A sensitivity analysis by using design of experiment and its application in stamping

Abstract

There are many manufacturing problems when considering sheet metal stamping parts. These may occur due to complexity of product geometry or formability related concerns such as tearing, wrinkling etc. Shape distortion problems like springback is another issue encountered during production. The design of sheet metal part and related process conditions are in fact a complete system involving many parameters that need to be chosen and should be optimized. This requires defining an approach and its systematic application in order to help solving real industrial stampings. One of the approaches in performing this task is to conduct a sensitivity analysis of a set of design/process parameters. This is usually done using design of experiments (DOE), and the most important sets are identified and represented with response surface type meta-models. Then, the optimum parameters are chosen by weighting their effects on the design/process. In this study, this methodology is applied to reduce the springback deformation of an industrial stamping part and its possible improvements using practical process parameters such as forming forces. Comparisons with stamping experiments were also conducted and results show the applicability of this approach in practice. Furthermore, a moderate amount of springback is achieved.