect

the mould design and construction method. Other factors that need consideration
include (but are not limited to) the speed and ease of construction, mould durability
and fabrication cost.
6.2 Composite Part Design
A composite structure (or part) must be able to be extracted from the mould tool,
usually without causing damage to either the composite or the mould [7]. In some
instances, this requirement may mean that a designer needs to invest extra time
to redesign some features. It is worth the time to reconsider the shape, functional
needs and surface finish of the part prior to making a mould, as some simple design
modifications can notably improve a composite part’s mouldability [5]. To minimise
moulding issues, composite structures should usually be designed to be as simple as
possible.
Draft Angles. A draft angle defines the degree of taper on a moulded part. It is
the angle between the sides of the part (or sides of the mould) and the extraction
direction—zero, positive and negative drafts are shown in Fig. 6.1. A positive draft
angle helps a part to be demoulded, whilst a negative draft angle (or undercut) restricts
demoulding. Larger draft angles are usually required on deeper parts to prevent
adhesion as the resin shrinks onto the surface during curing [9], whilst shallow
parts can be adequately accommodated with relatively small drafts (1–3◦ ) [10]. An
appropriate draft angle minimises the friction between the composite and mould,
and reduces mould wear (increasing mould life). Sensible draft angles reduce the
likelihood of damage to the mould surface (and/or part) during the demoulding
process.
As a rule of thumb, shallow parts (up to ≈75 mm in depth) should have a draft
angle of 1◦ [1] (or more) for easy release. An extra degree of draft is recommended
here for each additional ≈25 mm depth (or division thereof) [7], viz.
depth [mm] − 75
θ ≥ 1 + (6.1)
25
Of course, it should be noted that any draft angle (however small) is better for
demoulding parts than no draft angle at all, but it is also important to be aware that
reduced drafts are likely to cause difficulty during part removal and increase the
chance of mould damage. Negative drafts or undercuts should be avoided whenever
possible as these design features complicate part removal. For parts that must accom-
modate negative drafts, a combination mould [6] (also known as a split mould) can
be used—see Fig. 6.2. Alternatively, removable inserts [5, 11] may be introduced at
the mould design step (see Sect. 6.3). In a split mould, the interface where the two
6.2 Composite Part Design 83
Extraction
Sample
Mould
Zero draft Positive draft Negative draft
(a)
Sample Extraction
Mould
Zero draft Positive draft Negative draft
(b)
Fig. 6.1 Draft angles for a male moulds; and b female moulds
(a)
Parting line
(b)
Fig. 6.2 Moulding a part with an undercut: a problematic part; and b combination mould
halves of the mould meet is referred to as the parting line; it is positioned to enable
the mould tool halves to move independently of one another and hence, facilitates
demoulding.
Example 6.1
A FRC channel section is to be fabricated in a female mould—see Fig. 6.3.
What draft angle, θ , would you recommend for the channel to minimise any
demoulding issues? What alternatives should be considered if the recom-
mended draft angle cannot be easily accommodated in the design?
110 mm
Fig. 6.3 Schematic profile of the channel
Solution
The draft angle (in degrees) should be
depth [mm] − 75
θ ≥1 +
25
110 − 75
θ ≥+ ≥ 2.4◦ ≥ 3◦
25
If 3◦ cannot be accommodated,
• A redesign of the channel could be considered.
• It may be possible to mould with a reduced draft angle (< 3◦ ). A reduction
in the draft angle will, however, increases the friction between the channel
and the mould. This will increases the likelihood of damaging the channel
and/or mould tool during demoulding.
• A split mould tool (perhaps a vertical split at the mid-plane) could be
considered.
Sharp Corners. Sharp corners or tight radii on a part are design features that can
cause problems during composite moulding and should be avoided [1]. It is difficult
to ensure fibres are tightly positioned against a mould when sharp discontinuities
are present. Trying to create sharp corners on a part during the moulding process
can lead to bridging [12]; fibres span the sudden contour change, rather than being
firmly pressed against the mould surface—see Fig. 6.4a. As a bare minimum, a fillet
of at least 5 mm is recommended here—1/4 inch (6.35 mm) is suggested in [13]. To
6.2 Composite Part Design 85
Vaccum bag
Intensifier Force
Bridge
Bridge
(a) (b)
Fig. 6.4 Bridging at the corners of a mould: a bridge across a sharp corner; and b reduced bridge
by using an intensifier
minimise bridging for composite parts with tight radii, pressure intensifiers can be
used during the vacuum bagging process [2] as shown in Fig. 6.4b.
Surface Texture. A surface texture or pattern can (in many cases) be replicated on a
composite component, but fine detail can sometimes be difficult to reproduce accu-
rately. Thus, where possible, mirror-like finishes are recommended as parts can be
released more easily from a smooth mould surface [5, 7]. Whilst surface patterns
can camouflage imperfections, they offer an increase in surface area and (as a result)
an increase in the chance for part adhesion during moulding. As mentioned, increas-
ing the surface friction makes the release of a composite part from the mould more
problematic so if rougher surface textures are unavoidable, consideration should be
given to larger draft angles [11]. Peters [7] suggests an extra 1◦ of draft angle on
vertical surfaces for each additional 0.0254 mm (0.001 inch) of texture depth.
6.3 Mould Design and Construction
Here, for simplicity, mould design and mould construction are combined in a single
section.
The design and construction of a mould sometimes starts with a plug [14]; the
terms master or pattern are used interchangeably [2, 5]. Of course, a mould can be
manufactured directly, without first making a plug [1]. Direct mould manufacture is
often employed when CNC machine tools are available, or for mould shapes with
planar or simple curves. If a plug is to be used, its design is critical to the success of
the mould and hence, to the subsequent fabrication of the composite part. A plug may
be an original structure or part (as is often the case for the home fabricator), a model
to be replicated, or as simple as a basic representation of the most important features.
The mould is created around the plug and then used to make the finished composite
part as shown in Fig. 6.5. In the figure, a male plug (i.e. the representation of the
mouldable part) is used to construct a female mould and then finally the composite
Plug Plug Composite part

Mould Mould
Fig. 6.5 Mould-making stages (from left to right): plug, creating the mould and moulding the
composite part
part with a smooth convex surface. Since the plug exhibits the same features as the
finished part, the plug must address the same design constraints previously discussed
in Sect. 6.2.
In considering the plug and mould specifications, it should be evident that the
surface finish of the plug and mould should be, at least, as good as the finish expected
on the composite part [1, 4]. A plug is rarely used more than a few times, so its
durability is often considered to be of lower importance. In contrast, a mould’s
lifespan is often the main design consideration; a mould may be used only a few
times for prototyping or many hundreds (or even thousands) of times for high volume
production parts [15].
Simple Low-Cost Moulds. Mould tools can be designed and constructed from easily
accessible and low-cost materials such as plaster, foam and wood [4, 12, 16]; these
materials can work well for many applications, but often need to be sealed [1, 17] and
sanded to provide a smooth surface finish. Plaster, wood and foam all offer a relatively
simple and quick fabrication route that requires only basic hand tools (i.e. wood and
metal working tools [5]) to ensure successful mould construction. Nevertheless, they
are not as durable as other tooling options, and they therefore tend to be used only
for prototypes (perhaps 1–5 parts) or low volume production (5–50 parts) [15].
Silicone rubber offers an alternative to plaster, wood or foam. It is often preferred
when there is a need to replicate complex or high (fine) detail, low-volume parts [2].
The use of a metal (or thin plastic) sheet formed around a substructure (often con-
structed from wood) is a further low-cost option used to increase mould durability. A
metal sheet is non-porous and therefore does not need to be sealed post fabrication.
Additive manufacturing or 3D-printing typically uses thermoplastic polymers and
is growing in popularity as a relatively low-cost option for mould-making [15, 18]. It
offers many advantages for small parts, including the promise of tighter dimensional
tolerances.
In considering mould-making materials, it is sometimes suggested that a mould
should be 3–4 times (stiffer and) stronger than the resultant laminate [5]. Extra mould
strength can obviously be achieved simply by increasing the thickness of the mould.
In some instances, however, a better option is to support the mould with a network
of braces [2]. This bracing method minimises the mass of the mould which can be
beneficial where heat transfer to the mould (during the curing process) is important.
6.3 Mould Design and Construction 87
Example 6.2
A composite fabricator wants to construct a low-cost mould for prototyping
a FRC flying disc (i.e. a FrisbeeTM ). An existing off-the-shelf low-density
polyethylene (LDPE) disc will be used as the plug—see Fig. 6.6. The fabricator
will select one of the following mould-making materials for the prototype
tooling:
a. Plaster.
b. Silicone.
Discuss the technical issues that might be foreseen for each of these materials?
Fig. 6.6 An off-the-shelf plastic frisbee
Note. The issue of whether a FRC is the most appropriate material selection
for a flying disc (frisbee) is not considered here. The flying disc is chosen on
the basis that it offers some obvious moulding challenges for a composite part.
Solution
(a) Plaster
The frisbee has sensible draft angles (no undercuts) and is shallow without
sharp corners or tight radii. Nevertheless, it still offers some clear mouldability
concerns—fine detail is evident on the flying disc’s surface in the form of a
circular pattern, and the frisbee itself (although not evident just from the photo)
may be too flexible for use as a plug without modification.
Fig. 6.7 Issues in the

reproduction of fine detail
using a plaster
Surface textures or patterns that incorporate such fine detail can be problematic
to replicate, an issue exacerbated by the use of some mould-making materi-
als, including plaster. A plaster mould is not able to reproduce the frisbee’s
pattern—see Fig. 6.7. Thus, if a plaster mould is to be used, the frisbee design
needs some modification—a smooth composite frisbee without the pattern
is one option (see Sect. 6.4). This design change, however, will affect flight
stability and may not be the most sensible solution in all circumstances—it is
a design decision!
Note. The circular pattern introduces turbulence at the frisbee’s leading edge
which helps to stabilise the airflow over the top of the disc [19]. If it is essential
to maintain the same flight stability characteristics as the original frisbee,
consideration should be given to alternative design options [20] or to other
mould-making materials, such as silicone that can replicate fine detail.
The plug flexibility must also be addressed before mould construction com-
mences. The frisbee must be stiffened to ensure it retains its shape during the
mould-making process. A simple way to do this is to simply fill the frisbee
with plaster.
(b) Silicone
The use of a silicone mould enables fine detail to be accurately reproduced [2]
and therefore removes the problem of replicating the plug’s circular pattern.
To mirror the pattern from the mould to the composite prototype, a gelcoat
could be used to reproduce the fine detail and further provide a smooth surface
coating for the flying disc—of course, in this case, the frisbee pattern is filled
only with gelcoat and would contain no fibres. Moreover, it should be noted
that gelcoats are usually brittle and, as a result, the surface would be prone to
chipping.
6.3 Mould Design and Construction 89
Higher Volume Moulds. To produce more robust moulds, FRCs (composite tooling)
or metals (hard tooling) are chosen. Composite tooling is often easier to fabricate
than its metal counterpart and is a lower cost option that uses similar materials and
fabrication methods as the laminates they are used to create. The cost and speed
to build composite tooling is still higher than the aforementioned low-cost but less
durable options of plaster, wood, foam and silicone. The use of composite tooling is
usually suitable only for low to medium volume (in some cases up to several hundred)
parts [14]. In contrast, hard tooling can be used for several hundred to several thousand
composite parts [15] but it is heavy, more expensive and often needs specialist metal
machining tools (e.g. CNC machine tools) for in-house fabrication. Hard tooling is
therefore often out of reach for the small-scale composite fabricator.
Each of the mould-making materials available to the composite designer has its
own advantages and disadvantages (or limitations), and there are many different
approaches that will result in a successful outcome. The main advantages and disad-
vantages of each mould-making material, as well as recommended uses, are shown
in Table 6.1.
6.4 Practical Task: Making a Frisbee

