MM - 305 Polymeric Composite Materials

Dr. Kausar Ali Syed

Composite Materials
A composite material is a materials system composed of a mixture or combination of two or more micro or macro constituents that differ in form and chemical composition and which are essentially insoluble in each other. The new material so formed is better suited for a particular application than either of the original material alone. In other words, composites are structures in which two (or more) materials are
combined to produce a new material whose properties would not be attainable by conventional means. Examples include plywood, concrete, and steel belted tires.

The most prevalent applications for fiber reinforced composites are structural materials where rigidity, strength, and low density are important. Many tennis rackets, racing bicycles, and skis are now fabricated from a carbon fiber-epoxy composite that is strong, light, and only moderately expensive. In this composite, carbon fibers are embedded in a matrix of epoxy. The carbon fibers are strong and rigid but have limited ductility. Because of this brittleness, it would not be practical to construct a tennis racket or ski from carbon alone. The epoxy, which in itself is not very strong, plays two important roles. It acts as a medium to transfer load to the fibers, and the fiber-matrix interface deflects and stops small cracks, thus making the composite better able to resist cracks than either of its constituent components. The strength and rigidity of a composite can be controlled by varying the amount of carbon fiber incorporated into the epoxy. This ability to tailor properties, combined with the inherent low density of the composite and its (relative) ease of fabrication, makes this material an extremely attractive alternative for many applications. Many composite materials comprise just two phases the matrix, which is continuous and surrounds the second phase, often called the dispersed phase or the reinforcement. The properties of composites are a function of the properties of the constituent phases, their relative amounts, and the geometry of the dispersed phase. Composites are often classified according to forms of the reinforcement, particulate, fiber, flake, and laminar composites. Fiber composites can be further divided into those containing discontinuous or continuous. One simple scheme for the classification of composite materials is shown in Figure 1, which consists of three main divisions

Fig. 1 particle-reinforced, fiber-reinforced, and structural composites; also, at least two subdivisions exist for each. The dispersed phase for particle-reinforced composites is equiaxed (i.e., particle dimensions are approximately the same in all directions); for fiber-reinforced composites, the dispersed phase has the geometry of a fiber (i.e., a large length-to-diameter ratio). Structural composites are combinations of composites and homogenous materials.

Particle-Reinforced Composites
Large particle and Dispersion strengthened composites are two subclassifications of particle-reinforced composites. The distinction between these is based upon reinforcement or strengthening mechanism.

Dispersion-strengthened Composites
Particles are normally much smaller having diameters between 0.01 and 0.1 m (10 and 100 nm). Particle matrix interactions that lead to strengthening occur on the atomic or molecular level. Metals and metal alloys may be strengthened and hardened by the uniform dispersion of several volume percent of fine particles of a very hard inert material. The dispersed phase may be metallic or nonmetallic; oxide materials are often used. The mechanism of strengthening is like this that the matrix bears the major portion of an applied load, the small dispersed particles hinder or impedes the motion of dislocations. Thus, plastic deformation is restricted such that yield and tensile strengths, as well as hardness, improve but ductility is reduced. The high-temperature strength of nickel alloys may be enhanced significantly by the addition of about 3 vol % of thoria (ThO2 ) as finely dispersed particles; this material is known as thoria-dispersed (or TD) nickel. The same effect is produced in the aluminum -aluminum oxide system. A very thin and adherent alumina coating is caused to form on the surface of extremely small (0.1 to 0.2 m thick) flakes of aluminum, which are dispersed within an aluminum metal matrix; this material is termed sintered aluminum powder (SAP).

LARGE-PARTICLE COMPOSITES
In metallic matrices, these larger sized particles act to restrain movement of the matrix phase in the vicinity of the particle rather than interfere and halt dislocation mobility. The matrix then transmits some of the applied stress to the reinforcing particles and the degree of reinforcement depends on the strength of the bonding at the interface between the particle and the matrix. Particulate composites have been used with all three material types (metals, polymers, and ceramics). Perhaps the most industrially important particulate composites are cemented carbides that are basically ceramic-particle metal-matrix combinations. For example titanium carbide (TiC) and/or tungsten carbide (WC) are embedded in a matrix of nickel cobalt alloy and are extensively used for cutting tools and drill bits. The very hard carbide particles provide the cutting surface but, because they are brittle cannot withstand the impact loads and stresses without early fracture and failure. Resistance to fracture is enhanced by their inclusion in a ductile metalmatrix, which isolates the particles from one another, prevents particle to particle crack propagation, and thereby increases the toughness, or ability to absorb energy without failure. Some polymeric materials to which fillers have been added are really large-particle composites. Again, the fillers modify or improve the properties of the material and/ or replace some of the polymer volume with a less expensive material the filler. An outstanding example of polymeric materials reinforced with particulates to improve overall properties is that of carbon black in rubber. Our use of many of the modern rubbers would be severely restricted without carbon black as reinforcing particulate materials. Carbon black consists of very small and essentially spherical particles of carbon, produced by the combustion of natural gas or oil in an atmosphere that has only a limited air supply. When added to vulcanized rubber, this extremely inexpensive material enhances tensile strength, toughness, and tear and abrasion resistance. Automobile tires contain on the order of 15 to 30 vol% of carbon black. For the carbon black to provide significant reinforcement, the particle size must be extremely small, with diameters between 20 and 50 nm; also, the particles must be evenly distributed throughout the rubber and must form a strong adhesive bond with the rubber matrix. Particle reinforcement using other materials (e.g., silica) is much less effective because this special interaction between the rubber molecules and particle surfaces does not exist. Figure below is an electron micrograph of a carbon black-reinforced rubber.

