Polymer Matrix Composites: Matrices and Processing$

CJG Plummer, P-E Bourban, and J-A Månson, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland
r 2016 Elsevier Inc. All rights reserved.

1 Matrix Materials 1
1.1 Thermoset Resins 2
1.1.a Unsaturated polyesters 2
1.1.b Epoxy resins 2
1.1.c High-performance thermosets 2
1.2 Thermoplastic Resins 3
1.2.a Semicrystalline thermoplastics 3
1.2.b Amorphous glassy thermoplastics 3
2 Processing 3
2.1 Fibers, Prepregs, Feedstocks 3
2.2 Shaping and Forming Techniques 4
2.2.a Press molding 4
2.2.b Vacuum and autoclave molding 4
2.2.c Resin transfer molding, reaction injection molding 5
2.2.d Pultrusion 5
2.2.e Filament winding 5
2.2.f Resin infusion 5
2.2.g Other emerging technologies 6
2.3 Choice of Materials and Processing Routes: Composite Processing Maps 6
2.4 Cost Modeling, Life Cycle Engineering 6
3 Process–Structure–Property Relationships 7
4 Conclusion 8
References 8

Polymer matrix composites, and fiber-reinforced plastics (FRPs) in particular, are commonplace in the industrialized world. Their
applications range from primary structural aircraft components to tennis racquets, and many household appliances incorporate
some form of fiber reinforcement. The development of engineering FRPs began once significant quantities of glass fiber became
available in the 1940s, and was initially performance driven (e.g., for aerospace). Since then, reduction of materials costs has
greatly widened their range of applications, and the impetus for continued development stems mainly from economic factors,
such as efficient fabrication and transformation, and a need for the energy savings possible with low-weight, high-strength, and
high-modulus materials. However, FRPs have the major disadvantage with respect to metal- and ceramic-based structural materials
that their service temperatures are currently limited to below 400 1C, even for the most advanced matrices.
An overview is given in this article of the most widely used FRP matrix materials and their relative merits. The various stages of
processing to form composite parts are discussed, with reference to cost forecasting, life cycle engineering, and developing
technologies. Finally, the important question of process–structure–property relationships in thermoset- and thermoplastic-based
FRPs is addressed.

1 Matrix Materials

Although there has been a substantial increase in the use of thermoplastic matrices beyond the well-established market for
injection molding thermoplastics, which typically contain short (o1 mm) fibers, most FRPs continue to be based on thermoset
resins. Thermosets can be cured catalytically, thermolytically, or photolytically over a wide range of temperatures to give highly
cross-linked solids. One of their principal advantages for FRPs is their low initial viscosity, which facilitates fiber impregnation,
particularly where the base ‘A-stage’ resin consists of monomeric or oligomeric components. Partial cure may then be used to
produce solid ‘B-stage’ prepregs used as precursors for a variety of processing routes. There is nevertheless great interest in
thermoplastic polymers. Although the high molar mass of thermoplastics implies that higher processing temperatures or spe-
cialized techniques are needed for impregnation, this may be outweighed by other factors: (1) they are generally nonreactive,
implying an indefinite shelf life for precursors; (2) forming processes are reversible, facilitating recycling, and repair; (3) cycle times

☆
Change History: September 2015. C.J.G. Plummer updated the text and references to this entire article, and modified Figures 2–4.

Reference Module in Materials Science and Materials Engineering doi:10.1016/B978-0-12-803581-8.02386-9 1

2 Polymer Matrix Composites: Matrices and Processing

Table 1 Property and process characteristics of thermosets and thermoplastics

Property Thermoset Thermoplastic

Modulus High Medium

Service temperature High Medium
Toughness Medium High
Viscosity Low High
Processing temperature Low High
Recyclability Limited Good

are not limited by reaction kinetics, and mixing steps, which may involve toxic precursors and need careful control to ensure
reproducibility, are eliminated.
Table 1 summarizes the main trends, although it should be stressed that there is considerable overlap, the toughness of FRPs
based on both classes of resins being highly sensitive to fiber morphology, interfacial adhesion, and the presence of additives.
Moreover, each class of resins encompasses a wide variety of materials and materials properties, which will now be presented in
more detail along with typical applications.

