Short fibre polymer composites

Short fibre filled polymer composites form a relatively new family of materials,yet they are already well established in many applications. There is a vast range of materials in this category, some offering unique properties, some simply competing with other materials because of their relatively low cost. Their potential advantages are far from being fully realized and we anticipate continued growth in their use for many years to come. Research into these materials is crucial to their development and exploitation and will be for many years to come. Research continues into the design of short fibre reinforced composites and into the fundamental mechanisms that govern their behaviour, and also into methods of fabrication that will not only produce the required shape but will also result in the optimal properties being achieved.

Short fibre-polymer composites

Short fibre-polymer composites

Edited by

S K DE and J R WHITE

W O O D H E A D P U B L I S H I N G L I M I T E D

CAMBRIDGE ENGLAND

Published by Woodhead Publishing Limited,

Abington Hall, Abington, Cambridge CBl 6AH, England First published 1996, Woodhead Publishing Limited

0 1996, Woodhead Publishing Limited

Conditions of sale All rights reserved No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in

writing from the publisher

While a great deal of care has been taken to provide accurate and current information, neither the authors, nor the publisher, nor anyone else associated with this publication, shall be liable for any loss, damage or liability directly or indirectly caused,

or alleged to be caused, by this book

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

ISBN 1 85573 220 3

Cover design by the ColourStudio Typeset by Vision Typesetting, Manchester, England Printed by Galliard (Printers) Ltd, Great Yarmouth, England

Morphology of short fibre reinforced polymers

Mechanics of short fibre reinforced polymers

Measurement of fibre orientation distribution

Properties of fibre reinforced polymers

3 Thermosetting short fibre reinforced composites

S B WlLKlNSON AND J R WHITE

5 Composites of polychloroprene rubber with short fibres of

poly(ethy1ene terephthalate) and nylon

M ASHIDA

5.1 Introduction

5.2 Preparation of composites

5.3 Mechanical and viscoelastic properties

5.4 Effect of absorbed water

5.5 Dynamic fatigue of composites

7 Electrically conductive rubber and plastic composites

with carbon particles or conductive fibres

P B J A N A , A K MALLICK A N D S K DE

7.1 Introduction

7.2 Percolation phenomena in conductive polymer composites

7.3 Mechanism of electrical conduction

7.4 Effects of processing factors on electrical properties

7.5 Effects of polymer matrices on conductive network formation

7.6 Effects of the types of filler, their geometry and morphology

7.7 Effect of structural deformation on electrical resistivity

7.8 Hydrostatic pressure effect on resistivity

7.9 Temperature effects on volume resistivity

7.10 Galvanomagnetic properties of conductive rubber composites

7.1 1 EM1 shielding effectiveness

Stress-strain behaviour of unidirectional short fibre

reinforced elastomer composites

Fracture morphology, fibre configuration and

deformation mechanism

Stress-strain equation for unidirectional short fibre

reinforced elastomer composites

Modulus properties of unidirectional short fibre reinforced

elastomer composites and applications

Concluding remarks

References

9 Design considerations and end-use applications of

short fibre filled rubbers and thermoplastic elastomers

Short fibre filled polymer composites form a relatively new family of materials, yet they are already well established in many applications There is a vast range of materials in this category, some offering unique properties, some simply competing with other materials because of their relatively low cost Their potential advantages are far from being fully realized and we anticipate continued growth in their use for many years to come Research into these materials is crucial to their development and exploitation and will be for many years to come Research continues into the design of short fibre reinforced composites and into the fundamental mechanisms that govern their behaviour, and also into methods

of fabrication that will not only produce the required shape but will also result in the optimal properties being achieved The book reflects this and is offered as an introduction to anyone starting out in research into short fibre-polymer composites It also provides the background and bibliography for further reading It is intended to be of value to industrial technologists who are working with these materials and are seeking further insight into their manufacture and behaviour It is also meant to inspire materials users to consider new applications for these composites - even perhaps to formulate new ones with different

combinations of properties A special feature of the book is that it includes

significant discussion on rubber-matrix fibre composites, an important sub-class

of short fibre reinforced composites that is often neglected in reviews of polymer

composites

We are indebted to the contributors of the chapters for their co-operation We are grateful to them for letting us perform some editorial interventions with the aim of adding to the coherence of the text: we accept full responsibility if we have erred in this task! We wish to acknowledge also our students and colleagues at the Indian Institute of Technology, Kharagpur and the University of Newcastle upon Tyne who are responsible for stimulating and maintaining our interest in the area of short fibre filled polymer composites We are especially grateful to those students who have conducted research in this area under our supervision and whose work forms an important part of the chapters which we have authored

X Preface

This book would not have appeared without the efforts of Patricia Morrison and her colleagues at Woodhead Publishing We are indebted to our wives (Deya and Li Tong) for their expert technical help and advice during the preparation of the manuscript as well as for their patience and understanding Finally we are thankful to our children (Barna, Dominique, Michelle, Francine and Chris- topher) for their cheerful acceptance of one more task competing for our attention

S K De

J R White

M ASHIDA, 1-108 Andoji-cho, ltami 664, Japan (formerly Faculty of Engineer- ing, Kobe University, Rokkodai Nada, Kobe 657, Japan)

D M BIGG, R G Barry Corporation, PO Box 129, Columbus, Ohio 43216, USA

S K DE, Rubber Technology Centre, Indian Institute of Technology, Kharagpur

721302, India

A P FOLDI, C & C Consultants, 2833 W Oakland Drive, Wilmington, Delaware

B R GUPTA, Rubber Technology Centre, Indian Institute of Technology, Kharagpur 721302, India

P B JANA, Super Seals India Limited, Mathura Road, Faridabad 121003, India

M C H LEE, General Motors Research Laboratories, 30500 Mount Road, Warren, Michigan, 48090-9055, USA

A K MALLICK, Electrical Communication Engineering, Indian Institute of Technology, Kharagpur 721 302, India

G B N A N D O , Rubber Technology Centre, Indian Institute of Technology, Kharagpur 721302, India

J R WHITE, Materials Division, Department of Mechanical, Materials and Manufacturing Engineering, University of Newcastle upon Tyne, Newcastle upon Tyne NE1 7RU, UK

S B WI LKI NSON, Materials Division, Department of Mechanical, Materials and Manufacturing Engineering, University of Newcastle upon Tyne, Newcastle upon Tyne NE1 7RU, UK

19808-2422, USA

J R WHITE AND S K DE

I I Introduction

Short fibre reinforced composites are tinding ever-increasing applications in engineering and in consumer goods They can offer a unique combination of properties or may be used simply because they are more economical than competing materials The matrix material is usually polymeric and in many applications they compete with unreinforced polymers Fibre reinforcement improves the stiffness and the strength, and for many polymers it improves the toughness, though the toughness may decrease in polymers that are already tough in unreinforced form The dimensional stability is improved and, in the case of rubbery composites, better green strength is obtained Benefits such as creep resistance and better ageing and weathering properties may be crucial in some applications Conductive fibres may be added to change the electrical properties