Another familiar large-particle composite is concrete, being composed of cement (the matrix), and sand and gravel (the particulates). Particles can have quite a variety of geometries, but they should be of approximately the same dimension in all directions (equiaxed). For effective reinforcement the particles should be small and evenly distributed throughout the matrix. Properties are influenced by particle size and distribution as well as by the volume concentration employed, with stiffness or elastic modulus increasing with increased particulate content (provided of course that the particulate has a higher elastic modulus than the matrix phase. Two mathematical expressions have been formulated for the dependence of the elastic modulus on the volume fraction of the constituent phases for a two-phase composite. These rule of mixtures equations predict that the elastic modulus should fall between an upper bound represented by Ec And a lower value given by Ec = Em Ep / (Em Vp + Ep Vm) = Em Vm + Ep Vp

Where E is the elastic modulus, V is the volume fraction, and the subscripts c, m, and p represent composite, matrix, and particulate phases, respectively.

Fiber-Reinforced Composites
Technologically, the most important composites are those in which the dispersed phase is in the form of a fiber. Design goals of fiber-reinforced composites often include high strength and /or stiffness on a weight basis. These characteristics are expressed in terms of specific strength and specific modulus parameters, which correspond, respectively, to the ratios of tensile strength to specific gravity and modulus of elasticity to specific gravity. Fiber-reinforced composites with exceptionally high specific strengths and moduli have been produced that utilize low-density fiber and matrix materials.

Fiber-reinforced composites were developed in response to the needs of aerospace industry. They were first developed in 1940s. Fiberglass reinforced plastics were used successfully in many applications, including filament wound rocket motors. Now inexpensive fiberglass composites are used today in numerous consumer as well as aerospace products. Current availability of new and stronger fibers with properties superior to fiberglass has extended the performance range of fiber-reinforced composites, ensuring many future product developments.

Fiber materials
Almost all high-strength, high elastic modulus materials fail by catastrophic propagation of flaws. A fiber of such a material, however, displays higher strength because of what is called the size effect. Tensile strength is basically dependent on a statistical distribution of flaws, with larger forms of the same material having larger and more frequent flaws. They exhibit lower strength values than smaller cross-sectional forms. In addition, if equal volumes of fibrous and bulk material are compared, we would find that a crack due to a broken fiber would not easily propagate to cause the entire assemblage of fibers to fail, but failure would occur in bulk material from a similar flaw. On the basis of diameter and character, fibers may be grouped into three different classifications: whiskers, fibers, and wires. Whiskers are very small diameter single crystals that have large length-to-diameter ratios, for example, 20-50 m in length and 0.1-1 m in diameter. As a consequence of their small size and synthesis, they have a high degree of perfection for their exceptionally high strengths and, in this form, they are the strongest known materials. In spite of these high strengths, whiskers are not used extensively as a reinforcement medium because they are expensive, difficult to orient and classify into uniform diameters and lengths, difficult to deagglomerate and distribute uniformly in a matrix phase, and difficult to bond to in many instances. Nonetheless, several applications where they are currently employed involve metal and ceramic matrices to be described. Whisker materials include carbon (as the graphite structure), aluminum oxide (also known as sapphire whiskers), silicon carbide, and silicon nitride. Some of their characteristics are given in Table below.

