Ceramic Fibers and Coatings: Advanced Materials for the Twenty-First Century (1998)

The performance, manufacturing, and life cycle costs for systems containing CMC components, including ceramic fibers and coatings, are interconnected. The prices of ceramic fibers, fiber coatings, and composites, however, do not always reflect manufacturing costs. These issues have not been widely explored and are even less widely understood. Consequently, the influence of fiber and coating costs on the manufacture and implementation of CMCs remains ambiguous and system dependent. This chapter discusses the relationship between costs and prices, aspects of fiber costs, and the implications of CMC market volume on fiber prices.

The words price and cost are often used interchangeably. There are, however, important distinctions between price and cost in the commercialization of a new technology.

Price can be defined as what the market will bear for a product or service. Price levels for competitive goods and services under normal circumstances are set by the marketplace. For new materials, price levels are determined by assessing the price/performance ratio of existing materials. In general, the price and performance of CMCs will be compared to superalloys and other turbine engine materials, as well as to monolithic ceramics for the components of small industrial turbine generators.

Costs are determined by actual expenditures for producing a good. Costs are governed by physical realities, for example, the cost of buying machines and the cost of labor to operate them. Costs can fluctuate widely depending on how much is being produced per unit of time or how efficiently the manufacturer uses resources. In general, there are two types of cost: costs associated with production (i.e., manufacturing costs) and costs associated with operating a continuing business (i.e., administrative and marketing costs). The most important factor for commercializing a new technology is potential manufacturing costs. If the product can be manufactured efficiently, it is assumed that a suitable “business structure” can be put in place for delivering the product to the market.

As products flow along the supply chain, the prices for one entity become the costs for the next. Thus, for the commercialization of new technology, the word cost has different meanings for producers and users. For users, the price and cost of a product or service are the same. For producers, however, cost is the expense incurred to produce a product or supply a service. To earn a profit and a return on investment, the producer's aggregate costs over time must be lower than the aggregate revenue. Time becomes the key element in the equation. In the short run, costs are often higher than prices; but there must be an expectation that, in the long run, prices will exceed costs.

For producers, profit is the difference between cost and price. Making a profit is the goal of every commercial enterprise, and the definition of successful “commercialization” of a new technology should be when a new technology is either earning a profit or is on a trajectory to do so. Even if a new technology is available for consumption, it will eventually be removed from the market if it does not show promise to earn a sufficient profit.

Some costs are incurred all at once, but their effects are felt over a period of time. Other costs increase or decrease depending on the level of business. Because of the difference in how costs occur, categories have been developed (see, for example, Figure 7-1) to aid in the analysis of financial alternatives:

• Variable costs are costs that vary, in the aggregate, with production volume. Raw materials are the best example of a variable cost. As production increases, the aggregate cost of raw materials increases.

FIGURE 7-1 Elements in manufacturing costs. Source: IBIS Associates,Inc.

• Fixed costs are costs that, in the aggregate, are constant (fixed) regardless of the level of production. The best example is machinery, which is put in place with a one-time expense. No matter how much of the capacity of the machine is used, the aggregate expense does not change.

However, these distinctions can almost be reversed when considering the cost of a single unit. Variable costs are constant for each unit of production regardless of volume. For example, if the output of a brick plant doubles, the amount of clay consumed per brick is the same. The only time this principle doesn't hold is if something else changes, if the amount of scrap is somehow reduced, for example, along with the increase in volume. Fixed costs, however, account for less cost per unit of production as volume is increased to capacity. For commercializing a new technology in the manufacturing environment, the two types of cost have important implications.

The aggregate manufacturing costs that change with production volume fall into three basic categories: materials, direct labor, and utilities. Materials include the input materials that end up in the final product, as well as scrap losses and other consumable items affected by production volume. For example, mold release coatings, although they are not part of the final product, are still required to make the part and are accounted for as a materials cost. Direct labor is the cost of the manpower to operate the machinery in the factory; as production increases, more labor is required. It is assumed that if a machine is not running, the machine operator can “do something else” and that cost (associated with running that machine) is not incurred; in practice this is obviously not always the case. Also, workers can “learn” some processes to improve the production rate, for example, composite lay-up. In this case, the labor cost decreases on a per unit basis, although it still increases in the aggregate. Utilities are the flat-rate consumables needed to operate a factory, such as electric power and water provided by a utility. Once again, the aggregate costs increase with production volume, but the per unit costs are constant.

