investigation of deformation behaviour of dry textiles
TRANSCRIPT
Investigation of deformation behaviour of
dry textiles under forming forces by
computed tomography
A thesis submitted to The University of Manchester for the Degree of
Doctor of Philosophy
In the Faculty of Engineering and Physical Sciences
2014
Zeshan Yousaf
School of Materials
Department of Textile and Paper
The University of Manchester
Contents
2
Contents
List of Tables 6
List of Figures 7
Abstract 11
Declaration 12
Copyright Statement 13
Dedication 15
Acknowledgment 16
Chapter 1. Introduction 17
1.1. Introduction 17
1.2. Problem definition 17
1.3. Research aim 18
1.4. Research objectives 18
1.5. Thesis lay out 19
Chapter 2. Literature Review 22
2.1. Textile fabrics 22
2.1.1. Woven fabrics 22
2.1.1.1. 2D woven fabrics 22
2.1.1.1.1. Plain fabrics 23
2.1.1.1.2. Twill fabrics 23
2.1.1.1.3. Satin and sateen fabrics 24
2.1.1.2. 3D woven fabrics 25
2.1.1.2.1. Orthogonal fabrics 25
2.1.1.2.2. Angle-interlock fabrics 26
2.1.2. Non-crimp stitched fabrics 27
2.1.3. Braided fabrics 28
2.2. Composite manufacturing processes 28
2.2.1. Resin transfer moulding 28
2.2.2. Vacuum bagging 29
2.2.3. Autoclave processing 30
2.3. Compaction of textile preforms 31
Contents
3
Chapter 3. Deformation of single layer dry preforms under
compaction
64
3.1. Introduction 65
3.2. Material and mechanical testing 68
3.2.1. Mechanical testing 69
3.3. Tomography and in-situ compression rig 69
3.4. Results and discussion 72
3.4.1. Macroscopic deformation 72
3.4.2. Meso-structural analysis by computed tomography 77
3.4.3. Resin channels in dry preform 87
3.4.4. Yarn packing fraction 93
3.4.5. Yarn and fibre volume fraction of the preform 95
3.5. Conclusion 98
Chapter 4. Deformation of multilayer dry preforms under
compaction
105
4.1. Introduction 106
4.2. Nesting factor and tow geometry parameters 109
4.3. Material and mechanical testing 110
4.3.1. Material 110
4.3.2. Mechanical testing 111
4.4. Tomography and in-situ compression set up 112
4.5. Results and discussion 113
4.5.1. Macroscopic deformation 113
4.5.2. Meso-structural by computed tomography (CT) 114
4.5.2.1. Nesting factor analysis 122
4.5.2.2. Quantification of inter-tow voids using image analysis 123
4.5.2.3. Yarn packing fraction 126
4.6. Conclusions 127
Chapter 5. Compaction and nesting in textile preforms influenced
by tow architecture
134
5.1. Introduction 135
5.2. Material & Experimental details 136
Contents
4
5.2.1. Material 136
5.2.2. Experimental methodology 138
5.2.2.1. Mechanical testing 138
5.2.2.2. Computed tomography (CT) analysis 139
5.3. Analysis of macro-scale deformations 141
5.3.1. Woven fabrics 141
5.3.2. Non-crimp fabrics 146
5.3.3. Nesting of layers 148
5.4. Analysis of meso-scale deformations 151
5.4.1. Tow waviness 152
5.4.2. Inter-tow voids 155
5.5. Conclusions 158
Chapter 6. Deformation of dry and wet preforms under compaction 164
6.1. Introduction 165
6.2. Material and mechanical testing 166
6.2.1. Material 166
6.2.2. Mechanical testing 166
6.3. Results and discussion 167
6.3.1. Macroscopic deformation 167
6.3.2. Tomography and in-situ compression set up 173
6.3.3. Meso-structural analysis by computed tomography (CT) 175
6.3.3.1. Tow waviness 177
6.3.3.2. Inter-tow voids 179
6.4. Conclusions 181
Chapter 7. Development of an in-situ technique to analyse the
meso-structure of dry fabrics under biaxial loadings
184
7.1. Introduction 185
7.2. Development of biaxial testing machine 187
7.2.1. Load cell calibration 189
7.3. Tomography and in-situ loading rig 190
7.4. Conclusions 192
Chapter 8. Conclusions and future work 197
8.1. Conclusions 197
Contents
5
8.2. Future work 200
Publications 202
List of tables
6
List of Tables
Table 3.1. Tow geometry parameters (warp) 79
Table 3.2. Tow geometry parameters (weft) 79
Table 3.3. Tow area calculated by CT and Texgen 87
Table 3.4. Yarn packing fractions of warp cross-sections 95
Table 3.5. Yarn packing fractions of weft cross-sections 95
Table 4.1. Parameters used to describe the fabric geometry 115
Table 4.2. Tow geometry parameters measured by X-ray tomography 118
Table 4.3. Nesting measurements for the 6 layer preform 120
Table 4.4. Yarn packing fraction 127
Table 5.1. Woven fabric specifications 137
Table 5.2. Stitched non-crimp fabric (NCF) specifications 137
Table 6.1. Material used for mechanical testing 166
Table 6.2. Material used for meso-structure analysis 166
List of figures
7
List of Figures
Figure 2.1. Plain weave 23
Figure 2.2. Twill weave 24
Figure 2.3. Satin weave 24
Figure 2.4. Orthogonal weave 26
Figure 2.5. Angle-interlock 27
Figure 2.6. NCF fabrics 27
Figure 2.7. Braided fabric 28
Figure 2.8. Resin transfer moulding 29
Figure 2.9. Vacuum bagging 30
Figure 2.10. Autoclave processing 31
Figure 2.11. Typical pressure thickness cure 33
Figure 2.12. Stages of fabric compression 34
Figure 2.13. Schematic representation of time dependent behaviour of fibrous
materials
36
Figure 2.14. Main factors affecting the compaction behaviour of preforms 47
Figure 3.1. Pressure thickness curve for a woven fabric under compaction 66
Figure 3.2. Scanned image of plain weave (1/1) fabric 68
Figure 3.3. Schematic of the compression rig 70
Figure 3.4. (a) Compression rig accommodated on the tomography stage, and
(b) close-up of the rig
71
Figure 3.5. Pressure-thickness response for a single layer under dry and wet
conditions
73
Figure 3.6. Effect of loading time on the thickness of the single layer dry fabric 73
Figure 3.7. t/t0 plotted against time for different pressures 74
Figure 3.8. The constants plotted against pressure 75
Figure 3.9. Experimental and predicted thicknesses with time at different
pressures, (a) dry fabric and (b) wet fabric
76
Figure 3.10. Definition of the yarn geometry parameters 77
Figure 3.11. Segmented virtual slices of warp and weft yarns 78
Figure 3.12. 3D reconstruction of single layer fabric 78
Figure 3.13. Cross-sectional view of the preform at the centre of the tow 81
Figure 3.14. 3D reconstructions of yarns, (a) warp yarn, (b) weft yarn 81
List of figures
8
Figure 3.15. The crimp amplitude of the warp and the weft yarns 82
Figure 3.16. The crimp percentage of warp and weft yarns 82
Figure 3.17. The crimp angle of warp and weft yarns 83
Figure 3.18. Cross-sectional images of the warp and the weft yarns in between
two tows
85
Figure 3.19. Tow cross-section at the centre due to crimp angle 86
Figure 3.20. Inter-tow voids in 3D structure 89
Figure 3.21. Inter-tow voids at the centre of the tow intersections 89
Figure 3.22. Inter-tow voids in between two tows 90
Figure 3.23. Average void thickness with pressure 91
Figure 3.24. The frequency distribution of the void thickness 91
Figure 3.25. Resin channels in dry preform during compression 93
Figure 3.26. Yarn packing fractions of dry preform against different loads 94
Figure 3.27. Yarn packing fraction on application of pressure 94
Figure 3.28. Yarn volume fractions calculated at the centre of the tow
intersections
96
Figure 3.29. Yarn volume fractions calculated in between two tows 96
Figure 3.30. FVF calculated by mechanical testing and tomographic analysis 98
Figure 4.1. Typical pressure thickness curve for a woven fabric under
compaction
107
Figure 4.2. Layer thickness (a) Single layer, (b) 2 layer without nesting, (c) 2
layer with shifting and nesting
110
Figure 4.3. Compression Rig (a) Compression rig fixed on the tomography
stage, (b) close-up of the rig
113
Figure 4.4. Pressure thickness curves of different layers 114
Figure 4.5. Yarn geometry parameters 115
Figure 4.6. 3D view of the six layer dry preform 116
Figure 4.7. Tomographic sections through the stack of six layers at different
pressures
119
Figure 4.8. Yarn isolated from the tomograph showing crimp in crossing yarns 120
Figure 4.9. Tomographic weft cross-sections showing layers with and without
shift
121
Figure 4.10. Nesting factors from mechanical testing and tomographic analysis 123
List of figures
9
Figure 4.11. Flow channels at (a) 4 kPa, (b) 45 kPa, and (c) 100 kPa 124
Figure 4.12. 3D representations of inter-tow voids at (a) 4kPa, (b) 45kPa, and
(c) 100kPa
125
Figure 4.13. Inter-tow voids at different slices and average inter-tow voids 126
Figure 5.1. Woven fabrics (a) Plain, (b) 3/1 twill, (c) 5H sateen 137
Figure 5.2. Scanned images of woven fabrics and non-crimp stitched fabrics 138
Figure 5.3. Compression set up 139
Figure 5.4. Compression rig used for set up 140
Figure 5.5. Single layer thickness results against pressure 142
Figure 5.6. Compression behaviour of multi-layer fabrics, (a) plain, (b) twill,
(c) sateen
144
Figure 5.7. Different layers thickness results against pressure, (a) two layers (b)
six layers
145
Figure 5.8. Non-crimp fabric thickness results as a function of pressure, (a)
+45о/-45
о, (b) 0
о/90
о
147
Figure 5.9. Nesting factors (a) woven fabrics (b) non-crimp stitched fabrics 150
Figure 5.10. 3-D reconstruction of single layer (right) and 10 layers (left) 152
Figure 5.11. CT images of single layer 152
Figure 5.12. Tow showing crimp 153
Figure 5.13. Yarn crimp percentage (a) warp yarn, (b) weft yarn 153
Figure 5.14. CT images of ten layer stack 155
Figure 5.15. Inter-tow voids in single layer and multilayer stack 157
Figure 5.16. Inter-tow voids in single layer along slices (a) 100 kPa, (b) 300
kPa
157
Figure 5.17. Inter-tow voids in multilayer stack along slices (a) 100 kPa, (b)
300 kPa
157
Figure 5.18. Plain-weave repeat presenting, between two tows and centre of the
tow
158
Figure 6.1. Thickness vs pressure of single layer and two layers of P1 fabric 168
Figure 6.2. Thickness vs pressure of single layer and two layers of T1 fabric 168
Figure 6.3. Thickness vs pressure of single layer and two layers of S1 fabric 169
Figure 6.4. Thickness vs pressure of single layer and two layers of NCF 169
Figure 6.5. Thickness vs pressure of single layer and two layers of plain fabric 169
List of figures
10
P2
Figure 6.6. Td-Tw/T0 plotted against pressure for plain (P2) woven glass fabric 171
Figure 6.7. Experimental and predicted wet thickness of plain fabric (P1) 171
Figure 6.8. Experimental and predicted wet thickness of twill fabric (T1) 172
Figure 6.9. Experimental and predicted wet thickness of sateen fabric (S1) 172
Figure 6.10. Experimental and predicted wet thickness of NCF 172
Figure 6.11. Experimental and predicted wet thickness of plain fabric (P2) 173
Figure 6.12. Compression Rig (a) Compression rig fixed on the tomography
stage, (b) close-up of the rig
174
Figure 6.13. CT images showing slice of (a) warp cross-section, (b) segmented
inter-tow voids and (c) segmented yarns
175
Figure 6.14. Reconstructed image of single layer fabric compacted in
polycarbonate plate
175
Figure 6.15. Yarn geometry parameters 176
Figure 6.16. Cross- sectional images of single layer dry and wet fabrics 177
Figure 6.17. Tow waviness of dry and wet fabrics (a) warp yarns, b) weft yarns 178
Figure 6.18. Inter-tow voids in dry and wet fabric 180
Figure 6.19. Inter-tow voids in (a) dry fabric and (b) wet fabric at 10 kPa 181
Figure 7.1. Rig for biaxial shear and tensile testing 188
Figure 7.2. Biaxial rig with cruciform specimen 189
Figure 7.3. Rig to grip the biaxially loaded fabric 190
Figure 7.4. CT images of the biaxially loaded dry fabric, a) side view, b) top
view and crossing yarns in two slices
192
Abstract
11
Abstract
Textile preforms undergo different deformations during preforming including transverse
compression, biaxial tensile, in-plane shear and out-of-plane bending deformation. All
these deformations change the tow and resin channel geometry of the preform and
consequently affect the mechanical properties of the final product. To make simulation
tools and to execute structural analysis, accurate experimental data during these
deformations is required.
In the present work, an in-situ measurement technique has been developed to study the
deformation behaviour of textile preforms at meso-scale during the compression mode.
This technique for the first time, enables measurement of the meso-structure of dry and
wet fabrics under in-situ loadings. Tow geometry changes of single layer and multilayer
preforms have been captured under in-situ loadings using computed tomography (CT).
Resin channels in single layer and multilayer preforms have been investigated in detail.
In multilayer fabrics, it has been observed that nesting of the layers strongly influences
the inter-tow voids. It has been observed that crimp reduction in single layer fabric and
nesting of layers in multilayer fabric is responsible for the fabric thickness reduction in
a low load regime (~100 kPa).
During study of different fabric architectures, it has been observed that the float length
of the weave plays an important role in compression behaviour.
The deformation of wet fabrics has also been studied at macro and meso-levels using
mechanical test results and CT. It has also been observed that at low loads (~300 kPa)
there is higher crimp reduction in wet fabrics compared to dry fabrics.
Additionally, an in-situ measurement technique has also been developed to study the
meso-structure of dry fabrics under biaxial tensile and shear loading using CT.
Declaration
12
Declaration
No portion of the work referred to in the thesis has been submitted in support of an
application for another degree or qualification of this or any other university or other
institute of learning.
Copyright statement
13
Copyright Statement
I. The author of this thesis (including any appendices and/or schedules to this
thesis) owns certain copyright or related rights in it (the “Copyright”) and he has
given The University of Manchester certain rights to use such Copyright,
including for administrative purposes.
II. Copies of this thesis, either in full or in extracts and whether in hard or
electronic copy, may be made only in accordance with the Copyright, Designs
and Patents Act 1988 (as amended) and regulations issued under it or, where
appropriate, in accordance with licensing agreements which the University has
from time to time. This page must form part of any such copies made.
III. The ownership of certain Copyright, patents, designs, trademarks and other
intellectual property (the “Intellectual Property”) and any reproductions of
copyright works in the thesis, for example graphs and tables (“Reproductions”),
which may be described in this thesis, may not be owned by the author and may
be owned by third parties. Such Intellectual Property and Reproductions cannot
and must not be made available for use without the prior written permission of
the owner(s) of the relevant Intellectual Property and/or Reproductions.
IV. Further information on the conditions under which disclosure, publication and
commercialisation of this thesis, the Copyright and any Intellectual Property
and/or Reproductions described in it may take place is available in the
University IP Policy (see
Copyright statement
14
http://documents.manchester.ac.uk/DocuInfo.aspx?DocID=487), in any relevant
Thesis restriction declarations deposited in the University Library, The
University Library’s regulations (see
http://www.manchester.ac.uk/library/aboutus/regulations) and in The
University’s policy on Presentation of Theses.
Dedication
15
Dedication
Dedicated to the soul of my beloved mother
Acknowledgements
16
Acknowledgements
I would sincerely like to acknowledge my supervisor Prof. Prasad Potluri for his
guidance, help and encouragement for my research work, especially his motivation
towards my goals which I can never forget throughout my life. In fact, he has been a
source of motivation for me in all aspects.
Special thanks to my co-supervisor Prof. Philip Withers to guide me towards my
research work and providing me an excellent opportunity to perform my experimental
work in X-ray Imaging Facility which enabled me to successfully achieve my research
goals.
I also wish to thank X-ray imaging Facility team, especially Mr Fabien Leonard to help
me during the experimental work. Thanks to Stuart Morse in Material Science Lab and
Thomas Kerr in weaving Lab for their help.
My sincere thanks to Dr. Richard Kannon for his endeavouring support during my
write-up. Special thanks to Sabahat Nawaz, Khayale Jan and Ibraheem for their help.
I would also like to thank all my colleagues and friends for their help and support
during my research work especially Vivek, Haseeb, Alvaro, Dhaval, Mubeen,
M.Peerzada and Rishad.
I would like to thank Bhauuddin Zakria University Multan for funding my research.
Finally, I would like to express my thanks to my family, especially my father and sister
who continuously encouraged me throughout my work.
Chapter 1 Introduction
17
Chapter 1
Introduction
1.1 Introduction
Low-cost composites manufacturing techniques based on resin infusion of dry textile
preforms are becoming popular in the aerospace, automotive and energy sectors. In
these processes, several fabric plies are draped on the tool surface and are subjected to
transverse compaction and biaxial tensile and shear forces either by vacuum (Vacuum
Infusion) or external pressure (Resin Transfer Moulding), and at the same time infused
with liquid resin. These forces change the tow and resin channel geometry, resulting in
change in the mechanical properties of the final product. To make simulation tools and
to perform structural analysis, accurate experimental data for these geometrical changes
during forming is required.
1.2 Problem definition
Deformation of textile preforms under different loading conditions (transverse
compaction, biaxial tensile and shear) during forming plays an important role on tow
and resin channel geometry and on the mechanical properties of the composites.
Chapter 1 Introduction
18
Extensive research work has been performed on the deformation behaviour of textile
preforms during composite manufacturing. The limitation of the existing research work
is that structural analysis at meso-scale during forming is only available on laminated
composites. No research work is available on the deformation of dry and wet textile
preforms prior to curing during forming at meso-scale. Additionally, the effect of weave
structure of textile fabrics has been studied by many researchers but no attempt has been
taken to keep the tow spacing and tow count similar during the weave structure
comparisons. In order to make structural analysis and simulation tools accurate,
knowledge of deformations in dry textile fabrics under different loadings prior to curing
is required.
1.3 Research Aim
The aim of the present research work is to study deformation behaviour of dry and wet
textile fabrics under transverse compaction forces at macro and meso-scale and to
develop a technique to measure the dry fabric behaviour under in-situ biaxial tensile and
shear forces.
1.4 Research Objectives
Mechanical testing of single layer and multilayer dry and wet glass fabric
samples using an Instron testing machine.
Development of a compression rig calibrated by the mechanical test results to
measure the meso-structure of glass fabrics under in-situ compression loading.
Structural analysis of glass fabric samples under in-situ compression loading
using computed tomography (CT).
Chapter 1 Introduction
19
Development of different woven fabric architectures by keeping similar thread
densities and thread counts.
Macro-scale deformation study of developed woven fabrics and non-crimp
fabrics by testing of the fabric samples using an Instron testing machine.
Development of the biaxial testing machine and an in-situ rig to study the meso-
structure of dry textile preforms under biaxial tensile and shear loadings using
CT.
1.5 Thesis layout
Chapter-2: Literature review
This chapter gives an introduction of the basic textile architectures, different composite
manufacturing techniques and detail of work done on the deformation of textile
preforms during forming.
Chapter-3: Deformation of single layer dry preforms under compaction
This chapter presents an introduction to compaction; the methodology to investigate
mechanical testing and the computed tomography (CT) scanning and deformation
behaviour of single layer preforms at meso-level using computed tomography (CT).
Different tow geometry parameters of tow changes, the resin channels in the preform
during compaction, comparison of computed tow area calculated from Texgen software
and CT images and comparison of fibre volume fraction calculated from mechanical test
results and from computed tomography data are examined.
Chapter 1 Introduction
20
Chapter-4: Deformation of multilayer dry preforms under compaction
In this chapter the deformation behaviour of six layer plain woven preforms under
compaction has been presented at meso-scale.
A brief introduction of the deformation behaviour of multilayer preforms and the
nesting of layers has been discussed. The methodology used for the mechanical testing
and computed tomography has been presented.
Macro and meso-level deformation of six layer preforms have been presented and
different tow geometry parameters calculated for six layer preforms have been
discussed at low loads of 100 kPa.
Chapter-5: Compaction and nesting in textile preforms influenced by tow
architectures
In this chapter, the deformation behaviour of different textile preforms has been studied
at macro-level. The comparison of single layer and multilayer preforms of each
architecture under compaction has been presented and nesting factors have been
calculated for all these using mechanical test results. The tow waviness and resin
channels of single and ten layer specimens have been investigated using CT images.
Chapter-6: Deformation of dry and wet preforms under compaction
Here the deformation of dry and wet fabrics has been presented at macro and meso-
levels. The deformation behaviour of different fabrics has been studied using the
mechanical test results at macro level. Power law relations have been derived to predict
Chapter 1 Introduction
21
the wet thickness results from dry fabric thickness. Tow waviness and resin channels
have been investigated and compared for single layer dry and wet preforms.
Chapter-7: Development of an in-situ technique to analyse the meso-
structure of dry fabrics under biaxial loading
In this chapter, a method to measure the deformation of dry preforms under in-situ shear
and tensile loading has been presented. Plain woven fabric has been investigated under
biaxial tensile loading to verify the accuracy of the technique.
Chapter-8: Conclusions and future work
This chapter includes the conclusions and suggestions for future work.
Chapter 2 Literature review
22
Chapter 2
LITERATURE REVIEW
2.1 Textile fabrics
A textile fabric is defined as a manufactured assembly of fibres and/or yarns, which has
a substantial surface area in relation to its thickness and sufficient inherent cohesion to
give mechanical strength to the assembly [1].
Based on the manufacturing techniques, conventional textile fabrics can be divided into
woven, non-woven, knitted and braided formations. A brief introduction of these
structures is presented here.
2.1.1 Woven fabrics
Woven fabrics that are used in composites can be classified as two dimensional (2D)
and three dimensional (3D) structures [2].
2.1.1.1 2D woven fabrics
2D woven fabrics are made by the interlacement of two groups of threads, warp and
weft, which are placed at right angles to each other in the plane of the fabric. The warp
and the weft yarns are oriented along the length and the width of the fabric respectively.
Warp yarns are also known as ends and weft yarns as picks. Depending on the repeat
Chapter 2 Literature review
23
pattern of the interlacement, 2D woven fabrics can be further classified; some examples
are plain, twill and satin fabrics [3].
2.1.1.1.1 Plain fabrics
The plain weave is the simplest of all the available weaving patterns. This weave has the
highest possible interlacing frequency. Every weft yarn passes over and under the
successive warp yarn and repeats the same pattern with alternate yarns in the following
row of weft yarn. A single repeat of the plain weave is depicted in Figure 2.1.
Further derivatives of the plain weave fabric can be made by passing two or more
adjacent warp threads and/or two or more adjacent weft threads at the same time. In this
way it is possible to obtain larger warp and weft covered areas than in plain weave
fabrics [4].
Figure 2.1. Plain weave [5]
2.1.1.1.2 Twill fabrics
Twill weave fabrics can be identified by the diagonal lines on the face of the fabric due
to warp and weft floats [3, 4]. This weave can further be categorized by “S” or “Z” twill
depending on the direction of the diagonal lines. The twill weave can be produced on a
Chapter 2 Literature review
24
minimum of 3 ends and picks (2/1-twill) with no theoretical upper limit [3]. A single
repeat of 3/1-twill is presented in Figure 2.2.
Figure 2.2. Twill weave [5]
2.1.1.1.3 Satin and sateen fabrics
Satin weave is a warp faced weave whereas the sateen weave is a weft faced weave. In
both satin and sateen weaves, the weave interlacements are arranged in such a way that
a smooth surface is achieved free from twill lines [3]. The smallest repeat of satin and
sateen is 5, a repeat of 5 end satin is shown in Figure 2.3. The most popular weave
repeats are 5 and 8 [4].
Figure 2.3. Satin weave [5]
Chapter 2 Literature review
25
2.1.1.2 3D woven fabrics
3D fabrics are recognised by the presence of thickness of the fabric in the Z-direction in
addition to the X and Y directions. These fabrics are woven with multiple warp or weft
layers. The thickness of a 3D fabrics is considerable in comparison to 2D fabrics [2].
Orthogonal and angle interlock fabrics are the most renowned classes of 3D woven
fabrics.
2.1.1.2.1 Orthogonal fabrics
In orthogonal fabrics, the straight yarns are arranged perpendicular to each other in X, Y
and Z directions as shown in Figure 2.4. The absence of crimp in the warp and weft
yarns makes this structure ideal for applications where non-crimp features are required.
Both isotropic and anisotropic preforms can be achieved by arranging the number of
yarns in each dimension [2, 6]. The orthogonal fabrics can further be classified into two
types, through-the-thickness orthogonal (Figure 2.4a) and layer-to layer-orthogonal
(Figure 2.4b). In orthogonal through-the-thickness fabrics, binding warp travels from
one surface of the preform to the other, holding together all the layers of the preform.
The orthogonal layer-to-layer is a multilayer woven fabric in which binding warps
travel from one layer to the adjacent layer and back.
Chapter 2 Literature review
26
Figure 2.4. Orthogonal weave, (a) Through-the-thickness, (b) Layer-to-layer
2.1.1.2.2 Angle-interlock fabrics
In angle-interlock structures, the warp (or weft) yarns are used to bind many layers of
weft (or warp) yarns with weft/warp yarns being straight. A third set of yarns (stuffer
yarns) can also be added in angle-interlock fabrics to increase fibre volume fraction and
in-plane strength [2]. Figure 2.5 presents the angle-interlock structure with warp, weft
and binder yarns. Like orthogonal fabrics, angle interlock structure can also be divided
into through-the-thickness (Figure 2.5a) and layer-to-layer (Figure 2.5b) angle interlock
depending on the passage of binder yarn
(a) (b)
Chapter 2 Literature review
27
Figure 2.5. Angle-interlock, (a) Through-the-thickness, (b) Layer-to-layer
2.1.2 Non-crimp stitched fabrics (NCF)
In non-crimp stitched fabrics, a light weight fibre/yarn is used as a loop which is sewn
or knitted around the reinforcement tow to create the fabric as can be seen in Figure 2.6.
These fabrics are commonly known as non-crimp fabrics (NCF) as the reinforcement
tow remains straight without crimp [7].
Composites made of NCF are used in aerospace, automotive, civil engineering and the
wind turbine industry due to their high strength properties [8].
Figure 2.6. NCF fabric [5]
(a) (b)
Chapter 2 Literature review
28
2.1.3 Braided fabrics
In braided fabrics, three or more threads are interlaced in such a way that they cross one
another in diagonal formation to form the braided structure. A braided structure is
presented in Figure 2.7. Braiding is a simple form of narrow fabric construction and the
limitations of knitting and weaving make braiding an important method of fabric
construction in the textile composites industry [1, 9].
Figure 2.7. Braided fabric [5]
2.2 Composite manufacturing processes
There are a number of different techniques for fibre reinforced polymer (FRP)
composite manufacturing. A brief introduction of some main manufacturing processes
is presented here.
2.2.1 Resin transfer moulding process (RTM)
Resin transfer moulding is the most common manufacturing process for structurally
capable composites. RTM is capable of producing large, complex and highly integrated
components. Additionally it has advantages of low capital cost, low mould cost and a
good work environment. In this process, the dry reinforcement is placed into a closed
mould and liquid resin is injected into the closed mould to impregnate the reinforcement.