In this section, we expand on Example 6.2. A step-by-step guide to the design and
manufacture of a prototype frisbee (including mould design and construction) is
introduced for the plaster mould only—see Fig. 6.8a. A frisbee fabricated using a
silicone mould (see Fig. 6.8b) was produced in a similar way, but a white gelcoat
was used to create the surface pattern and to offer a smooth finish for the glass fibre
substrate.
Design Specification. The aim is to produce a simple low-cost FRC frisbee that is
easy to manufacture. A target mass of 175–200 g is identified for the frisbee [21] as
well as an improvement in disc stiffness (elastic modulus) compared to the off-the-
shelf LDPE disc (as shown in Fig. 6.6). Design improvements for the frisbee in terms
of aerodynamics and flight stability are beyond the scope of this project, but if these
issues are of interest, the reader is directed to reference [22] for further information.
Before the frisbee prototype can be fabricated (using our six-step process—see
Fig. 3.1 ), a mould is needed.
Making the Mould. Here, a plaster mould is constructed. The off-the-shelf frisbee
is used as the plug, but as previously mentioned, plaster will be unable to replicate
the circular pattern on the frisbee’s surface; hence, the pattern needs to be removed.
The pattern removal can be treated in one of the two ways:
• The pattern can be removed (smoothed) on the plug and the mould constructed
from the smoothed frisbee.
• The mould can be formed with the pattern included on the plug and then the
imperfections filled and sanded to produce a smooth mould surface.
Table 6.1 Advantages and limitations of mould-making materials [2, 5, 6, 12]
90
Material Advantages Disadvantages Used for:

Plaster
• Easy to make. • Short lifespan (1–5 parts). • Plugs and one-off moulds.
• Lowest cost option. • Needs drying and sealing. • Breakdown moulds (see Chap. 7).
• Low temperature moulding (thermal shock). • Planar surfaces and surfaces with either simple or compound
• Low (course) detail only. curves.
• Low detail moulds.
Wood
• Easy to construct. • Relatively short lifespan. • Plugs, prototyping or low volume moulds (typically <50
• Low-cost. • Needs drying and sealing. parts), or bracing structures.
• Low to moderate moulding temperatures only (heat distortion). • Planar surfaces and surfaces with simple curves.
• Complex shapes can be difficult to produce.
Sheet metal
• Easy to cut and form. • Performance depends on substructure (often wood). • Low to medium volumes.
• Complex shapes are difficult to produce. • Planar surfaces and surfaces with simple curves.
• Low detail moulds.
Foam
• Easy to shape. • Relatively short lifespan—low density foams can be easily dam- • Plugs, one-off moulds or prototyping (usually 1–5 parts).
• Low-cost. aged. • Planar surfaces and surfaces with simple or complex curves.
• Complex shapes (skill-dependent). • Needs sealing and sanding. • Low detail moulds.
• Low-medium temperature moulding. • Higher volumes are not usually possible.
Silicone Rubber
• Easy to make. • Relatively short lifespan (typically <50 parts). • Low volume moulds.
• Low-cost. • High coefficient of thermal expansion (can also be a positive). • Planar surfaces and surfaces with simple or complex curves.
• Complex shapes and fine (high) detail. • High (fine) detail moulding applications.
• Low to high moulding temperatures are possible.
Composite Tool-
• Easy to make. • Less durable than hard tooling. • Moderate volume moulds (often >50 parts).
ing
• Cheaper and lower mass than hard tooling. • Low to moderate moulding temperatures only (resin dependent). • Planar surfaces and surfaces with simple or compound
curves.
Hard Tooling
• Long lifespan. • Heavy (high density). • High volume moulds (100–1000 parts or more).
• Low to high temperature moulding. • Expensive (highest cost). • Planar surfaces and surfaces with simple or compound
• Complex shapes and high (fine) detail are possible. • Specialist machine tools are required. curves.
• High coefficient of thermal expansion (can also be a positive). • High (fine) detail moulds.
6 Moulding Composite Parts
6.4 Practical Task: Making a Frisbee 91
(a) (b)
Fig. 6.8 Fibre-reinforced composite frisbee fabricated using a a plaster mould; and b silicone
mould
Fig. 6.9 Smoothed frisbee

plug (circular pattern
removed)
It is usually better to resolve issues when making a plug, rather than later making
significant amendments to a mould (or multiple moulds). Thus, smoothing the frisbee
plug is preferred here. Sanding will smooth the surface but it will not result in the
same high-gloss finish as the original frisbee—see Fig. 6.9. The dull sanded finish
is adequate for a plaster mould as there will still be a need for some minor mould
detailing, i.e. the plaster mould will need to be sealed and sanded after it has cured.
For mould-making materials that require no sealing or sanding, the finish on the
plug should be at least to the same standard as that needed on the finished composite
part [5].
The plug flexibility must also be addressed before mould construction commences.
To increase the frisbee stiffness and ensure it retains its shape during the mould-
making process, the smoothed frisbee is filled with plaster; a plaster-based cement
(cornice cement) is used here in preference to plaster of Paris. The cornice cement
is mixed thoroughly and then poured into the frisbee base. It is left to cure as shown
in Fig. 6.10. A wet (low viscosity) mix is used to help the plaster flow and fill the
cavity.
Fig. 6.10 Plug stiffened

with cornice cement
Fig. 6.11 Frisbee plug,

finished and ready for mould
creation
Next, the stiffened and smoothed frisbee is temporarily bonded to a glass plate—a
gap of approximately 5 mm is created between the frisbee and plate. This gap is used
to lift the disc off the glass, providing a deeper cavity in the mould that facilitates
trimming and finish of the FRC frisbee after demoulding. Modelling clay is used to
bridge the gap and remove the sharp discontinuity at the frisbee-plate interface as
shown in Fig. 6.11.
The complete frisbee plug (stiffened frisbee and glass plate) is released with at
least four coats of wax [3], and a wood structure is constructed around the glass plate
to define the mould boundary—see Fig. 6.12a. To create the plaster mould, another
batch of cornice cement is mixed to the consistency of honey and then slowly poured
inside the wood boundary and over the stiffened frisbee—see Fig. 6.12b. The cornice
cement is then left to cure and hence, form the mould. The setting time will depend
upon the volume of the cement and the ambient temperature. Do not be tempted to
rush this step when making a mould from plaster; the mould may take several days
(or more) without intervention to set hard and allow the removal of the plug without
causing damage to the mould; the goal is simply to produce a smooth negative of the
frisbee plug.
Once the plaster has cured, the plug can be carefully extracted from the mould and
the surrounding wood structure can be disassembled. The unsealed plaster mould is
(a) (b)
(c) (d)
Fig. 6.12 Mould construction: a frisbee plug and wood ‘boundary’ structure; b pouring the cornice
cement; c unsealed mould; and d sealed and sanded mould
shown in Fig. 6.12c, whilst Fig. 6.12d shows the mould after it has been sealed, sanded
and polished. Here, two coats of low-viscosity (200–340 mPa · s) epoxy resin [23] are
used to seal the mould. The epoxy is left to fully cure before the surface is carefully
sanded to a smooth finish.
We now have the plaster mould, so we can make the frisbee prototype.
Frisbee Prototype. The frisbee is manufactured using wet layup with glass fibres
(chopped strand random mat, 450 g/m2 ) in a polyester matrix [24]. These material
selections offer the simplest and lowest cost options for a FRC frisbee—herein, the
focus is on satisfying the low-cost requirement but it is recognised that other alter-
native material selections are just as valid. The number of plies are chosen to satisfy
the mass and stiffness criteria—see Example 6.3. Two random mat plies (low Vf ) are
used to create the main structure of the flying disc, and a thickened rim is replicated
with three extra random mat plies—note, a heavier rim is known to improve flight
stability [22].
Example 6.3
How many plies of glass fibre random mat (Aw = 450 g/m2 ) are needed to
manufacture a FRC frisbee with a target mass of 175–200 g, and an elastic
modulus in excess of a typical LDPE flying disc. Assume the reinforced rim
has, at least, twice as many plies as the main structure of the flying disc—see
Fig. 6.13.
275 mm
250 mm
Fig. 6.13 Schematic cross-section of the frisbee
Note. The frisbee surface will need a decorative finish. To meet the frisbee’s
target mass, consideration of the weight of any surface finish is also required.
Solution
The range of fibre volume/weight fraction for wet layup is obtained from
Table 3.2, i.e. Vf = 0.1 − 0.3 and Wf = 0.14 − 0.48. Based on fabrication
experience from earlier frisbee prototyping activities using the same con-
stituents, volume fraction and weight fractions at the medium to high range are
targeted here. As a first step, we therefore assume Vf = 0.25 and Wf = 0.41;
for fabricators with less experience, consideration should be given to lower
fibre volume and weight fractions.
Mass. The target mass of the frisbee is 175–200 g. Therefore, to allow for a
painted surface finish, an unpainted composite mass of approximately 175 g is
considered.
Based on a uniform thickness, we can calculate the mass of the main structure.
For a glass fibre random mat composite (Wf = 0.41), the mass can be calculated
as
π D2 g π 0.2752
Mass of fibre (per ply) = Aw = 450 2 × = 26.7 g
4 m 4
0.59
Mass of resin (per ply) = 26.7 g × = 38.4 g
0.41
Mass of each ply = 26.7 + 38.4 = 65.1 g
Mass for 2 plies = 2 × 65.1 = 130.2 g
Assuming two plies, the thickness of the main structure is calculated from
n Aw
t=
ρf Vf
2 × 0.450
= = 1.4 mm (i.e. 0.7 mm per ply)
2550 × 0.25
Now, considering the extra rim reinforcement (width = 20 mm)
Mass of extra fibre (per ply) = Aw × π Drim × width

g
= 450 2 × π 0.25 × 0.02 = 7.1 g
m
0.59
Mass of resin (per ply) = 7.1g × = 10.2 g
0.41
Mass of each rim ply = 7.1 + 10.2 = 17.3 g
Mass for 3 plies = 3 × 17.3 = 51.9 g
Total frisbee mass = mass of main structure + mass of rim reinforcement