For analytical purposes, it is best to treat the per-unit variable costs as constant throughout the range of production volumes. Factors that will change the per-unit variable costs are theoretically independent of production volumes, although in practice they become easier to affect as volumes increase. For example, reductions in scrap losses lower the cost of material per unit, as well as in the aggregate. It is easier to identify the root causes of scrap losses when capacity utilization is high, but in theory there is no correlation between scrap losses per unit and production volume. Although the “learning curve” effect reduces direct labor content per unit, manufacturing philosophies dramatically affect the extent of this phenomenon.

Manufacturing costs that do not change, in the aggregate, with production volume include: the cost of equipment; the cost of tooling; the cost of the building; the time value of the money invested in equipment, tooling, and the building; and the labor and maintenance expenses needed to operate the plant and equipment.

The first four items are relatively easy to understand. Let's say General Motors decides to build a new electric vehicle. It builds a new plant with special machinery and orders tooling that can only be used to make that specific car. It builds a new building for the assembly of this unique vehicle and invests money borrowed from a bank or shareholders, on which it will owe interest or dividends, to put everything in place.

If the new technology (i.e., the electric vehicle) is accepted by the public, the factory will increase production, and the cost of the investment in the fixed plant will be spread over a great many cars. If the electric vehicle is a flop in the market, the investment costs will be spread over many fewer cars, and per unit costs will be higher. The number of cars produced doesn't affect the aggregate spending on facilities; but it does affect the amount of spending per car.

It's easy to see that fixed facilities require up-front cash expenditures and that only future events will determine if the investments were sound. Thus, fixed costs are the most significant for commercialization; companies must take risks that fixed costs, investments, will be paid back at a later date.

Some costs are generally regarded as fixed costs but are in somewhat of a “gray area.” Examples are factory “overhead,” labor, and maintenance costs. A scheduler is necessary regardless of the plant's utilization, whereas an extra press operator is not. Thus the scheduler is dubbed a fixed cost, and the press operator a variable cost. This designation is made even though the press operator may be given something else to do, and the cost still incurred, if production is not up to capacity. Similarly, maintenance costs can be based on time or on production (akin to changing your car's oil every 3,000 miles or every three months, whichever comes first). Accountants generally put maintenance in the “fixed” category, although it often functions more like a variable cost.

Time is the governing issue in profitability; how much can be produced and sold in a given time for a given investment. Let's say engineers figure out a way to halve the time it takes to produce a unit on a given machine. The output of that

machine per unit time doubles; the fixed cost per unit is halved. The full capacity of the machine, however, must be used to achieve the cost savings. For example, if throughput is doubled, but sales remain constant, the only effect would be to idle the machine half the time. Only if the number of workers to operate the machine and the amount of power consumed by the machine remain constant, do the elements of variable cost get halved (as a result of halving production time per unit). Labor and power costs are reduced regardless of production volume (after all, these are variable costs).

Cycle time is, therefore, one of the key cost predictors of a new technology. As an example, car companies have looked towards polymer-based composite structures to reduce the weight of vehicles. Steel panels, however, can be formed in just a few seconds. Although much larger, vastly more expensive stamping machines are required for steel panel production, high sales volumes easily allow the cost of these stamping presses to be minimized on a per unit basis. Thus steel has remained the low-cost way to build automobiles.

The manufacture of fibers presents three unusual cost issues: fibers have a very high length per unit mass; the machinery used for fiber production is highly specialized; and materials to be formed into fibers require special treatments or additives.

The cost of most materials is scaled by cost per unit of mass. Fibers have a tremendous amount of length per unit mass, and processing that length incurs additional costs relative to mass (see, for example, Figure 7-2). The additional cost can be seen as a cycle time penalty. For example, it takes less time to produce a kilogram of bulk polymer than a kilogram of the same polymer spun into a fiber. The cost premium for engineering materials that have been “stretched”

to fiber diameters is significant, especially in the linear density range of interest for composite materials.

The machinery used for fiber production is highly specialized; there is little in the way of off-the-shelf fiber spinning, drawing, or heat treating machinery. The low volume, customized nature of this machinery extracts a considerable investment premium. For example, plastic injection molding machines are available on a turn-key basis in a range of sizes from a large number of suppliers. Plastic fiber spinning machines, however, are only available on a custom basis from a limited number of suppliers.

Finally, there is an added premium to materials prices that are formed into fibers. This is because fiber drawing materials (e.g., spin dope) must be exceptionally pure and require special treatments or additives. For example, the cost of carbon pitch is measured in dollars per ton. The purification required to spin carbon pitch into carbon fiber, however, raises the raw materials costs to dollars per pound, an increase of three orders of magnitude.

