G N Morscher et aL/ Composites Science and Technology 68(2008)3305-3313 No evidence of oxidation embrittlement 110 MPa creep for 2036 h followed by room temper The higher magnification images show no oxidation which is best seen by the presence of the Bl run-out specimen did not exhibit any oxidation crack regions at the fracture surface or along the polished sections. 5 Conclusions Excellent creep-rupture and fatigue properties were demon- strated for the Sylramic-iBN fiber-reinforced melt-infiltrated com- posite system. Two clear regimes of elevated temperature strength degradation were observed: (1)oxidation-induced unbridged crack growth emanating from composite surfaces, most notably a corner of the specimen, and (2) fiber strength degradation probably due to an intrinsic creep-controlled flaw growth mechanism or attack of he fibers from Si diffusion through the Cvi Sic portion of the m (>179 MPa): whereas the second mechanism was dominant at 100um ower stresses(<165 MPa)and longer times. It was demonstrated that the damage development in this site was the growth of matrix cracks and increasing number of matrix cracks with stress and time. The matrix crack densities were very similar to matrix Fig. 14. Micrograph of 165 MPa creep specimen that failed after 1500 h that shows crack densities measured at room temperature on as-produced the presence of some free Si in the 90 and 0 incomposite portions of an bridged crack well away from the fracture surface. specimens. Therefore, the relationships already established for stress-dependent matrix crack density [11 provides a good basis for mechanistic-based models for this material system. However, hance diffusion of free Si and lead to a similar effect. For example, the 1204C matrix crack densities for creep and fatigue specimens Fig. 14 shows evidence of Si diffusion into some of 90 not appear to propagate through the cross-section except for minicomposite microcracks and the unbridged oo the final failure crack. incomposite crack for a specimen that ruptured In order to further improve these composites, higher matrix at 165 MPa cr racking stresses would be required to counter oxidation-induced For the two specimens discussed above that ruptured at 110 unbridged crack growth. This could be achieved with fiber-archl- and 165 MPa after long times, there was a small oxidation-induced tecture modifications if desired [19 In order to improve long unbridged region of the fracture surface. These cracks probably be- time, low stress properties, an improvement in fiber creep rupture gan as 90 tunnel cracks that after long periods of time propagated properties would be required, which has been demonstrated for into a few load-bearing tows and led to a combination of both the further modifications to the Sylramic-iBN fiber[20J external and internal degradation mechanisms. That is, the envi- ronmental synthesis of a small unbridged crack and the internal Acknowledgments degradation of fibers in the bulk of the composite led to rupture or significant strength degradation after long periods of time. It is The Materials Manufacturing Directorate, Air Force interesting to note that the 110 MPa specimens that did not fail Laboratory(aFrlRXL), Wright-Patterson AFB sponsored th during creep for long times and the 165 MPa 30 Hz fatigue under contracts F33615-01-C-5234 and F33615-03-D-2354-1hance diffusion of free Si and lead to a similar effect. For example, Fig. 14 shows evidence of Si diffusion into some regions of 90 minicomposite microcracks and the unbridged region of a 0 minicomposite crack for a specimen that ruptured after 1500 h at 165 MPa creep. For the two specimens discussed above that ruptured at 110 and 165 MPa after long times, there was a small oxidation-induced unbridged region of the fracture surface. These cracks probably be￾gan as 90 tunnel cracks that after long periods of time propagated into a few load-bearing tows and led to a combination of both the external and internal degradation mechanisms. That is, the envi￾ronmental synthesis of a small unbridged crack and the internal degradation of fibers in the bulk of the composite led to rupture or significant strength degradation after long periods of time. It is interesting to note that the 110 MPa specimens that did not fail during creep for long times and the 165 MPa 30 Hz fatigue run-out specimen did not exhibit any oxidation-induced unbridged crack regions at the fracture surface or along the polished sections. 5. Conclusions Excellent creep-rupture and fatigue properties were demon￾strated for the Sylramic-iBN fiber-reinforced melt-infiltrated com￾posite system. Two clear regimes of elevated temperature strength degradation were observed: (1) oxidation-induced unbridged crack growth emanating from composite surfaces, most notably a corner of the specimen, and (2) fiber strength degradation probably due to an intrinsic creep-controlled flaw growth mechanism or attack of the fibers from Si diffusion through the CVI SiC portion of the ma￾trix. The first mechanism was dominant at higher stress conditions (P179 MPa); whereas the second mechanism was dominant at lower stresses (6165 MPa) and longer times. It was demonstrated that the damage development in this composite was the growth of matrix cracks and increasing number of matrix cracks with stress and time. The matrix crack densities were very similar to matrix crack densities measured at room temperature on as-produced specimens. Therefore, the relationships already established for stress-dependent matrix crack density [11] provides a good basis for mechanistic-based models for this material system. However, the 1204 C matrix crack densities for creep and fatigue specimens did not appear to propagate through the cross-section except for the final failure crack. In order to further improve these composites, higher matrix cracking stresses would be required to counter oxidation-induced unbridged crack growth. This could be achieved with fiber-archi￾tecture modifications if desired [19]. In order to improve long￾time, low stress properties, an improvement in fiber creep rupture properties would be required, which has been demonstrated for further modifications to the Sylramic-iBN fiber [20]. Acknowledgments The Materials & Manufacturing Directorate, Air Force Research Laboratory (AFRL/RXL), Wright-Patterson AFB sponsored this work under contracts F33615-01-C-5234 and F33615-03-D-2354-D004. Fig. 13. Fracture surface of 1300-01-006-p02, 110 MPa creep for 2036 h followed by room temperature residual strength test, showing no oxidation at the fracture surface. The higher magnification images show no oxidation which is best seen by the presence of the BN layer around the fibers and no oxidation of SiC matrix. Fig. 14. Micrograph of 165 MPa creep specimen that failed after 1500 h that shows the presence of some free Si in the 90 and 0 minicomposite portions of an unbridged crack well away from the fracture surface. 3312 G.N. Morscher et al. / Composites Science and Technology 68 (2008) 3305–3313