1.1 Thermoset Resins

There are several main types of thermosets, distinguished by their chemistry, curing mechanisms, and final properties (Charrier,
1990). Those discussed here are the most important for FRPs, namely polyesters, epoxies, and high-performance thermosets,
although other thermosets, including polyurethanes (PUs), silicone rubbers, and phenol–formaldehydes (PFs), are also sometimes
used as FRP matrices (e.g., PF–aramid composites for aircraft interiors). Although most of these resins are petroleum-based, there
is increasing interest in wholly or partly bio-sourced alternatives, including epoxy and polyester resins (Zini and Scandola, 2011).

1.1.a Unsaturated polyesters

Unsaturated polyesters (UPs) are the most widely used thermosets for FRPs thanks to their low cost and ease of handling. They
consist of prepolymers cross-linkable by addition, initiated photolytically or by thermal decomposition of peroxides. Processing
involves most of the methods to be described in the next section, and indeed UPs were originally developed for use in FRPs.
However, to ensure adequate flow behavior, up to 50 wt.% styrene co-reagent is often added. Environmental concerns have led to
the use of low styrene emission additives in UPs. These segregate to, and seal the surface during the final stages of cure,
compromising adhesion in some cases. Another potential difficulty is excessive shrinkage, but this is largely avoided by the use of
thermoplastic low profile additives. Once cured, UPs have heat deflection temperatures (HDTs) ranging from 80 up to 150 1C in
certain vinylester resins (the HDT is a softening temperature measured according to industrial norms and marks the approximate
upper limit to the range of service temperatures). Vinylester resins are generally more expensive than UPs but are more durable and
resistant to water and chemically aggressive environments. The main composite applications of UPs and vinylesters are to be found
in the construction and transport industries (boats, car body panels), furniture, and chemical storage tanks.

1.1.b Epoxy resins

Epoxy resins may also be used in conjunction with most composite processing techniques and are particularly suited to B-stage
processing. Cure generally involves polyaddition with a multifunctional co-reagent, referred to as the ‘hardener,’ to produce a
three-dimensional network with low shrinkage, and excellent adhesion and environmental resistance. Drawbacks of conventional
resins include somewhat higher viscosities than for UPs, long cure times, and moisture uptake. A wide range of resins and
co-reagents is nevertheless available, products ranging from tough low-temperature epoxies to more brittle, high-temperature
epoxies for advanced FRPs (prepregs for aircraft components account for around 30% of total production). Commonly used
aliphatic amine hardeners permit room temperature cure, but the reaction is strongly exothermic, causing problems in thick
moldings. Aromatic amine hardeners require higher curing temperatures but are better suited to large parts and give relatively
high HDTs (up to 230 1C). Anhydride hardeners are less toxic and less polar than amines, but need an accelerator. With
suitable catalysis, epoxies may also be cured thermally or photolytically without a co-reagent.

1.1.c High-performance thermosets

High-performance thermosets include allyls, cyanate esters, benzoxazine, bismaleimide, phthalonitrile, and liquid crystalline
thermosets. Among the most successful are addition polyimides, developed mainly for the aerospace and electronics industries.
Their HDT may reach 400 1C after cure, and they have excellent stiffness and environmental resistance. The cure mechanism is
thermally induced free radical addition of a polyimide precursor via reactive end groups in the presence of peroxide, avoiding
problems with the volatile by-products of conventional condensation reactions. Applications include jet and automobile engine
 Polymer Matrix Composites: Matrices and Processing 3

components, heat-resistant panels, etc. However, some potential future applications of high-performance FRPs, such as reusable
launch vehicles, will require HDTs in excess of 500 1C, setting the standard for the next generation of thermoset matrix materials.

1.2 Thermoplastic Resins

Both semicrystalline and amorphous glassy polymers form viscous melts that need to be processed at temperatures well above the
glass transition temperature, Tg. However, because semicrystalline polymers melt well above Tg, it is the melting point, Tm, that
defines the processing window. Whereas thermoset designations tend to be generic, most thermoplastics have well-defined linear
architectures that undergo little change during processing. Their property profiles are generally adjusted for different processing
routes by varying the molar mass or additive package (Charrier, 1990). However, it should be borne in mind that other important
properties such as toughness and ultimate tensile strength are also highly sensitive to changes in molar mass. As with thermo-
plastic resins, there is substantial interest in thermoplastic bioresins, the most important of these being polylactides and various
other bio-sourced polyesters (Zini and Scandola, 2011).