Taken together, these three factors make fiber production for composites a dubious business proposition. The time it takes to learn how to use the fiber and qualify the material for specific applications has resulted in slow market development, during which time the manufacturer must hold a tremendous amount of underutilized invested capital. The fiber manufacturer is thus in a precarious situation. If the fiber manufacturer produces fiber at full capacity to reduce manufacturing costs per unit, he will produce more fiber than can be sold and will lose money through excessive variable costs. If the fiber manufacturer produces only the amount of fiber that can be sold, he will incur high underutilized fixed costs, resulting again in his losing money.

Take, for example, the two major commercialized advanced composite fibers, Kevlar and carbon fiber. When Kevlar was first developed, DuPont believed it would be used as a replacement for fiberglass in ropes, tire cords, and many

FIGURE 7-2 Fiber price premiums. Source: IBIS Associates, Inc.

other applications. Therefore, DuPont spent heavily on research and development and production capacity, but a large market did not materialize rapidly enough for DuPont to make a return on its investment. If DuPont had simply put all of the money it invested in Kevlar “in the bank,” it would be earning more in interest, from doing nothing, than it earns in profits from Kevlar production.

A similar situation occurred in the carbon fiber industry. Carbon fiber was expected to be the material of choice for aircraft structures, but the drop-off in the production of military aircraft and the hurdles of qualifying for civil aircraft production resulted in inadequate market development and large losses for the early developers. Over time, however, the understanding of carbon fiber has matured and manufacturers have improved processing, thus reducing production costs. As a result, carbon fiber is now a large, growing, profitable business. It currently has important uses in commercial aerospace, automotive, and electronic markets and may even become a significant material in civil engineering. But its “commercialization” took 25 years, well beyond the purview of any single commercial enterprise.

Ceramic fiber producers are in the same situation that DuPont was in with Kevlar and that several manufacturers were in with carbon fibers. The fixed costs of producing ceramic fibers are high. The only way to reduce the fixed cost per pound of fiber is to produce more fiber than can be sold because the market is not significant, regardless of price. Offering a long-run price to the market would certainly help the market grow, but market development still takes time, and manufacturers would incur losses until the market becomes substantial.

There are three basic steps in the manufacture of high performance, continuous ceramic fibers, all of which have important cost considerations: (1) preparation of bulk, preceramic material to be spun; (2) spinning the bulk material into a green fiber; and (3) heat treating the spun green fiber to convert it into ceramic fiber.

The preparation of preceramic material involves processing a high purity, easily spinnable material, usually a preceramic polymer or sol. In either case, specialized equipment is required to produce the material because there is no other market for these exact materials (although there are some close relatives). The high level of purity, the delicate balance of material properties, and the dedicated, customized equipment make preparation an expensive process.

Spinning involves pumping the material through very fine, precise holes in a plate (called a spinneret) into a controlled environment where the material is solidified. This environment can be gaseous (dry spinning) or liquid (wet spinning). As the material is forced through the spinneret, it takes fibrous form. The diameter of the fiber is usually reduced by taking up the bundle of fiber on a spool at a faster rate than it is being produced at the spinneret.

At this point the fiber is called green fiber and is usually extremely delicate and difficult to handle, resulting in reduced yield. Controlling the spinning environment is challenging. If the fibers are too dry, they will break apart. If they are too wet, they will stick together and form a useless cord when they are squeezed into a bundle on the spool. The lower yield that results from these processing complexities results in increased costs.

Heat treating involves ramping the temperature of the green fiber to the temperature of its conversion to ceramic material, holding it at that temperature until the fibers are fully converted, and ramping the fiber temperature back down. The environment must be carefully controlled throughout this process. The time and temperature required to convert the fiber, the controlled environment, and the custom furnaces all make this an expensive process.

The combination of high fixed costs for all three steps and the long cycle times of the last two steps result in extraordinarily high costs per unit mass for low volumes (often thousands of dollars per pound). The most effective way to lower the costs of ceramic fibers is to increase the volume of production. Other efforts to reduce fiber cost will have far less impact than increases in production volume.

Oxides appear to have a distinct manufacturing cost advantage over non-oxides. Oxides generally require shorter heat treatments at lower temperatures in less difficult environments (e.g., the presence of oxygen is not a complicating factor). This translates into significant savings in fixed costs. Lower temperature furnaces, for example, are less expensive than higher temperature furnaces. Furthermore, eliminating oxygen from the cure environment for non-oxides is expensive and adds costs that are not incurred for oxide fibers. Shorter heat treatment times for oxides result in faster cycle times. There is also a strong possibility that the cost of raw materials for producing oxide fibers can be reduced to well below the cost of raw materials for non-oxide fibers.

