X-ray tomography-assisted modeling of unidirectional mini SiC_f/SiC composites considering microstructure defects

SiC fiber reinforced SiC matrix composites (SiC_f/SiC) are prime candidate materials in aerospace due to their excellent properties. For SiC_f/SiC composites, high-precision geometric models which reveal their micro-structures are the key to predicting their mechanical properties. This work conducted extensive CT tests of unidirectional mini SiC_f/SiC composites to gain insights into their micro-structural details after the CVD process. Three primary defects: fiber misaligning, fiber oval-shaped features and voids, were observed and statistically analyzed comprehensively. A finite element mesh reconstruction method and voids implantation method: Spatial Mesh Mapping (SPAMM) were proposed based on the CT images for the mini SiC_f/SiC composite. A neural network model was tailored to speed up the CT image segmentation process. Subsequently, an in-situ test was conducted on a unidirectional mini SiC_f/SiC specimen to verify the accuracy of the modeling approach. The modeling and simulation approach provides new insights into damage and failure analyses of ceramic-matrix composites.