1.2.a Semicrystalline thermoplastics

Semicrystalline thermoplastics are tough and resistant to organic solvents, but many inexpensive and easy-to-process low Tg
polymers, such as polyethylene, also show poor adhesion owing to their lack of functionality, and low softening temperatures,
often associated with excessive creep. A good compromise between cost and performance is provided by polypropylene (PP,
TmE170 1C), for which functionalized additives are available to improve adhesion. PP is widely used in FRP applications in the
automotive industry, for example, owing notably to its use in glass mat thermoplastic (GMT) technology. Polyamides (PAs or
nylons), such as PA6 and PA66, are also popular, having higher HDTs than most thermosets, although they are moisture sensitive
and more than twice as expensive as PP. Another polyamide of interest for FRPs is PA12, whose versatile synthesis and func-
tionality make it especially attractive for reactive FRP processing, in which the thermoplastic matrix is polymerized in situ after
impregnation of the fiber bed with the low-viscosity monomer, thus avoiding difficulties associated with the liquid processing of
highly viscous molten thermoplastics. This approach has also been used in conjunction with anionically polymerized PA6 and
cyclic butylene terephthalate monomer (CBT), for example (van Rijswijk and Bersee, 2007).
High Tg semicrystalline polymers such as poly(phenylene sulfide) (PPS) and poly(ether ether ketone) (PEEK) require corre-
spondingly high processing temperatures, PEEK melting at about 345 1C, for example. These polymers have outstanding prop-
erties, including exceptional toughness and fatigue resistance. Particular effort has gone into PEEK–carbon fiber composite
technology, leading to the so-called aromatic polymer composite (APC) prepregs, considered to be a major step in the devel-
opment of advanced thermoplastic matrix FRPs (Cogswell, 1992). However, PEEK is nearly two orders of magnitude more
expensive than PP (Charrier, 1990).

1.2.b Amorphous glassy thermoplastics

Aromatic high-performance amorphous thermoplastics such as poly(ether sulfone) (PES, TgE230 1C) or poly(ether imide) (PEI,
TgE215 1C) have also attracted interest as advanced FRP matrix materials. They are less expensive than PEEK and tougher than
thermosets, but there remain concerns over their solvent resistance (Penn and Muzzy, 1989).

2 Processing

Figure 1 summarizes the various routes by which raw materials (resin plus fiber) are transformed into the final part (Månson,
1994). A description of the most important of these now follows.

2.1 Fibers, Prepregs, Feedstocks

Fibers are usually supplied in the form of rovings (glass fibers) or tows (carbon fibers). Rovings consist of straight continuous glass
fiber strands or bundles of about 200 filaments; the number of strands depends on the end use, and these may be several
kilometers long. Tows are likewise available in various configurations. The fibers are typically sized (coated) during production to
promote wetting and adhesion, silane coupling agents being widely used for glass fibers. They may be woven into fabrics if
required, including ad hoc fiber preforms for specific applications. Chopped fiber lengths range from less than 1 mm for injection,
to around 50 mm in randomly oriented mats for laminates.
Of major concern in FRP processing are trapped air and volatiles, because voids substantially reduce the final strength. Low-
viscosity resins, vacuum outgassing, and consolidation all help reduce porosity. Use of preconsolidated sheets or prepregs also
facilitates certain processes (Figure 1). These are available as impregnated fabrics or continuous warp sheet typically containing
60–65 vol.% uniaxially oriented fibers. Impregnation with thermosets may be achieved by dipping or hot melt processing
(rolling of the fibers together with a resin film). When the viscosity is high, flow distances must be kept to a minimum, and
4 Polymer Matrix Composites: Matrices and Processing

Figure 1 Alternative manufacturing routes: from resin and fiber to finished part (after Månson, 1994).

