negotiations between the students and the corporations are handled through the professor.

Fibers have been developed to meet the current and anticipated needs of the consumer. At this time the emphasis is to further identify the most desirable precursors by chemical analysis. and other characterization methods that are used to assist in the definition of the optimum carbonization cycles. The Toho fibers have a variety of properties, e.g., strengths from 15,000 to 600,000 psi and modulus approaching 90 million psi. The strains available are as high as 2%. No further improvement is going to be made on this property until the customers indicate there is a need to do so. Interestingly, a curve was shown of strength plotted against modulus values and families of isostrain curves as the third variable. It appears that the maximum strain of 2% occurs at a modulus of 45 million psi with strength in excess of 600,000 psi. For strains in excess of or less than 2%, the strength and strain values are lower, independent of the modulus values.

There are no production problems concerning the manufacturing procedures related to rayon-based fibers as there are strict environmental standards and control procedures in effect.

Architecture and weaving requirements are dictated by the utilization need. At this time resin matrices are the controlling matrix dictating the type of architecture, especially as it applies to the aircraft industry. The analysis used for selecting the architecture is usually done by the customer or some other organization outside of Toho, although a general design code has been developed at Toho expense that is licensed outside the company for the use of its customers. There are a number of composites made from low cost laminates of two-dimensional (2D) material with tensile strengths from 15,000 to 30,000 psi, and there are high strength laminates made of long fibers that contain yarns that have 0.2-mm

spacings. There are tubes made with wall thicknesses of 1.5 mm and outside diameters (ODS) of about 1.5 inches that are similar to the Naval Surface Warfare Center (NSWC) tubes to be used for space structures. Other laminates are made of chopped fibers, which are matted and the surfaces of the fibers are activated for use as absorbent material. The densities of these various composites can range from 1.3 to over 1.7 g/cc depending on the application. The most complex shape of C/C that was seen was a turbine wheel with thin blades that were integrally woven into the rotor. Its density was in excess of 1.7 g/cc and it had a fine weave spacing of about 2 mm. All the C/C development work began at Toho about 10 years ago. work began at Toho about 10 years ago.

One concern in the processing of these different composites is to increase the utilization of the fiber properties that are in the composites which, at this time, is only approximately 50%. A high strength 2D composite can have high strength 2D composite can have tensile strengths as high as 100 kpsi at a modulus of 26 Mpsi, a density of 1.63 g/cc, and a fiber volume of 57%.

Densification methods by both liquid and chemical vapor deposition (CVD) are being evaluated at this time to find the most economical approach that takes into account the specific application that is under consideration. At this time, there is a concerted effort to determine if coal tar or petroleumbased pitches are the best for densification of the types of preforms that are being sold. One technical difficulty that Toho has found with the CVD method of densification is the inability to obtain a uniform density of deposition even with cross-section thicknesses of only a few millimeters. This, of course, is a classic difficulty that is found with all those that use the CVD method.

Interfacial bonds between the matrix and the filament are considered to be very important for composites. However, the major emphasis at this time is with resin matrix composites. Studies about interfaces of C/Cs will occur at a

future time. Toho is oxidizing its fibers at 300 °C as a means of increasing the bond strength between the fiber and the matrix.

Methods for enhancing the oxidation resistance of C/Cs were begun 2 to 3 years ago. At this time the protective layers are composed of SiC and matrix inhibitors of silicon carbide or boron carbide. Toho is also looking at how to protect the fibers either by increasing the perfection of the fiber's microstructure and thereby reducing the number of sites for oxidation or depositing inhibitors on the surface of the fibers. The degree of protection is close to zero weight loss at 1,200 °C for a duration of 200 minutes. However, no thermal cycling was performed prior to this oxidation resistance test. There are two staff people working on this area of technology.

Microstructural evaluation of samples has been limited to optical examination, both with polarized and bright field illumination. In addition, the SEM has been used to examine the topography of C/Cs and their fractured surfaces. But no etching of samples has been performed for examining the various types of matrices and their interactions with the fibers that they surround.