Fiber prices could be reduced by a market-driven increase in production volume, which would lower the fixed cost per unit. Although current fiber manufacturers and independent consultants have conducted studies to predict fiber prices and cost levels, exact predictions are elusive because of the difficulty

in precisely forecasting future yields and process improvements. Furthermore, prices will depend on the pricing policies of individual manufacturers as much as on actual costs.

Assuming that the performance of ceramic fibers is suitable for thermostructural applications, implementing these materials still poses great business risks. CMC manufacturers have said that fibers that meet their performance requirements must be priced at approximately $1,100/kg ($500/lb) for large volumes to be in demand. Ceramic fiber developers/manufacturers, however, have indicated that ~ 23,000 kg (50,000 lbs) of fiber must be sold annually to attain a price of $1,100/kg ($500/lb) (see Figure 7-3). The math is simple; 23,000 kg (50,000 lbs) of fiber at $1,100/kg ($500/lb) results in a total of ~ $25 million in revenue for the fibers alone. Even if an assumption is made that the cost of the fibers constitutes 25 to 50 percent of the cost of the final composite, $50 million to $100 million in composite sales will be necessary to support $25 million in fiber sales.

The problem is that no single current application of continuous fiber composite material within one company totals $100 million. This includes applications that utilize carbon fibers, Kevlar, or even fiberglass. In fact, a large application for a composites manufacturer is $10 million. The largest single applications of continuous fiber composites (by dollar value), within one company are carbon fiber golf clubs, pultruded fiberglass ladder rails, and certain aerospace applications (e.g., Boeing's 777 empennage structure). Even by the standard of the jet engine component industry, $100 million represents a tremendous volume of sales. Current suppliers of metal jet engine components, such as Wyman Gordon and Howmet, measure sales in hundreds of millions of dollars, but these sales are spread over several individual components for a large number of engines.

Therefore, it is extremely unlikely that the commercialization of CMCs and their constituent fibers can be supported by one application. Several applications will have to be developed to reach $50 million to $100 million in CMC sales.

FIGURE 7-3 Projected costs of fibers as a function of annual production. Source: IBIS Associates Inc.

Materials suppliers typically solve this problem by offering new materials to the market at a long-run price below cost (on the assumption that several applications will be developed) and incur losses until a market develops.

If a ceramic fiber manufacturer were to take this approach, he would face a daunting task. For example, if Dow Corning were to offer Hi-Nicalon fiber at a price of $500/kg ($227/lb) and could develop a CMC market of $20 million the first year (a tremendous accomplishment by the standards of the composite materials industry), this would equate to fiber sales of roughly 10,000 kg (22,000 lbs) in the first year. Dow Corning, however, projects that the cost of manufacturing Hi-Nicalon fiber, at 10,000 kilograms (22,000 lbs) of annual production, would be in the range of $1,250/kg ($568/lb) (a dramatic drop from its current cost of more than $3,000/kg [$1,363/lb]), as can be seen in Figure 7-3. Thus the payoff for developing the market is a loss of $750/kg ($340/lb) of fiber and a total loss of $7.5 million to Dow Corning in the first year alone.

Given the typical rate of market development for composite materials, fiber manufacturers may face many years of these large losses. Clearly, one of two situations must develop for the commercialization of successful ceramic fibers: (1) high-value applications must be developed that support fiber prices of much more than $500/kg ($227/lb); or (2) very large volume applications (by composites standards) must be developed that allow fiber manufacturers to produce fibers efficiently and offer them at a low price (e.g., $500 per kilogram [$227/lb]). Given the historical rate of market development for composites and the niche applications targeted for CMCs, the first situation is more likely to develop.

Figure 7-2 shows that the Nextel ceramic oxide fibers are currently available within the price range of commercially available carbon fibers. That is, they cost more than low modulus carbon fibers but less than high modulus carbon fibers. There are even specialty grades of carbon fiber that sell for much more than the carbon fiber prices shown in Figure 7-2. Thus, by current standards, the Nextel family of ceramic fibers cannot truly be labeled expensive.

