A Projective Transformation-Based Topology Optimization Using Moving Morphable Components

Rixin Wang,Xianmin Zhang,Benliang Zhu

DOI: https://doi.org/10.1016/j.cma.2020.113646

IF: 6.588

2021-01-01

Computer Methods in Applied Mechanics and Engineering

Abstract:The moving morphable components (MMC)-based topology optimization method makes it possible to directly connect theoretical designs to computer-aided design (CAD) systems. To improve the manufacturability of the optimal result, this paper proposes a projective transformation-based topology optimization using moving morphable components (PMMC), which can reduce the performance loss when the optimized design is ported to practical manufacturing. The basic idea is to treat an arbitrary component as a projection of a component template independent of design variables. A hierarchical feature construction method is used to reconstruct the internal details within an element while reducing the interpolation error, in order to achieve a high-precision feature representation and sensitivity analysis that are not affected by the quality of the background mesh. The proposed method not only ensures the consistency of the geometric model and the analysis model, but also flexibly addresses components that are difficult to describe explicitly. The effectiveness of the proposed method is verified through a series of numerical examples.