= 130.2 + 51.9 = 182.1g
Thus, the frisbee mass prediction should be within the target mass range (based
on the design specification), assuming not more than 17.9g is used for finishing
(painting).
Stiffness. The elastic modulus of the composite frisbee must be greater than
that of an off-the-shelf LDPE frisbee.
The elastic modulus of the glass fibres and polyester are assumed to be
70 GPa and 3 GPa, respectively, since no supplier property data was readily
available for either constituent.
Thus,
E c = E m (1 − Vf ) + E f Vf
3
= 3(1 − 0.25) + × 70 × 0.25 = 8.8 GPa E LDPE ( 0.172–0.282 GPa [27])
8
Fig. 6.14 Materials and equipment for wet layup
The stiffness prediction therefore satisfies the design specification, i.e. E c >
E LDPE .
Summary
Based on the initial design (sizing) calculations:
• Two plies (glass fibre random mat) should be used for the main structure
of the frisbee (1.4 mm thick).
• An extra three random mat plies should be used to thicken the disc’s rim
(an extra 2.1 mm).
• To meet the frisbee target mass (175–200 g), the painted coating should
have a mass not more than 17.9 g.
Prior to layup, at least 4 coats of wax release agent should be applied to the surface of
a new mould [3]. Here, six coats are used as, from previous experience, demoulding
parts from plaster moulds can prove troublesome. The wax is buffed to a sheen
between coats in accordance with the supplier’s instructions.
After waxing is complete, the layup and consolidation step commences. All mate-
rials and equipment are initially set up in a well-ventilated area—see Fig. 6.14. The
fibres are precut to the requisite size, and the resin is premixed with 1.5% catalyst (in
accordance with the supplier’s instructions) assuming a wastage factor of 1.25—see
Example 6.4. Here, the fibres are cut oversize and will be trimmed back later after
demoulding.
(a) (b)
(c) (d)
Fig. 6.15 Lamination of a frisbee: a wet out of random mat; b roller consolidation; c complete wet
out; and d rim reinforcement
To create the frisbee, a random mat ply is positioned in the mould and the polyester
is applied with a brush to wet out the random fibres as shown in Fig. 6.15a. The random
mat is worked with a stippling action to ensure the fibres are pushed flat against the
mould surface. This is repeated for the second layer of fibres. After each ply, a roller
is used to remove air trapped between the random fibres as shown in Fig. 6.15b. The
wet-out and rolled layup is shown in Fig. 6.15c. Lastly, the extra plies are added to
reinforce the disc’s rim (see Fig. 6.15d)1 . This is a tricky task to perform at this stage
so if necessary, the rim reinforcement can be added after the two plies (for the main
structure) have cured.
1Note. The image in Fig. 6.15d is not from the same frisbee fabrication process as Fig. 6.15a, 6.15b
and 6.15c.
Example 6.4
Consider Example 6.3. Assuming the resin is to be premixed with 1.5% catalyst
using a wastage factor of 1.25, how much polyester resin and catalyst are
needed?
Solution
Based on previous experience with the glass fibres and polyester, a medium-
high fibre volume fraction (for wet layup) was chosen in the earlier example,
i.e. Vf = 0.25.
Thus, Wf = 0.41 and hence
Resin and catalyst mix (g) = 182.1 × (1 − 0.41) = 107.4 g
To take account for wastage, a factor of 1.25 is applied, viz.
Resin and catalyst mix (g) = 107.4 × 1.25 ≈ 134.3 g
The polyester resin used here is mixed with only 1.5% catalyst, i.e. 98.5/100 of
resin to 1.5/100 of catalyst. Thus
Resin (g) = 134.3 × 0.985 = 132.3 g

Catalyst (g) = 134.3 × 0.015 = 2.0 g
The frisbee is left for several hours until it reaches a solid cure before it is removed
from the mould. In some instances, demoulding can be performed using a plastic
wedge to tease the disc’s rim away from the mould, allowing the plaster mould to be
reused. Of course, a notable benefit of using a plaster mould is that (in extreme cases)
it can be broken to release the composite part—others refer to this as a breakdown
method [12]. The demoulded frisbee is shown in Fig. 6.16.
To finish the frisbee, the disc’s rim is trimmed and lightly sanded to create an
even rounded edge—see Fig. 6.17a. The disc’s surface is sanded, cleaned and finally
painted. Sanding commences with using 220 grit and then continues to progress to
finer grades (i.e. P220–≈P600). A primer is used to ensure maximum adhesion and to
remove any minor surface imperfections that may still remain. The frisbee is finished
with a black acrylic paint.
Fig. 6.16 Demoulded FRC frisbee with excess fibres
(a) (b)
Fig. 6.17 Frisbee post-processing: a trimmed; and b sanded and painted
Physical Testing. The final mass of the composite frisbee is measured at 182.0 g. This
is not too dissimilar to the estimated unpainted mass of 182.1 g—see Example 6.3.
By flexing the composite flying disc, it is clear that the prototype is much stiffer than
the off-the-shelf LDPE frisbee that was used to make the plaster mould. So, now all
that is left to do is to test the flight characteristics of our FRC prototype!
6.5 Summary
To create useful composite structures or parts, consideration should be given to part

design in the context of mouldability as well as to mould design and construction
issues.
In terms of composite part design, the following questions (as a minimum) should
be reviewed:
• Are the draft angles appropriate?
• Are there negative drafts (undercuts)?
• Are there any surface texture issues?
• Are there any sharp corners or tight radii?
For mould design and construction, the following questions are mentionable:
• Is there a need for a plug? If yes, will an existing part be used, or does the plug
need to be fabricated?
• How many parts will be moulded?
• How accurate does the finished laminate (and hence, mould) need to be? If
there are tight tolerances, specialist machine tools may be needed for mould
construction.
• How much time is available to build the mould tool, and what is the budget?
When considering mould-making materials, the best selection depends on many fac-
tors including the speed and ease of manufacture; surface finish requirements; mould
durability; and fabrication cost.
Mould-making materials for simple low-cost tooling include:

• Plaster.
• Wood.
• Foam.
• Silicone.
For higher volume moulds, materials typically used are:
• Composite tooling.
• Hard (metal) tooling.
6.6 Questions
Question 6.1 What is a draft angle and why is it important in moulding FRCs?
Question 6.2 What is the minimum (recommended) draft angle for shallow com-
posite parts? For deeper parts, what should the draft angle be?
6.6 Questions 101
Question 6.3 Why is it important to avoid sharp corners on composite parts? What
minimum fillet radius would you recommend?
Question 6.4 What is bridging and why should it be avoided?
Question 6.5 Why should surface textures on composite parts be avoided?
Question 6.6 What is a plug? What other terms are used interchangeably to mean
plug?
Question 6.7 Why is the durability of a plug usually of lesser importance than that
of a mould?
Question 6.8 List four low-cost mould-making materials and two materials that
might be used for higher volume moulds.
6.7 Problems
Problem 6.1 Which mould in Fig. 6.18 is male, which is female and which is a
matched mould?
Answer. a. Female; b. male; c. matched mould.
Problem 6.2 Which of the moulds in Fig. 6.19 are likely to allow the composite part
to be demoulded from the tool? If demoulding is possible, comment on how difficult
this might be.
Problem 6.3 The cross-section of a thin-walled composite part and its mould tool
are shown in Fig. 6.20. Identify issues that might arise when moulding the part and
then make suggestions (in terms of part and/or mould design) to ensure a successful
composite project.
Answer. Issues include mould stiffness(?), sharp corners, no drafts on some sections
and an undercut (negative draft).
Problem 6.4 Identify the most appropriate mould-making material based on the
moulding requirements in Fig. 6.21.
Answer. Top-to-bottom: sheet metal (wood substructure); metal; and composite.
(a) (b) (c)
Fig. 6.18 Male, female and matched moulds (mould types to be identified)
Sample
Mould
(a) (b)
Parting line
(c) (d)
Fig. 6.19 Demoulding (demoulding issues, if any, to be identified)
Fig. 6.20 A composite part and its mould tool (moulding issues to be identified)
Mould requirements Mould material
• Low-cost. Plaster
• Simple and quick to fabricate.
• Low to medium volumes.
Wood
• Cost is not the most important factor. Composite

• Mould has complex geometry.
• High volumes (1000+ composite parts).
Metal
• Low to moderate cost. Sheet metal (wood

• Medium volumes (100–200 composite parts). substructure)
Fig. 6.21 Mould requirements and material selection (materials to be identified)

