T Nozawa et aL /Joumal of Nuclear Materials 384(2009)195-211 Fig. 13. Typical tensile fracture surfaces of multilayer interphase composites. that interfacial shear properties are significantly improved when the bonded F/M interface case 53 The matrix fracture energy is applying advanced Sic fibers with rough surface 50 or modified unaffected or slightly increased by irradiation 4, 54, resulting in SiC fibers with surface treatment [51. From Fig 8, it is no doubt minor contribution for the pls decrease. In contrast, the interfacial that the frictional stress of multilayer composites at 800C up to debond energy closely related to the interfacial debond shear 7.7 dpa decreased remarkably due to the fact that the crack path strength, varies by irradiation and this would therefore be a possi changed from the rough surface within Pyc to the smooth fiber ble contributor. Recent updates in Tables 3 and 5 provide a quali- surface. However, this data point is quite questionable because tatively good correlation among the interfacial debond shear there is no reasonable answer whether it is actually a phenomenon strength and the Pls. Additionally, the reportedly higher CtE of induced by neutron irradiation by considering scarcity of data and Hi-Nicalon Type-s as compared to CvD-SiC may cause an intrinsic by simply extrapolating the data trend. Also, no clear mechanism thermally-induced residual stress, providing a misfit stress at the has yet been proven. It is necessary to clarify the detailed crack interface(Fig 9). A thermally-induced compressive axial stress in propagation behavior to conclude it. the matrix theoretically increases PLS. Under this condition, irradi Another important point of discussion is the observed deterio- ation creep should eliminate this stress to compensate it. Thus, ration of interfacial shear properties by high-temperature neutron irradiation creep may decrease PLs, ignoring the differential axial irradiation From Fig. 12, no significant difference of irradiation- in- swelling. This is qualitatively in good agreement with the results. duced residual stresses was generally expected with the change of Potentially, the irradiation-induced residual stress( Fig. 12)might irradiation temperatures. However, in Fig. 7, it is apparent that increase the PLS. This is, however, quite opposite from the actual irradiation-induced deterioration of interfacial shear stresses en- trend. Stress relaxation by irradiation creep may also affect the re- hanced at >1000C. According to the finding by Kondo et al. 52, sults even in this case. The Pls also depends on Youngs moduli of this temperature(1050 C)agrees well with the temperature at the fiber, matrix, and composites. The change of Young,'s moduli of which irradiation-induced voids can be detected, however, it is Sic by irradiation is, however, statistically very small at high-tem proven that the void contribution to swelling is negligibly sn peratures x1000C (4 and this effect would be negligible. under this condition. Accordingly, no deterioration of pure Sic The effect of neutron irradiation on interfacial shear properties was identified [4. In contrast, there is a report that the strength was also discussed from the tensile unloading-reloading hysteresis of bare Hi-Nicalon Type-S fiber itself deteriorates by irradiation loop analysis for the same-grade composites [55 In that study, en- near 1000C due probably to the presence of uncertain impurities larged hysteresis loop width first implied a probable decrease of identified by mass analysis 41. However, there is also evidence interfacial friction by irradiation. However, considering the fact that the fiber strength was retained during irradiation in composite that push-out tests in this study show no significant deterioration ples [ 41]. The fiber deterioration by irradiation would there- of interfacial friction up to 7.7 dpa at 800C for monolayer com- fore be unlikely. Alternatively, irradiation creep, which has re- posites [13]. it was concluded that this mechanism is insufficient cently been seen for Sic [43], may influence the irradiation to explain the phenomenon completely. Alternatively, it is believed behavior of SiC/SiC composites. Relaxation of residual stresses by that the hysteresis loop widening is due primarily to the difference irradiation creep would be greater in case of lower temperature in matrix crack density The higher matrix crack density for irradi irradiation. Otherwise, deterioration of the Pyc interphase itself ated samples gives the wide hysteresis loop. The matrix crack den- at higher temperature irradiation is possible but there is no evi- sity increases with decreasing interfacial debond shear strength. ence for this at present. For multilayer composites, because sufficient hysteresis informa- tion could not be provided due to the extremely low strain after 4.3. Effect of interfacial shear properties on tensile properties matrix cracking, much discussion could not be done based on that. It is well known that a weak interface leads to lower composite The composites'overall performance relies on the F/M interfa- strengths since some fibers are never loaded to their full capacity al properties, as well as fiber strength itself. In particular for uni- and that some fibers fail prior to achieving full fracture strength directional composites, the proportional limit stress can be if their interface is too strong. The fracture strength is, therefore, described as a function of the fiber volume fraction, an interfacial dependent on the interfacial shear stresses as well as the fiber debond energy, a matrix fracture energy and a misfit stress for strength. This is reasonable based on the analytical result by Curtinthat interfacial