You are on page 1of 21

4.

1 An Introductory Overview of Metal Matrix Composites Systems, Types and


Developments
T William Clyne, University of Cambridge, Cambridge, United Kingdom
r 2018 Elsevier Ltd. All rights reserved.

4.1.1 An Introductory Survey of Metal Matrix Composite Characteristics 1


4.1.1.1 The Historical Context of Metal Matrix Composites 1
4.1.1.2 Property Prediction for MMCs and the Concept of Load Transfer 2
4.1.1.3 Microstructural Features and Property/Ease of Processing Combinations 4
4.1.1.4 Microstructural Quality Control 5
4.1.1.5 Damage, Ductility and Fracture 5
4.1.2 Discontinuously-Reinforced MMCs 6
4.1.2.1 Particulate MMCs 6
4.1.2.2 Short Fiber-Reinforced MMCs 9
4.1.3 Continuously-Reinforced MMCs 9
4.1.3.1 Multifilament MMCs 9
4.1.3.2 Monofilament MMCs 10
4.1.4 Layered MMC Systems 11
4.1.5 Special Category MMCs 12
4.1.5.1 Metallic Foams 12
4.1.5.2 Cermets 12
4.1.5.2.1 Carbide-based cermets 13
4.1.5.2.2 Oxide-based cermets 13
4.1.5.2.3 Boride-based cermets 13
4.1.5.2.4 Carbon-containing cermets 13
4.1.5.3 Reactively Processed MMCs 15
4.1.5.4 IMC 16
4.1.5.5 Graded MMC Coatings and Structures 17
4.1.5.6 MMCs Containing Nanoscale Reinforcement 17
References 18

4.1.1 An Introductory Survey of Metal Matrix Composite Characteristics

4.1.1.1 The Historical Context of Metal Matrix Composites


Reinforced materials based on metals have long been of technological significance. For example, layered structures of metallic and non-
metallic constituents, made by repeated lamination and hammering, were produced in several ancient civilizations.1 Of course, only
recently has the significance of microstructural features in metals been fully appreciated, allowing systematic development of mono-
lithic and composite metal-based materials. Dispersion-hardened metals, and precipitation-hardening (age-hardening) systems, were
developed several decades ago. In both cases, the primary basis of the strengthening mechanism is to impede dislocation motion with
small particles. Dislocations are forced to bow around these obstacles, via a mechanism first identified by Orowan, requiring higher
applied loads. These barriers are either fine oxide particles or precipitates nucleated and grown within a metallic matrix. They must be
closely spaced (oB1 mm apart) if significant strengthening is to be achieved. Since it is generally possible to achieve finer distributions
(by nucleation and growth) in precipitation-hardened systems, these normally exhibit higher strengths at room temperature. However,
dispersion-strengthened systems show advantages at elevated temperature, because of the high thermal stability of the oxide particles.
These materials would not, however, normally be classified as true composites. While there is no universally accepted definition
of a composite, it is commonly assumed that it is only when significant load transfer occurs between matrix and reinforcement
when subjected to an applied stress that the term can properly be applied. When a composite is externally loaded, the matrix is
relieved of a substantial proportion of that load by the presence of the reinforcement. On this basis, conventional dispersion and
precipitation hardened systems are not composites, since they typically contain only around 1% or less of second phase and at
such levels reinforcing constituents cannot significantly reduce the stress borne by the matrix, irrespective of their size and shape.
The development of true metal matrix composite (MMC) materials, such as Al or Cu reinforced with 30–70% of continuous
tungsten or boron fibers, was initiated in the 1960s. As is the case with most polymeric composites (PMCs), an applied load is
largely borne by the fibers in such a material and the matrix microstructure and strength are relatively unimportant. However, it
should be recognized that, in contrast to the corresponding PMCs, commercial usage of long fiber-reinforced MMCs has not really
reached levels which could be considered industrially significant, at least in terms of tonnage consumption, and it is by no means

Comprehensive Composite Materials II, Volume 4 doi:10.1016/B978-0-12-803581-8.09961-6 1


2 An Introductory Overview of Metal Matrix Composites Systems, Types and Developments

certain that such levels will be reached in the future. Certainly, commercial development of such systems has been minimal since
the turn of the millennium.
Discontinuously-reinforced metal composites were developed during the 1980s, with attention focussed on Al-based matrices
reinforced with SiC particles, or Al2O3 particles or short fibers. A combination of good properties, low cost and relatively high
workability has made them attractive for many applications and commercial exploitation has become significant. These materials
fall somewhere between the dispersion-strengthened and fiber-strengthened extremes. They differ from dispersion-hardened
systems in having large (B1–100 mm diameter) reinforcing particles, which contribute negligible Orowan strengthening, and in
containing a relatively high (B5–50%) volume fraction of reinforcement, such that load transfer from the matrix is significant.
However, unlike (axially-loaded) long fiber-reinforced systems, the matrix does bear a substantial load and its strength is relevant.
It is also important in this context to mention cermets, which were developed in the 1950s and have constituted a techno-
logically significant class of materials ever since. Their usage is substantial and steadily increasing, with about 80% of production
going to cutting and machining applications. They certainly should be considered as (particulate) MMCs. While the metal content
is relatively low (below 20%, and commonly no more than about 10%), the matrix is normally contiguous and it does play an
important role. Some information about them is provided in the present chapter (Section 4.1.5.2).
These distinctions concerning the role of the “reinforcement” are highlighted by the schematic plots2 in Fig. 1, illustrating how
strengthening is strongly dependent on reinforcement size for dispersion- and precipitation-hardened metals, but is sensitive to
reinforcement content and aspect ratio for “genuine” MMCs. This is not, however, to say that the presence of the reinforcement has
no influence on fine scale (dislocation) structures within MMCs. Furthermore, the reinforcement may influence characteristics such
as the nucleation and growth of precipitates, and also the creation of voids under an applied load.
It is probably fair to say that the overall rate of commercial development of MMCs, with the notable exception of cermets, has
been relatively slow over the past 15–20 years, although both scientific and technological interest has broadened in terms of the
range of materials and systems being explored, which might be classed as MMCs. For example, various kinds of metallic “foams,”
including fiber network materials, are in use for certain types of application and there are composite materials in commercial use
that contain metal fiber reinforcement. In fact, metal fiber-reinforced ceramic composites do offer a lot of promise for high
temperature applications requiring good toughness and, while they would not always be classed as MMCs.
In general, most aspects of the behavior of MMCs (of various types) are now fairly well understood, with the limitations and
attractions of their processing and performance characteristics being reasonably clear. The present volume aims to provide detailed
and up-to-date information about these characteristics. While processing limitations and economic factors mean that large
tonnage exploitation looks unlikely in the near future, MMCs represent a potentially viable option for a wide range of applications.
The second half of this volume is oriented toward processing and applications of MMCs, constituting a higher proportion of the
volume than for the first edition (published in 2000). This covers graded components, highly porous metallic systems, metal fiber-
reinforced material, intermetallic matrix composites (IMC) and molecular scale metal-organic framework (MOF) materials, as well
as aerospace usage of long fiber MMCs, MMCs for thermal management applications, and design aspects of MMC usage. It is also
worth mentioning that there is coverage of processing aspects of nano-particulate MMCs. There has been a lot of interest in ultra-
fine scale reinforcement of composites (including MMCs) over the past couple of decades, although the theoretical justification for
expecting such refinement to lead to an enhanced set of properties (particularly when toughness is taken into account) is far from
clear (see Section 4.1.5.6) and in general the property combinations obtained have been disappointing. Of course, there are also
challenges in their processing.

4.1.1.2 Property Prediction for MMCs and the Concept of Load Transfer
When designing an MMC, an objective might be to combine the high ductility and formability of the matrix with the stiffness and
load-bearing capacity of the reinforcement, or perhaps to unite the high thermal conductivity of the matrix with the low thermal
expansion of the reinforcement. In attempting to identify attractive matrix/reinforcement combinations, it is often illuminating to
derive a “merit index” for the performance required, in the form of a specified combination of properties. Appropriate models can
then be used to place upper and lower bounds on the composite properties involved in the merit index, for a given volume
fraction of reinforcement. The use of “maps,” with material properties as axes, can then be very useful in highlighting how the
combinations offered by different classes of material compare with each other. The framework for such comparisons and pre-
dictions, covering a range of areas within materials science, has been clearly set out by Ashby in a seminal series of publications. In
order to implement these, a systematic and reliable database of properties is needed. Shortcomings of the currently available MMC
database and predictive capacity, in terms of both scope and reliability, are responsible for some of the caution sometimes
expressed by engineers concerning wider MMC usage.
For many composite properties, upper and lower bounds can be identified, based on corresponding properties of the con-
stituents. In practice, however, there is often interest in establishing composite properties to a greater precision than is possible by
the use of bounds, which in many cases are widely separated. This can be relatively complex, particularly for MMCs – in which
certain matrix properties may be significantly affected by the presence of the reinforcement. Nevertheless, reliable approaches have
been developed for prediction of many of the basic properties of MMCs. These include elastic stiffness, thermal expansion, the
onset of plasticity and work hardening characteristics. These areas are systematically covered in textbooks such as those of Clyne
and Withers2 and Chawla and Chawla.3 The present volume does not include comprehensive treatment of these areas, although
some of them are covered in other volumes of Comprehensive Composite Materials. Chapters in the present volume are largely
An Introductory Overview of Metal Matrix Composites Systems, Types and Developments 3

Fig. 1 Schematic illustration of the nature and magnitude of composite strengthening and stiffening mechanisms, as a function of reinforcement
aspect ratio (s), size (d), and volume fraction (f). The strength (composite yield stress divided by that of the matrix) is plotted in (a), while in (b)
both this and the corresponding stiffness ratios are shown. In (a), matrix strengthening dominates, with the “reinforcement” volume fraction (1%)
being too low for it to carry a significant proportion of the load. In (b), both strengthening and stiffening effects are a consequence of load
transfer to the reinforcement, which is too coarse (dB10 mm) to strengthen the matrix by affecting the motion of individual dislocations.
Reproduced from Clyne, T.W., Withers, P.J., 1993. An introduction to metal matrix composites. In: Davis, E., Ward, I. (Eds.), Cambridge Solid
State Science Series. Cambridge: Cambridge University Press.

directed toward more complex performance characteristics of MMCs, such as their fracture, wear, thermal response, including
creep, residual stresses, and corrosion resistance. These are in practice often pivotal to the question of whether available MMCs are
suitable for particular applications, or could be made so by appropriate microstructural control.
Since the present volume does not include treatment of factors affecting the stiffness, yielding and work hardening of MMCs, it
is appropriate here to give a brief outline of the concept of load sharing between matrix and reinforcement, which is central to an
understanding of these characteristics. Under an applied mechanical load, the stress within an MMC may vary from point to point,
but the proportion of the external load borne by each of the individual constituents (matrix and reinforcement) can be evaluated
by finding the volume-averaged stress within each of them. The external load must equal the sum of the volume-averaged loads
borne by the constituents, so that
ð1  f Þs m þ f s r ¼ sA ð1Þ
which relates the volume-averaged matrix and reinforcement stresses sm and sr to the applied stress s , with a volume fraction f of
A

reinforcement. A certain fraction of the applied load will thus be borne by the reinforcement and the remainder by the matrix.
Provided the response of the composite remains elastic, this proportion will be independent of the magnitude of the applied load
and it represents an important characteristic of the material. It depends on the volume fraction, shape and orientation of the
reinforcement and on the elastic properties of both constituents. The reinforcement may be regarded as acting efficiently if it carries
4 An Introductory Overview of Metal Matrix Composites Systems, Types and Developments

a relatively high proportion of the applied load. For example, stiff fibers aligned in the loading direction carry a relatively high
load, whereas particles and transversely-oriented fibers do not. The concept of elastic load transfer has long been familiar to those
working with (polymer-based) fiber composites. It is readily translated to the elastic behavior of MMCs, although calculation of
the load partitioning is often more complex as a result of the greater interest in discontinuous (short fiber and particulate)
reinforcement, as opposed to continuous fibers. These cases can, however, be treated using numerical techniques or analytical
methods such as the Eshelby approach (which is described in some MMC textbooks2).
Efficient load transfer often results in higher strength, as well as greater stiffness, because the reinforcement is usually stronger,
as well as stiffer, than the matrix. While stiffening effects are straightforward, strengthening is more complex, since there may be
contributions both from load transfer and from in situ matrix strengthening (see Fig. 1). While the latter can often be predicted
using well-established laws and correlations drawn from dislocation theory and metallurgical experience, strengthening by load
transfer is less simple, particularly when the matrix starts to undergo plastic deformation. This is expected to cause rapid transfer of
load to the reinforcement, but in practice this often stimulates other phenomena, such as internal damage development (leading
to fracture) and/or stress relaxation effects such as creep.