Materials that are classified as fibers are either polycrystalline or amorphous and have diameters larger than those of whiskers, that is, about 10 m, and are produced as continuous filaments and are wound on spools. These include such compositions as glass, carbon, silicon carbide, aluminum oxide, boron, and polymer aramids. Two types of SiC are available, one of which results from the processing and controlled decomposition of polycarbosilane to yield continuous filaments of small diameter, essentially amorphous SiC. The other process involves the chemically vapor deposition of SiC onto a carbon filament substrate to yield a fine-grained polycrystalline product approximately 125 m in diameter and also of continuous length. The same manufacturer of this larger diameter SiC filament produces boron fiber in a similar cross-sectional diameter. The boron is chemically vapor deposited onto a moving tungsten filament to provide a continuous boron fiber. Data for these materials are also provided in Table above. Glass fibers are produced by drawing monofilaments of glass from a furnace containing molten glass, coating the monofilaments with a polymer to "dull" any surface cracking, and gathering a large number of these filaments to form a strand of glass fibers, as depicted in Figure 2. The strands then are used to make glass fiber yams, or rovings, that consist of a collection of bundles of continuous filaments. Considering the data of above Table, it can be seen that glass fibers have the lowest elastic modulus, if not the lowest tensile strength, compared to the other fibers. However, because of their much lower cost and ready availability, glass fibers are by far the most commonly used reinforcing fibers for plastics.

Fig. 2

Fiberglass forming process

Carbon fibers have a combination of very high strength, low density, and high elastic modulus. These properties make the use of carbon fiber-plastic composite materials especially attractive for aerospace applications. Carbon fibers are produced mainly from two sources, polyacrylonitrile (PAN) and pitch, which are called precursors. In general, carbon fibers are produced from PAN precursor fibers by three processing stages: (1) stabilization, (2) carbonization, and (3) graphitization. In the stabilization stage, the PAN fibers are first stretched to align the fibrillar networks within each fiber parallel to the fiber axis. Then they are oxidized in air at about 200-220C while held in tension in order to provide cross-linking between the fibrils to avoid melting at the next stage. In the second stage, carbonization, the cross-linked fibrils are pyrolated (heated) until they become transformed into carbon fibers by the elimination of O, H, and N from the precursor fiber. This carbonization treatment is carried out in an inert atmosphere in the range of 1000 1500C. During the carbonization process, graphite like fibrils, or ribbons, are formed within each fiber and provide the great increase in tensile strength observed. The third stage involves complete conversion of the fiber to oriented graphite crystal form by heating the fiber while under tension to temperatures above 2000C. Nitrogen is removed and chains are joined into graphite planes. This graphitization procedure can raise the elastic modulus to levels exceeding 50 x 106 psi (250 GPa). Carbon fibers produced from PAN precursor material have tensile strengths that range from about 450 to 650 ksi (3.10 to 4.45 GPa) and elastic moduli that range from 28 to 50 x 106 psi (193 to 250 GPa). The density of the carbonized and graphitized PAN fibers is usually about 1.8 g/cm3 with fiber diameters of 7 m to 10 m.

Aramid fibers were introduced commercially by the DuPont Corporation in 1972 under the trade name Kevlar and currently these are available in two commercial types, Kevlar 29 and Kevlar 49. Kevlar 29 is a lowdensity, high-strength fiber designed for such applications as ropes, cables, and even ballistic armor. The properties of Kevlar 49 make its fibers useful as reinforcement for plastics in composites for the aerospace, marine, and automotive industries as well as many other applications.

Fig. 3 Kevlar polymer chain repeating unit, or mer The chemical repeating unit of the Kevlar polymer chain is that of an aromatic polyamide as shown in Figure 3. The aromatic ring structure gives high rigidity to the polymer chains, causing them to have a rodlike structure. Strong covalent bonding in the polymer chains provides the strength and high elastic modulus character. Kevlar aramid is used for high-performance composite applications where light weight, high strength, stiffness, and impact and fatigue resistance are important.

Polymer-Matrix Fiber Reinforced Composites

The purpose of matrix polymers is to bind fibers together by virtue of their adhesive characteristics so that mechanical loads may be transferred from the weak matrices to the higher strength fibers. In these composites,

binding strength between fibers and matrices must be high to minimize fiber pullout, which causes premature failure. The matrices can also serve to protect the fibers from handling damage and environmental degradation, and matrix selection generally determines overall service temperature limitations of a composite as well as processing conditions during fabrication. Polyester resins are the most commonly used matrices for fiberglass-base composites. These thermosetting resins offer a combination of low cost, versatility in many processes, and good property performance. Even though reinforcement efficiency is lower for discontinuous than for continuous fibers, discontinuous (short length) fiber composites are becoming important in the commercial market because of lower processing costs. Chopped glass fibers are used extensively in polyester matrices. These short fiber composites can be produced to have tensile strengths that approach 50% of their continuous fiber counterparts.

