Optimum structural topology and shape design using three phase hybrid GA and artificial density method

In this paper an efficient GA based Three Phase Topology Optimization algorithm which is an improvement over the Two-Phase Method developed by the team at UT Delhi is presented. The Solid Isotropic Material with Penalization approach is outlined and Sequential Linear Programming method with move limits is utilized for solving the compliance minimization and volume minimization problems with volume and displacement constraints respectively. The need for parametric studies with various penalty factors and initial starting constraints are emphasized. The availability of accurate constraint derivatives with minimal effort is a boon but as in any NLP problem with a large number of variables, the global nature of the optimum is somewhat suspect in SIMP based method. A detailed comparison of the performances of the two phase and three phase methods and the SIMP method is presented for problems of different size and general conclusions are drawn.