Chapter 2 Literature review
29
A vacuum can also be applied in the mould cavity to infiltrate resin into the fabric,
which is known as vacuum assisted resin transfer moulding. During RTM, the injection
pressures range from 0.1 MPa to over 1 MPa depending on fibre content and resin type
[10, 11]. Figure 2.8 presents a schematic representation of RTM.
Figure 2.8. Resin transfer moulding [10]
2.2.2 Vacuum bagging
In the vacuum bagging process pressure is applied to the laminate once it is laid-up in
order to improve its consolidation. This is achieved by sealing a plastic film over the
dry laid-up laminate and onto the tool. To remove the air under the bag, a vacuum pump
is used and a pressure of one atmosphere is applied to achieve the consolidation of the
laminate. Figure 2.9 represent a typical vacuum bagging set up.
Mould tool
Mould tool
Resin
injection
under
pressure
Compression to secure
the mould
Optional
vacuum
assistance
Dry reinforcement preform
Chapter 2 Literature review
30
Figure 2.9. Vacuum bagging [12]
2.2.3 Autoclave processing
To achieve the maximum performance of thermoset composite materials, an increase in
the fibre to resin ratio and removal of all air voids is required. This can be obtained by
subjecting the material to elevated pressures and temperatures. Vacuum bagging can
apply atmospheric pressure equal to one bar. To increase pressure to more than one bar,
additional external pressure is required. This is achieved by autoclave processing in
which pressures up to 5~7 bar can be applied [13]. In the autoclave, a composite is laid
up and enclosed in a vacuum bag. A full or partial vacuum is drawn within the bag and
the bag is kept inside the autoclave chamber where the pressure is kept at more than one
bar. This extra pressure is added on the exterior of the bag. Simultaneously the
temperature is raised which results in reduction in viscosity of the polymer. Thus the
wetting of the reinforcing fibres and the consolidation becomes better. A schematic
representation of the autoclave process is presented in Figure 2.10.
To vacuum pump To vacuum gauge
Vacuum bagging film
Sealant tape
Release film Release coated
mould Laminate
Peel ply
Breather/absorption
fabric
Chapter 2 Literature review
31
Figure 2.10. Autoclave processing [14]
2.3 Compaction of textile preforms
During the manufacturing of composites, the dry preform is compacted under a certain
level of pressure, which changes the tow geometry of the dry preform, the resin
permeability and the final mechanical properties of the composites [15]. Hence, this
compaction process is considered an important parameter of the manufacturing
processes [16]. An intensive study is important to keep a record of the changes of tow
geometry at each and every stage of composite manufacturing so that reliable data can
be extracted for making accurate simulation tools [15].
Extensive research has been carried out on the compaction behaviour of different fabrics
using mechanical test methods and structural analysis [15-49].
Van Wyk [27] is considered the first researcher who treated the fibres under the
compression as a system of bending units. He derived a relationship between the
pressure and volume of a mass of wool fibres by assuming that compression of the wool
Bleeder
Vacuum connection Autoclave
Prepreg
Sealant tape Peel ply
Release film
perforated
Mould
Chapter 2 Literature review
32
fibres is due to bending of the wool fibres. He also suggested that compressibility
measurements constitute a convenient method of comparing the flexural elastic
properties of the fibres composing different wool samples.
Gutowski et al. [50] studied the compression behaviour of aligned fibre bundles. They
proposed a simple elastic deformation model for the transverse compression of aligned
fibre bundles by using Van Wyk’s idea of deformation. Later on, the deformation of
lubricated fibre bundles was also studied by Gutowski et al. [51] and they proposed a
method to measure the fibre volume fraction during consolidation.
De Jong et al. [21] developed a mechanical model for the lateral compression of wool
woven fabrics based on Van Wyk’s compression law of fibre assemblies. They
observed that the compressible surface layers were following Van Wyk’s law.
Compression behaviour of cotton woven fabrics in low load regions was studied by Hu
et al. [52]. They expressed the pressure thickness relationship of woven cotton fabrics
by a mathematical expression in the low load regime. In their research, they observed
that the equation proposed by Van Wyk [27] can be applied to cotton woven fabrics,
which was previously applied to the compression of wool fabrics by De Jong et al. [21].
A detailed study of the fabric compression behaviour by using a KES compression
tester and microscope images of fabric samples was carried out by Matsurdaira et al.
[53]. On the basis of their research, they divided the fabric pressure thickness curve into
three parts. The first and third steps of the fabric compressional curve were seen to obey
a linear relationship, whereas the second part of the curve was following an exponential
relationship. They attributed the first step of the curve to the bending of the fibres on the
Chapter 2 Literature review
33
fabric surface, due to which the curve behaved linearly. This typical behaviour is shown
in Figure 2.1.
Figure 2.11. Typical pressure thickness curve
The second step of the compression curve, which has an exponential behaviour, was
related with the hardness in compression due to friction between fibres. The third linear
step of the compressional curve was attributed to the fibre material and was explained
by the initial lateral compressional modulus of the fibres.
Pearce et al. [24] conducted experimental work on plain woven glass fibre
reinforcement to study the compressibility of plain fabric. They fitted the loading cycle
response of the fabric to the power law relationship and the relaxation cycles were fitted
to an exponential decay function. They also observed that there was more reduction in
thickness of single layers compared to the average layer thickness of a 5 layer stack.
They concluded that the reason for higher single layer thickness reduction was due to
fabric-to-fabric interaction rather than fabric-to-platen interactions.
Chapter 2 Literature review
34
Potluri et al. [15] characterised the different structural changes, which take place during
compression of fabric as shown in Figure 2.1. According to this, during the initial stage
of compression the gap between the plates and the fibre surface gradually reduces to
zero, which results in a decrease in crimp in one set of yarns and an increase in the
second set of yarns. When the gap between the plates and the fabric surface is zero, any
additional compression results in yarn flattening accompanied by crimp reduction in the
crossing yarns. This process of yarn flattening continues until the fibres are maximally
packed into each other.
Figure 2.12. Stages of fabric compression [15]
The compaction of fabrics is affected by different processing parameters. After resin
infusion during composite manufacturing, the dry fabric changes to a wet state. The
compaction behaviour of wet fabrics is different from that of dry fabrics. There is more
thickness reduction of the wet fabric compared to dry fabric due to compaction as
studied by various researchers [18, 25, 29, 38, 39, 54-60].
Chapter 2 Literature review
35
Robitaille et al. [18] performed experimental work on woven fabric to study the
compaction and relaxation behaviour of dry and wet fabrics. Distilled water was used to
saturate the fabric samples to study the wet compaction. They concluded that saturation
of fabric samples with water plays an important role on the compaction behaviour.
Kelly et al. [38] studied the response of dry and wet fibrous materials during
compaction. In their research, they performed experimental work on glass fibre
continuous filament mats. They observed the different compression responses of dry
and wet reinforcements.
Saunders et al. [25] studied the compaction response of glass woven fabric samples
after application of different resins. They applied three types of resin to the fabric
surface with higher to lower viscosity and observed the change in the thickness of the
layers. They concluded that there was no significant effect of the difference of the resin
viscosity on the compaction behaviour of the fabrics.
The compaction behaviour of fibrous materials is also affected by loading time as
observed by many researchers that pressure drops to maintain a specific preform
thickness under compression loading with time or thickness drops to maintain a
constant load with time [18, 22, 26, 29, 38, 39, 61].
This behaviour is depicted in Figure 2.13. As can be seen from this figure there is a
drop in thickness or stress with time to maintain constant pressure or constant thickness.
Chapter 2 Literature review
36
Figure. 2.13 Schematic representation of time dependent behaviour of fibrous
materials
Gutowski et al. [51] studied the elastic deformation of lubricated carbon bundles. In
their work they plotted the experimental results of fibre volume fractions against the
time scale of the experiment and observed that there was strong correlation between
fibre volume fraction and time. The experimental results showed that higher amounts of
fibre volume fraction can be achieved by longer processing times.
Kim et al. [22] performed experimental work on fibrous reinforcements to study the
compression and relaxation behaviour of reinforcement material. In their research, they
studied E-glass, graphite cloth, mat, unidirectional material and combinations of two
different fibre orientations. They observed that there was stress relaxation during
compression. They related the amount of stress relaxation to fibre orientation and
observed that relaxation behaviour decreased with fibre alignment. They explained the
relaxation behaviour of fibre reinforcements during compression with the Maxwell-
Wiechert viscoelastic model and fitted their data using this model and observed good
conformity between the model and experimental results.
Time Time
Th
ick
nes
s
Str
ess
Chapter 2 Literature review
37
Francois et al. [16] collected the experimental results of compression and relaxation
from several previous researchers and fitted the relaxation data using a power law
equation in which pressure was taken as a function of time as in equation 2.1.
(2.1)
In this equation P is the final compaction pressure, P0 is the initial compaction pressure,
C is the pressure decay after 1 second and D is the relaxation index.
Luo et al. [23] studied the compression and relaxation response of a new sandwich
textile preform for liquid composite moulding and fitted the experimental data with the
power law equation in the following form as in equation 2.2.
(2.2)
Where, P and P0 are the compression pressure before relaxation and 1000 seconds after
relaxation respectively. is the relaxation index and is defined by the following
equation 2.3.
(2.3)
During the comparison study of the dry and wet fabric samples different wetting agents
were used to wet out the fabric samples. Glucose syrup was used by Kelly et al. [38] to
wet the glass fibre filament mat samples. Bickerton et al. [29] used a non-reactive corn
syrup solution as the test fluid for the wet test experiments. Robitaille et al. [18]
performed experimental work on different fabrics in dry and wet conditions to study
their compaction response. They used distilled water to saturate the fabric samples.
Chapter 2 Literature review
38
Lawrence et al. [54] studied the compression of the fibrous preform materials by using
experimental and simulation tools. In the research they studied the wet preform samples
saturated with water and oil for the compression behaviour. They observed that in water
saturated samples, there were no viscous forces present. They suggested that by using
water as a lubricating fluid, the force exerted on the fabric can be directly measured.
The fibre volume fraction of the preform during compaction loading can be varied by
changing the cross-head movement of the compression machine. It was noticed by
different researchers that cyclic compaction increases the fibre volume fraction of the
preform by decreasing the fabric thickness [19, 22, 39].
Kruckenberg et al. [39] studied the effect of vibration compaction on plain woven
fabrics. They observed that vibration compaction plays a significant role in the
compaction response of plain woven fabrics.
It was noticed by Kim et al. [22] that there was a change in the thickness of the fibre
reinforcement with change in the tensile machine cross-head speed. The thickness of the
preform was seen to decrease with slower speed. They discussed this behaviour with the
rearrangement of the fibres with a prolonged period at a slow speed. Saunders et al. [25]
performed compression testing on different woven fabrics and observed that there was
no significant effect of machine speed on the thickness of the fabrics in the dry state but
they observed significant effects of machine speed on the compression of wet samples.
Compaction behaviour of single layer and multilayer preforms was also studied by
using theoretical models. Chen et al. [30] developed a 3D model of plain woven fabric
to predict the compressive behaviour of single layer preforms. For this model, the yarns
were treated as transversely isotropic solids. It was assumed that there were no voids
Chapter 2 Literature review
39
and gaps between the yarns and there is deformation only in the yarn shape without
change in the yarn cross-sectional area. Using this model, they established analytical
expressions for the fibre volume fraction, the applied compressive force and the preform
thickness.
The micromechanical model of Chen et al. [34] describes the compaction behaviour of
single layer fabrics with plain, twill and satin weaves. The deformation mechanism was
studied at two different levels including micro deformation of yarn cross-section and
macro deformation of yarn bending along with yarn waveform flattening. The
relationship between thickness and pressure for plain, twill and satin fabrics was also
investigated. They observed that single layer plain weave fabric is most difficult to
compact whereas the single layer satin fabric is the easiest to compact. The least
deformation of single layer plain fabric was attributed to the presence of its highest ratio
of curved parts to the straight parts due to which there was more contribution of macro-
bending deformation in resisting compaction of single layer plain weave fabric. In
addition it was shown that compression behaviour of single layer fabric was affected by
the initial fibre packing ratio of the yarns.
An analytical model for compaction of multilayer woven preforms was presented by
Chen et al. [31] in which they established a relationship between fibre volume fraction,
the applied compressive force and the preform thickness for nested and non-nested
cases.
A micromechanical model was developed by Chen et al. [35] to investigate the
compaction behaviour of multilayer plain woven fabric preforms. The deformation
mechanism of fabrics at different hierarchical levels was studied. The deformation and
Chapter 2 Literature review
40
compaction of yarn cross-section, the flattening of the yarn waveform, nesting between
adjacent layers and inter-layer packing were taken into consideration to develop this
model. They showed in their model that the nesting of the layers is affected by the
number of layers. Also the shifting of the layers was shown as an important factor in
defining the compaction behaviour of the multilayer preforms.
The compaction behaviour of multilayer preforms differs from single layer preforms
due to the presence of nesting in the multilayer preforms. Nesting of the layers in a
multilayer preform is a geometrical and mechanical phenomenon which influences the
thickness of the preform during the composite manufacturing process and the
mechanical properties of the composite [44]. The study of the compaction of multilayer
preforms has been investigated by a number of researchers [15, 20, 24, 25, 31-33, 35, 39,
40, 44, 48, 57, 62, 63].
Nesting of the layers reduces the gaps between neighbouring layers in the stack and
increases the gap between the mould and the topmost layer during composite
manufacturing [64]. The effect of the nesting on the permeability of the layers was
studied by various researchers [62, 64-66].
Karahan [67] performed a damage study on 2/2 twill woven carbon multilayer
composites and observed that the nesting of the layers reduces the delamination of the
layers in the multilayer stack.
Tekalur et al. [68] studied the mechanical behaviour of glass and carbon composites.
The glass and carbon composites were subjected to quasi-static and high strain loading
to study the tensile, compressive and shear properties of the composites. They observed
that the inter-laminar shear properties of the glass composites were better than for the
Chapter 2 Literature review
41
carbon composites. This was due to higher levels of nesting in glass composites
compared with carbon composites.
Ebraheem et al. [69] developed a geometrical model for plain woven fabric to analyse
the inter-laminar shear distribution between nested layers of composites. By using this
geometrical model they concluded that nesting caused a reduction in the inter-laminar
shear stress and compaction of nested layers reduced the variation in the stress
distribution within the layers.
Hoes et al. [66] performed experimental work on plain woven glass fabric to study the
effect of nesting on the permeability scatter using electrical resistance sensors. In their
research, they studied the scatter in permeability with minimum, random and maximum
nested cases. They came to the conclusion that the nesting of the individual layers was
the major source of the scatter of the permeability measurement data.
Nesting of the layers in a multilayer stack depends upon different parameters. It was
observed by Kruckenberg et al.[39] that nesting of the layers depends upon the shifting
of the layers.
Potluri et al. [15] described cross-sectional shape of yarn, yarn spacing, applied pressure
and surface waviness as parameters which effect the nesting of the layers.
Lomov et al. [44] established a geometrical model for the nesting of the layers in textile
laminates. They studied the layer nesting in multilayer preform by varying the different
parameters. The effects of flatness of the yarns, tightness and balance of the fabric, the
number of layers in the multilayer stack and the effect of weaving/braiding and knitted
pattern was studied for the nesting phenomenon. They observed that higher thread
spacing in the fabric enhanced the nesting of the layers in the stack and vice versa.
Chapter 2 Literature review
42
Higher nesting was seen with a decrease in the float length of the weave. In the case of
non-crimp stitched fabric, they defined the nesting by the stitching pattern.
Nesting of the layers in the multilayer stack affect the stack layer thickness [20, 25, 31,
33, 41, 44, 57] and the permeability of the preform [62, 66].
Nesting of the layers was calculated by different researchers in terms of nesting factor
or nesting coefficients [15, 41, 70, 71]. Potluri et al.[15, 70] defined the nesting factor
by equation 2.4.
∑
(2.4)
In equation 2.4, higher the nesting factor, the lower will be the nesting of the layers and
vice versa. They described that the nesting factor as 1 if the layers sit exactly on top of
each other without nesting. But in actual cases, the layers do not sit exactly upon each
other due to shifting of the layers so the nesting factor will be less than 1.
By using equation 2.4, they calculated the nesting factor for two layer stacks of glass
plain woven fabric. They observed that the nesting factor at the start of the compression
without any pressure was 0.83, which decreased to 0.77 with the application of 5 kPa
pressure [15] which shows nesting increases with increase in pressure.
Karin et al.[71] studied the nesting of the layers in biaxial and triaxial braids. They used
the thickness of a single layer and the laminate thickness for the calculation of the
Chapter 2 Literature review
43
nesting factor. The nesting factors for biaxial and triaxial braids were calculated by
using the equations 2.5 and 2.6.
(2.5)
(2.6)
In these equations, nl is the number of layers in the laminate, tSLb is the single layer
thickness of biaxial braid, tSLt is the single layer thickness of triaxial braid and tlam is the
laminate thickness. The single layer thickness of biaxial and triaxial braids was
described by the thickness of the crossing yarns.
Lomov et al. [41] calculated the nesting of the multilayer stack in terms of the nesting
coefficient by using the following equation 2.7.
(2.7)
In the above equation, N is the number of plies, t is the ply thickness and p is the
pressure applied. In this equation, the higher the nesting coefficient, the higher will be
the nesting of the plies and vice versa. In their experimental work, they calculated the
nesting coefficients for 4 layer fabrics. The observed higher nesting coefficient for the
plain woven fabric (16.2 % ), whereas the nesting coefficient for the satin fabric was
5.2 %, which shows that plain fabrics have better nesting compared to satin fabrics.
Yurgartis et al. [63] developed a method to quantify yarn shape and nesting in plain
woven composites. In their method they used the angle match data to characterize the
nesting of the layers.
Chapter 2 Literature review
44
Kruckenberg et al. [39] performed work on fibreglass plain woven fabrics and studied
the nesting of the layers. In their work, they studied the nesting of the layers using the
microstructure of the composite laminates and nesting shift was calculated using the
following equation 2.8.
(2.8)
In equation 2.8, ‘a’ is half of the yarn width, ‘d’ is the tow shift and ‘s’ is the spacing
between the yarns. They calculated these values from the microstructure images of the
laminates.
As already discussed the compaction and nesting of the layers affect the resin
permeability during composite manufacturing. The resin permeability has been studied
extensively by a number of researchers using experimental and simulation tools [36, 40,
64, 65, 72-88]. The resin flow during composite manufacturing depends on different
parameters. Grujicic et al. [62] studied the effect of shear, compaction and nesting of the
layers on the permeability of orthogonal plain woven fabrics. According to them, the
resin flows mainly through the pores and channels or inter-tow voids in the
neighbouring layers. Yu and Lee [83] developed a simple in-plane permeability model
for textile fabrics. They calculated the flow in the channels or gaps between fibre tows
by using the one-dimensional Stokes equation and used the one-dimensional Brinkman
equation for flow in the fibre tows. From the analysis of their research results, they
concluded that the permeability of fibre preforms is mainly determined by gaps or
channels between the fibre tows whereas the effect of intra tow voids on the
permeability of the preform is negligible. It was emphasised by Chen et al. [33] that
Chapter 2 Literature review
45
the understanding the existence of resin channels is essential for making resin
simulation tools.
Hoes et al. [66] studied the effect of layer nesting on the resin permeability of the
woven glass fabrics and concluded that layer shifting and nesting in multilayer preforms
are the major sources of variations in permeability values.
The effect of compaction on tow geometry and voids during composite manufacturing
has been studied by different researchers [17, 20, 25, 33, 39, 89, 90]. Potluri et al. [17]
developed a novel stress freezing technique for studying the compressional behaviour of
woven fabrics. In their method, fabric impregnated with clear resin was subjected to
compressive stress and held under load allowing the resin to cure. The cured fabric
samples were analysed using a scanning electron microscope (SEM) and different tow
geometry parameters were investigated. They observed a decrease in crimp percentage
and crimp amplitude of the tows while an increase was observed in yarn spacing on
application of pressure.
Saunders et al. [20] studied the microstructure of plain woven fabrics under
compression loading. The resin impregnated fabric samples were cured and studied
using the SEM. The study contains 20 ply laminate samples compressed at different
pressure levels (4.4 kPa to 1768 kPa). On the basis of their analysis, they concluded that
the compression of woven cloths can be modelled as three different modes. The first
compression mode was dominated by the nesting of the layers in which layers came
closer to each other. In the second mode of compression, the yarns were deformed by
yarn amplitude reduction and the thickness of the individual plies reduced. In the third
mode of compression, they observed that fibrous yarns were deformed and compressed
Chapter 2 Literature review
46
individually. According to this study, the first mode of compression was dominant over
a wide range of pressures in the low and intermediate pressure regions while the second
and third modes were significant at higher pressures .
Kruckenberg et al. [39] studied the microstructure of plain woven composite laminates.
In this research, they applied vibration compaction to four ply plain woven fabric
samples at different loads and cured the resin impregnated samples in order to study the
microstructure of the fabric. They observed that nesting of the layers is dominant at low
loads. The yarn width remained constant throughout the compaction and there was
decrease in yarn height.
Saunders et al. [25] performed microstructural analysis on laminates of plain, twill and
satin fabrics by using SEM. They studied average area porosity (resin rich areas), areas
pore structure and voids for all the fabrics.
Chen et al. [33] conducted experimental and theoretical studies on the compaction
behaviour of fabric preforms in the resin transfer moulding process. The research was
conducted on continuous strand mats, plain woven fabrics and unidirectional knitted
materials. From analysing the results, several factors were identified which were the
main contributors to the preform compaction. These factors are presented in Figure 2.14.
Chapter 2 Literature review
47
Figure 2.14. Main factors affecting the compaction behaviour of preform [33]
Different image analysis techniques are being used for microstructural study of fibre
composite materials. A review of different computer based techniques for image
analysis of fibre reinforced composites can be seen in literature [91].
Schell et al. [92] conducted a study of the meso-structure analysis of glass fibre
reinforced polymers by using computed tomography (CT). They studied the resin
channels in fibre reinforced composites and analysed the voids in the composite
samples by selecting the voids by a threshold method.
Computed tomography is a non-destructive technique, which can provide geometrical
features of the internal structure of an object. CT was initially used in the biomedical
industry [93]. The use of CT became popular in the composites industry due to its
ability to study the objective in 3D. Several researchers have used CT for their research
work in the composites field [45, 89, 92, 94-122].
Desplentere et al. [106] used CT to characterise the micro-structural variation of 3D
fabrics. They determined the accuracy of the CT technique by comparing the measured
values of yarn thickness, width and spacing between the yarns obtained from CT with
Chapter 2 Literature review
48
measured values of the same parameters obtained from optical micrographs of the same
textiles. They observed that the difference between the data obtained by these two
techniques was not significant. They concluded that CT is a reliable technique to obtain
the input for modelling textiles or textile composites.
Schell et al. [92] studied glass fibre reinforced polymer samples to investigate the
geometry of the fibre bundles and voids in the samples using micro CT. They described
CT as an appropriate technique to investigate the fibre bundle structure.
Little et al. [123] conducted a study to compare the accuracy and reliability of various
void characterisation techniques in composite materials. They studied voids in the
samples of carbon fibre reinforced composites using microscopy, Archimedes density
measurements and micro CT. After their study they concluded that micro CT analysis is
the most accurate and reliable technique for characterising voids in composite materials.
Kastner et al. [124] used CT for the characterisation of carbon fibre reinforced
composites and compared the results with other different segmentation methods
including ultrasonic testing and acid digestion. They observed that X-ray CT is a
powerful non-destructive technique for characterisation and especially an excellent tool
for measuring the volumetric porosity of carbon fibre reinforced composites. They also
mentioned that CT is not only limited to measuring the porosity but additionally the size,
shape and position of the pores can also be studied using this technique. Finally they
concluded that CT can be used as a substitute method for non-destructive measurement
of porosity in composites.
Computed tomography has limitation that it requires a rigid sample that does not move
during the CT scan. In some cases, particularly when going to very high resolution,
Chapter 2 Literature review
49
special care is needed to achieve this, e.g. by consolidation of fixation of the sample.
This can be considered as destruction of the original sample. Due to variety in sample
size, shape and composition, no fixed and generally accepted protocols exist for CT
scanning. A large number of free parameters can be chosen arbitrary, such as tube
voltage , exposure time, etc., all influencing the final results and making this technique
more manual dependent [125]. Additionally the operational cost of this technique is
high.
References
[1] Textile Terms and Definitions. The Textile Institute.
[2] Badawi SS. Development of the weaving machine and 3D woven spacer fabric
structures for lightweight composites materials. Dr-Ing Dissertation. 2007.
[3] Robinson A.T.C MR. Woven cloth construction. The Textile Institute. 1973.
[4] Horrocks A R ASC. Hand book of technical textile. The Textile Institute.
[5] http://texgen.sourceforge.net/index.php/Main_Page.
[6] http://www.texeng.co.uk/orthogonal.html.
[7] Plymouth Uo. Composites design and manufacture.
http://wwwtechplymacuk/sme/mats324/.
[8] Lomov SV. Non-crimp fabric composites: manufacturing, properties and
applications. Woodhead publishing.
[9] Gopalakrishnan D B, Ramakrishnan V , Sabarinath T. Braiding & narrow width
fabrics. The Indian Textile Journal. 2007.
Chapter 2 Literature review
50
[10] <http://www.carbonfiberguru.com/> carbon fibre processing part 3 of 12-Resin
transfer molding ( accessed: 12 February 2014).
[11] Astrom BT. Manufacturing of polymer composites. 1997.
[12] <http://www.netcomposites.com/guide/vacuum-bagging/53>, Vaccum bagging,
( accessed 12 Feburary 2014).
[13] <http://www.netcomposites.com/guide/autoclave-design/59> Autoclave design,
(accessed: 12 Feburary 2014).
[14] <http://www.rampf-tooling.de/en/products-solutions/composites-a-fiber-
composite-material-with-a-promising-future/raku-tool-composites-production-
processes/>, Autoclav process, ( accessed 12 Feburary 2014)
[15] Potluri P, Sagar TV. Compaction modelling of textile preforms for composite
structures. Composite Structures. 2008;86(1-3):177-85.
[16] Robitaille F, Gauvin R. Compaction of textile reinforcements for composites
manufacturing. I: Review of experimental results. Polymer Composites.
1998;19(2):198-216.
[17] Potluri P, Wilding MA, Memon A. A novel stress-freezing technique for studying
the compressional behavior of woven fabrics. Textile Research Journal.
2002;72(12):1073-8.
[18] Robitaille F, Gauvin R. Compaction of textile reinforcements for composites
manufacturing. II: Compaction and relaxation of dry and H2O-saturated woven
reinforcements. Polymer Composites. 1998;19(5):543.
Chapter 2 Literature review
51
[19] Robitaille F, Gauvin R. Compaction of textile reinforcements for composites
manufacturing. III: Reorganization of the fiber network. Polymer Composites.
1999;20(1):48.
[20] Saunders RA, Lekakou C, Bader MG. Compression and microstructure of fibre
plain woven cloths in the processing of polymer composites. Composites Part A:
Applied Science and Manufacturing. 1998;29(4):443-54.
[21] De Jong S, Snaith JW, Michie NA. A mechanical model for the lateral compression
of woven fabrics. Textile Research Journal. 1986;56(12):759-67.
[22] Kim YR, McCarthy SP, Fanucci JP. Compressibility and relaxation of fiber
reinforcements during composite processing. Polymer Composites. 1991;12(1):13-9.