Continuous and Aligned Fiber Composites

Tensile StressStrain BehaviorLongitudinal Loading Mechanical responses of this type of composite depend on several factors to include the stressstrain behaviors of fiber and matrix phases, the phase volume fractions, and, in addition, the direction in which the stress or load is applied. Furthermore, the properties of a composite having its fibers aligned are highly anisotropic, that is, dependent on the direction in which they are measured. Let us first consider the stressstrain behavior for the situation wherein the stress is applied along the direction of alignment, the longitudinal direction, which is indicated in Figure 4a.

Figure 4 Schematic representations of fiber reinforced composites.

(a) continuous and aligned, (b) discontinuous and aligned, and (c) discontinuous and randomly oriented

To begin, assume the stress versus strain behaviors for fiber and matrix phases that are represented schematically in Figure 5a; in this treatment we consider the fiber to be totally brittle and the matrix phase to be reasonably ductile. Also indicated in this figure are fracture strengths in tension for fiber and matrix, f* and m* respectively, and their corresponding fracture strains, f* and m*; furthermore, it is assumed that m* > f*, which is normally the case.

Figure 5 (a) Schematic stressstrain curves for brittle fiber and ductile matrix materials. Fracture stresses and strains for both materials are noted. (b) Schematic stressstrain curve for an aligned fiber-reinforced composite that is exposed to a uniaxial stress applied in the direction of alignment; curves for the fiber and matrix materials shown in part (a) are also superimposed.

A fiber-reinforced composite consisting of these fiber and matrix materials will exhibit the uniaxial stress strain response illustrated in Figure 5b; the fiber and matrix behaviors from Figure 5a are included to provide perspective. In the initial Stage I region, both fibers and matrix deform elastically; normally this portion of the curve is linear. Typically, for a composite of this type, the matrix yields and deforms plastically (at Figure 5b) while the fibers continue to stretch elastically, in as much as the tensile strength of the fibers is significantly higher than the yield strength of the matrix. This process constitutes Stage II as noted in the figure; this stage is ordinarily very nearly linear, but of diminished slope relative to Stage I. Furthermore, in passing from Stage I to Stage II, the proportion of the applied load that is borne by the fibers increases. The onset of composite failure begins as the fibers start to fracture, which corresponds to a strain of approximately as noted in Figure 5b. Composite failure is not catastrophic for a couple of reasons. First, not all fibers fracture at the same time, since there will always be considerable variations in the fracture strength of brittle fiber materials . In addition, even after fiber failure, the matrix is still intact inasmuch as (Figure 5a). Thus, these fractured fibers, which are shorter than the original ones, are still embedded within the intact

matrix, and consequently are capable of sustaining a diminished load as the matrix continues to plastically deform.

Elastic BehaviorLongitudinal Loading (isostrain)

Let us now consider the elastic behavior of a continuous and oriented fibrous composite that is loaded in the direction of fiber alignment. First, it is assumed that the fibermatrix interfacial bond is very good, such that deformation of both matrix and fibers is the same (an isostrain situation). Under these conditions, the total load sustained by the composite is equal to the sum of the loads carried by the matrix phase Fm and the fiber phase Ff , or Fc = Fm + Ff

From the definition of stress, = F / A, and thus expressions for Fc , Fm , and Ff in terms of their respective stresses (c, m, and f ) and cross-sectional areas ( Ac, Am, and Af ) are possible. Substitution of these into above Equation yields c Ac = m Am + f Af

and then, dividing through by the total cross-sectional area of the composite, Ac, we have c = m Am / Ac + f Af / Ac

where Am / Ac and Af / Ac are the area fractions of the matrix and fiber phases, respectively. If the composite, matrix, and fiber phase lengths are all equal, Am / Ac is equivalent to the volume fraction of the matrix, Vm and likewise for the fibers, Vf = Af /Ac. The above equation now becomes c = m Vm + f Vf

The previous assumption of an isostrain state means that c = m = f

and when each term in Equation (c = c / c

m Vm + f Vf ) is divided by its respective strain, = m / m x Vm + f / f x Vf m / m = Em,

Furthermore, if composite, matrix, and fiber deformations are all elastic, then c / c = Ec,

and f / f = Ef. and the Es being the moduli of elasticity for the respective phases. Substitution into above Equation yields an expression for the modulus of elasticity of a continuous and aligned fibrous composite in the direction of alignment (or longitudinal direction), as Ecl = Em Vm + EfVf