4.1.1.3 Microstructural Features and Property/Ease of Processing Combinations


Generic classification as an MMC encompasses a wide range of scales and microstructures. Common to most of them is a metallic
matrix, which is normally contiguous. The reinforcing constituent is in most cases a ceramic, although there are exceptions to this
and MMCs can be taken to encompass materials “reinforced” with relatively soft and/or compliant phases, such as graphitic flakes,
lead particles or even gases. It is also possible to use refractory metals, intermetallics or semiconductors, rather than true ceramics.
This flags up a number of uncertainties about whether various materials should be considered as MMCs. Examples include Al–Si
eutectic, cast irons and the so-called “dual phase” steels (which contain about 20% of large, hard martensitic particles in a ferritic
matrix). There is no “correct” answer to this question, but it can certainly be argued that such materials are closer to conventional
MMCs in terms of behavior and performance characteristics than are the precipitation-hardened and dispersion-strengthened
systems referred to above in Section 4.1.1.1.
It is also worth noting that there are some classes of materials, such as the MOF materials, that are composites on the molecular
scale. This introduces major conceptual changes, since the idea of constituent phases, which is rather central to most analyses of
composites, is lost, or at least substantially modified. Nevertheless, considering their properties, particularly their thermo-
mechanical characteristics, by using ideas drawn from the approaches employed for composites is helpful in several respects.
The main MMC types are commonly subdivided, as depicted in Fig. 2, according to whether the reinforcement is in the form of
(a) long aligned fibers (allowing high reinforcement contents), (b) short fibers (with or without a degree of alignment), or (c)
particles, which are at least approximately equiaxed. A similar classification is commonly applied to PMCs. A major difference
between the two matrix types is that, while the progression (a)–(c) is in general one of decreasing industrial significance for PMCs,
the emerging picture for MMCs is that levels of commercial interest rise for the same sequence. It might be considered that, for
both PMCs and MMCs, the broad trends on moving from (a) to (c) would be for reductions in performance, but compensatory
improvements in formability and ease of processing. However, while this is probably a fair generalization for PMCs, certain points
can be identified for MMCs that alter the picture somewhat. Firstly, metals are highly formable, making secondary processing very
attractive, provided the reinforcement content is such as to allow this, whereas for PMCs it is only with thermoplastic matrices that
post-consolidation forming is possible. Secondly, long fibers, particularly when aligned and present at high volume fractions,
severely constrain plastic deformation of the matrix in MMCs, inhibiting the highly efficient energy absorption mechanism of
dislocation motion and thus substantially reducing the toughness of the metal. In PMCs, the same inhibition of matrix plasticity

Fig. 2 A schematic depiction of the main metal matrix composite (MMC) systems, classified according to the type of reinforcement. Reproduced
from Clyne, T.W., Withers, P.J., 1993. An introduction to metal matrix composites. In: Davis, E., Ward, I. (Eds.), Cambridge Solid State Science
Series. Cambridge: Cambridge University Press.
An Introductory Overview of Metal Matrix Composites Systems, Types and Developments 5

occurs, but the matrix toughness is in any event very low and energy absorption via interfacial debonding and fiber pullout
becomes highly significant. Such fiber pullout is often rather more difficult to promote in MMCs, for complex reasons concerned
with interfacial properties and fracture behavior, but in any event the inescapable fact is that, while polymeric matrices are
substantially toughened by the introduction of long fibers, most metallic matrices are substantially embrittled. This effect is largely
responsible for the trends in popularity between different MMC types noted above. It does not, of course, eliminate the possibility
of applications being identified in which this shortcoming of long fiber MMCs is outweighed by attractive aspects of their
performance.
Of course, in examining the possibility of using novel materials such as MMCs for particular applications, the economic factors
involved are often of paramount importance. These may be difficult to quantify in some cases, since there are often indications
that production and processing costs would fall significantly as usage became well established, but it may be difficult to obtain
guarantees or accurate projections. Nevertheless, cost-benefit analyses for a switch to a different material can be attempted. It is
emphasized that the property profiles obtainable with MMCs often fall outside the envelopes available from conventional
(monolithic) materials and that this might have particular relevance and attractions for certain applications. Costs, however, are
sometimes inherently high and this must be balanced against performance benefits considered in the context of the application.
Realistic quantitative assessments of this type will be essential if MMC usage is to become fully established in the applications for
which they would be genuinely beneficial.

4.1.1.4 Microstructural Quality Control


A recurrent issue for manufacturers of MMC material and products concerns identification of appropriate quality control para-
meters, particularly relating to microstructural features. Some of these can be drawn from standard practice for unreinforced
metals. For example, levels of impurity or porosity in the matrix can be fairly easily measured and specification maxima for these
can be defined. While there may be problems in establishing the precise effects of these parameters on the performance of
particular types of MMC, there are no conceptual difficulties in setting specifications of this type.
However, it is becoming clear that there are certain specific microstructural characteristics peculiar to the presence of the
reinforcement that can have a strong influence on the behavior of MMCs. For example, not only is the size, shape and volume
fraction of the reinforcing constituent often significant, but also its spatial distribution and the nature of the interfacial bonding.
Moreover, there are two separate types of problem associated with such attributes: firstly their effect on the performance of the
material may be incompletely understood and secondly the way in which they are best characterized may be far from obvious. It is
now fairly well established that inhomogeneous distributions of reinforcement, particularly the presence of clustering (giving rise
to ceramic-depleted and ceramic-rich areas), can have a deleterious effect on the mechanical properties. This is particularly true for
the ductility and fracture toughness. Systematic analysis requires the application of tesselation procedures, in which a metallo-
graphic section is divided up into a series of polygons, each containing a single reinforcing particle or fiber. In the present state of
MMC development, it would be unusual for such procedures to be systematically used as a quality control tool, although with the
advance of automatic image analysis facilities they are becoming easier to implement and they may be used in the future.
A similar problem relates to the interfacial bond strength, except that it is quite difficult to even approximately estimate this via
a study of the microstructure. In general, it is now clear that the properties of most MMCs are enhanced by the promotion of a high
interfacial bond strength and toughness.4 Such good bonding is certainly present in successful high ceramic volume fraction
MMCs, such as cermets (see Section 4.1.5.2), and there are many indications that good bonding is beneficial for lower ceramic
content MMCs. It is also fairly clear that the presence of brittle interfacial reaction products, which tend to lower the toughness of
the interfacial region, is normally undesirable. While this can be monitored by microstructural examination, quantitative eva-
luation of bond strength requires some sort of mechanical interrogation of the material. A variety of techniques have been
developed5–9 for doing this, most of which are much more easily applied to long fiber systems (particularly monofilament-
reinforced) than to particulate or short fiber MMCs.

4.1.1.5 Damage, Ductility and Fracture


Perhaps the area of greatest concern in the performance of MMCs is that there is a tendency for the ductility and toughness to be
relatively low. In fact, MMCs can exhibit fracture toughness values of well above 20 MPa √m, which is certainly adequate for many
critical load-bearing applications, and ductilities of at least 5% or so are routinely obtainable. However, many MMCs exhibit lower
values than these. Furthermore, the toughness of nominally similar types of MMC, perhaps made by different routes or with
different processing conditions, can vary significantly. Of course, materials engineers have long been familiar with such variations
in the context of metallic alloys, in which microstructural changes during production or as a consequence of thermo-mechanical
heat treatments can lead to substantial changes in mechanical properties. However, while the mechanisms responsible for these
changes are well established and understood for unreinforced metals, cause and effect are in some cases rather unclear for MMCs.
In particular, while there has certainly been extensive work in these areas,10–16 there is still a degree of uncertainty surrounding
the effects of reinforcement size, degree of clustering, interfacial bond strength, residual stress distribution and matrix toughness
on the ductility and fracture toughness of MMC systems. Broadly speaking, it is clear that uniform spatial distribution of
reinforcement and strong interfacial bonding both lead to enhanced fracture toughness. Finer particles also tend to promote a
higher toughness in MMCs, although the reverse appears to be true for cermets (see Section 4.1.5.2) and in fact it is commonly
6 An Introductory Overview of Metal Matrix Composites Systems, Types and Developments

observed in a range of composite materials that, for given fracture geometry, a coarser scale of structure (e.g., fiber diameter) leads
to tougher material. Clearly the behavior depends on the predominant mechanisms of energy absorption during crack propagation
in the composite concerned. For fibrous MMCs, while situations can arise with relatively brittle metallic matrices when fiber
pullout contributes significantly to the net energy absorption during fracture,17 this is in general rather unusual and factors
controlling the matrix plasticity constitute the most important consideration for the toughness of all types of MMCs. Work remains
to be done on various aspects of this, including the precise role of residual stresses within the matrix, reinforcement volume
fraction and the effect of having bimodal distributions of particle size. On the other hand, in ceramic-based composites reinforced
with metallic fibers, contributions from pullout and plastic rupture of fibers dominate the work of fracture, in such a way that
coarser fibers toughen more effectively.

4.1.2 Discontinuously-Reinforced MMCs

4.1.2.1 Particulate MMCs


This type of MMC has been the subject of extensive study for a range of industrial applications. While much of this interest centers
around Al alloy matrices, there has also been work on Ti-, Fe- and Mg-based systems. The particulate material employed is most
commonly SiC or Al2O3, but many others have been investigated. These include TiB2, B4C, SiO2, TiC, WC, BN, ZrO2 etc. In general,
the main attributes ideally exhibited by particulates for MMC reinforcement are:

• high stiffness,
• low density,
• high hardness, and
• good availability/low cost for suitably-sized powder (B1–20 mm diameter).

An overview of how some candidate materials compare with metals, and with each other, in terms of some simple properties is
provided by the maps in Fig. 3. Fig. 3(a) is a familiar plot of stiffness against density. The attractions of stiff, light reinforcements,
such as boron and diamond, are clear. Of course, such materials tend to be expensive, with diamond (and other forms of carbon,
such as carbon fibers and graphene) presenting the additional problem of a strong likelihood of (excessive) reaction with the
matrix during processing and/or in service – many metallic matrices are strong carbide-formers (although Mg and Cu are not). In
practice, carbonaceous materials are not widely used in MMCs and boron is only used (in the form of monofilaments) in some
rather exotic variants for very high performance applications. However, it can be seen in Fig. 3(a) that other (more readily
available) materials, such as SiC, TiB2 and Al2O3, do offer potential for raising the specific stiffness of metals considerably. On the
other hand, it is also clear that there is little point in adding glass (fibers) to a metal if the main objective is to raise the (specific)
stiffness.
Of course, most metals already have a fairly high stiffness and in any event there are many applications in which stiffness is not
the main concern. There is scope for designing MMCs so as to offer attractive combinations of (elastic) thermo-mechanical
properties, such as a high conductivity and a low thermal expansivity, and some such issues are addressed in other chapters, and
are also outlined below. However, plastic deformation, and associated issues of ductility and toughness, are often of central
importance in metals, and hence in MMCs. Of course, there is interest in making metals more resistant to plastic deformation, for
example, so as to improve the wear resistance. In any event, the relative hardness of metals and potential reinforcements (Fig. 3(b))
is clearly of interest.
It should, of course, be emphasized that hardness cannot be considered to be a “primary” property – it depends on both yield
stress and work-hardening characteristics, in a way that is not well-defined. Measured values thus vary between types of hardness
test (with different indenter shapes), and also with applied load (and hence levels of plastic strain) for a given type of test.
Furthermore, the hardness of a given type of material can vary substantially as its microstructure is changed, particularly for metals –
in which the mobility of dislocations is a key factor. The plot in Fig. 3(b) reflects this, with a wide range of HV values being shown in
many cases, particularly for the metals. Nevertheless, it is quite clear that most candidate reinforcements for MMCs are much harder
than the matrix. Under most circumstances, they will undergo little or no plastic deformation within an MMC, while constraining the
plasticity of surrounding matrix. This may, of course, be beneficial, although a major concern is with the danger that this constraint
will substantially impair the toughness of the material. Still, it is important to note that this kind of “hardening” of the metal is
different from more conventional mechanisms, such as precipitation hardening, and is likely to exhibit different characteristics – one
example being freedom from the danger of precipitate-free zones forming near grain boundaries. The plasticity that occurs in an
MMC is likely to be more inhomogeneous than in a conventionally hardened metal, which may be advantageous or problematic,
depending on a number of factors.
Further characteristics of potential interest include thermal/electrical conductivity, thermal expansion, chemical compatibility
with the matrix during processing and the ease with which a strong interfacial bond can be formed. While all ceramics can in
general be taken as electrical insulators in comparison with metals, their thermal conductivity can be comparable to, or even
greater than, those typical of metallic systems. This arises because heat can be conducted by phonons, as well as electrons, and
phonon transport is favoured by a light, stiff crystal lattice. The thermal conductivities of SiC and diamond are particularly high.
However, it should be noted that phonons are readily scattered by defects and experimental values are often relatively low as a
consequence of the presence of grain boundaries and other microstructural features.
An Introductory Overview of Metal Matrix Composites Systems, Types and Developments 7

Fig. 3 Maps of (a) stiffness and (b) hardness against density, for a few common metals and for some (ceramic) materials that might be
considered as candidate reinforcements for them.