The laboratory contains complete chemical and physical property equipment for evaluation of raw materials as well as finished products. The equipment includes SEMS, TEMS, Instrons that test samples from -60 to 300°C, IR spectrometers, thermal conductivity and expansion testing devices, high precision electrical property measuring equipment, electron spectroscopy for chemical analysis (ESCA), liquid and gas chromatographs, and other equipment for chemical analysis. At this time these laboratories are equipped with an emphasis towards organic matrix composites because this is the area that contains the major portion of the company's product line. These laboratory facilities occupy two floors of an

laboratory is capable of producing 50 tons of prototype fibers in a year. Nippon Oil produces 50 tons/yr of high modulus fibers that are derived from pitch and 500 tons/yr of low modulus fibers. According to the company, these are the largest quantities of pitch fibers produced in Japan. But these are relatively small amounts compared to the quantity of PAN-based fibers that are produced. It is estimated that more pitch-based fibers will not be used until the cost of their production decreases another 25% to 50%.

All sorts of weaving capabilities exist in this organization, including threedimensional (3D) preforms that are 6 inches long by 4 by 4 inches in cross section. There is a unique ability for forming tapes of fibers that are used to produce aluminum and ceramic matrix composites. The fabrication process is performed by first spreading the fibers within yarns and then plasma spraying coatings on to them. Next, the coated yarns are stacked and hot pressed into final composite shapes.

Densification studies are underway to determine the best process for densifying preforms either by liquid pitch or CVD. In the latter case, there is concern that CVD may take too long (up to a year) to obtain the proper microstructure in a C/C preform. The predictability of the processing time is also difficult to estimate to know if the process is economical. The alternative process for densifying preforms is to infiltrate the preforms with liquid pitch. There is autoclave equipment available for high pressure impregnation [hot isostatic pressing (HIP)] of preforms to over 14 kpsi. This enhances the ability to increase the billet's density with fewer impregnation steps by increasing the carbon yield by 10% to 15% for each impregnation cycle, thereby reducing the cost of densification and fabrication. In a smaller autoclave it was found that the pore sizes can be decreased if the preform is subjected to about 28 kpsi gas pressures.

This smaller autoclave can reach temperatures of 1000 °C, which also results in a smaller amount of preform shrinkage.

There is not much research underway on evaluating the microstructure of the matrix, although it is recognized that this could be an important factor in controlling the properties of C/Cs.

finished products as well as for determining their electrical and mechanical properties. Some of the outstanding equipment included: electron probe, electron spin resonance (ESR) with a special probe that can operate at 600 °C, in-situ or real time fractography as viewed with the SEM, 130-keV TEM with a resolution capability of 2 Å, Fourier transform infrared (FTIR) spectroscopy, nuclear magnetic resonance (NMR) with a superconductivity magnet that produces a magnetic field of 9.3 T or over 80,000 G/cm2, other NMRS that can be used for measuring solids up to temperatures of 450 °C, high precision mass spectrometer, Instrons for testing composites at up to 2,000 °C, hot presses that can be used at 2,300 °C, CVD furnaces that can be top or bottom loaded, equipment for measuring thermal expansion up to 700 °C, ultrasonic equipment to detect defects in the C/Cs, and a finite element code for predicting the stresses and strains in polymeric and metal matrix composites. Capital equipment can be bought by the director up to $70,000; otherwise, it must be approved at the corporate level.

Oxidation protection for C/Cs is being investigated whereby protection is derived from layers of SiC or boron carbide that are vapor deposited on the surfaces of the C/C substrates. There is a concerted effort to find a method of minimizing the problem of cracking of the ceramic coatings due to the differential thermal expansion mismatch between the C/C substrate and the oxidation protection layer. The general approach is to vapor deposit six or more layers that each have a different coefficient of thermal expansion (CTE). These values are adjusted so that there is a transition of expansivities that is compatible with both the oxidation protective layer and the carbon substrate. This is called the functional gradient materials (FGM) method, which was successfully developed some years ago by Nippon Oil to form ceramicto-metal seals. The principle is to adjust the porosity of a particular layer in the Assessment transition zone by selecting the proper mixture of gases and temperatures while the layer is being deposited. For example, the carbon content of a gas mixture with SiC can result in different degrees of porosity and variations of CTES from 1 to 4 by 1/1,000,000 per °C. This is the first definitive and more fundamental approach that has been presented by any of the European or Asian organizations that have been visited on this assessment study up to this time.

mixture of gases

Laboratory equipment and facilities are extensive and occupy three floors in one building and one floor of another building for a total floor space that is estimated at more than 80,000 ft2. There were the usual types of equipment for performing chemical analysis and characterization of the precursors and

Concerning the interchange of information with the United States, Nippon Oil feels that years ago the United States was ahead of Japan in the technology of C/Cs and would not talk to people outside of the country. However, now there is more equality of knowledge between Japan and the United States; therefore, Nippon Oil believes that there should be more exchange of information than there is now for the benefit of both countries.

