Comparisons between algorithms for structural topology optimization using a series of benchmark studies

Abstract

Using a series of benchmark studies this paper compares algorithms for structural topology optimization based on (i) the artificial material model and (ii) a special microstructure with square voids. The new `constrained adaptive topology optimization' or CATO algorithm combines ideas from the two main groups of procedures which have emerged in topology optimization: the more mathematically rigorous homogenization or h-methods and the more intuitive evolutionary or e-methods. The CATO algorithm is compared with an algorithm which is a development of an h-method which uses an artificial material model as well as a variation of the original classic microcell model. The discriminating benchmark problems are selected to determine the robustness of the two methods with respect to (a) discretization aspects such as finite element mesh refinement and mesh distortion (b) solution parameters such as non-uniform starting density distributions.