Cellular Ceramics / 5

571

5.9

Interpenetrating Composites

Jon Binner

5.9.1

Introduction

The conventional approaches to producing composites consisting of two phases typically result in materials that consist of a discrete phase dispersed in an otherwise homogeneous matrix material. That is, while the matrix phase is continuous throughout the material, the dispersed phase is either not interconnected in any of the three dimensions or, if continuous fibers are used, connected in a single dimension. Such composites are defined as 0–3 and 1–3 respectively in Fig. 1 [1]. However, a change in connectivity between the phases, from 0–0 (zero connectivity between either of the phases present) through to 3–3 (both phases interpenetrating in all three dimensions) can result in substantially different properties. Consider as an example a mixture of electrically conducting and insulating phases. The two extremes types of connectivity, 0–0 and 3–3, and the intermediate 0–3, are likely to display isotropic behavior with the 0–0 composite being insulating and the 3–3 composite conducting, whilst the other seven levels of connectivity will show different degrees of anisotropy. Truly interpenetrating composites are those that are defined as 3–3.

Although 3–3 connectivity is fairly common in natural composites such as bone and wood, there have been relatively few attempts at creating it synthetically [2]. The difficulty with fabricating a truly interpenetrating network lies in achieving the required connectivity and spatial distribution of the phases, especially on a fine scale. However, the ability to fabricate by design such an interpenetrating microstructure raises the possibility of developing materials with truly multifunctional properties; each phase contributes its own characteristics to the macroscopic properties of the composite. For instance, one phase might provide high strength or wear resistance, while the other contributes a different property such as electrical conductivity. In addition, since many continuum properties change abruptly at the percolation threshold (the limit where connectivity is gained or lost), possibly the greatest potential for making materials capable of exhibiting novel behavior lies with those that have marginally interconnected microstructures [2]. This necessitates the ability to control the fraction and structure of both phases, preferably with a degree of independence. However, until more of these interpenetrating composites are produced and their properties systematically investigated, there is little other guidance as to how the interconnectivity might affect the overall properties.

Cellular Ceramics: Structure, Manufacturing, Properties and Applications.

Michael Scheffler, Paolo Colombo (Eds.)

Copyright 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

ISBN: 3-527-31320-6

572 Part 5 Applications

Fig. 1 The ten different levels of connectivity that can exist in composites, as defined by Newnham [1] (Newnham, R.E., Skinner, D.P., Cross, L.E., Mat. Res. Bull., 1978 13 525–536). Reprinted with permission from Elsevier.

One route to achieving tailorable 3–3 composites is infiltration of a second phase into porous materials that display complete pore interconnectivity. Provided the structure of the initial porous material can be precisely controlled in terms of the degree of porosity, the size and shape of the pores, the size of the windows between them, and the nature of the struts separating them, then there is the opportunity to design and fabricate interpenetrating composites with customized structures. Hence, the infiltration of cellular ceramics offers the potential for producing tailored ceramic-based interpenetrating composites with 3–3 interconnectivity.

5.9.2

Metal–Ceramic Interpenetrating Composites

A number of methods are available for fabricating cellular ceramics, and some of these yield structures that exhibit connected, open-celled porosity (see Part 2, and especially Chapters 2.1, 2.6, 2.7, and 2.8 of this book). Such structures can then be infiltrated with metals, polymers, or different ceramics by a number of techniques. Perhaps the majority of work involves introducing molten metal alloys by squeeze or die casting [3–8], the size of the porosity determining the degree of pressure re-

5.9 Interpenetrating Composites 573

quired, though it must also be born in mind that the cellular ceramic preform must have the required strength to survive the process.

In this field, most work has involved the use of aluminum and aluminum alloys, for example, the work of Mattern et al. [6]. In this case, alumina cellular ceramics produced by using starch-based sacrificial pore-forming agents were infiltrated with the aluminum–silicon alloy AlSi12 by squeeze casting. A pressure of 100 MPa was found to be sufficient to infiltrate even the microporous walls of preforms sintered at low temperatures, and thus fully interpenetrating composites were achieved both on the macroand microscale.