Excessive chemical reaction during processing can occur with some particulate reinforcements. For example, SiC
can be particularly problematic when incorporated into Al- and Ti-based MMCs. There is extensive documentation18,19
concerning the reaction between SiC and Al melts, including the possibility of the particles becoming pitted.20 Titanium is
more likely to be processed into MMCs in the solid state (via powder routes), but SiC commonly reacts with Ti during such
processing,21–23 to a greater extent than occurs with C-free alternatives such as TiB2.24 (Of course, Ti is a strong carbide-
former.) Coatings (such as B4C25 or TiB226) are sometimes employed on SiC for use in Ti, in order to control interfacial
reactions.
Reactions also occur between SiC reinforcement and ferrous matrices,27 sometimes leading to complete dissolution,28 although
again TiB2 has proved effective as a stable particulate reinforcement in ferrous MMCs. In fact, there is considerable current interest
in “high modulus steels” containing about 10–20 vol% of TiB2 particulate.29–33 The Young’s modulus can thus be raised from 210
GPa to around 250–300 GPa. This system might turn out to be one of the most successful types of particulate MMC and industrial
applications are now starting to become significant.
Alumina is usually less reactive than SiC in Al, but it does react quite strongly with Ti at elevated temperature. Magnesium is
rather different from Al and Ti, in that it does not form a stable carbide, but it does have a high affinity for oxygen. The greater
stability of Al2O3, compared with SiC, in Al is therefore reversed for Mg matrices.34 In general, while coatings or other surface
treatments may be worth considering for fibers (particularly monofilaments), economic and practical considerations mean that
particulate reinforcement is normally introduced into MMCs in the virgin state. This may, however, be such that a surface oxide
layer is present and deliberate thickening of this layer, for example, by heat treatment in air, has in some instances been found to
have a beneficial effect on interfacial bonding or other characteristics.
Particulate MMCs are usually manufactured on a commercial basis either by melt incorporation and casting or by powder
blending and consolidation. More specialist production routes, which are less widely used at industrial production levels, involve
reactive processing (see Secion 4.3) or spray co-deposition. Quality control objectives include the elimination of excessive
interfacial reaction during processing, particularly for melt routes, and also the avoidance of microstructural defects such as poor
interfacial bonding, internal voids and clustering16 of the reinforcement. Typically, reinforcement particles are about 10–20 mm in
diameter and constitute about 10–30% by volume of the material, although MMCs in which the values concerned lie outside of
8 An Introductory Overview of Metal Matrix Composites Systems, Types and Developments

Fig. 4 Optical micrographs of (a) a particle-reinforced metal matrix composite (MMC) (Al-10% Al2O3) and (b) a cermet (Co-90%WC).

these ranges have been studied and are available commercially (particularly finer particles and higher particle contents). Of course,
cermets constitute particulate MMCs with a very high loading (B90%) of reinforcement: some information about them is
presented in Section 4.1.5.2.
The two micrographs in Fig. 4 allow comparison between the structures typical of most particulate MMCs and cermets. Fig. 4(a)
shows an MMC produced by blending of (spheroidized) Al and Al2O3 particles, followed by extrusion and heat treatment, as part
of a study on the effect of reinforcement in MMCs on recrystallization.35 The degree of inhomogeneity (clustering) apparent here is
typical of both powder and melt route particulate MMCs. As is the case with the material shown, it’s quite common for individual
grains to contain a (large) number of particles. Also, such material is usually more or less free of porosity. The cermet shown in
Fig. 4(b), on the other hand, was produced by blending of WC and Co powders, followed by molding to a green compact and
heating so as to melt the Co matrix. The volume fraction of ceramic particles is so high that they are mostly in contact with each
other and there is little inhomogeneity. On the other hand, the solidification shrinkage of the Co almost inevitably leaves some
residual porosity, which can be seen in this micrograph. Also, while it’s not entirely clear in this micrograph, individual matrix
grains rarely envelope even a single ceramic particle. Despite these differences, however, it’s clear that cermets are essentially a type
of particulate MMC.
The range of applications for which particulate MMC components are being developed is quite wide. As an example of this,
there have been extensive trials involving various automotive components, such as brake calipers, disks and pads.36–38 The
superior wear performance of MMCs is of prime interest, although stiffness enhancement is also beneficial. Replacement of
conventional cast iron disks, giving a weight saving of several kg, is very attractive, particularly for high performance cars. However,
the temperature of the Al matrix must be kept fairly low, since it cannot be permitted to soften extensively (and precipitation
hardening is ineffective above B2001C). The high thermal conductivity of Al (which is retained when SiC reinforcement is
introduced) favours the avoidance of “hot spots” on the disk and cooling fins can help to dissipate the heat. Nevertheless, this
issue has caused problems (under “alpine descent” conditions). Steel-based MMCs are potentially attractive in terms of durability
and stiffness, but the weight-saving is then much less and of course the cost is always an issue for automotive components. There
has been some work39,40 on MMC brakes for railway applications.
An Introductory Overview of Metal Matrix Composites Systems, Types and Developments 9

Fig. 5 SEM micrograph of “Saffil” short fibers of Al2O3.

4.1.2.2 Short Fiber-Reinforced MMCs


Short fiber-reinforced MMCs first attracted widespread attention in the mid-1980s with development of Al-based diesel engine
pistons selectively reinforced with short alumina (“Saffil”) fibers having diameters in the approximate range of 1–5 mm.41–43 These
and similar (alumino-silicate) fibers are manufactured mainly for insulation purposes, but they have been widely explored as
MMC reinforcement and indeed pistons containing them are commercially available. Components are commonly produced by
melt infiltration. These fibers have a fine-grained polycrystalline microstructure44 and they are manufactured via a melt spinning
route. They have been employed extensively in Mg-based MMCs,45–47 as well as in Al alloys. A micrograph of some (partially
milled) Saffil fibers is shown in Fig. 5.
Short fiber MMCs can offer attractive combinations of properties and ease of processing, either via a melt route, such as squeeze
infiltration, or by powder blending and consolidation. Interfacial characteristics are partly dependent on the degree of reaction
during processing, which is in turn affected by the surface chemistry of the fibers: for example, Saffil fibers have a thin silica-rich
surface layer which tends48,49 to react with an Al melt during processing, particularly if Mg is present. Some secondary processing,
such as forging and extrusion, can be carried out under appropriate conditions, although their formability is in general inferior to
that of particulate MMCs. On the other hand, certain property advantages over particulate MMCs are common, particularly in
terms of resistance to creep and, to a lesser extent, wear (Fig. 6).
The mechanical properties of fibers such as Saffil are relatively good, with failure stress values of the order of 1 GPa, but even
higher strengths can often be obtained if the fine-grained structure can be replaced by that of a single crystal, while also reducing
the fiber diameter (and hence the size of flaws likely to be present). There has, of course, been an explosion of interest over the past
couple of decades in the general concept of nanotechnology and in the particular idea that exceptional properties can be obtained
from ultra-fine structures such as carbon nanotubes. In fact, carbon is not an attractive reinforcement for most metals, in view of
the likelihood of reaction (carbide formation) during production, which obviously becomes increasingly problematic as the
diameter of the fiber is reduced. This particular problem does not apply to all reinforcement/metal combinations and there has
been considerable interest in use of fine, single crystal fibers (often called “whiskers”) in MMCs. However, their usage has not in
fact become widespread, for reasons (common to all ultra-fine reinforcement) that are outlined in Section 4.1.5.6.

4.1.3 Continuously-Reinforced MMCs

4.1.3.1 Multifilament MMCs


A number of long fiber MMCs systems have been investigated and some of these have been used in certain industrial and military
applications. However, as a consequence of processing difficulties and of the constraints on ductility and toughness outlined above in
Section 4.1.2.1, this usage has in general remained rather specialised and limited. The term “multifilament” refers to relatively small
diameter (B5–30 mm diameter) fibers that, as a consequence of their low beam stiffness (which scales as the fourth power of the
10 An Introductory Overview of Metal Matrix Composites Systems, Types and Developments

Fig. 6 Measured and predicted fracture energy values for failure of metal-ceramic laminates, plotted against the thickness of the metal layers. The
experimental data refer to specimens produced with a fixed ceramic layer thickness, so that the metal volume fraction, fm, is different in each case.
Reproduced from Pateras, S.K., Howard, S.J., Clyne, T.W., 1997. The contribution of bridging ligament rupture to energy absorption during
fracture of metal-ceramic laminates. Key Engineering Materials 127–131, 1127–1136.

diameter), are flexible enough to be handled as tows or bundles that can be woven, braided, filament wound, etc. The materials
concerned include SiC fibers and various oxide fibers. There are certain other types of multifilament in common use, but these are
either unable to survive the elevated temperatures involved in MMC production (e.g., polymeric and organic fibers) or are of limited
interest for MMCs because of relatively poor mechanical characteristics such as stiffness or creep resistance (e.g., most glass fibers).
Multifilament MMCs can be produced by melt infiltration, although problems arise with unidirectionally aligned fibers, in that
the applied melt pressure transverse to the fiber axis tends to bring them into close contact and reduce the channels between them
to such small dimensions that the melt is unable to penetrate. This problem can be reduced by introducing particulate or
transverse fibers, although it is also possible in some cases to arrange for the infiltration to take place in the axial direction.50 There
has also been work51 on MMCs produced by infiltrating Al into woven alumina textiles. Other processing techniques, including
powder metallurgy approaches, have been employed, but are often rather unsatisfactory.
Carbon fibers are not popular as MMC reinforcement, primarily as a consequence of excessive interfacial reaction during
processing and the galvanic corrosion effects that can take place in service. Chemical reaction problems are severe for Al, Ti and Fe
alloys, but much less pronounced with Mg, which does not form a stable carbide. Mg-based carbon multifilament MMCs have
therefore received some attention.52–55 Attempts have been made to protect carbon fibers with a surface coating, such as titanium
nitride,56 but in general this is difficult and expensive for multifilaments. There is a particular problem with Al alloys, in that the
interfacial reaction product, Al3C4, is hygroscopic, so that carbon fiber-reinforced Al tends to undergo rapid corrosion in aqueous
environments. Although there have been claims57 that alloys and processing conditions can be identified which lead to carbon
fiber reinforced Al which is quite corrosion-resistant, the problem has severely limited the development of this type of MMC.
While SiC has been a successful reinforcement in particulate-reinforced MMCs, there is a shortage of multifilament SiC fibers
that are suitable for incorporation into metallic matrices. Multifilaments are available commercially, under tradenames such as
Nicalon, which are derived from polycarbosilane (PCS) precursors in a similar way to the production of carbon fibers from
polyacrylonitrile (PAN). However, although these are nominally SiC, they actually contain considerable amounts of free carbon
and silica, leading to excessive reaction with most metallic matrices during processing. There are, however, a number of oxide
multifilaments which are fairly resistant to attack by molten metals. Notable among these are polycrystalline alumina fibers,
usually composed primarily of the stable a phase,58 which have been studied for use in Al59,60 and, to a lesser extent, Ti61 matrices.
Commercial interest in all such systems has remained low, although there has been some activity concerning reinforcement of
intermetallic matrices with oxide multifilaments.

4.1.3.2 Monofilament MMCs


Monofilaments are large diameter (B100–150 mm diameter) fibers, most of which are produced by chemical vapor deposition
(CVD) of either SiC or boron onto a core of carbon fiber or tungsten wire. A consequence of the large diameter is that
An Introductory Overview of Metal Matrix Composites Systems, Types and Developments 11

monofilaments are much less flexible than multifilaments, so that they normally need to be handled as single fibers rather than
bundles and precautions are necessary to avoid causing damage by the imposition of sharp curvature during processing operations.
There are, however, advantages in having such a large diameter. One of these is that interfacial reaction consumes a much smaller
proportion of the fiber than would be the case for multifilaments, simply because the interfacial area is much smaller. There is also
much greater scope for tailoring the surface chemistry and introducing surface coatings. This can often be done as part of the fiber
production process. A lot of effort went into development of coatings on SiC monofilaments for incorporation into titanium and
titanium aluminide matrices. Virtually all fiber materials react with Ti at elevated temperature, so that use of coated monofilaments is
one of the few approaches offering scope for control of this problem. Thick graphitic coatings (which are progressively consumed, but
prevent defects from forming on the fiber itself) have been popular and there has also been work62–64 on various duplex layers, such
as TiB2/C and Y/Y2O3, which are designed to slow interfacial reaction rates down to very low levels.
Monofilament-reinforced MMCs have mainly been produced by the foil-fiber-foil (diffusion bonding) route or by the eva-
poration of relatively thick layers of matrix material onto the surface of the fiber, followed by hot pressing. Work in this area is very
much oriented toward Ti-based matrices. Fortunately, Ti diffusion bonds to itself very readily, mainly because it dissolves its own
surface oxide layer at elevated temperature in controlled atmosphere. The evaporation method is slower and more expensive than
the foil-fiber-foil route, but it does produce material in which the fiber distribution is more uniform. This is an important
advantage in view of extensive evidence for various types of MMC that the toughness and ductility are impaired by clustering of the
reinforcement, although it should be recognized that matrix plasticity is heavily constrained by the presence of the fibers in this
type of MMCs, so that toughness values tend to be relatively low even if there is no clustering.
There has been a lot of interest in the selective use of monofilament-reinforced titanium for critical components in aeroengines. The
presence of the SiC monofilaments confers a dramatic improvement in the creep resistance of titanium and there is also a substantial
enhancement to the stiffness. It may also be noted that the resistance to compressive failure and buckling collapse is considerably
improved.65 These are all valuable property improvements for titanium in demanding applications within a gas turbine engine.
However, such enhancement occurs exclusively or predominantly in the direction of fiber alignment and careful account must be taken
of the nature of the imposed stress field, and the effect of internal stresses from differential thermal contraction, when designing the
fiber orientation within the component. Nevertheless, it has been concluded that the enhancements in critical properties are such that
there could be complete redesign of certain gas turbine components, with dramatic weight savings and consequent benefits.