References 103
References
3. Wanberg J (2009) Composite materials: fabrication handbook #1, composite garage series, vol
1. Wolfgang Publications, Stillwater, Minnesota
4. Fibreglast (2019) Mold construction guide. https://www.fibreglast.com/product/mold-
construction/LearningCenter
6. Wang R, Zheng S, Zheng Y (2011) Polymer matrix composites and technology. Woodhead
publishing in materials, Woodhead Publication and Science Press, Oxford and Philadelphia
and Beijing
7. Peters ST (1998) Handbook of composites, 2nd edn. Chapman & Hall, London
9. Biron M (2004) Thermosets and composites: technical information for plastics users/Michel
Biron. Elsevier, Oxford
10. Mallick PK (1997) Composites engineering handbook, materials engineering, vol 11. Marcel
Dekker, New York
11. Carlsson LA, Gillespie JW (1989-91) Delaware composites design encyclopedia. Technomic,
Lancaster
13. Aird F (2014) Fiberglass and other composite materialshp1498: a guide to high performance
non-metallic materials for automotiveracing and marine use. includes fiberglass, kevlar, carbon
fiber, molds, structures and materials. HP Books, New York
14. Pauer R, Pokelwaldt A (2017) Composite molds: choices and considerations. http://
compositesmanufacturingmagazine.com/2017/11/best-practices-for-choosing-composite-
molds/
15. CompositesWorld (2016) Tooling. https://www.compositesworld.com/articles/tooling
17. Weatherhead RG (1980) FRP technology: fibre reinforced resin systems. Springer, Netherlands
18. CompositesWorld (2015) A growing trend: 3d printing of aerospace tooling. https://www.
compositesworld.com/articles/a-growing-trend-3d-printing-of-aerospace-tooling
19. Motoyama E (2002) The physics of flying discs. URL:
https://www.people.csail.mit.edu/jrennie/discgolf/physics.pdf
20. Polar Manufacturing (2020) Inspiration. www.polar-manufacturing.com/inspiration/
21. Carter J (2020) 10 best frisbees in 2020. www.gearhungry.com/best-frisbee/
22. Potts J, Crowther W (2002) Frisbee(tm) aerodynamics. In: 20th AIAA applied aerodynamics
conference, American Institute of Aeronautics and Astronautics, Reston. https://doi.org/10.
2514/6.2002-3150
23. Kinetix (2019) Laminating: R118 infusion. http://atlcomposites.com.au/icart/products/14/
images/main/KINETIXR118&H115-H120-H126-H103.pdf
24. Protite (2019) Fibreglass resin. https://www.protite.com.au/products/fibreglass/fibreglass-
resin/
edn. Wiley, Hoboken
Chapter 7
Hollow Sections—How to Make
Composite Tubes
Abstract This chapter considers the manufacture of hollow sections. Mandrel lam-
ination (wrapping) and bladder moulding are described and then used to create
composite tubes. A step-by-step guide to tube manufacture is presented for prepreg
moulding, but the fabrication methods introduced are more widely applicable (and
hence, can be adapted for use in wet layup processes). Whilst the focus is on cylin-
drical tubes, the fabrication process can be simply modified to create square or
rectangular hollow sections or more complex tubular structures. A helical spring
and a bicycle handlebar are used as more complex examples. Demoulding methods
for mandrel wrapped tubes are discussed. A specific focus is given to mechanical
extraction and thermal (heating and cooling) mechanisms. The concept of mouldless
composite construction is introduced. The importance of a mandrel’s coefficient of
thermal expansion is considered in the context of demoulding using thermal methods.
7.1 Introduction
The design and manufacture of composite laminates has been considered in earlier
chapters. A composite can have a high stiffness and strength at fibre orientations in
the plane of the plies, but tends to offer minimal resistance to out-of-plane loads [1].
To improve the flexural and torsional performance of a composite laminate without
adding unnecessary weight, the laminate can be effectively ‘thickened’ by using a
hollow section.
Hollow steel sections are widely used in structural applications. The volume of
material is located at the outer extremes of the cross-section and this maximises the
section properties [2], improving bending and torsional performance per unit mass.
The manufacture of composite hollow structures offers weight benefits over those
from traditional materials, and hence enables fabrication of even lighter and more
economical structures.
There are numerous methods for manufacturing fibre-reinforced composite (FRC)
hollow sections, but two simple methods are mandrel lamination (wrapping) [3] and
bladder moulding [4]. In this chapter, a step-by-step approach for creating a cylin-
drical tube is presented for these two simple methods using prepreg. The approach
https://doi.org/10.1007/978-3-030-78807-0_7
106 7 Hollow Sections—How to Make Composite Tubes
is similar for wet layup. The methods can be easily adapted for square or rectangular
cross-sections, or complex hollow shapes with cross-sectional properties that vary
along the length of the composite. To illustrate this, the chapter finishes with two more
complex design and manufacture examples: a helical spring and a bicycle handlebar.
7.2 Mandrel Lamination
In mandrel lamination, the fibres are wrapped around a mandrel [3], as shown in
Fig. 7.1. A wet layup method [3] or prepreg moulding [1, 5] may be used. If wet
layup is preferred, the resin can be introduced to the fibres either prior to or whilst
wrapping each layer of fibre. The mandrel acts as the mould tool to create a smooth
inner surface on the tube, similar to a male mould [6]. To create open-ended parts,
a mandrel is often simply a cylindrical or tapered (conical) mould tool constructed
from steel or aluminium [1, 7]. The mandrel defines the internal tube diameter and
surface finish, but the external features of the tube are not precisely controlled [8].
The open-ends facilitate the removal of the finished (cured) composite from the
mandrel [9].
Mandrel lamination is more complicated for closed-ended parts [1], or for parts
with zero or negative drafts. Herein, we will consider tubes with no draft angle. For
closed-ended parts, the simplest method is to leave the mandrel (often made of foam)
inside the part after curing. If mandrel removal is needed, one method is to use a
dissolvable mandrel [7]; sometimes, this is referred to as mouldless composite con-
struction [10]. Polystyrene foam is compatible with epoxy resins (but not polyester)
and can be dissolved with acetone [3] whilst wax mandrels (candle or paraffin wax)
can be melted at elevated temperatures (ranging 50–90 ◦ C). A low-cost and water-
soluble mandrel can be easily created by combining sand and polyvinyl alcohol [11],
or alternatively breakdown materials such as plaster can be used [7].
To demonstrate the mandrel lamination process, a cylindrical aluminium (6061)
mandrel is used to create a straight hollow tube comprising four layers of 200 g/m2
carbon twill weave prepreg; of course, a similar lamination outcome is possible with
a wet layup process or by using other fibre types and forms. It should be noted that
Fig. 7.1 Mandrel lamination Composite layers

for an open-ended hollow
tube
Mandrel
7.2 Mandrel Lamination 107
Fig. 7.2 The aluminium mandrel: 25mm OD with a smooth (mirror) finish
the six steps to design and make a FRC (see Fig. 3.1) are once again followed; the
steps are not explicitly referenced herein, as it should (by now) be trivial to relate
the specific fabrication detail to the relevant step in the composite manufacturing
process.
The mandrel has an outer diameter of 25 mm and the surface is polished to a
mirror finish—see Fig. 7.2. The warp fibres in each ply are aligned with the mandrel
axis, whilst the weft fibres are orientated circumferentially around the tube.
For each of the four plies, a rectangle 250 mm long (warp) and 90 mm wide
(weft) is cut from the roll of prepreg. The rectangle’s width is slightly greater than
the circumference of the mandrel (approximately 10 mm extra) to allow one ply
of fibres to exceed one complete rotation around the mandrel circumference. This
method is preferred here but alternative processes can also be used, depending on the
fabricator’s preference; for example, a Swiss roll layup [12] or a multi-layer wrapping
process (see Sect. 7.3).
To prepare the mandrel, it is first cleaned with warm water and a lint-free cloth,
then left to dry. At least four layers of high-temperature wax (suitable up to 120 ◦ C)
are applied to the mandrel. Between each layer, the wax is carefully buffed to an
even sheen in accordance with the supplier’s instruction.
During layup, the prepreg is carefully wrapped around the mandrel—see Fig. 7.3.
Tension is applied to the prepreg during roll wrapping to prevent creases. Here, the
wrapping process focuses on one layer at a time. After the first ply is complete and
smoothed against the mandrel’s surface, the second, third and finally the fourth plies
are wrapped. The start of each ply commences close to the finishing position of the
previous one.
In the earlier flat plate demonstration of prepreg moulding (see Chap. 4), the com-
posite was cured in an oven and consolidation was performed using a vacuum bag.
Here, an opportunity is taken to introduce an alternative consolidation method using
a heat shrink tube [13]—normally, this method of consolidation is used with wet
resin. The shrink tube is placed over the fibres and mandrel (as shown in Fig. 7.4a),
and is then either placed in the oven to shrink onto the tube during curing, or gently
and uniformly heated (prior to curing) with a heat gun. The shrink tube should have
an appropriate contraction temperature to ensure adequate consolidation as com-
posite performance and surface finish are significantly influenced by consolidation
pressure [14]. Here, the shrink tube and prepreg are simply placed directly in the
(a) (b)
Fig. 7.3 Mandrel lamination: a prepreg tension; and b wrapping process
(a) (b)
Fig. 7.4 Consolidation using a heat shrink tube; and b shrink tape
oven, allowing the shrink tube to consolidate the prepreg during curing at a dwell
temperature of 120 ◦ C.
A similar alternative to shrink tube is a shrink tape [15]. Shrink tape works in an
equivalent manner to its tubular counterpart but is wrapped in a spiral manner around
the preform, rather than slid over the top of the fibres and mandrel. Each spiral wrap
of shrink tape should overlap the previous ones by at least half of the tape width [3]
as shown in Fig. 7.4b.
Here, the composite tube is left to cure on the mandrel and the shrink tube is carefully
cut and removed. A post-cure is performed before the composite part is extracted
(demoulded) from the mandrel.
If some encouragement is needed during demoulding, three methods can be used to

assist mandrel removal from an open-ended part:
• Mechanical extraction [3].
• Thermal (heating and cooling) methods [12].
• Compressed air-assisted demoulding [16].
Mechanical Extraction. Demoulding is often achieved by mechanical extraction.
It is particularly common when wet layup is used. A constant force is applied to
the mandrel (often using a press [3]) whilst the tube is suitably restrained (or vice
versa). An alternative method is to introduce a gentle tap (dynamic force) with a
mallet. The amount of mechanical force needed will depend upon the interference
fit (between the FRC and mandrel) that results from the resin shrinkage during the
curing process [11, 17].
Thermal (Heating and Cooling) Methods. Heating causes an aluminium mandrel
to expand and cooling results in contraction. The amount of expansion or contrac-
tion is identified by the coefficient of thermal expansion (CTE). If a mandrel has a
higher CTE than the composite, it will expand more when heated and contract more
when cooled. This can be used to release the interference fit caused by the shrinking
resin [12].
For a unidirectional FRC, the axial (longitudinal) CTE (αcl ) is given by [11]
αm E m (1 − Vf ) + αf E f Vf
αcl = (7.1)
E m (1 − Vf ) + E f Vf
where αm and αf are the coefficients of thermal expansion of the matrix and fibres,
respectively, E m and E f are the elastic moduli values, and Vf is the fibre volume
fraction.
The transverse CTE (αct ) for a unidirectional composite is given by [11]
αct = αm (1 + νm )(1 − Vf ) + αf Vf (7.2)
where νm is Poisson’s ratio of the matrix.

The fibres in a woven and random fibre composite are not all orientated in one
direction. Woven fabrics have fibres orientated in both the warp and weft directions
and hence, the CTE of the composite in each of these orientations will be similar,
and will lie between the longitudinal and transverse threshold coefficients. Typically,
the CTE is much closer to the longitudinal values [18]. The coefficients for random
FRCs will be similar in all in-plane orientations and will also lie between the threshold
values.
Compressed Air-Assisted Demoulding. The use of compressed air can be a simple
and effective method for demoulding a composite part. A hole (or a series of holes)
is drilled in the mandrel so that low-pressure air can be blown between the surface
of the mandrel and the composite. This air pressure helps to release the composite
from the mandrel.
In our mandrel lamination demonstration, the prepreg (and hence, mandrel) is heated
during the curing process and then cooled to room temperature prior to demoulding.
The elevated cure temperature for prepreg therefore effectively introducing a thermal
assisted demoulding step as part of the usual manufacturing process. Thus, minimal
force is needed to break any remaining seal and the tube simply slides off the mandrel
despite the zero draft angle. However, if the curing process was performed at room
temperature (as is the case for wet layup—see Chap. 3), the thermal demoulding
method would require a reduction of temperature to shrink the mandrel from the
FRC. In this instance, the composite can be placed in cold storage (−18 ◦ C) or
introduced to liquid nitrogen (about −200 ◦ C) [19].
Example 7.1
Mandrel lamination is carried out with woven carbon fibre-reinforced epoxy
prepreg on an aluminium mandrel. Assume αc is about 4 µm/m◦ C [18] for the
composite (i.e. for both the longitudinal and transverse orientations), and that
of the aluminium mandrel αman is 23.6 µm/m◦ C [20]. If the prepreg is cured at
120 ◦ C and then cooled to a room temperature of 20 ◦ C prior to demoulding,
calculate the clearance between the mandrel and the tube. Assume no adhesion
occurs and neglect the shrinkage of the epoxy during curing.
[Note. This is an estimation for the carbon fibre tube we are fabricating here.]
Solution 7.1
During the curing process, the mandrel and composite are heated up from room
temperature (20 ◦ C) to 120 ◦ C, and then cooled back down to 20 ◦ C. The diam-
eter of the mandrel at room temperature is dman,20 . The diameter expands as
the temperature increases to dman,120 and finally returns to its original diameter
dman,20 when cooled back to room temperature.
During heating, it is assumed that the FRC can expand freely since it has not
yet reached a solid cure. In other words, the composite diameter dc,120 will
follow that of the mandrel.
Thus, the expansion of the mandrel diameter is
dman = dman,20 αa t
= 25 × 10−3 × 23.6 × 10−6 × (120 − 20)
= 59 × 10−6 m = 0.059 mm
and hence, the composite diameter after it is fully cured is
→ dc,120 = dman,120 = dman,20 + dman = 25.059 mm
During the cooling phase, the mandrel returns to its original diameter of 25 mm,
whilst the composite tube undergoes a contraction related to its coefficient of
thermal expansion, αc ,
dc = dc,120 αc t
= 25.059 × 10−3 × 4 × 10−6 × (20 − 120)
= −10.0236 × 10−6 m = −0.0100236 mm
→ dc,20 = dc,120 + dc = 25.059 − 0.0100236 = 25.049 mm
The final diameter of the FRC after the curing process is dc,20 = 25.049 mm.
The clearance between the mandrel and composite will therefore be
0.049 mm.
To produce a smooth surface finish on the outside of a tube fabricated using mandrel
lamination requires secondary processing tasks. Here, the ends of the composite tube
are carefully trimmed using a band saw. The outer surface is then sanded, moving
through a series of grits until greater than 400 grit, sanding finished at P800 (i.e.
400–500 grit). The surfaces are cleaned with a solvent (acetone) and rinsed with
water then left to air dry before a clear spray paint (clear coat) is used as a decorative
finish. Several layers of clear coat are applied in thin coats to produce a smooth and
glossy finish. The carbon tube is shown in its near-finished state in Fig. 7.5—only a
few minor surface imperfections remain.
Fig. 7.5 Carbon fibre-reinforced tube manufactured using mandrel lamination

