August 1997 Tensile Stress Rupture of sic/siCm Minicomposites with C and B Interphases at Elevated Temperatures in Air 2037 with carbon interphases. The minicomposites seem to have better rupture properties than the 0/90 Nic-SiC composites The average fiber volume fraction was determined to be% in the loading direction for these composites. 4 If the volume fraction of fibers was less at the fracture surface. the data for the 0/90 Nic-SiC would improve in Fig. 13(a). Also, because these composites were 0/90, other factors, such as the mor phology of the porosity and crossply cracks, could lead to greater embrittlement compared to the"simple" minicompos- ites of this stud The BN-interphase minicomposite data mimics the fiber rupture data, although the BN-minie is lower at room temperature. To compare the rupture behavior of Hi-Nicalon fibers to that of the Hn minicomposites, the fiber data was normalized to the room-temperature strength of the bn minicomposites (Fig. 13(b)). Two regions represente the different mechanisms that were considered to cause fiber PBN7 2 5.0 kv x1 0a Weiderhorn et al.25). The lower-temperature re gion(mild dependence on q) is due to slow crack growth, and the higher-temperature region(strong dependence on g) is due to creep rupture(Fig t lower temperatures(≤950°C b data), the minicomposite rupture data follows the fiber data (the same slope), indicating that the rupture strength decreases because of a slow-crack-growth mechanism(in air) that is in- (1200C), the minicomposite rupture strength was within the scatter of the single-fiber rupture data and appeared to have the same dependence on q. These two regions are designated"Fi- ber Slow-Crack Growth Controlled Degradation"andFiber Creep-Rupture Controlled Degradation, respectively, in Fig The BN-HN minicomposite rupture strengths between he low-and high-temperature regions fall below the fiber rupture strength. Thi region is designated""Compos Embrittlement due to Stressed-Oxidation'in Fig. 13(b)and must be caused by the interaction of the precracked mini composite with the environment. The 3MBN-Nic mini- PBN8 x4.00k7. 5eu however,more tests would need to be performed&. 13(a): composites seem to have the same three regions(Fi this observation. The significant features of these obse Fig 9. Higher. cation micrographs of PBN-HN mi ite that failed at for carbon- and BN-interphase minicomposites are of the bright fibers in the middle of are oxidized; i.e, the fibers had failed prior to minicomposit (2) C-Nic Mimicomposite Time-Dependent Failure The observation that the carbon interphase is lost via oxida T(log tR+C) ion and that the fibers eventually contact the matrix was as xpected5-7 Because the interphase was vacated and SiO, for mation was negligible at 700%C. embrittlement that is caused C the Larson-Miller constant, which was 22 for both Nicalon bers is considered unlikely. The fact that fiber fracture the fi where T is temperature(in K), Ig the time to rupture (in h), and by strong bonding and severe stress concentrations on and Hi-Nicalon fibers. 8 The minicomposite stress-rupture data do not occur most often at the point of fiber-matrix can then be related to the fiber stress-rupture data via the Lar- indicates that the fibers either are not bonded or son-Miller relationship, using the same C value for the mini- strongly bonded to the matrix. In addition, the fiber surface mirror regions increase(the fiber strength decreases) F13)0)dp如的 ther flaws are growing or new flaws are created od q. The individual fiber data from the literature are from stress- fiber sur rupture experiments(1180%-1400C). 9,20 The scatter in the rupture must be due to fiber degradation. data is large for the stress-rupture data, as represented in Fig Also shown in Fig. 13(a)are composite stress-rupture data Time-Dependent Failure composite 13(b) 3) BN-Interphase Min for Nicalon/C/CVI-SiC composites, 23 and Nicalon/BN/BMAS The BN-interphase minicomposites showed significantly glass-ceramic composites. Both of the composites were pre- better time-dependent failure properties, in comparison to the cracked; however, the exterior skin of the BMAS composite carbon-interphase minicomposites. This was observed sealed"the interior of the composite from the environment, though significant microstructural changes had occurred as described earlier, enabling this composite to survive for (A) Effect of Humidity on BN: The BN volatilization was more than 1.5 years surprising, especially at such low temperatures($500C). It is The rupture-strength decrease of C-interphas I known that BN may react with water vapor, even at room es(ig. 13(a))with q is significantly greater than that of mperature, depending on the purity and crystallinity of the BN-interphase minicomposites(Figs. 13(a)and(b). The rup- BN 26,27 However. there is recent evidence that bN can be ture strengths of the carbon-interphase minicomposites are very consumed at intermediate temperatures in humid environ- similar to the data in the literature for Nic SiCm composites ments, depending on the crystallinityq 4 T(log tR + C) (2) where T is temperature (in K), tR the time to rupture (in h), and C the Larson–Miller constant, which was 22 for both Nicalon and Hi-Nicalon fibers.18 The minicomposite stress-rupture data can then be related to the fiber stress-rupture data via the Lar￾son–Miller relationship, using