Multi-objective optimization of cellular scanning strategy in selective laser melting
Abstract
The scanning strategy for selective laser melting - an additive manufacturing process - determines the temperature fields during the manufacturing process, which in turn affects residual stresses and distortions, two of the main sources of process-induced defects. The goal of this study is to develop a multi-objective approach to optimize the cellular scanning strategy such that the two aforementioned defects are minimized. The decision variable in the chosen problem is a combination of the sequence in which cells are processed and one of six scanning strategies applied to each cell. Thus, the problem is a combination of combinatorial and choice optimization, which makes the problem difficult to solve. On a process simulation domain consisting of 32 cells, our multi-objective evolutionary method is able to find a set of trade-off solutions for the defined conflicting objectives, which cannot be obtained by performing merely a local search. Possible similarities in Pareto-optimal solutions are explored.
Authors: Ali Ahrari, Kalyanmoy Deb, Sankhya Mohanty, Jesper Henri Hattel
Published in: IEEE Congress on Evolutionary Computation (CEC) (2017)