There is a strong possibility that some form of cooperative program can be worked out between Nippon Oil and organizations in the United States so long as it involves research that is being sponsored by the company. Even samples might be exchanged. But if the

work is being sponsored by the Japanese Government, then a whole set of new approvals must be obtained from all the parties in the consortium that are involved in the work. Clearly the outcome is not as predictable.

It appears there is more extensive research going on at Nippon Oil than at many other Japanese industrial organizations. Therefore, it seems prudent to remain in communication with this

fabrication process, establish the market in Japan, and then take it abroad in the future.

The laboratory effort in the C/C area consists of about 10 persons, of which half are professionals and half are for support. Additional persons are doing developmental, prototype, or processing studies in connection with production.

NIPPON STEEL

R&D Laboratories-I

Central R&D Bureau
1618 Ida

Nakahara-ku, Kawasaki 211, Japan
Tel: (044) 777-4111
Date: 18 Dec 1990

Background

The host for this visit was Dr. Ken-ichi Fujimoto, who is the general manager of the chemicals research laboratory, and the arrangements were made by Nabuhiko Narita.

The development of C/Cs began here in 1982 for developing a fabrication process for automobile brakes. Governmental financial support for this work started about 5 years ago. Prof. Yasuda of the Tokyo Institute of Technology, who has been an advisor to the R&D Laboratories, started his research in this area about 10 years ago. So far brakes for cars are not a commercial item because of the cost to fabricate them. So there is a significant effort directed toward finding ways of reducing these fabrication costs. Nevertheless, a significant amount of R&D in the general area of carbon and C/Cs has taken place, especially in the area of aircraft brakes. At this time there is only a small market in Japan for this material. The plan is to perfect the

Pitch-based fibers are being processed at Nippon Steel and sold only in Japan at a rate of about 40 tons/yr; the fibers have a modulus of 600 GPa (87 Mpsi) and a strength of 3.5 GPa (505 kpsi). In addition, about 60 tons of fibers with lower property values are being sold where the modulus is about 75 Mpsi and the strengths vary between 35 and 140 kpsi. Some of the fibers are chemically etched to improve their bonding strengths for use with the different types of organic, metallic, or carbonaceous matrices and even in

concrete.

A major emphasis during these discussions was the research efforts on developing the most effective and efficient methods of densifying carbon/ graphite fiber preforms with an appropriate carbonaceous microstructure that provides the desired physical properties. Generally, three types of preforms are used: unidirectional, 2D, and randomly oriented chopped fibers. For densification purposes, two impregnation processes are being used, gaseous CVD or liquid pitch followed by pyrolysis and additional heat treatment steps.

The CVD densification method was discussed in some detail, especially with respect to the processing conditions and the resulting characteristics of the C/Cs. One of the major factors being evaluated is the weight change or carbon uptake as a function of the deposition time. Nippon Steel has had some success in modeling the deposition phenomena so that the predicted weight

gain as a function of deposition time correlates well with the experimental values for randomly oriented fiber samples that are relatively thin, i.e., 2 mm. For unidirectional samples that contain 55% by volume of fibers, the agreement between theoretical and experimental values is good for the initial part of the process. However, after about two-thirds of the predicted weight pickup has occurred, there is a saturation effect whereby no more weight gain occurs in the experimental samples. Now the investigators are evaluating the changes to the pore structure of the matrix with the processing time because this effect appears to be caused by a premature closing of the pores. This effect is not accounted for in the model as the assumption is made that there is uniform diffusion of the methane throughout the sample during the entire deposition period. For an open structure, like that in the randomly oriented chopped fiber mat, the diffusion coefficient will probably be nearly constant for the 200 or more hours of deposition time as the pores remain relatively open during most of the process. But in the case of the unidirectional samples, the pores are smaller and variable in their cross sections. This means the carbon, which is being uniformly deposited on all the surfaces of the pores, will first close the narrowest diameter pores and prevent further diffusion of the vapors into the remaining pores. Experimental determination of the pore diameters, by the mercury porosimetry method, verifies that this premature pore closure phenomenon is occurring at the end of the impregnation cycle. Consequently, the weight gain will be terminated before the samples are fully densified and the experimental values will be less than the predicted values. It also explains the density gradients that are being observed by these investigators. Therefore, the predictive model must be changed to account for these effects. This work will continue in order to

obtain the best and most uniform matrix densities at the most economical cost. These results using CVD are being compared with the liquid impregnation method of densification.

