Robust design of a sheet stamping process: approaches to control the inner process variability

Abstract

The robustness investigation has become a crucial issue in the design of the sheet metal forming operations. The uncontrollable effects of some noise variables heavily influence the final process results, often evolving up to the failure of the process design. Therefore, tools able to handle such variability are worth developing with the aim to minimize the scattering of the investigated process performances despite of the process design. Moreover, the stamping process operations are usually characterized by objectives with conflicting behaviors and a set of compromise solutions, referred as Pareto front, is finally obtained instead of a global optimum. Therefore, the question becomes more about how the noise variables effects could influence the obtained Pareto solutions. And accordingly, which is the most robust operative window? In this paper, the stamping of the IFU Hishida part is designed by the calibration of the blank holder force, with the aim to minimize both thinning and wrinkling occurrences. Coil to coil variation of the material properties is accounted for. Two different robust approaches are applied on a multi-objective sheet stamping process design: a hybrid deterministic-stochastic approach and a stochastic framework. The former approach is based on a proper integration among Finite Element simulation, Response Surface Methodology and Monte Carlo simulation. A probability distribution of the final process performances is achieved and a robustness evaluation of an investigated Pareto solution is then performed. Within the second procedure, a Dual Response Surface approach is applied. The set of Pareto solutions is obtained under the influence of the material properties variation. A comparison between the two approaches is finally proposed in order to assess their capabilities in terms of robust design and variability investigation.