Explicit control of 2D and 3D structural complexity by discrete variable topology optimization method

• Published in: Computer methods in applied mechanics and engineering 2022-02, Vol.389, p.114302, Article 114302
• Main Authors: Liang, Yuan, Yan, XinYu, Cheng, GengDong
• Format: Article
• Language: English
• Subjects: Algorithms; Apexes; Complexity; Density; Design optimization; Euler and Betti number; Formulations; Graph theory; Integer programming; Manufacturability; Optimization; Redundancy; Structural complexity control; Topology invariants; Topology optimization
• ISSN: 0045-7825; 1879-2138
• DOI: 10.1016/j.cma.2021.114302

The structural complexity (the number of holes) of the 2D or 3D continuum structures can be measured by their topology invariants (i.e., Euler and Betti numbers). Controlling the 2D and 3D structural complexity is significant in topology optimization design because of the various consideration, including manufacturability and necessary structural redundancy, but remains a challenging subject. In this paper, we propose a programmable Euler–Poincaré formula to efficiently calculate the Euler and Betti numbers for the 0–1 pixel-based structures. This programmable Euler–Poincaré​ formula only relates to the nodal density and nodal characteristic vector that represents the nodal neighbor relation so that it avoids manually counting the information of the vertices, edges, and planes on the surfaces of the structure. As a result, the explicit formulations between the structural complexity (the number of holes) and the discrete density design variables for 2D and 3D continuum structures can be efficiently constructed. Furthermore, the discrete variable sensitivity of the structural complexity is calculated through the programmable Euler–Poincaré formula so that the structural complexity control problem can be efficiently and mathematically solved by Sequential Approximate Integer Programming and Canonical relaxation algorithm Various 2D and complicated 3D numerical examples are presented to demonstrate the effectiveness of the method. We further believe that this study bridges the gap between structural topology optimization and mathematical topology analysis, which is much expected in the structural optimization community.