[23] Luo Y, Verpoest I. Compressibility and relaxation of a new sandwich textile
preform for liquid composite molding. Polymer Composites. 1999;20(2):179-91.
[24] Pearce N, Summerscales J. The compressibility of a reinforcement fabric.
Composites Manufacturing. 1995;6(1):15-21.
[25] Saunders RA, Lekakou C, Bader MG. Compression in the processing of polymer
composites 1. A mechanical and microstructural study for different glass fabrics and
resins. Composites Science and Technology. 1999;59(7):983-93.
[26] Seung Hwan L, Jeong Hwa H, Seong Yun K, Jae Ryoun Y, Young Seok S.
Compression and relaxation behavior of dry fiber preforms for resin transfer molding.
Journal of Composite Materials.44(15):1801-20.
[27] Van Wyk CM. Note on the compressibility of wool. Journal of the Textile Institute
1946;37(12):T285-92.
Chapter 2 Literature review
52
[28] Acheson JA, Simacek P, Advani SG. The implications of fiber compaction and
saturation on fully coupled VARTM simulation. Composites Part A: Applied Science
and Manufacturing. 2004;35(2):159-69.
[29] Bickerton S, Buntain MJ, Somashekar AA. The viscoelastic compression behavior
of liquid composite molding preforms. Composites Part A: Applied Science and
Manufacturing. 2003;34(5):431-44.
[30] Chen B, Chou T-W. Compaction of woven-fabric preforms in liquid composite
molding processes: single-layer deformation. Composites Science and Technology.
1999;59(10):1519-26.
[31] Chen B, Chou T-W. Compaction of woven-fabric preforms: nesting and multi-
layer deformation. Composites Science and Technology. 2000;60(12-13):2223-31.
[32] Chen B, Cheng AHD, Chou TW. A nonlinear compaction model for fibrous
preforms. Composites Part A: Applied Science and Manufacturing. 2001;32(5):701-7.
[33] Chen B, Lang EJ, Chou T-W. Experimental and theoretical studies of fabric
compaction behavior in resin transfer molding. Materials Science and Engineering A.
2001;317(1-2):188-96.
[34] Chen Z-R, Ye L, Kruckenberg T. A micromechanical compaction model for woven
fabric preforms. Part I: Single layer. Composites Science and Technology.
2006;66(16):3254-62.
[35] Chen Z-R, Ye L. A micromechanical compaction model for woven fabric preforms.
Part II: Multilayer. Composites Science and Technology. 2006;66(16):3263-72.
[36] Endruweit A, Long AC. Analysis of compressibility and permeability of selected
3D woven reinforcements. Journal of Composite Materials. 2010;44(24):2833-62.
Chapter 2 Literature review
53
[37] Kelly PA. A viscoelastic model for the compaction of fibrous materials. Journal of
The Textile Institute.102(8):689-99.
[38] Kelly PA, Umer R, Bickerton S. Viscoelastic response of dry and wet fibrous
materials during infusion processes. Composites Part A: Applied Science and
Manufacturing. 2006;37(6):868-3.
[39] Kruckenberg T, Ye L, Paton R. Static and vibration compaction and microstructure
analysis on plain-woven textile fabrics. Composites Part A: Applied Science and
Manufacturing. 2008;39(3):488-502.
[40] Liu XL, Falzon PJ, Sweeting R, Paton R. Effective compressibility and
permeability of multi-layer non-crimp fiberglass reinforcements. Journal of Reinforced
Plastics and Composites. 2004;23(8):861-79.
[41] Lomov SV, Gorbatikh L, Kotanjac Z, Koissin V, Houlle M, Rochez O, et al.
Compressibility of carbon woven fabrics with carbon nanotubes/nanofibres grown on
the fibres. Composites Science and Technology. 2011;71(3):315-25.
[42] Lomov SV, Gorbatikh L, Verpoest I. Compression behaviour of a fibre bundle with
grafted carbon nanotubes. Carbon. 2011;49(13):4458-65.
[43] Lomov SV, Verpoest I. Compression of woven reinforcements: A mathematical
model. Journal of Reinforced Plastics and Composites. 2000;19(16):1329-50.
[44] Lomov SV, Verpoest I, Peeters T, Roose D, Zako M. Nesting in textile laminates:
geometrical modelling of the laminate. Composites Science and Technology.
2003;63(7):993-1007.
[45] Mahadik Y, Hallett SR. Effect of fabric compaction and yarn waviness on 3D
woven composite compressive properties. Composites Part A: Applied Science and
Manufacturing.42(11):1592-600.
Chapter 2 Literature review
54
[46] Potluri P, Parlak I, Ramgulam R, Sagar TV. Analysis of tow deformations in textile
preforms subjected to forming forces. Composites Science and Technology.
2006;66(2):297-305.
[47] Yang B, Kozey V, Adanur S, Kumar S. Bending, compression, and shear behavior
of woven glass fiber–epoxy composites. Composites Part B: Engineering.
2000;31(8):715-21.
[48] Nguyen QT, Vidal-Sallé E, Boisse P, Park CH, Saouab A, Bréard J, et al.
Mesoscopic scale analyses of textile composite reinforcement compaction. Composites
Part B: Engineering. 2013;44(1):231-41.
[49] Williams CD, Grove SM, Summerscales J. The compression response of fibre-
reinforced plastic plates during manufacture by the resin infusion under flexible tooling
method. Composites Part A: Applied Science and Manufacturing. 1998;29(1–2):111-4.
[50] Gutowski T, Kingery J, Boucher D. Experiments in composites consolidation: fiber
deformation. ANTEC 86. 1986:1316-20.
[51] Gutowski TG, Dillon G. The elastic deformation of lubricated carbon fiber bundles:
Comparison of theory and experiments. Journal of Composite Materials.
1992;26(16):2330-47.
[52] Hu J, Newton A. Low-load lateral-compression behaviour of woven fabrics.
Journal of the Textile Institute. 1997;88(3):242-54.
[53] Matsudaira M, Qin H. Features and mechanical parameters of a fabric's
compressional property. Journal of the Textile Institute. 1995;86(3):476-86.
[54] Lawrence JMS, Pavel; Frey, Pierre; Bhat, Prabhas; Gebauer, Timo; Advani, Suresh
G. The compaction behavior of fibrous preform materials during the VARTM infusion.
10th Esaform conference on material forming 2007;907:1039-45.
Chapter 2 Literature review
55
[55] Francucci G, Vázquez A, Rodríguez ES. Key differences on the compaction
response of natural and glass fiber preforms in liquid composite molding. Textile
Research Journal. 2012;82(17):1774-85.
[56] Debnath S, Madhusoothanan M. Studies on compression properties of polyester
needle-punched nonwoven fabrics under dry and wet conditions. Journal of Industrial
Textiles. 2012;41(4):292-308.
[57] Hammami A. Effect of reinforcement structure on compaction behavior in the
vacuum infusion process. Polymer Composites. 2001;22(3):337-48.
[58] Yenilmez B, Senan M, Murat Sozer E. Variation of part thickness and compaction
pressure in vacuum infusion process. Composites Science and Technology. 2009;69(11-
12):1710-9.
[59] Yuexin D, Zhaoyuan T, Yan Z, Jing S. Compression responses of preform in
vacuum infusion process. Chinese Journal of Aeronautics. 2008;21(4):370-7.
[60] Francucci G, Rodríguez ES, Vázquez A. Experimental study of the compaction
response of jute fabrics in liquid composite molding processes. Journal of Composite
Materials. 2012;46(2):155-67.
[61] Kim YR, McCarthy SP, Fanucci JP. Compressibility and relaxation of fiber
reinforcements during composite processing. Polymer Composites. 1991;12(1):13.
[62] Grujicic M, Chittajallu KM, Walsh S. Effect of shear, compaction and nesting on
permeability of the orthogonal plain-weave fabric preforms. Materials Chemistry and
Physics. 2004;86(2-3):358-69.
[63] Yurgartis SW, Morey K, Jortner J. Measurement of yarn shape and nesting in
plain-weave composites. Composites Science and Technology. 1993;46(1):39-50.
Chapter 2 Literature review
56
[64] Dungan FD, Senoguz MT, Sastry AM, Faillaci DA. Simulations and experiments
on low-pressure permeation of fabrics: Part I-3D modeling of unbalanced fabric. Journal
of Composite Materials. 2001;35(14):1250-84.
[65] Simacek P, Advani SG. Permeability model for a woven fabric. Polymer
Composites. 1996;17(6):887-99.
[66] Hoes K, Dinescu D, Sol H, Parnas RS, Lomov S. Study of nesting induced scatter
of permeability values in layered reinforcement fabrics. Composites Part A: Applied
Science and Manufacturing. 2004;35(12):1407-18.
[67] Karahan M. Investigation of damage initiation and propagation in 2 × 2 twill
woven carbon/epoxy multi-layer composites. Textile Research Journal. 2011;81(4):412-
28.
[68] Tekalur SA, Shivakumar K, Shukla A. Mechanical behavior and damage evolution
in E-glass vinyl ester and carbon composites subjected to static and blast loads.
Composites Part B: Engineering. 2008;39(1):57-65.
[69] Shady E, Gowayed Y. Interlaminar shear stress distribution between nested layers
of plain weave composites. Polymer Composites. 2010;31(11):1838-45.
[70] Potluri P, Manan A, Francke M, Day RJ. Flexural and torsional behaviour of
biaxial and triaxial braided composite structures. Composite Structures. 2006;75(1–
4):377-86.
[71] Birkefeld K, Röder M, Von Reden T, Bulat M, Drechsler K. Characterization of
biaxial and triaxial braids: Fiber architecture and mechanical properties. Applied
Composite Materials.1.
[72] Adams KL, Rebenfeld L. In-plane flow of fluids in fabrics: Structure/Flow
characterization. Textile Research Journal. 1987;57(11):647-54.
Chapter 2 Literature review
57
[73] Chen R, Dong C, Liang Z, Zhang C, Wang B. Flow modeling and simulation for
vacuum assisted resin transfer molding process with the equivalent permeability method.
Polymer Composites. 2004;25(2):146-64.
[74] Chen Z, Ye L, Liu H. Effective permeabilities of multilayer fabric preforms in
liquid composite moulding. Composite Structures. 2004;66(1-4):351-7.
[75] Gebart BR. Permeability of unidirectional reinforcements for RTM. Journal of
Composite Materials. 1992;26(8):1100-33.
[76] Griffin PR, Grove SM, Russell P, Short D, Summerscales J, Guild FJ, et al. The
effect of reinforcement architecture on the long-range flow in fibrous reinforcements.
Composites Manufacturing. 1995;6(3–4):221-35.
[77] Mogavero J, Advani SG. Experimental investigation of flow through multi-layered
preforms. Polymer Composites. 1997;18(5):649-55.
[78] Pillai KM, Advani SG. Numerical simulation of unsaturated flow in woven fiber
preforms during the resin transfer molding process. Polymer Composites.
1998;19(1):71-80.
[79] Senoguz MT, Dungan FD, Sastry AM, Klamo JT. Simulations and experiments on
low-pressure permeation of fabrics: Part II ”The variable gap model and prediction of
permeability. Journal of Composite Materials. 2001;35(14):1285-322.
[80] Simacek P, Advani SG. Modeling resin flow and fiber tow saturation induced by
distribution media collapse in VARTM. Composites Science and Technology.
2007;67(13):2757-69.
[81] Song YS, Youn JR. Flow advancement through multi-layered preform with
sandwich structure. Composites Part A: Applied Science and Manufacturing.
2007;38(4):1082-8.
Chapter 2 Literature review
58
[82] Verleye B, Lomov SV, Long A, Verpoest I, Roose D. Permeability prediction for
the meso-macro coupling in the simulation of the impregnation stage of Resin Transfer
Moulding. Composites Part A: Applied Science and Manufacturing. 2010;41(1):29-35.
[83] Yu B, James Lee L. A simplified in-plane permeability model for textile fabrics.
Polymer Composites. 2000;21(5):660-85.
[84] Xiao X, Zeng X, Long A, Lin H, Clifford M, Saldaeva E. An analytical model for
through-thickness permeability of woven fabric. Textile Research Journal.
2012;82(5):492-501.
[85] Arbter R, Beraud JM, Binetruy C, Bizet L, Bréard J, Comas-Cardona S, et al.
Experimental determination of the permeability of textiles: A benchmark exercise.
Composites Part A: Applied Science and Manufacturing. 2011;42(9):1157-68.
[86] Govignon Q, Bickerton S, Kelly PA. Simulation of the reinforcement compaction
and resin flow during the complete resin infusion process. Composites Part A: Applied
Science and Manufacturing. 2010;41(1):45-57.
[87] Ahn KJ, Seferis JC, Berg JC. Simultaneous measurements of permeability and
capillary pressure of thermosetting matrices in woven fabric reinforcements. Polymer
Composites. 1991;12(3):146-52.
[88] Amico S, Lekakou C. An experimental study of the permeability and capillary
pressure in resin-transfer moulding. Composites Science and Technology.
2001;61(13):1945-59.
[89] Mahadik Y, Brown KAR, Hallett SR. Characterisation of 3D woven composite
internal architecture and effect of compaction. Composites Part A: Applied Science and
Manufacturing.41(7):872-80.
Chapter 2 Literature review
59
[90] Jortner J. Microstructure of cloth-reinforced carbon-carbon laminates. Carbon.
1992;30(2):153-63.
[91] Guild FJ, Summerscales J. Microstructural image analysis applied to fibre
composite materials: a review. Composites. 1993;24(5):383-93.
[92] Schell JSU, Renggli M, Van Lenthe GH, Müller R, Ermanni P. Micro-computed
tomography determination of glass fibre reinforced polymer meso-structure.
Composites Science and Technology. 2006;66(13):2016-22.
[93] Muller R. The zurich experience: One decade of three-dimensional high-resolution
computed tomography. Top Magn Reson Imaging. 2002;13(5):307-22.
[94] Tan KT, Watanabe N, Iwahori Y. X-ray radiography and micro-computed
tomography examination of damage characteristics in stitched composites subjected to
impact loading. Composites Part B: Engineering. 2011;42(4):874-84.
[95] Tiseanu I, Tsitrone E, Kreter A, Craciunescu T, Loarer T, Pégourié B, et al. X-ray
micro-tomography studies on carbon based composite materials for porosity network
characterization. Fusion Engineering and Design. 2011;86(9–11):1646-51.
[96] Seltzer R, González C, Muñoz R, Llorca J, Blanco-Varela T. X-ray
microtomography analysis of the damage micromechanisms in 3D woven composites
under low-velocity impact. Composites Part A: Applied Science and Manufacturing.
2013;45(0):49-60.
[97] McCombe GP, Rouse J, Trask RS, Withers PJ, Bond IP. X-ray damage
characterisation in self-healing fibre reinforced polymers. Composites Part A: Applied
Science and Manufacturing. 2012;43(4):613-20.
Chapter 2 Literature review
60
[98] Schilling PJ, Karedla BR, Tatiparthi AK, Verges MA, Herrington PD. X-ray
computed microtomography of internal damage in fiber reinforced polymer matrix
composites. Composites Science and Technology. 2005;65(14):2071-8.
[99] Parnas RS, Wevers M, Verpoest I. Using textile topography to analyze X-ray CT
data of composite microstructure. Polymer Composites. 2003;24(2):212-20.
[100] Latil P, Orgéas L, Geindreau C, Dumont PJJ, Rolland du Roscoat S. Towards the
3D in situ characterisation of deformation micro-mechanisms within a compressed
bundle of fibres. Composites Science and Technology. 2011;71(4):480-8.
[101] Soltani P, Johari MS, Zarrebini M. Tomography based determination of
transverse permeability in fibrous porous media. Journal of Industrial Textiles. 2013.
[102] Badel P, Vidal-Sallé E, Maire E, Boisse P. Simulation and tomography analysis of
textile composite reinforcement deformation at the mesoscopic scale. Composites
Science and Technology. 2008;68(12):2433-40.
[103] Pandita SD, Verpoest I. Prediction of the tensile stiffness of weft knitted fabric
composites based on X-ray tomography images. Composites Science and Technology.
2003;63(2):311-25.
[104] Cosmi F, Bernasconi A, Sodini N. Phase contrast micro-tomography and
morphological analysis of a short carbon fibre reinforced polyamide. Composites
Science and Technology. 2011;71(1):23-30.
[105] Feng Y, Feng Z, Li S, Zhang W, Luan X, Liu Y, et al. Micro-CT characterization
on porosity structure of 3D Cf/SiCm composite. Composites Part A: Applied Science
and Manufacturing. 2011;42(11):1645-50.
Chapter 2 Literature review
61
[106] Desplentere F, Lomov SV, Woerdeman DL, Verpoest I, Wevers M, Bogdanovich
A. Micro-CT characterization of variability in 3D textile architecture. Composites
Science and Technology. 2005;65(13):1920-30.
[107] Pazmino J, Carvelli V, Lomov SV. Micro-CT analysis of the internal deformed
geometry of a non-crimp 3D orthogonal weave E-glass composite reinforcement.
Composites Part B: Engineering. (0).
[108] Chateau C, Gélébart L, Bornert M, Crépin J, Boller E, Sauder C, et al. In situ X-
ray microtomography characterization of damage in SiCf/SiC minicomposites.
Composites Science and Technology. 2011;71(6):916-24.
[109] Scott AE, Mavrogordato M, Wright P, Sinclair I, Spearing SM. In situ fibre
fracture measurement in carbon–epoxy laminates using high resolution computed
tomography. Composites Science and Technology. 2011;71(12):1471-7.
[110] Pandita SD, Falconet D, Verpoest I. Impact properties of weft knitted fabric
reinforced composites. Composites Science and Technology. 2002;62(7–8):1113-23.
[111] Wright P, Moffat A, Sinclair I, Spearing SM. High resolution tomographic
imaging and modelling of notch tip damage in a laminated composite. Composites
Science and Technology. 2010;70(10):1444-52.
[112] Rinaldi RG, Blacklock M, Bale H, Begley MR, Cox BN. Generating virtual
textile composite specimens using statistical data from micro-computed tomography:
3D tow representations. Journal of the Mechanics and Physics of Solids.
2012;60(8):1561-81.
[113] Blacklock M, Bale H, Begley M, Cox B. Generating virtual textile composite
specimens using statistical data from micro-computed tomography: 1D tow
Chapter 2 Literature review
62
representations for the Binary Model. Journal of the Mechanics and Physics of Solids.
2012;60(3):451-70.
[114] Ullah H, Harland A, Silberschmidt V. Experimental and Numerical Analysis of
Damage in Woven GFRP Composites Under Large-deflection Bending. Applied
Composite Materials. 2012;19(5):769-83.
[115] Sket F, Seltzer R, Molina-Aldareguía JM, Gonzalez C, Llorca J. Determination of
damage micromechanisms and fracture resistance of glass fiber/epoxy cross-ply
laminate by means of X-ray computed microtomography. Composites Science and
Technology. 2012;72(2):350-9.
[116] Kim KJ, Yu W-R, Lee JS, Gao L, Thostenson ET, Chou T-W, et al. Damage
characterization of 3D braided composites using carbon nanotube-based in situ sensing.
Composites Part A: Applied Science and Manufacturing. 2010;41(10):1531-7.
[117] Scott AE, Sinclair I, Spearing SM, Thionnet A, Bunsell AR. Damage
accumulation in a carbon/epoxy composite: Comparison between a multiscale model
and computed tomography experimental results. Composites Part A: Applied Science
and Manufacturing. 2012;43(9):1514-22.
[118] Djukic LP, Herszberg I, Walsh WR, Schoeppner GA, Gangadhara Prusty B, Kelly
DW. Contrast enhancement in visualisation of woven composite tow architecture using
a MicroCT Scanner. Part 1: Fabric coating and resin additives. Composites Part A:
Applied Science and Manufacturing. 2009;40(5):553-65.
[119] Amenabar I, Mendikute A, López-Arraiza A, Lizaranzu M, Aurrekoetxea J.
Comparison and analysis of non-destructive testing techniques suitable for delamination
inspection in wind turbine blades. Composites Part B: Engineering. 2011;42(5):1298-
305.
Chapter 2 Literature review
63
[120] Hufenbach W, Böhm R, Gude M, Berthel M, Hornig A, Ručevskis S, et al. A test
device for damage characterisation of composites based on in situ computed
tomography. Composites Science and Technology. 2012;72(12):1361-7.
[121] Bull DJ, Helfen L, Sinclair I, Spearing SM, Baumbach T. A comparison of multi-
scale 3D X-ray tomographic inspection techniques for assessing carbon fibre composite
impact damage. Composites Science and Technology. 2013;75(0):55-61.
[122] Faessel M, Delisée C, Bos F, Castéra P. 3D modelling of random cellulosic
fibrous networks based on X-ray tomography and image analysis. Composites Science
and Technology. 2005;65(13):1931-40.
[123] Little JE, Yuan X, Jones MI. Characterisation of voids in fibre reinforced
composite materials. NDT & E International. 2012;46(0):122-7.
[124] Kastner J PB, Salaberger D, Sekelja J. Defect and porosity determination of fibre
reinforced polymers by X-ray computed tomography. 2nd International Symposium on
NDT in Aerospace. 2010.
[125] Cnudde V, Boone MN. High-resolution X-ray computed tomography in
geosciences: A review of the current technology and applications. Earth-Science
Reviews. 2013;123(0):1-17.
Chapter 3 Deformation of single layer dry preforms under compaction
64
Chapter 3
DEFORMATION OF SINGLE LAYER DRY
PREFORMS UNDER COMPACTION
Z.Yousaf, P.Potluri, P.Withers
Abstract
Meso-scale geometrical changes during transverse compression of a dry textile preform
have a significant influence on resin permeability during infusion as well as on the
mechanical properties of the resulting polymer composite. Here tow deformations have
been analysed in detail by x-ray computed tomography (CT) for dry fabric preforms as a
function of compressive loading. With the aid of an in-situ loading device, 3D images
have been captured at each pressure increment up to 600 kPa, to simulate the forces
applied by a range of manufacturing processes including vacuum infusion and autoclave
curing. The evolutions of meso-scale geometrical features such as tow cross-section,
waviness, and inter-tow channels have been quantified. Good contrast has allowed the
geometry of the individual tows, as well as inter-tow channels to be described by 3D
solid objects. This technique, for the first time, facilitates the capture of detailed
preform geometry under compressive loading that can be used for resin flow
simulations as well as for structural analysis.
Keywords: compaction, tow geometry, woven textile, resin infiltration, inter-tow voids
Chapter 3 Deformation of single layer dry preforms under compaction
65
3.1 Introduction
Recent accelerated growth in the use of advanced composites is sustainable only
through continuous development of advanced manufacturing techniques matched by
improved design and simulation tools. Prepreg hand lay-up techniques developed
primarily for high-value markets such as military airframes and Formula 1 cars are
being gradually replaced by automated tape-laying and tow-placement techniques.
Further expansion to cost-sensitive high-volume markets, such as the passenger car
industry, will only be feasible through a significant reduction in manufacturing and
product development costs. Liquid resin infusion technologies in conjunction with dry
fibre preforming have been widely recognised as a way forward for achieving the next
level of affordable composites manufacturing. Dry textile preforms have meso-scale
features due to the interlacement of tows that affect the ease of resin infiltration and the
final properties of the manufactured composite. These meso-scale tow-level geometric
features need to be captured as a function of external loading so that they may be used
further for structural analysis and for making resin simulation tools.
For composite manufacturing, a number of different techniques are employed. These
manufacturing processes involve compaction of dry preform to a certain level of
pressure which changes the tow and the resin channel geometry of the preform affecting
the fibre volume fraction and the resin permeability. Based upon these facts, the
compaction process is taken as an important parameter of the manufacturing process
[1]. Extensive research has been performed on the compaction behaviour of different
fabrics [2-15]. It is considered that Van Wyk [3] was probably the first to treat the fibres
under compression as a system of bending units. Pearce et al. [11] performed
Chapter 3 Deformation of single layer dry preforms under compaction
66
experimental work on the compressibility of the reinforcement fabric under loading and
relaxing cycles, and attributed the loading cycle response to a power-law relationship. A
typical pressure - thickness curve was proposed by Hu et al. and Mastudaira et al. [16,
17], as shown in Figure 3.1. The curve can be divided into three regimes; namely two
linear stages separated by an exponential stage. Mastudaira et al. attributed the first
stage to the bending of the fibres, the second to friction between the fibres, and the third
part to the lateral compressional modulus of the fibres themselves. According to Chen et
al. [8], the third stage of the compression curve was due to bending deformation of the
yarn. A compaction model for a single layer woven fabric preform was developed by
Chen et al. [10]. In the compaction model, micro-deformation of the yarn cross-section
and macro-deformation by yarn bending accompanied by yarn waveform flattening
were described. In addition, the effects of single layer woven fabric microstructure on
the compaction behaviour were evaluated. It was shown that the macro-bending
stiffness of the fibre and the initial fibre packing ratio of the yarn affect the compaction
behaviour of single layer fabric.
Figure 3.1 Pressure - thickness curve for a woven fabric under compaction
Chapter 3 Deformation of single layer dry preforms under compaction
67
Saunders et al. [18] studied the compaction of multilayer plain woven cloths using
mechanical testing and microstructural images of the laminates. They divided the fabric
compression into three modes of deformation: the first mode being the nesting of cloths,
the second mode the deformation of the yarn waveform and the third mode as the
deformation of the yarn cross-section.
Microstructural analyses on multilayer woven preforms for different tow geometry
parameters under compaction were conducted by Saunders et al. and Kruckenberg et al.
[18-20]. They observed the nesting of the layers as the dominant factor in the initial
mode of compression for multilayer fabrics. Mahadik et al. [21-23] investigated the
effect of compaction on 3D woven fabric’s internal architecture and composite
compressive properties. They found the resin channels are highly influenced by the
weave style of the fabrics.
As already mentioned, for calculating the permeability of the reinforcements during
composite manufacturing, the study of the meso-structure of the reinforcement is
necessary. The meso-structure of the glass laminate for resin channels and inter-tow
voids was studied by Schell et al. [24] using computed tomography (CT). Saunders et
al. [19] studied the different glass fabrics under compression and provided data for
average area porosity, area pore structure and voids for different type of fabrics using
microstructure images of composites. The area pore distribution in cross-sections of
laminates was measured by Griffin et al. [25]
The deformation behaviour of single layer fabrics has been studied extensively using
mechanical methods while the microstructure of single layer fabrics has been analysed
only on composite laminates [12]. In the previous literature, no study has been reported
Chapter 3 Deformation of single layer dry preforms under compaction
68
on the meso-structure of single layer dry fabrics. The laminate structure may not
represent the meso-level features of dry fabric. This limitation has been pointed out by
Schell et al. [24]. Thus in order to validate tow geometry and permeability simulations,
the meso-structure analysis of dry fabric is necessary. In the present work, the
deformation behaviour of single layer dry preforms was investigated in-situ under
increasing compressive loading using CT. The changes in tow cross-section, yarn
waviness, yarn packing fraction, yarn volume fraction and resin channels were analysed
at different pressure levels. To carry out this study, the fabric preform was compressed
in a rig comprising two clear polycarbonate plates, and then scanned by CT for meso-
structural analysis.
3.2 Material and mechanical testing
The material under observations was E-glass plain woven fabric having a warp density
of 4.8 /cm and a weft density of 4.4 /cm. The warp and the weft counts were 650 Tex,
and the density of the fibres was 2.60 g/cm3. The initial unloaded thickness of the fabric
was 0.92 mm. The scanned image of the glass fabric used is presented in Figure 3.2.