The fibre reinforced composites with the best mechanical properties are those with continuous fibre reinforcement Such materials cannot be adapted easily to mass production and are generally confined to products in which the property benefits outweigh the cost penalty Short fibre reinforced composites can be processed in a manner similar to the matrix In the case of thermoplastics this means that methods such as injection moulding are available, allowing mass production of components with quite intricate shapes Reinforced thermosets suitable for injection moulding have also been developed, whereas thermoform- ing (using ‘sheet moulding compound’) is another alternative for short fibre-thermoset composites Reinforced natural and synthetic rubbers can be processed by the usual rubber processing methods such as calendering, extrusion and injection moulding, and uneconomic methods such as dipping, wrapping and laying are not required Many of the processes cause fibre alignment, which is often beneficial The properties are partly determined by the composition (fibre volume fraction) and partly by the processing, giving a wide range of property (and cost) combinations to which both designer and fabricator should be alert Part of the property advantage of continuous fibre composites derives from the

2 Short fibre-polymer composites

continuous nature of the reinforcement but part is a consequence of the highly parallel fibre orientation In short fibre composites the fibre orientation distribu- tion is far less perfect and is often random As a result, the degree of anisotropy is generally less than in continuous fibre composites, but it is often significant and must not be overlooked by product designers Processing windows may have to

be quite narrowly defined when making components for critical applications in order that the correct fibre orientation distribution is maintained

By adding suitable fibres and by controlling factors such as the aspect ratio, the dispersion and orientation of fibres, and the fibre-matrix adhesion, significant improvements in property can be achieved with thermoplastic, thermosetting and rubbery polymers

This book presents an up-to-date review of short fibre reinforced composites which covers thermoplastics, thermosets and rubbers The emphasis is on the research that has underpinned their development and discussion is centred on understanding the origins of their properties The aim is to provide an introduction to these ‘designer materials’ that will be helpful to both the component designer and the fabricator

I .2 Composition

1.2 I fibres

The most common fibre reinforcement is glass, usually E-glass (for ‘electrical

grade’) Glass fibres are usually 5-20pm in diameter and are round and fairly smooth, though their surfaces are never completely defect-free For some short fibre reinforced composites they are provided pre-chopped in lengths of a few millimetres For other applications they are provided already dispersed in a polymer matrix in the form of dough moulding compound (DMC) or granules In one of the specialized forms, the granules are formed by chopping a continuous fibre reinforced pultruded rod into lengths of lOmm or so in which the fibre length equals the granule length Yet another form of starting material is sheet moulding compound (SMC) in which short glass fibres are compacted into a

continuous sheet and impregnated with a thermosetting resin at room tempera- ture When the compound is processed the fibres become further shortened, especially in the case of injection moulding in which severe stresses are applied to the melt

A coating (‘size’) is commonly applied to glass fibres to improve processability

and to reduce damage during handling Surface coatings may also be applied to improve the adhesion between the matrix and the fibre

It is possible to improve the properties of composites by using higher performance fibres such as carbon, boron or polyaramid fibres (KevlarTM) Such fibres are often utilized in special applications of continuous fibre composites, in which the property benefits of the high performance fibres are exploited fully In short fibre composites, the matrix dilutes the properties much more than in continuous fibre composites and the cost of a high performance reinforcing fibre

is rarely justified At the other end of the cost and property spectrum, sometimes a fibrous filler that is obtained quite cheaply may be used to obtain a useful improvement in property compared with the unfilled polymer, even though the fibres are not easy to manipulate for optimum reinforcement Such is the case with sepiolite, a clay-like mineral.'-3

Metal fibres can be used to provide electrical conductivitỵ The major application for thermoplastics made conductive by ađing metal fibres is as housings for computers and other electrical goods requiring protection against electromagnetic interference and electrostatic dischargẹ Stainless steel fibres are used and loadings of the order of 1 YO by volume (depending on the application) are common Commercial grades are available based on several polymers including acrylonitrile butadiene styrene (ABS), various nylons, polycarbonate and polypropylenẹ This subject is discussed in Chapter 6

In the case of soft rubbery composites cellulose fibres have been found to give better reinforcement than glass or carbon fibrệ^ The reason for this is probably that the flexibility of cellulose fibres results in less breakage during processing than happens with the brittle glass or carbon fibres which have less resistance to bending The mixing processes applied to rubber compounds are quite intensive and are conducted while the compound possesses high viscosity, giving high bending stresses and causing severe fibre breakage with brittle fibres The fibre length used in the preparation of rubber composites is critical It should not be too long or the fibres will get entangled, causing problems with dispersion; if it is too short the stress transfer area is too small and the fibres do not provide effective reinforcement Short jute and silk fibres can be used to reinforce

Electrical conductivity can again be obtained by the use of ađitives and this is dealt with in Chapter 7

1.2.2 Polymers

Short fibre-polymer composites based on thermoplastics, thermosets and rubbers are available and are discussed in the following chapters

The most common thermoplastic short fibre reinforced composites are based

on polypropylene and nylon Higher performance thermoplastic composites use poly(ether sulphone) or poly(ether ether ketone) (PEEK), giving higher stiffnesses and higher working temperatures Also with thermoplastic behaviour are the thermotropic liquid crystal polymers in which the self-reinforcing property is further enhanced by the inclusion of fibres Polyimides, which can show either thermoset or thermoplastic characteristics, are also available in reinforced form Thermoset matrices include polyesters, epoxies and phenolics Thermosets are generally less suitable than thermoplastics for mass production but the improved properties achieved by ađing fibre reinforcement has provided ađed incentive

to develop a suitable means of fabrication

Both natural and synthetic rubbers are candidates for short fibre reinforce- ment Considerable attention is paid to short fibre-rubber composites in this book, a feature that distinguishes it from many other books in this areạ

4 Short fibrepolymer composites

1.2.3 Other ingredients

In addition to the major ingredients (fibre plus polymer matrix) the composites contain several other components, the precise nature and concentration of which will normally be a commercial secret The matrix may contain other fillers, such

as talc or fly-ash, to improve the mechanical properties, and many other additives may be present as processing aids, plasticizers, stabilizers, curing agents and mould-release agents Furthermore, the matrix may contain additives to enhance the fibre-matrix bond The purpose and effect of such additives are dealt with as appropriate in the following chapters

1.2.4 Preparation of short fibre composites 1.2.4 I Thermoplastics

Short fibre reinforced thermoplastics are most commonly supplied in the form of granules suitable for use in injection moulding machines The granules are roughly cylindrical, measuring 3-5 mm long and about 3 mm in diameter and the fibres are dispersed fairly uniformly within them To make the granules, chopped fibres and unfilled thermoplastic polymer powder or granules are fed into an extruder Mixing occurs on the passage through the extruder and the melt is passed through a die to make a lace that is chopped into granules of the required length Fibre breakage is a significant problem during this process and the design

of the screw is of some importance, deep flights being required Twin screw extruders are often preferred and the fibres are sometimes fed into the barrel part way along so that they enter directly into melt: this is found to reduce fibre breakage