4.1.4 Layered MMC Systems

A class of MMCs which has attracted attention is that in which the two constituents are in the form of alternate layers of some sort.
Such arrangements might range in scale from layer thicknesses of a few nanometer up to macroscopic laminates made by bonding
together plates that are several centimeter thick. Many of the mechanical properties of such systems, particularly those relating to elastic
behavior, can readily be predicted using the simple concept of either an equal strain (loading parallel to the plane of the layer) or an
equal stress (transverse loading) being imposed on both constituents. There is also likely to be relatively high constraint on the plastic
deformation of the metallic layers, leading in many cases to high work-hardening rates. In addition to simple bonded layers of
monolithic ceramic and metal layers, there has also been interest in other arrangements, such as alternate layers of metal and polymer-
based long fiber composite – which is the basis of the so-called “ARALL” material,66,67 and also the related “GLARE” material.68,69
One of the attractions of layered systems is that, in comparison with the corresponding fibrous or particulate composites, it is
often relatively easy to manufacture components with this geometry. A range of fabrication methods can be used to produce metal-
ceramic laminates. These include vapor deposition or sputtering to produce very thin layered structures, such as are used in a
variety of device and electronic applications. Particular attention must be paid to the danger of excessive interdiffusion and
chemical reaction during deposition for such structures. Furthermore, these techniques are rather slow if three-dimensional
components rather than thin films are required. Faster production is possible using various diffusion bonding methods and, for
cases in which it is acceptable for the ceramic layers to become discontinuous, it may be possible to use roll bonding of thin metal
strips with surface oxide layers or to roll layered material down to reduce the layer thickness.
Fine layered metal/ceramic structures often have interesting physical and functional properties. They may also exhibit certain
attractive mechanical properties, such as a high yield stress. However, coarser scale materials are quite likely to be of more interest
for mechanical and structural purposes. This is partly a consequence of factors related to the processing militating against the
economic production of very fine scale structures, but in fact certain key mechanical characteristics are often superior with coarser
structures. For example, consider the issue of energy absorption during fracture of laminated ceramic/metal materials. For these
materials, the toughness is dominated70–73 by the energy absorbed as the metallic layers undergo plastic deformation during
necking and rupture. Since the volume of material in which plastic deformation occurs, per unit area of fracture surface, is larger
when the layer thickness (hm) is greater, coarser structure have higher fracture energies. Modeling of the energy absorption process
leads to an equation of the form
Z χmax
hm sN
Gc ¼ ð1  fm ÞGcer þ fm sY w; where w ¼ dχ ð2Þ
2 0 sY

in which Gcer is the fracture energy of the ceramic, fm is the metal volume fraction, sY is the effective yield stress of the metal, sn is
the nominal stress on the ligament during traction, χmax is the normalized crack opening displacement at which the metal ligament
12 An Introductory Overview of Metal Matrix Composites Systems, Types and Developments

fractures and the integral, w, is the area under a plot of sN/sY against χ. The broad validity of this treatment is confirmed by the
predictions and experimental data shown in Figs. 1–6, which refer to Al–Al2O3 laminates tested at room temperature. In fact, this
tendency for the toughness to rise as the scale of the structure becomes coarser is also observed for all contributions to the fracture
energy associated with fiber pullout74 and with any plasticity of fiber rupture. Both types of contribution are significant in metal
fiber-reinforced ceramics and the dependence of their toughness on fiber diameter is described in other chapter. It’s also worth
noting that this effect (i.e., a tendency for the toughness of most fiber composites to decrease as the fiber diameter is reduced) has
created problems during attempts to make new types of composite reinforced with fibers having diameters in the nanometre range
(see Section 4.1.5.6).

4.1.5 Special Category MMCs

4.1.5.1 Metallic Foams


It is possible to consider a metallic foam as a composite with metallic and gaseous constituents, i.e., an MMC with a gas as the
“reinforcing” constituent. Indeed, this is a sensible approach to the prediction of many of the properties of metallic foams. If the
gas is taken to have a Young’s modulus of zero and a very low bulk modulus, then predictions of the elastic properties of the foam,
using well-established methods, should be quite reliable. For other properties, such as the thermal conductivity, the fact that the
reinforcement is a gas becomes relevant, since convection may have a significant effect – particularly in open cell foams.75
However, it should be noted that treating the metallic framework as a continuum, within which gaseous inclusions are
dispersed, may be rather inappropriate for many purposes, particularly if the gas volume fraction is high (4B80%). It has thus
been common to employ a framework for mechanical property prediction which is based on an assembly of structural elements –
usually the edges and faces of the cells, taken to have a certain geometrical shape. This may range from a simple cube to more
complex polyhedra such as tetrakaidecahedra or pentagonal dodecahedra. The resulting relationships between porosity content
and mechanical properties have been studied in considerable detail. These are covered in the text of Gibson and Ashby.76 Simple
analytical expressions are available76–78 for the stiffness of fibrous metallic networks.
Metallic foams have attracted considerable industrial and scientific interest, partly because of advances in processing
technology.79–81 This has led to the prospect of stock and shaped metallic foam components becoming available at prices
comparable with corresponding bulk material (on a weight basis). Also, it has been appreciated that there is potential for the
attainment of attractive property combinations, particularly in terms of specific stiffness82 and specific energy absorption.83
Furthermore, there are many functions, ranging from heat exchange capacity to magnetically-induced bone growth stimulation, for
which metallic fiber networks offer attractive performance. Sound absorption in the exhaust of gas turbine engines is another
example of an application in which metallic fiber network materials perform well.84
As it happens, some types of metallic foam are MMCs in two senses – firstly because all metallic foams can be treated as
constituting a special class of MMC and secondly because the cell walls are often made up of MMC material. This is the case
because, depending on the processing methodology employed, it is often necessary to stabilize the foam against cell coarsening
and collapse while the metal is in a liquid or semiliquid state. This is commonly achieved by introducing a dispersion of ceramic
into the melt, either as oxide films or as ceramic particles, which in effect raise the viscosity of the melt. It may be noted, however,
that the fact that the cell walls of such foams are actually MMCs, and hence have limited ductility, is likely to have an adverse
effect85 on their toughness (and energy-absorbing capability).

4.1.5.2 Cermets
While the term “cermet” is simply a portmanteau word, from “ceramic” and “metal,” so that various materials incorporating both of
these constituents could in principle be encompassed, in practice it is normally used86 to designate an assembly of ceramic particles
bonded together by a small amount of a metallic phase. Some authorities have specified that the ceramic should constitute at least
70% by volume and there should be little solubility between metallic and ceramic phases at the preparation temperature. There is
thus a strong case for regarding cermets as a special class of MMCs. While they could be considered as ceramics that have
been toughened by the presence of a small proportion of metal, in practice the metallic phase often forms some sort of partially
interlinked network, so that they are in effect particulate-reinforced MMC with an exceptionally high proportion of ceramic particles
(see Fig. 4(b)). The literature on cermets is very large and advances continue to be made on their processing and formulation.87–90
On the topic of nomenclature, it should be noted that the terms “hardmetal” and “cemented carbide” are also commonly used
in describing products within this class of material. The differences between these are sometimes unclear and largely arise from
historical origins.86 The term “hardmetal” was originally reserved for the carbides, nitrides, borides and silicides of the metals of
the fourth to sixth group of the periodic table of elements. Prominent among these are ceramics such as tungsten carbide (WC).
These exhibit relatively high thermal and electrical conductivity, but they have mechanical properties, such as high hardness, which
are typical of ceramics. The term has come to signify material produced by bonding together such ceramic particles with a metal
binder, which is usually an alloy of Co, Ni, or Fe. Sometimes, hardmetals based on WC (commonly WC/Co) are called cemented
carbides, while those based on TiC (commonly TiCxN1–x/Mo–Ni) are termed cermets, to indicate their more ceramic character
(e.g., lower electrical conductivity). However, it should be noted that this terminology is not very logical or consistently used. For
example, the thermal conductivity of ceramics can be higher than those of metals, as a consequence of efficient heat transport via
An Introductory Overview of Metal Matrix Composites Systems, Types and Developments 13

phonons, so that metallic character is not necessarily implied by a high value. The electrical conductivity of ceramics, on the other
hand, is normally negligible compared with metals, so that it is the connectivity of the metallic constituent that will be the most
significant factor in determining the apparent electrical resistivity.
It is possible91 to classify cermets into four main groups, as outlined in the paragraphs below.

4.1.5.2.1 Carbide-based cermets


Most cermets in industrial use are of this type. They include cemented carbides based on WC, which are widely used for cutting
tools and wear-resistant parts, TiC cermets for high temperature components in propulsion systems, and chromium carbide
(Cr3C2) cermets for applications requiring good corrosion resistance. The metallic binder may be Ni-, Mo-, Co-, or Al-based. The
system in most common use is WC-Co. It has been estimated86 that 65% of the total world production of tungsten is devoted to
the fabrication of cermets for cutting tools.

4.1.5.2.2 Oxide-based cermets


These include materials comprising particles of Al2O3 in a Cr or Cr–Mo matrix, which is often used when resistance to contact with
liquid metals is required, and SiO2 in brass, bronze or lead, for components to be exposed to high frictional forces. There are also
more specialised systems, such as uranium dioxide (UO2) or thorium dioxide (ThO2) in aluminum, stainless steel, or tungsten,
which are used as fissile constituents in nuclear reactor fuel elements.

4.1.5.2.3 Boride-based cermets


In this case the ceramic is a boride of one of the transition metals. These cermets are stable against attack by reactive metals in the
molten or vapor state. For example, a combination of ZrB2 and SiC, in a boron matrix, is resistant against erosion from the
propulsion gases of chemical rockets.91

4.1.5.2.4 Carbon-containing cermets


These cover materials which contain free graphite, usually present to enhance electrical contact or to provide surface lubrication.
The metal binder is usually tin, lead, or zinc. However, the class also encompasses material comprising diamond particles in
metallic matrices, which are used in special tools.91 There has been interest in Al-based MMCs containing diamond reinforcement,
in the form of both particles92 and fibers,93 although, as with other carbonaceous reinforcements in carbide-forming metals, there
is a serious danger of excessive reaction during processing.
One of the keys to the successful exploitation of cermets in industrial applications concerns their ease of processing. They are
commonly produced (as are some particulate MMCs) by blending of ceramic and metallic powders, followed by liquid phase
sintering, although many processing variants are of interest, including spark plasma sintering, application of pressure, etc. Typi-
cally, the ceramic particles used in cermet production are 1–10 mm in diameter. Blending involves a milling operation that tends to
coat the ceramic particles with metal. This is usually followed by cold isostatic pressing, or injection molding, to give the required
shape and then holding at a suitable temperature under vacuum, inert gas or hydrogen. During liquid phase sintering, particle
rearrangement occurs,94 driven by capillarity forces. This may or may not lead to effective densification, depending on the degree
of wetting in the system. In some cases, notably for the oxide-based cermets, it is often necessary to impose uniaxial or hydrostatic
pressure in order to eliminate porosity. This is reminiscent of the melt infiltration process commonly used to form MMCs, but the
very high ceramic content and fine scale of the structure means that, in the absence of wetting, very high pressure might be
necessary to ensure that liquid flows into all the cavities.
One approach to the problem of inadequate wetting is deliberately to promote selected chemical reactions during sintering.
More details of reactive processing techniques are given below in Section 4.1.5.3. However, in general the occurrence of pro-
nounced chemical reaction leads to difficulties in the form of uncontrolled volume changes, heat release, and undesirable phase
formation. Fortunately, the carbides and nitrides of the hardmetals group are in general thermodynamically more stable than the
carbides or nitrides of most binder metals, so that there are few problems of significant chemical attack during sintering. Some
metal ion exchange may occur, and in certain cases there is a degree of dissolution of the ceramic in the metal and reprecipitation
on solidification. Such processes probably have the effect of raising the bond strength, but for most systems there is little danger of
thick, brittle interfacial reaction zones forming during sintering. Of course, this is largely because cermet systems have been chosen
partly with this chemical compatibility in mind, since there are not the same constraints of economics and matrix mechanical
properties which operate when choosing the constituents for MMCs: the Al-, Ti- and Mg-based matrices commonly used for MMCs
are attractive in these terms, but are very reactive in the liquid state, which often leads to problems during this type of processing.
One point which has become well established95–97 is that even cermet systems that are quite stable during sintering, such as WC-
Co, can undergo various chemical reactions if exposed, even for a short time, to very high temperatures (well above the fusion point
of the matrix). This happens, for example, during thermal spraying of cermet coatings, which is in extensive industrial use. Such
reactions are particularly likely in an oxidizing environment, so that the chemistry of the flame and surrounding atmosphere may be
important. Excessive reaction can lead to degradation of coating properties, particularly the resistance to abrasion and erosion.
It may be noted that, even in the absence of noticeable chemical reaction during sintering, the characteristics of the binder
within cermets may be rather different from that in the monolithic state. The binder is commonly in an effectively elastically strain
hardened state,98 probably as a consequence of the high triaxial constraint and the plastic deformation imposed during cooling. It
14 An Introductory Overview of Metal Matrix Composites Systems, Types and Developments