The SiC fibers shown in Figure 7-2 (the upper line shows the projected price level at 1,000 kilograms [2,200 lbs] per year; the lower line shows 100,000 kilograms [220,000 lbs] per year) do indeed appear to be “expensive,” well above the price line for carbon fibers. It may not be appropriate, however, to compare the cost of SiC fibers, which are produced at a rate of 100,000 kilograms (220,000 lbs) per year to the cost of carbon fibers, which are typically produced in millions of kilograms per year per manufacturer. Note that when carbon fibers were first used in aerospace, the price was several hundred dollars per kilogram in late 1960s dollars, which corresponds to much more than $1,000/kg ($455/lb) in

1997 dollars. If aerospace grade carbon fibers were made in a factory today with only tens of thousands of kilograms of capacity, the cost would rise to hundreds of dollars per kilogram.

Finally, the metal parts in aerospace applications (e.g., turbine engines) that CMCs would replace are already expensive. Given the projected prices for SiC fibers of $1,250/kg ($568/lb) (at sales of 10,000 kilograms [22,000 lbs] per year) and the weight savings and improvements in performance offered by CMCs, one must ask if cost is really the limiting factor? The answer is probably not. It seems more likely that the limiting factor is lack of confidence by potential CMC end users that the application will be successful and reluctance to incur the initial costs of implementation.

Given the difficulties of fiber manufacturing (high fixed capital cost and slow cycle times) and the low sales volume, the only way to lower prices (at low sales volumes) is to produce the fiber with much less capital equipment. Even if cycle time is dramatically improved, the only way to capitalize would be to set up a factory that could produce small volumes of fiber quickly with a small amount of equipment. After 30 years of research and development on ceramic fibers, however, there is no indication that such a development is forthcoming.

Over the past 30 years, however, the performance of ceramic fibers has continued to improve. Today's low modulus carbon fiber has a modulus that was attainable only by the highest grades of early carbon fiber. The latest version of 3M Nextel oxide ceramic fibers show thermal stability far above the Nextel grades that were first commercialized. The Hi-Nicalon SiC fibers manufactured by Nippon Carbon Company are superior to the original Nicalon fibers. It is possible, therefore, that ceramic fibers will eventually reach a level of performance that will justify their price to design engineers of thermostructural components.

The fiber is only one of three constituents of a CMC system, which also includes an interface and a matrix. Matrix processing is clearly beyond the scope of this report, but fiber-matrix interfaces (or interphases) are not. The manufacturing cost of potential interphase materials and the processes used to apply them to fibers and preforms is not well defined or understood. Interphase coatings are not mature at this point, and the cost impact of the interphase application process should be an important consideration in future efforts to commercialize CMCs. Therefore, the committee concludes that efforts should be made to determine the impact of the cost of fiber interphases on the cost of the total CMC system.

Despite the high cost of producing fibers and composites, some applications appear to offer favorable cost/benefit ratios for incorporating CMCs. The reason these applications have not been commercialized is most likely the cost and risk of implementation. Implementation costs include all qualification costs, such as generating an engineering database, design and development costs, and the cost of gaining regulatory approval. The probability of unforeseen problems that would require additional time and money to address add a considerable risk to those attempting to qualify CMC applications. The combination of known qualification costs and a risky payback have made the incorporation of CMCs, even when applications appear to be favorable, an unacceptable business proposition.

High manufacturing costs are barriers to the commercialization of ceramic fibers for advanced composites. Barring a tremendous technical breakthrough, manufacturers will likely view entering the market as a highly risky investment that is more likely to lose money than make it. The up-front costs and competition versus the potential future payback presents a stark contrast. That is, a material supplier would have to invest significant capital up-front for development and capital equipment and would face certain competition from currently available, relatively inexpensive, well understood materials. In addition, the product may not work, the market may take a long time to develop, and favorable pricing and volumes may never materialize. Early producers of advanced fibers for composite materials, such as Kevlar and carbon, lost money despite the high capabilities and promise of those materials. These experiences are ominous reminders of the precarious nature of the advanced fibers business.

Despite the tremendous financial risks and, in some cases, financial losses incurred by fiber manufacturers, society has benefited significantly from the development of advanced fibers. For example, Kevlar fibers have saved lives, in bullet-proof vests for policemen, in helmets for soldiers, in rescue hoists on helicopters, but almost none of that value has been translated into profits for DuPont. Carbon fibers are making light, efficient satellites possible that will allow worldwide portable telecommunications. Despite the tremendous market value of this innovation, however, Union Carbide, the original pioneer of carbon fibers in the United States, has long since gone out of the carbon fiber business after writing off the financial losses incurred during development. The lack of short-term, or even longer-term, return on investment from fiber development makes it difficult for a commercial enterprise to justify developing advanced fibers independently.