In most of the common short fibre thermoplastic moulding materials the original fibre length is around 3 mm but few fibres survive mixing and moulding intact, with most fibres present in mouldings measuring fractions of a millimetre Another class of compound is available known as ‘long fibre’ thermoplastic, in which the granules measure about 10 mm and contain parallel continuous fibres that span the full length These granules are produced by chopping up pultruded rod in which a continuous tow of fibres is wetted by the polymer when forced through a heated die.8 The long fibre compounds are found to behave surprisingly well in conventional injection moulding machinery and the fibre lengths are preserved more completely than is the case with compounds in which the fibres are shorter to begin with

Other methods of mixing fibres into thermoplastic polymers are given in the book by F ~ l k e s , ~ including the use of kneaders and in situ polymerization

More details of the processing of reinforced thermoplastic materials are given

in Chapter 2

1.2.4.2 Thermosets

Chopped fibres can be combined with thermosetting resin at the fabrication stage

by spraying both ingredients on to a mould simultaneously In other processes

the fibres may be formed into a mat prior to impregnation with polymer resin In SMC the fibres are pressed into asheet and impregnated with polyester resin then stored until the final fabrication stage when the sheet is shaped in a hot press causing the resin to cure D M C also contains both uncured resin and filler, and has sufficiently good flow properties to be introduced into an injection moulding machine, though this is usually assisted by a 'stuffer' rather than by gravity feed as

is the case with thermoplastic granules

More details of the processing of reinforced thermoietting polymers are given

in Chapter 3

1.2.4.3 Rubbers

Short fibres can be mixed into rubbers using any of the rubber compounding machinery, including mills, calenders and internal mixers Fibre breakage is a problem and the mixing conditions selected must take this, as well as other factors, into account Mill mixing and calendering cause fibre orientation, giving rise to anisotropy and other property changes that are especially important if the final product is made this way

Details of reinforced rubber processing for various products are given in Chapters 4, 5, 7 and 8

I .3 Morphology of short fibre reinforced polymers

The detailed morphological characteristics of the different classes of short fibre reinforced polymers are described in the relevant chapters It is necessary only to give a brief general introduction here

The properties of short fibre-polymer composites are strongly dependent on the fibre volume fraction and on the fibre orientation distribution The fibre volume fraction is usually fairly tightly controlled, though some segregation of fibres and polymer may occur during fabrication Mouldings sometimes have a stratified morphology in which the most prominent feature is the fibre orienta- tion distribution which varies strongly from one depth to another within thermoplastic injection moulded short fibre-polymer composites.".' ' The fibre volume fraction may also vary with depth, though there are few reports of measurements of this in the literature In addition to variations with depth, there may also be localized regions at different sites along the flow path with significantly higher or lower fibre content than the average value Segregation of

fibres in thermoplastic matrices is discussed in section 2.3.2.3

The fibre orientation distribution changes when the moulding conditions change, but it is difficult to control Enormous benefits would be yielded if methods could be developed for exercising tight control over fibre orientation distribution in mouldings made from short fibre-polymer composites

In order to devise methods of controlling fibre orientation distribution it is necessary first to have a model for the dependence of fibre orientation distribu- tion on processing conditions Several studies have been made of this t o p i ~ ' ~ - ' ' The approach by Bay and Tucker is set out in some detail' and involves a finite

6 Short fibre-polymer composites

difference method in which temperature and velocity variations are calculated along the flow direction and through the thickness of the part The limitations of the method are discussed16 and the predictions are then compared with experimental results in a companion paper.' The general behaviour is found to follow the predictions reasonably well but the detailed agreement is not particularly good The motivation for studies of this kind is to enable computer modelling of processes such as injection moulding so that the design of the moulding machinery can be optimized before it is manufactured Thus in the case

of an injection moulding tool, the geometry of runner systems, the size and positioning of gates, and the arrangement of cooling channels are chosen with the assistance of computer modelling Even though careful experimental analysis shows that the results obtained with commercial predictive software are not particularly accurate and that further refinement should be sought, the wide- spread adoption of computer-aided design of moulds has resulted in a consider- able reduction in the fraction of tools that require adjustment before entering service after a trial period

Most attention has been focused on the fibre orientation distribution in injection mouldings but Advani and Tucker have developed a numerical simulation of the compression moulding of sheet moulding compound and have demonstrated a reasonably good agreement between their theoretical predictions and measurements made on real mouldings.'*

Many products benefit from having enhanced preferential fibre orientation in the direction of flow The SCORIMN process has been developed to maximize

this type of orientation This is achieved by repeatedly passing melt through the mould cavity using a 'live feed' push-pull arrangement with counteracting plungers which ensures that solidification is always in a high shear zone in which orientation is

During processing and subsequent fabrication of short fibre filled rubber composites, the fibres orient preferentially in one direction and the ultimate properties of the composites depend mainly on the angle between the fibre axis and the applied stress The effects ofmill parameters such as the number of passes, nip gap and roll speed ratio have been studied and the nip gap is known to have significant effect on the fibre orientation The greatest amount of fibre orientation takes place in the first pass However, the orientation is also strongly influenced

by the manner in which the sheet is folded and care should be taken to ensure that maximum orientation is obtained along the machine direction

Although the properties of fibre-reinforced composites depend mainly on the fibre orientation distribution, the adhesion between the fibre and the matrix is important Loads are not applied directly to the fibres but to the matrix For high performance, loads must be transferred effectively to the fibres, which demands a strong fibre-matrix bond This can be controlled by surface treatments applied to the fibre and/or by modifying the polymer to provide functional groups that bond

to the fibre or to the surface coating In rubbers, the dry bonding system known as HRH, consisting of hydrated silica, resorcinol (or a resorcinol derivative) and hexamethylene tetramine (a methylene donor) is used to promote adhesion

between fibre and rubber matrix.2 For applications requiring toughness rather than strength a weaker bond between fibre and matrix may be preferred since this will encourage the fibres to pull out of the matrix against frictional forces, giving large energy dissipation, rather than breaking at low composite deformation Some details of the chemical nature of the bonds are given in Chapters 2 , 3 , 4 and

6

In the case of crystallizing polymer matrices, the fibre may have a strong influence on the crystal morphology In some fibre-polymer pairings this is because crystal nucleation occurs on the fibre surface, leading to row nucleation that develops into a quite different morphology from that which forms in the absence of fibres in other fibre-polymer pairings crystal orientation follows molecular orientation that is influenced by the fibres.22 This morphology is likely

to enhance the fibre-matrix bond and hence influence the properties of the composite The changes in crystallinity and crystal orientation distribution that occur when fibres are included alter the properties of the matrix and this also modifies the composite properties

I .4 Mechanics of short fibre reinforced polymers

In a fibre-polymer composite the fibres are stiffer than the matrix and the proportion of the load that they support is greater than their volume fraction The overall elastic properties of a composite are relatively easy to compute from the elastic properties of the components when the fibres are continuous and parallelz3 and it is worth considering briefly this special case before moving on to the much more challenging problem presented by short fibre-polymer compos- ites The derivations of the relationships presented below are found in any standard text on composites (e.g the book by P i g g ~ t t ~ ~ ) and will not be reproduced here