7.3 Bladder Moulding
Bladder moulding [4] uses a cavity mould tool and an inflation bladder—see Fig. 7.6.
Silicone or latex are commonly used elastomeric bladders [21], but a nylon film sheet,
cut to shape and bonded on its edges has been used as a low-cost option [1, 4]. During
moulding, the composite layup is positioned between the bladder and mould cavity.
The bladder is internally pressurised and expands to fill the void and the composite
is consolidated against the inner surface of the mould. The cavity therefore defines
the external shape and surface finish of the hollow structure. To remove the bladder,
the inflation pressure is simply reduced and the bladder returns to its undeformed
shape.
Here, a hollow tube is created using a split (two-part) aluminium mould (6061)—
see Fig. 7.7a. The internal mould cavity is used to create an external tube diameter
of about 27 mm. Each of the split mould halves has four alignment holes. Moreover,
there are four threaded holes in the upper half of the mould and four blind recesses in
the lower half (at matching locations). The threaded holes and the matching recesses
are to be used in demoulding and will be revisited later. To ensure the mould halves
are correctly located, alignment pins are inserted into the alignment holes during
the layup and the moulding process. The pins retain the two mould halves in place,
ensuring their cavities align. A silicone tube, knotted at one end, is used as the bladder.
The other open-end of the bladder is sealed against two matching bevelled flanges—
see Fig. 7.7b. The mating surfaces compress the silicone and seal the bladder. A
similar arrangement is shown in Ref. [4]. At the knotted end, a blanking plate and an
end cap with a hemispherical cavity are inserted in the mould to control the expansion
(and prevent splitting) of the silicone bladder during the moulding process.
To create the tube, four layers of twill weave prepreg are again used—see Sect. 7.2.
The prepreg plies are placed on top of each other with a slight offset (see Fig. 7.8a),
rolled to remove trapped air and then loosely wrapped into a tubular preform around
the bladder. The offset for each ply ensures a gradual transition between the prepreg
Split mould Bladder
Air pressure
Prepeg layup
Fig. 7.6 Bladder moulding (schematic)

7.3 Bladder Moulding 113
Blanking plate
End cap
Silicon tube
Bladder inflation seal
(a) (b)
Fig. 7.7 Aluminium split mould: a mould and silicone bladder; b bladder inflation seal
Layer 3 Layer 2
Layer 4 Layer 1
(a) (b)
Fig. 7.8 Bladder Moulding: a prepreg layup; and b tube preform
layers. The prepreg layup and silicon bladder are then positioned in the mould as
shown in Fig. 7.8b.
The mould is assembled and clamped shut. The inflation pressure is gradually
increased up to 2 bar and the bladder is checked to ensure there are no leaks. The
(a) (b)
Fig. 7.9 Demoulding: a mould separation; and b cured composite tube
Fig. 7.10 Finished carbon fibre tube manufactured using bladder moulding
mould is placed in the oven and the prepreg is cured according to the supplier’s
instructions. Bladder pressure is continuously monitored to ensure no leaks occur
during the curing process.
After curing, the carbon fibre-reinforced tube is demoulded. The silicone bladder
is deflated, and the mould is deconstructed. The mould is separated here with minimal
effort, but if some difficulty is experienced splitting the mould halves, four socket
head cap screws can be inserted into the threaded holes in the upper mould half. The
screws can be tightened against the recesses in the lower mould half and the stubborn
mould can be jacked apart (Fig. 7.9a) to reveal the composite tube (see Fig. 7.9b).
The cured part has a smooth surface finish with few imperfections and thus,
minimal post-processing is needed for a bladder moulded part. The ends of the tube
are trimmed to length and the part is lightly sanded (to P800, or 400–500 grit). To
finish the tube, the same clear coat used earlier in the mandrel lamination process
is introduced to again provide a smooth surface over the carbon fibre aesthetic—see
Fig. 7.10.
7.4 Practical Task: a Helical Spring 115
Fig. 7.11 Schematic of a

helical spring (with
dimensions)
25 mm
40 mm
7.4 Practical Task: a Helical Spring
In this section, we extend the manufacture of hollow structures to consider a more

complicated shape—a helical (coil) spring. To create a hollow tube, we use carbon
fibre braided sleeves over a wax rod to create a mouldless composite construction.
The spring coil preform is then achieved by spiralling the mouldless construction
around a 3D printed mandrel.
Design Specification. The helical compression spring shown in Fig. 7.11 is to be
manufactured using a carbon fibre-reinforced epoxy composite. The spring design
requires a compression stiffness (referred to as the spring rate), k, of greater than
20 N/mm and a displacement or travel, x, of at least 5 mm. The spring should have at
least 3 turns (windings), a mean coil diameter of 40 mm and a spring pitch of 25 mm.
Composite Design. The helical spring is fabricated using a wet layup process with
carbon fibre (braided sleeve, 350 g/m2 [22]) in an epoxy matrix [23]. Two plies of
carbon fibre are used to fabricate a hollow circular cross-section—see Example 7.2.
Example 7.2
How many carbon fibre plies are needed to manufacture the helical spring in
Fig. 7.11? The spring rate, k, must be greater than 20 N/mm and the deflection
should be at least 5 mm.
For a helical spring with a hollow cross-section, assume the spring rate and
deflection can be approximated by
G c × (do4 − di4 )
k=
8 C3 N
and
τc π C 2 N
x=
do G c
Note. G c is the shear modulus of the composite and τc is the shear stress. The
parameters do and di are the outer and inner diameters of the spring cross-
section, respectively, C is the coil mean diameter and N is the number of
turns.
Solution 7.2
In this example, we assume the helical spring has an inner diameter of 10 mm.
Carbon fibre braided sleeves are used as the reinforcing fibres, but it is noted
that other fibre selections could produce an equally successful outcome. The
braids are selected in this instance to provide fibre angles of approximately
±45◦ to the spiralling spring axis. In doing so, the fibres offer the maxi-
mum resistance to the torsion load applied to the spring’s cross-section—see
Sect. 2.6.
The fibre and matrix properties assumed here are shown in Table 7.1.
Table 7.1 Fibre and matrix properties
Material Elastic modulus Tensile strength Poisson’s ratio
(GPa) (MPa)
Carbon fibre 240 [22] 3800 [22] 0.22 [24]
Epoxy 3.65 [23] 83.3 [23] 0.39 [24]
The range of fibre volume fractions for wet layup of woven fabrics (including
braids) is 0.2–0.4—see Table 3.1. Therefore, we initially take a mid-range
Vf = 0.3 for our design estimates.
Since the shear performance of a laminate with fibres orientated at +45. or -45.
is the same, we approximate the woven laminate as a UD composite (with the
correct Vf) offset at 45. to the shear loading.
Spring Rate. The spring rate is dependent upon the shear modulus of the FRC,
G c45 , via
G c45 × (do4 − di4 )

k=
8 C3 N
The shear modulus (at ±45◦ ), G c45 , is calculated from
1 1 + 2νc 1
= +
G c45◦ E cl E ct
(7.3)
where the longitudinal modulus, E cl , is
E cl = E m (1 − Vf ) + E f Vf
= 3.65 × 109 (1 − 0.3) + 240 × 109 × 0.3
= 74.6 × 109 N/m2 = 74.6 GPa
and the transverse modulus, E ct , is
1 (1 − Vf ) Vf
= +
E ct Em Ef
1 (1 − 0.3) 0.3
= +
E ct 3.65 × 10 9 240 × 109
E ct = 5.2 × 10 /m = 5.2 GPa
9 N 2
Poisson’s ratio, νc , is given by
νc = νm (1 − Vf ) + νf Vf
= 0.39 × (1 − 0.3) + 0.20 × 0.3
= 0.33
Thus,
1 1 + (2 × 0.33) 1
= +
G c 45◦ 74.6 × 10 9 5.2 × 109
G c 45◦ = 4.7 × 10 /m = 4.7 GPa
9 N 2
(7.4)
Taking di = 10 mm, we can rearrange the spring stiffness equation to calculate

the minimum outer diameter, do , to be

8 k C3 n
do ≥ + di4
4
G c 45◦

8 × 20000 × 0.043 × 3
≥ + 0.014
4
4.7 × 109
≥ 0.0113 m = 11.3 mm
The thickness of each ply (sleeve) is
n Aw
t=
ρf Vf
1 × 350
= = 0.66 mm
1780 × 0.3
and hence, the number of sleeves is calculated to be
(11.3 − 10)/2
n= = 0.98
0.66
Two carbon fibre sleeves are conservatively selected here. This is to account
for a factor of safety (see Sect. 5.5) that addresses design and manufacturing
inaccuracies and other uncertainties. The selection will yield an outer diameter
of approximately 12.64 mm and a spring rate of
4.7 × 109 × (0.012644 − 0.014 )

k=
8 × 0.043 × 3
= 47509 N/m = 47.5 N/mm > 20 N/mm
Deflection. The spring deflection at failure is dependent upon the shear mod-
ulus and shear strength of the braided composite, viz.
τc∗45◦ π C 2 N
x=
do G c45◦
As previously mentioned, the estimation of composite strength under off-axis

loading is a complex task that is beyond the scope of this textbook—see
Sect. 2.6. However, it has also been noted that the maximum shear strength
occurs with fibre orientations at ±45◦ . Since the braided sleeves have fibres at
±45◦ , we can use the in-plane shear strength (τc∗ ) as a conservative estimate
∗
of τc45◦.
The in-plane shear strength is estimated from
τc∗ ≈ τm∗ = 0.5 × σm∗ ,

