February 2000 Fugitive Interfacial Carbon Coatings for Oxide/Oxide Composites (B) Strength: As expected, the strengths of the uncoated I. Summary/Conclusions omposites are very low, -25 MPa, in both the 0 and +45 orientations. The strengths of the "C composites are significantly Fugitive carbon coatings have been shown to be useful in dense igher in the 0 orientation, and they increase with"C thickness This may be due to the nonuniformity of the coatings; the carbon 80% of their as-processed strength after long-term heat treatment coverage of the fiber surface may not be complete and may in air at 1000C for 500 h. The coating thickness appeared to affect increase with coating thickness. This is also consistent with th the composite properties, with the slightly thicker coating produc- fact that the strengths of the"C composites are low compared to ing better results in unid nal composites, while the thinner what might be expected from the tow strength of the fibers. The oating was more advantageous in the +45 samples. For a given stem, the coating thickness had to be optimized with respect to "C"tows exhibit strengths ranging from -477(0.04 um"C) to off-axis properties and high-temperature behavior. The overall 750 MPa(0.02 um"C ) hence, in a composite with a fiber volume percentage of 32.5, composite strengths of 155-244 MPa trengths of the dense matrix composites in this work were ignificantly lower than anticipated because of fiber strength loss are expected. Incomplete coverage or coating damage might result during"C"coating and, presumably, because of chemical interac- in fiber/matrix bonding, which could then cause premature fiber tion between the fiber and matrix in regions where the coating was failure. The strengths of the composites should be increased by discontinuous. This was supported by the continued strength loss preventing fiber/matrix interaction through improved fiber coat- in the samples containing uncoated Nextel 720. To avoid this problem, "C coatings with complete fiber coverage should be The fugitive composites are somewhat inferior to the"C used, along with a more compatible matrix for the Nextel 720 containing composites. This can be explained based on th ossibility that as the"C is removed, some regions of the fiber For porous matrix composites, it was shown that composite and matrix touch and bond, again causing premature fiber failure. strength, after a long-term exposure at elevated temperatures, was (2 Porous Matrix Composites not dependent on the state of the interface. This confirmed existin iews conceming these materials. Fugitive coatings might be The strengths of the porous matrix composites with"C" are beneficial in porous matrix composites for exposures at higher found to be completely retained after the carbon is removed. This temperatures or for reactive fiber/matrix combinations. This shows that, within experimental error, there is no difference again, will be system-specific between having a"C layer and having a gap at the interface in paring the fugitive dense th the porous matrix composites. The strength difference between the matrix composites showed that the latter exhibited superior uncoated Nextel 720/porous matrix composites and the " C trengths. It is anticipated, however, that the strengths of composites is unexpected. However, this can be rationalized if the fugitive dense matrix composites can be increased by choosing experimental tow strengths of the fibers, the UTS of the compos- fugitive d ces of fiber degradation. Ginn by elimae"C"cov- test data are compared after normalizing the ultimate tensile better fiber/matrix chemistries and by having complete"Ccov- me.ngth(UTS) with respect to tow strength. Table II lists the age on the fibers during processing, there ing addi- asured fiber volume fraction in the loading direction, the same strength, the composites are expected to be superior to ites, and the calculated ratio of UTS to the expected strength, vo posites in applications requiring matrix (rule of mixtures). The first two data sets in the table refer to the dominated orous matrix