intimate mixing of fine polymer powders or fibers with the fiber reinforcement often precedes melt processing of thermoplastic
composites. Solvent processing is one alternative; it may lead to problems with residual solvent in the later stages of fabrication
but has been used with success for PEI prepregs, for example. Another possibility is to chain extend relatively low molar mass
precursors in situ by reaction of suitable end groups, an approach frequently used for high-performance thermoplastic FRPs.
Sheet and dough molding compounds are mostly based on unsaturated polyester resins containing filler and chopped fibers.
Sheet molding compounds (SMCs) are produced from chopped strand mat placed between layers of filled resin containing
thickening agents to promote moldability. Impregnation involves rolling of the resulting sandwich structure. SMCs usually contain
25–35 wt.% fibers but the fiber content can be up to 65 wt.%, in which case they are called HMCs. XMCs are similar to SMCs, but
incorporate surface layers containing continuous fiber rovings. Dough molding compounds (DMCs) or bulk molding compounds
(BMCs), contain shorter fibers than SMCs and are produced by mechanical mixing, which results in a more markedly three-
dimensional fiber orientation distribution. Glass mat thermoplastics (GMTs) can be produced in a similar manner to SMCs, but
the high matrix viscosity means that success is very dependent on the choice of fiber mat. Powder processing routes are an
interesting alternative in this case.
It is sometimes advantageous to preassemble reinforcing fibers to fit a given shape, rather than rely on flow to distribute them
throughout a mold. If displacement of the fibers during subsequent introduction of the resin is a problem, they can be anchored in
place by spraying with a binder, or substituted by a pre-impregnated (prepreg) fabric shaped into a suitable perform.

2.2 Shaping and Forming Techniques

The simplest forming technique is hand placement, where a dry woven fabric or chopped strand mat is placed on a nonstick
former and premixed liquid thermoset resin added with a brush or by spraying, although it is also possible to place a prepreg
(wet layup). A roller may then be used to compress the laminate to consolidate the fibers and remove trapped air (debulking).
Hand placement is versatile and requires low investment, but the resulting mechanical properties are limited and the manufacturer
may fall foul of legislation against volatiles and hazardous air pollutants such as styrene. It is widely used for repair or for one-off
or short production runs, including large objects, such as boat hulls. However, for reproducible, high-quality, cost-effective
production other techniques must be considered.

2.2.a Press molding

Press molding using matched male and female tools is the most widely used technique for volume production of FRPs, being
suitable for most types of precursor. With SMCs or DMCs, complex shapes can readily be formed at pressures of 3–15 MPa with
cycle times of the order of minutes depending on the cure rate and the part size. It is also used for continuous fiber FRPs, but the
limited flow restricts the practically attainable range of wall thickness and curvature. For mass production, steel tools and hence
additional capital investment may be necessary, but for short runs, epoxy composite tools are often adequate. Variations include
rubber-pad press molding, hydroforming, diaphragm forming, and cold stamping.

2.2.b Vacuum and autoclave molding

In vacuum molding, atmospheric pressure consolidates the material during cure, avoiding the need for a press. A prepreg or fabric
laminate is placed on a mold, covered with a flexible membrane or vacuum bag (VB), and the assembly sealed and evacuated.
It provides a low-cost alternative to press molding for large, thin shells. If further consolidation is needed, cure may be done in an
autoclave. Autoclaves usually operate at up to 1.5 MPa pressure and are widely used in the aerospace industry, having the
 Polymer Matrix Composites: Matrices and Processing 5

advantage over press molding of cheap tooling for large parts and being ideal for complex structures such as honeycomb core
sandwich composites. Indeed almost all prepreg material used in the aerospace industry is autoclaved. It is less suited to high-
volume manufacture, although progress in automatic cutting and layup operations has led to considerable cost reductions and
new market opportunities.

2.2.c Resin transfer molding, reaction injection molding

In resin transfer molding (RTM), the fiber reinforcement (either dry fibers or a preform) is placed in a closed mold cavity, and the
resin then introduced by injection. RTM generally requires low-viscosity resins, but impregnation can be improved by vacuum-
assisted injection (although the widely used term ‘vacuum-assisted resin transfer molding’ (VARTM) usually refers to vacuum
infusion without positive pressure – see Section 2.2.e below). The consequent reduction in mold closure pressure makes the
process more suitable for large moldings such as car body shells, and implies that the molds themselves may be made from lighter,
cheaper materials, with considerable cost savings (light RTM). Another recent development is to inject into a partially closed mold
and complete the cycle with a compression step, allowing not only increased injection speeds but also higher fiber volume
fractions in the finished part. However, although it is cheaper than prepreg processing, RTM remains more limited in its range of
application.
Reaction injection molding (RIM) makes use of rapidly reacting components mixed just prior to injection using a high-pressure
impinging mixer, permitting very short cycle times, and short fiber reinforcement may also be introduced in one or both reagents
(reinforced reaction injection molding (RRIM)). It is competitive with press molding for the manufacture of a variety of small to
medium sized automotive and domestic components.

