« PreviousContinue »
When used to incorporate glass formers, the matrix chemical modification technique has the advantage of producing increasing quantities of protective material as oxygen intrusion proceeds. This process results in the filling of surface pores and the formation of a coherent glass layer beneath the external coating. Particulate additions can be made simply and in relatively high concentrations. However, powders have the disadvantage of not penetrating the fiber yarn bundles and may mechanically damage fibers during composite consolidation. Segregation of the particulate between the fabric plies of CC laminates also increases the ply spacing and results in reduced inplane mechanical properties. These problems currently are being addressed by the use of fiber coatings and optimization of powder particle sizes.
Although molecular alterations to the resin or pitch matrix precursors have the advantage of penetrating the yams to produce more uniform protection, the concentrations of active species that can be incorporated by this technique are much lower than can be achieved by powder loading. For this reason, the particulate loading and molecular alteration techniques may be most effective in combination with one another.
The CVI method for providing internal oxidation protection is extremely versatile; it can be used to chemically modify the carbon matrix, to coat the internal surfaces, or to completely replace the carbon matrix with another material. This method is limited to nonoxides when carbon is present and is a time-consuming and inefficient process as normally practiced when used to produce a composite matrix. However, recent work has demonstrated that satisfactory densification of fiber preforms can be accomplished in periods of 24 hours or less if the reactive gases are forced through the fibers and a thermal gradient is maintained to create a front of preferred deposition within the composite (refs. 112 to 114). Figure 14 shows the type of forced-flow thermal gradient arrangement currently being used for rapid CVI within deposition chambers at General Atomics.
Although actual temperature limitations on the coatings currently being used for CC oxidation protection will vary depending on specific operating conditions and component performance requirements, general guidelines can be established based on experimental observations and the known thermochemical properties of the coating constituents. For example, unprotected B2O3 glass layers were shown to be effective in retarding the oxidation of CC substrates for extended periods at temperatures up to about 1000°C, but volatilization of the glass is limiting at higher temperatures (ref. 77). When B2O3 was used as a crack sealant for hard outer coatings or as a binder in coatings composed of hard oxide particles, oxidation protection was extended to temperatures of more than 1200°C (ref. 78). Tests have
demonstrated that coating systems of the type shown in figure 6. in which a boronrich inner layer provides glass to seal cracks in a CVD SiC outer layer, can provide hundreds of hours of oxidation protection under thermal cycling conditions with maximum temperatures close to 1400°C (ref. 65).
The maximum temperature at which BoO.j can be used in contact with carbon at atmospheric pressure is approximately I575°C because of the disruption of the glass by equilibrium CO reaction product pressures greater than 10 MPa (I atm) at higher temperatures (ref. 79). The vapor pressure of B2O3 at this temperature also is projected to be about 10 MPa under dry conditions and close to 10~1 MPa in moist environments (rets. 74 and 77). These factors, combined with the tendency previously described for BoO;; to corrode the siliconbased outer coatings, suggest maximum temperatures for the borate glass internal and glass-sealed external coatings of 1500°C to 1550°C even for short periods of performance.
Certain applications do not require glass-sealed outer coatings because the component is rapidly heated to and cooled from temperatures high enough to close the coating coefficient of thermal expansion (CTE) mismatch cracks. Relative to this situation, pure CVD SiC and Si;jN| coatings are viable in the I700CC to 1800°C range under strongly oxidizing conditions (refs. 65 and 115). This temperature limit exists because at higher temperatures the SiOo layers that normally protect these materials are disrupted by CO and No interfacial reaction product pressures that become greater than l()~1 MPa. thus causing the coatings to erode by uncontrolled oxidation. Under less oxidizing conditions, the temperature limit can be reduced to the 1500°C 10 I600°C range by active oxidation (refs. 116 and 117). Gaseous SiO is then formed instead of the protective Si02, and coatings can fail by rapid loss of material.