shear properties are significantly improved when applying advanced SiC fibers with rough surface [50] or modified SiC fibers with surface treatment [51]. From Fig. 8, it is no doubt that the frictional stress of multilayer composites at 800 C up to 7.7 dpa decreased remarkably due to the fact that the crack path changed from the rough surface within PyC to the smooth fiber surface. However, this data point is quite questionable because there is no reasonable answer whether it is actually a phenomenon induced by neutron irradiation by considering scarcity of data and by simply extrapolating the data trend. Also, no clear mechanism has yet been proven. It is necessary to clarify the detailed crack propagation behavior to conclude it. Another important point of discussion is the observed deterio￾ration of interfacial shear properties by high-temperature neutron irradiation. From Fig. 12, no significant difference of irradiation-in￾duced residual stresses was generally expected with the change of irradiation temperatures. However, in Fig. 7, it is apparent that irradiation-induced deterioration of interfacial shear stresses en￾hanced at >1000 C. According to the finding by Kondo et al. [52], this temperature (1050 C) agrees well with the temperature at which irradiation-induced voids can be detected, however, it is proven that the void contribution to swelling is negligibly small under this condition. Accordingly, no deterioration of pure SiC was identified [4]. In contrast, there is a report that the strength of bare Hi-NicalonTM Type-S fiber itself deteriorates by irradiation near 1000 C due probably to the presence of uncertain impurities identified by mass analysis [41]. However, there is also evidence that the fiber strength was retained during irradiation in composite samples [41]. The fiber deterioration by irradiation would there￾fore be unlikely. Alternatively, irradiation creep, which has re￾cently been seen for SiC [43], may influence the irradiation behavior of SiC/SiC composites. Relaxation of residual stresses by irradiation creep would be greater in case of lower temperature irradiation. Otherwise, deterioration of the PyC interphase itself at higher temperature irradiation is possible but there is no evi￾dence for this at present. 4.3. Effect of interfacial shear properties on tensile properties The composites’ overall performance relies on the F/M interfa￾cial properties, as well as fiber strength itself. In particular for uni￾directional composites, the proportional limit stress can be described as a function of the fiber volume fraction, an interfacial debond energy, a matrix fracture energy, and a misfit stress for the bonded F/M interface case [53]. The matrix fracture energy is unaffected or slightly increased by irradiation [4,54], resulting in minor contribution for the PLS decrease. In contrast, the interfacial debond energy, closely related to the interfacial debond shear strength, varies by irradiation and this would therefore be a possi￾ble contributor. Recent updates in Tables 3 and 5 provide a quali￾tatively good correlation among the interfacial debond shear strength and the PLS. Additionally, the reportedly higher CTE of Hi-NicalonTM Type-S as compared to CVD-SiC may cause an intrinsic thermally-induced residual stress, providing a misfit stress at the interface (Fig. 9). A thermally-induced compressive axial stress in the matrix theoretically increases PLS. Under this condition, irradi￾ation creep should eliminate this stress to compensate it. Thus, irradiation creep may decrease PLS, ignoring the differential axial swelling. This is qualitatively in good agreement with the results. Potentially, the irradiation-induced residual stress (Fig. 12) might increase the PLS. This is, however, quite opposite from the actual trend. Stress relaxation by irradiation creep may also affect the re￾sults even in this case. The PLS also depends on Young’s moduli of the fiber, matrix, and composites. The change of Young’s moduli of SiC by irradiation is, however, statistically very small at high-tem￾peratures 1000 C [4] and this effect would be negligible. The effect of neutron irradiation on interfacial shear properties was also discussed from the tensile unloading–reloading hysteresis loop analysis for the same-grade composites [55]. In that study, en￾larged hysteresis loop width first implied a probable decrease of interfacial friction by irradiation. However, considering the fact that push-out tests in this study show no significant deterioration of interfacial friction up to 7.7 dpa at 800 C for monolayer com￾posites [13], it was concluded that this mechanism is insufficient to explain the phenomenon completely. Alternatively, it is believed that the hysteresis loop widening is due primarily to the difference in matrix crack density. The higher matrix crack density for irradi￾ated samples gives the wide hysteresis loop. The matrix crack den￾sity increases with decreasing interfacial debond shear strength. For multilayer composites, because sufficient hysteresis informa￾tion could not be provided due to the extremely low strain after matrix cracking, much discussion could not be done based on that. It is well known that a weak interface leads to lower composite strengths since some fibers are never loaded to their full capacity and that some fibers fail prior to achieving full fracture strength if their interface is too strong. The fracture strength is, therefore, dependent on the interfacial shear stresses as well as the fiber strength. This is reasonable based on the analytical result by Curtin Fig. 13. Typical tensile fracture surfaces of multilayer interphase composites. 208 T. Nozawa et al. / Journal of Nuclear Materials 384 (2009) 195–211