Computational homogenization of thermomechanical properties for Fully Ceramic Microencapsulated fuel with 3D printed SiC matrix

Fully Ceramic Microencapsulated (FCM) fuel, fabricated using 3D printing technology, incorporates traditional Tristructural-Isotropic (TRISO) fuel particles embedded within a 3D-printed SiC matrix. This fuel type shows great potential for compact, high-temperature nuclear reactors. However, comprehensive performance analysis remains challenging due to the multilayered structure, complex material properties, and the irradiation behavior of numerous TRISO particles within the matrix. To address these challenges, this study aims to apply finite element (FE)-based computational homogenization to evaluate the fuel's thermomechanical properties. First, a detailed finite element analysis (FEA) of a single TRISO particle's performance was conducted, focusing primarily on the Buffer-IPyC gap size change during irradiation, a critical factor influencing the fuel's thermal performance. Second, the homogenization of a single TRISO particle was performed using FEA, accounting for the changes in the Buffer-IPyC gap size. Finally, homogenized TRISO particles were randomly distributed and perfectly bonded within a 3D-printed SiC matrix to form a Representative Volume Element (RVE). The FE homogenization of the RVE was then conducted to derive the effective thermomechanical properties of the FCM fuel. The packing fraction of TRISO particles in the matrix ranged from 10 % to 50 %, with temperature and burnup conditions spanning 800–1600 K and 0–16 % FIMA, respectively. Results show that irradiation significantly affects the effective properties of the fuel, though this impact diminishes as burnup increases. Additionally, the effective thermal conductivity of the FCM fuel decreases with increasing TRISO packing fraction, assuming the thermal conductivity of the SiC matrix exceeds that of TRISO particles. This study provides a valuable reference for the design and optimization of FCM fuel for future nuclear applications.