In the following discussion the properties of the fibre will be indicated by subscript f, the polymer matrix by m and the composite by c Directions in the composite are indicated by subscript notation as described, for example, by Young and L0ve11.~~ We have chosen to represent the fibre axis direction by ‘3’: the reader is warned that this convention is not universally applied and some authors, including P i g g ~ t t , ’ ~ use ‘1’ to represent the fibre axis direction Thus if the Young’s modulus is represented by E and the volume fraction by V the

Young’s modulus in the direction of the fibres, E c , 3 3 , is:

This expression follows from the assumption that uniform strain is present in the composite when it is loaded in this way When an aligned fibre composite is loaded in the transverse direction, most of the deformation is taken up by the soft polymer phase and it is assumed that a state of uniform stress prevails This leads

to the following expression from which can be calculated the transverse Young’s modulus for the composite ( E c , l

8 Short fibre-polymer composites

If typical values for E,, Em, V, and V,, are substituted into equations [1.1] and [1.2], it is confirmed that the aligned fibre composite is very anisotropic Once the values for Ec.33 and E c , , , are known, the Young's modulus, 4 , at an angle 0 to.the fibre axis, can be calculated and is given by:

where the shear modulus G is given by:

and \ ' 1 3 is Poisson's ratio defined in the following way:

strain along libre axis transverse strain

1.4 I Stress transfer at the interface between a short fibre and the matrix

Consider a single iso!ated short fibre of diameter rl within a continuous matrix If

a homogeneous strain is applied to the matrix parallel to the fibre axis, load is transferred to the fibre by means of shear stresses at the interface (Fig 1.1) The shear stress varies along the fibre axis and, a s a consequence, the axial stress within the fibre varies If a small tensile deformation is applied parallel to the fibre axis (s-axis), the stress in the fibre at s o,(.s), and the interfacial shear stress acting

on the surface of the fibre, x i s ) , can be related by solving the force balance equation, which reduces to:23

The stress distribution along the fibre is derived using Cox’s ‘shear-lag

a n a l y ~ i s ” ~ in which the fibre is considered to be surrounded by a cylinder of polymer matrix of radius R where 2R is the average fibre separation distance (Fig 1.2) If the shear stress on a cylindrical surface radius r ( < R ) in the matrix is T then for a fibre of length I a n d diameter d, force balance requires that:

dT

2 n d ~ = ndlri i.e T = -

If u is the axial displacement at axial position x on a cylindrical surface within

the matrix at radius r, then the shear strain in the cylinder of matrix is given by:

10 Short fibrepolymer composites

where c, has been replaced by o,/E,

T h e general solution to equation C1.133 is

o f = E,E, + C sinh(Ax) + D cosh(Ax) C1.141 where:

If we now place the centre of the fibre a t x = 0 and assume that no axial stress is

and x = 3, a n d the transmitted across the ends of the fibre then o f = 0 at x = -

solution becomes:

[1.16]

T h e stress in the fibre increases from the fibre end and reaches a maximum a t the centre unless failure occurs T h e interfacial shear stress is obtained by differenti- ating equation C1.161 (see equation C1.61):

t = C1.171

’ 4cosh( A;)

Note that equations [ 1.161 and [ 1.171 show the dependence of the stress transfer within the composite on the fibre aspect ratio, l/d (recall that A is a function of d )

and confirm that it should be as high as possible

The maximum magnitude o f t , (equation [ 1.171) occurs at the fibre ends and, as

a consequence, this is where the first event leading to failure will occur by debonding from the matrix or by shear failure in the matrix Thus ifthe composite

is deformed sufficiently for this to occur, a slipped region develops and equations [l.l6]and C1.171 nolongerapply Adjacentto theslipped region thefibrestressis obtained by solving equation [1.6] for constant t i ( = T?, say) The result can be rearranged to give the length of the slipped region at each end of the fibre, I,, in terms of the stress in the fibre in the unslipped region (taken to be a constant value, c,,J i.e I , = o ~ ~ ~ / ~ T T The fibre will break if ( T ~ , ~ reaches the breaking stress, but this can happen only if the fibre length exceeds a critical length, length should exceed I,,,,, but if the composite is overloaded fibre fracture will occur until the fibres degrade to this value

The analysis presented above shows that both the fibre length and the aspect ratio should be as large as possible The dimensions of the short fibres for reinforced composites must be chosen accordingly and steps must be taken during processing to maintain lengths at adequate levels

AdE,E, sinh Ax

1 = 2 1 cr,, s,cr,, = cr,crild/2~: Generally, for the best mechanical properties, the fibre

1.4.2 Young’s modulus of a short fibre-polymer composite

Mathematical formulae for predicting the stiffness of a composite possessing an idealized fibre orientation distribution are well established The most commonly studied distributions are (i) all fibres parallel to the test direction, (ii) all fibres transverse to the test direction and (iii) random orientations The models also assume that the matrix is fully isotropic Such models are therefore not applicable

to injection moulded composites

Several mathematical models have been developed to predict Young’s modulus from the fibre orientation distribution; the one presented below2‘ has been developed from methods described previously by other workers I t is essentially an adaptation of the aggregate theory of Brody and Ward,27 incorporating the depth dependency of both the fibre orientation and the matrix stiffness

In injection moulded bars the overall anisotropy is expected to result from the preferred orientation of fibres and from anisotropy within the polymer phase Here we attempt to account for the latter by relating the matrix anisotropy to the local fibre orientation

In the aggregate theory the composite is considered to behave as an aggregate

of identically anisotropic sub-units, each with the elastic properties of fully oriented material The matrix is assumed to be anisotropic with transverse

12 Short fibrepolymer composites

The relationship between the global reference frame of the composite (as defined by the injection moulding process) and the local sub-unit reference frame (as defined by the fibre alignment) is shown in Fig 1.3

The properties of the sub-unit are represented in the usual mannerz8 using unprimed suffixes The elastic constants of the composite ('aggregate') are represented using primed suffixes It has been shownz8 that the aggregate axial compliance '53,38 is given by:

s 3 , 3 , = IIS, 1 + fzS,, + 1 3 ( 2 ' 5 1 3 + '544) E1.181 where:

&+)sin4 4 d 4

The elastic constants of the sub-unit were determined from the Halpin-Tsai

14 Short fibrepolymer composites

[ 1.271

Within a short fibre reinforced injection moulding there exists a distribution in fibre length This must be incorporated into the aggregate theory and this can be done by taking the maximum and minimum values for fibre length, I,, to give upper and lower bounds for the composite stiffness.z6 O’Donnell and Whitez6 showed that the calculations of composite stiffness are not very sensitive to the fibre length when it is in the range found in the typical injection moulded composites used in their studies Some of their results are given in Chapter 2 O’Donnell and White perfomed a layer-wise analysis and obtained Young’s modulus as a function of depth through the mouldings The value of the Young’s modulus of the matrix inserted into the equations was taken to be that measured

in mouldings made from the same polymer in unfilled form This varied with depth and it was shown that this produced better agreement with the measured Young’s modulus d’stribution than when the calculations used a uniform matrix Young’s