2.2.d Pultrusion
Pultrusion is the continuous fiber FRP equivalent of extrusion, permitting continuous production of constant profile parts such as
fishing rods, antennas, and channel profiles (or constant section parts, if the profile is modified in situ). Unlike in extrusion,
however, the reinforcing fibers themselves are used to pull the material through a heated forming die, as shown schematically in
Figure 2. Rovings or tows are passed through a resin bath followed by a series of carding plates or ‘preformers,’ which force the
resin into the fibers. The impregnated fibers are then pulled through the heater die, where the resin gels and cures. Preheating by
radio-frequency methods may be used to increase throughput, resulting in production rates of up to several meters a minute.
In some cases the resin is also injected directly into the die, eliminating the need for a preimpregnation step.

2.2.e Filament winding

Strong, stiff shell structures can be made by winding bands of continuous fibers impregnated with resin onto mandrels, which are
removed after cure. Filament winding is widely used for the manufacture of reinforced plastic pipe and other cylindrical objects.
However, modern filament winding techniques can also be used to produce complex structures such as aircraft fuselages. In such
cases, the fibers must be wound along constantly changing non-repetitive geodesic paths to prevent slippage. Automation of this
technique has led to the development of advanced computer-based integrated design and manufacturing methods.

2.2.f Resin infusion

Vacuum infusion (VARTM) or ‘resin infusion with flexible tooling’ (RIFT), and its many variants (see Beckwith (2007) for a
detailed discussion of the terminology), is a recent development, in which a vacuum is used to draw resin into a preform or fiber
bed placed on a rigid or flexible mold covered with a plastic bagging film, have been successfully applied to large objects such as
boat hulls and wind turbine blades, and are currently under development for aerospace structures. In resin film infusion (RFI),
B-stage resin films are stacked with dry fabrics, greatly reducing the infusion distances during subsequent processing (Marsh,
2002). Finally, in membrane assisted low pressure infusion (VAP), a flexible microporous membrane that is impermeable to the
resin may be used to exert a more uniform vacuum over the entire part area, resulting in faster infusion and reduced porosity
(Filsinger et al., 2001).

Figure 2 Schematic of the pultrusion process.

6 Polymer Matrix Composites: Matrices and Processing

2.2.g Other emerging technologies

Large injection molding body technology (LIMBT), offers a viable alternative to press molding of SMCs and GMTs, contrary to the
widely held notion that injection molding facilities become prohibitively expensive as part sizes increase. It has been used, for
example, to make large automotive body components from various thermoplastic FRPs (Argeropoulos et al., 1999). Integrated
processing (IP) is aimed at improving manufacturing flexibility and adding value to composite products by combining different
materials and automated processes such robot-assisted placement; selective positioning of reinforcement and integration of
functions are thus achieved in short cycle times (Månson et al., 1995; Bourban et al., 1998, 2001). Moreover, automated tape lying
(ATL) and automated fiber placement (AFP) have themselves undergone significant development in recent years, enabling new
technologies that go well beyond the automatization of traditional hand layup techniques (Lukaszewicz et al., 2012).
Developments in reactive thermoplastics processing using monomeric precursors for impregnation followed by polymerization
in situ, have enabled extension of liquid molding techniques such as RTM and VARTM to thermoplastics (Schmid and Eder, 1997;
Bourban, 2000; van Rijswijk and Bersee, 2007). There has also been increasing interest in natural fiber-reinforced composites and
hence the particular constraints associated with hydrophilic and temperature sensitive materials, such as the need for moisture
removal during the manufacturing cycle (Summerscales and Grove, 2014).