Once the temperature limit for SiC and Si3N4 coatings is exceeded, identifying coatings with acceptably low oxygen permeabilities becomes the issue. For an external 100-/xm-thick coating, estimated coating permeabilities of approximately 10~8 and 10-10 g-cm -s or less would be required to provide acceptable shortand long-term oxidation protection of CC composites (ref. 79). Figure 11 shows that at temperatures above 1800°C, the materials expected to meet these criteria are Si(\>. AI2O3, and the high-temperature noble metals, rhodium and iridium. At such high temperatures, materials that form refractory oxides such as ZrO2 and HfC>2 oxidize very rapidly because the oxides are inherently permeable to oxygen. Based on previous work and the considerations discussed here, current ultrahigh-temperature coating development, for all but very short-term applications, is focused on concepts that use iridium and SiOi as essential constituents.
More than 20 years ago, extensive work was conducted on iridium and iridium alloy coatings to be used to protect synthetic graphite bodies from oxidation at very high temperatures (ref. 87). Iridium is attractive because of its 2440°C melting point, extremely low oxygen and carbon permeabilities, and inertness to carbon. Coatings were made by a variety of techniques including CVD, electroplating, and a combination of plasma spraying and hot-isostatic pressing. Although excellent protection was demonstrated for short times in the 2000°C to 2100°C range, difficulties in fabricating high-quality coatings and the cost and availability of iridium were deterrents to further development. Iridium and rhodium are also susceptible to erosion by the formation of volatile oxides and have CTE values that are extremely high compared with those of CC composite materials (ref. 118). Coating erosion may or may not be an issue depending on oxygen partial pressure, gas flow conditions, and required time of performance. Oxide overcoatings have been proposed to inhibit erosion, but the CTE mismatch problem is profound and might very well prohibit the use of these materials as external coatings on CC composites. This same problem exists for AI2O3 and most other oxides.
As an external coating, the advantages of Si02 are a very low CTE and an ability to flow at high temperatures to accommodate differential strains. The viscosity of Si02 is about 107 dPa-s at 1800°C (ref. 74). Then, the glass will adhere on contact, begin to flow under its own weight, and can be considered to behave as a sealant. Because Si02 is reduced by carbon and is unstable in contact with carbides at very high temperatures, Si02 glass must be separated from the CC substrate by a hard oxide or some other compatible material. An outer coating must be used for most applications to prevent flow erosion and excessive vaporization. The vapor pressure of Si02 under strongly oxidizing conditions is projected to be somewhat less than 10-5 MPa at 2000°C (ref. 119). Low oxygen pressures increase the volatility of SiO2 by promoting SiO formation.
Inner and outer coatings are required to envelop and protect the SiC«2, which invariably leads to the consideration of materials that are unattractive from the CTE standpoint. It is interesting to note that several refractory oxides are known to exhibit very low CTE values. This phenomenon is a result of lowtemperature microcracking. which produces extreme crystallographic anisotropy. The recombination of the microcracks produces significantly increased CTE values at higher temperatures (refs. 4 and 120). Furthermore, any appropriate oxide in contact with carbon at very high temperatures will form an interfacial carbide. This reaction raises the issue of continued conversion of the oxide to the carbide by transport of carbon through the carbide. The presence of the carbide also reintroduces the CTE problem. The most refractory of the low-CTE oxide is HfTi04, which melts incongruently at about 1980°C (ref. 121).
Materials selection for internal coatings to perform at very high temperatures is even more limited than the selection for external oxygen and carbon diffusion barriers. The coatings need to be chemically stable when in contact with carbon, and it is likely that this stability must result from thermodynamic compatibility rather than transport limitations because of the very short diffusion distances. Because the oxides are reduced by carbon at the temperatures in question and because rhodium forms a eutectic melt with carbon at about 1700°C, the only material that clearly meets the above criteria is iridium (ref. 118). Totally replacing the carbon matrix with a ceramic would prevent matrix gasification, but experience has shown that this replacement would do little over extended periods of time to protect the carbon fibers from oxidation (ref. 108). Fiber protection would require thin, adherent barrier coatings that are chemically compatible with the fibers and matrix. Again, iridium appears to be the preferred material.