In principle, a wide range of metals and their alloys can be used provided the ceramic has sufficient thermal shock resistance to survive contact with the molten metal and does not react with it to form unwanted reaction products that deleteriously affect the properties of the resulting material. For example, Zeschky et al. [7] produced interpenetrating magnesium/ceramic composites by infiltrating polysilox- ane-derived ceramic foam performs (82 % porous, 800 mm average cell diameter) with AZ 31 magnesium alloy by squeeze casting at 680 C. Interfacial bonding was achieved by formation of spinel and cordierite in a reaction layer at the metal/ceramic interface. The dense metal–ceramic composites had significantly higher elastic modulus, yield strength, and creep resistance than the monolithic alloy at both room temperature and 135 C. The results suggested that filler-loaded preceramic polymers have a high potential for optimization of lightweight cast metal components.

Di et al. produced Mg-alloy-based interpenetrating composites using “ecoceramics” as performs [8]. These are porous carbon materials that are obtained by impregnating wood-based waste materials with phenolic resin by using an ultrasonic impregnation system and then carbonized at a high temperature under vacuum. These cellular ceramic performs were then infiltrated with ZK60A Mg alloy in a vacuum high-pressure infiltration furnace to yield dense and homogenous interpenetrating composites with good mechanical properties and excellent damping capability.

As indicated above, 3–3 composites can display a range of useful properties, typically including increased strength, stiffness, hardness, wear and abrasion resistance, lower mass and thermal expansion coefficient, and better resistance to elevated temperatures and creep compared to the matrix metal while retaining adequate electrical and thermal conductivity. However, a disadvantage seems to be that they are inherently brittle as a result of the continuous nature of the ceramic phase. Whilst this limits their potential applications, one for aluminum-alloy-infiltrated ceramic foams is as brake disks for automobiles and trains.

This application in particular has been investigated during a research and development project involving Aachen University, Thyssen Guss AG, and SAB Wabco BSI in Germany [9]. The primary goal was to reduce the mass of the bogie of highspeed railway coaches (Fig. 2a). Since the brake disks and ancillary equipment comprise about 20 % of the total bogie mass, any reduction could have significant consequences. A range of different A356 aluminum-alloy-based metal-matrix composites were compared to the spheroidal-graphite iron disks currently used, and potential

5.9 Interpenetrating Composites 575

5.9.3

Polymer–Ceramic Interpenetrating Composites

Similar approaches can also be applied to the infiltration of cellular ceramics with liquids that can be polymerized or cross-linked to result in interpenetrating poly- mer–ceramic composites. As for the metal–ceramic composites, the distribution of the two phases can be tailored by controlling the degree of porosity, pore size, and pore shape in the precursor ceramic foam. For example, anisotropy can be created by squeezing the ceramic foam while it is still flexible. Production also requires only three steps: 1) casting of the foam to the desired shape, 2) drying and sintering of the preform, and 3) infiltration of the preform with a suitable polymer.

Regarding the third step, evidence suggests that the use of (low) pressure to force the polymer into the ceramic preform results in a higher overall density, and hence superior properties, compared to sucking the polymer into the ceramic by using a vacuum [11]. The latter method can result in entrainment of air in the polymer.

Potential applications for polymer-infiltrated ceramic foams can be found in the building, automotive, and sensor industries [12]. For the first-named, the combination of their light weight, waterproofness, and capability for net-shape manufacture could result in a number of applications if manufacturing costs can be kept low enough. Possibilities include fire-resistant internal walls, cladding panels, and roofing tiles. In the automotive industry, while the inherently brittle nature of the continuous ceramic phase again prohibits structural applications, a potentially major use could be as fire-resistant underhood insulation.

However, the first application for interpenetrating polymer-ceramic composites will probably be in the field of sonar hydrophones. As long ago as 1978 it was shown that these type of composites can display a larger and more tailorable piezoelectric coefficient than dense piezoelectric ceramics, should have better acoustic coupling to water, and their buoyancy can be more easily adjusted than that of high-density lead zirconate titanate (PZT) ceramic. Furthermore, a more compliant material is more resistant towards mechanical shock and has a higher damping, which is desirable in a passive device [13].

