Data-driven Oxidation Kinetic Model Construction of Silicon Carbide Composites

¹ School of Mathematics and Statistics, Northwestern Polytechnical University, Xi’an, 710021, China
² Science and Technology on Thermostructural Composite Materials Laboratory, Northwestern Polytechnical University, Xi’an, 710021, China

^a) Corresponding author: manyuxiao@nwpu.edu.cn;
^b) zwq2023@mail.nwpu.edu.cn;
^c) xgluan@nwpu.edu.cn;
^d) xuxinming@mail.nwpu.edu.cn

Abstract

Silicon carbide fiber-reinforced silicon carbide matrix composites (SiC_f/SiC) exhibit exceptional properties, including lightweight, high strength, superior heat resistance, and high fracture toughness. However, the material's inherent inhomogeneity, nonlinearity, anisotropy, and diverse failure modes make the damage and failure mechanisms of SiC_f/SiC composites highly complex. This research integrates experimental methods with intelligent data analysis techniques to accurately characterize and predict the oxidation behavior and mechanical property evolution of SiC_f/SiC composites under complex service conditions. By leveraging mechanical performance evolution data of SiC_f/SiC in various single environments, along with Gaussian functions and parameter optimization methods, and incorporating the composite material's chemical reaction characteristics, an automatic construction of the oxidation kinetics model for SiC_f/SiC under complex multi-field coupling environments was achieved based on data-driven strategy. The constructed SiC_f/SiC oxidation kinetics model precisely delineates the relationship between the mechanical performance parameters after high-temperature oxidation in different environments and their service time. Furthermore, a complex coupling relationship model was developed that links the mechanical performance parameters with service time and service temperature. Experimental verification confirmed the accuracy and effectiveness of the constructed oxidation kinetics model in predicting the performance evolution of SiC_f/SiC composites under various environmental conditions.