Figure 3.2 Scanned image of plain weave (1/1) fabric
Chapter 3 Deformation of single layer dry preforms under compaction
69
3.2.1 Mechanical testing
An Instron 5569 machine with a 5 kN load cell was used for the mechanical testing of
the single layer fabric samples. The sample area was 15 cm2 and the cross-head speed
during testing was 1 mm/min. In glass yarn fibres tend to rearrange themselves under
load with the passage of time. Consequently, either the pressure decreases at constant
thickness or the thickness decreases at constant pressure when the fabric is kept under
load for long periods [15, 26-29]. For the mechanical testing of the fabric samples, a
static test method was used to maintain constant thickness at constant pressure. In this
method, the machine crosshead was moved to apply the desired constant pressure at a
constant speed of 1 mm/min and was held for 5 minutes at that point before moving to
the next position. The holding time of 5 minutes is considered adequate for these types
of loadings [20, 29]. At each target load, the initial and final thickness readings were
recorded. Prior to testing, the compliance curve was taken to adjust the minor
compliances during the compression. The compliance values were then removed from
the thickness readings, so that the compliance effect due to the machine was balanced.
3.3 Tomography and in-situ compression rig
A compression rig, shown in Figure 3.3, was developed to compress dry preforms in-
situ within the tomography machine. The rig comprises two clear polycarbonate plates
60 x 35 x 12 mm in length, width and thickness, respectively. Two side screws were used
to compress the plates from both sides, and two thickness gauges were placed on either
side to maintain uniformity on both sides. The edge-to-edge distance of the two side
screws was 35 mm, and the sample size of dry fabric compressed between the two plates
was 40 x 35 mm. At each pressure level, a slip gauge of known thickness was placed
Chapter 3 Deformation of single layer dry preforms under compaction
70
between the two plates on each side as illustrated in Figure 3.3. The bending strength of
the plates was estimated by applying beam bending theory to ensure that there was no
significant bending of the plates under load. Once mounted on the x-ray scanner, the
side screws were tightened to compress the fabric preform to the desired pressure. From
the pressure - thickness curve, the thickness value against desired pressure was taken
and the slip gauge of thickness corresponding to that pressure was put on both sides of
the plates. Figure 3.4 presents the rig mounted on the x-ray scanner.
Figure 3.3. Schematic of the compression rig
Tightening screw Top plate
Thickness gauge
Bottom plate Dry fabric Nut
Chapter 3 Deformation of single layer dry preforms under compaction
71
Figure 3.4. (a) Compression rig accommodated on the tomography stage, and (b)
close-up of the rig
The CT method involves collecting a large set of radiographs (projections) of the
sample as it is rotated through 180°. Together with a small number of calibration
images, these images are reconstructed into a 3D volume, which represents the
attenuation through the sample. A Nikon Metris 225/320 kV Custom Bay system was
used for scanning. The current and voltage were adjusted to 115 µA and 85 kV,
respectively. For each scan, a total of 3142 projections were taken. These images were
then reconstructed using Metris X-Tek CT Pro software, and the data processed using
VSG Avizo Fire 6.3 software.
Chapter 3 Deformation of single layer dry preforms under compaction
72
3.4 Results and discussion
3.4.1 Macroscopic deformation
The compaction of the single layer fabric was performed in both dry and wet conditions.
For wet compaction the fabric was impregnated with water and was compacted to
measure the pressure - thickness response. The thickness results for dry and the wet
fabrics at different pressures are shown in Figure 3.5. Two curves were recorded for
each fabric, one represents the thickness immediately after the load is applied, and the
other is the thickness after 5 minutes of holding at that loading point. From Figure 3.5, it
can be seen that the compaction of the wet fabric was higher than the dry fabric at the
same pressures which is in agreement with the previous findings [30, 31]. Significant
compaction occurs over time in both cases at constant load, the change having stabilised
after 5 minutes. It is clear from Figure 3.6 that the displacement has essentially
plateaued within 5 minutes of loading for both wet and dry preforms. Unsurprisingly,
the change of thickness with time was greatest during the early stages of compaction;
this reflects the fact that the fibres relax less at the higher loads.
Chapter 3 Deformation of single layer dry preforms under compaction
73
Figure 3.5. Pressure - thickness response for a single layer under dry and wet
conditions
Figure 3. 6. Effect of loading time on the thickness of single layer dry fabric
The thickness curves with time at particular pressures as illustrated in Figure 3.7 were
fitted with a power-law of the form shown in Equation 3.1:
(3.1)
Chapter 3 Deformation of single layer dry preforms under compaction
74
Where t/t0 is the final thickness to initial thickness ratio, ‘T’ is the time and ‘a’ and ‘c’
are the empirical constants.
Figure 3.7. t/t0 plotted against time for different pressures
The constants ‘a’ and ‘c’ can further be varied against different pressures and it was
found that ‘a’ takes the trend of a power law with pressure while ‘c’ is almost constant.
The experimental and fitted curves of constant ‘a’ against pressure are presented in
Figure 3.8.
Chapter 3 Deformation of single layer dry preforms under compaction
75
Figure 3.8. The constants plotted against pressure
The fitted power follows the form shown in Equation 3.2 for ‘a’ in the above equation
(3.2)
So the final relationship between t/to with pressure and time becomes that of Equation
3.3
(3.3)
The term ‘q’ was found to vary with the wettability of the fabric.
Figure 3.9 shows the experimental and fitted values of a single layer of dry and wet
fabrics at different pressures, which are in close agreement with each other.
Chapter 3 Deformation of single layer dry preforms under compaction
76
(a)
(b)
Figure 3.9. Experimental and predicted thicknesses with time at different
pressures, (a) dry fabric and (b) wet fabric
Chapter 3 Deformation of single layer dry preforms under compaction
77
3.4.2 Meso-structural analysis by computed tomography (CT)
Single layer dry fabric compressed between polycarbonate plates was scanned by CT
for its meso-structural analysis at different pressures. The parameters describing the
yarn geometry obtained from the meso-structural analysis are illustrated in Figure 3.10.
The thread spacing, P, is the distance between two consecutive yarns, the crimp
amplitude is represented by h, the crimp angle by θ, the yarn width by a, and yarn
thickness by b. The crimp percentage of the yarn was calculated by using the equation
3.4.
(3.4)
Figure 3.10. Definition of the yarn geometry parameters
The tomography technique has the advantage that the scanned structure can be
visualised over any region of interest. To investigate the behaviour of the fabric, both
the warp and the weft yarns have been extracted as virtual slices from the centre of the
tow intersections and in between two tows to study the tow at the interlacing and non-
interlacing points for both tow geometry and the void-space data as shown in Figure
3.11.
Chapter 3 Deformation of single layer dry preforms under compaction
78
Figure 3.11. Segmented virtual slices of warp and weft yarns
Figure 3.12 represents the 3D structure of the fabric sample, segmented using Avizo 6.3
on the basis of simple thresholding.
Figure 3.12. 3D reconstruction of single layer fabric
In between two tows Centre of tow
Chapter 3 Deformation of single layer dry preforms under compaction
79
Table 3.1. Tow geometry parameters (warp).
Pressure
(kPa)
Tow
area
(mm2)
Tow
width
(mm)
Tow
height
(mm)
Thread
spacing
(mm)
10 0.47 2.07 0.31 2.03
50 0.46 2.11 0.30 2.01
100 0.44 2.11 0.28 2.03
300 0.41 2.12 0.26 2.02
600 0.38 2.12 0.24 2.03
Table 3.2. Tow geometry parameters (weft).
Pressure
(kPa)
Tow
area
(mm2)
Tow
width
(mm)
Tow
height
(mm)
Thread
spacing
(mm)
10 0.45 1.78 0.40 2.25
50 0.44 1.77 0.39 2.26
100 0.42 1.78 0.37 2.26
300 0.39 1.79 0.34 2.25
600 0.37 1.79 0.31 2.24
Chapter 3 Deformation of single layer dry preforms under compaction
80
Tomographic slices were taken at the centre of the intersections of the warp and the
weft yarns, and various tow geometry parameters of both the warp and the weft yarns
were calculated at different pressure levels as shown in Tables 3.1 and 3.2. The
corresponding cross-sectional images of the warp and the weft yarns are displayed in
Figure 3.13. The range of pressures studied varied from 10 kPa to 600 kPa to investigate
the behaviour of the geometrical changes for both vacuum infusion and autoclave
curing. During the meso-structural analysis of the compressed dry fabric preform, it was
observed that initially, the shape of the warp yarn was elliptical and the weft yarn was
lenticular. The warp yarn was highly crimped as shown in Figure 3.14a and the yarn
followed a sinusoidal path, whereas the weft yarn exhibited very low crimp as shown in
Figure 3.14b and the yarn path was nearly straight. The warp tows were in contact with
the compression plate whereas there was separation between the weft tows and the
compression plate at the initial loading. Warp tows were spreading to the available gaps
between the tows and there was a slight overlap between the warp tows at the edges of
the cross-sections .Due to the higher warp yarn crimp percentage, the crimp amplitude
and the crimp angle were also higher in the warp yarns than the weft yarns.
During the first step of compression the amplitude of the warp yarns reduced as the
pressure increased from 10 kPa to 50 kPa; by contrast, the amplitude of the weft yarns
increased as seen in Figure 3.15. Similarly, both the crimp and the crimp angle reduced
for the warp yarns and increased for the weft yarns with increasing the pressure as in
Figures 3.16 to 3.17. It should be noted that the yarn crimp reduces the compressive
strength of the laminates [32] and the magnitude of the crimp angle is important in
predicting the in-plane compressive strength of the composite. The phenomenon of
crimp behaviour may be attributed to the balancing of the warp and the weft yarns in
Chapter 3 Deformation of single layer dry preforms under compaction
81
which the crimp in one yarn increases and the other yarn decreases to reach the
balancing position. This balancing of crimp was discussed in detail by Potluri et al. and
Lomov et al. [33, 34].This crimp interchange is an important phenomenon which affects
the laminate tensile moduli [35]. It is worth mentioning that the other geometrical
parameters (area, height) of dry fabric were only slightly changed by increasing the
pressure.
Weft cross-section Warp cross-section
Figure 3.13. Cross-sectional view of the preform at the centre of the tow
intersections
Figure 3.14. 3D reconstruction of yarns, (a) warp yarn, (b) weft yarn
Chapter 3 Deformation of single layer dry preforms under compaction
82
Figure 3.15. The crimp amplitude of the warp and the weft yarns
Figure 3.16. The crimp percentage of warp and weft yarns
Chapter 3 Deformation of single layer dry preforms under compaction
83
Figure 3.17. The crimp angle of warp and weft yarns
When the pressure was increased from 50 kPa to 100 kPa, the same phenomenon of
crimp balancing continued in which the crimp percentage, the crimp amplitude, and the
crimp angle decreased in the warp yarns and increased in the weft yarns. In all cases, the
change in the yarn height and area was not significant. Figure 3.15 indicates that for the
regime of pressure up to 100 kPa corresponding to vacuum infusion, compaction is
mainly dominated by the crimp amplitude reduction. In this region, the main factor
contributing to fabric thickness reduction was also identified, which may return to the
crimp break of warp yarns for the fabric. A similar behaviour was observed by Potluri et
al. [12] for the single layer plain woven fabric laminate in which crimp reduction was
responsible for the compression at low loads. This observation shows that the thickness
reduction in the low load regime for single layer dry fabric is different from the
multilayer laminate in which the nesting of the layers is the main factor contributing to
the thickness reduction of the laminate as observed by Saunders et al. and Kruckenberg
et al. [18, 20].
Chapter 3 Deformation of single layer dry preforms under compaction
84
In cases where higher pressures (300 kPa to 600 kPa) were applied, the further reduction
in yarn amplitude and yarn crimp was insignificant, and the deformation in the tow
occurred at the current pressures as in Figures 3.15 to 3.17 and Tables 3.1& 3.2. In
contrast to the lower pressures, there was greater reduction in yarn height and yarn area.
Additionally, a slight increase was noticed in the yarn width as the pressure changed to
600 kPa. Contrary to the initial pressures, the weft yarn crimp percentage, crimp
amplitude and crimp angle were decreased at a pressure of 600 kPa as seen in Figures
3.15 to 3.17. The observation of decrease in yarn area is in contrast to the assumption of
the model made by Chen et al. [8] in which yarn shape deforms but the yarn area
remains the same. These findings indicate that after the yarn attains the balancing
condition, the crimp in both types of crossing yarns starts to decrease with the
application of further load. It should be pointed out that no significant trend was
observed for the thread spacing with the pressure increasing. The overall behaviour of
the fabric under compaction shows that applying low pressures at the start, the single
layer thickness reduction is maximised which is due to crimp interchange and at higher
pressures the thickness reduction is due to yarn deformation. This is also in agreement
with the pressure thickness curve in Figure 3.1 where a small pressure is required in the
initial mode of compression and then a higher amount of force is required even for
small amounts of thickness reduction.
Chapter 3 Deformation of single layer dry preforms under compaction
85
Weft cross-section Warp cross-section
Figure 3.18. Cross-sectional images of the warp and the weft yarns in between two
tows
In addition to the centre of the interlacing points, tow cross-sections were also analysed
in between two consecutive warp and weft yarns. For this purpose, yarn slices were
taken in between two tows using Avizo software and the change in the corresponding
tow area was investigated at each pressure level. Figure 3.18 shows the images taken in
between the two tows. In the warp cross-sections only warp tows are visible whereas in
the weft cross-sections both the warp and the weft tow are visible. The reason for this
was when the tows cross-sections were taken for the warp cross-sections in between two
weft tows, due to the higher weft spacing only warp cross-sections were present at the
slices and when the weft cross-sections were taken in between the two warp yarns, the
warp yarns were overlapping on the edges and on the slices they also contributed with
the weft yarns.
As tows in between warp and weft yarns were not straight due to the crimp angle, so
when slices were taken at the centre point they gave different area values due to
additional crimp angle. As shown in Figure 3.19, the slice taken in between two tows
gave a value of bd instead of ae due to the crimp angle θ1. This crimp effect was
Chapter 3 Deformation of single layer dry preforms under compaction
86
eliminated by using the trigonometric cosine function and the actual value of tow cross-
section area ( ae ) at the central point was measured.
Figure 3.19. Tow cross-section at the centre due to crimp angle
It was observed that the tow area in between the two tows was higher than the
corresponding area at the centre of the crossing point of both the warp and the weft
yarns. This may be attributed to the fact that the forming forces acting on the crossing
points are higher than those in between the two tows. This area in between the two tows
changed negligibly during the application of initial loads and decreased slightly as
pressure approached 600 kPa, giving room for more intra-tow voids and, consequently,
a lower yarn packing fraction, compared to that at the crossing point of the yarns.
The tow area calculated from the CT images was compared with the tow area calculated
by the Texgen software developed by The University of Nottingham; the figures appear
in Table 3.3. It was observed that the area calculated by Texgen and tomography differ
from each other; the reason might be that the Texgen software calculates the ideal tow
Chapter 3 Deformation of single layer dry preforms under compaction
87
area whilst due to variations in the manufactured yarns, the area calculated from the CT
images is different from the Texgen area.
Table 3.3 Tow area calculated by CT and Texgen
Area (mm2)
Pressure(kPa)
By CT
By Texgen
10
0.47
0.50
50
0.46
0.50
100
0.44
0.48
300
0.41
0.44
600
0.37
0.40
3.4.3 Resin channels in dry preform
The compaction of the fabrics leads to a change in the fabric permeability [36]. This is
taken to arise because of the reduction in the inter-tow voids or the resin channels; while
the intra-tow voids are believed to exert a negligible effect [37]. Because of the
importance of the fabric permeability, several models have been developed for both
single and multi-layer fabrics [2, 37-39].
Studying the resin channels or inter-tow voids in dry fabrics is significant in developing
accurate permeability models. In the current work, the resin channels were examined in
detail at different pressure levels, and the corresponding behaviour analysed.
Volumetric as well as areal channels were analysed for both warp and weft cross-
sections, as a function of the applied pressure.
Chapter 3 Deformation of single layer dry preforms under compaction
88
As seen from Figure 3.20, the inter-tow void percentage decreases exponentially with
increasing the pressure from 10 kPa to 600 kPa. The reduction is steep over the pressure
range 10-100 kPa, which corresponds to the vacuum infusion region; while in the
pressure range 300-600 kPa, the reduction was insignificant. Perhaps surprisingly, the
void percentage was markedly higher in between two tows than at the centre of the
intersections of the warp and the weft cross-sections as illustrated in Figures 3.21 &
3.22, giving larger passages for resin flow in between two tows rather than at the centre
of the intersecting point of the tows. Further, the void percentage was higher for the
weft cross-section at the centre of the tow intersections than that in the warp cross-
section. This might be due to the higher number of warp yarns than weft yarns.
However, the reverse behaviour was seen for the void percentage in between the tow
where the void percentage was larger in magnitude in the warp cross-section compared
with the weft cross-section, suggesting better resin transfer in the warp cross-section
than in weft cross-section. The reason for these larger inter-tow voids in between the
tows in the warp cross-section was that when the slices were taken in between two weft
tows in the warp cross-section, only warp yarns were present there due to the greater
yarn spacing between the weft yarns while on the other side, when the yarn slices were
taken for the weft cross-sections in between two warp yarns, the weft slices at that point
also contained warp yarns which were overlapping at the edges of each other.
Chapter 3 Deformation of single layer dry preforms under compaction
89
Figure 3.20. Inter-tow voids in 3D structure
Figure 3.21. Inter-tow voids at the centre of the tow intersections
Chapter 3 Deformation of single layer dry preforms under compaction
90
Figure 3.22. Inter-tow voids in between two tows
The thickness of the inter-tow voids and its full distribution were measured at different
pressure levels as shown in Figures 3.23 & 3.24. Similarly to the void percentage-
pressure relationship, the void thickness decreased exponentially with increasing
applied pressure as seen in Figure 3.23. It was found in the full distribution of the resin
channel heights that there were two peaks for each curve as shown by Figure 3.24. From
50 kPa to 600 kPa, the intensity of the peak at low thickness values increased whereas
the intensity of the peak at higher thickness values decreased. The evaluation of the
curves is consistent from 50 kPa up to 600 kPa; however, the 10 kPa curve is singular:
the peak at lower thickness values is larger than that at higher thickness values. Also,
the first peak is at the lowest thickness value. This needs to be investigated further. Also
the thickness of the resin channels was found to be different at the centre of the tow
cross-sections and in between tows.
Chapter 3 Deformation of single layer dry preforms under compaction
91
Figure 3.23. Average void thickness with pressure
Figure 3.24. The frequency distribution of the void thickness
From the above observations it is clear that the macro-voids not only differed in the
warp and weft cross-sections but also in the centre of the tow and in between two tows
because of the low cover factor.
Figure 3.25a shows the 3D reconstructed image of the compressed fabric sample with
the resin channels segmented separately in the Avizo software using the threshold
Chapter 3 Deformation of single layer dry preforms under compaction
92
method. The volumetric resin channels for the compressed dry fabric preform at 10 kPa
and 600 kPa are pictorially represented in Figures 3.25b and 3.25c, respectively. The
resin channels were also observed for the connectivity of the pores using the threshold
method and it was seen that these channels were connected at each pressure level. At the
higher pressure level of 600 kPa there were thin connections of area even up to 0.01
mm2 at the centre of the tow giving the least passage to resin flow in those areas.
(a)
(b)
Chapter 3 Deformation of single layer dry preforms under compaction
93
(c)
Figure 3.25. Resin channels in dry preform during compression: (a) dry preform
with resin channels, (b) channels at 300 kPa, and (c) channels at 600 kPa
3.4.4 Yarn packing fraction
The packing fraction of the yarns (YPF) was calculated by using equation 3.5
(3.5)
In equation 3.5 ‘Vf’ is the volume of the fibres in the yarn, and ‘Vy’ is the total volume
of the yarns (fibres plus air) in the stack. Herein, the yarn volume (Vy) was calculated
with the help of the Avizo software in which the yarns were selected using the threshold
method and the volume was measured using the material statistics tool in the Avizo
software. Whereas, the fibre volume (Vf) was calculated from the density of the fibre
given in the material section.
Chapter 3 Deformation of single layer dry preforms under compaction
94
Figure 3.26. Yarn packing fractions of dry preform against different loads
Figure 3.26 represents the yarn packing fractions at various pressure levels. It can be
seen from Figure 3.26 that there was only a slight change in the packing fraction when
the applied pressure increased from 10 kPa to 100 kPa. However, in case where the
applied pressure increased from 100 kPa to 600 kPa, the packing fraction increased
gradually. Figure 3.27 depicts an increase in the packing fraction in the cross-sectional
area with the application of pressure.
Figure 3.27. Yarn packing fraction on application of pressure
Also when the yarn packing fractions were studied in the centre of the tows and in
between two tows, it was observed that the change in yarn packing fractions with
pressure was insignificant in between the tows for both warp and weft yarns compared
After compression
Chapter 3 Deformation of single layer dry preforms under compaction
95
to the packing fraction in the centre of the tows for which data are collected in Tables
3.4 & 3.5.
Table 3.4. Yarn packing fraction of warp cross-sections
Packing fraction
Pressure(kPa) Centre of tow In between tows
10 0.53 0.53
50 0.54 0.53
100 0.57 0.53
300 0.61 0.54
600 0.66 0.56
Table 3.5. Yarn packing fractions of weft cross-sections
Packing fraction
Pressure(kPa) Centre of tow In between tows
10 0.56 0.56
50 0.57 0.56
100 0.6 0.56
300 0.64 0.57
600 0.68 0.57
3.4.5 Yarn and fibre volume fraction of the preform
The yarn volume fraction of dry preform was calculated for the warp and the weft cross-
sections at the centre of the intersections as seen in Figure 3.28 and in between two tows
as seen in Figure 3.29.
Chapter 3 Deformation of single layer dry preforms under compaction
96
Figure 3.28. Yarn volume fractions calculated at the centre of the tow intersections
Figure 3.29. Yarn volume fractions calculated in between two tows
It was seen that the yarn volume fractions in the slices taken at the centre of the tow
intersections were much higher than those taken in between the two tows for both the
warp and the weft yarns and these are illustrated in Figures 3.28 & 3.29 giving more
passage to resin flow in between the two tows than at the centre of the tow cross-
Chapter 3 Deformation of single layer dry preforms under compaction
97
sections. In the slices at the centre of the tows, the volume fraction for the warp yarns
was higher than that at the centre of the weft yarns and this is shown in Figure 3.28.
Whereas, in the slices in between the two tow, the opposite trend was observed and may
be seen in Figure 3.29. This phenomenon again relates to the point that the warp yarn
density was higher than the weft yarn so a higher yarn volume fraction was seen in the
warp direction than that of the weft direction in the centre of the tow while due to
overlapping of the warp yarns on the edges; the weft yarn slices contained some
portions of warp yarns at that point giving a greater yarn volume fraction in the slices
taken in between two warp tows in the weft cross-section.
Also, the fibre volume fraction was calculated and compared using both tomographic
and mechanical test results. From the tomographic results the fibre volume fraction
(FVF) was calculated using equation 3.6.
(3.6)
Where: YPF is the yarn packing fraction and YVF is the yarn volume fraction.
Whereas for the fibre volume fractions from mechanical test results, equation 3.7 was
employed.
(3.7)
Here in equation 3.7 ‘ ’ is the number of fabric layers; is the areal weight of the
fabric, the density of the fibre and ‘ ’ is the thickness of the fabric under
compression.
Chapter 3 Deformation of single layer dry preforms under compaction
98
Figure 3.30. FVF calculated by mechanical testing and tomographic analysis
The fibre volume fraction results calculated both from mechanical testing and
tomographic analysis are depicted graphically in Figure 3.30. Close agreement was
observed between the FVF calculated by mechanical test result and the tomography
analysis confirming the accuracy of the imaging/segmentation process. So with good
contrast exact tow geometry can be achieved and FVF and improved geometry models
can also be exported to Abaqus and other analysis software using CT.
3.5 Conclusions
In this work, the fabric tow cross-sectional changes and the geometry of the resin
channels of a single layer dry fabric are followed in-situ by X-ray tomography during
increasing compressive loading. Pressures ranging from 10 kPa to 600 kPa were applied.
It was seen that for pressures up to 100 kPa, the reduction in fabric thickness was
mainly due to reducing crimp and crimp amplitude. In this regime, the crimp balancing
Chapter 3 Deformation of single layer dry preforms under compaction
99
position was initially attained by a decrease in the crimp of the warp yarns, followed by
an increase in the crimp for the weft yarns. The reduction in the resin channels was also
prominent in this pressure range of 10-100 kPa, whereas, yarn deformation and packing
fraction were not significant contributors. At higher pressures (300 kPa to 600 kPa) yarn
deformation became prominent, and contribution from the reduction in crimp and crimp
amplitude exhausted. In addition, the macro-void percentage in between two tows was
much higher than that at the centre of the crossing of both the warp and the weft yarns.
Additionally, CT was seen as an accurate tool to find the fibre volume fractions.
Acknowledgements
The authors are grateful to the staff of the Henry Moseley X-ray Imaging Facility which
was funded by the EPSRC under EP/F007906, EP/I02249X, EP/J021229/1 with
additional support from the University of Manchester;
References
[1] Robitaille F, Gauvin R. Compaction of textile reinforcements for composites
manufacturing. I: Review of experimental results. Polymer Composites.
1998;19(2):198-216.
[2] Nguyen QT, Vidal-Sallé E, Boisse P, Park CH, Saouab A, Bréard J, et al.
Mesoscopic scale analyses of textile composite reinforcement compaction. Composites
Part B: Engineering. 2013;44(1):231-41.
[3] Van Wyk CM. Note on the compressibility of wool. Journal of the Textile Institute
1946;37(12):T285-92.
Chapter 3 Deformation of single layer dry preforms under compaction
100
[4] Gutowski TG, Dillon G. The elastic deformation of lubricated carbon fiber bundles:
Comparison of theory and experiments. Journal of Composite Materials.
1992;26(16):2330-47.
[5] De Jong S, Snaith JW, Michie NA. A mechanical model for the lateral compression
of woven fabrics. Textile Research Journal. 1986;56(12):759-67.
[6] Karbhari VM, Simacek P. Notes on the modeling of preform compaction: II-Effect
of sizing on bundle level micromechanics. Journal of Reinforced Plastics and
Composites. 1996;15(8):837-61.
[7] Simacek P, Karbhari VM. Notes on the modeling of preform compaction: I -
Micromechanics at the fiber bundle level. Journal of Reinforced Plastics and
Composites. 1996;15(1):86-122.
[8] Chen B, Chou T-W. Compaction of woven-fabric preforms in liquid composite
molding processes: single-layer deformation. Composites Science and Technology.
1999;59(10):1519-26.
[9] Chen B, Chou T-W. Compaction of woven-fabric preforms: nesting and multi-layer
deformation. Composites Science and Technology. 2000;60(12-13):2223-31.
[10] Chen Z-R, Ye L, Kruckenberg T. A micromechanical compaction model for woven
fabric preforms. Part I: Single layer. Composites Science and Technology.
2006;66(16):3254-62.
[11] Pearce N, Summerscales J. The compressibility of a reinforcement fabric.
Composites Manufacturing. 1995;6(1):15-21.
[12] Potluri P, Wilding MA, Memon A. A novel stress-freezing technique for studying
the compressional behavior of woven fabrics. Textile Research Journal.
2002;72(12):1073-8.