or

Ecl

Em ( 1 Vf ) + Ef Vf

since the composite consists of only matrix and fiber phases; that is, Vm + Vf = 1. Thus, Ecl is equal to the volume-fraction weighted average of the moduli of elasticity of the fiber and matrix phases. Other properties, including density, also have this dependence on volume fractions. It can also be shown, for longitudinal loading, that the ratio of the load carried by the fibers to that carried by the matrix is Ff / Fm Longitudinal Tensile Strength We now consider the strength characteristics of continuous and aligned fiber-reinforced composites that are loaded in the longitudinal direction. Under these circumstances, strength is normally taken as the maximum stress on the stressstrain curve, Figure 5b; often this point corresponds to fiber fracture, and marks the onset of composite failure.Table below lists typical longitudinal tensile strength values for three common fibrous composites. Failure of this type of composite material is a relatively complex process, and several different failure modes are possible. The mode that operates for a specific composite will depend on fiber and matrix properties, and the nature and strength of the fibermatrix interfacial bond. If we assume that f* < m* (Figure 7a), which is the usual case, then fibers will fail before the matrix. Once the fibers have fractured, the majority of the load that was borne by the fibers is now transferred to the matrix. This being the case, it is possible to adapt the expression for the stress on this type of composite, Equation c = m Vm + f Vf into the following expression for the longitudinal strength of the composite. Here m is the stress in the matrix at fiber failure (as illustrated in Figure 5a) and, as previously, f* is the fiber tensile strength. cl* = m (1 Vf) + f*Vf Case 2: When m* < f*, then matrix will fail before the fibers. cl* = *m (1 Vf) + fVf Here f is the stress in the fiber at matrix failure and m* is the matrix tensile strength. = Ef Vf / Em Vm

Elastic BehaviorTransverse Loading (Isostress)

A continuous and oriented fiber composite may be loaded in the transverse direction; that is, the load is applied at a 90 angle to the direction of fiber alignment. For this situation the stress to which the composite as well as both phases are exposed is the same, or c = m = f = This is termed an isostress state. Also, the strain or deformation of the entire composite c is c but, since = / E, / Ect = ( / Em) (Vm) + ( / Ef ) (Vf) where Ect is the modulus of elasticity in the transverse direction. Now, dividing through by yields 1 / Ect which reduces to Ec = Em Ep / (Em Vp + Ep Vm) = Vm / Em + Vf / Ef = m Vm + f Vf

Above Equation is analogous to the lower-bound expression for particulate composites. Example Problem A continuous unidirectionally aligned glass fiber-reinforced composite consists of 45 volume percent of glass fibers having a modulus of elasticity of 10 x 106 psi (69 x 103 MPa) and 55 volume percent of a polyester resin that has a modulus of 5 x 105 psi (3.5 x 103 MPa). a. Compute the modulus of elasticity of this composite in the fiber aligned direction. b. If the cross-sectional area is 0.5 in.2 (323 mm2) and a stress of 10,000 psi (69 MPa) is applied in this direction, calculate the load carried by each of the fiber and matrix phases. c. Determine the strain that is sustained by each phase when the stress in part b is applied. d. Assuming tensile strengths of 500,000 psi (3.5 x 103 MPa) for glass fibers and 10,000 psi (69 MPa) for polyester resin, determine the tensile strength of this fiber composite in the fiber aligned direction. Solution Basic relations for longitudinally reinforced fiber composites with well-bonded axially aligned fibers are Ec = EfVf + EmVm c = f Vf + m Vm c = f =m

Where E is the elastic modulus, is the tensile strength, and is the strain. The subscripts c, f, and m represent composite, fiber, and matrix (polymer), respectively.
a.

Ec = 69 X 103 MPa (0.45) + 3.5 x 103 MPa (0.55) = 33 x 103 MPa or 33 GPa

b. To solve this section of the problem, we must find the ratio of fiber load to matrix load. Since the strains are equal when stress is applied longitudinally, c = m = f Since stress / strain = modulus ( / = E) and load / area = stress (F / A = ), we have c = Fc / (Ac Ec) = Ff / (Af Ef) = Fm / (Am Em) or Ff / Fm = Ef x Af / Em x Am = Ef x Af / Ac] / Em x Am/ Ac (If the composite, matrix, and fiber phase lengths are all equal, Am / Ac is equivalent to the volume fraction of the matrix, Vm and likewise for the fibers, Vf = Af /Ac). Therefore, Ff / Fm = Ef Vf / Em Vm = 69 GPa x 0.45 / 3.5 GPa x 0.55 = 16.09 Ff = 16.09 Fm Total force sustained by the composite is Or Fc = Ac c Fc = 323 mm2 x 69 MPa = 22,287 N Fc Fc Whereas = = Fm + Ff 16.09 Fm + Fm = 22,287 or Fm = 1304 N

However this total load is just the sum of the loads carried by fiber and matrix phases; that is

17.09 Fm

Ff = 22287 1304 = 20,983 N

Thus the fiber phase supports most of the applied load.