I 5 Measurement of fibre orientation distribution

It is evident f i J m the discussion presented in section 1.4 that the characteristic that is most important in determining composite mechanical properties is the fibre orientation distribution In some cases it is sufficient to have a qualitative assessment of the fibre orientation distribution, as, for example, near a weld line in

a multiply-gated injection moulding In mouldings having relatively simple shapes and gating arrangements it is an advantage to have quantitative measurement methods of fibre orientation distribution so that an objective assessment may be made of the effect of different processing procedures designed

to control the fibre orientation distribution It is equally important to be able to determine the parameters describing the orientation distribution for insertion into the equations that describe the mechanical behaviour This is normally done

by image analysis, which can also provide measurement of the fibre length distribution, also needed for the mechanics analyses

I .5 I Image analysis of seaions

The first step is normally to prepare a section of the sample cut in a chosen direction, often normal to the flow direction in an injection moulding (but see discussion later on) The section is then polished using metallographic pro- cedures, developing improved polishes through the successive use of abrasive papers followed by a polishing wheel with cloth-borne abrasive particles of diminishing s i ~ e ~ ’ ~ ’ This requires much care with unfilled polymers because of their softness and tendency to heat when abraded; the use of a cooling mixture of

water and detergent is r e ~ o m m e n d e d ~ ' With fibre reinforced materials the large difference in stiffness or hardness between the fibre and matrix can lead to problems and care must be taken to ensure that differential removal of the two phases does not occur and that fibre breakage does not take place Toll and Andersson claim that uneven cutting occurs if the final polish is made with particles that are too fine and recommend a final polish with 2pm diamond particles.32 Examples of sections through short fibre reinforced composites are given in Chapter 2 (thermoplastics) section 2.3.2.3, Fig 2.2, and in Chapter 3

(thermosets) section 3.6, Fig 3.2 and 3.3

When viewed in the light microscope, the contrast between the fibre and the matrix may not be sufficient, especially for automated image analysis, and when this is thecase an etch may be used, generally to erode the polymer matrix slightly

in the case of thermoplastics Alternative procedures for thermoset matrix composites include etching the sample either with hydrofluoric acid (which attacks the fibres) or with oxygen plasma Another method for enhancing contrast is by applying a vapour deposited coating, usually gold or platinum instead of an etch Touching fibres can present a problem in automatic image analysis but instructions can be included in the software for recognition and suitable a c t i ~ n ~ ' ? ~ ~

The most common method of fibre orientation measurement is by using the light microscope and an image analysis package The image analyser measures the aspect ratio (maximum/minimum diameter) of the elliptical image of each fibre section in the field of view; the reciprocal of this quantity equals cos 4, where

4 is the angle between the fibre axis and the normal to the sampling plane A field typically contains 512 x 512 pixels on a video screen Each pixel is examined separately by the image analyser and described by its x and y co-ordinates plus a digitized grey level which describes its shade usually in 256 steps between black

(0) and white (255) The grey level assigned to an individual pixel may be altered

to enhance the image contrast The enhanced grey image is then available for image analysis The orientations of the fibres are determined from their shape and the orientation of their elliptical intersections on the plane of the section The error involved in digitizing the diameters when they are expressed in terms of pixels is important when the fibre is closely parallel to the normal to the

because of the shape of the cos 4 function for small values of 4, and for typical image magnifications the error in 4 is of the order of 10" To avoid this

problem, Hine et al recommended the use of oblique sections, cut at an angle

between 35" and 75" to the flow direction, when a strong preferred orientation is present.36

The objective of the image analysis is usually to produce a volume average of the fibre orientation distribution rather than a description of the fibre orientation distribution of those fibres that happen to intersect the sampling plane Those fibres that are oriented normal to the sampling plane are most likely to intersect it and a weighting function of l/cos$ or similar should be applied to compensate for this.32*34*37 Fibres that are inclined to the normal to the sampling plane and are intersected near to one end appear as truncated ellipses, and ways of dealing

16 Short fibre-polymer composites

with this are discussed by Bay and Tucker34 and by Fischer and E ~ e r e r ~ * Another problem occurs when a fibre that lies nearly parallel to the sampling plane is curved along its length and this is discussed by Toll and A n d e r ~ s o n ~ ~ Recent developments in image analysis coupled with the rapid measurement of thousands of elliptical intersections, render sampling errors negligible, though this is not generally available on commercial image analysis packagệ^^.^^ In

another development, the use of confocal microscopy, producing images from different selected planes beneath the polished surface, allows 3D orientations to

be measured from a single scan;35 again this procedure is not in routine usẹ Conventional commercial image analysis equipment has been used by several workers to investigate the orientation of fibres that intersect a plane in a short fibre The fibre orientation can be represented by an orientation parameter as recommended by Fakirov and F a k i r ~ v a , ~ ’ who used two orienta- tion parameters introduced by Hermans4’ to describe the orientation in crystalline polymers Assuming that there is a planar distribution of fibres, the Hermans orientation parameter f, is of the form:

f p = 2 < c O s ’ + > - 1 where:

The modified Hermans orientation parameterfb is of the form:

f b = 2 c o s 2 < 4 > - 1 where:

+ is shown in Fig 1.3

The angle that the major axis of the ellipse makes with the y direction is equal

to 0 The angle that the fibre axis, OF, makes with the major flow direction (the x-axis) is equal to + and is determined by calculating aspect ratios of the fibre intersections:

r length of minor axis

R length of major axis cos+ = - =

The orientation parameters are scaled so that for a composite material where the fibres are randomly oriented f, andf, are both zerọ If the fibres are aligned parallel to the flow direction then bothf, andfp = 1 and if the fibres are aligned perpendicular to the flow direction.f, = ,rp = - 1 For a section taken at an angle

TX to the vertical axis (as in the work by Hine et it is possible to obtain both 8

and + using simple trigonometric transformations

1.5.2 Contact microradiography

If a thin section can be made, the fibre orientation distribution can be revealed by microradiography whereby X-rays are passed through it and their transmission is recorded on an X-ray sensitive film.43 The shadows thrown by the fibres d o not

permit the assessment of their inclination to the plane of the section As an

alternative to performing image analysis on the microradiographs, McGee and McCullough suggested using optical diffraction to obtain measurements of average orientation^.^^

I .6 Properties of fibre reinforced polymers

I .6 I Average properties

Table 1.1 lists some of the key properties of structural materials Short fibre reinforced thermosetting polymers are represented by SMC and DMC, and short fibre reinforced thermoplastics by GFPP and GFN6,6 Much of the property range given for SMC and D M C is due to differences in composition from different sources (largely undeclared) The same is true with thermoplastics: glass fibre reinforced nylon 6,6 with 20% by weight of glass is much closer in property

to the polypropylenecompound with a similar glass content than it is to GFN6,6 with 40% by weight of glass The stiffness can be almost doubled by replacing glass fibres by carbon fibres; this also causes significant improvement in the strength and the density (which is lower in carbon fibre reinforced compounds) The true value of reinforcement can be judged first of all by comparing the composites with the corresponding unreinforced polymers in Table 1.1, remem-

Table 1 I Properties of selected materials Young’s modulus,

Material G N m-* Strength, M N m - ’ Density, kgm-’