1. Blakslee, O. L.; Proctor, D. C; Seldin, E. J.; Spence, G. B.; and Weng, T.: Elastic Constants of Compression-Annealed Pyrolytic Graphite. J. Appl. Phys., vol. 41, no. 8, July 1970, pp. 3373-3382.
2. Riley, W. C.: Graphite. Ceramics for Advanced Technologies, J. E. Hove and W. C. Riley, eds., John Wiley & Sons, Inc., c.1965, pp. 14-75.
3. Riley, W. C.: Pyrolytic Materials. Ceramics for Advanced Technologies, J. E. Hove and W. C. Riley, eds., John Wiley & Sons, 1965, pp. 235-260.
4. Buessem, W. R.: Internal Ruptures and Recombinations in Anisotropic Ceramic Materials. Mechanical Properties of Engineering Ceramics, W. Wurth Kriegel and Hayne Palmour ITI, eds., Interscience Publ. Inc., 1961, pp. 127-148.
5. Chen, K. J.; LeMaistre, C. W.; Wang, J. H.; and Diefendorf, R. J.: The Consequences of Residual Stresses in High Modulus Carbon Fibers on Composite Properties. Polymer Prepr., vol. 22, no. 2, 1981, pp. 212-215.
6. Bacon, R.: Carbon Fibers From Rayon Precursors. Chemistry and Physics of Carbon, Volume 9, P. L. Walker and P. A. Thrower, eds., Marcel Dekker, 1973, pp. 1-102.
7. Edie, D. D.: Research Into Pitch-Based Carbon Fibers. Recent Research Into Carbon-Carbon Composites, M. Genisio, ed., Southern Illinois Univ., 1987, pp. 16-41.
8. Diefendorf, Russell J.: Carbon/Graphite Fibers. Composites, Volume 1 of Engineered Materials Handbook?^, ASM International, c.1987, pp. 49-53.
9. Riley, W. C.: Carbon-Carbon Space Structures. Paper presented at the 10th Annual Conference on Composites and Advanced Ceramic Materials (Cocoa Beach, Florida), 1986.
10. Judd, N. C. W.: The Chemical Resistance of Carbon Fibers and a Carbon Fiber Polyester Composite. Proceedings of the First International Conference on Carbon Fibres, Plastics Inst. (London), 1971, pp. 258-265.
11. McKee, D. W.; and Mimeault, V. J.: Surface Properties of Carbon Fibers. Chemistry and Physics of Carbon—A Series of Advances, Volume 8, P. L. Walker, Jr., and Peter A. Thrower, eds., Marcel Dekker, Inc., 1973, pp. 151-245.
12. Gibbs, Hugh H.; Wendt, Robert C.; and Wilson, Frank C.: Carbon Fiber Structure and Stability Studies. Polymer Eng. & Sci., vol. 19, no. 5, Apr. 1979, pp. 342-349.
13. Eckstein, Bernard H.: The Weight Loss of Carbon Fibers in Circulating Air. Materials for Space—The Gathering Momentum, Volume 18 of International SAMPE Technical Conference, J. T. Hoggatt, S. G. Hill, and J. C. Johnson, eds., Soc. for the Advancement of Material and Process Engineering, 1986, pp. 149-160.
14. Amateau, M. F.: Progress in the Development of Graphite-Aluminum Composites Using Liquid Infiltration Technology. J. Compos. Mater., vol. 10, Oct. 1976, pp. 279-296.
15. Weisweiler, W.; Fitter, E.; Nagel, G; and Jager, H.: R.F.-Sputtered SiC Coatings on Carbon Fibres. Thin Solid Films, vol. 148, no. 1, Mar. 30, 1987, pp. 93-108.