c. The stress for both fiber and matrix phases must first be calculated. Then by using the elastic modulus

for each (from part a), the strain values may be determined. For stress calculations, phase crosssectional areas are necessary and for continuous equal length fibers, the volume fraction is equal to the area, not a real fraction, so that Am = Vm Ac Af = Vf Ac Thus = = 0.55 x 323 = 177.65 mm2 0.45 x 323 = 145.35 mm2

m = Fm / Am = 1304 / 177.65 = 7.34 N/mm2 or 7.34 MPa f = Ff / Af = 20983 / 144.35 = 144.36 MPa

Finally strains are calculated m = m / Em = 7.34 / 3.5 x 103 = 0.0021

f d

= f / Ef

= 144.36 / 69 x 103

= 0.0021

Therefore, strains for both matrix and fiber phases are identical, which they should be. For tensile strength, we have c = = m Vm + f Vf 3.5 x 103 x 0.45 + 69 x 0.55 = 1613 MPa

The importance of adding glass fiber to increase strength is clearly shown. NOTE: In fact for tensile strength of composite, first determine which of the two constituents, fiber or matrix will fail first. At that strain point then determine the corresponding stress on matrix or fiber. That value of stress is then used in the rule of mixtures to calculate the tensile strength of the composite.

Discontinuous parallel fibers

The mechanical characteristics of a fiber-reinforced composite depend not only on the properties of the fiber, but also on the degree to which an applied load is transmitted to the fibers by the matrix phase. Important to the extent of this load transmittance is the magnitude of the interfacial bond between the fiber and matrix phases. Under an applied stress, this fibermatrix bond ceases at the fiber ends, yielding a matrix deformation pattern as shown schematically in Figure 4; in other words, there is no load transmittance from the matrix at each fiber extremity. Some critical fiber length is necessary for effective strengthening and stiffening of the composite material. This critical length lc is dependent on the fiber diameter d and its ultimate (or tensile) strength f* and on the fiber matrix bond strength (or the shear yield strength of the matrix, whichever is smaller) according to lc = f* d / 2c
Fig. 8 The deformation pattern in the matrix surrounding a fiber that is subjected to an applied tensile load.

(1)

For a number of glass and carbon fibermatrix combinations, this critical length is on the order of 1 mm, which ranges between 20 and 150 times the fiber diameter.

When a stress equal to is applied to a fiber having just this critical length, the stressposition profile shown in Figure 5a results; that is, the maximum fiber load is achieved only at the axial center of the fiber. As fiber length l increases, the fiber reinforcement becomes more effective; this is demonstrated in Figure 5b, a stress axial position profile for l > lc when the applied stress is equal to the fiber strength. Figure 5c shows the stressposition profile for l > lc. Fibers for which l >> lc (normally l >15lc ) are termed continuous; discontinuous or short fibers have lengths shorter than this. For discontinuous fibers of lengths significantly less than lc the matrix deforms around the fiber such that there is virtually no stress transference and little reinforcement by the fiber. These are essentially the particulate composites as described above. To affect a significant improvement in strength of the composite, the fibers must be continuous. Figure 5
Stressposition profiles when fiber length l (a) is equal to the critical length (b) is greater than the critical length, and (c) is less than the critical length for a fiber-reinforced composite that is subjected to a tensile stress equal to the fiber tensile strength sf *.

INFLUENCE OF FIBER ORIENTATION AND CONCENTRATION The arrangement or orientation of the fibers relative to one another, the fiber concentration, and the distribution all have a significant influence on the strength and other properties of fiber-reinforced composites. With respect to orientation, two extremes are possible: (1) a parallel alignment of the longitudinal axis of the fibers in a single direction, and (2) a totally random alignment. Continuous fibers are normally aligned (Figure 4a), whereas discontinuous fibers may be aligned (Figure 4b), randomly oriented (Figure 4c), or partially oriented. Better overall composite properties are realized when the fiber distribution is uniform.