Fig. 7 Dependence on Co content of: (a) hardness, (b) abrasion rate, (c) density, and (d) fracture toughness of sintered WC-Co cermets. These
data were compiled from the work of Santhanam, A., Tierney, P., Hunt, J.L., 1990. Cemented carbides. In: Properties and Selection: Non-Ferrous
Alloys and Special Purpose Materials. Materials Park, OH: ASM, pp. 950–977 and Lueth, R.C., 1974. Determination of fracture toughness
parameters for tungsten carbide-cobalt alloys. In: Bradt, R.C. (Ed.), Fracture Mechanics of Ceramics. New York, NY: Plenum Press, pp. 791–806.

may also exhibit metastable phases. Nevertheless, in general the presence of the binder does confer a substantial increase in
toughness, as well as facilitating the consolidation process.
In fact, the mechanical properties of a typical cermet, such as sintered WC-Co, can cover quite a wide range, depending primarily
on the Co content, as well as the details of the manufacturing method. This is illustrated by the data in Fig. 7. Both hardness and
abrasion resistance fall off with increasing Co content, and also with increasing particle size. Of course, density varies inversely with
Co content and is independent of particle size. Fracture toughness,99–102 on the other hand, rises with increasing Co content and also
with WC particle size. These trends are rather similar to those exhibited by conventional particulate MMCs, except for the effect of
particle size on toughness. In particulate MMCs (with relatively low volume fractions of reinforcement), the matrix regions close to
ceramic particles tend to experience high plastic strains during loading, often leading to cavitation and cracking, and this is more
pronounced with larger particles. In cermets, however, all of the matrix (binder) is close to ceramic particles and the key issue for
An Introductory Overview of Metal Matrix Composites Systems, Types and Developments 15

toughness is the degree to which the matrix is constrained by the presence of the ceramic from undergoing plastic straining. As with
other toughening mechanisms in composites, such as fiber pullout and plastic deformation of bridging fiber ligaments, coarsening of
the scale tends to cause an increase in the work done during fracture. It would appear that a coarser particle size in cermets leads to an
increase in the degree of plastic work (ahead of a crack tip) that takes place in the matrix. Of course, it is possible that other factors
could contribute, such as weaker interfacial bonding with finer particles, due to incomplete penetration of the binder phase.
In fact, it seems likely that research into optimization of particulate MMCs could benefit from improved familiarity with the
extensive experience and expertise associated with cermet technology. For example, the promotion of higher interfacial bond
strengths is probably desirable for many MMCs and study of the mechanisms by which this is achieved in certain cermets may be
instructive. On the other hand, it is possible that the understanding of areas such as processing, residual stresses, constraint effects
and fracture mechanisms, acquired during study of MMCs, may lead to further improvements in the economics of production or
the mechanical performance characteristics of cermets.
Irrespective of the details of the dependence of properties on particle size and binder content, it can be seen from Fig. 7 that
very attractive combinations of room temperature hardness and toughness are obtainable with sintered cermets. Of course, for
applications such as cutting tools, the behavior at high temperature is also very important. The binder will tend to soften as the
temperature is raised, but this will not necessarily impair the hardness very much, particularly at relatively low binder levels, since
it is heavily constrained by the presence of the surrounding particles. Cermets thus exhibit excellent hot hardness (markedly
superior to high speed tool steels) in the range 700–10001C. However, at temperatures above this, grain boundary sliding of the
carbides starts to occur, leading to extensive plastic deformation. This is pronounced in WC-Co cermets. The TiC cermets,
commonly having TiCN–Mo–Ni formulations, are more resistant to this effect and are thus commonly used in high speed cutting
operations which may raise the temperature of the tool above 10001C. For such applications, their exceptional hot hardness offsets
the rather lower hardness and toughness these systems exhibit at room temperature, when compared with WC-Co. Of course,
many other (more complex) cermet formulations are also available.

4.1.5.3 Reactively Processed MMCs


A concept that has attracted interest is that of MMCs produced by passage of a molten metal infiltration front through a packed
ceramic bed of some sort, with or without associated chemical reaction. Such materials sometimes bear a marked resemblance to
the cermets described in the preceding section, but a distinction can be drawn in terms of certain differences in processing
conditions and in the alloy and ceramic systems which are commonly involved.
Much of the work done in this area has its origins in attempts to facilitate the melt infiltration process, particularly with Al-based
melts. When liquid metal is injected into a relatively fine array of particles or fibers, the applied pressure needed to generate the
necessary meniscus curvature at the infiltration front can be very large42,103,104 – typically at least several MPa. If the scale of the
reinforcement were submicron, then even higher pressures would be needed. Such high pressures are difficult to apply safely using
pneumatic systems and often require the introduction of hydraulic ram systems, adding considerably to processing cost and com-
plexity. Pressure application via centrifuging has also been explored,105 although it’s difficult to generate very high pressures this way.
For these reasons, much effort has been invested in a search for methods by which this requirement for the application of pressure
can be reduced or eliminated. Ideally, “wetting” and/or chemical reaction occurring at the infiltration front would be such as to
promote spontaneous infiltration without the need for external application of pressure. While it has proved difficult to promote rapid
infiltration under these conditions, MMC products can be made in this way with good near-net-shape characteristics, particularly
when the ceramic content is high. The basic problem when attempting to infiltrate liquid Al into Al2O3 or SiC preforms is one of
poor wetting. The introduction of wetting agents such as K2ZrF6 was found106 to be quite effective, but such additions are cum-
bersome and inconvenient. It became clear, however, that wetting could be strongly promoted by chemical reactions arising from the
introduction of Mg into the melt and nitrogen into the surrounding atmosphere (preferably in the absence of oxygen). Such
conditions are experimentally fairly easy to arrange. Compounds such as AlN and MgAl2O4 are then formed.107–109
While the details of the thermodynamic and kinetic characteristics involved are complex, the important point is that the reactions
involved take place at locations and at rates such as to promote spontaneous penetration of an Al-based melt into an array of ceramic
particles or fibers. Commonly, the melt employed is a binary or multicomponent Al–Mg alloy and the process is carried out under a
nitrogen atmosphere. This type of processing, for which trademarks such as “DIMOX” and “PRIMEX” have been used, are described
in some detail.110–116 There have also been claims117 that arrays of carbon nanotubes can be infiltrated (by metallic melts) without
the application of pressure, although an obvious problem with ultra-fine reinforcement is that almost any reaction (needed for
infiltration) will entirely consume it. This is particularly problematic for carbon in Al, which rapidly forms a carbide (Al4C3) that is
hygroscopic. There has been work118 on limiting the formation of this carbide in such systems. This is difficult, although a possible
approach is to incorporate a very strong carbide-former, leading to the in situ creation of, say, TiC reinforcement in an Al MMC.119
Unfortunately, the reaction kinetics and local melt flow characteristics are such that, even when infiltration does occur
spontaneously, it tends to be relatively slow. Typically it might take many minutes, or even hours, for a preform to become fully
infiltrated. This problem, together with the fact that the presence of relatively high reaction product contents may impair
mechanical properties, has meant that reactive processing is not extensively used for production of “conventional” low cost MMCs,
for applications such as pistons or brake disks. However, the process has clear attractions for more specialized applications,
particularly those requiring high ceramic contents – which are rather difficult to produce by standard infiltration or powder
processing methods. Again, there is clear overlap with cermet technology here. For example, the Al  70 vol%SiC particulate MMC
16 An Introductory Overview of Metal Matrix Composites Systems, Types and Developments

Fig. 8 Optical micrograph showing the microstructure of an Al-70 vol% SiC MMC produced using the PRIMEX process. Reproduced from Singh,
J.R., 1999. The role of composites as an enabling materials technology for transition to 300 mm precision systems. Future Fab 2 (5). Available at:
http://www.future-fab.com.

Fig. 9 Components of a chuck for securing Si wafers during processing of electronic devices, made from PRIMEX processed Al-70 vol% SiC
metal matrix composite (MMC) material. Reproduced from Singh, J.R., 1999. The role of composites as an enabling materials technology for
transition to 300 mm precision systems. Future Fab 2 (5). Available at: http://www.future-fab.com.

shown in Fig. 8 was produced using the PRIMEX process, in the form of the components shown in Fig. 9. These are part of a
specialized chuck used in the electronics industry for wafer handling, which has demanding stiffness, conductivity and thermal
expansivity requirements.120 Note the bimodal size distribution of the SiC particles in this micrograph, which facilitates the
generation of high ceramic contents in the initial powder compact.

4.1.5.4 IMC
Several intermetallic compounds have attractive combinations of properties, particularly in terms of their high temperature
resistance to creep, oxidation121 and wear. The major drawback concerning their use for structural purposes is commonly their
relatively poor fracture toughness at relatively low temperatures. They therefore share many characteristics with ceramics, although
the partially metallic nature of the interatomic bonding leads to relatively high values for properties such as electrical conductivity
and there is often at least some scope for promoting dislocation mobility and designing the microstructure so as to toughen the
material somewhat. The production of composites with intermetallic matrices122,123 thus usually has the primary aim of raising
the fracture energy of the material relative to that of the unreinforced matrix.
An Introductory Overview of Metal Matrix Composites Systems, Types and Developments 17

There are two broad approaches to the toughening of intermetallics by the introduction of a reinforcing constituent, which is
usually in fibrous form. The first is to add a relatively tough and ductile fiber,124 commonly a high melting point metal. There may
be problems of interfacial reaction during processing and coated fibers have been explored125 in attempts to control this. Basically,
the toughening in these systems comes simply from energy absorbed during plastic deformation of the reinforcement: fibers are
preferred since, depending on their geometrical arrangement, it is difficult for crack propagation to occur without fiber defor-
mation. This is similar to the toughening.
The second approach is to introduce a reinforcement that does not itself have a high toughness, but raises the fracture energy by
promoting interfacial bonding and consequent crack deflection. This mechanism can operate even when the reinforcement is in
particulate form, although with both fibers and particles the potential for toughening tends to be lower than with ductile
reinforcement. The detailed mechanics of toughening by crack deflection with a brittle matrix, and optimization of processing126
so as to promote microstructural features that enhance the toughness via this mechanism, are covered in several chapters other
volumes of the present series. There is a degree of uncertainty about whether IMCs are more appropriately considered as a subset of
MMCs or as part of the ceramic matrix composites (CMC) family.
The intermetallic systems that have received most attention are the aluminides, particularly Ti3Al, Ti2AlNb, TiAl, Ni3Al and NiAl.
There is also interest in certain metal silicides, such as MoSi2, which have excellent oxidation resistance. In general, the commercial
exploitation of IMCs has remained at a low level, although there have been extensive demonstrator trials. This may be partly because
in many cases optimization of the microstructural features exhibited by the intermetallics themselves is not yet complete. This is quite
a complex issue, with substantial changes in properties often arising from rather subtle microstructural modifications, and the
presence of reinforcement often tends to disturb the evolution of microstructure during processing. There are also problems of added
cost and complexity, given that most intermetallics are themselves rather expensive and difficult to process.

4.1.5.5 Graded MMC Coatings and Structures


Materials in which metallic and ceramic constituents are incorporated into planar structures, with continuous or discrete changes
in the proportions of the two, have attracted interest over an extended period. These are often in the form of thin coatings on
massive substrates, although it is also possible to build layered structures up into bulk laminated material of the type outlined in
Section 4.1.4. Reviews cover processing aspects127 and also thermo-mechanical characteristics128 of such systems. In many cases,
the aim is to combine desirable features of a ceramic, such as good wear resistance or thermal insulation properties, with the
toughness and adhesion to a substrate that are often more readily achievable with metallic materials. There may also be objectives
related to the establishing of control over residual stress distributions by using constituents with selected thermal expansivities or
creep characteristics. An example is provided by thermal barrier coating systems, which conventionally consist of a metallic bond
coat and a ceramic top coat. Attempts have been made to improve the mechanical stability of such coatings by grading the
composition between entirely ceramic at the free surface and entirely metallic at the substrate, although in this particular case there
is often a problem of enhanced oxidation rates of the metallic constituent with such structures. However, in general it is often
found that tailored distributions of metal/ceramic proportions through the thickness of a graded structure can lead to improve-
ments in the achievable property combinations exhibited by the system.