= 0.5 × 83.3 × 106
= 41.7 × 106 = 41.7 MPa
Thus, at failure, the spring deflection is defined as

τc∗ π C 2 N
x>
do G c 45◦
(41.7 × 106 ) × π × 0.042 × 3
>
0.01264 × (4.7 × 109 )
> 0.0106 m = 10.6 mm
This conservative (low bound) estimate of spring travel (using two braided
sleeves) exceeds the minimum of 5 mm.
Summary
The design estimates have indicated that two carbon fibre braided sleeves
(Aw = 350 g/m2 ) will exceed the minimum stiffness of 20 N/mm, and permit more
than 5 mm of spring travel (displacement). The estimates are based on a helical
(coil) spring with a circular hollow cross-section and inner diameter of 10 mm.
Manufacturing Process. To make the mouldless construction for the helical spring,
a wax rod of approximately 10 mm is used as the inner hollow core of the helical
spring—see Fig. 7.12a. The carbon fibre braided sleeves (10 mm diameter, 350 g/m2 )
[22] are precut and the epoxy resin [23] is mixed in accordance with the supplier’s
instructions, assuming a wastage factor of 1.25—see Example 7.3. It should be noted
that the braided sleeves are cut slightly longer than needed to allow the spring to be
trimmed to length after the composite is demoulded.
Example 7.3
Consider Example 7.2. How much resin and how much hardener are needed
to fabricate the helical spring (with two braided sleeves)?
Solution 7.3
The two carbon fibre braided sleeves (10 mm diameter) were cut to 700 mm
long. The surface area of each sleeve was therefore π × 0.01 m × 0.7 =
0.022 m2 .
The fibre volume fraction was previously chosen based on the woven carbon
fibre values for wet layup in Table 3.2, i.e. Vf = 0.3.
For two sleeves, Wf = 0.38 and hence
(a)
(b) (c)
(d)
Fig. 7.12 Spring layup process: a wax rod; b wet out of braids; c 3D printed mandrel; and d
creating the helical preform
A × n × Aw × (1 − Wf )
Resin and hardener mix (g) =
Wf
0.022 × 2 × 350 × (1 − 0.38)
= = 25.1 g
0.38
To account for wastage, a factor of 1.25 (between 1 and 1.5) is used for the
resin and hardener mix. Thus
Resinandhardenermix(g) = 75 × 1.25 = 31.4 g
The epoxy resin used in this case has a resin to hardener ratio of 4:1 [23], i.e.
4/5 of resin and 1/5 of hardener
Resin (g) = 31.4 × 4/5 = 25.1 g

Hardener (g) = 31.4 × 1/5 = 6.3 g
The carbon fibre braids are initially wet out (see Fig. 7.12b) before the sleeves are
carefully slid over the top of the wax rod, one at a time. To assist with the layup, the
sleeve diameter can be increased by compressing the tube’s length and then, once in
place, decreased by stretching the sleeve. Excess resin is removed before a spring
preformed is created.
To make the spring preform, the carbon fibre sleeves (and internal wax rod) are
gently formed around a 3D printed mandrel which has been pre-released with a
wax release agent. The mandrel has a semi-circular groove that spirals around its
circumference to define the spring shape—see Fig. 7.12c. The spiral preforming
process is illustrated in Fig. 7.12d. The spring is left to cure on the mandrel for
at least 24 h before demoulding. During demoulding, the spring is simply rotated
(unscrewed) from the mandrel.
A post-cure temperature of 80 ◦ C is used to melt the wax rod and hence, produce
the hollow coil. The helical spring is then trimmed to length. Here, a spring that is
slightly longer than three turns is manufactured—see Fig. 7.13. The excess length is
used as a means to locate the spring in a prefabricated housing for physical testing.
Physical Testing. As mentioned, the FRC spring is located in a spring housing
for testing on an Instron Universal Testing System (UTS)—see Fig. 7.14a. The 3D
printed housing is used to constrain the lateral movement and rotation of the spring;
motion is only unrestrained along the spring’s compression axis. After a small preload
is applied to the spring, the load-deflection characteristics are recorded at a displace-
ment rate of 2 mm/min (up to 5.5 mm). The experimental measurements are compared
to the spring rate and deflection requirements (from the design specification) in
Fig. 7.14b. In the figure, the spring’s performance is shown to exceed the design
requirements.
Fig. 7.13 Helical spring after demoulding and trimming
Fig. 7.14 Spring testing: a spring with the housing under compression, and b the mechanical
response of the spring
7.5 Extension Task: a Bicycle Handlebar
Design Specification. As a further example, a carbon fibre-reinforced mountain bike

(MTB) handlebar is fabricated (see Fig. 7.15) to satisfy the standard lateral bending
test described in BS EN ISO 4210-5:20141 [25]. In the lateral bending test for an
1 The lateral bending test is one of several tests specified in BS EN ISO 4210-5:2014 [25]. To be
compliant, a bicycle handlebar must satisfy all relevant test procedures listed in the safety standard.
7.5 Extension Task: a Bicycle Handlebar 123
+0.1
1
22.2 - 0.
17
1°
HANDLEBAR STEM
WILL ATTACH HERE
+2
680 - 2
200 70 200
+0.1
31.8 - 0.1
Fig. 7.15 MTB handlebar (all dimensions in millimetres). Reproduced from [26] with permission
from N. Emerson
MTB, a force of 1000 N must be applied (for 1 min) at a distance of 50 mm from the
free end of the handlebar. In a successful test, the FRC handlebars must not fail (i.e.
either crack or fracture) under the loading conditions.
Composite Design. To create the MTB bicycle handlebar, six unidirectional plies
and four plies of twill weave carbon fibre prepreg (both 200 g/m2 ) are chosen. This
selection is made based on design calculations and some initial assumptions. The
result is far from an optimised solution but offers a starting point—see Example 7.4.
Here, we make and evaluate a handlebar prototype using carbon fibres orientated
along the handlebar length and around the circumference only. Despite the obvious
benefit for shear (transverse shear and torsion) performance, no ±45◦ fibres are used
in this example. The inclusion of off-axis fibres is essential to improve handlebar
performance and is left as an optimisation exercise for the reader. A more compre-
hensive analysis is possible with classical laminate theory [27], extended layer-wise
theory [28] or finite element analysis (FEA) [29–31], but these more complex meth-
ods are beyond the scope of this textbook.
Example 7.4
A lateral bending test for the MTB bicycle handlebar (see Fig. 7.15) is schemat-
ically shown in Fig. 7.16. Carbon fibre prepreg will be used to manufacture the
MTB handlebar. A load of 1 kN is applied at a distance of 50 mm from the free
end of the handlebar (with a 28 mm eccentricity). What ply selection should be
used? In this example, consideration should only be given to fibres orientated
along the handlebar length and/or circumferentially around the tube.
Fig. 7.16 Schematic of lateral bending test of MTB handlebar (top view) where a downward
force F is being applied
Solution 7.4
Assumptions. Herein, the UD prepreg used in Chap. 5 is considered and hence,
the mechanical properties in Ref. [32] are used for the design estimates—see
Table 5.2. To account for design and manufacturing inaccuracies and other
uncertainties, a factor of safety (FoS) of 2.0 is selected as a safety net.
The mechanical properties are:
E cl = 131 GPa
σcl∗ = 2575 MPa
∗
σcl, comp = 1235 MPa
G = 3.9 GPa
τc∗ = 85.7 MPa
Ply thickness, t ≈ 0.21 mm (...based on measurements in Table 5.1)
Structural Analysis. The 1 kN load causes bending, transverse shear and

torsion of the MTB handlebar. The bending moment, shear force and torsion
moment diagrams are shown in Fig. 7.17. In the same figure, the normal stress
(from bending) and shear stresses (from transverse shear force and torsional)
are also illustrated.
Normal (Bending) Stresses. The bending moment, M, increases linearly from
the load application point to the clamp (constrained face). Along the length of
the handlebar, the cross-sectional properties change. The diameter at the free
end is 22.2 mm, whilst the diameter is 31.8 mm at the stem clamp. Thus, as
a first approximation, it is conservative to assume that the maximum bending
moment (at the stem clamp) acts on the minimum tube diameter (i.e. 22.2 mm).
Fig. 7.17 Simplified model of the handlebar and the respective shear force (Q), bending
moment (M) and torsion (T ) diagrams
Assuming a unidirectional composite:
My
σ =
I
σcl∗ My My
∴ > = π(do4 −di4 )
FoS I
64

64M y × FoS ∗
di < 4
do4 − ∗ (assuming σcl,comp as the lower value)
π σcl,comp

64 × 290 × 11.1 × 10−3 × 2.0
< (22.2 × 10−3 )4 − = 0.01923 m
4
π × 1235 × 106
0.0222 − 0.01923
t= = 0.00149 m = 1.49 mm
2
1.49
∴ No. of UD plies = = 7.1 plies (assuming ≈0.21 mm/ply)
0.21
Thus, eight UD plies (or the equivalent) are needed to resist the bending
moment.
Transverse Shear Stresses. The shear force, Q, is constant (1 kN) along the
length of the handlebar so we can assume this load acts on the smallest section
(22.2 mm). Note that τc∗ is assumed to be 85.7 MPa.
For a unidirectional composite
Q
τavg =
A
2Q
τmax ≈ (assumingthetubeisthin − walled)
A
τ∗ 2Q 2Q
∴ c > = π(d 2 −d 2 )
FoS A o i
4

8Q × FoS
di < do2 − = 0.02082 m = 20.82 mm
π τc∗
0.0222 − 0.02082
t= = 0.00069 m = 0.69 mm
2
0.69
∴ No. ofUDplies = ≈ 3.3plies
0.21
Note. To simplify the calculations, the MTB handlebars are considered to be
analogous to a thin-walled structure. This assumption is usually only applied
when the inner tube radius is at least 10 times larger than the thickness (r/t > 10
[33]).
Torsional Stresses. The torsional moment rises to a maximum (28 N·m) at the
start of the tapered section and then remains at this maximum up to the stem
clamp. Again, we assume the maximum load acts on the minimum section.
Thus, for a unidirectional composite
Tr
τ=
J
τc∗ Tr Tr
∴ > = π(do4 −di4 )
FoS J
32

32T r × FoS
di < 4
do4 −
π τc∗

32 × 28 × 11.1 × 10−3 × 2.0
di < (22.2 × 10−3 )4 −
4
π × 85.7 × 106
di < 0.02028 m = 20.28 mm
0.0222 − 0.02028
t= = 0.00096 m = 0.96 mm
2
0.96
∴ No. ofUDplies = ≈ 4.6plies
0.21
Now, considering the Principle of Superposition for both sources of shear
stresses, it follows that
No. of UD plies = 3.3 (transverse shear) + 4.6 (torsion) = 7.9
Thus, eight UD plies (or the equivalent) are needed to resist the shear stresses.
Normal (Hoop) Stresses. In addition to the normal (bending) and shear
stresses mentioned above, consideration must be given to the stress concentra-
tions that arise at the stem clamp and directly under the applied 1 kN load. Both
of these are likely to cause normal stresses in the hoop (transverse) direction.
These stresses could cause localised crushing issues. To minimise these local
effects, we must introduce extra plies with fibres orientated circumferentially
around the tube. Here, we assume that the equivalent of two UD plies will be
sufficient.
So, summarising the ply selection... To resist the bending moment, we
require at least eight UD plies, whilst shear (due to torsion and transverse
shear) also requires the equivalent of eight UD plies. To account for localised
crushing, we assume the equivalent of two circumferentially orientated UD
plies is needed.
Failure analysis. Laminate failure is assumed to occur when a stress, either
parallel or perpendicular to the fibre orientations (namely, σcl , σct or τc ),
exceeds the critical (strength) value; this is the basis of the maximum stress
criterion [24]. The normal and shear stresses are treated independently from
one another, viz.:
σcl < σcl∗ .