the same C value for the mini￾composites. Figures 13(a) and (b) show plots of the Nicalon and Hi￾Nicalon minicomposite and fiber data on applied stress versus q. The individual fiber data from the literature are from stress￾rupture experiments (1180°–1400°C).19,20 The scatter in the data is large for the stress-rupture data, as represented in Fig. 13(b). Also shown in Fig. 13(a) are composite stress-rupture data for Nicalon/C/CVI-SiC composites8,23 and Nicalon/BN/BMAS glass-ceramic composites.2 Both of the composites were pre￾cracked; however, the exterior skin of the BMAS composite ‘‘sealed’’ the interior of the composite from the environment, as described earlier, enabling this composite to survive for more than 1.5 years. The rupture-strength decrease of C-interphase minicompos￾ites (Fig. 13(a)) with q is significantly greater than that of BN-interphase minicomposites (Figs. 13(a) and (b)). The rup￾ture strengths of the carbon-interphase minicomposites are very similar to the data in the literature for Nicf /SiCm composites with carbon interphases. The minicomposites seem to have better rupture properties than the 0/90 Nic–SiC composites. The average fiber volume fraction was determined to be ∼22% in the loading direction for these composites.24 If the volume fraction of fibers was less at the fracture surface, the data for the 0/90 Nic–SiC would improve in Fig. 13(a). Also, because these composites were 0/90, other factors, such as the mor￾phology of the porosity and crossply cracks, could lead to greater embrittlement compared to the ‘‘simple’’ minicompos￾ites of this study. The BN-interphase minicomposite data mimics the fiber￾rupture data, although the BN-minicomposite ultimate strength is lower at room temperature. To compare the rupture behavior of Hi-Nicalon fibers to that of the HN minicomposites, the fiber data was normalized to the room-temperature strength of the BN minicomposites (Fig. 13(b)). Two regions represented the different mechanisms that were considered to cause fiber rupture (after Weiderhorn et al.25). The lower-temperature re￾gion (mild dependence on q) is due to slow crack growth, and the higher-temperature region (strong dependence on q) is due to creep rupture (Fig. 13). At lower temperatures (#950°C data), the minicomposite rupture data follows the fiber data (the same slope), indicating that the rupture strength decreases because of a slow-crack-growth mechanism (in air) that is in￾herent to the fibers themselves. At the higher temperatures (1200°C), the minicomposite rupture strength was within the scatter of the single-fiber rupture data and appeared to have the same dependence on q. These two regions are designated ‘‘Fi￾ber Slow-Crack Growth Controlled Degradation’’ and ‘‘Fiber Creep-Rupture Controlled Degradation,’’ respectively, in Fig. 13(b). The BN-HN minicomposite rupture strengths between the low- and high-temperature regions fall below the fiber rupture strength. This region is designated ‘‘Composite Embrittlement due to Stressed-Oxidation’’ in Fig. 13(b) and must be caused by the interaction of the precracked mini￾composite with the environment. The 3MBN-Nic mini￾composites seem to have the same three regions (Fig. 13(a)); however, more tests would need to be performed to confirm this observation. The significant features of these observations for carbon- and BN-interphase minicomposites are discussed below. (2) C-Nic Minicomposite Time-Dependent Failure The observation that the carbon interphase is lost via oxida￾tion and that the fibers eventually contact the matrix was as expected.5–7 Because the interphase was vacated and SiO2 for￾mation was negligible at 700°C, embrittlement that is caused by strong bonding and severe stress concentrations on the fi￾bers is considered unlikely. The fact that fiber fracture origins do not occur most often at the point of fiber–matrix contact indicates that the fibers either are not bonded or are not strongly bonded to the matrix. In addition, the fiber fracture surface mirror regions increase (the fiber strength decreases) with longer rupture times (Fig. 12), from which one can infer that either flaws are growing or new flaws are created on the fiber surfaces. Therefore, the cause of C-Nic minicomposite rupture must be due to fiber degradation. (3) BN-Interphase Minicomposite Time-Dependent Failure The BN-interphase minicomposites showed significantly better time-dependent failure properties, in comparison to the carbon-interphase minicomposites. This was observed even though significant microstructural changes had occurred. (A) Effect of Humidity on BN: The BN volatilization was surprising, especially at such low temperatures (#500°C). It is well known that BN may react with water vapor, even at room temperature, depending on the purity and crystallinity of the BN.26,27 However, there is recent evidence that BN can be consumed at intermediate temperatures in humid environ￾ments, depending on the crystallinity.28 Fig. 9. Higher-magnification micrographs of PBN-HN minicompos￾ite that failed at 1200°C after 36 h (Fig. 5(c)). EDS analysis indicates that the fracture surface of the bright fibers in the middle of Fig. 9(a) are oxidized; i.e., the fibers had failed prior to minicomposite failure. August 1997 Tensile Stress Rupture of SiCf/SiCm Minicomposites with C and BN Interphases at Elevated Temperatures in Air 2037