4.1.5.6 MMCs Containing Nanoscale Reinforcement


The so-called “whiskers” that result when fine, slender fibers are produced as monocrystals created considerable interest as early as
the 1960s. Whiskers are usually submicron, or perhaps around a micron, in diameter, with aspect ratios of up to several hundred.
Their tensile strengths are often very high and, depending on the material and crystallographic growth direction, they may also
exhibit higher elastic moduli along the fiber axis than the corresponding polycrstal.129 There is particular interest in ceramics, and
whiskers of materials such as SiC, Si3N4 and Al2O3 have been subjected to widespread investigation. Some of these can be
produced quite cost-effectively, an example being extraction of SiC whiskers from rice husks.130 This led to a surge in interest for
their use in various types of application, including incorporation into MMCs.131–137 Of course, such studies have been followed
over the past couple of decades by an explosion of interest in ultra-fine graphene-based material, including carbon nanotubes
(CNTs). This has encompassed extensive attempts to incorporate such constituents into MMCs, despite the fact that, at least for
many of the most relevant metals, the tendency to form carbides at elevated temperature means that it is very difficult to avoid
degradation of the reinforcement during production of the composite. More recently, a process has been developed138 for the bulk
(kg per hour) production (from a large Al melt) of single crystal g-alumina fibers with diameters B10 nm and lengths in the cm
range. There has also been a lot of interest in nano-particles (for reinforcement of composites).
However, leaving aside the issues of cost-effective production of such ultra-fine reinforcement, there are some serious diffi-
culties with the concept of their incorporation into composite materials, and particularly into MMCs. Firstly, they are difficult to
handle, and potentially hazardous. Particles and, particularly, fibers in the submicron size range can readily become airborne, and
tend to remain suspended in air for prolonged periods. Furthermore, they are likely to reach the lungs, since they tend to evade the
natural protective mechanisms operative in the nasal passages and throat. Such dangers were recognized several decades ago for
asbestos fibers and were highlighted in the 1980s for the particular case of SiC whiskers.139 Of course, it is possible to handle such
materials safely, but the necessary precautions do add to the cost. Secondly, it is often very difficult to incorporate ultra-fine fibers
18 An Introductory Overview of Metal Matrix Composites Systems, Types and Developments

or particles into a (metallic) matrix so that they are uniformly dispersed and well bonded to the matrix. As mentioned above, there
is often a particular problem of chemical reaction with carbon-based reinforcement, but the difficulty of dispersion is common to
all nano-particles and nano-fibers.
Perhaps the most significant issue, however, and one that is not addressed as frequently or clearly as it should be, is that of
whether use of ultra-fine reinforcement is in fact expected to result in composites with superior (mechanical) properties. It is, of
course, true that the tensile strength of ceramic constituents, particularly fibers, tends to rise as their scale is reduced. This is simply a
consequence of the critical flaw size effect, with the largest crack present in a fiber expected to scale with its diameter. However, the
tensile strength of the fibers (or particles) in a composite material, while clearly of significance, is far from being the only, or even the
most important “micro-property” of the system. Firstly, having very strong fibers is of little or no benefit if the interfacial bond
strength and/or the fiber aspect ratio are such that they will not become heavily stressed when the composite is put under load. This is
clearly an issue for particles, with their “strength” being of little relevance, but it also applies to many “short” fibers. Secondly, the
most important (mechanical) property of composites is often their toughness (rather than their “failure strength”). In fact, a major
attraction of (polymer) composites is their high toughness, which is largely attributable74 to the energy absorbed during fiber pullout.
It turns out that, for a given pullout aspect ratio, this contribution to the work of fracture is directly proportional to the fiber diameter.
It is arguable that a “stronger” fiber may lead to a higher pullout aspect ratio, although in fact this is rather doubtful, and a low
Weibull modulus (i.e., a scatter in strength due to a distribution of flaw sizes) favours a high pullout aspect ratio. It is therefore
expected, and in general quite widely observed, that the toughness of a composite structure decreases as it becomes very fine. This
apparently applies to MMCs, including particulate MMCs, as well as to polymer and CMC. As outlined in Section 4.1.5.2 (Fig. 7
(d)), it certainly applies to cermets, at least over the range 1–10 mm, and it also applies140 to metal fiber-reinforced ceramics. The
underlying reason is the same in all cases, which is that the volume of material at the crack tip, within which energy-absorbing
processes (pullout, fiber rupture and/or matrix plasticity) take place, becomes smaller as the scale of the structure is reduced.
Of course, the obvious way to check on this is measure the toughness (and other properties) of composites containing ultra-
fine reinforcement (fibrous or particulate). Unfortunately, there is very little information of this type available in the literature (for
MMCs or other types of composite). There are a few papers141–144 in which it is claimed that composites containing ultra-fine
reinforcement have exhibited good toughness (or at least, for PMCs, an improvement on the unreinforced matrix), but these are
very limited in scope and often refer to material with low reinforcement contents (B1%), which should not really be considered to
be composites at all. Very few such studies involve a comparison with the effect of corresponding coarser scale reinforcement.
Some work145 has been done involving rubber nano-particles in epoxy, giving legitimate toughening, but the mechanisms
involved in this case are rather different and in fact were established some time ago (in the development of “High Impact
Polystyrene”): they do not really depend on the particles being ultra-fine.
There appear to be virtually no publications in which the toughness of MMCs containing nanoscale reinforcement has been
reported. Undoubtedly a lot of work has in fact been done that yielded low toughness values, but such results are very rarely
reported (with the conclusion perhaps being drawn that it was attributable to poor microstructural quality, and hence not worth
reporting). In any event, it would be timely now for these fundamental difficulties associated with ultra-fine reinforcement to be
more widely recognized.

See also: 4.2 Effect of Interface Strength on Metal Matrix Composites Properties. 4.3 Void Formation in Metal Matrix Composites. 4.4 Fracture
Toughness and Fatigue of Particulate Metal Matrix Composites. 4.5 Wear of Particulate Metal Matrix Composites. 4.6 Mechanical Properties of
Metallic Fiber Network Materials. 4.7 Thermal and Electrical Conduction in Metal Matrix Composites. 4.8 Thermal Stresses and Thermal
Expansion in Metal Matrix Composites. 4.9 Microstructural Design of Metal Matrix Composites for High Temperature Strength and Superplastic
Behavior by Strain Mismatch. 4.10 Residual Stresses in Metal Matrix Composites. 4.11 Aqueous Corrosion of Metal Matrix Composites. 4.12
Processing of Nanoparticulate Metal Matrix Composites. 4.13 Gradient Metal Matrix Composites. 4.14 Production of Metal Foams. 4.15 Metal
Matrix Syntactic Foams. 4.16 Metal Matrix Composite Thermal Management Materials. 4.17 Metal Fiber Network Materials for Compact Heat
Exchangers. 4.18 Metal Fiber Network Materials for Magnetically-Induced Bioactivation. 4.19 Development of Continuously-Reinforced Metal
Matrix Composites for Aerospace Applications. 4.20 Metal Fibre-Reinforced Ceramic Composites and Their Industrial Usage. 4.21 Intermetallic
Matrix Composites. 4.22 Metal–Organic Framework Based Composites. 4.23 Design Aspects of Metal Matrix Composite Usage

References

1. Muhly, J.D., 1988. The beginnings of metallurgy in the old world. In: Maddin, R. (Ed.), The Beginning of the Use of Metals & Alloys. Cambridge, MA: MIT Press, pp. 2–20.
2. Clyne, T.W., Withers, P.J., 1993. An introduction to metal matrix composites. In: Davis, E., Ward, I. (Eds.), Cambridge Solid State Science Series. Cambridge: Cambridge
University Press.
3. Chawla, N., Chawla, K.K., 2014. Metal Matrix Composites. New York, NY: Springer.
4. Clyne, T.W., 1996. The effect of interfacial characteristics on the mechanical performance of particulate, fibrous and layered metal matrix composites – A review of some
recent work. Key Engineering Materials 127–131, 81–98.
5. Kallas, M.N., Koss, D.A., Hahn, H.T., Hellmann, J.R., 1992. Interfacial stress state present in a “thin slice” fiber push-out test. Journal of Materials Science 27, 3821–3826.
6. Roman, I., Aharonov, R., 1992. Mechanical interrogation of interfaces in monofilament model composites of continuous SiC fiber-aluminium matrix. Acta Metallurgica et
Materialia 40 (3), 477–485.
7. Watson, M.C., Clyne, T.W., 1993. The tensioned push-out test for measurement of fibre/matrix interfacial toughness under mixed mode loading. Materials Science and
Engineering: A 160, 1–5.
An Introductory Overview of Metal Matrix Composites Systems, Types and Developments 19