composites. It is seen that, based on the ratio UTS/o the fugitive composite is actually slightly better than the uncoated composite. Therefore, the present work shows that the porous matrix composites, with or without a fugitive gap, are The authors would like to thank Dr. Kenneth Chyung stable at s1150C for a service life of 500 h. In summary, one can supplying the CAS onclude that porous matrix composites do not gain significanth siting cvd carbon coatings 720 fabric from a weak interfacial layer; however, in some cases, a weak Thanks also to mr Cook for sample preparation and interlayer or a fugitive gap may be useful in extending the life of the composite by preventing fiber/matrix reactions References R.J. Kerans, R. S Hay, N. J. Pagano, and T. A. Parthasarathy," The Role of the aber-Matrix Interface in Ceramic Composites, Am. Ceram. Soc. Bull, 68[21 ( Fugitive Dense Matrix versus Porous Matrix Composites 429-42(1989) 2A. G. Evans, F. W. Zok, and J. B. Davis, "The Role of Interfaces in Fiber It is worth comparing the data of the fugitive dense matrix Reinforced Brittle Matrix Composites, "Compos. Sci. Technol., 42, 3-24(1991) opposites with those of porous matrix composites to determine BR. J. Kerans, "Issues in the Control of Fiber-Matrix Interface Properties in which approach offers more engineering value. Porous matrix Ceramic Composites, "Scr. Metall. Mater, 31[8]1079-1084(1994).Annu.Rev composites likely suffer from poor matrix-dominated properties, Mater. Sci., 27, 499-232419o>posite Interfaces: Properties and Design," such as transverse strength/creep, low thermal conductivity, and and Oxide Coatings on Continuous Ceramic Fibers". pp. 377-82 in Ceramic Matrix ear/abrasion. The fugitive dense matrix composites are expected to have better thermal conductivity and wear/abrasion resistance Commposites-Adanced High Temperature Structural Materials, Materials Research sium Proceedings, Vol. 365(Boston, MA, December 1994). Edited by Thus, for engineering use, one might select one over the other, R.A. Lowden, M. K Ferber, J. R. Hellmann, and S G. DiPetro, Materials Research depending on the specific application needs Because the fiber strengths are not the same in all composites, Hermes, "Sol-Gel Coatings on Continuous Ceramic Fibers, Ceram. Eng. Sci. Proc 11 19-10) 1526-32(1990) it is appropriate to compare the ratio of the UTs normalized with E. Boakye, M. D Petry, and R. S. Hay, "Porous Aluminum Oxide and Lanthanun respect to the tow strengths. Table II includes the data on the Phasphate Fiber Coatings," Ceram, Eng. Sci. Proc, 17 (4)53-60(1996xe-Matrix Contrasting these results with those of the porous matrix compos- 1233-46(1996) erived from Sol-Gel Fiber Coatings, J. Amm. Ceram. Soc., 79 [51 tes, it is apparent that the porous matrix composites retain better trengths than the fugitive dense matrix composites. As mentioned Fiber-Oxide Matrix Composites, Ceram. Eng. Sci. Proc., 15 [51743-52(1994). arlier, the lower strength of the fugitive dense matrix composites L. U. J. T. Ogbuji, "A Porous, Oxidation Resistant Fiber Coating for CMC Interface, " Ceram. Eng. Sci. Proc., 16 14]497-505(1995). may be due to fiber/matrix interaction, which can be eliminated by Carpenter and J. Bohlen, "Fiber Coatings for Ceramic-Matrix Composites, a better choice of chemistries and/or complete "C" coverage on th 8-56(1992) F. Rebillat, A. Bleier, T. M fibers. The possibility of such a reaction, however, precludes Lara-Curzio and p. K Liaw. "Oxidation-Resistant Interfacial Coatings for Contin- definitive comparison of these composite types uous Fiber Ceramic Composites, Ceram. Eng. Sci. Proc., 16[41389-99(1995)(B) Strength: As expected, the strengths of the uncoated composites are very low, ;25 MPa, in both the 0° and 645° orientations. The strengths of the “C” composites are significantly higher in the 0° orientation, and they increase with “C” thickness. This may be due to the nonuniformity of the coatings; the carbon coverage of the fiber surface may not be complete and may increase with coating thickness. This is also consistent with the fact that the strengths