Chapter 3 Deformation of single layer dry preforms under compaction
101
[13] Chen B, Cheng AHD, Chou TW. A nonlinear compaction model for fibrous
preforms. Composites Part A: Applied Science and Manufacturing. 2001;32(5):701-7.
[14] Chen B, Lang EJ, Chou T-W. Experimental and theoretical studies of fabric
compaction behavior in resin transfer molding. Materials Science and Engineering A.
2001;317(1-2):188-96.
[15] Bickerton S, Buntain MJ, Somashekar AA. The viscoelastic compression behavior
of liquid composite molding preforms. Composites Part A: Applied Science and
Manufacturing. 2003;34(5):431-44.
[16] Hu J, Newton A. Low-load lateral-compression behaviour of woven fabrics.
Journal of the Textile Institute. 1997;88(3):242-54.
[17] Matsudaira M, Qin H. Features and mechanical parameters of a fabric's
compressional property. Journal of the Textile Institute. 1995;86(3):476-86.
[18] Saunders RA, Lekakou C, Bader MG. Compression and microstructure of fibre
plain woven cloths in the processing of polymer composites. Composites Part A:
Applied Science and Manufacturing. 1998;29(4):443-54.
[19] Saunders RA, Lekakou C, Bader MG. Compression in the processing of polymer
composites 1. A mechanical and microstructural study for different glass fabrics and
resins. Composites Science and Technology. 1999;59(7):983-93.
[20] Kruckenberg T, Ye L, Paton R. Static and vibration compaction and microstructure
analysis on plain-woven textile fabrics. Composites Part A: Applied Science and
Manufacturing. 2008;39(3):488-502.
[21] Mahadik Y, Hallett SR. Finite element modelling of tow geometry in 3D woven
fabrics. Composites Part A: Applied Science and Manufacturing.41(9):1192-200.
Chapter 3 Deformation of single layer dry preforms under compaction
102
[22] Mahadik Y, Hallett SR. Effect of fabric compaction and yarn waviness on 3D
woven composite compressive properties. Composites Part A: Applied Science and
Manufacturing.42(11):1592-600.
[23] Mahadik Y, Brown KAR, Hallett SR. Characterisation of 3D woven composite
internal architecture and effect of compaction. Composites Part A: Applied Science and
Manufacturing.41(7):872-80.
[24] Schell JSU, Renggli M, Van Lenthe GH, Müller R, Ermanni P. Micro-computed
tomography determination of glass fibre reinforced polymer meso-structure.
Composites Science and Technology. 2006;66(13):2016-22.
[25] Griffin PR, Grove SM, Russell P, Short D, Summerscales J, Guild FJ, et al. The
effect of reinforcement architecture on the long-range flow in fibrous reinforcements.
Composites Manufacturing. 1995;6(3–4):221-35.
[26] Seung Hwan L, Jeong Hwa H, Seong Yun K, Jae Ryoun Y, Young Seok S.
Compression and relaxation behavior of dry fiber preforms for resin transfer molding.
Journal of Composite Materials.44(15):1801-20.
[27] Robitaille F, Gauvin R. Compaction of textile reinforcements for composites
manufacturing. II: Compaction and relaxation of dry and H2O-saturated woven
reinforcements. Polymer Composites. 1998;19(5):543-57.
[28] Kelly PA, Umer R, Bickerton S. Viscoelastic response of dry and wet fibrous
materials during infusion processes. Composites Part A: Applied Science and
Manufacturing. 2006;37(6):868-3.
[29] Kim YR, McCarthy SP, Fanucci JP. Compressibility and relaxation of fiber
reinforcements during composite processing. Polymer Composites. 1991;12(1):13-9.
Chapter 3 Deformation of single layer dry preforms under compaction
103
[30] Arbter R, Beraud JM, Binetruy C, Bizet L, Bréard J, Comas-Cardona S, et al.
Experimental determination of the permeability of textiles: A benchmark exercise.
Composites Part A: Applied Science and Manufacturing. 2011;42(9):1157-68.
[31] Kelly PA, Umer R, Bickerton S. Viscoelastic response of dry and wet fibrous
materials during infusion processes. Composites Part A: Applied Science and
Manufacturing. 2006;37(6):868-73.
[32] Yang B, Kozey V, Adanur S, Kumar S. Bending, compression, and shear behavior
of woven glass fiber–epoxy composites. Composites Part B: Engineering.
2000;31(8):715-21.
[33] Lomov SV, Verpoest I. Compression of woven reinforcements: A mathematical
model. Journal of Reinforced Plastics and Composites. 2000;19(16):1329-50.
[34] Potluri P, Sagar TV. Compaction modelling of textile preforms for composite
structures. Composite Structures. 2008;86(1-3):177-85.
[35] Potluri P, Parlak I, Ramgulam R, Sagar TV. Analysis of tow deformations in textile
preforms subjected to forming forces. Composites Science and Technology.
2006;66(2):297-305.
[36] Grujicic M, Chittajallu KM, Walsh S. Effect of shear, compaction and nesting on
permeability of the orthogonal plain-weave fabric preforms. Materials Chemistry and
Physics. 2004;86(2-3):358-69.
[37] Yu B, James Lee L. A simplified in-plane permeability model for textile fabrics.
Polymer Composites. 2000;21(5):660-85.
[38] Simacek P, Advani SG. Permeability model for a woven fabric. Polymer
Composites. 1996;17(6):887-99.
Chapter 3 Deformation of single layer dry preforms under compaction
104
[39] Badel P, Vidal-Sallé E, Maire E, Boisse P. Simulation and tomography analysis of
textile composite reinforcement deformation at the mesoscopic scale. Composites
Science and Technology. 2008;68(12):2433-40.
Chapter 4 Deformation of multilayer dry preforms under compaction
105
Chapter 4
DEFORMATION OF MULTILAYER DRY
PREFORMS UNDER COMPACTION
Z.Yousaf, P.Potluri, P.Withers
Abstract
Preform compression is an important mechanism during the resin infusion of the
composite manufacturing process. Here, the analysis of tow cross-section, tow
waviness, nesting of the layers and inter-tow voids in a 2D multilayer woven preform
under compressive loading in-situ was done using computed X-ray tomography (CT).
Low pressures up to 100 kPa, being representative of Vacuum Infusion (VI) were
applied to dry preform and changes in tow cross-section, yarn waviness and inter-tow
voids were observed. Nesting between individual layers was calculated and nesting
factors calculated by mechanical testing and tomographic images were compared. Yarns
packing fractions of dry preform were also calculated for this study. It was seen that the
inter-tow voids reduced drastically for this low load regime and the nesting of the layers
was observed as major contributor to the stack thickness reduction for pressures up to
100 kPa.
Keywords: dry preform, inter-tow voids, nesting factor, computed tomography (CT)
Chapter 4 Deformation of multilayer dry preforms under compaction
106
4.1 Introduction
The vacuum infusion process (VI) is becoming a promising manufacturing technique in
both automotive and aerospace industry for manufacturing composite parts due to its
low cost, better fibre to resin ratio and environmental benefits. A discussion about the
VI technique can be found in the literature [1]This manufacturing process involves the
compaction of the preform to a certain level of pressure. The compaction of the preform
not only affects the tow geometry of the yarn but also changes the permeability and
mechanical properties of the final product and is taken as an important parameter of the
manufacturing process [2]
Since, in most of the practical applications the use of more than one layer is needed so
the study of the compaction of multilayer stacks of preforms is significant. The
compression behaviour of different fabrics has been studied by a number of researchers
[3-15]. Van Wyk [3] was probably the first to consider fibres under compression as a
system of bending units. Gutowski at el. [4] proposed a theory of elastic deformation of
carbon fibre bundles and the compaction of the fibre bundles was attributed to the
bending of the curved filaments. De Jong et al. [5] derived a mechanical model from
Van Wyk’s compression law to describe a relationship between pressure and thickness
of woven fabrics. Pearce et al. [11] performed experimental work on the
compressibility of the reinforcement fabric under loading and relaxing cycles. They
fitted the loading cycle response to the power law relationship and the relaxation cycles
to an exponential decay functions. Saunders at el. [16] made a mechanical and micro
structural study of different fabrics under compression and provided data of the average
area porosity, area pore structure and voids of the fabric. A detailed study on the
Chapter 4 Deformation of multilayer dry preforms under compaction
107
internal architecture, tow geometry and effect of yarn waviness on the mechanical
properties for the 3D woven angle interlock fabric under compaction was made by
Mahadik et al. [17-19]. A typical pressure thickness curve was given by Hu et al. and
Mastudaira et al. [20, 21] as shown in Figure 4.1. The curve can be divided into three
parts, two linear parts and one exponential part. Mastudaira et al. attributed the first part
of the curve to the bending of the fibres, the second part with the hardness in
compression because of the friction between the fibres and the third part of the curve to
the lateral compressional modulus of the fibres itself. According to Chen et al. [8] the
preform compaction in the initial mode is due to the reduction of pores and gaps among
the fibres and yarns; while in the third mode it is due to the bending deformation of
yarns. Saunders et al. [22] studied the compaction of plain woven cloth using
mechanical testing and micro structural images of the compressed resin impregnated
samples. They divided the fabric compression into three modes of deformation, the first
mode as nesting of cloths, the second mode as deformation of the yarn waveform and
the third mode as the deformation of the yarn cross- section.
Figure 4.1. Typical pressure thickness curve for a woven fabric under compaction
The compaction behaviour of a multilayer preform is different from a single layer
preform due to shifting and nesting between the individual layers. The nesting between
Chapter 4 Deformation of multilayer dry preforms under compaction
108
individual layers plays an important role on the compressibility and permeability of the
reinforcement. It also effects the fibre volume fraction and thickness of the preform [9].
The effect of nesting was also studied with regards to the permeability of the woven
fabrics [23-25]. It was shown by Hoes et al. [26] that nesting is a major source of
variation in experimentally determined permeabilites. Keeping in view these points,
study of the nesting phenomenon in textile preforms during compaction is essential to
predict the behaviour of the composites. Yurgartis et al. [27] developed methods of
quantifying yarn shape and nesting using the inclination angle, crimp angle and angle
match. Chen et al. [9] developed an analytical methodology for predicting the elastic
deformation and nesting of multilayer woven fabric preforms under compaction using
their 3D model. Rong et al. [28] developed a micromechanical compaction model for
multilayer woven fabric preforms in which both the shifting and the number of layers
were shown as having significant effect on the compaction behaviour. Also, an increase
in nesting was seen with an increase in the number of fabric layers. Lomov et al. [29]
made a 3D geometrical model to study the nesting of the reinforcement in textile
laminates. In this model the nesting was associated with the float length and tightness of
the weave. According to this model the nesting will increase with an increase in yarn
spacing and with a decrease in float length of the weave. Teresa et al. [30] analysed the
resin impregnated plain weave fabric samples after static and vibration compaction and
studied the degree of nesting along with different tow geometry parameters. They also
observed a decrease in nesting with a decrease of yarn spacing.
In the previous research work, multilayer fabric compression has been studied on the
macro-level using mechanical methods for both dry and wet samples but the meso-
structure of multilayer fabrics has only been investigated for the laminated composite
Chapter 4 Deformation of multilayer dry preforms under compaction
109
samples. No work has yet been reported on the meso-structure of dry fabrics under
compressive loading. As dry and wet fabrics have different compaction behaviour [31-
33], so the study of dry fabric for the meso-structure is important to predict the
structural changes during compression and for making accurate simulation tools. This
research work concentrates on the analysis of deformation behaviour of dry fabric
samples at macro and meso-scale using mechanical and geometrical analysis of the
fabric samples. The scope of the present work is to study the changes in the tow cross-
section, tow waviness, nesting of the layers, yarn packing fraction and inter-tow voids
in the six layer dry fabric stack using CT under in-situ loading at low pressures. For this
objective a dry preform was compressed between two clear polycarbonate plates at low
pressures and scanned by CT to study the above mentioned parameters.
4.2 Nesting factor and tow geometry parameters
The nesting of layers in a laminate is an important effect which changes the thickness
and the permeability of the laminate resulting a change in fibre volume fraction and
mechanical properties of the final product [29]. The nesting is defined in terms of
nesting factor [34] or nesting coefficient [35].Nesting factor can be calculated by using
equation 4.1.
∑ (4.1)
Where NF is the nesting factor, Ts is the stack thickness and Ti is the individual ply
thickness.
In Figure 4.2a, the thickness of single ply is represented by “H”. In case of two plies of
individual thickness H sit exactly upon top of each other without shifting, the nesting
Chapter 4 Deformation of multilayer dry preforms under compaction
110
factor will be 1 i.e. no nesting (Figure 4.2b). The thickness of the stack in this case will
be 2H, the sum of the thickness of the individual plies. If there is shifting and nesting
between individual plies the nesting factor will be less than 1 (Figure 4.2c). The stack
thickness will not be equal to the sum of the individual ply thicknesses, but due to layer
shifting and nesting it will be less than the sum of the individual layer thickness by a
factor “N”, where N is the nesting of the layers. The stack thickness will be 2H-N
(Figure 4.2c). Nesting can be calculated by using the stack thickness (Ts) and the
individual layer thickness (Ti) as given in equation 4.2.
∑ (4.2)
Figure 4.2. Layer thickness (a) Single layer, (b) 2 layer without nesting, (c) 2 layer
with shifting and nesting
4.3 Material and mechanical testing
4.3.1 Material
The material used for the mechanical and microstructure analysis was E-glass plain
woven fabric with a warp density of 4.8/cm and weft density 4.4/cm. The warp and weft
(a)
(b) (c)
Chapter 4 Deformation of multilayer dry preforms under compaction
111
count was 650 Tex and density of the fibres was 2.60 g/cm3. The numbers of filaments
in the yarn were 1960.
4.3.2 Mechanical testing
Mechanical testing was done by using Instron testing machine. On the Instron 5569, two
metal platens were used for compaction of the woven fabric sample. The area of the
sample was 15 cm2. Testing of the samples was done by using static test method instead
of dynamic control. In dynamic testing the machine is moved continuously at a constant
speed up to certain pressure level and the pressure thickness curve is recorded. Whereas
in static test method the machine is moved to a defined pressure level and stopped at
that pressure level for certain defined period of time before proceeding to the next
position. As already discussed in previous research that due to relaxation or
rearrangement of the fibres the load decreases with time to maintain a constant
thickness or thickness decreases to maintain constant load due to relaxation or
rearrangement of the fibres [36-38]. In the present work, static test method was adopted
with pressure control to test the fabric sample so that at constant pre-defined pressures,
a constant thickness of dry preform may be achieved. In this method the machine was
moved to the desired pressure level at a constant speed of 1mm/min and then it was
hold for 5 minutes at that pressure level before moving to the next position. This 5
minute hold was considered adequate for this type of loading [30, 39]. Additionally a
compliance curve was also taken before the testing of the fabric samples so that minor
compliances due to the machine may be removed. The pressure thickness results for
single layer and for the stack of six layers were recorded.
Chapter 4 Deformation of multilayer dry preforms under compaction
112
4.4 Tomography and in-situ compression set up
For the compression of dry preform a compression rig (Figure 4.3) was made to
compress the preform within tomography machine for a meso-structure analysis of the
fabric sample. The rig comprised two polycarbonate compression plates 60x30x12 mm
in length, width and thickness respectively. Two side screws were used to compress the
plates from both sides and two thickness gauges were placed on either side to keep the
thickness uniform on both sides. The edge to edge distance of the two side screws was
40 mm, The sample size of dry fabric compressed between two plates was 40x30 mm.
The bending strength of the plates was estimated by applying beam bending theory to
ensure that there was no significant bending of the plates under load. Once mounted on
the x-ray scanner, the side screws were tightened to compress the fabric preform to the
desired pressure. From the pressure - thickness curve, the thickness value against
desired pressure was taken and the slip gauge of thickness corresponding to that
pressure was put on both sides of the plates.
Chapter 4 Deformation of multilayer dry preforms under compaction
113
Figure 4.3. Compression Rig (a) Compression rig fixed on the tomography stage,
(b) close-up of the rig
The process of CT involves collecting a large set of radiographs (projections) of the
sample during a single rotation. Together with a small number of calibration images,
these images are reconstructed into a 3D volume which represents the attenuation
through the sample. A Nikon Custom 320 Bay tomography system was used for this
analysis. The current was then adjusted to 110µA and the voltage was adjusted to 80 kV
and the white grey level was kept at 62500: 3500 projections were taken for each
tomograph.These images were then reconstructed using Metris X-Tek CT Pro and taken
to the Avizo 6.3 software for analysis.
4.5 Results and discussion
4.5.1 Macroscopic deformation
Figure 4.4 shows the average thickness results of the one and six layer fabric with the
pressure up to 600 kPa tested on Instron testing machine with static test method. It is
clearly evident from Figure 4.4 that the thickness of both the single layer and the six
Chapter 4 Deformation of multilayer dry preforms under compaction
114
layers preform decreased with application of pressures. The average thickness per layer
of the six layers stack is lower than the single layer at the same pressures which is
considered due to nesting of the layers. It can also be seen from the Figure 4.4 that the
thickness of the single layer and six layer stack is different at the start of the loading and
after 5 minutes hold at the same pressure. This is due to relaxation of the fibre as
already discussed. The thickness results changed at the start and became constant with
in 5 minutes of hold.
Figure 4.4. Pressure thickness curves of different layers
4.5.2 Meso-structural analysis by computed tomography (CT)
Figure 4.5 represents the different tow geometry parameters which were calculated
during the meso-structure analysis using CT. These parameters are also described in the
table 4.1 for both warp and weft yarns separately. The thread spacing P calculated is
also equal to half of the wavelength (λ/2).A sample size of 80 yarns for both warp and
weft yarns at different slices was observed for the tow geometry parameters.
Chapter 4 Deformation of multilayer dry preforms under compaction
115
Figure 4.5. Yarn geometry parameters
The tomography technique has the advantage that the specimen can be seen throughout
the 3D thickness at any slice so both the warp and the weft yarns were seen at different
slices with in the compressed fabric stack to check the tow parameters and the nesting
of the layers.
Table 4.1. Parameters used to describe the fabric geometry
Parameter Warp Weft
Yarn width a1 a2
Yarn thickness b1 b2
Yarn area A1 A2
Crimp
c1 c2
Crimp amplitude h1 h2
Yarn shift S1 S2
Thread spacing P1 P2
Crimp percentage is calculated by using the equation 4.3.
(4.3)
Chapter 4 Deformation of multilayer dry preforms under compaction
116
Figure 4.6 represents the 3D stack of dry preform which was reconstructed using Avizo
software.
Figure 4.6. 3D view of the six layer dry preform
Table 4.2 presents the mean values of various tow geometry parameters calculated
during compaction at different pressure levels. For the multilayer preform, the stack
thickness reduction at the initial load of 4 kPa was due to nesting of the layers and as
the pressure was increased from 4 kPa to 45 kPa (Figure 4.7a, b), there was slight
reduction in the thickness of the single layer which was due to crimp amplitude
reduction in the warp yarns, whereas the change in the yarn height and area was
insignificant. The crimp was higher in the warp yarns compared to the weft yarns which
can also be seen in the Figure 4.8. When the pressure level was increased to 45kPa
(Figure 4.7b) there was a decrease in the warp crimp percentage , and the crimp
amplitude whereas there was a slight increase in weft crimp percentage and crimp
amplitude, which is most likely due to the crimp balance or crimp interchange of the
warp and the weft yarns, this phenomenon of crimp interchange during compression
when crimp in one type of yarn is more than the other type of yarn was discussed in
detail by lomov et al. [40]. There was no change in yarn widths of both the warp and
weft tows at this stage and only a slight decrease in the yarn height and yarn area of
Chapter 4 Deformation of multilayer dry preforms under compaction
117
both warp and weft yarns was seen. At this pressure level nesting was the prominent
reason for the stack thickness reduction. The phenomenon of nesting was varying at
different slices with in the stack resulting a change in the single layer thickness at
different points with in the compressed fabric sample. At the third pressure level of 100
kPa (Figure 4.7c) there was a decrease in yarn area and yarn height and again there was
a reduction in the crimp and crimp amplitude of the warp yarns and an increase in weft
yarns at this pressure level. Also fabric layers were packed further and nested closer to
each other. In all these pressure levels the effect of layer nesting on stack thickness
reduction was dominant and the single layer reduction was mainly due to crimp
amplitude reduction at the 45 and 100 kPa pressure levels. This observation is in
agreement to the experimental results of Saunders et al. [16]and Teresa at al. [30] on
plain woven resin impregnated samples in which they also reported the nesting of the
layers as dominant factor in the initial mode of compression. When the pressure level
was increased to 100 kPa (Figure 4.7c), inter- tow voids were reduced. No significant
change was found in yarn widths upon an increase in pressure even to 100 kPa. From
Table 4.2, if the warp data (area, height) is compared with the weft data, the warp yarn
is likely to deform more as compared to the weft yarn. The reason may be due to the
number of the warp yarns per unit area being greater than the weft yarns so that there
was less nesting in the warp yarns compared to the weft yarns, giving rise to more
deformation of the warp yarns.
Chapter 4 Deformation of multilayer dry preforms under compaction
118
Table 4.2. Tow geometry parameters measured by X-ray tomography
Warp Weft
4 kPa 45 kPa 100 kPa 4 kPa 45 kPa 100 kPa
(STD) (STD) (STD) (STD) (STD) (STD)
a (mm) 2.08 2.07 2.1 1.65 1.65 1.66
0.16 0.2 0.18 0.08 0.08 0.08
b (mm) 0.33 0.32 0.29 0.41 0.4 0.38
0.03 0.03 0.03 0.03 0.03 0.03
A (mm2) 0.52 0.51 0.48 0.49 0.48 0.46
0.05 0.04 0.04 0.03 0.03 0.03
C (%) 3.5 2.8 2.3 1.02 1.03 1.05
1.01 1.12 1.16 0.67 0.61 0.59
h (mm) 0.57 0.54 0.49 0.23 0.22 0.22
0.07 0.06 0.06 0.06 0.05 0.03
P (mm) 2.1 2.1 2.11 2.27 2.26 2.28
0.11 0.16 0.15 0.09 0.12 0.13
S (mm) 0.53 0.71 0.72 1.1 1.15 1.12
Chapter 4 Deformation of multilayer dry preforms under compaction
119
Figure 4.7. Tomographic sections through the stack of six layers at different
pressures with warp cross-sections on the LHS and weft cross-sections on the RHS
at (a) 4 kPa, (b) 45 kPa, and (c) 100 kPa
Chapter 4 Deformation of multilayer dry preforms under compaction
120
Figure 4.8.Yarn isolated from the tomograph showing crimp in crossing yarns, (a)
warp yarn, (b) weft yarn
Table 4.3. Nesting measurements for the 6 layer preform
Nesting
Pressure Shift 1st 2nd 3rd 4th 5th 6th Stack
(kPa) (mm) layer layer layer layer layer layer Nesting
4 1.1 0.92 0.91 0.92 0.9 0.88 0.9 0.91
45 1.15 0.77 0.74 0.72 0.75 0.75 0.78 0.75
100 1.12 0.73 0.67 0.68 0.66 0.68 0.75 0.69
Table 4.3 presents the nesting measurements for 6 layers made from tomographic
images with random nesting at different slices. There was only a minor change in the
nesting of the individual layers at a pressure level of 4kPa but at the increased pressure
levels of 45kPa and 100kPa there was more deviation in the nesting results for the 6
individual layers. For the top and bottom layers at these two pressure levels least nesting
was observed, which may be due to them being in contact with the compression plates.
(a)
(b)
Chapter 4 Deformation of multilayer dry preforms under compaction
121
(a) (b)
Figure 4.9. Tomographic weft cross-sections showing layers with and without shift
at a) 45 kPa and b) 100 kPa
To investigate the effect of layer shifting and nesting on inter-tow voids, slices of dry
preform were taken into consideration where there was no shifting and nesting of the
tows and where some tows were shifted and nested closer to each other within the stack
of the six layers at pressure levels of 45 kPa and 100 kPa. It can be seen that at 45 kPa,
there was no shifting in the three bottom layers where the tows sit exactly upon each
other (Figure 4.9a), and there was shifting in the top three layers. In the bottom layers
bigger flow channels were present. The maximum area of the channels in the bottom
layers with no shifting was 0.43 mm2 while in the top layers where there was shifting,
the maximum area of the channels was 0.25 mm2. At the pressure of 100 kPa (Figure
4.9b), the layer were sitting upon each other with no shifting. The maximum area of the
channels in the top three layers was 0.23 mm2. In the bottom three layers where there
was the highest level of shifting and nesting, the maximum flow channel area was
0.1mm2. It can also be seen from the Figure 4.9 that at the point of perfect nesting, the
adjacent layers are giving almost no gaps for the resin infiltration. The results show that
channel area for resin flow decreases with the nesting of the layers. The higher the
nesting, the smallest the channel size and vice versa. The same behaviour was observed
Chapter 4 Deformation of multilayer dry preforms under compaction
122
by Schell et al. [41] for the resin impregnated sample. It also shows that when the
pressure increased from 45 kPa to 100 kPa the layers packing became more prominent.
4.5.2.1 Nesting factor analysis
The nesting factors were calculated from the tomographic images and the mechanical
testing separately at each pressure level using equation 4.1.For the tomographic nesting
factors, the single layer thickness was calculated from individual layers from the stack
of six layers and average layer thickness was taken from the average stack layer
thickness. In mechanical testing the nesting factors were calculated from the thickness
results of a single layer and stack of six layers taken from the Instron testing machine.
The nesting factor results calculated from both the tomography and the mechanical
testing were compared. In Figure 4.10, the nesting factors from mechanical testing and
tomographic images for six layer stack are presented. It can be seen that the nesting
factors calculated from the mechanical compressions are higher than the nesting factors
calculated from the tomographic images. The reason for this difference may be because
the single layer thickness for both mechanical and tomographic results differs due to
nesting of the layers within the stack of six layers in tomography. Also the nesting
factors for top and bottom layers are higher which mean lower nesting in both the top
and the bottom layers than the stack of six layers which may be the top and bottom layer
being in contact with the plate.
Chapter 4 Deformation of multilayer dry preforms under compaction
123
Figure 4.10. Nesting factors from mechanical testing and tomographic analysis
From Figure 4.10, it can also be seen that the decrease in the nesting factors was higher
when pressure was increased from 4 kPa to 45 kPa and there was lesser decrease upon
increase of pressure to 100 kPa for tomographic images. For the mechanical testing
there is no such sharp decrease in nesting factors. Also it was observed that the
deviation at different slices in the nesting factors of six layers was high when seen
throughout the sample in the 3D stack. The tomographic analysis of dry preform gives
an actual picture of the nesting in the preform during compaction.
4.5.2.2 Quantification of inter-tow voids using image analysis
The permeability of the fibre preform is mainly determined by the inter-tow voids and
the intra-tow voids has only a negligible effect on the permeability of the fibre preforms
[42] so the study of the inter-tow voids is significant to make accurate permeability
models. In the previous chapter 3, the inter-tow voids were calculated for single layer
Chapter 4 Deformation of multilayer dry preforms under compaction
124
preform. Here in this chapter the inter-tow voids or channels at different slices for both
warp and weft yarns and for the 3D structure of six layer preform were analysed and
quantified in Avizo 6.3 software by applying a threshold method as both the yarns and
the voids have different densities so they were easily distinguished by the threshold
method. They were labelled as different materials and in material statistics their
respective volumes were calculated.