σct < σct∗ .
τc < τc∗ .
Here, to satisfy the maximum stress criteria, we select:

• six UD plies with the fibres oriented along the handlebar length.
• four woven plies with warp fibres along the length and weft circumferential.
In making this layup selection, some judgement has been made with regard
to the equivalence of UD and woven plies in the context of the handlebar
stresses. A woven ply is assumed to offer about half the bending performance
of a UD ply along the length of the handlebar, as well as half the UD stiffness
and strength in the hoop orientation (neglecting the reduction in Vf for woven
prepreg). Moreover, the woven plies are assumed to provide the equivalent
shear performance of a UD layer [34]—see Sect. 2.7 for further information.
Summary. Based on our design estimates, the handlebar layup (i.e. 6 UD and
4 woven prepreg plies) should satisfy the requisite lateral bending load case,
offering the equivalent of
• eight UD plies to resist the bending moment, meeting the requisite eight
plies: 6 UD +(4 woven × 0.5) = 8
• Ten UD plies to resist the shear stresses, exceeding the requisite eight plies:
6 UD + (4 woven × 1) = 10
• Two UD plies to resist localised crushing: (4 woven × 0.5) = 2
Note. This is only an initial prototyping exercise and therefore many approxi-
mations (and judgements) have been made to simplify the above calculations.
Manufacturing Process. The composite handlebar is manufactured using the blad-

der moulding method described in Sect. 7.3. The mould used is a two-part split
aluminium mould (6061) with a similar bladder inflation seal, blanking plate and
end cap arrangement—see Fig. 7.18. The split mould is manufactured using special-
ist CNC machine tools. Prior to prepreg layup, the mould is released with six coats
of high-temperature wax suitable for temperatures up to 120 ◦ C.
Here, the handlebar is considered as three sections: the central (stem) clamping
section; and the two handle sections, including the tapered region connecting the
handles to the stem section—see Fig. 7.19. To minimise the effects of sectioning, the
prepreg is cut oversize; the prepreg is cut to be longer and wider than needed for the
handlebar. An overlap of at least 0.5 inches (≈12.5 mm) [35] is recommended in a
sectioned layup process to maintain the strength of the composite. Here, the over-
Socket head jacking screws Alignment pins Bladder inflation seal blanking plate
Fig. 7.18 Handlebar ‘split’ mould (only half is shown) and ancillaries
Fig. 7.19 Prepreg sectioning of handlebar (no dimensions)
laps are cut slightly larger than these minimum requirements, adding weight to the
handlebar but offering confidence in the sectioning method used during fabrication.
The plies for each of the prepreg sections are initially consolidated on a flat glass
plate with a roller (see Fig. 7.20a) and then preformed into a tube—see Fig. 7.20b.
The inflation bladder is threaded through each of the tube preforms. The preforms
are positioned in one half of the handlebar mould with the wider end of the handle
sections inserted into the stem section as shown in Fig. 7.20c. The mould is closed
and the locating pins are inserted to align the mould halves. The inflation assembly
and blanking plate are then installed as shown in Fig. 7.20d. Finally, the two mould
halves are clamped shut.
The inflation bladder is gradually pressurised to consolidate the prepreg against
the mould cavity. The bladder is checked to ensure no leaks are present before
commencing the curing process.
The mould is placed in an oven, and the prepreg is cured and post-cured in line
with the supplier’s instructions. The inflation pressure is continuously monitored to
ensure no leaks occur during the curing process.
During the demoulding process, the silicone bladder is deflated and the mould
halves are separated. If the split mould is difficult to part, the cap head bolts can be
inserted into the threaded holes on one of the mould halves and tightened against the
recess of the other mould half to jack the mould apart—as previously discussed; see
Fig. 7.9a. The composite handlebar is then easily removed from the mould as shown
in Fig. 7.21.
The handlebar is cut to length and weighed before physical testing commences;
no surface finish (paint or clear coat) is applied in this instance. The mass of the
handlebar (without optimisation) is measured to be approximately 230 g. This is
comparable to other carbon fibre handlebars currently on the market [36, 37], and
is notably lighter than aluminium handlebars which often weigh in excess of 300 g
[38, 39].
Physical Testing. The handlebar is clamped in a stem extension on a lateral bending
test rig. The test rig is bolted to an Instron UTS as shown in Fig. 7.22a. The handlebar
is aligned perpendicular to the stem axis as described in the standard [25], and
a monotonic load is applied (up to 1000 N) at a distance of 50 mm from the free
end. Load-deflection measurements are recorded at a displacement rate of 5 mm/min.
(a) (b)
(c)
(d)
Fig. 7.20 Handlebar layup process: a initial ply consolidation; b a tube preform; c preform layup;
and d mould assembled
Fig. 7.21 Demoulded handlebar

Fig. 7.22 Lateral bending test: a testing setup; b Load-deflection characteristics of composite
handlebar
The experimental data is shown in Fig. 7.22b. The maximum load is maintained
for at least 1 min without the handlebar cracking or fracturing. This is a successful
outcome for a first structural prototype, but a lighter handlebar is possible if ±45◦
fibres are used.
7.6 Summary
Hollow sections minimise the volume of material, and hence can be used to make
lightweight and economical structures.
Two methods for creating hollow FRC tubes are:
• Mandrel lamination.
• Bladder moulding.
Mandrel lamination usually has lower tooling (mandrel) costs, but it tends to pro-
duce a hollow form with a less than ideal visual appearance on the external surface.
In contrast, the bladder moulding process produces tight (external) tolerances and a
near-perfect visual appearance, but usually at the expense of higher mould tool costs.
7.7 Questions
Questions 7.1 Define mandrel lamination and bladder moulding.
Questions 7.2 Describe three methods that could be used to assist mandrel removal
in open-ended parts.
Questions 7.3 Why is mandrel lamination more complicated for closed-ended struc-
tures than for their open-ended counterparts?
Questions 7.4 Describe two methods that could be used to remove a mandrel from
a closed-ended part.
Questions 7.5 What is shrink tube and how does it work? How does it differ from
shrink tape?
Questions 7.6 In bladder moulding, why are silicone and latex excellent choices
for a bladder?
Questions 7.7 In a split mould, what is the purpose of the alignment pins? Why is
it still important to clamp the mould halves shut?
7.8 Problems
Problem 7.1 A unidirectional glass fibre-reinforced square hollow section (SHS) is

40 mm × 40 mm × 1.2 mm. The glass fibres have an elastic modulus of 69 GPa and
a tensile strength of 2500 MPa; the polyester matrix has a modulus of 2.5 GPa and a
strength of 50 MPa. If the SHS is subjected to a tensile load of 10 kN, calculate
a. The axial stress in the SHS.
b. The Factor of Safety (FoS), assuming a fibre volume fraction of 0.35.
Answer. a. 53.7 MPa; and b. 16.3.
Problem 7.2 A unidirectional hollow circular tube is clamped at one end (i.e. a
cantilever). If the outer diameter of the tube is 30 mm and the wall thickness is 1 mm,
calculate the stress in the thin-walled tube when a torque of 10 N·m is applied at the
free end. If the in-plane shear strength of the composite is 40 MPa, will the composite
fail?
π
4
Note. The polar second moment of area for a tube is J = 32 do − di4 , where do is
the outer diameter and di is the inner diameter of the tube.
Answer. 7.8 MPa. No (τc < τc∗ ).
Problem 7.3 A tubular shaft is to be designed with an outside diameter of 25 mm and

a length of 0.5 m. The tube is manufactured from a unidirectional fibre-epoxy prepreg.
The tensile modulus of the carbon-epoxy prepreg is 110 GPa, and the tensile strength
and in-plane shear strength are 1800 MPa and 60 MPa, respectively. In service, the
tubular shaft is subjected to a three-point bending test with a midspan load of 1kN.
Determine the minimum number of 0.21 mm plies that are needed, to ensure
7.8 Problems 133
a. Longitudinal (tensile or compressive failure) does not occur.

b. Shear failure of the thin-walled tube does not occur.
c. The tubular shaft does not exhibit a midspan deflection of more than 10 mm.
d 4 −d 4
Note. The second moment of area for a shaft is I = π o64 i , where di and do are,
respectively, the inner and outer diameters; maximum deflection at the midspan of
Fl 3
the three-point bending test is δmax = 48E I
, where l is the length of the beam, and F
is the point load.
Answer. a. 4 plies; b. 2 plies; and c. 7 plies.
References

