Numerical study on the effect of mechanical properties variability in sheet metal forming processes

Abstract

This work presents a numerical study on the influence of variability in the mechanical properties of metal sheets in predicting the occurrence of springback and maximum thinning, for a sheet metal forming process. For this purpose, a mild steel was selected as reference material and the U-channel forming process was considered. The factors selected for this variability study were Young’s modulus, the isotropic hardening law parameters, the anisotropy coefficients and the initial thickness of the sheet metal. The variability of all factors was described by a probabilistic normal distribution, with well-defined mean and standard deviation values. Firstly, one-factor-at-a-time, 1/16th fractional factorial and Plackett-Burman screening designs were tested, using FEA results of the U-channel forming process, at two levels of blank-holder force. Sensitivity studies based on analysis of main effects, index of influence and variance (ANOVA) were performed, for each screening design approach, to identify the most relevant factors concerning the occurrence of springback and maximum thinning in the forming process. The comparison between the different screening design approaches indicates that they can lead to similar results, or to the selection of different factors, depending on the sensitivity analysis procedure selected to determine the most relevant factors. Then, metamodels were established using the most relevant factors to build the response surfaces, i.e. analytical relationships between the identified relevant factors and the responses (springback and maximum thinning); consequently, the responses variability can be expeditiously evaluated.