8. Kalton, A.F., Miracle, D.B., Clyne, T.W., 1997. The effect of interfacial strength on the response of Ti MMCs to single fibre push-out and transverse tensile testing. Key
Engineering Materials 127–131, 659–670.
9. Kalton, A.F., Howard, S.J., Janczak-Rusch, J., Clyne, T.W., 1998. Measurement of interfacial fracture energy by single fibre push-out testing and its application to the
titanium–silicon carbide system. Acta Materialia 46, 3175–3189.
10. Lewandowski, J.J., Liu, C., Hunt, W.H., 1989. Effects of matrix microstructure and particle distribution on fracture of an Al MMC. Materials Science and Engineering: A
107, 241–255.
11. Manoharan, M., Lewandowski, J.J., 1989. Effects of aging condition on the fracture toughness of 2XXX & 7XXX series Al alloy composites. Scripta Metallurgica 23, 301–304.
12. Singh, P.M., Lewandowski, J.J., 1993. Effects of heat treatment and reinforcement size on reinforcement fracture during tension testing of a SiCp dispontinuously
reinforced aluminium alloy. Metallurgical and Materials Transactions A 24, 2531–2543.
13. Marchal, Y., Delannay, F., Froyen, L., 1996. The essential work of fracture as a means for characterizing the influence of particle size and volume fraction on the fracture
toughness of plates of Al/SiC composites. Scripta Materialia 35, 193–198.
14. Maire, E., Wilkinson, D.S., Embury, J.D., Fougeres, R., 1997. Role of damage on the flow and fracture of particulate reinforced alloys and metal matrix composites. Acta
Materialia 45, 5261–5274.
15. Wilkinson, D.S., Maire, E., Embury, J.D., 1997. The role of heterogeneity on the flow and fracture of two-phase materials. Materials Science and Engineering: A 233, 145–154.
16. Murphy, A.M., Howard, S.J., Clyne, T.W., 1998. Characterisation of the severity of particle clustering and its effect on the fracture of particulate MMCs. Materials Science
and Technology 14, 959–968.
17. Dellis, M.A., Keustermans, J.P., Delannay, F., Wegria, J., 1991. Zn–Al matrix composites – Investigation of the thermal expansion, creep resistance and fracture
toughness. Materials Science and Engineering: A 135, 253–257.
18. Skibo, M., Morris, P.L., Lloyd, D.J., 1988. Structure and properties of liquid metal processed SiC reinforced aluminium. In: Fishman, S.G., Dhingra, A.K. (Eds.), Cast
Reinforced Metal Composites. Chicago: ASM.
19. Lloyd, D.J., Lagace, H., McLeod, A., Morris, P.L., 1989. Microstructural aspects of aluminium-silicon carbide particulate composites produced by a casting method.
Materials Science and Engineering: A 107, 73–80.
20. Sritharan, T., Chan, L.S., Tan, L.K., Hung, N.P., 2001. A feature of the reaction between Al and SiC particles in an MMC. Materials Characterization 47 (1), 75–77.
21. Kooi, B.J., Kabel, M., Kloosterman, A.B., de Hosson, J.T.M., 1999. Reaction layers around SiC particles in Ti: An electron microscopy study. Acta Materialia 47 (10),
3105–3116.
22. Zou, Y., Sun, Z.M., Hashimoto, H., Cheng, L., 2010. Reaction mechanism in Ti–SiC–C powder mixture during pulse discharge sintering. Ceramics International 36 (3),
1027–1031.
23. Zhang, W., Yang, Y.Q., Zhao, G.M., et al., 2013. Investigation of interfacial reaction in SiC fiber reinforced Ti–43Al–9V composites. Intermetallics 33, 54–59.
24. Turner, S.P., Taylor, R., Gordon, F.H., Clyne, T.W., 1996. Thermal properties of Ti–SiC and Ti–TiB2 reinforced composites. International Journal of Thermophysics 17, 239–251.
25. Zhang, W., Yang, Y.Q., Zhao, G.M., et al., 2014. Interfacial reaction studies of B4C-coated and C-coated SiC fiber reinforced Ti–43Al–9V composites. Intermetallics 50, 14–19.
26. Ferraris, M., Badini, C., Marino, F., Marchetti, F., Girardi, S., 1993. Interfacial reactions in a Ti–6Al–4V based composite: Role of the TIB2 coating. Journal of Materials
Science 28 (7), 1983–1987.
27. Pelleg, J., 1999. Reactions in the matrix and Interface of the Fe–SiC metal matrix composite system. Materials Science and Engineering: A 269, 225–241.
28. Buytoz, S., Ulutan, M., 2006. In situ synthesis of SiC reinforced MMC surface on AISI 304 stainless steel by TIG surface alloying. Surface & Coatings Technology 200
(12–13), 3698–3704.
29. Tanaka, K., Saito, T., 1999. Phase equilibria in TiB2-reinforced high modulus steel. Journal of Phase Equilibria 20 (3), 207–214.
30. Aparicio-Fernández, R., Springer, H., Szczepaniak, A., Zhang, H., Raabe, D., 2016. In-situ metal matrix composite steels: Effect of alloying and annealing on morphology,
structure and mechanical properties of TiB2 particle containing high modulus steels. Acta Materialia 107, 38–48.
31. Springer, H., Fernandez, R.A., Duarte, M.J., Kostka, A., Raabe, D., 2015. Microstructure refinement for high modulus in-situ metal matrix composite steels via controlled
solidification of the system Fe–TiB2. Acta Materialia 96, 47–56.
32. Zhang, H., Springer, H., Aparicio-Fernandez, R., Raabe, D., 2016. Improving the mechanical properties of Fe–TiB2 high modulus steels through controlled solidification
processes. Acta Materialia 118, 187–195.
33. Baron, C., Springer, H., Raabe, D., 2016. Efficient liquid metallurgy synthesis of Fe–TiB2 high modulus steels via in-situ reduction of titanium oxides. Materials & Design
97, 357–363.
34. Hallstedt, B., Liu, Z.K., Agren, J., 1993. Reactions in Al2O3–Mg metal-matrix composites during prolonged heat-treatment AT 400-degrees-C, 550-degrees-C and 600-
degrees-C. Materials Science and Engineering A – Structural Materials Properties Microstructure and Processing 169 (1–2), 149–157.
35. Shahani, R.A., Clyne, T.W., 1991. Recrystallization in fibrous and particulate metal matrix composites. Materials Science and Engineering: A 135, 281–285.
36. Nakanishi, H., Kakihara, K., Nakayama, A., Murayama, T., 2002. Development of aluminum metal matrix composites (Al-MMC) brake rotor and pad. JSAE Review 23 (3),
365–370.
37. Shorowordi, K.M., Haseeb, A.S.M., Celis, J.P., 2004. Velocity effects on the wear, friction and tribochemistry of aluminum MMC sliding against phenolic brake pad. Wear
256 (11–12), 1176–1181.
38. Gultekin, D., Uysal, M., Aslan, S., et al., 2010. The effects of applied load on the coefficient of friction in Cu-MMC brake pad/Al-SiCp MMC brake disc system. Wear
270 (1–2), 73–82.
39. Jin, Y.X., Lee, J.M., Kang, S.B., 2008. Wear characteristics of dry friction between the A356/SiCp composite and applied organic brake pad of medium speed train. Rare
Metal Materials and Engineering 37 (11), 1956–1960.
40. Pavlov, A.V., Kudelnikova, S.P., Vicharev, A.N., 2015. On the corrosion resistance of half-metallic composite brake pads for railroad cars. Journal of Friction and Wear 36
(2), 123–126.
41. Clyne, T.W., Bader, M.G., Cappleman, G.R., Hubert, P.A., 1985. The use of a d-alumina fibre for metal matrix composites. Journal of Materials Science 20, 85–96.
42. Clyne, T.W., Mason, J.F., 1987. The squeeze infiltration process for fabrication of metal matrix composites. Metallurgical Transactions A 18, 1519–1530.
43. Clegg, W.J., Horsfall, I., Mason, J.F., Edwards, L., 1988. The tensile deformation and fracture of al-Saffil metal matrix composites. Acta Metallurgica 36 (8), 2151–2159.
44. Birchall, J.D., 1983. The preparation and properties of polycrystalline aluminium oxide fibres. Transactions of the British Ceramic Society 82, 143–145.
45. Jayalakshmi, S., Kailas, S.V., Seshan, S., Fleury, E., 2006. Properties of squeeze cast Mg–6Zn–3Cu alloy and its Saffil alumina short fibre reinforced composites. Journal
of Materials Science 41 (12), 3743–3752.
46. Mondal, A.K., Rao, B., Kumar, S., 2007. Wear behaviour of AE42 þ 20% Saffil Mg-MMC. Tribology International 40 (2), 290–296.
47. Trojanova, Z., Lukac, P., Riehemann, W., 2009. Damping behaviour of a Mg–Al–Ca alloy reinforced by short Saffil fibres. Materials Science and Engineering A –
Structural Materials Properties Microstructure and Processing 521–522, 314–317.
48. Cappelman, G.R., Watts, J.F., Clyne, T.W., 1985. The interface region in squeeze-infiltrated composites containing δ-alumina fibre in an aluminium matrix. Journal of
Materials Science 20, 2159–2168.
49. Kaufmann, H., Mortensen, A., 1992. Wetting of Saffil alumina fiber preforms by aluminium at 973K. Metallurgical Transactions A 23, 2071–2073.
50. Liu, H.N., Miyahara, H., Ogi, K., 1998. Fabrication of Al2O3 continuous fibre reinforced Al–Cu alloys by axial infiltration process. Materials Science and Technology 14, 292–298.
51. Bobzin, K., Bagcivan, N., Parkot, D., Kashko, T., 2010. Calculation of effective properties of textile reinforced aluminum alloy by a two-step homogenization procedure.
Computational Materials Science 47 (3), 801–806.
52. Hall, I.W., 1991. The interface in carbon-magnesium composites: Fibre and matrix effects. Journal of Materials Science 26, 776–781.
20 An Introductory Overview of Metal Matrix Composites Systems, Types and Developments

53. Kagawa, Y., Nakata, E., 1992. Some mechanical properties of carbon fibre reinforced magnesium matrix composites fabricated by squeeze casting. Journal of Materials
Science Letters 11, 176–178.
54. Russell-Stevens, M., Todd, R., Papakyriacou, M., 2005. Microstructural analysis of a carbon fibre reinforced AZ91D magnesium alloy composite. Surface and Interface
Analysis 37 (3), 336–342.
55. Huang, S.J., Ho, C.H., Feldman, Y., Tenne, R., 2016. Advanced AZ31 Mg alloy composites reinforced by WS2 nanotubes. Journal of Alloys and Compounds 654, 15–22.
56. Popovska, N., Gerhard, H., Wurm, D., et al., 1997. Chemical vapor deposition of titanium nitride on carbon fibres as a protective layer in metal matrix composites.
Materials & Design 18, 239–242.
57. Wendt, R.G., Moshier, W.C., Shaw, B., Miller, P., Olson, D.L., 1994. Corrosion-resistant aluminium matrix for graphite aluminium composites. Corrosion 50, 819–826.
58. Lavaste, V., Berger, M.H., Bunsell, A.R., Besson, J., 1995. Microstructure and mechanical characteristics of alpha-alumina-based fibres. Journal of the American Ceramic
Society 30, 4215–4225.
59. Zhu, J., Hihara, L.H., 2010. Corrosion of continuous alumina-fibre reinforced Al-2 wt% Cu–T6 metal-matrix composite in 3.15 wt% NaCl solution. Corrosion Science 52
(2), 406–415.
60. Moser, B., Weber, L., Mortensen, A., 2005. Damage accumulation during cyclic loading of a continuous alumina fibre reinforced aluminium composite. Scripta Materialia
53 (10), 1111–1115.
61. Warren, J., Elzey, D.M., Wadley, H.N.G., 1995. Fiber damage during the consolidation of PVD Ti–6Al–4V coated nextel 610(TM) alumina fibers. Acta Metallurgica et
Materialia 43, 3605–3619.
62. Kieschke, R.R., Clyne, T.W., 1991. Development of a diffusion barrier for SiC monofilaments in titanium. Materials Science and Engineering: A 135, 145–149.
63. Haque, S., Choy, K.L., 2000. Push-out testing of SiC monofilaments with a TiC based functionally graded coating. Journal of Materials Science 35 (17), 4225–4229.
64. Shatwell, R.A., 1995. Adhesion of SM1240 þ coatings to silicon carbide substrate in sigma monofilament. Materials Science and Technology 10, 552–557.
65. Spowart, J.E., Clyne, T.W., 1998. The axial compressive failure of titanium reinforced with silicon carbide monofilaments. Acta Materialia 47, 671–687.
66. Bucci, R.J., Mueller, R.N., Vogelesang, L.B., Gunnink, J.W., 1987. ARALL laminates – Properties and design update. Journal of Metals 39, A58.
67. Qaiser, H., Umar, S., Nasir, A., Shah, M., Nauman, S., 2015. Optimization of interlaminar shear strength behavior of anodized and unanodized ARALL composites
fabricated through VARTM process. International Journal of Material Forming 8 (3), 481–493.
68. Vasek, A.W., Polak, J., Kozak, V., 1997. Fatigue crack initiation in fiber-metal laminate glare-2. Materials Science and Engineering: A 234, 621–624.
69. Kotik, H.G., Ipina, J.E.P., 2017. Short-beam shear fatigue behavior of fiber metal laminate (glare). International Journal of Fatigue 95, 236–242.
70. Shaw, M.C., Marshall, D.B., Dadkhah, M.S., Evans, A.G., 1993. Cracking and damage mechanisms in ceramic/metal multilayers. Acta Metallurgica et Materialia 41 (11),
3311–3320.
71. Shaw, M.C., Clyne, T.W., Cocks, A.C.F., Fleck, N.A., Pateras, S.K., 1996. Cracking patterns in ceramic/metal laminates. Journal of the Mechanics and Physics of Solids
44 (5), 801–814.
72. Pateras, S.K., Howard, S.J., Clyne, T.W., 1997. The contribution of bridging ligament rupture to energy absorption during fracture of metal-ceramic laminates. Key
Engineering Materials 127–131, 1127–1136.
73. Hwu, K.L., Derby, R., 1999. Fracture of metal/ceramic laminates – II. Crack growth resistance and toughness. Acta Materialia 47 (2), 545–563.
74. Hull, D., Clyne, T.W., 1996. An introduction to composite materials. In: Clarke, D.R., Suresh, S., Ward, I.M. (Eds.), Cambridge Solid State Science Series. Cambridge:
Cambridge University Press.
75. Lu, T.J., Stone, H.A., Ashby, M.F., 1998. Heat transfer in open-cell metal foams. Acta Materialia 46 (10), 3619–3635.
76. Gibson, L.J., Ashby, M.F., 1997. Cellular Solids: Structure and Properties, second ed. Cambridge: Cambridge University Press, p. 357.
77. Markaki, A.E., Clyne, T.W., 2004. Magneto-mechanical stimulation of bone growth in a bonded array of ferromagnetic fibres. Biomaterials 25 (19), 4805–4815.
78. Markaki, A.E., Clyne, T.W., 2005. Magneto-mechanical actuation of bonded ferromagnetic fibre arrays. Acta Materialia 53 (3), 877–889.
79. Banhart, J., Baumeister, J., 1998. Production methods for metallic foams. In: Shwartz, D.S., Shih, D.S., Evans, A.G., Wadley, H.N.G. (Eds.), Porous and Cellular Materials
for Structural Applications. San Francisco: MRS.
80. Gergely, V., Clyne, T.W., 2000. The formgrip process: Foaming of reinforced metals by gas release In precursors. Advanced Engineering Materials 2, 175–178.
81. Gergely, V., Curran, D.C., Clyne, T.W., 2003. The FOAMCARP process: Foaming of aluminium MMCS by the chalk-aluminium reaction in precursors. Composites
Science and Technology 63, 2301–2310. (Special issue on porous materials).
82. Simone, A.E., Gibson, L.J., 1997. Efficient structural components using porous metals. Materials Science and Engineering: A 229, 55–62.
83. Fusheng, H., Zhengang, Z., Junchang, G., 1998. Compressive deformation and energy absorbing characteristic of foamed aluminium. Metallurgical Transactions A 29, 2497–2502.
84. Golosnoy, I.O., Tan, J.C., Clyne, T.W., 2008. Ferrous fibre network materials for noise reduction in gas turbine aeroengines, part I: Acoustic effects. Advanced Engineering
Materials 10, 192–200.
85. Markaki, A.E., Clyne, T.W., 2001. The effect of cell wall microstructure on the deformation and fracture of aluminium-based foams. Acta Materialia 49 (9), 1677–1686.
86. Mari, D., 2001. Cermets and hardmetals. Composite Materials In: Mortensen, A. Encyclopaedia of Materials. Oxford: Elsevier, pp. 1118–1123.
87. Venkateswaran, T., Basu, B., Raju, G.B., Kim, D.Y., 2006. Densification and properties of transition metal borides-based cermets via spark plasma sintering. Journal of the
European Ceramic Society 26 (13), 2431–2440.
88. Kim, H.C., Shon, I.J., Yoon, J.K., Doh, J.M., Munir, Z.A., 2006. Rapid sintering of ultrafine WC-Ni cermets. International Journal of Refractory Metals & Hard Materials
24 (6), 427–431.
89. Zhang, S.C., Hilmas, G.E., Fahrenholtz, W.G., 2007. Zirconium carbide-tungsten cermets prepared by in situ reaction sintering. Journal of the American Ceramic Society
90 (7), 1930–1933.
90. Deorsola, F.A., Vallauri, D., Villalba, G.A.O., De Benedetti, B., 2010. Densification of ultrafine WC-12Co cermets by pressure assisted fast electric sintering. International
Journal of Refractory Metals & Hard Materials 28 (2), 254–259.
91. Goetzel, C.G., 1984. Cermets. In: Metals Handbook: Powder Metallurgy. Materials Park, OH: American Society of Metals, pp. 798–814.
92. Johnson, W.B., Sonuparlak, B., 1993. Diamond/aluminium metal matrix composites formed by the pressureless metal infiltration process. Journal of Materials Research 8,
1169–1173.
93. May, P.W., Rego, C.A., Thomas, R.M., et al., 1994. Preparation of solid and hollow diamond fibers and the potential for diamond fiber metal matrix composites. Journal
of Materials Science Letters 13 (4), 247–249.
94. German, R.N., 1996. Sintering Theory and Practice, first ed. New York, NY: John Wiley & Sons.
95. Ramnath, V., Jayaraman, N., 1989. Characterization and wear performance of plasma sprayed WC-Co coatings. Materials Science and Technology 5, 382–388.
96. Stoica, V., Ahmed, R., Itsukaichi, T., 2005. Influence of heat-treatment on the sliding wear of thermal spray cermet coatings. Surface & Coatings Technology 199 (1), 7–21.
97. Zhu, H.B., Li, H., Li, Z.X., 2013. Plasma sprayed TiB2–Ni cermet coatings: Effect of feedstock characteristics on the microstructure and tribological performance. Surface
& Coatings Technology 235, 620–627.
98. Sigl, L.S., Fischmeister, H.F., 1988. On the fracture toughness of cemented carbides. Acta Metallurgica 36 (4), 887–897.
99. Lueth, R.C., 1974. Determination of fracture toughness parameters for tungsten carbide-cobalt alloys. In: Bradt, R.C. (Ed.), Fracture Mechanics of Ceramics. New York,
NY: Plenum Press, pp. 791–806.
100. Pickens, J.R., Gurland, J., 1978. The fracture toughness of WC-Co alloys measured on single-edge notched beam specimens precracked by electron discharge machining.
Materials Science and Engineering 33, 135–142.
An Introductory Overview of Metal Matrix Composites Systems, Types and Developments 21