Two approaches have been investigated with respect to replacing dense ceramics in hydrostatic applications. The first involved finding the desired properties in a single material, something that was more or less achieved with poly(vinylidene fluoride), PVF. However, whilst the hydrostatic voltage coefficient, compliance, and flexibility are desirably high and the density suitably low, a low piezoelectric strain coefficient means that the material is not of interest as an active device. In addition, while a high voltage sensitivity means it should be good for passive devices, when used as a hydrophone the material must be fixed to a curved surface that can flex in response to pressure changes. The difficulty lies in designing a sealed flexible mount for the polymer that will function when exposed to the high pressures that exist deep in the ocean and still retain sensitivity near the surface [13].

The second approach therefore involved the development of piezoelectric cera- mic–polymer composites in which the properties of both materials are combined. Wenger et al. investigated the potential of such composites with 0–3 and 1–3 connec-

576 Part 5 Applications

tivity [14–17]. While they found them to be quite suitable for passive sensors, and they had the advantage of being flexible, composites with higher level connectivity offered superior responses.

Skinner et al. [13] used the replamine form process to produce 3–3 piezoelectric composites (Fig. 3) and investigated their properties. The process consisted of vacuum impregnating a coral skeleton with wax and then leaching away the skeleton with hydrochloric acid to leave a wax negative of the original coral. This was vacu- um-infiltrated in turn with a PZT slip and then the wax was burned out at 300 C to leave a replica of the original coral structure made of PZT. This was sintered and again infiltrated with a high-purity silicone rubber. The resultant rigid composite was connected with electrode layers and poled in an electric field; a flexible version of the composite was produced by crushing it to break the ceramic struts. While this lowered the permittivity by interrupting the electric flux, because the ceramic pieces were still held in position by the polymer matrix, stresses could still be transmitted. The net result was a composite with a high hydrostatic charge coefficient, while the permittivity was lowered and therefore the piezoelectric voltage coefficient was greatly enhanced. For the unbroken composites, a rigid polymer such as epoxy or polyester could be used, and a low-density, high-coupling resonator could be fabricated.

Fig. 3 Piezoelectric ceramic–polymer composite made by the replamine form process, from Ref. [13] (Skinner, D.P., Newnham, R.E. & Cross, L.E., Mat. Res. Bull. 1978 13 599–607). Reprinted with permission from Elsevier.

Shrout et al. [18] also investigated a PZT–polymer composite with 3–3 connectivity, was produced by mixing a PZT powder with poly(methyl methacrylate) spheres of 50–150 mm diameter in a 30:70 volume ratio. The mixture was pressed to form green bodies that were slowly heated to 400 C to volatilize the spheres. The porous ceramic was then sintered and vacuum-infiltrated with silicone rubber or an epoxy resin. The resulting piezoelectric characteristics were found to be similar to those of the replamine form process described above, but the microstructure was more randomly orientated. The main advantage of this process lay in the fewer processing steps required.

578 Part 5 Applications

to achieve composites that could have a range of different connectivities ranging from 1–3, shown in the figure, to 3–3. While this approach clearly yielded the greatest degree of sophistication it is by no means a simple route and a large number of process steps are required, increasing the cost and reducing the applicability for mass production.

5.9.4

Summary

Interpenetrating composites are now the subject of increasing research around the world, and new and more sophisticated routes to their fabrication are under development. They offer the promise of unusual combinations of properties and the ability to tailor properties to an extent that has not been available before. To this end, the infiltration of cellular ceramics with molten metals or polymers offers the chance for a simple and easy fabrication route to ceramic-based interpenetrating composites. In addition, since many continuum properties change abruptly at the percolation threshold, possibly the greatest potential for making materials capable of exhibiting novel behavior lies with those that have marginally interconnected microstructures. In this respect, the precise control over the cellular structure that is increasingly possible, as the other chapters in this book demonstrate, offers tremendous advantages.

Acknowledgements

The author would like to acknowledge the work of some of his former research students, Mr. Lars Monson and Mr. Suresh Talluri.

References