18 Short fibre-polymer composites

bering that the reinforced materials can be processed using the same methods Secondly, when assessing the suitability of the materials for certain structural applications it is often instructive to divide the key property by the density to obtain a figure of merit or ‘performance index’ For example, aerospace (and, to a lesser extent, automotive) applications normally require stiffness and at the same time weight saving When this is taken into account the polymer composites can often compete very strongly with high performance metal alloys The derivation

of appropriate performance indices and their application is discussed by Ashby.*’

1.6.2 Depth variation of properties The properties discussed above are those averaged through the section of the moulding If there is fibre segregation or if the fibre Orientation distribution varies through the depth of the moulding then there will be a variation in property through the thickness For example, Young’s modulus and thermal expansion coefficient both depend sensitively on the fibre orientation distribution and will vary through the depth accordingly A straightforward average can be expected for the overall values in the preferred fibre orientation direction but other properties may not be as easily derived The stiffness in bending is more sensitive

to the distribution in Young’s modulus It is given by the summation through the depth of the product of the second moment of area and the (depth-varying) Young’s modulus, i.e bjE(z)zdz where b is the width of the bar This is the key parameter when designing sandwich structures for load-bearing applications, for which the departure of its value from that calculated assuming a uniform average Young’s modulus is very great indeed In a similar way, a reinforced moulding with a high Young’s modulus near the surface is stiffer in bending than would be predicted if only the average Young’s modulus were available

Sharp changes in property occur whenever there is a sudden change in the fibre orientation distribution Thus in injection mouldings with a stratified morphol- ogy there will occur narrow boundaries between different regions with a significant mismatch of properties such as Young’s modulus and linear thermal expansion coefficient Under certain loadings or thermal conditions this may promote delamination and such a morphology should be avoided if possible Examples of measurements of depth variation in properties are given in Chapters 2 and 3

I .7 Conclusions Reinforcement of polymers using short fibres can lead to unique property combinations Even if the properties of the short fibre-polymer composite are not unique, reinforcement can produce significant property benefits at a modest cost

In most cases the composites are processed easily using conventional polymer or rubber processing methods Further improvements in property could be achieved by exercising more control over the fibre orientation distribution Variations in fibre orientation distribution commonly occur within mouldings,

particularly through the wall thickness This is unlikely to produce optimal

properties and may even lead to failure

References

1 Acosta J L, Morales E, Ojeda M C and Linares A, J Muter Sci, 21 (1986) 725

2 Acosta J L, Morales E, Ojeda M C and Linares A, Angew Makromol Chem, 138 (1986)

3 Morales E and White J R, J Muter Sci, 23 (1988) 4525

4 Murthy V M and De S K, Polym Eng Revs 4 (1984) 313

5 Murthy V M and D e S K , J Appl Polym Sci 29 (1984) 1355

6 Akhtar S, De P P and De S K, J Appl Polym Sci 32 (1986) 5132

7 Setua D K and De S K, Rubber Chem Tech 56 (1983) 804

8 Crosby J M, Chapter 5 in Thermoplastic Composite Materials, Ed L A Carlsson,

9 Folkes M J, Short Fibre Reinforced Thermoplastics, Research Studies Press/Wiley,

103

Elsevier, Amsterdam (1991) 139

Chichester (1982)

10 Darlington M W, McGinley P L and Smith G R, J Muter Sci, 1 1 (1976) 877

11 Bright P F , Crowson R J and Folkes M J, J Muter Sci, 13 (1978) 2497

12 D e Frahan H H, Verleye V, Dupret F and Crochet M J, Polym Eng Sci, 32 (1992) 254

13 MatsuokaT, Takabatake J-I, Inoue Y and Takahashi H, Polym Eng Sci, 30 (1990) 957

14 Matsuoka T, Polypropylene: Structure, Blends and Composites, Vol3: Composites, Ed J

15 Lockett F J, Plast Rubber-Proc, 5 (1980) 85

16 Bay R S and Tucker C L , 111, Polym Compos, 13 (1992) 317

17 Bay R S and Tucker C L, 111, Polym Compos, 13 (1992) 332

18 Advani S G and Tucker C L, 111, Polym Comp, 11 (1990) 164

19 Allan P S and Bevis M J, Plast Rubber Proc Applics, 7 (1987) 3

20 Gibson J R, Allan P S and Bevis M J, Plast Rubber Inr 16(5) (May 1991) 12

21 Murthy V M and De S K, Rubber Chem Technol, 55 (1982) 287

22 Campbell D and White J R , Angew Makromol Chem, 122 (1984) 61

23 Piggott M R, Load Bearing Fibre Composites, Pergamon, Oxford (1980)

24 Young R J and Love11 P A , Introduction to Polymers, 2nd Ed, Chapman and Hall,

25 Cox H L, Brit J Appl Phys, 3 (1952) 72

26 O’Donnell B and White J R, Plast Rubber Compos Proc Applics, 22 (1994) 69

27 Brody H and Ward I M, Polym Eng Sci, I 1 (197 1) 139

28 Curtis A C, Hope P S and Ward I M, Po1,ym Combos, 3 (1982) 138

29 Halpin J C and Tsai S C, Eflects of environmental factors on composite materials,

30 Sawyer L C and Grubb DT, Polymer Microscopy, Chapman and Hall, London, New

31 Bartosiewicz L and Mencik Z, J Polym Sci Polyrn Phys Ed, 12 (1974) 1163

32 Toll S and Andersson P - 0 , Composites, 22 (1991) 298

33 Clarke A, Davidson N and Archenhold G Proc Int Con! Transpuring ’91 vol 1,IOS

34 Bay R S and Tucker C L, 111, Polym Eng Sci 32 (1992) 240

35 Clarke A, Davidson N and Archenhold G , J Microsc, 171 (1993) 69

Karger-Koksis, Chapman & Hall, London (1995) 1 13

London (1991)

AFML-TR-67, (1969)

York (1987)

Press, California (199 1) 3 1

20 Short fibrepolymer composites

36 Hine P, Duckett R A, Davidson N and Clarke A R, Compos Sci Technol, 47 (1993) 65

37 Moginger B and Eyerer P, Composites, 22 (1991) 394

38 Fischer G and Eyerer P, Polyrn Compos, 9 (1988) 297

39 Clarke A, Davidson N and Archenhold G, Trans Roy Microsc SOC, 1 (Proc Conf Micro

40 Fakirov S and Fakirova C, Polyrn Compos, 6 (1986) 41

41 Vaxmann A, Narkis M, Seigmann A and Kenig S J, J Muter Sci Letts, 7 (1988) 25

42 Hermans P H, Contributions to the Physics of Cellulose Fibre, Elsevier, Amsterdam

43 Darlington M W and McGinley P L, J Muter Sci, 10 (1975) 906

44 McGee S H and McCullough R L, J Appl Phys 55 (1984) 1394

45 Ashby M F, Materials Selecfion in Mechanical Design, Pergamon Press, Oxford (1992)

'90, London), Adam Hilger, Bristol (1990) 305

(1946)

be taken into consideration when determining the commercial aspects of material selection

Injection moulding permits rapid production of components with complex shapes and reinforced thermoplastic mouldings are used in a wide variety of applications Examples of gears and drive shaft components are shown in Fig 2.1