Developers of ceramic fibers for CMCs are in a similar situation. The use of CMCs in turbine engines can both improve efficiency and lower polluting emissions. Although real value can be put on these advantages, in the form of lower electricity costs and cleaner air, it is not likely that enough of this value would be realized in the form of profits to motivate a company to single-handedly invest in making ceramic fibers a viable commercial business. Without cooperation between the commercial and governmental entities that stand to benefit from the development of advanced ceramic fibers and composites, the technology will almost certainly remain in the laboratory, and the potential benefits will not be realized.

Although advanced fibers and composites offer tremendous potential value to the marketplace, historically fiber manufacturers have not realized enough profits to justify their investments. This has resulted in difficulty in maintaining a reliable fiber supply. In the absence of outside support, fiber manufacturers will continue to lose money until a viable market for CMCs develops. Because they cannot control how the market develops, however, they are in a more tenuous business position than other CMC stakeholders.

The successful commercialization of ceramic fibers will require a price level at which initial commercial sales are feasible. In other words, ceramic fibers, and ultimately CMCs, must be judged by end users to be economically and functionally viable to justify the cost of qualifying and implementing them. The price of CMCs (and therefore fibers) must be at a level that will supply the market initially (perhaps at a loss to the CMC and fiber producers) with hopes that the market will grow to the point where a profitable return on investments can be made. This condition has not occurred at the price/performance levels of currently available ceramic fibers. The committee has identified three mechanisms for improving the market potential of ceramic fibers and CMCs: (1) improving fiber performance; (2) reducing the risk of commercializing ceramic fibers; and (3) expanding the potential market.

The committee has identified two models for the business development of advanced materials that might be applicable to improving the performance of ceramic fibers. The first model is creating a new material with exceptional performance properties and marketing this material to high value applications that can afford to pay the high initial price. As the market grows, prices can be lowered as a function of reduced fixed costs per unit and the development of lower value, higher volume markets. The second model is improving the performance level of an existing commercial base of technology. In this model, fixed cost investments have already been made, which allows for lower fixed costs at initial volumes.

The first model is based on the carbon fiber experience. Once the high-performance aerospace business had been established, other markets, such as racing cars, sporting goods, and industrial products, were developed. All of these applications have grown in volume as prices have come down. The second model is based on the Kevlar fiber experience, in which DuPont was skilled in polymer fiber technology prior to the development of Kevlar.

Both models have established track records in the aerospace industry, which is a well known proving ground for truly new materials. Carbon fiber, glass fiber, titanium, and even aluminum owe their initial commercialization to the aerospace industry. All of them were supported, however, by military demand that could pay high initial prices. The losses faced by the materials supplier, however, are inevitable, and only if they appear to be short-lived and not excessive will the manufacturer take the risk of supplying the market. These two models suggest several recommendations for future investment.

Given the current state of the art, neither model is producing ceramic fiber suitable for applications that would justify paying the short-run price and the high cost of implementation. The goal is to improve performance, either in a unique environment or in an environment that can leverage existing assets. Therefore, the committee recommends research that focuses on improving the properties of polycrystalline ceramic fibers, recognizing that leveraging existing assets will help reduce fixed costs.

The costs and risks of incorporating CMCs into existing applications can be addressed by obtaining data from tests of CMC components to alleviate concerns about encountering “unknowns” during materials qualification. The focus should be on bench testing and insertion programs because data from coupon tests of composites are already widely available. Thus far, the results of bench testing and field testing, such as the tests of the Dornier material in Germany (see Chapter 2), have been promising. Therefore, the committee recommends that low risk government sponsored insertion programs for CMCs be expanded to demonstrate the field performance of CMC components.

Another way to advance the commercialization of ceramic fibers is for the initial high costs of fiber production to be subsidized, thus reducing the risk incurred by fiber developers and manufacturers. In effect, this is what occurred as a result of the early military procurement of carbon fibers. The current low build rate of new military aerospace products and a procurement environment that favors lower purchase costs over higher performance do not bode well for this approach. The outright subsidization of production, without a military demand,

has generally not been a policy of the United States although other countries, notably Japan, have employed this policy.

CMCs are targeted largely toward applications in turbine engines. In terms of material consumption, this is a relatively small market. Because of the large production volumes needed to support fiber manufacture, every potential outlet for the use of ceramic fibers should be investigated. Multiple applications must be developed because individual applications can constitute only a small part of the production necessary to support the development and manufacture of advanced ceramic fibers.