A combination of phenolic and pitch precursors is being used to densify the substrates, which contain either unidirectionally or 2D oriented fibers. In all cases the fibers are prepregged with phenolic. These are then laid up into stacks of plies, which are carbonized at about 1,300 °C. Next, the preforms are further densified by multiple impregnation cycles of pitch, then carbonized at about 1,300 °C and graphitized above 2,200 °C. Frequently, CVD is used as the final densification step. This has an advantage because the vapor can enter smaller diameter pores than the liquid, so the density of the C/Cs will be increased further. In addition, the mechanical strengths will be improved because better bonding strengths will occur because of the more penetrating vapors. The densities of these pitch and CVD impregnated samples can be in excess of 1.8 g/cc with only three impregnation cycles at about 7 kpsi or 500 atm of pressure above 250 °C. At this time. the preferred impregnation method is liquid pitch, as CVD takes too long and therefore is too costly.

The physical properties of these C/Cs vary with the processing conditions. For example, Nippon Steel has found that the maximum bending strength of unidirectional samples is 1,300 MPa (188 kpsi) at a CVD deposition time of 70 hours. Beyond this time, the strength begins to decline continuously to 300 hours of deposition time. For ribbon type samples a maximum strength of 250 MPa (36,250 psi) is attained at 170 hours. The cause for this maximum strength at an optimum deposition time is not understood and investigations are continuing to identify the influential parameters. Data indicate that the bending strengths of 2D composites are 10% to 20% higher than those that have been densified by the liquid pitch

method even if the latter samples have a slightly higher density.

Nippon Steel's research shows that the microstructure of the as-deposited CVD matrix is important in determining the final type of microstructure in the matrix after final heat treatment of the C/C has occurred. So if it is desirable to have a highly graphitic structure matrix, the as-deposited matrix must have a rough laminar type of microstructure and the final heat treatment temperature needs to be in the range of 2,500 °C. Also, the strength of such samples is reduced by the high temperature heat treatments.

is a graphitization furnace for heat treating samples as large as 100 mm OD and 30 mm thick in a nitrogen atmosphere to >2,400 °C. The CVD equipment is used to prepare the abovementioned samples, which incidently, have a geometry that is similar to that of brake disks. It also appears that these laboratories are well equipped to perform evaluations of these materials as there are SEMS, TEMs, laser Raman interferometers, and x-ray units, as well as the standard types of laboratory equipment needed to do chemical analysis and measure thermalmechanical properties of materials. During this tour various types of materials were evident, which indicated the diversity of applications being investigated in which carbon fibers or composite materials are evaluated for their applicability and marketability, such as for the Japanese space plane where the top use temperature may be 1,700 °C, brakes for high speed trains, fibers (4%) in concrete to strengthen it, complex shapes of C/Cs for high strength and temperature applications such as for dies or fasteners, and thermal insulation where strength is required.

Oxidation protection layers of silicon carbide and other unidentified coatings are being deposited on C/Cs by the CVD method. Studies are underby the CVD method. Studies are underway to minimize the differential therway to minimize the differential thermal expansion problem that exists between these oxidation protection between these oxidation protection layers and the carbonaceous substrates. One approach that is being employed is similar to that which is used for making ceramic-to-metal seals, namely, a number of layers are being deposited on the substrate that contain different values of coefficients of thermal expansion. In this manner a transition region is formed that accommodates the differences of expansion that exist between the C/C and the protection layer, reducing the mechanical stresses and cracking of the protective layer. This approach Assessment is similar in concept to the FGM one that is being studied by Nippon Oil.

Nippon Steel is the only organization that was visited in Japan or in Asia tion that was visited in Japan or in Asia that is conducting research on the creep characteristics of C/Cs. This is being done using unidirectional samples to determine the influence of temperature, stress, and geometry of the sample. A negligible creep effect was found ple. A negligible creep effect was found when four-point bend tests were used below 1,500 °C.

A brief laboratory facilities tour was conducted with particular emphasis on the special equipment that is used to prepare the laboratory samples. There

These laboratory facilities are extensive, as they take up three floors of space with an estimated area over 80,000 ft2.

Collaboration with U.S. organizations is definitely desired by Nippon Steel, according to Dr Fujimoto.

Future directions of C/C research efforts will probably remain the same for the next 8 to 10 years. Fujimoto also noted that 29 organizations are currently developing methods of fabricating C/Cs so they will not be left out of the race if this material suddenly becomes needed to fill a big commercial market. But in the next 5 or more years, this number of companies will be sharply reduced to a few when it is clear that the market is much smaller than was