Figure 4.11. Flow channels at (a) 4 kPa, (b) 45 kPa, and (c) 100 kPa
Figure 4.11 shows the presence of flow channels (inter-tow voids) in between the tows
at pressure levels of 4 kPa (Figure 4.11a), 45 kPa (Figure 4.11 b) and 100 kPa (Figure
4.11 c). These images were taken from the material segmentation tool in the Avizo 6.3
software for the weft cross-section at different slices. From these images it can be seen
that the flow channels are not connected in the series and there is a gap in these
channels which is due to nesting of the layers that results in breakage in pores
connectivity in these images. When the layers nest perfectly, the pores connectivity
breaks at that point, resulting no passage for resin flow.
Chapter 4 Deformation of multilayer dry preforms under compaction
125
Figure 4.12. 3D representations of inter-tow voids at (a) 4kPa, (b) 45kPa, and (c)
100kPa
Figure 4.12 represents the 3D view of the inter-tow voids at the three different pressure
levels. It can be seen that flow channels reduced when the pressure was increased from
4 kPa to 45 kPa and 100 kPa. It was noticed that these voids decreased sharply with an
increase in pressure. Almost 60~70 percent decrease in inter-tow voids was observed. It
was also seen that voids were higher in the area between the two tows and lesser in the
centre of the tow intersection when these voids were checked for different slices for
both the warp and the weft yarn cross-sections. Also interestingly the connectivity of
these resin channels was found to break at points of perfect nesting giving no passage to
resin flow. Figure 4.13 shows the inter-tow voids for both warp and the weft yarn cross-
sections at different slices and also the total inter-tow voids present in the 3D stack at
different pressure levels.
Chapter 4 Deformation of multilayer dry preforms under compaction
126
Figure 4.13. Inter-tow voids at different slices and average inter-tow voids
4.5.2.3 Yarn packing fraction
Packing fraction of the yarn was calculated by using equation 4.4
(4.4)
where ‘Vf’ is the volume of the fibres in the yarns and ‘Vy’ is the total volume of the
yarns in the stack. Here also the yarn volume was calculated by selecting the yarns by
applying the threshold and fibre volume was calculated by the density of the fibre given
in the material section. Table 4.4 shows the yarn packing fraction against the
corresponding pressures. There was only a slight increase in the yarn packing fraction
with increase in pressure from 4-100 kPa, which shows that the composites
manufactured using vacuum infusion will provide low packing fraction materials and
Chapter 4 Deformation of multilayer dry preforms under compaction
127
the thickness reduction and fibre volume fraction change is mainly due to reduction of
inter tow voids and nesting of the layers.
Table 4.4. Yarn packing fraction
Pressure (kPa) Yarn packing fraction
4 0.49
45 0.51
100 0.55
4.6 Conclusions
In the present work, changes in the tow cross-section, waviness, layer nesting and inter
tow voids has been studied in dry woven preform under in-situ compressive loading
using CT up to pressure range of 100 kPa being representative of VI. It was seen that at
initial pressure level of 4 kPa the thickness reduction of the preform was due to layer
nesting though the nesting was low in magnitude. Higher inter-tow voids were present
in the layers at this pressure level. As the pressure was raised to 45 kPa the layers nested
closer to each other and the thickness reduction was mainly due to this nesting effect.
There was no significant change in the tow cross-section of both the warp and the weft
at this pressure level. The thickness reduction of the individual layers was due to crimp
amplitude reduction which was again very low. At increased pressure level of 100 kPa,
it was observed that the tow cross-section deformed and again the nesting of the layers
was major contributor to stack thickness reduction. Here also the tow waviness was
responsible for single layer thickness reduction. At the increment of pressures from 4
kPa to 45kPa and 100 kPa, it was observed that the inter-tow voids reduced drastically.
The layer nesting was playing an important role on the magnitude of the resin channels.
Chapter 4 Deformation of multilayer dry preforms under compaction
128
It was noticed that at pressure levels of 45 & 100 kPa and at the point of perfect nesting
there was no space for resin infiltration. However, bigger resin channels were found
when layers were sitting upon each other without nesting. Yarn packing fraction was
also calculated at these pressures and it was seen that there is no significant increase in
the yarn packing fraction for the vacuum infusion region. The nesting factors were also
calculated from average layer and stack thickness using tomographic images and these
nesting factors were compared with nesting factor values from mechanical testing. It
was found that the nesting factors calculated from tomographic analysis are lower than
the nesting factor calculated from the mechanical test results. It is evident from the
tomographic study of dry preform that nesting and inter-tow void reduction are major
sources of stack thickness reduction at low pressures. This work will help the
researchers to understand the compaction process of dry preform at low load regime in a
better way and data of the inter-tow voids will be used to make permeability model in
the future.
References
[1] Summerscales J, Searle TJ. Low-pressure (vacuum infusion) techniques for
moulding large composite structures. Proceedings of the Institution of Mechanical
Engineers, Part L: Journal of Materials Design and Applications. 2005;219(1):45-58.
[2] Robitaille F, Gauvin R. Compaction of textile reinforcements for composites
manufacturing. I: Review of experimental results. Polymer Composites.
1998;19(2):198-216.
[3] Van Wyk CM. Note on the compressibility of wool. Journal of the Textile Institute
1946;37(12):T285-92.
Chapter 4 Deformation of multilayer dry preforms under compaction
129
[4] Gutowski TG, Dillon G. The elastic deformation of lubricated carbon fiber bundles:
Comparison of theory and experiments. Journal of Composite Materials.
1992;26(16):2330-47.
[5] De Jong S, Snaith JW, Michie NA. A mechanical model for the lateral compression
of woven fabrics. Textile Research Journal. 1986;56(12):759-67.
[6] Karbhari VM, Simacek P. Notes on the modeling of preform compaction: II-Effect
of sizing on bundle level micromechanics. Journal of Reinforced Plastics and
Composites. 1996;15(8):837-61.
[7] Simacek P, Karbhari VM. Notes on the modeling of preform compaction: I -
Micromechanics at the fiber bundle level. Journal of Reinforced Plastics and
Composites. 1996;15(1):86-122.
[8] Chen B, Chou T-W. Compaction of woven-fabric preforms in liquid composite
molding processes: single-layer deformation. Composites Science and Technology.
1999;59(10):1519-26.
[9] Chen B, Chou T-W. Compaction of woven-fabric preforms: nesting and multi-layer
deformation. Composites Science and Technology. 2000;60(12-13):2223-31.
[10] Chen Z-R, Ye L, Kruckenberg T. A micromechanical compaction model for woven
fabric preforms. Part I: Single layer. Composites Science and Technology.
2006;66(16):3254-62.
[11] Pearce N, Summerscales J. The compressibility of a reinforcement fabric.
Composites Manufacturing. 1995;6(1):15-21.
[12] Potluri P, Wilding MA, Memon A. A novel stress-freezing technique for studying
the compressional behavior of woven fabrics. Textile Research Journal.
2002;72(12):1073-78.
Chapter 4 Deformation of multilayer dry preforms under compaction
130
[13] Chen B, Cheng AHD, Chou TW. A nonlinear compaction model for fibrous
preforms. Composites Part A: Applied Science and Manufacturing. 2001;32(5):701-7.
[14] Chen B, Lang EJ, Chou T-W. Experimental and theoretical studies of fabric
compaction behavior in resin transfer molding. Materials Science and Engineering A.
2001;317(1-2):188-96.
[15] Bickerton S, Buntain MJ, Somashekar AA. The viscoelastic compression behavior
of liquid composite molding preforms. Composites Part A: Applied Science and
Manufacturing. 2003;34(5):431-44.
[16] Saunders RA, Lekakou C, Bader MG. Compression in the processing of polymer
composites 1. A mechanical and microstructural study for different glass fabrics and
resins. Composites Science and Technology. 1999;59(7):983-93.
[17] Mahadik Y, Hallett SR. Effect of fabric compaction and yarn waviness on 3D
woven composite compressive properties. Composites Part A: Applied Science and
Manufacturing.42(11):1592-600.
[18] Mahadik Y, Hallett SR. Finite element modelling of tow geometry in 3D woven
fabrics. Composites Part A: Applied Science and Manufacturing.41(9):1192-200.
[19] Mahadik Y, Brown KAR, Hallett SR. Characterisation of 3D woven composite
internal architecture and effect of compaction. Composites Part A: Applied Science and
Manufacturing.41(7):872-80.
[20] Hu J, Newton A. Low-load lateral-compression behaviour of woven fabrics.
Journal of the Textile Institute. 1997;88(3):242-54.
[21] Matsudaira M, Qin H. Features and mechanical parameters of a fabric's
compressional property. Journal of the Textile Institute. 1995;86(3):476-86.
Chapter 4 Deformation of multilayer dry preforms under compaction
131
[22] Saunders RA, Lekakou C, Bader MG. Compression and microstructure of fibre
plain woven cloths in the processing of polymer composites. Composites Part A:
Applied Science and Manufacturing. 1998;29(4):443-54.
[23] Grujicic M, Chittajallu KM, Walsh S. Effect of shear, compaction and nesting on
permeability of the orthogonal plain-weave fabric preforms. Materials Chemistry and
Physics. 2004;86(2-3):358-69.
[24] Senoguz MT, Dungan FD, Sastry AM, Klamo JT. Simulations and experiments on
low-pressure permeation of fabrics: Part II ”The variable gap model and prediction of
permeability. Journal of Composite Materials. 2001;35(14):1285-322.
[25] Dungan FD, Senoguz MT, Sastry AM, Faillaci DA. Simulations and experiments
on low-pressure permeation of fabrics: Part I-3D modeling of unbalanced fabric. Journal
of Composite Materials. 2001;35(14):1250-84.
[26] Hoes K, Dinescu D, Sol H, Parnas RS, Lomov S. Study of nesting induced scatter
of permeability values in layered reinforcement fabrics. Composites Part A: Applied
Science and Manufacturing. 2004;35(12):1407-18.
[27] Yurgartis SW, Morey K, Jortner J. Measurement of yarn shape and nesting in
plain-weave composites. Composites Science and Technology. 1993;46(1):39-50.
[28] Chen Z-R, Ye L. A micromechanical compaction model for woven fabric
preforms. Part II: Multilayer. Composites Science and Technology. 2006;66(16):3263-
72.
[29] Lomov SV, Verpoest I, Peeters T, Roose D, Zako M. Nesting in textile laminates:
geometrical modelling of the laminate. Composites Science and Technology.
2003;63(7):993-1007.
Chapter 4 Deformation of multilayer dry preforms under compaction
132
[30] Kruckenberg T, Ye L, Paton R. Static and vibration compaction and microstructure
analysis on plain-woven textile fabrics. Composites Part A: Applied Science and
Manufacturing. 2008;39(3):488-502.
[31] Arbter R, Beraud JM, Binetruy C, Bizet L, Bréard J, Comas-Cardona S, et al.
Experimental determination of the permeability of textiles: A benchmark exercise.
Composites Part A: Applied Science and Manufacturing. 2011;42(9):1157-68.
[32] Robitaille F, Gauvin R. Compaction of textile reinforcements for composites
manufacturing. II: Compaction and relaxation of dry and H2O-saturated woven
reinforcements. Polymer Composites. 1998;19(5):543.
[33] Kelly PA, Umer R, Bickerton S. Viscoelastic response of dry and wet fibrous
materials during infusion processes. Composites Part A: Applied Science and
Manufacturing. 2006;37(6):868-3.
[34] Potluri P, Sagar TV. Compaction modelling of textile preforms for composite
structures. Composite Structures. 2008;86(1-3):177-85.
[35] Lomov SV, Gorbatikh L, Kotanjac Z, Koissin V, Houlle M, Rochez O, et al.
Compressibility of carbon woven fabrics with carbon nanotubes/nanofibres grown on
the fibres. Composites Science and Technology. 2011;71(3):315-25.
[36] Kim YR, McCarthy SP, Fanucci JP. Compressibility and relaxation of fiber
reinforcements during composite processing. Polymer Composites. 1991;12(1):13-9.
[37] Kelly PA, Umer R, Bickerton S. Viscoelastic response of dry and wet fibrous
materials during infusion processes. Composites Part A: Applied Science and
Manufacturing. 2006;37(6):868-73.
Chapter 4 Deformation of multilayer dry preforms under compaction
133
[38] Robitaille F, Gauvin R. Compaction of textile reinforcements for composites
manufacturing. II: Compaction and relaxation of dry and H2O-saturated woven
reinforcements. Polymer Composites. 1998;19(5):543-57.
[39] Kim YR, McCarthy SP, Fanucci JP. Compressibility and relaxation of fiber
reinforcements during composite processing. Polymer Composites. 1991;12(1):13.
[40] Lomov SV, Verpoest I. Compression of woven reinforcements: A mathematical
model. Journal of Reinforced Plastics and Composites. 2000;19(16):1329-50.
[41] Schell JSU, Renggli M, Van Lenthe GH, Müller R, Ermanni P. Micro-computed
tomography determination of glass fibre reinforced polymer meso-structure.
Composites Science and Technology. 2006;66(13):2016-22.
[42] Yu B, James Lee L. A simplified in-plane permeability model for textile fabrics.
Polymer Composites. 2000;21(5):660-85.
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
134
Chapter 5
COMPACTION AND NESTING IN TEXTILE
PREFORMS INFLUENCED BY TOW
ARCHITECTURE
Z.Yousaf, P.Potluri, P.Withers
Abstract
This chapter reports macro and meso-scale deformations in textile preforms due to
compaction pressures applied during processing techniques such as vacuum infusion
and autoclave curing. 2D fabrics with a variety of interlacement patterns - plain, twill
and sateen- as well as stitched non-crimp (NCF) fabrics have been investigated. The
study demonstrates the influence of interlacement topology on the compaction and
nesting behaviour of individual plies. In-situ compression of single and multi-layer
fabrics in X-ray Computed Tomography shows interesting meso-scale features within
each ply such as tow waviness and inter-tow voids/channels for resin flow. Composite
laminate thickness, fibre volume fractions and the mechanical properties are influenced
by nesting efficiency as well as tow compaction behaviour under process pressure.
Keywords: weave architecture, compaction, nesting, computed tomography (CT), non-
crimp fabrics
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
135
5.1 Introduction
Low-cost composites manufacturing techniques based on resin infusion of dry textile
preforms are becoming popular in aerospace, automotive and energy sectors. In these
processes, several fabric plies are draped on the tool surface and subjected to transverse
compaction forces either by vacuum (Vacuum Infusion) or external pressure (in match
mould RTM), and at the same time infused with liquid resin. Compaction of the
preform changes the thickness and the resulting change in the fibre volume fraction of
the composite laminate. Preform compaction changes the tow waviness, which in turn
changes the mechanical properties of the composites [1]. Compaction of the preform
also effects its permeability to resin flow [2]. During autoclave curing, significantly
higher pressures (around 7 bar) are applied resulting in thickness reduction and a
corresponding increase in fibre volume fractions. As a result it is important to
characterise the compaction behaviour during the composite manufacturing process [3].
Extensive research work has been reported in the literature on the compaction behaviour
of different fabrics [4-15]. Van Wyk [16] was probably one of the first researchers to
treat the fibres under compression as a system of bending units. A typical pressure-
thickness curve was given by Mastudaira et al. [5]. The curve consists of three parts,
two linear and one exponential part. The effect of the reinforcement structure on the
compaction behaviour has been studied by various researchers [8, 13, 17-19].
Experimental research on the compression of plain weave fabric can be found in
previous literature [4, 8, 11, 17, 20, 21]. The compression response of non-crimp
stitched fabrics has also been reported [22-25]. The compression response of multilayer
preforms differs from that of a single layer due to nesting of the layers. The nesting
behaviour was discussed by Lomov et al. [26]. Resin permeability through fabrics due
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
136
to the nesting effect has been studied by various researchers [27-29]. It was observed by
Hoes et al. that nesting is a major source of variation in experimentally determined
permeability values [30]. The nesting of layers has been studied in terms of a nesting
factor [31] and a nesting coefficient [32]. The majority of the existing literature is
focused on plain weave, with very little reported on the influence of interlacement
architecture on compaction behaviour. Additionally, tow geometry analysis found in
the literature was conducted primarily on composite laminates not on dry fabrics.
Commercial fabrics have been used in previous compaction studies with no control on
weave parameters. In this study, several weave architectures were produced on a rapier
loom by keeping the remaining parameters such as inter-tow spacing and tow linear
density constant. This procedure enabled the study of the influence of interlacement
geometry on compaction behaviour, while keeping all the other parameters constant.
5.2 Material & Experimental details
5.2.1 Material
In this study, woven fabrics including plain (1/1), twill (3/1) and 5-harness sateen
weaves were produced with 600 tex, E-glass yarn in both warp and weft directions. In
addition, 0о/90
о and +45
о/-45
о non-crimp fabrics were included in the study [33]. The
fabrics were manufactured such that the yarn linear density (Tex) and inter-tow spacing
(ends or picks/cm) were identical for plain, twill and sateen fabrics. In this way the
influence of weave style on the compaction behaviour can be studied. The specifications
of the fabrics, used are listed in tables 5.1 & 5.2. The unit repeats of the plain, twill and
sateen weaves are as shown in Figure 5.1; the images were created using the TexGen
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
137
software developed by University of Nottingham [34]. The scanned images of woven
fabrics and non-crimp stitched carbon fabrics are presented in Figure 5.2.
Figure 5.1. Woven fabrics (a) Plain, (b) 3/1 twill, (c) 5H sateen
Table 5.1. Woven fabric specifications
Sample
#
Yarn
count
Type Weave
Ends
/cm
Picks/
cm
Areal weight
(g/m2)
WV-1 650 Tex E-glass 1/1 Plain 4.8 4.4 620
WV-2 600 Tex E-glass 1/1 Plain 4.8 4.4 548
WV-3 600 Tex E-glass 3/1 Twill 4.8 4.4 546
WV-4 600 Tex E-glass 5H Sateen 4.8 4.4 545
Table 5.2. Stitched non-crimped fabric (NCF) specifications
Sample
Yarn
count
Plies
orientation
Tows/cm Stitch yarn
Arial
weight
(g/m2)
NCF-1
12k-
carbon
+45о/-45
о 1.5/1.5
55 dtex
(polyester)
246
NCF-2
12k-
carbon
0о/90
о 2.5/2.5
80 dtex
(polyester)
550
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
138
Figure 5.2. Scanned images of woven fabrics and non-crimp stitched fabrics
5.2.2 Experimental Methodology
5.2.2.1 Mechanical testing
Mechanical testing of single and multilayer fabrics was conducted using an Instron
5569 universal testing machine (Figure 5.3) with a small-capacity load cell (5kN). The
surface area of the top and the bottom plate was 50 cm2. The fabric samples were cut
into 10x10 cm pieces. As the fabric thickness is small in comparison to the machine
stroke, the accuracy with which the compression strain measured becomes important. In
order to minimise errors, machine compliance as a function of the applied load was
measured and accounted in fabric starin calculations. Two different test methods were
used, namely continuous and static testing. In the continuous testing method, the cross-
head was moved continuously at 1mm/min and the resulting pressure thickness curve
was recorded. During the static test method, the machine cross-head was moved at a
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
139
speed of 1 mm/min to the desired load and then it was stopped for five minutes. Fibres
tend to relax during compression which result in a decrease in fabric thickness with
time, so five minute hold time was given to allow the fibres to relax so that there is no
further relaxation in the fibres with time and consequently, a stable, uniform thickness is
achieved. This five minute period was considered to be adequate for these type of
loadings [20, 21] . These final thickness values at each pressure level were recorded.
These values were also used to caliberate the compression rig for the meso-structure
analysis.
Figure 5.3. Compression set up
5.2.2.2 Computed tomography (CT) analysis
Meso-scale tow geometry has been studied by several researchers using SEM or
Computed Tomography, primarily on laminate samples. Tow geometry in dry fabrics
has not been studied so far as the fabrics are limp and hence difficult to section or scan.
In the present work, a novel in-situ loading rig has been developed (Figure 5.4) in order
to support dry fabric as well as to apply known compressive loads. The rig consists of
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
140
two clear polycarbonate plates, 60x35x12 mm in length, width and thickness directions
respectively. Two thickness gauges (slip gauges) were placed between the two plates on
either side in order to control the degree of compression. The fabric was placed
between these two plates and compressed to the desired thickness by tightening the side
screws. The gauge thickness was decided according to the pressure level required, based
on the mechanical tests.
Figure 5.4. Compression rig for in-situ loading
The fabric specimens were scanned on the Nikon Metrology 225/320 kV Custom Bay
system at the Henry Moseley X-ray Imaging Facility, University of Manchester.
Scanning was performed with a silver target using a voltage of 85 kV and a current of
115 μA. The number of projections was set to 3142 acquired of 360. The 3D data set
was reconstructed at full resolution with a voxel size of 13.2 μm along the x, y,
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
141
and z directions. Image analysis was performed using Avizo® Fire version 7.0.1
software.
The dataset were filtered using a non-local means filter in order to remove noise from
the data. The segmentation, i.e. separation of the pore space and fibres, has been
achieved by a global thresholding approach based on seed region growing. The seed
region growing is one of the simplest region based segmentation methods. It performs a
segmentation of an image wich examines the neighbouring pixels of a set of points,
known as seed points, and determines whether the pixels could be classified to the
cluster of seed points or not. The threshold is made by the user and it is usually based on
intensity, grey level, or colour values. The regions are chosen to be as uniform as
possible.
5.3 Analysis of macro-scale deformations
5.3.1 Woven fabrics
Fabric compression tests were performed on single layers, as well as stacks of four and
six layer fabrics. Figure 5.5 shows the pressure thickness curves for single layer plain,
twill and sateen fabrics. It was observed that, in case of single layer, the thickness of the
plain woven fabric was lowest of the three weaves, under a very small load; twill and
sateen weaves have nearly same thickness (Figure 5.5).
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
142
Figure 5.5. Single layer thickness results against pressure
With an increase in pressure, the thickness of all the fabrics started decreasing and at a
pressure of 20 kPa, the sateen fabric thickness became less than the plain fabric and at a
pressure of 30 kPa, the twill weave thickness becomes less than the plain fabric. Upon
increasing the pressure further, the least reduction was observed for the plain fabric and
the highest thickness reduction was evident for the sateen fabric. The plain fabric may
have the lowest initial thickness due to having the highest number of interlacement
points compared to the twill and sateen fabrics. Interlacements in the plain weave result
in higher inter-tow compaction forces and a corresponding reduction in thickness.. Twill
and sateen fabrics have fewer interlacements per unit area and hence a higher initial
thickness. During compression, plain weave offers the greatest resistance due to the
tightness of the weave giving the least reduction in thickness. Chen et al. [13] attributed
this to the bending stiffness of the interlacing tow segments. Of the two remaining
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
143
structures, sateen has the least number of interlacements and hence the highest reduction
in thickness.
(a)
(b)
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
144
(c)
Figure 5.6. Compression behaviour of multi-layer fabrics, (a) plain, (b) twill, (c)
sateen
In the case of multilayer preforms, stacks of two and six layers were tested for plain,
twill and sateen fabrics. The average layer thickness results against different pressures
are presented in Figure 5.6 along with single-layer curve. From Figure 5.6, it can be
seen that there was a reduction in the average layer thickness of all these fabrics with
increase in number of layers, which is considered to be due to nesting of the layers. This
observation of reduction of average layer thickness with increasing number of layers is
in agreement with the literature [7, 11, 17, 26] as already mentioned in chapter 4.
It can be observed from Figure 5.7 that prior to the application of pressure the thickness
of the multi-layer plain weave is the lowest. With increasing pressure, the rate of
reduction in thickness was least for the plain fabric and highest for the sateen fabrics
giving rise to the lowest fibre volume fraction in plain fabrics compared to twill and
sateen fabrics at high pressures. Consequently, the thickness of the twill and sateen
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
145
weaves is almost 20 percent higher than the plain weave at a pressure level of 5 kPa for
both two and six layers stacks. In the case of the two-layer stack, the thickness of the
sateen and twill fabrics became equal to that of the plain weave at pressures of 25 kPa
and 35 kPa respectively. But in the case of the six layer stacks, considerably higher
pressures of 57 kPa and 85 kPa were required to bring the sateen and twill fabrics
respectively equal to the plain weave fabric thickness.
(a)
(b)
Figure 5.7. Different layers thickness results against pressure, (a) two layers (b) six
layers
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
146
This observation can be attributed to the tight structure of the plain fabric, which
hinders compression compared to the twill and sateen fabrics. This observation is in
contradiction to the study of Saunders et al. [17], in which they found twill fabric to be
the most resitant to compression compared to plain and sateen fabrics. But in their
study, the yarn densities and the counts were not same for all the three type of weaves
which could result in different compaction behaviour of these fabrics. Overall, from
single layer and multilayer fabric study of plain, twill and sateen structures under
compression, it was seen that the plain fabric is the most compression resistant due to its
tight structure and consequently it results in the lowest fibre volume fraction at a given
pressure even though the pressures are characteristically low (<100 kPa) for vacuum
infusion.
5.3.2 Non-crimp fabrics
In the case of non-crimp stitched carbon fabrics, two configurations: +45о/-45
о and
0о/90
о were tested for compression behaviour. In both cases, the plies were stitched
together by a polyester thread. The pressure thickness results are presented in Figure
5.8.
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
147
(a)
(b)
Figure 5.8. Non-crimp fabric thickness results as a function of pressure, (a) +45о/-
45о, (b) 0
о/90
о
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
148
For NCF-1 (+45о/-45
о), it was observed that there was a reduction in the average layer
thickness upon movimg from a single layer to 2 but that the addition of further layers
resulted in only a marginal decrease in thickness. In the case of NCF-2 (0о/90
о), which
is similar to woven broadcloth, the thickness reduction with increasing number of layers
was prominent. For non-crimp stitched fabrics, different compaction behaviours were
observed in literature. Only a small thickness reduction was noticed in stitched non-
crimp fabrics for both 0о/90
о and +45
о/-45
о by Hammani [18] while the thickness
reduction was found to be higher in the 0 о
/-45 о
/90 о
/+45 о
plies compared to +45 о
/-45 о
plies [22]. Saunders et al. [17] observed that non-crimp stitched bonded fabrics are easy
to deform compared to plain and satin fabrics.
5.3.3 Nesting of layers
Layer nesting describes how the “hills” and “valleys” of two adjacent layers embed one
into another reducing the average layer thickness relative to that of an individual layer
[32]. The nesting of the layers is quantified in terms of the nesting coefficient [32] or
nesting factor [31]. The nesting factor can be calculated using the equation 5.1 [31].
∑
⁄ (5.1)
is the nesting factor, tack thickness and individual layer
thickness.
If the layers sit upon each other without nesting, the thickness of the two layers stack
will be equal to sum of the individual layer thickness and the nesting factor will be one.
But in the case where the layers shift and come closer to each other, the nesting factor
will be less than one and the thickness of the stack will be smaller than the sum of the
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
149
thickness of the individual layers such that a lower nesting factor indicates more
efficient nesting.
In this study, the nesting of the layers was calculated in terms of nesting factor (NF) for
multilayer stacks using test results obtained from the compaction of dry preform on
Instron testing machine. Figure 5.9 represents the results of nesting factors calculated
for woven and non-crimp stitched fabrics.