2.2 Materials 2.2 I Polymers

Most thermoplastics are candidates for fibre reinforcement, although the end use has a significant influence on the list of common grades that have been made available commercially Fibre filled grades are opaque, and this excludes them from many applications Although products are often made from polystyrene because of its high transparency, its cheapness and excellent processing proper- ties have resulted in the development of glass fibre reinforced grades for other

22 Short fibre-polymer composites

2.1 Gear and drive shaft mouldings made from fibre reinforced thermo- plastics (Photograph courtesy of ICI Chemicals and Polymers Limited.) quite different applications Reinforced acrylics are not as well known Short fibre reinforcement is used to enhance the properties of other amorphous polymers including polycarbonate, poly(ether sulphone) and poly(ethy1ene terephthalate) Unfilled crystallizing thermoplastics are normally translucent or opaque and the addition of fibres is rarely restricted by optical property requirements The most common short fibre reinforced polymers are polypropylene and the nylons (mainly nylon 6,6 and nylon 6) For high performance, composites based on

poly(ether ether ketone) (PEEK) or poly(ether ketone) (PEK) are available To maintain good processing properties, with melt flow behaviour suitable for injection moulding, it is sometimes necessary to use a lower molecular weight than that preferred in an unfilled grade Finally it is noted that fibre reinforced thermotropic liquid crystal poiymers have been made available commercially The combination of the fibre reinforcement and the self-reinforcing property of the liquid crystal polymer produces a particularly stiff composite

2.2.2 Fibres

The moulding materials are normally supplied in the form of granules with the fibres already mixed into the polymer, usually at a concentration within the range 10-40% by weight, which corresponds very roughly to 5-20% by volume, depending on the densities of the polymer and the fibres The granules are normally cylindrical with length 3-5 mm and 3 mm diameter, and the fibres

may span their length or, more usually, are much shorter The fibre length becomes significantly degraded during pro~essing’-~ but the residual length still provides useful reinforcement A class of materials has been developed in which the granules and the fibres measure 10 mm in length, and, with some relatively simple modifications to the moulding conditions and the geometry of the flow path, it is possible to preserve quite high fibre lengths in the moulded products, giving superior reinforcement These compounds are often referred to as ‘long

fibre’ composites (e.g Verton, originally developed by ICI and now available

from DuPont) but this is simply to distinguish them from the more common short fibre composites and it is entirely appropriate to include them in this text in which

the term ‘short fibre’ is meant to exclude continuousfibre composites, but not to

restrict consideration to fibre lengths lower than some arbitrary value (say 5 mm)

The mouldings shown in Fig 2.1 are made from Verton and it is evident that the use of relatively long fibres does not prevent the manufacture of quite intricate shapes

Glass fibres are overwhelmingly the most common fibre reinforcement, both in short fibre composites and in ‘long fibre’ composites Their properties are good and they are available at consistent quality and reasonable cost Carbon fibres have higher stiffness and strength but their substitution for glass fibres produces only modest improvements in the properties of the composite because the matrix properties dominate Carbon fibre reinforcement is therefore used only in applications where the need for performance outweighs the need for cost control, such as aerospace components and specialized sports equipment There are no special advantages offered by ceramic fibres as reinforcement for thermoplastics, though the prospect of thermoplastics containing fibres made from high temperature superconductor ceramics is intriguing Metal fibres can be used to provide electrical conductivity’ (see Chapter 6) The use of polymer fibres is restricted by the desire to fabricate using melt processing There are many polymers with processing temperatures low enough to use fibres made from lyotropic liquid crystal polymers such as polyaramids (e.g poly(parapheny1ene

terephthalamide), Keular, or poly(parapheny1ene benzobisthiozole), PBZT),6-8

but the cost of these materials is rarely justified except when they are used as continuous fibres, when the maximum benefit can be obtained

2.2.3 Particulate reinforcement

The use of particulate reinforcement might appear to be outside the scope of this book, but in the case of thermoplastics it demands consideration because particulate reinforcement is often used for the same reasons as fibre reinforcement and the materials produced offer similar properties and often compete in the market-place In some cases the particulate is fibrous, as for example, sepiolite, a clay-like mineral which has been investigated as a filler for p o l y p r ~ p y l e n e ~ * ’ ~ Other minerals such as talc and mica come in the form of platelets and share with short fibres the ability to align during moulding as a consequence of flow, and produce a generally similar anisotropic reinforcement Mouldings reinforced

24 Short fibrepolymer composites

with glass fibre tend to be prone to warping but this is less of a problem with polypropylenefilled with mica;' ' the inclusion of mica in combination with fibres can help to counteract warpagẹ12 Composites made from equiaxed particulates such as chalk generally display less reinforcement and are less anisotropic than unfilled injection mouldings, but particulate fillers are effective in improving dimensional stability and resistance to elevated temperatures

2.2.4 Coupling agents

Coupling agents are frequently employed to improve the adhesion between the filler and the polymer Some of the basic chemistry of coupling systems is dealt with in Chapter 3 (section 3.7.1) For some combinations of filler and polymer the composition of the coupling agent (and hence its detailed action) is protected by commercial secrecỵ The most popular group of coupling agents are the silanes which can bond both to the inorganic glass or mineral filler and to the organic polymer The filler surface usually has hydroxide groups to which the coupling agent bonds, forming a bridge across to the polymer matrix (see section 3.7.1) Acrylic acid is often used to improve fibre-matrix adhesion13 and a combination

of polypropylene with acrylic acid grafted on to it with silane-treated mica has been studied by Chiang and Yang.14

The inclusion of a coupling agent normally produces a significant improve- ment in stiffness and strength The improvement in flexural strength is typically from 10 to 90%." One of the main causes of weakening of fibre composites is the destruction of polymer-matrix adhesion by the ingress of water, and particular attention has been paid to developing coupling agents that resist this a t t a ~ k ' ~ The effectiveness of a coupling agent is quite sensitive to the method used to fabricate the compositẹ The improvement in property is much greater with compression moulding than with injection moulding.' This is not simply because the injection moulding has better properties to begin with, but is largely because the high shcar conditions that prevail during moulding discourage the formation of the interfacial bonds Ionomeric coupling agents have been developed that partly overcome this problem "

Coupling agents are also used to advantage with particulate fillers, including mica, talc and glass flakệ'^.'^ * O