101. Bouaouadja, N., Hamidiuche, M., Osmani, H., Fantozzi, G., 1994. Fracture toughness of WC-Co cemented carbides at room temperature. Journal of Materials Science
Letters 13, 17–19.
102. Ravichandran, K.S., 1994. Fracture toughness of two phase WC-Co cermets. Acta Metallurgica et Materialia 42, 143–150.
103. Mortensen, A., Wong, T., 1990. Infiltration of fibrous preforms by a pure metal: Part III. Capillary phenomena. Metallurgical Transactions A 21, 2257–2263.
104. Manu, K.M.S., Raag, L.A., Rajan, T.P.D., Gupta, M., Pai, B.C., 2016. Liquid metal infiltration processing of metallic composites: A critical review. Metallurgical and
Materials Transactions B – Process Metallurgy and Materials Processing Science 47 (5), 2799–2819.
105. Wannasin, J., Flemings, M.C., 2005. Fabrication of metal matrix composites by a high-pressure centrifugal infiltration process. Journal of Materials Processing
Technology 169 (2), 143–149.
106. Schamm, S., Fedou, R., Rocher, J.P., Quinisset, J.M., Naslain, R., 1991. The K2ZrF6 wetting process: Effects of surface chemistry on the ability of a SiC fibre preform to
be impregnated by aluminium. Metallurgical Transactions A 22, 2133–2139.
107. Aghajanian, M.K., Rocazella, M.A., Burke, J.T., Keck, S.D., 1991. The fabrication of metal matrix composites by a pressureless infiltration technique. Journal of Materials
Science 26, 447–454.
108. Scholz, H., Greil, P., 1991. Nitridation reactions of molten Al-(Mg, Si) alloys. Journal of Materials Science 26, 669–677.
109. Scholz, H., Gunther, R., Rodel, J., Greil, P., 1993. Formation of Al2O3 fibre reinforced AlN/Al matrix composites by Al(Mg) melt nitridation. Journal of Materials Science
Letters 12, 939–942.
110. Schiroky, G.H., Miller, D.V., Aghajanian, M.K., Fareed, A.S., 1997. Fabrication of CMCs and MMCs using novel processes. In: Fuentes, M., Martinez-Esnaola, J.M.,
Daniel, A.M., (Eds.), Cmmc 96 – Proceedings of the First International Conference on Ceramic and Metal Matrix Composites, Pts 1 and 2. pp. 141–152. Clausthal
Zellerfe: Trans Tech Publications.
111. Cui, C.X., Shen, Y.T., Meng, F.B., Kang, S.B., 2000. Review on fabrication methods of in situ metal matrix composites. Journal of Materials Science & Technology 16
(6), 619–626.
112. Rao, B.S., Jayaram, V., 2001. Pressureless infiltration of Al–Mg based alloys into Al2O3 preforms: Mechanisms and phenomenology. Acta Materialia 49 (13), 2373–2385.
113. Ren, S.B., He, X.B., Qu, X.H., Humail, I.S., Li, Y., 2007. Effect of Mg and Si in the aluminum on the thermo-mechanical properties of pressureless infiltrated SiCp/Al
composites. Composites Science and Technology 67 (10), 2103–2113.
114. Wittig, D., Glauche, A., Aneziris, C.G., et al., 2008. Activated pressureless melt infiltration of zirconia-based metal matrix composites. Materials Science and Engineering
A – Structural Materials Properties Microstructure and Processing 488 (1–2), 580–585.
115. Chen, J.C., Hao, C.Y., Zhang, J.S., 2006. Fabrication of 3D-SiC network reinforced aluminum-matrix composites by pressureless infiltration. Materials Letters 60 (20), 2489–2492.
116. Nakae, H., Hiramoto, Y., 2011. Spontaneous infiltration of al melts into sic preform. International Journal of Metalcasting 5 (2), 23–28.
117. Zhou, S.M., Zhang, X.B., Ding, Z.P., et al., 2007. Fabrication and tribological properties of carbon nanotubes reinforced Al composites prepared by pressureless infiltration
technique. Composites Part A – Applied Science and Manufacturing 38 (2), 301–306.
118. Rodríguez-Reyes, M., Pech-Canul, M.I., Rendón-Angeles, J.C., López-Cuevas, J., 2006. Limiting the development of Al4C3 to prevent degradation of Al/SiCP composites
processed by pressureless infiltration. Composites Science and Technology 66 (7–8), 1056–1062.
119. Samer, N., Andrieux, J., Gardiola, B., et al., 2015. Microstructure and mechanical properties of an Al–TiC metal matrix composite obtained by reactive synthesis.
Composites Part A – Applied Science and Manufacturing 72, 50–57.
120. Singh, J.R., 1999. The role of composites as an enabling materials technology for transition to 300 mm precision systems. Future Fab 2 (5), Available at: http://www.
future-fab.com.
121. Grabke, H.J., Brumm, M., Steinhorst, M., 1992. Development of oxidation-resistant high temperature intermetallics. Materials Science and Technology 8, 339–344.
122. Kumar, K.S., Bao, G., 1994. Intermetallic-matrix composites composites: An overview. Composites Science and Technology 52, 127–150.
123. Koch, C.C., 1998. Intermetallic matrix composites prepared by mechanical alloying – A review. Materials Science and Engineering A – Structural Materials Properties
Microstructure and Processing 244 (1), 39–48.
124. Deve, H.E., Maloney, M.J., 1991. On the toughening of intermetallics with ductile fibres: Role of interfaces. Acta Metallurgica et Materialia 39, 2275–2284.
125. Brunet, A., Valle, R., Vassel, A., 2000. Intermetallic TiAl-based matrix composites: Investigation of the chemical and mechanical compatibility of a protective coating
adapted to an alumina fibre. Acta Materialia 48 (20), 4763–4774.
126. Stoloff, N.S., Alman, D.E., 1990. Innovative processing techniques for intermetallic matrix composites. MRS Bulletin 15, 47–53.
127. Mortensen, A., Suresh, S., 1995. Functionally graded metals and metal-ceramic composites: Part I. Processing. International Materials Reviews 40 (6), 239–265.
128. Suresh, S., Mortensen, A., 1997. Functionally graded metals and metal-ceramic composites. Part II – Thermomechanical behaviour. International Materials Reviews 42 (3), 85–116.
129. Parvizi-Majidi, A., 1993. Fibres and whiskers. In: Chou, T.W. (Ed.), Structure and Properties of Composites. New York, NY: VCH.
130. Nutt, S.R., 1988. Microstructure and growth model for rice-hull-derived SiC whiskers. Journal of the American Ceramic Society 71, 149–156.
131. Nardone, V.C., Strife, J.R., 1987. Analysis of the creep behaviour of silicon carbide whisker reinforced 2124 Al (T4). Metallurgical Transactions A 18, 109–114.
132. Matsubara, H., Nishida, Y., Yamada, M., Shirayanagi, I., Imai, T., 1987. Si3N4 whisker reinforced Al alloy composite. Journal of Materials Science Letters 6, 1313–1315.
133. Daehn, G.S., González-Doncel, G., 1989. Deformation of whisker-reinforced metal-matrix composites under changing temperature conditions. Metallurgical Transactions A
20, 2355–2368.
134. Mason, J.F., Warwick, C.M., Smith, P., Charles, J.A., Clyne, T.W., 1989. Magnesium–lithium alloys in metal matrix composites – A preliminary report. Journal of
Materials Science 24, 3934–3946.
135. Dutta, I., Sims, J.D., Seigenthaler, D.M., 1993. An analytical study of residual stress effects on uniaxial deformation of whisker reinforced metal matrix composites. Acta
Metallurgica et Materialia 41, 885–908.
136. Christman, T., Needleman, A., Suresh, S., 1990. Microstructural development in an Al-alloy SiC whisker composite. Acta Metallurgica 37, 3029–3050.
137. Mei, Z., Zhu, Y.H., Lee, W.B., Yue, T.M., Pang, G.K.H., 2006. Microstructure investigation of a SiC whisker reinforced eutectoid zinc alloy matrix composite. Composites
Part A – Applied Science and Manufacturing 37 (9), 1345–1350.
138. Su, V.M.T., Terehov, M., Clyne, T.W., 2012. Filtration performance of membranes produced using nanoscale alumina fibres (NAF). Advanced Engineering Materials
14 (12), 1088–1096.
139. Birchall, J.D., Stanley, D.R., Mockford, M.J., Pigott, G.H., Pinto, P.J., 1988. The toxicity of silicon carbide whiskers. Journal of Materials Science Letters 7, 350–352.
140. Pemberton, S.R., Oberg, E.K., Dean, J., et al., 2011. The fracture energy of metal fibre reinforced ceramic composites (MFCs). Composites Science and Technology 71
(3), 266–275.
141. Brunner, A.J., Necola, A., Rees, M., et al., 2006. The influence of silicate-based nano-filler on the fracture toughness of epoxy resin. Engineering Fracture Mechanics 73
(16), 2336–2345.
142. Yilmaz, C., Korkmaz, T., 2007. The reinforcement effect of nano and microfillers on fracture toughness of two provisional resin materials. Materials & Design 28 (7), 2063–2070.
143. Kim, B.C., Park, S.W., Lee, D.G., 2008. Fracture toughness of the nano-particle reinforced epoxy composite. Composite Structures 86 (1–3), 69–77.
144. Vu, C.M., Choi, H.J., 2016. Fracture toughness and surface morphology of micro/nano sized fibrils-modified epoxy resin. Polymer Science Series A 58 (3), 464–470.
145. Quan, D., Ivankovic, A., 2015. Effect of core-shell rubber (CSR) nano-particles on mechanical properties and fracture toughness of an epoxy polymer. Polymer 66, 16–28.

You might also like