(a)
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
150
(b)
Figure 5.9. Nesting factors (a) woven fabrics (b) non-crimp stitched fabrics
In the case of woven fabrics, It was observed that the nesting factors are lower for plain
woven fabrics, reflecting better nesting compared to twill and sateen fabrics. The
highest nesting was seen in the six layer plain woven fabric which again can be
attributed to the presence of the shorter float length as already discussed. The highest
nesting factors were observed for sateen fabric. Higher nesting in plain woven fabrics
can be attributed to a shorter float length which allows better nesting . The reason of
better nesting due to shorter float length can be explained by the presence of more peaks
and valleys in the shorter floatt compared to longer float which allow neighbouring
yarns to locate in these valleys as mentioned by Lomov et al.[26]. On the other hand,
the sateen fabric,which have the largest float length, have minimum nesting. This
observation agrees with the study of Lomov et al. [26], according to which the presence
of longer yarn float results in minimum nesting and vice versa. As already mentioned in
the discussion of results of woven fabrics that least fabric thickness reduction was
witnessed in plain fabrics, in both single and multilayer cases. Whereas, better nesting is
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
151
observed in plain fabric compared to twill and sateen fabrics during nesting factor
calculations. Ideally, the better nesting in plain multilayer fabrics should result into
reduced stack thickness of plain fabric compared to multilayer twill and sateen fabrics
but opposite behaviour was noticed. The reason of higher stack thickness of plain
woven fabric, even in presence of higher nesting, can be returned to least single layer
thickness deformation due to tight structure of plain fabrics. Consequently, higher stack
thickness even in presence of higher nesting in plain fabrics.
For non-crimp fabrics, the measured nesting factor results (Figure 5.9), present better
nesting in case of 0о/90
о compared to +45
о/-45
о configuration. Again the reason may be
due to the difference in tow spacing. The nesting in 0о/90
о configured fabric is even
better than woven fabrics. This can be attributed to the presence of stitching yarn on the
face and the surface of the non woven stitched fabric,which allowed the yarns of the
other layers to nest better. Also, the shifting pattern of the layers during lay-up, plays an
important role in nesting. In ideal situations, the shifting of the layers should also be
considered using CT images during nesting comparison which can give a better
understanding of the nesting phenomenon.
5.4 Analysis of meso-scale deformations
For meso-deformation analysis, Wv-1 fabric was used. Both single layer and multi-layer
stacks were studied for geometrical changes using CT. The CT images were analysed
for tow waviness and inter-tow voids at different cross sections. Additionally, the void
behaviour was investigated for the 3-D structure of single layer and multilayer stack.
Figure 5.10 shows the 3D reconstruction of the single and multi-layer fabric samples.
The cross-sectional images of single layer and multi-layer stack are presented in Figure
5.11 and 5.14. The obtained images were studied for inter-tow voids using Avizo
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
152
software and the yarn waviness in the cross-sectional images was calculated using
measure software.
Figure 5.10. 3-D reconstruction of single layer (right) and 10 layers (left)
Figure 5.11. CT images of single layer
5.4.1 Tow waviness
Interlacement of warp and weft yarns imparts a certain amount of waviness to these
yarns in the fabric. This waviness is termed as yarn crimp. Due to this crimp, the actual
length of the yarn in the fabric exceeds the length of the fabric. Yarn crimp is calculated
using the equation 5.2.
(5.2)
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
153
Where P and L are actual yarn length and crimp yarn length respectively as seen in
Figure 5.12. The tow waviness results are presented in Figure 5.13.
Figure 5.12. Tow showing crimp
Figure 5.13. Yarn crimp percentage (a) warp yarn, (b) weft yarn
During the tow waviness study, it was observed that the crimp percentage is higher for
the warp yarns compared to the weft yarns in both single layer and multilayer cases at a
pressure of 100 kPa (Figure 5.13). Additionally, the crimp percentage of ten layer stack
was higher than that of single layer. This is most likely due to the nesting of the layers
in multilayer stack, as yarns of one layer are embedded in the gaps between the adjacent
yarns of the other layer which results in less crimp reduction in the nested areas.
Whereas, in the case of single layer fabric, the warp and weft yarns are in direct contact
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
154
to the compression plate, giving rise to a higher crimp reduction on application of
pressure. After increasing pressure from 100 kPa to 300 kPa, there was a reduction in
the crimp percentage of the warp yarns and an increase was observed for the weft yarns
for both single layer and ten layer stack (Figure 5.13). This phenomenon can be
attributed to the balancing of the yarns, in which the crimp percentage decreases in
yarns in one direction and increases in yarns going in the other direction on application
of pressure, if the two are not at same crimp levels. This behaviour was discussed in
detail by Potluri et al. [31]. Also at the pressure level of 300 kPa, the crimp percentage
of multilayer stack was higher than the single layer which again can be due to the
nesting of the layers in multilayer stack as already discussed. Moreover, the results of
the crimp percentage showed higher deviation for the multilayer stack compared to
single layer (Figure 5.13). Again the reason for this is due to the presence of nesting in
some areas and absence in other areas within the multilayer stack of ten layers. This
results in different crimp percentage at nested and non-nested regions, consequently,
giving more scattered results of crimp percentage for the multilayer stack.
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
155
Figure 5.14. CT images of ten layer stack
5.4.2 Inter-tow voids
The study of the inter-tow voids or the resin channels in textile preforms is important
for making accurate simulation tools as the resin permeability mainly depends on inter-
tow voids [2]. Here in this work, a study of inter-tow voids in 650 Tex, E-glass dry
plain woven fabric, single layer and multi-layer stacks conducted under in-situ
compaction has been presented. The inter-tow voids were investigated in 3D structure
as well as at different 2D slices with in the preform and the behaviour was compared for
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
156
single layer and multilayer. A slice is a CT image which corresponds to a certain
thickness of the object being scanned. A slice has a thickness of one voxel and the
length and width of the slice depend on the field of view of the sample. It was observed
that the inter-tow voids were higher in single layer compared to multi-layer stack at
pressure level of 100 kPa. At a pressure of 300 kPa, the inter-tow voids reduced in both
single and multilayer fabrics. The magnitude of voids reduction was higher in
multilayer stack compared to single layer (Figure 5.15). This higher voids reduction in
multilayer stack can be attributed to the nesting of the layers in the multilayer stack.
Also, in the multilayer stack, larger resin channels were present in non-nested regions
compared with the nested regions. The inter-tow voids were also investigated at
different slices in single layer and multilayer fabrics. The results of voids percentage,
slice by slice, are presented in Figure 5.16 and 5.17. It was found that the deviation of
the inter-tow voids from the mean results was higher in single layer compared to
multilayer stack. The reason of this higher deviation of the voids in single layer was the
different behaviour of voids in the centre of the tow and in between two tows. As
depicted in Figure 5.18, the absence/lower amount of fibres in-between the two tows
results in higher voids in this region compared to the centre of the tow. This gives
higher deviation of the void results from the mean value in single layer. Whereas, in
case of multi-layer stack, due to shifting and nesting of the layers these spaces between
the two tows are occupied by the top and bottom layers. This results in less deviation of
void percentage from the mean value, compared to single layer. Also, the voids
percentage along the slices represents a periodic shape which is due to sinusoidal
structure of the plain weave (1/1). The minimum voids are found at the intersecting
point of the tows, compared with between the tows where maximum voids are observed.
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
157
This behaviour is clearly visible in the periodic behaviour of inter-tow voids through the
samples (Figure 5.16 & 5.17).
Figure 5.15. Inter-tow voids in single layer and multilayer stack
Figure 5.16. Inter-tow voids in single layer along slices (a) 100 kPa, (b) 300 kPa
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
158
Figure 5.17. Inter-tow voids in multilayer stack along slices (a) 100 kPa, (b) 300 kPa
Figure 5.18. Plain-weave repeat presenting, between two tows and centre of the tow
5.5 Conclusions
In this work, 2D woven fabrics having different interlacement patterns with identical
tows and tow spacing were developed and tested for compaction behaviour.
Additionally, non-crimp fabrics with two different tow orientations were included in the
study. In woven fabrics compression of plain, twill and sateen fabrics and highest
compression was observed in sateen fabrics and least in plain woven fabrics. Opposite
behaviour was present in case of nesting where plain woven fabric gave better nesting
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
159
results. This showed that fabric compression and nesting are influenced by float length
of weave. In case of non-crimp fabric, highest nesting was present in fabrics placed at
+00/90
0. The tow waviness was higher in multilayer layers stacks compared to single
layer at 100 kPa showing higher tow waviness reduction in single layer in vacuum
infusion region.
Increase in pressure beyond 100 kPa resulted more reduction in tow waviness in
multilayer stack.The inter-tow voids were lower in multilayer stack compared to single
layer at 100 kPa, which was due to nesting of layers which reduced inter-tow voids in
multilayer stack. The reduction in voids was also higher in multilayer stack at
application of 300 kPa pressure, here again the nesting of the layers was the reason of
higher voids reduction.
References
[1] Yang B, Kozey V, Adanur S, Kumar S. Bending, compression, and shear behavior
of woven glass fiber–epoxy composites. Composites Part B: Engineering.
2000;31(8):715-21.
[2] Yu B, James Lee L. A simplified in-plane permeability model for textile fabrics.
Polymer Composites. 2000;21(5):660-85.
[3] Robitaille F, Gauvin R. Compaction of textile reinforcements for composites
manufacturing. I: Review of experimental results. Polymer Composites.
1998;19(2):198-216.
[4] Pearce N, Summerscales J. The compressibility of a reinforcement fabric.
Composites Manufacturing. 1995;6(1):15-21.
[5] Matsudaira M, Qin H. Features and mechanical parameters of a fabric's
compressional property. Journal of the Textile Institute. 1995;86(3):476-86.
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
160
[6] Chen B, Chou T-W. Compaction of woven-fabric preforms in liquid composite
molding processes: single-layer deformation. Composites Science and Technology.
1999;59(10):1519-26.
[7] Chen B, Chou T-W. Compaction of woven-fabric preforms: nesting and multi-layer
deformation. Composites Science and Technology. 2000;60(12-13):2223-31.
[8] Chen B, Lang EJ, Chou T-W. Experimental and theoretical studies of fabric
compaction behavior in resin transfer molding. Materials Science and Engineering A.
2001;317(1-2):188-96.
[9] Chen B, Cheng AHD, Chou TW. A nonlinear compaction model for fibrous
preforms. Composites Part A: Applied Science and Manufacturing. 2001;32(5):701-7.
[10] De Jong S, Snaith JW, Michie NA. A mechanical model for the lateral compression
of woven fabrics. Textile Research Journal. 1986;56(12):759-67.
[11] Saunders RA, Lekakou C, Bader MG. Compression and microstructure of fibre
plain woven cloths in the processing of polymer composites. Composites Part A:
Applied Science and Manufacturing. 1998;29(4):443-54.
[12] Simacek P, Karbhari VM. Notes on the modeling of preform compaction: I -
Micromechanics at the fiber bundle level. Journal of Reinforced Plastics and
Composites. 1996;15(1):86-122.
[13] Chen Z-R, Ye L, Kruckenberg T. A micromechanical compaction model for woven
fabric preforms. Part I: Single layer. Composites Science and Technology.
2006;66(16):3254-62.
[14] Chen Z-R, Ye L. A micromechanical compaction model for woven fabric
preforms. Part II: Multilayer. Composites Science and Technology. 2006;66(16):3263-
72.
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
161
[15] Potluri P, Parlak I, Ramgulam R, Sagar TV. Analysis of tow deformations in textile
preforms subjected to forming forces. Composites Science and Technology.
2006;66(2):297-305.
[16] Van Wyk CM. Note on the compressibility of wool. Journal of the Textile Institute
1946;37(12):T285-92.
[17] Saunders RA, Lekakou C, Bader MG. Compression in the processing of polymer
composites 1. A mechanical and microstructural study for different glass fabrics and
resins. Composites Science and Technology. 1999;59(7):983-93.
[18] Hammami A. Effect of reinforcement structure on compaction behavior in the
vacuum infusion process. Polymer Composites. 2001;22(3):337-48.
[19] Luo Y, Verpoest I. Compressibility and relaxation of a new sandwich textile
preform for liquid composite molding. Polymer Composites. 1999;20(2):179-91.
[20] Kim YR, McCarthy SP, Fanucci JP. Compressibility and relaxation of fiber
reinforcements during composite processing. Polymer Composites. 1991;12(1):13-9.
[21] Kruckenberg T, Ye L, Paton R. Static and vibration compaction and microstructure
analysis on plain-woven textile fabrics. Composites Part A: Applied Science and
Manufacturing. 2008;39(3):488-502.
[22] Lomov SV, Belov EB, Bischoff T, Ghosh SB, Truong Chi T, Verpoest I. Carbon
composites based on multiaxial multiply stitched preforms. Part 1. Geometry of the
preform. Composites Part A: Applied Science and Manufacturing. 2002;33(9):1171-83.
[23] Lomov SV, Verpoest I, Barburski M, Laperre J. Carbon composites based on
multiaxial multiply stitched preforms. Part 2. KES-F characterisation of the
deformability of the preforms at low loads. Composites Part A: Applied Science and
Manufacturing. 2003;34(4):359-70.
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
162
[24] Liu XL, Falzon PJ, Sweeting R, Paton R. Effective compressibility and
permeability of multi-layer non-crimp fiberglass reinforcements. Journal of Reinforced
Plastics and Composites. 2004;23(8):861-79.
[25] Lomov SV, Barburski M, Stoilova T, Verpoest I, Akkerman R, Loendersloot R, et
al. Carbon composites based on multiaxial multiply stitched preforms. Part 3: Biaxial
tension, picture frame and compression tests of the preforms. Composites Part A:
Applied Science and Manufacturing. 2005;36(9):1188-206.
[26] Lomov SV, Verpoest I, Peeters T, Roose D, Zako M. Nesting in textile laminates:
geometrical modelling of the laminate. Composites Science and Technology.
2003;63(7):993-1007.
[27] Grujicic M, Chittajallu KM, Walsh S. Effect of shear, compaction and nesting on
permeability of the orthogonal plain-weave fabric preforms. Materials Chemistry and
Physics. 2004;86(2-3):358-69.
[28] Senoguz MT, Dungan FD, Sastry AM, Klamo JT. Simulations and experiments on
low-pressure permeation of fabrics: Part II—The variable gap model and prediction of
permeability. Journal of Composite Materials. 2001;35(14):1285-322.
[29] Dungan FD, Senoguz MT, Sastry AM, Faillaci DA. Simulations and experiments
on low-pressure permeation of fabrics: Part I-3D modeling of unbalanced fabric. Journal
of Composite Materials. 2001;35(14):1250-84.
[30] Hoes K, Dinescu D, Sol H, Parnas RS, Lomov S. Study of nesting induced scatter
of permeability values in layered reinforcement fabrics. Composites Part A: Applied
Science and Manufacturing. 2004;35(12):1407-18.
[31] Potluri P, Sagar TV. Compaction modelling of textile preforms for composite
structures. Composite Structures. 2008;86(1-3):177-85.
Chapter 5 Compaction and nesting in textile preforms influenced by tow architecture
163
[32] Lomov SV, Gorbatikh L, Kotanjac Z, Koissin V, Houlle M, Rochez O, et al.
Compressibility of carbon woven fabrics with carbon nanotubes/nanofibres grown on
the fibres. Composites Science and Technology. 2011;71(3):315-25.
[33] http://www.sigmatex.com/Home.
[34] http://texgen.sourceforge.net/index.php/Main_Page.
Chapter 6 Deformation of dry and wet preforms under compaction
164
CHAPTER 6
DEFORMATION OF DRY AND WET PREFORMS
UNDER COMPACTION
Z.Yousaf, P.Potluri, P.Withers
Abstract
In the present work, the compaction behaviour of textile preforms has been studied for
dry and wet fabrics. Mechanical test results have been used to study the compaction
response of dry and wet preforms at the macro-level. A power-law relation has been
derived to predict the wet fabric thickness from the dry fabric thickness results. Meso-
scale analysis of single layer dry and wet fabrics has been conducted using computed
tomography (CT) under in-situ loading. Tow waviness and resin channels of dry and
wet fabrics have been studied for single layer preforms using CT images. It has been
observed that tow waviness of the fabrics decreases drastically with wettability at low
loads.
Keywords: dry preform, wet preform, compaction, computed tomography (CT)
Chapter 6 Deformation of dry and wet preforms under compaction
165
6.1 Introduction
The use of composite materials is becoming popular rapidly as a key structural element
in civil and the aerospace industries as they are easy to handle and have high strength to
weight ratio. For composite manufacturing a number of different techniques are
employed amongst which Resin Transfer Moulding (RTM) and Vacuum Infusion (VI)
are most common. A detailed review of both of these manufacturing processes can be
found in literature [1, 2]. These manufacturing processes involve the compaction of dry
fabrics, the infusion of the resin and compaction under wet conditions and subsequent
curing of the laminate to obtain the final composite material. Compaction of the fabrics
under dry and wet conditions changes the preform thickness which results in a change
in the fibre volume fraction of the preform. The compaction of the preform also affects
the inter-tow voids. The compression behaviour of different fabrics has been studied by
a number of researchers [3-6]. The compaction behaviour of wet fabrics is different
from dry fabric compaction as observed by various researchers [7-11]. In the previous
research work, the compaction behaviour of dry and wet fabrics has been investigated
using mechanical test results. There is no comparative study available on the structural
analysis of dry and wet fabrics. In the present work, the compaction behaviour of dry
and wet fabrics has been studied at macro-level using mechanical test results. A power-
law relationship has been used to predict wet fabric thickness from dry fabric test results.
The meso-structure of single layer dry and wet fabrics has been studied using CT
images.
Chapter 6 Deformation of dry and wet preforms under compaction
166
6.2 Material and mechanical testing
6.2.1 Material
The specifications of fabrics used for the macro study are presented in Table 6.1.
Table 6.1 Material used for mechanical testing
Structure Yarn type Yarn count GSM Fabric ID
Plain weave (1/1) Glass yarn 600-tex 548 P1
Twill weave (3/1) Glass yarn 600-tex 546 T1
Satin weave (5-harness) Glass yarn 600-tex 545 S1
Non crimp fabric ( +45/-45) Carbon yarn 12-k 246 NCF
The specifications of the material used for meso-structure analysis of dry and wet
fabrics are presented in Table 6.2.
Table 6.2 Material used for meso-structure analysis
Structure Yarn type Yarn count GSM Fabric ID
Plain weave (1/1) Glass yarn 650-tex 620 P2
6.2.2 Mechanical testing
For the mechanical testing of the fabric samples, an Instron 5569 machine was used.
Samples with an area of 50 cm2 were used and the speed of the machine during testing
was 1 mm/min. The load cell used was of 5kN. For the testing of wet fabrics, the
samples were saturated with water before compression. To study the macro-level
Chapter 6 Deformation of dry and wet preforms under compaction
167
deformation of the fabrics, a continuous test method was used. In this method the upper
compression plate was moved continuously at a speed of 1mm/min until the final load
was achieved and the pressure thickness curve was recorded. For meso-level
deformation study of the glass fabrics, a static test method was used. The reason for
using the static test method for this study was that there is thickness reduction in the
preform under loading with time due to relaxation of the fibres. Therefore, 5 minute
hold time was given for the fibres to relax so that an unchanged, stable final thickness
may be achieved. It had been observed in previous studies that after 5 minutes loading
at a constant pressure there is no further thickness reduction with time [12]. These
thickness values were also used for calibration of the compression rig used for the CT
analysis. Prior to the testing of the samples, the compliance curve was recorded. The
compliance values were removed from the thickness readings so that the compliance
effect due to the machine was balanced.
6.3 Results and discussion
6.3.1 Macroscopic deformation
The compaction of single layer fabric was performed in dry and wet conditions. For wet
compaction the fabric, saturated with water, was compacted to measure the pressure-
thickness response. The thickness results for dry and wet fabrics at different pressures
are presented in Figures 6.1 to 6.5.
Chapter 6 Deformation of dry and wet preforms under compaction
168
Figure 6.1. Thickness vs pressure of single layer and two layers of P1 fabric
Figure 6.2. Thickness vs pressure of single layer and two layers of T1 fabric
Figure 6.3. Thickness vs pressure of single layer and two layers of S1 fabric
Chapter 6 Deformation of dry and wet preforms under compaction
169
Figure 6.4. Thickness vs pressure of single layer and two layers of NCF
Figure 6.5. Thickness vs pressure of single layer and two layers of plain fabric P2
Two different thickness curves were recorded for each fabric; the thickness curve for
wet fabrics represents the lower thickness values, which shows there was higher
thickness reduction in wet fabrics compared to dry fabrics under the same compression.
The thickness reduction on being wetted occurs on compression of fabric which may be
associated with the crimp amplitude reduction; just lubrication of fibre without
compression does not reduce the fabric thickness. The thickness decrease on wettability
was higher in the low pressure regions compared to higher pressures. The highest fabric
Chapter 6 Deformation of dry and wet preforms under compaction
170
thickness reduction in wet fabrics was observed in sateen fabrics and the least thickness
reduction in plain fabrics. The reason for higher thickness reduction in sateen fabrics on
wettability can be returned to the presence of higher float length of the yarns and fewer
intersecting points giving more space for tow spreading resulting in higher thickness
reductions. This phenomenon of higher thickness reduction was observed in all the
compacted fabrics. This study is in agreement with previous research [13, 14].
A power-law relationship was obtained to predict wet preform thickness from dry
thickness results. To achieve this relationship, the negative stretch ratio (in the
compression direction) Td/T0 and Tw/T0 were calculated for both dry and wet fabric,
where T0 is the initial thickness at a low pressure of 3 kPa, which is assumed to be same
for both dry and wet fabrics and Td and Tw are dry and wet fabric thicknesses
respectively at increased pressure levels. The stretch ratio difference of both dry and wet
fabrics (Td-Tw/T0) was plotted against pressure and fitted to the power-law as shown in
equation 6.1. Both experimental and power equation curves were in good agreement
with each other as seen from Figure 6.6
(6.1)
Where X is the pressure, ‘Y’ is the stretch ratio difference (Td-Tw/T0) and ‘a’ & ‘b’ are
constants, equation 6.1 can be modified to derive equation 6.2.
Chapter 6 Deformation of dry and wet preforms under compaction
171
Figure 6.6. (Td-Tw)/T0 plotted against pressure for plain (P2) woven glass fabric
(6.2)
From this relation, wet thickness values (Tw) of different fabrics were calculated and
compared with the experimental results. A good agreement was observed between the
experimental and predicted values of these fabrics. The experimental and fitted values
of these fabrics are presented in Figures 6.7 to 6.11.
Figure 6.7. Experimental and predicted wet thickness of plain fabric (P1)
Chapter 6 Deformation of dry and wet preforms under compaction
172
Figure 6.8. Experimental and predicted wet thickness of twill fabric (T1)
Figure 6.9. Experimental and predicted wet thickness of sateen fabric (S1)
Figure 6.10. Experimental and predicted wet thickness of NCF
Chapter 6 Deformation of dry and wet preforms under compaction
173
Figure 6.11. Experimental and predicted wet thickness of plain fabric (P2)
6.3.2 Tomography and in-situ compression set up
For compression of dry and wet preforms a compression rig (Figure 6.12) was
developed to compress a dry preform so that the fabric under in-situ loading may be
scanned within the tomography machine for meso-structure analysis. The rig comprised
two polycarbonate compression plates, each measuring 60x35x12 mm in length, width
and thickness respectively. Two side screws were used to compress the plates from both
sides and two thickness gauges were placed on either side to keep the thickness uniform.
The edge to edge distance of the two side screws was 35 mm and the sample size of dry
fabric compressed between the two plates was 40x30 mm. For compression at each
pressure level a slip gauge of known thickness was placed between the two plates on
each side. Once on the scanner the side screws were tightened to compress the fabric
preform to the required pressure. From the pressure thickness curve the thickness
against desired pressure was taken and a thickness gauge corresponding to that pressure
was set on both sides of the plates.
Chapter 6 Deformation of dry and wet preforms under compaction
174
Figure 6.12. Compression Rig (a) Compression rig fixed on the tomography stage,
(b) close-up of the rig
The process of CT involves collecting a large set of radiographs (projections) of the
sample during a single rotation. Together with a small number of calibration images,
these images are reconstructed into a 3D volume which represents the attenuation
through the sample. A Nikon Custom 320 Bay tomography system was used for this
analysis. The current was then adjusted to 110µA and the voltage was adjusted to 80 kV
and the white grey level was kept at 62500: 3500 projections were taken for each
tomograph. These images were then reconstructed using Metris X-Tek CT Pro and
taken to the Avizo 7.1 software for analysis. Figure 6.13 represents the sectional views
of the preform, the segmented resin channels and yarns. The 3D reconstruction of the
fabric with resin channels compacted in the plates is shown in Figure 6.14.
Chapter 6 Deformation of dry and wet preforms under compaction
175
Figure 6.13. CT images showing slice of (a) warp cross-section, (b) segmented
inter-tow voids and (c) segmented yarns
Figure 6.14. Reconstructed image of single layer fabric compacted in between
polycarbonate plates
6.3.3 Meso-structural analysis by computed tomography (CT)
The yarn geometry parameters which were calculated during CT analysis of the fabric
sample are shown in Figure 6.15.
Upper and bottom plate
Yarns
Inter-tow voids
a
b
c
Chapter 6 Deformation of dry and wet preforms under compaction
176
Figure 6.15. Yarn geometry parameters
The tomography technique has the benefit that the sample can be analysed through the
3D structure at different slices. In the present study, the warp and weft tows were
studied for tow waviness at different slices. The resin channels in single layer dry and
wet fabrics were studied along the slices and in the 3D structure. The results of tow
waviness of dry and wet fabrics were compared at two different pressure levels. Tow
waviness was calculated using equation 6.3.
(6.3)
In equation 6.3, ‘L’ is the crimped length and ‘P’ is the original yarn length without
crimp.
The cross- sectional images of single layer dry and wet preforms are presented in Figure
6.16.
Chapter 6 Deformation of dry and wet preforms under compaction
177
Figure 6.16. Cross-sectional images of single layer dry and wet fabrics
6.3.3.1 Tow waviness
The tow waviness results of dry and wet fabrics at 10 and 300 kPa are presented in
Figure 6.17. It can be seen from the results that initially at a pressure level of 10 kPa,
the crimp percentages of the warp yarns in dry fabric were higher than for the wet fabric.
The reason might be that the coefficient of friction decreases in wet fabric on being
wetted due to lubrication of fibres at this lower pressure which results in greater
reduction in the warp yarn crimp percentage of wet fabric compared with dry fabrics.
Also, the crimp percentage yarns in both dry and wet fabrics were higher than the weft
yarns. At a pressure level of 10 kPa, relatively higher crimp percentage was present in
Chapter 6 Deformation of dry and wet preforms under compaction
178
the weft yarns of the wet fabric compared to dry fabric. This might be due to the
application of load, the phenomenon of crimp interchange occurs, where crimp in higher
crimped yarn decreases and increases in the less crimped yarn on application of pressure
when both the yarns are oriented in opposite directions. Therefore, at low load there was
higher reduction in the warp yarn crimp of the wet fabric and consequently an increase
in the weft yarn crimp of the same fabric.
(a)
(b)
Figure 6.17. Tow waviness of dry and wet fabrics (a) warp yarns, b) weft yarns
Chapter 6 Deformation of dry and wet preforms under compaction
179
At a pressure level of 300 kPa, there was again a decrease in the crimp percentages of
the warp yarns of both dry and wet fabrics. The decrease was higher in the warp yarn
crimp of wet fabric compared to dry fabric, but the difference in crimp percentages of
dry and wet fabrics at this pressure level was low compared to initial pressure level of
10 kPa. At 10 kPa pressure level, the crimp percentage of the warp yarns of wet fabric
was almost 25 % less than the crimp percentage of dry fabric, whereas at 300 kPa, the
difference in crimp percentage for dry and wet warp yarn was 9 %. This shows that at
low pressure, there is more crimp reduction for wet fabrics which contributes towards
the higher fabric thickness reduction in wet fabrics compared to dry fabrics.
