Converting pixel‐type topology optimization results to MMC‐representation based on sparse optimization and its applications

Abstract

AbstractHow to realize the switching between various topology optimization approaches such as SIMP and moving morphable component (MMC) method, is a crucial challenge in the field of structural design. In this article, a robust conversion framework is proposed to convert a pixel‐type topology optimization result to MMC representation. Based on the sparse optimization approach, the framework enables the determination of the minimum number of components with a specified shape error. This method provides an efficient bridge for these two types of geometric descriptions, and promotes the free switching between the topology optimization frameworks with pixel‐based and MMC‐based design domains. The proposed procedure contains a pre‐processing of resolution improvement, symmetry axis extraction with sparse optimization, and variational shape approximation. Two practical applications are demonstrated using the proposed framework. First, it can be applied to the intermediate results of SIMP, to achieve faster optimization convergence. Furthermore, a stress‐based shape optimization approach can be applied to the obtained MMCs, and novel progressive continuity constraints are also introduced to maintain boundary continuity. Several examples demonstrate advantages of the proposed framework.