of the “C” composites are low compared to what might be expected from the tow strength of the fibers. The “C” tows exhibit strengths ranging from ;477 (0.04 mm “C”) to 750 MPa (0.02 mm “C”); hence, in a composite with a fiber volume percentage of 32.5, composite strengths of 155–244 MPa are expected. Incomplete coverage or coating damage might result in fiber/matrix bonding, which could then cause premature fiber failure. The strengths of the composites should be increased by preventing fiber/matrix interaction through improved fiber coat￾ings. The fugitive composites are somewhat inferior to the “C”- containing composites. This can be explained based on the possibility that as the “C” is removed, some regions of the fiber and matrix touch and bond, again causing premature fiber failure. (2) Porous Matrix Composites The strengths of the porous matrix composites with “C” are found to be completely retained after the carbon is removed. This shows that, within experimental error, there is no difference between having a “C” layer and having a gap at the interface in porous matrix composites. The strength difference between the uncoated Nextel 720/porous matrix composites and the “C” composites is unexpected. However, this can be rationalized if the test data are compared after normalizing the ultimate tensile strength (UTS) with respect to tow strength. Table II lists the measured fiber volume fraction in the loading direction, the experimental tow strengths of the fibers, the UTS of the compos￾ites, and the calculated ratio of UTS to the expected strength, Vf sf (rule of mixtures). The first two data sets in the table refer to the porous matrix composites. It is seen that, based on the ratio UTS/Vf sf , the fugitive composite is actually slightly better than the uncoated composite. Therefore, the present work shows that the porous matrix composites, with or without a fugitive gap, are stable at #1150°C for a service life of 500 h. In summary, one can conclude that porous matrix composites do not gain significantly from a weak interfacial layer; however, in some cases, a weak interlayer or a fugitive gap may be useful in extending the life of the composite by preventing fiber/matrix reactions. (3) Fugitive Dense Matrix versus Porous Matrix Composites It is worth comparing the data of the fugitive dense matrix composites with those of porous matrix composites to determine which approach offers more engineering value. Porous matrix composites likely suffer from poor matrix-dominated properties, such as transverse strength/creep, low thermal conductivity, and wear/abrasion. The fugitive dense matrix composites are expected to have better thermal conductivity and wear/abrasion resistance. Thus, for engineering use, one might select one over the other, depending on the specific application needs. Because the fiber strengths are not the same in all composites, it is appropriate to compare the ratio of the UTS normalized with respect to the tow strengths. Table II includes the data on the control (uncoated and “C”) and fugitive dense matrix composites. Contrasting these results with those of the porous matrix compos￾ites, it is apparent that the porous matrix composites retain better strengths than the fugitive dense matrix composites. As mentioned earlier, the lower strength of the fugitive dense matrix composites may be due to fiber/matrix interaction, which can be eliminated by a better choice of chemistries and/or complete “C” coverage on the fibers. The possibility of such a reaction, however, precludes definitive comparison of these composite types. IV. Summary/Conclusions Fugitive carbon coatings have been shown to be useful in dense matrix composites. Nextel™ 720/“C”/CAS composites retained ;80% of their as-processed strength after long-term heat treatment in air at 1000°C for 500 h. The coating thickness appeared to affect the composite properties, with the slightly thicker coating produc￾ing better results in unidirectional composites, while the thinner coating was more advantageous in the 645° samples. For a given system, the coating thickness had