6.3.3.2 Inter-tow voids
Dry and wet preforms were also investigated for the resin channels/inter-tow voids and
the results of the inter-tow voids of dry and wet preforms at a pressure level of 10 kPa
and 300 kPa are presented in Figure 6.18. It can be seen from Figure 6.18 that there
were more inter-tow voids in the dry preform compared to the wet preform at a pressure
level of 10 kPa. The greater thickness reduction in the wet preform at this pressure level
also affected the inter-tow voids.
Chapter 6 Deformation of dry and wet preforms under compaction
180
Figure 6.18. Inter-tow voids in dry and wet fabric
On application of a higher pressure of 300 kPa, the inter-tow voids reduced in both dry
and wet preforms. Again, at this pressure level, the inter-tow voids were smaller in the
wet preform compared to the dry preform and the reason again is due to the greater
thickness reduction in wet preforms on the application of pressure, which resulted in
more reduction of the inter-tow voids in wet preforms.
(a)
Chapter 6 Deformation of dry and wet preforms under compaction
181
(b)
Figure 6.19. Inter-tow voids (a) dry (b) wet fabric at 10 kPa
6.4 Conclusions
The aim of this work was to study the compaction of textile preforms for dry and wet
fabrics. It was observed that there was more deformation in textile preforms upon being
wetted. The fabric compression results at macro-level were fitted with a power-law to
predict wet thickness values from dry fabric results. There was good agreement between
the experimental and fitted result values. At meso-scale, single layer dry and wet fabrics
were studied for tow waviness and inter-tow voids. It was seen that there was a greater
decrease in tow waviness in wet fabrics compared to dry fabrics, which contributed to
the higher thickness reduction in wet fabrics. Also, there was a greater reduction in
inter-tow voids in wet fabric compared to dry fabric.
References
[1] Summerscales J, Searle TJ. Low-pressure (vacuum infusion) techniques for
moulding large composite structures. Proceedings of the Institution of Mechanical
Engineers, Part L: Journal of Materials Design and Applications. 2005;219(1):45-58.
Chapter 6 Deformation of dry and wet preforms under compaction
182
[2] Potter KD. The early history of the resin transfer moulding process for aerospace
applications. Composites Part A: Applied Science and Manufacturing. 1999;30(5):619-
21.
[3] Robitaille F, Gauvin R. Compaction of textile reinforcements for composites
manufacturing. I: Review of experimental results. Polymer Composites.
1998;19(2):198-216.
[4] Van Wyk CM. Note on the compressibility of wool. Journal of the Textile Institute
1946;37(12):T285-92.
[5] Gutowski TG, Dillon G. The elastic deformation of lubricated carbon fiber bundles:
Comparison of theory and experiments. Journal of Composite Materials.
1992;26(16):2330-47.
[6] De Jong S, Snaith JW, Michie NA. A mechanical model for the lateral compression
of woven fabrics. Textile Research Journal. 1986;56(12):759-67.
[7] Hammami A. Effect of reinforcement structure on compaction behavior in the
vacuum infusion process. Polymer Composites. 2001;22(3):337-48.
[8] Kelly PA, Umer R, Bickerton S. Viscoelastic response of dry and wet fibrous
materials during infusion processes. Composites Part A: Applied Science and
Manufacturing. 2006;37(6):868-3.
[9] Robitaille F, Gauvin R. Compaction of textile reinforcements for composites
manufacturing. II: Compaction and relaxation of dry and H2O-saturated woven
reinforcements. Polymer Composites. 1998;19(5):543.
[10] Debnath S, Madhusoothanan M. Studies on compression properties of polyester
needle-punched nonwoven fabrics under dry and wet conditions. Journal of Industrial
Textiles. 2012;41(4):292-308.
Chapter 6 Deformation of dry and wet preforms under compaction
183
[11] Francucci G, Vázquez A, Rodríguez ES. Key differences on the compaction
response of natural and glass fiber preforms in liquid composite molding. Textile
Research Journal. 2012;82(17):1774-85.
[12] Kim YR, McCarthy SP, Fanucci JP. Compressibility and relaxation of fiber
reinforcements during composite processing. Polymer Composites. 1991;12(1):13-9.
[13] Arbter R, Beraud JM, Binetruy C, Bizet L, Bréard J, Comas-Cardona S, et al.
Experimental determination of the permeability of textiles: A benchmark exercise.
Composites Part A: Applied Science and Manufacturing. 2011;42(9):1157-68.
[14] Kelly PA, Umer R, Bickerton S. Viscoelastic response of dry and wet fibrous
materials during infusion processes. Composites Part A: Applied Science and
Manufacturing. 2006;37(6):868-73.
Chapter 7 Development of an in-situ technique to analyse
meso-structure of dry fabrics under biaxial loadings
184
Chapter 7
DEVELOPMENT OF AN IN-SITU TECHNIQUE TO
ANALYSE THE MESO-STRUCTURE OF DRY
FABRICS UNDER BIAXIAL LOADINGS
Z.Yousaf, P.Potluri, P.Withers
Abstract
During the composite manufacturing process textile preforms experience different
forces. Due to the nature of woven fabrics, biaxial forces play an important role on the
tow geometry of the preform. In this research, a biaxial testing machine has been
designed which can measure the biaxial tensile and shear forces applied to the fabric.
An in-situ technique has also been developed to study the meso-structure of the fabrics
under biaxial loading. With the aid of the designed rig, the dry fabric can be scanned
using computed tomography (CT) under in-situ biaxial loadings. CT images of biaxially
loaded glass fabric were captured using this technique. The quality of the CT images
shows the validity of the developed technique for meso-structure analysis.
Keywords: woven preforms, biaxial rig, computed tomography, tow waviness.
Chapter 7 Development of an in-situ technique to analyse
meso-structure of dry fabrics under biaxial loadings
185
7.1 Introduction
Textile composites are becoming popular in aerospace, automotive and civil industries
due to their light weight, high strength to weight ratio and ease of handling during
composite manufacturing. In comparison to unidirectional composites, woven fabrics
have advantages due to their higher dimensional stability, better impact resistance and
damage tolerance [1], better drapeability and suitability to manufacture doubly curved
surfaces [2]. Several composite manufacturing techniques such as RTM and
thermoforming involve complex deformations of textile preforms including biaxial
tensile, in-plane shear, transverse compaction and out of plane bending deformations [3].
Various researchers have studied the shear and tensile behaviour of such high
performance fabrics [4-18].
Picture frame [8, 9, 19, 20], Bias extension [3, 16, 18, 21, 22] and Biaxial tests [3, 9]
have been used for shear testing of the fabrics. The deformation behaviour of woven
fabrics is generally of biaxial nature due to the presence of two mutually perpendicular
yarns known as the warp and the weft in the orthogonal fabric. Several researchers have
studied the deformation of textiles under biaxial shear and tensile loadings [9, 23-34].
The study of the biaxial nature of the interactions between yarns in two dimensional
woven fabrics is considered important as it includes the process of crimp interchange
between the yarns, which can affect many structural applications [32]. The study of the
meso-scale geometry during shear and tensile loadings is important during structural
analysis and for making simulation tools.
Hofstee at al. [35] investigated the influence of thermo-forming on the geometry of a
plain woven carbon fabric laminate. They studied the effects of shear deformation on
Chapter 7 Development of an in-situ technique to analyse
meso-structure of dry fabrics under biaxial loadings
186
yarn crimp and cross sectional geometry. The fabric samples were stretched and sheared
in a heat chamber above the melting temperature of the resin and the geometry of the
laminate was studied along the warp and weft cross sections using optical microscopy.
Tow geometry under deformation was also studied by Chang et al. [36]. In their study,
they pasted the resin on the surface of the deformed fabric samples under bias extension,
picture frame and biaxial testing. Fabric samples under loading were cured and with the
help of scanning electron microscopy (SEM) images of the samples, different tow
geometry parameters were investigated.
Potluri et al. [3] developed an optical technique for in-plane tow deformation in which
the textile preform was subjected to in-plane stresses ( tensile and shear). The deformed
images were scanned with a flatbed scanner for each deformed state. After application
of a specified biased load, a metallic clamp was placed around the pure shear region
before removing the applied load. Subsequently the fabric samples were impregnated
with epoxy resin and the load was removed after curing of the resin. The laminated
samples were then scanned by SEM and different tow geometry parameters were
calculated.
Analysis of tow deformation in textile preforms subjected to forming forces was
performed by Potluri et al. [28] by using images of laminates obtained by a video
microscope.
All previous research related with the geometry of the fabrics mentioned previously
deals with the analysis of cured fabric geometry after resin infiltration. The laminated
samples may not accurately represent the geometry of the dry fabric as the introduction
Chapter 7 Development of an in-situ technique to analyse
meso-structure of dry fabrics under biaxial loadings
187
of the resin into the fabric samples can change the tow geometry of the dry samples [3].
In the present research work, the focus was to establish a technique which may enable
the meso-scale geometry of the dry fabric to be measured under in-situ axial loadings.
To achieve this objective a biaxial testing machine was designed to load the fabric
samples in biaxial directions; this machine is shown in Figure 7.1. In addition an in-situ
rig shown in Figure 7.3 was devloped to hold the deformed geometry between two
frames so that the deformed geometry may be scanned using computed tomography
(CT).
7.2 Development of biaxial testing machine
For testing of the fabric samples under biaxial loadings for both shear and tensile tests, a
biaxial machine was designed by modification to the existing set up of the shear
machine designed by Ciurezu [37]. The machine consists of four linear actuators, which
can move in orthogonal directions and two small load cells of capacity 2.5 kN, which
are mounted on the two perpendicular jaws as shown in Figure 7.1. In the previous set
up of the machine, there was a limitation where it could only control one actuator at a
time and it was not possible to measure the fabric under biaxial loading by moving two
or more actuators simultaneously. In the modified set up of the machine, the
programming has been done using LabVIEW software from National Instruments [38]
to move the actuators with the following combinations:
individual actuators;
two actuators simultaneously;
four actuators simultaneously in the crossing direction.
Chapter 7 Development of an in-situ technique to analyse
meso-structure of dry fabrics under biaxial loadings
188
The software, FCT, developed by Festo can also be used to control the linear actuator
motions but it has a limitation in that it can only move one actuator at a time. The
actuators, speed can be varied from 0.1-9.1 mm/min.
Data acquisition from load cells is performed by using the data acquisition facility
(cDAQ 9172) of LabVIEW. Data from the load cells is received in mV/V, which is
converted to Newton force.
Figure 7.1. Rig for biaxial shear and tensile testing
Chapter 7 Development of an in-situ technique to analyse
meso-structure of dry fabrics under biaxial loadings
189
Figure 7.2 Biaxial rig with cruciform specimen
7.2.1 Load cell calibration
The loads cells used for acquiring the load values during testing of the samples for shear
and tensile testing need to be calibrated correctly prior to testing in order to achieve
accurate measurements. Here the load cells were calibrated by compressing a 10 kN
load cell in between the two actuators. The LabVIEW software was used for data
acquisition of the output voltage from the load cells .By using the output voltage of the
load cells and the value of the load obtained by the compressed load cell, the calibration
process was completed.
Different readings up to the load value of 2500 N were taken and by applying a trend
line to the curve of the values of the output voltage and load cell readings, the calibrated
values were calculated.
Chapter 7 Development of an in-situ technique to analyse
meso-structure of dry fabrics under biaxial loadings
190
7.3 Tomography and in-situ loading rig
Figure 7.3 Rig to grip the biaxial loaded fabric
A rig was designed as shown in Figure 7.3 to clamp the fabric sample under biaxial
loading (Figure 7.2) in such a way that it can be scanned in the computed tomography
machine under in-situ loading without distortion of the biaxial loaded geometry of the
fabric. The rig consists of two square frames of polycarbonate material with dimensions
75 mmx75mm inside the edges. The reason for selecting polycarbonate material was its
low density (1.22 g/cm3) compared to glass yarn (2.60 g/cm
3). The advantage of the low
density of polycarbonate is that fewer x-rays will be absorbed by the frame allowing
more x-rays to pass through the glass fabric giving better quality images. Four holes
were made in the four corners of each frame so that the two frames on the opposite sides
may be tightened with the loaded fabric inside. After the fabric was loaded under a
Chapter 7 Development of an in-situ technique to analyse
meso-structure of dry fabrics under biaxial loadings
191
certain load, glue was pasted on the edges of the frame in order to avoid any slippage of
the deformed fabric samples and the screws were tightened on all four sides of the
frames. This clamped fabric was left for 8 hours to allow the glue to dry. During pasting
of the glue on the frame edges, care was taken to avoid any contact of the glue with the
fabric inside the hollow area. After the glue was dried, the fabric outside the clamps was
trimmed and the clamped fabric was taken to the CT for analysis.
The CT analysis was performed on the Nikon Metris 225/320 kV Custom Bay system in
the Materials Science Centre in The University of Manchester. The current and voltage
were adjusted to 100 µA and 60 kV, respectively. For each scan, a total of 3142
projections were taken. These images were then reconstructed using Metris X-Tek CT
Pro software, and the data processed using VSG Avizo 7.1 software.
Chapter 7 Development of an in-situ technique to analyse
meso-structure of dry fabrics under biaxial loadings
192
Figure 7.4. CT images of the biaxially loaded dry fabric: a) side view, b) top view
and crossing yarns in two slices
7.4 Conclusion
An in-situ measurement technique has been designed and CT images have been
captured by using this technique which shows the validity of the methodology.
References
[1] Tabiei A, Yi W. Comparative study of predictive methods for woven fabric
composite elastic properties. Composite Structures. 2002;58(1):149-64.
[2] Page J, Wang J. Prediction of shear force using 3D non-linear FEM analyses for a
plain weave carbon fabric in a bias extension state. Finite Elements in Analysis and
Design. 2002;38(8):755-64.
[3] Potluri P, Perez Ciurezu DA, Ramgulam RB. Measurement of meso-scale shear
deformations for modelling textile composites. Composites Part A: Applied Science and
Manufacturing. 2006;37(2):303-14.
[4] Potluri P, Thammandra VS. Influence of uniaxial and biaxial tension on meso-scale
geometry and strain fields in a woven composite. Composite Structures.
2007;77(3):405-18.
Chapter 7 Development of an in-situ technique to analyse
meso-structure of dry fabrics under biaxial loadings
193
[5] Boubaker BB, Haussy B, Ganghoffer J-F. Consideration of the yarn–yarn
interactions in meso/macro discrete model of fabric: Part II: Woven fabric under
uniaxial and biaxial extension. Mechanics Research Communications. 2007;34(4):371-8.
[6] Quaglini V, Corazza C, Poggi C. Experimental characterization of orthotropic
technical textiles under uniaxial and biaxial loading. Composites Part A: Applied
Science and Manufacturing. 2008;39(8):1331-42.
[7] T.V. Sagar PP, Hearle JWS. Energy approach to predict uniaxial/biaxial load-
deformation of woven preforms. ECCM 10. 2002.
[8] Mohammed U, Lekakou C, Dong L, Bader MG. Shear deformation and
micromechanics of woven fabrics. Composites Part A: Applied Science and
Manufacturing. 2000;31(4):299-308.
[9] Sharma SB, Sutcliffe MPF, Chang SH. Characterisation of material properties for
draping of dry woven composite material. Composites Part A: Applied Science and
Manufacturing. 2003;34(12):1167-75.
[10] Harrison P, Potluri P, Bandara K, Long AC. A normalisation procedure for biaxial
bias extension tests. International Journal of Material Forming. 2008;1(1):863-6.
[11] Sharma SB, Sutcliffe MPF. A simplified finite element model for draping of woven
material. Composites Part A: Applied Science and Manufacturing. 2004;35(6):637-43.
[12] Hivet G, Vidal-Salle E, Boisse P. Analysis of the stress components in a textile
composite reinforcement. Journal of Composite Materials. 2012;47(3):269-85.
[13] Badel P, Vidal-Sallé E, Boisse P. Computational determination of in-plane shear
mechanical behaviour of textile composite reinforcements. Computational Materials
Science. 2007;40(4):439-48.
Chapter 7 Development of an in-situ technique to analyse
meso-structure of dry fabrics under biaxial loadings
194
[14] Launay J, Hivet G, Duong AV, Boisse P. Experimental analysis of the influence of
tensions on in plane shear behaviour of woven composite reinforcements. Composites
Science and Technology. 2008;68(2):506-15.
[15] Dong L, Lekakou C, Bader MG. Processing of Composites: Simulations of the
Draping of Fabrics with Updated Material Behaviour Law. Journal of Composite
Materials. 2001;35(2):138-63.
[16] Harrison P, Clifford MJ, Long AC. Shear characterisation of viscous woven textile
composites: a comparison between picture frame and bias extension experiments.
Composites Science and Technology. 2004;64(10–11):1453-65.
[17] Boisse P, Zouari B, Gasser A. A mesoscopic approach for the simulation of woven
fibre composite forming. Composites Science and Technology. 2005;65(3–4):429-36.
[18] Cao J, Akkerman R, Boisse P, Chen J, Cheng HS, de Graaf EF, et al.
Characterization of mechanical behavior of woven fabrics: Experimental methods and
benchmark results. Composites Part A: Applied Science and Manufacturing.
2008;39(6):1037-53.
[19] Peng XQ, Cao J. A continuum mechanics-based non-orthogonal constitutive model
for woven composite fabrics. Composites Part A: Applied Science and Manufacturing.
2005;36(6):859-74.
[20] Nguyen M, Herszberg I, Paton R. The shear properties of woven carbon fabric.
Composite Structures. 1999;47(1–4):767-79.
[21] Lebrun G, Bureau MN, Denault J. Evaluation of bias-extension and picture-frame
test methods for the measurement of intraply shear properties of PP/glass commingled
fabrics. Composite Structures. 2003;61(4):341-52.
Chapter 7 Development of an in-situ technique to analyse
meso-structure of dry fabrics under biaxial loadings
195
[22] Wang J, Page JR, Paton R. Experimental investigation of the draping properties of
reinforcement fabrics. Composites Science and Technology. 1998;58(2):229-37.
[23] Potluri P DA, Young RJ. Biaxial shear testing of textile preforms for formability
analysis. ICCM 16. 2007.
[24] Boisse P, Gasser A, Hivet G. Analyses of fabric tensile behaviour: determination of
the biaxial tension–strain surfaces and their use in forming simulations. Composites Part
A: Applied Science and Manufacturing. 2001;32(10):1395-414.
[25] Pan N. Analysis of woven fabric strengths: Prediction of fabric strength under
uniaxial and biaxial extensions. Composites Science and Technology. 1996;56(3):311-
27.
[26] Buet-Gautier K, Boisse P. Experimental analysis and modeling of biaxial
mechanical behavior of woven composite reinforcements. Experimental Mechanics.
2001;41(3):260-9.
[27] Gasser A, Boisse P, Hanklar S. Mechanical behaviour of dry fabric reinforcements.
3D simulations versus biaxial tests. Computational Materials Science. 2000;17(1):7-20.
[28] Potluri P, Parlak I, Ramgulam R, Sagar TV. Analysis of tow deformations in textile
preforms subjected to forming forces. Composites Science and Technology.
2006;66(2):297-305.
[29] Boisse P, Borr M, Buet K, Cherouat A. Finite element simulations of textile
composite forming including the biaxial fabric behaviour. Composites Part B:
Engineering. 1997;28(4):453-64.
[30] Willems A, Lomov SV, Verpoest I, Vandepitte D. Drape-ability characterization of
textile composite reinforcements using digital image correlation. Optics and Lasers in
Engineering. 2009;47(3–4):343-51.
Chapter 7 Development of an in-situ technique to analyse
meso-structure of dry fabrics under biaxial loadings
196
[31] Boisse P, Buet K, Gasser A, Launay J. Meso/macro-mechanical behaviour of
textile reinforcements for thin composites. Composites Science and Technology.
2001;61(3):395-401.
[32] Zhu D, Mobasher B, Vaidya A, Rajan SD. Mechanical behaviors of Kevlar 49
fabric subjected to uniaxial, biaxial tension and in-plane large shear deformation.
Composites Science and Technology. 2013;74(0):121-30.
[33] Xue P, Peng X, Cao J. A non-orthogonal constitutive model for characterizing
woven composites. Composites Part A: Applied Science and Manufacturing.
2003;34(2):183-93.
[34] Hivet G, Boisse P. Consistent mesoscopic mechanical behaviour model for woven
composite reinforcements in biaxial tension. Composites Part B: Engineering.
2008;39(2):345-61.
[35] Hofstee J, de Boer H, van Keulen F. Elastic stiffness analysis of a thermo-formed
plain-weave fabric composite: Part I: geometry. Composites Science and Technology.
2000;60(7):1041-53.
[36] Chang SH, Sharma SB, Sutcliffe MPF. Microscopic investigation of tow geometry
of a dry satin weave fabric during deformation. Composites Science and Technology.
2003;63(1):99-111.
[37] Ciurezu DAP. Characterization of large shear deformation in textile preforms. PhD
thesis. 2007.
[38] http://www.ni.com/labview/.
Chapter 8 Conclusions and future work
197
Chapter 8
Conclusions and Future work
8.1 Conclusions
In the present research work, the deformation behaviour of textile preforms was
studied under transverse compaction and an in-situ technique was introduced to
measure the meso-structure of dry textile preforms during biaxial tensile and shear
loading.
The deformation under compaction was studied at macro and meso-levels for single
layer and multilayer stacks of different architectures. Glass fabrics and non-crimp
fabrics (NCF) were used in this study. To investigate the meso-structure of glass
fabrics, an in-situ rig was designed which can be used to study the meso-structure of
textile fabrics under compaction loadings using computed tomography (CT).
Additionally, comparisons of dry and wet fabrics were made to investigate the
compaction behaviour of the fabrics after resin infiltration.
The study of single layer preforms was conducted for plain woven glass fabric
(chapter 2). Macro-scale deformation was studied by using mechanical testing of
single layer fabrics on an Instron testing machine using static loadings. The fabric
was tested in the dry and wet states. Higher deformation was observed in wet fabric
Chapter 8 Conclusions and future work
198
compared to dry fabric. Also it was observed that fabric thickness dropped to
maintain a constant pressure which was considered to be due to rearrangement of the
fibres. A power-law relationship was derived to predict the final fabric thickness
values for every load and good agreement was observed between the predicted and
the experimental results.
The meso-structure of single layer fabric was studied by compressing the fabric
sample in the designed rig and scanning it by computed tomography (CT) at
increased pressure levels up to 600 kPa. It was observed that fabric thickness
reduction at low loads up to 100 kPa was mainly due to reduction in tow waviness.
The phenomenon of crimp interchange was also observed at low loads. Changes in
the other parameters like tow area, tow thickness and tow width were not prominent
during these loadings. The reduction of inter-tow voids with application of pressure
was greater for pressures up to 100 kPa. On increment of pressure to 300 and 600
kPa, there was noticeable deformation in the tow area and tow thickness which
resulted in higher fibre volume fractions at these loads. CT was seen as an accurate
tool to calculate fibre volume fraction (FVF) as FVF calculated by mechanical test
results and CT were in good agreement.
The deformation behaviour of multilayer stacks was studied for 6 layer plain woven
glass fabric (chapter 4). Tow geometry and inter-tow voids were analysed for a
pressure range up to 100 kPa. It was seen that nesting of the layers was a major
source of stack thickness reduction in this low load regime. The main reason
controlling single layer thickness in this pressure range was reduction in tow
waviness. As for single layer preforms, crimp interchange was noticed in the
multilayer preform. Also it was seen that the resin channels in multilayer preforms
were highly influenced by the presence of nesting in layers. Maximum resin channels
Chapter 8 Conclusions and future work
199
were seen in slices where fabric layers were perfectly sitted upon each other and vice
versa. Due to the small decrease in tow area and thickness at these low pressure
values, the increase in yarn packing fractions was not prominent in multilayer stacks.
The deformation behaviour of textile fabrics with different architectures was also
studied at macro-level. Woven fabrics developed with plain, twill and sateen weaves
with the same tow count and tow spacing and non-crimp fabrics (NCF) were
included in this study (chapter 5). It was observed that the float length of the weave
plays an important role in the compression of woven fabrics. Increase in nesting was
observed with shorter float length and with increasing numbers of layers for woven
fabrics. In the case of NCF, it was seen that the fabrics placed at 00/90
0 orientation
exhibited higher nesting. However, no significant increase was observed when the
number of layers was increased to more than two. NCF placed at 00/90
0 showed even
better nesting than plain woven fabrics.
In addition to dry fabrics, the compaction behaviour of wet fabrics was also studied
at macro and meso-scales (chapter 6). Plain, twill and sateen fabrics of glass yarn and
non-crimp fabric of carbon yarn were included in the deformation study at macro-
level. It was observed that there was greater thickness reduction in wet fabrics
compared to dry fabrics. The power-law relation derived to predict wet fabric
thickness from dry thickness results showed good agreement between experimental
and predicted thickness results. The deformation behaviour of single layer wet
fabrics at meso-level was investigated using CT and compared with dry fabric meso-
structures for tow waviness and resin channels. The results showed that there was
higher reduction in tow waviness and inter-tow voids in wet fabrics.
Chapter 8 Conclusions and future work
200
In addition to study of the deformation behaviour of textile preforms under
transverse compression, an in-situ measurement technique was designed to study the
deformation behaviour of textile preforms during biaxial tensile and shear loadings
(chapter 7). CT images of plain woven glass fabric captured under in-situ biaxial
loading showed that this technique can be used to accurately measure the meso-
structure of dry textile preforms under biaxial tensile and shear loadings.
8.2 Future work
Research work on the deformation behaviour of dry textiles under forming forces can
be further carried out to analyse the changes at macro and meso scales. Following are
the main aspects of the future research work.
8.2.1 Simulation of the resin permeability during composite manufacturing
Experimental data have been achieved for the voids in single and multilayer fabrics
that now can be used to develop resin simulation tools during the composite
manufacturing process.
8.2.2 Meso-study of NCF and different woven architectures
A macro level deformation study of NCF and different woven fabrics has been
conducted and compared for thickness and nesting of layers. Further studies at meso
level can be performed to analyse the structural changes during compaction and to
study the effect of tow interlacement on the nesting of layers and resin channels.
8.2.3 Deformations of textile fabrics under biaxial shear and tensile loadings
Macro-structure analysis of different fabrics can be performed by using the
developed biaxial rig under biaxial shear and tensile loadings. For biaxial shear
Chapter 8 Conclusions and future work
201
loadings, existing high capacity load cells can be replaced by small capacity load
cells so that deformations at low loads may be recorded accurately. With the aid of
in-situ loading device meso structure analysis of fabrics for tow geometry changes
under biaxial shear and tensile loadings can also be performed using computed
tomography (CT).
Publications
202
Publications
1. Conference paper presented in Deformation and Fracture of Composites (DFC-
12) & Structural Integrity and Multi-scale Modelling (SI-6) at Queens’ College
Cambridge 8-11 April 2013University of Cambridge with title “Meso-scale
geometry and nesting of 2D woven fabrics during compaction”.
2. Conference paper presented in ICCM 19, July 2013, Montreal, Quebec, Canada
with title “Meso-scale analysis of 2D glass woven preform under compaction”.
3. “Deformation of single layer dry preforms under compaction” ready to submit in
Composites Part-A.
4. “Deformation of multilayer dry preforms under compaction” ready to submit in
Composites Part-A
5. “Compaction and nesting in textile preforms influenced by tow architecture”
ready to submit in Applied Composite Materials.