2.3 Fabrication - injection moulding

2.3 I Introduction

The vast majority of articles made from fibre reinforced thermoplastics are produced by injection moulding Most injection moulding machines are of the reciprocatingscrew type in which the polymer is melted in a barrel within which is

an Archimedian screw The material in the form of granules is fed into the barrel via a hopper and falls on to one end of the screw The flights at this end of the

screw are deep to permit conveyance of solids and as the screw turns the material

is moved forwards into a zone heated by electric resistance heaters attached around the outside of the barrel The polymer melts and further heating is obtained from mechanical work as the screw turns through the viscous melt At the forward end, the shank of the screw has a larger diameter so that the channel defined by the flights is shallower and the work done on the material intensifies This assists mixing but also increases fibre length degradation As material is fed forward the screw is allowed to move backwards along the barrel axis and a charge of homogenized melt gathers at the front end of the barrel The resistance

to the motion of the screw during this part of the process, or screw-back pressure, can be adjusted and is likely to influence the degree of fibre degradation When sufficient material is present to fill the mould cavity plus the runner system the screw is thrust forwards and, acting as a ram, propels the melt into the mould via a nozzle that is held tightly against the entrance to the mould The injection speed, or rate at which the screw is thrust forward, is normally under the control of the operator and is another parameter that may influence fibre degradation The material in the mould must now be allowed to cool until it is sufficiently solidified The force on the screw is maintained for a significant fraction of the cooling time so that thermal shrinkage of the melt can be combated and the mould is kept ‘topped up’ This is often most satisfactorily achieved using

a holding pressure that is less than the injection pressure The melt is admitted into the mould cavity from the runner system through a narrow constriction called a gate (or a set of gates in the case of large mouldings or mouldings with complicated shapes) Once the material in the gate has frozen, no more topping-up can occur, the holding pressure can be released, and the screw can start to turn again and prepare the next charge Further cooling time must be allowed to elapse before the moulding has solidified sufficiently to be ejected The common commercial grades of fibre reinforced thermoplastics are designed for use in standard injection moulding machines and work satisfactorily

at temperatures, pressures, injection rates and timings similar to those used with unfilled thermoplastics There are differences between fibre reinforced materials and unfilled thermoplastics, however, and these must be heeded Firstly, it should

be noted that the fibres are generally abrasive and cause wear of the injection moulding machine and tool; this should be taken into account when selecting the material for a product, though the choice is usually dictated by the property requirements of the product, and the end-use properties are normally very different for unfilled and fibre filled materials Secondly, it is prudent to attempt

to minimize fibre length degradation by appropriate design of the tool and selection of operating conditions This is particularly important when moulding

‘long’ (10mm) fibre reinforced grades, and is discussed later

2.3.2 Morphology in injection moulded and fibre filled thermoplastics

2.3.2 I Introduction

Unfilled injection moulded thermoplastics generally contain a multilayered structure Injection mouldings made from glassy polymers usually possess a

26 Short fibre-polymer composites

skin-core morphology with a surface region of the order of 0.1-0.4 mm thick that

is quite different from the interior The reason for this is that the surface cools rapidly when the melt contacts the mould cavity wall, whereas the material in the interior cools much more slowly because of the low thermal conductivity of the polymer The cooling rate near the surface is normally too fast to permit much recoil of the molecules that have become oriented during the mould filling process Thus, in the skin, much of the flow-induced molecular orientation is

retained in a 'frozen-in' state As long as the moulding remains below the polymer

glass transition temperature, T,, there will be very little change in the conforma- tion of the molecule backbone, though some localized reorganization may occur, leading to density changes In the interior, cooling takes place sufficiently slowly

to allow the molecules to recoil significantly, though not necessarily completely, and the core is characterized by a much lower degree of orientation

The skin-core morphology is generally more evident in semi-crystalline polymers than in non-crystalline polymers This is because the differences in molecular orientation through the depth of the moulding lead to different crystal morphologies, and sometimes even different crystal structures in the case of a polymorphic polymer such as polypropylene Near the surface the pronounced molecular orientation usually causes a preferred orientation in the crystals that grow there In the interior, however, the molecules recoil before they cool sufficiently to form crystals and when crystallisation does eventually commence the melt is effectively isotropic, and, in unfilled polymers, equiaxed spherulites form The spherulite size may vary with cooling rate and may change through the depth of the moulding In addition to the two main regions so identified, there is often evidence of the presence of a third located right at the surface, which is indicated to be amorphous, having cooled too rapidly to form crystals In some studies, more than three distinct regions have been found.2'*22 More references

to the skin-core morphology in unfilled polymers are listed by Chen and White.23

Reinforcement can influence the morphology in three different ways Firstly, the fibres or particulates modify the flow properties of the polymer and as a consequence may modify the morphology of the polymer itself Secondly, in the case of crystallizing polymers, the additive may act as a crystal nucleant, influencing the crystal size and (local) orientation Finally the distribution of the additive itself within the moulding (if segregation occurs) and the variation of its orientation distribution (if the additive is not equiaxed) has a marked influence over the morphology and properties

Sometimes a skin can be identified that corresponds closely to the skin in the unfilled polymer In the case of fibre reinforced injection mouldings a stratified layer morphology is found to occur Various characteristic layer types have been observed including (i) regions in which the preferred orientation of the fibres is parallel to the melt flow direction; (ii) regions in which the preferred orientation is transverse to the melt flow direction; (iii) regions in which the fibres tend to lie in planes parallel to a surface of the moulding but do not display any preferred orientation within the plane; (iv) regions in which the fibre

orientation is random in three dimensions This is reviewed below Plate-like fillers such as mica and talc can show similar features but have not been studied

by identifying the colour produced when using white light, then machining away

a thin layer and measuring the relative retardation of the remainder This process

is repeated until the remainder is too thin to handle A plot of relative retardation

versus depth removed is then generated and the birefringence at any chosen depth

is given by the gradient of this plot.24-27 The source of the birefringence in an injection moulding treated in this manner will normally be molecular orientation, though residual moulding stresses will also contribute The problem of separat- ing the two contributions is discussed e l ~ e w h e r e ~ ~ - ~ '

Although semi-crystalline polymers are birefringent the observation of fringe patterns when viewed between crossed polars is only possible with thin sections because of the light scattering that occurs as the result of the presence of two phases (amorphous and crystalline) of very different refractive indices Thin sections are normally prepared using a microtome and the transition between the spherulitic core and the oriented skin is usually very clear when the section is viewed between crossed polars in a low magnification microscope The skin-core boundary is often visible on fracture surfaces, both in the light microscope and in the scanning electron microscope The most important morphological variation

is crystal orientation and is best investigated by X-ray techniques Thin sections can again be used to isolate a chosen region For the most comprehensive analysis, pole should be constructed for several prominent crystal direct ions

With filled polymers, light microscopy of a polished section is again a valuable technique, but the preparation of satisfactory sections is difficult because of the vast difference in hardness between the polymer and the filler Thin sections for transmission microscopy are particularly difficult to make The layer structure is normally well defined once a satisfactory section has been prepared

Observation of the skin-core morphology

28 Short fibrepolymer composites

2.2 Fibre cross-sections a t different depths of a glass reinforced nylon 6,6 moulding: (a) near to the surface, where fibres are predominantly parallel to the flow direction and lie mainly normal t o the section, presenting nearly circular profiles; (b) in the interior in a region in which many fibres lie a t oblique angles to the section (Micrographs courtesy of B O’Donnell.)

2.3.2.3 Fibre orientation distribution

The fibre orientation distribution has a strong influence over the properties of the material The reinforcement provided by each individual fibre depends on the orientation with respect to the loading axis, while electrical and thermal properties are likewise affected A short review of theoretical predictions of fibre orientation distribution is-given in Chapter 1, section 1.3 These are required for computer modelling in which the objective is to design the mould geometry to give a component with the target properties For the purposes of testing the validity of theoretical predictions of fibre orientation distribution and of the