ABSTRACT

High-resolution topology optimization ensures highly accurate solutions and structural details; however, it also imposes a heavy computational burden. In this study, an efficient topology optimization method based on Cube-Tet mesh is proposed that significantly shortens the duration of finite element analysis (FEA) by compressing the size of the sparse matrix. First, the uniform cubic voxel set is obtained from the model. Each cubic voxel is transformed into six tetrahedral voxels without changing the total number of nodes; then, the FEA is performed. By transforming the mesh style, the coupling terms between nodes are decreased, the difficulty in solving the sparse matrix is reduced, and the iterative solution time is shortened, thus increasing the efficiency of topology optimization.

摘要

高分辨率拓扑优化确保高精度解决方案和结构细节；然而，它也带来了沉重的计算负担。在本研究中，提出了一种基于 Cube-Tet 网格的高效拓扑优化方法，通过压缩稀疏矩阵的大小，显着缩短了有限元分析 (FEA) 的持续时间。首先，从模型中获得均匀立方体素集。在不改变节点总数的情况下，将每个立方体素转化为六个四面体体素；然后，执行 FEA。通过网格样式的变换，减少了节点之间的耦合项，降低了稀疏矩阵的求解难度，缩短了迭代求解时间，从而提高了拓扑优化的效率。