THE MATRIX PHASE

The matrix phase of fibrous composites may be a metal, polymer, or ceramic. In general, metals and polymers are used as matrix materials because some ductility is desirable; for ceramic-matrix composites, the reinforcing component is added to improve fracture toughness. The discussion of this section will focus on polymer and metal matrices. For fiber-reinforced composites, the matrix phase serves several functions. First, it binds the fibers together and acts as the medium by which an externally applied stress is transmitted and distributed to the fibers; only a very small proportion of an applied load is sustained by the matrix phase. Furthermore, the matrix material should be ductile. In addition, the elastic modulus of the fiber should be much higher than that of the matrix. The second function of the matrix is to protect the individual fibers from surface damage as a result of mechanical abrasion or chemical reactions with the environment. Such interactions may introduce surface flaws capable of forming cracks, which may lead to failure at low tensile stress levels. Finally, the matrix separates the fibers and, by virtue of its relative softness and plasticity, prevents the propagation of brittle cracks from fiber to fiber, which could result in catastrophic failure; in other words, the matrix phase serves as a barrier to crack propagation. Even though some of the individual fibers fail, total composite fracture will not occur until large numbers of adjacent fibers, once having failed, form a cluster of critical size. It is essential that adhesive bonding forces between fiber and matrix be high to minimize fiber pull-out. In fact, bonding strength is an important consideration in the choice of the matrix-fiber combination. The ultimate strength of the composite depends to a large degree on the magnitude of this bond; adequate bonding is essential to maximize the stress transmittance from the weak matrix to the strong fibers. Problems with adhesion to carbon and aramid fibers have discouraged the development of polyester composites that use these fibers. When property requirements justify the additional costs, epoxies and other resins are used in high-performance commercial applications such as sporting goods (tennis rackets, fishing rods), printed circuit boards, and chemical piping. Epoxy resins are used far more than all other matrices in advanced structural applications. Although epoxies are sensitive to moisture in both their cured and uncured states, they are generally superior to polyesters in resisting moisture and other environmental influences and offer better mechanical properties at acceptable cost. Currently, carbon fiber-reinforced epoxy matrices are the most widely used composites for aerospace structural and other high-performance applications. The main advantage of carbon fibers is that they have very high strength coupled with high elastic moduli and low density. Carbon fiber-epoxy matrix composites have replaced much of the aluminum used in modern aircraft structures where weight reduction is so important. In engineering-designed structures, fiber-polymer matrix material is laminated using precast fabric plies of fibers in desired orientations so that tailor-made strength requirements are met. Figure 9 compares schematics

of unidirectional and multidirectional plies for a composite laminate. Fibers are oriented along 0, 90, and 45 directions. The thin sheets called the laminae, can be stacked together in a regular arrangement to make up the laminate composite. If the fibers are stacked 0 and 90 directions only, the strength of the composite is high along these directions, but such composites have poor shear resistance. To obtain good shear resistance, the fibers must also be stacked along the 45 orientations. Such laminates are strong in all directions within the plane containing the fibers but are weak in the direction perpendicular to the fiber planes.

Figure 9. Unidirectional and multidirectional fabric ply laminate composites. Fibers are oriented along 0, 90, and 45 directions. Although the preceding discussion has centered on single-fiber single-matrix combinations, we should be aware that fiber-reinforced matrix structures can take many hybrid forms. This is particularly true in the bonded abrasives industry. Case Study below bears out the importance of selecting the correct composite structure for the job at hand. Effort is ongoing to extend the service temperature limit of 120C for epoxy resin systems by investigating other formulations. Some high temperature resins possess many of the same desirable features as epoxies, such as fair handle ability, relative ease of processing, and excellent composite bonding behavior. They are also superior in extending the safe in-service temperature to about 220C; however, they have a lower elongation to failure than epoxies and are quite brittle. Work is still continuing to improve this system. Polyimide resins are available with a maximum in-service temperature of about 260C. These thermosetting resins, unlike the others, normally cure by a condensation reaction that releases volatiles. This creates a problem because the released volatiles produce undesired voids in the composite. Recent effort has been directed at reducing this problem by using an addition reaction during curing that does not release volatiles. Although these resins will produce low void content composite parts, they are also brittle, with poor impact

resistance. The attempts to improve thermosetting resins continue with major efforts focused on raising temperature stability and decreasing brittleness, which would improve impact resistance. The dual goal of improving temperature resistance and impact resistance of composite matrices has led to the development and limited use of new high-temperature thermoplastic resin matrices. These materials are very different from the well-known thermoplastics, such as polyethylene, polyvinyl chloride, and polystyrene, that are commonly used as plastic bags, piping, and tableware and have little resistance to elevated temperature. These high-temperature thermoplastics are tougher and offer the potential of improved temperature resistance over epoxies. They also exhibit higher strains to failure that should improve impact resistance. These materials include such resins as polyetherketone (PEEK), polyphenylene sulfide (PPS), and polyetherimide (PEl), all of which maintain thermoplastic character after processing. The fabrication procedures necessary for low-cost manufacture of thermoplastic matrix composites are being thoroughly investigated and major efforts are concerned with determining and understanding the mechanical properties and behavior of fabricated composites.

PROCESSING OF FIBER-REINFORCED COMPOSITES

To fabricate continuous fiber-reinforced plastics that meet design specifications, the fibers should be uniformly distributed within the plastic matrix and, in most instances, all oriented in virtually the same direction. Below several techniques (pultrusion, filament winding, and prepreg production processes) by which useful products of these materials are manufactured will be discussed.

