A voxel-based method of multiscale mechanical property optimization for the design of graded TPMS structures

Abstract Functionally graded triply periodic minimal surface (TPMS) structures are characterized by their various structures and mechanical properties, thereby exhibiting better mechanical performance than uniform structures. At present, most graded TPMS structures in studies possess unidirectional variable densities. However, when performing lattice optimization on an actual engineering model, the required multidirectional graded TPMS structure still relies on manual work, including parameter adjustment and modelling, so the optimization process is cumbersome and the design is suboptimal. In this paper, a new design method is proposed that realizes the stable and adaptive generation and optimization of functionally graded TPMS structures. This procedure is divided into two stages. In the preparation stage, the performance regulation of the TPMS and the values of von Mises stress are obtained from mesoscopic and macroscopic scales, respectively. In the optimization stage, optimization algorithms based on voxels are proposed so that functionally graded TPMS structures can be automatically obtained in different situations. Based on the simulation results, the compressive properties of our optimized design improves by 25.23% compared with that of the uniform structure and 8.63% compared with the functionally graded TPMS structure with topology optimization, which confirms the effectiveness and robustness of the proposed method.