CAE-Based six sigma robust optimization for deep- drawing process of sheet metal

Abstract

Optimization method and numerical simulation technology have been applied in sheet metal forming process to improve design quality and shorten design cycle. However, deterministic optimization may lead to unreliable and non-robust design due to not considering the fluctuation of design variables, environments and operation conditions, etc. In addition to that, iterations in optimization process may cause numerous simulation time or expensive experiment cost. A CAE-based six sigma robust design procedure is developed in this paper with the purpose to eliminate effect of uncertainties in design so as to improve quality. By applying Design of Experiment, Analysis of Variance, and Dual Response Surface Model, this procedure integrates Design for Six Sigma (DFSS), reliability optimization and robust design together to perform optimization. Deep-drawing processes of square cup (NUMISHEET’93) and cylindrical cup (NUMISHEET’2002) are studied as examples, in which thickness variation is taken as the optimization objective, and wrinkle and rupture criteria are taken as constraints. The results show that this design optimization method not only improves significantly the reliability and robustness of the forming quality, but also increases design efficiency with approximate model.