Pore classification and performance prediction of 2D carbon fiber reinforced silicon carbide composites

To facilitate more precise and rapid pore modeling, it is crucial to determine the critical volume that delineates between large and small pores. In this research, micro-CT technology and three-dimensional reconstruction technique were employed to extract and reconstruct the microstructure within 2D C_f/SiC composites. In terms of pore modeling, a novel approach was introduced to avoid pore overlap and fiber bundle obstruction, and established corresponding unit cell models using two classification criteria: Pore volume and aspect ratio (length-to-width ratio). By utilizing finite element analysis, the volume threshold for large and small pores was defined. According to the volume classification, pores with volumes less than 0.04 mm³ exert a minimal influence on the mechanical properties of C_f/SiC unit cells. Furthermore, based on the aspect ratio classification, it is founded that the shape of pores with volumes less than 0.021 mm³ has a negligible impact on material properties. Taking 0.021 mm³ as the critical volume, larger than the critical volume is modelled accurately, and smaller than the critical volume is modelled randomly based on the statistical data, the macroscopic modulus of the tensile specimen is calculated to be 110 GPa, with an error of 2.7% from the experimentally obtained 113 GPa, which proves that this partitioning method can increase the efficiency of the modelling while maintaining the accuracy. The effects of volume element selection and fibre bundle corrugation ratio on the modulus are explored on the basis of pore classification, and it is found that the single-cell model is more efficient while maintaining the computational accuracy, whereas an increase in the corrugation ratio leads to a decrease in the modulus of the composite.