to be optimized with respect to off-axis properties and high-temperature behavior. The overall strengths of the dense matrix composites in this work were significantly lower than anticipated because of fiber strength loss during “C” coating and, presumably, because of chemical interac￾tion between the fiber and matrix in regions where the coating was discontinuous. This was supported by the continued strength loss in the samples containing uncoated Nextel 720. To avoid this problem, “C” coatings with complete fiber coverage should be used, along with a more compatible matrix for the Nextel 720 fibers. For porous matrix composites, it was shown that composite strength, after a long-term exposure at elevated temperatures, was not dependent on the state of the interface. This confirmed existing views concerning these materials. Fugitive coatings might be beneficial in porous matrix composites for exposures at higher temperatures or for reactive fiber/matrix combinations. This, again, will be system-specific. Comparing the fugitive dense composites with the porous matrix composites showed that the latter exhibited superior strengths. It is anticipated, however, that the strengths of the fugitive dense matrix composites can be increased by choosing better fiber/matrix chemistries and by having complete “C” cov￾erage on the fibers during processing, thereby eliminating addi￾tional sources of fiber degradation. Given the same strength, the fugitive dense matrix composites are expected to be superior to porous matrix composites in applications requiring matrix￾dominated properties. Acknowledgments: The authors would like to thank Dr. Kenneth Chyung from Corning, Inc., for supplying the CAS glass-ceramic powder and Dr. Theodore Besmann at Oak Ridge National Laboratory for depositing CVD carbon coatings on Nextel 720 fabric. Thanks also to Mr. Marlin Cook for sample preparation and microscopy. References 1 R. J. Kerans, R. S. Hay, N. J. Pagano, and T. A. Parthasarathy, “The Role of the Fiber-Matrix Interface in Ceramic Composites,” Am. Ceram. Soc. Bull., 68 [2] 429–42 (1989). 2 A. G. Evans, F. W. Zok, and J. B. Davis, “The Role of Interfaces in Fiber￾Reinforced Brittle Matrix Composites,” Compos. Sci. Technol., 42, 3–24 (1991). 3 R. J. Kerans, “Issues in the Control of Fiber-Matrix Interface Properties in Ceramic Composites,” Scr. Metall. Mater., 31 [8] 1079–1084 (1994). 4 K. T. Faber, “Ceramic Composite Interfaces: Properties and Design,” Annu. Rev. Mater. Sci., 27, 499–524 (1997). 5 R. S. Hay, M. D. Petry, K. A. Keller, M. K. Cinibulk, and J. R. Welch, “Carbon and Oxide Coatings on Continuous Ceramic Fibers”; pp. 377–82 in Ceramic Matrix Composites—Advanced High Temperature Structural Materials, Materials Research Society Symposium Proceedings, Vol. 365 (Boston, MA, December 1994). Edited by R. A. Lowden, M. K. Ferber, J. R. Hellmann, and S. G. DiPetro. Materials Research Society, Pittsburgh, PA, 1995. 6 R. S. Hay and E. E. Hermes, “Sol–Gel Coatings on Continuous Ceramic Fibers,” Ceram. Eng. Sci. Proc., 11 [9–10] 1526–32 (1990). 7 E. Boakye, M. D. Petry, and R. S. Hay, “Porous Aluminum Oxide and Lanthanum Phosphate Fiber Coatings,” Ceram. Eng. Sci. Proc., 17 [4] 53–60 (1996). 8 M. K. Cinibulk and R. S. Hay, “Textured Magnetoplumbite Fiber-Matrix Interphase Derived from Sol–Gel Fiber Coatings,” J. Am. Ceram. Soc., 79 [5] 1233–46 (1996). 9 L. C. Lev and A. S. Argon, “Development of Oxide Coatings for Matching Oxide Fiber-Oxide Matrix Composites,” Ceram. Eng. Sci. Proc., 15 [5] 743–52 (1994). 10L. U. J. T. Ogbuji, “A Porous, Oxidation Resistant Fiber Coating for CMC Interface,” Ceram. Eng. Sci. Proc., 16 [4] 497–505 (1995). 11H. Carpenter and J. Bohlen, “Fiber Coatings for Ceramic-Matrix Composites,” Ceram. Eng. Sci. Proc., 13 [7–8] 238–56 (1992). 12S. Shanmugham, D. P. Stinton, F. Rebillat, A. Bleier, T. M. Besmann, E. Lara–Curzio, and P. K. Liaw, “Oxidation-Resistant Interfacial Coatings for Contin￾uous Fiber Ceramic Composites,” Ceram. Eng. Sci. Proc., 16 [4] 389–99 (1995). February 2000 Fugitive Interfacial Carbon Coatings for Oxide/Oxide Composites 335