2.3 Choice of Materials and Processing Routes: Composite Processing Maps

Figure 3 shows a map of various composite processing methods arranged according to the combination of resin viscosity and
reinforcement aspect ratio to which they are most suited. If viscosity is an overriding consideration it will generally dictate the
choice between thermoplastics and thermosets, but it is immediately clear from Figure 3 that the range of processes is broad
enough for solutions to be available for thermoplastics or thermosets for any required reinforcement aspect ratio. Such ‘composite
processing maps’ may be based on other parameters, such as the potential for control of reinforcement orientation and aspect
ratio; injection molding, for example, gives little orientation control and is limited to low aspect ratios, whereas three-dimensional
control and long fiber orientation require techniques such as filament winding, hand lay-up, and RTM/preform technologies.
Figure 4 gives a further example that is particularly important for costing (see below), namely the relationship between the
suitability of a given process route for complex shapes and its suitability for high-volume production, as defined by the cycle time.

2.4 Cost Modeling, Life Cycle Engineering

Because material costs are now substantially irreducible, and given that manufacturing often represents close to 60% of expen-
diture on FRP composite parts, the processing route is generally the most critical factor in cost-effective production. Assessing the
economic viability of a processing route requires comparative cost analysis and hence reliable data and cost estimation methods.
A variety of approaches for evaluating manufacturing costs have been proposed (Månson et al., 2000): comparative studies,
process-oriented cost models, parametric cost models, and process flow simulation. Comparative studies are based on historical
data and may suffice when envisaging the use of a standard technique for a component similar to those already in production.

Figure 3 Composite processing map: viscosity and reinforcement aspect ratio (after Månson, 1994).
 Polymer Matrix Composites: Matrices and Processing 7

Figure 4 Composite processing map: complex shaping and volume production (after Månson, 1994).

Process-oriented cost models are better adapted to new processes, and incorporate identification and quantification of cost drivers,
but require detailed process information and a precise definition of the components.
Parametric models are more versatile and allow easy manipulation of process and economic factors; however, they are based on
the assumption of independent manufacturing steps, and may consequently underestimate costs. Such difficulties may be avoided
by using process flow simulations, which take into account interactions between the different manufacturing steps, and provide
forecasting tools that are also suitable for improvement studies, identifying bottlenecks, and estimating repair costs. Given the
growing concern for environmental issues, the successful introduction of a new product or process may also require establishment
of a life cycle plan, particularly for the large-volume applications. As well as the impact of material on the environment during
manufacture, service, and disposal, it is important to consider conservation of raw material value. This is a potentially important
issue when choosing between thermosets and thermoplastics, because recycling of the former is severely limited. Furthermore,
recovery from continuous fiber FRPs is difficult even for thermoplastic matrices. One alternative under consideration is recovery of
the monomer via tertiary recycling.

3 Process–Structure–Property Relationships

The thermal cure history shown in Figure 5(a) is typical of prepreg and autoclave processing sequences for a thermoset FRP. During
impregnation, the resin viscosity first decreases owing to the rise in temperature, then increases as cross-linking proceeds.
The reaction is interrupted by cooling, and the resulting B-stage material is stored at low temperature to inhibit further reaction.
During final processing, the viscosity again passes through a minimum, tending to infinity as the gel point is approached. A two-
step final cure is often used to improve consolidation and dimensional stability, after which the part is cooled and removed from
the mold or former. The final properties of thermoset FRPs are strongly dependent on the details of such time–temperature
trajectories. Figure 5(b) shows a schematic time–temperature–transformation (TTT) diagram illustrating the main trends (Pang and
Gillham, 1990). During isothermal cure at temperatures between Tg and Tg1 , the polymer gels and then vitrifies, defining
gel

