The class of materials known as ceramic matrix composites, or CMCs, shows considerable promise for providing fracture-toughness values similar to those for metals such as cast iron. Two kinds of damage-tolerant ceramic-ceramic composites are being developed. One incorporates a continuous reinforcing phase, such as a fiber; the other, a discontinuous reinforcement, such as whiskers. The major difference between the two is in their failure behavior. Continuous-fiber-reinforced materials do not fail catastrophically. After matrix failure, the fiber can still support a load. A fibrous failure is similar to that which occurs in wood.

Incorporating whiskers into a ceramic matrix improves resistance to crack growth, making the composite less sensitive to flaws. These materials are commonly described as being flaw tolerant. However, once a crack begins to propagate, failure is catastrophic.

Of particular importance to the technology of toughened ceramics has been the development of high-temperature silicon carbide reinforcements. Although other reinforcement materials are available, such as glass and carbon fiber, metal whiskers, and alumina-based products, this discussion focuses on SiC-based products because they are more applicable to high-temperature use.

SiC fibers, which are capable of withstanding temperatures to about 1,200 °C, are manufactured from a polymer precursor. The polymer is spun into a fine thread, then pyrolized to form a 15- ∝ m ceramic fiber consisting of fine SiC crystallites and an amorphous phase. An advantage of the process is that it uses technology developed for commercial fiber products such as nylon and polyester. Two commercial SiC fiber products are Ube Industries' Tyranno fiber and Nippon Carbon's Nicalon fiber, both from Japan.

SiC filaments: are prepared by chemical vapor deposition. A thick layer of silicon carbide is deposited on a thin fiber substrate of tungsten or carbon. Diameter of the final product is about 140 ∝ m.

Although developed initially to reinforce aluminum and titanium matrices, SiC filaments have since been used as reinforcement in silicon nitride. The material is manufactured by Avco Specialty Materials/Textron in the U.S. and by Sigma Composite Materials in the Federal Republic of Germany.

SiC whiskers consist of a fine (0.5- ∝ m-diameter) single-crystal structure in lengths to 100 ∝ m. The material is strong (to 15.9 GPa) and is stable at temperatures to 1,800°C. Whiskers can be produced by heating SiO2 and carbon sources with a metal catalyst in the proper environments. These reinforcements are manufactured on a commercial scale in Japan by Tateho Chemical Industries and Tokai Carbon Co.

Although these materials are relatively new, at least one successful commercial product is already being marketed. An SiC-whisker-reinforced alumina cutting-tool material is being used to machine nickel-based superalloys. In addition, considerable interest has been generated in reinforcing other matrices such as mullite, silicon carbide, and silicon nitride for possible applications in automotive and aerospace industries.

Interface conditions: In addition to developments in reinforcement materials, advances in controlling the interfacial bond between matrix and reinforcement have led to further mechanical property improvements of ceramic-ceramic composites. The interfacial bond must be optimized to promote favorable toughening mechanisms such as crack deflection and crack bridging. Without proper interface control, a brittle polyphase material results, rather than a toughened composite.

Toughness improvements by interfacial modifications have been made in both fiber and whisker-reinforced systems. Interface control has resulted in the development of toughened fiber-reinforced glass-ceramic matrix materials at the United Technologies Research Center and of toughened fiber-reinforced zirconia-based materials at the Naval Research Laboratories. At present, interfacial control is more advanced for fiber-reinforced composites than for whisker-reinforced materials.

One current approach is to design the interface so it has a parting layer that promotes crack deflection parallel to the fiber length. The parting layer protects the fiber from damage by deflecting cracks, enabling the undamaged reinforcement to support load and bridge cracks during matrix failure. Thus, the composite does not fail catastrophically. Fracture morphology is comparable to that of the fibrous fracture of wood structures. Current materials being used for such interfaces are boron nitride and carbon, materials that have weak crystallographic orientations that preferentially delaminate.

Modifications to the interfacial zone of whisker-reinforced composites are in their developmental infancy because of the difficulty of applying thin coatings on fine whiskers. Studies at Oak Ridge National Laboratories have demonstrated that thermal treatments of whiskers prior to their incorporation into an alumina matrix can increase fracture toughness of the composite. In those materials, best toughness -- about 8.0 MPavm -- results with whisker surfaces modified to be carbon rich and oxygen poor.

Current mechanistic studies at the University of California/Santa Barbara are directed toward understanding the role of interfacial structure on toughness of ceramic-matrix composites. In addition, some investigators feel that the approaches used by the carbon-carbon community, such as applying various CVD coatings to seal off the fibers, may result in near-term solutions for improving toughness of fiber-reinforced ceramics.