2. Hibbeler RC (2014) Statics and mechanics of materials, 4th edn. Pearson, Upper Saddle River
6. Hollaway L (1994) Handbook of polymer composites for engineers. Woodhead Publishing Ltd,
Cambridge
Technology, Oxford
10. Aird F (2014) Fiberglass and other composite materialshp1498: a guide to high performance
non-metallic materials for automotiveracing and marine use includes fiberglass, kevlar, carbon
fiber, molds, structures and materials. HP Books, New York
11. Barbero EJ (2017) Introduction to composite materials design, composite materials, 3rd edn.
12. Silcock MD, Garschke C, Hall W, Fox BL (2007) Rapid composite tube manufacture uti-
lizing the quicksteptm process. J Compos Mater 41(8):965–978. https://doi.org/10.1177/
0021998306067261
13. Leslie JC (1970) Precision molding of advanced fiber structures. In: Volume 1B: general.
American Society of Mechanical Engineers. https://doi.org/10.1115/70-GT-126
14. Francucci G, Palmer S, Hall W (2018) External compaction pressure over vacuum-bagged
composite parts: effect on the quality of flax fiber/epoxy laminates. J Compos Mater 52(1):3–
15. https://doi.org/10.1177/0021998317701998
15. Mazumdar SK (2002) Composites manufacturing: materials, product, and process engineer-
ing/Sanjay K. CRC Press, Boca Raton
16. Shenoi RA, Wellicome JF (1993) Composite materials in maritime structures, Cambridge ocean
technology series, vols 4,5. CUP, Cambridge
17. Fiore V, Valenza A (2013) Epoxy resins as a matrix material in advanced fiber-reinforced poly-
mer (frp) composites. In: Advanced fibre-reinforced polymer (FRP) composites for structural
applications, Elsevier, pp 88–121. https://doi.org/10.1533/9780857098641.1.88
18. Ahmed A, Tavakol B, Das R, Joven R, Roozbehjavan P, Minaie B (2012) Study of thermal
expansion in carbon fiber-reinforced polymer composites. In: SAMPE international symposium
proceedings
19. Summerscales J, Cullen R (02/10/2019) Private correspondence

edn. Wiley, Hoboken
21. Anderson JP, Altan MC (2014) Bladder assisted composite manufacturing (bacm): challenges
and opportunities. In: Polymer processing society Europe-Africa conference. https://doi.org/
10.13140/2.1.2139.6169
22. Dowaksa (2016) 3k a-38: Technical data sheet. https://www.carbononline.nl/wp-content/
uploads/2020/05/Dowasksa-3K-A-38-7.pdf
23. Kinetix (2018) Engineering data: R240 wet-preg. http://atlcomposites.com.au/icart/products/
7/images/main/KINETIX240Pre-PregEngineeringData.pdf
25. British Standards (2014) Cycles - safety requirements for bicycles
26. Emerson N (08/06/2018) Private correspondence
27. Altenbach H, Altenbach JW, Kissing W (2004) Mechanics of composite structural elements.
Foundations of engineering mechanics. Springer, Berlin
28. Javanbakht Z, Aßmus M, Naumenko K, Öchsner A, Altenbach H (2019) On thermal strains and
residual stresses in the linear theory of anti–sandwiches. ZAMM - J Appl Math Mech/Zeitschrift
für Angewandte Mathematik und Mechanik 99(8). https://doi.org/10.1002/ZAMM.201900062
29. Javanbakht Z, Öchsner A (2017) Advanced finite element simulation with MSC Marc: appli-
cation of user subroutines. Springer, Cham
30. Javanbakht Z, Öchsner A (2018) Computational statics revision course. Springer International
Publishing, Cham. https://doi.org/10.1007/978-3-319-67462-9
31. Javanbakht Z, Hall W, Öchsner A (2017) Computational evaluation of transverse thermal con-
ductivity of natural fiber composites. In: Öchsner A, Altenbach H (eds) Improved performance
of materials: design and experimental approaches. Advanced structured materials, vol 72,
Springer, Cham, pp 197–206. https://doi.org/10.1007/978-3-319-59590-0
32. Solvay (2021) Technical data sheet: Vtm(mohana)260 series (prepeg). https://catalogservice.
solvay.com/
33. Govindjee S (2012) Engineering mechanics of deformable solids: a presentation with exercises.
OUP Oxford, Oxford
34. Medina C, Canales C, Arango C, Flores P (2014) The influence of carbon fabric weave on the
in-plane shear mechanical performance of epoxy fiber-reinforced laminates. J Compos Mater
48(23):2871–2878. https://doi.org/10.1177/0021998313503026
35. Jin H, Nelson K, Werner BT, Briggs T (2018) Mechanical strength of composites with different
overlap lengths: (No. SAND2018-10594). Sandia National Lab (SNL-CA), Livermore, CA
(United States). https://doi.org/10.2172/1488647
36. Hope Technology (2020) Carbon handlebar 31.8mm. https://www.hopetech.com/products/
controls/handlebars/carbon-handlebar-318mm/
37. Renthal (2020) Fatbar carbon 10mm rise. https://www.renthal.com/cycle/handlebars/fatbar-
carbon/fbc-10mm
38. Deity Components (2020) Deity blacklabel 800 handlebar: 15mm rise. https://www.
deitycomponents.com/store/p8/DEITY_BLACKLABEL_800_HANDLEBAR_15mm_
RISE.html
39. Renthal (2020) Fatbar 10mm rise. https://www.renthal.com/cycle/fatbar/fatbar-10
Index
A unidirectional, 14, 16, 21, 25, 35

Additive manufacturing, 86 Consolidation pressure, 35
Air-assisted demoulding, 109 Coupling stress, 6
Anisotropy, 13, 14 Crosslinking, 42, 43
Axial, see Longitudinal Curing, 42, 107, 114
B D
Bladder, 112 Debulking, 56
Bladder moulding, 105, 112, 128 Demoulding, 42, 46, 108, 109, 112, 129
Bleeder, 56 Dissolvable mandrel, see Mouldless con-
Bleeder/breather fabric, 59, 63 struction
Breach unit, 59 Draft angle, 82, 85
Breakdown method, 98, 106 Drapeabilty, 6
Breather, 56 Dry fibre weight, 26
Bridging, 84 Dwell time, 63
C E
Catalyst, 41, 44 Epoxy, 2, 35, 42, 45, 56
Caul plate, 37 Exothermic reaction, 42
Classic laminate theory, 123
Coefficient of Thermal Expansion (CTE),
109 F
Composite Fabric, 5
consolidation, 56 braided, 7
design, 13 plain, 6
failure mode, 21 random mat, 5, 7, 13
machining, 47 satin, 6
shear failure, 21 twill, 6
shear modulus, 18, 20 woven, 6, 25
shear strength, 23 Factor of safety, 77
strength, 21, 23 Failure strain, 15
tooling, 89 Fibre
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer 135
Nature Switzerland AG 2021
https://doi.org/10.1007/978-3-030-78807-0
136 Index
aramid, 5 Modelling clay, 92

bundle, 5 Mould, 37, 82
carbon, 4, 35, 56 female, 37, 81, 85
glass, 4 matched-die, 37
microbuckling, 22 Mouldability, 81
orientation, 7, 20, 35, 55 Mouldless construction, 106
random, 7, 25, 97
unidirectional, 5
volume fraction, 16, 35, 72 O
Finite element analysis, 123 Orientation efficiency, 25
Free radical, 44 Out-time, 62
Freezer-life, 62
P
G Packing arrangement, 28
Gelcoat, 9, 39, 48 Parting agent, see Release agent
Green stage, 42 Pattern, see Plug
Pattern removal, 89
Pleat, 61
H Plug, 85
Hand lamination, see Wet layup flexibility, 88, 91
Hardener, 41, 46 Poisson’s ratio, 20
Hard tooling, 89 Polyester, 2, 43
Heat shrink tube, 107 Polymer
Hollow section, 105 thermoplastic, 2, 86
thermoset, 2
Polymerisation, 43
I Polyvinyl Alcohol (PVA), 38
Initiator, 44 Post-curing, 42
Interface failure, 23 Prepreg, 62, 105, 107, 112
Inverse rule of mixture, 17, 76 Prepreg layup, 34
Prepreg moulding, see Prepreg layup
K
Kelly-Tyson model, 21, 22, 76 Q
Quasi-isotropy, 13
L
Laminate, 7, 13, 16 R
Layup moulding, see Wet layup Reactive site, 45
Leak rate, 60 Release agent, 38
Longitudinal, 14 Resin wastage, 41
axis, 7 Resin weight fraction, 26
elastic modulus, 14, 20 Reuss model, see Inverse rule of mixture
failure, 21, 22 Risk, 33
load, 24 Rule of mixture, 15, 76
modulus, 72
strength, 14, 22, 72
S
Sealant, 56, 61
M Seasoned mould, 38
Mandrel, 106, 108 Shrink tape, 108
Mandrel lamination, 105 Styrene, 43
Master, see Plug Surface finish, 37, 48, 85, 88, 89, 111
Index 137
Surface texture, see Surface finish Unsaturated polymer, 43

Swiss roll layup, 107
T V
Tensile test, 72 Vacuum bagging, 34, 56, 61
Tooling, 37 Vacuum connector, see Breach unit
Transverse, 14 Vacuum valve, see Breach unit
elastic modulus, 17, 18, 20 Vinylester, 2, 45
failure, 21, 23 Voigt model, see Rule of mixture
load, 24
strength, 72
Tresca critertion, 23
W
Wet layup, 34, 35, 55
U Wet out issue, 6
Unidirectional laminates, 72 Wrapping, see Mandrel lamination

Hall 2021

Uploaded by

Copyright:

Available Formats

Hall 2021

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hall 2021

Uploaded by

Copyright:

Available Formats

Advanced Structured Materials

Advanced Structured Materials is indexed in Google Scholar and Scopus.

More information about this series at http://www.springer.com/series/8611

Design and Manufacture

ISSN 1869-8433 ISSN 1869-8441 (electronic)

To my parents and my brother

Gold Coast, Australia Wayne Hall

Latin Symbols (Capital Letters)

Latin Symbols (Small Letters)

Greek Symbols (Capital Letters)

Denotes the change in a variable.

Greek Symbols (Small Letters)

α Coefficient of thermal expansion

. . .∗ Strength (or failure)

e.g. for example (from Latin ‘exempli gratia’)

Abstract The concept of fibre-reinforced polymer composites is introduced in this

A composite is a material comprising two or more constituent materials. The com-

The polymer matrix of a FRC can be classified as thermoplastic or thermoset [3].

1.3 Fibre Types

1.4 Fibre Forms

Fig. 1.1 Unidirectional

Three common woven fabrics are [4]:

Fig. 1.2 Warp and weft of a plain weave fabric

Fig. 1.4 Braided ±45◦

Fig. 1.5 Random fabrics: a chopped strand mat; and b veil

The most common thermosetting resins are:

Table 1.3 Some common fibre properties

Tensile stress (MPa) 4000

Fig. 1.6 Tensile stress-strain behaviour of a HS carbon-epoxy composite

a. Identify the stress-strain curve for each fibre type.

Problem 1.2 From the stress-strain data in Problem 1.1, identify

Abstract This chapter addresses the mechanical properties of a fibre-reinforced

The design of a FRC requires:

2.2 Axial (Longitudinal) Modulus

A fibre-reinforced composite subjected to an axial (tensile or compressive) load will

Fibre orientation Fibre orientation

4,000 HS carbon fibre

Fig. 2.2 Tensile stress-strain behaviour of a HS carbon-epoxy composite

The longitudinal modulus of the composite, E cl , is calculated by the rule of mixtures

Fig. 2.3 Composite member F

2.3 Transverse Modulus

The transverse modulus is calculated from

2.4 Shear Modulus

Fig. 2.4 A unidirectional

The minimum volume fraction to satisfy the shear stiffness requirement is

2.5 Effect of Fibre Orientation

Fibre orientation Fibre orientation

Elastic or shear modulus (GPa)

2.6 Strength of Composites

(a) (b) (c)

σcl∗ = σm(f∗ ) (1 − Vf ) + σf∗ Vf (2.5)

σcl∗ = σm(f∗ ) (1 − Vf ) + σf∗ Vf

If the matrix stress at fibre failure was unknown, we could conservatively

τc∗ ≈ τm∗ = 0.5 × σm∗ (2.8)

Failure Under Off-Axis Loading. Evaluating the strength of a UD composite sub-

2.7 Woven and Random Fabrics

and the warp/weft strengths are estimated from

2.8 Volume Fraction

The calculation of elastic properties for a unidirectional fibre composite requires