No abstract

No abstract

Silicon carbide (SiC) matrix composite reinforced by both SiC and carbon fibers ([SiC-C]/pyrolytic carbon [PyC]/SiC) was fabricated by chemical vapor infiltration for reducing matrix microcracks. Microstructure, mechanical properties, and oxidation resistance of the composite were compared with those of C/PyC/SiC and C/PyC HT /SiC composites in which PyC interphase was untreated and heat-treated at 18001C in argon, respectively. Compared with the C/PyC/SiC composite, (SiC-C)/PyC/SiC composite shows considerable improvements in flexural strength, fracture toughness, and oxidation resistance. The different mechanical behaviors of the three composites were analyzed based on the He and Hutchinson's model.

Accurate estimation of the permeability of fibrous media is extremely important for designing and optimizing composites processing. Compared with the liquid processing of composites, the researchers on chemical vapor infiltration (CVI) are more concerned with detailed knowledge of the variation in permeability during the densification. Although, many models and theories have been developed to relating the permeability of porous medium to their structural properties, the ones can be directly applied to CVI are very limited. Thus, in this study, we numerically calculate the permeability of six types of fiber or fiber bundle structures in the high and intermediate density domain (porosity is less than 70%) and the applicabilities of Tomadakis's and Gebart's permeability models to CVI are studied. Our study shows that the original form of Tomadakis's model overpredicts the viscous tortuosity (underpredicts the permeability) in most cases, thus his model should be modified by introducing a new parameter, whereas the general form of Gebart's model with three adjustable parameters is able to predict the evolution of permeability fairly well. Furthermore, our prediction is compared well with the reported data from various references not limited to CVI, thus could also be useful for other engineering applications.

No abstract