In whisker-reinforced materials, the matrix usually seals off the interface region from the composite exterior. This protects the interface from oxidizing environments. However, once cracks are initiated, they allow access of atmospheric elements into the interior. As with fiber-reinforced materials, new interface compositions must be developed that are stable in oxidizing environments.

In addition, there is still a need to develop further understanding of the role of whisker interfaces on toughening mechanisms for ceramics. The requirements of fiber and whisker-reinforced systems appear to have many similarities.

Reinforcement needs: Although the current interest in ceramic-matrix composites has resulted from improved reinforcements, there is still a need for further developments. Specifically, reinforcements are needed for ceramic matrices for service at temperatures greater than 1,800°C.

Currently available polymer-derived fibers are limited because they deteriorate above 1,200°C. A program aimed at developing higher temperature fiber has been sponsored by the Dept. of Defense, combining the expertise of Dow Corning in silicon-based materials with that of Celanese in fiber technology. From this program has come a new fiber material that has higher thermal stability than commercially available fibers.

SiC filament material has limitations in oxidizing environments due to its carbon core and carbon surface coatings that oxidize above 600°C. These filaments are designed for use in aluminum and titanium matrices. A similar product, engineered for ceramic matrices, is needed.

SiC whiskers are a nearer-term reinforcement for commercial ceramic-matrix composites, having already demonstrated success in reinforcing alumina. As with the other reinforcing materials, the whiskers currently being produced are more appropriate for reinforcing metals. Current theory indicates that thicker whiskers (1 to 3 ∝ m) are more appropriate for ceramics. Such materials are now under development.

Dimox process: Ceramic matrix composites are steadily moving from the laboratory to initial commercial applications. For example, engineers are currently evaluating these materials for use in the hot gas zones of gas turbine engines, because ceramics are known for their strength and favorable creep behavior at high temperatures. Advanced ceramics, for example, can potentially be used at temperatures 400 to 900°F above the maximum operating temperature for superalloys.

Until recently, however, there has been more evaluation than implementation of advanced ceramics for various reasons. Monolithic or single-component ceramics, for example, lack the required damage tolerance and toughness. Engine designers are put off by ceramic material's potential for catastrophic, brittle failures. While many CMCs have greater toughness, they are also difficult to process by traditional methods, and may not have the needed long-term high-temperature resistance.

A relatively new method for producing CMCs developed by Lanxide Corp., Newark, Del., promises to overcome the limitations of other ceramic technologies. Called the Dimox directed metal oxidation process, it is based on the reaction of a molten metal with an oxidant, usually a gas, to form the ceramic matrix. Unlike the sintering process, in which ceramic powders and fillers are consolidated under heat, directed metal oxidation grows the ceramic matrix material around the reinforcements.

Examples of ceramic matrices that can be produced by the Dimox directed metal oxidation process include Al2O3 , Al2Ti)5, AlN, TiN, ZrN, TiC, and ZrC. Filler materials can be anything chemically compatible with the ceramic, parent metal, and growth atmosphere.

The first step in the process involves making a shaped preform of the filler material. Preforms consisting of particles are fabricated with traditional ceramic forming techniques, while fiber preforms are made by weaving, braiding, or laying up woven cloth. Next, the preform is put in contact with the parent metal alloy. A gas-permeable growth barrier is applied to the surfaces of this assembly to limit its shape and size.

The assembly, supported in a suitable refractory container, is then heated in a furnace. For aluminum systems, temperatures typically range from 1,650 to 2,100°F. The parent metal reacts with the surrounding gas atmosphere to grow the ceramic reaction product through and around the filler to form a CMC.

Capillary action within the growing ceramic matrix continues to supply molten alloy to the growth front. There, the reaction continues until the growing matrix reaches the barrier. At this point, growth stops, and the part is cooled to ambient temperature. To recover the part, the growth barrier and any residual parent metal are removed. However, some of the parent metal (5 to 15% by volume) remains within the final composite in micron-sized interconnected channels.

Traditional ceramic processes use sintering or hot pressing to make a solid CMC out of ceramic powders and filler. Part size and shapes are limited by equipment size and the shrinkage that occurs during densification of the powders can make sintering unfeasible. Larger parts pose the biggest shrinkage problem. Advantages of the directed metal oxidation process include no shrinkage since matrix formation occurs by a growth process. As a result, tolerance control and large part fabrication can be easier with directed metal oxidation.

In addition, the growth process forms a matrix whose grain boundaries are free of impurities or sintering aids. Traditional methods often incorporate these additives, which reduce high-temperature properties. And cost comparisons show the newer process is a promising replacement for traditional methods.