optimum conditions for final processing. Isothermal cure between Tg1 and Tg0 , the initial glass transition temperature (or melting
point), would cause the polymer to vitrify prior to gelation, hindering further reaction.
In a typical processing cycle for a thermoplastic FRP, the components are first heated to above Tm of the resin (or Tg for a fully
amorphous resin). Pressure is applied for long enough to give adequate consolidation, and the resulting composite is cooled with
the pressure maintained. Optimum processing conditions may be defined with the aid of processing windows, such as that shown
schematically in Figure 6. The lower bound for the cooling (and heating) rate is fixed by economic constraints on the cycle time.
Slow cooling and heating may also result in prolonged exposure of the resin to high temperatures, leading to thermal degradation.
At very high cooling rates, however, considerable thermal gradients may develop through the thickness of the component, leading
to pronounced skin–core morphologies in semicrystalline resins, and excessive thermal stresses, as well as changes in primary
morphological parameters such as spherulite size, lamellar thickness, and the degree of crystallinity. Increasing the hold pressure
may help suppress defects resulting from internal stresses, but a maximum pressure will be imposed either by the performance of
the press, or by the cost of high-pressure molds. High pressure may also damage the fiber bed and/or cause resin starvation.
8 Polymer Matrix Composites: Matrices and Processing

Figure 5 (a) Typical thermal cure history for a thermoset composite. (b) Isothermal TTT diagram associated with the cure of a thermoset
polymer.

Figure 6 Generic processing window for a thermoplastic composite (after Månson, 1994).

4 Conclusion

Although the initial driving force for the development of FRP composites was performance, economics is now the overriding
consideration. Raw material costs are generally higher than for metals, but this is offset by weight reduction, increased design
freedom, and reduced ‘part count’ in a given application; even image can be a strong selling point for sports goods. Nowadays, it is
innovation in processing that offers the greatest scope for increasing the engineering viability of polymer matrix composite
materials, and for paving the way to new applications of these materials.

References

Argeropoulos, J.G., Fielding, C.R., Oswald, L.J., 1999. SAE 1999-01-3244.

Beckwith, S.W., 2007. SAMPE J. 43 (1), 61; 43 (3), 46; 43 (4), 6 and 66−70.
 Polymer Matrix Composites: Matrices and Processing 9

Bourban, P.-E., 2000. Liquid molding of thermoplastic composites. In: Kelly, A., Zweben, C. (Eds.), Comprehensive Composite Materials, vol. 2. Oxford: Pergamon,
pp. 965–978.
Bourban, P.-E., Bernet, N., Zanetto, J.-E., Månson, J.-A., 2001. Compos. Part A 32, 1045–1057.
Bourban, P.-E., Bögli, A., Bonjour, F., Månson, J.-A., 1998. Compos. Sci. Technol. 85 (5), 633–637.
Charrier, J.-M., 1990. Polymer Materials and Processing. Munich: Hanser Verlag.
Cogswell, F.N., 1992. Thermoplastic Aromatic Polymer Composites. Oxford: Butterworth-Heineman.
Filsinger, J., Lorenz, T., Stadler, F., Utecht, S., 2001. German Pat. WO 01/68353 A1.
Lukaszewicz, D.H.-J.A., Ward, C., Potter, K.D., 2012. Compos. Part B 43, 997–1009.
Månson, J.-A., Bourban, P.-E., Bonjour, F., 1995. Pat. PCT, WO 96/36477.
Månson, J.-A., Chawla, K.K., 1994. New Demands on Composite Materials. Warrendale, PA: TMS.
Månson, J.-A., Wakeman, M.D., Bernet, N., 2000. Composite processing and manufacturing − An overview. In: Kelly, A., Zweben, C. (Eds.), Comprehensive Composite
Materials, vol. 2. Oxford: Pergamon, pp. 577–608.
Marsh, G., 2002. Reinforced Plast. 46, 44–49.
Pang, K.P., Gillham, J.K., 1990. J. Appl. Polym. Sci. 39, 909–933.
Penn, L., Muzzy, J., 1989. Thermoplastic composites. In: Lee, S.M. (Ed.), Reference Book for Composites Technology, vol. 1. Lancaster: Technomic, pp. 49–60.
van Rijswijk, K., Bersee, H.E.N., 2007. Compos. Part A 38, 666–681.
Schmid, E., Eder, R., 1997. German Pat. 19 602 684 C1.
Summerscales, J., Grove, S., 2014. Manufacturing methods for natural fibre composites. In: Hodzic, A., Shanks, R. (Eds.), Natural Fibre Composites: Materials, Processes and
Properties. Cambridge: Woodhead Publishing, pp. 176–215.
Zini, E., Scandola, M., 2011. Polym. Compos. 32, 1905–1915.