PULTRUSION
Pultrusion is used for the manufacture of components having continuous lengths and a constant cross-sectional shape (i.e., rods, tubes, beams, etc.). With this technique, illustrated schematically in Figure 10, continuous fiber rovings, or tows are first impregnated with a thermosetting resin; these are then pulled through a steel die that preforms to the desired shape and also establishes the resin /fiber ratio. The stock then passes through a curing die that is precision machined so as to impart the final shape; this die is also heated in order to initiate curing of the resin matrix. A pulling device draws the stock through the dies and also determines the production speed. Tubes and hollow sections are made possible by using center mandrels or inserted hollow cores. Principal reinforcements are glass, carbon, and aramid fibers, normally added in concentrations between 40 and 70 vol%. Commonly used matrix materials include polyesters, vinyl esters, and epoxy resins.

Figure 10. Schematic diagram showing the pultrusion process Pultrusion is a continuous process that is easily automated; production rates are relatively high, making it very cost effective. Furthermore, a wide variety of shapes are possible, and there is really no practical limit to the length of stock that may be manufactured

PREPREG PRODUCTION PROCESSES

Prepreg is the composite industrys term for continuous fiber reinforcement preimpregnated with a polymer resin that is only partially cured. This material is delivered in tape form to the manufacturer, who then directly molds and fully cures the product without having to add any resin. It is probably the composite material form most widely used for structural applications. The prepregging process, represented schematically for thermoset polymers in Figure 11, begins by collimating a series of spool-wound continuous fiber tows.

Figure 11. Schematic diagram illustrating the production of prepreg tape using thermoset polymers These tows are then sandwiched and pressed between sheets of release and carrier paper using heated rollers, a process termed calendering. The release paper sheet has been coated with a thin film of heated resin solution of relatively low viscosity so as to provide for its thorough impregnation of the fibers. A doctor blade spreads the resin into a film of uniform thickness and width. The final prepreg product the thin tape

consisting of continuous and aligned fibers embedded in a partially cured resin is prepared for packaging by winding onto a cardboard core. As shown in Figure 7, the release paper sheet is removed as the impregnated tape is spooled. Typical tape thicknesses range between 0.08 and 0.25 mm (3x10 3 and 102 in.), tape widths range between 25 and 1525 mm (1 and 60 in.), whereas resin content usually lies between about 35 and 45 vol %. At room temperature the thermoset matrix undergoes curing reactions; therefore, the prepreg is stored at 0 C (32 F) or lower. Also, the time in use at room temperature (or out-time) must be minimized. If properly handled, thermoset prepregs have a lifetime of at least six months and usually longer. Both thermoplastic and thermosetting resins are utilized; carbon, glass, and aramid fibers are the common reinforcements. Actual fabrication begins with the lay-up laying of the prepreg tape onto a tooled surface. Normally a number of plies are laid up (after removal from the carrier backing paper) to provide the desired thickness. The lay-up arrangement may be unidirectional, but more often the fiber orientation is alternated to produce a crossply or angle-ply laminate. Final curing is accomplished by the simultaneous application of heat and pressure. The lay-up procedure may be carried out entirely by hand (hand lay-up), wherein the operator both cuts the lengths of tape and then positions them in the desired orientation on the tooled surface. Alternately, tape patterns may be machine cut, then hand laid. Fabrication costs can be further reduced by automation of prepreg lay-up and other manufacturing procedures (e.g., filament winding, as discussed below), which virtually eliminates the need for hand labor. These automated methods are essential for many applications of composite materials to be cost effective.

FILAMENT WINDING
Filament winding is a process by which continuous reinforcing fibers are accurately positioned in a predetermined pattern to form a hollow (usually cylindrical) shape. The fibers, either as individual strands or as tows, are first fed through a resin bath and then continuously wound onto a mandrel, usually using automated winding equipment (Figure 12). After the appropriate number of layers have been applied, curing is carried out either in an oven or at room temperature, after which the mandrel is removed. As an alternative, narrow and thin prepregs (i.e., tow pregs) 10 mm or less in width may be filament wound. Various winding patterns are possible (i.e., circumferential, helical, and polar) to give the desired mechanical characteristics. Filament-wound parts have very high strength-to-weight ratios. Also, a high degree of control over winding uniformity and orientation is afforded with this technique. Furthermore, when automated, the process is most economically attractive. Common filament-wound structures include rocket motor casings, storage tanks and pipes, and pressure vessels. Manufacturing techniques are now being used to produce a wide variety of structural shapes that are not

necessarily limited to surfaces of revolution (e.g., beams). This technology is advancing very rapidly because it is very cost effective.

Figure 12. Schematic representations of helical, circumferential, and polar filament winding techniques.