Topology optimization for continuum structure with numerical methods

Abstract

This thesis investigates the use of numerical methods for topology optimization of continuum structures. Topology optimization is a process in which materials are distributed in a way that the structure achieves its best performance. Topology optimization of continuum structures is one of the most-challenging research field of structural optimization. By using numerical methods which is based on finite element analysis, it is able to explore the most innovative and efficient structural forms at free constraint of traditional structural profiles and units. Many research works has been carried out in the last a few decades to search for efficient numerical algorithms this field. Among all, Bi-directional Evolutionary Structural Optimization (BESO) proposed by Xie and Huang (2010) is one of the most popular approach. This method is adopted as the research methodology in this thesis and a chapter is designated for detailed introduction. Two case studies are provided to proof the optimality of BESO design results. Besides, a new technique is proposed by the author to promote the computational efficiency of BESO method, which is named as Progressive Refinement Technique. The modified BESO operates topology optimization with two different finite element models, first one with coarse meshes and second with fine meshes. Adopting the new technique, four case studies were carried out in this thesis. It is proved that with proper use of progressive refinement method, BESO method can significantly reduce computational time without sacrifice of design quality. In the end, extended studies of BESO method are briefly described. Furthermore, several cases of practical application of BESO design are introduced.