Abstract
This study proposes a method for integrating finite element analysis and response surface methodology to predict tearing locations in the deep drawing of a rectangular-shape AISI 304 cup. The finite element model was confirmed by a deep-drawing experiment with nitrogengas springs, and the finite element analysis results were in agreement. Moreover, the regression model was developed using response surface methodology to predict final thickness at various locations on the rectangular-shape deep-drawn cup based on 40 finite element cases containing varied blank holder forces. The maximum percentage of thinning values was obtained from the response surface methodology results. The relationships among the blank holder forces and the thinning values at different locations were then further developed. The maps of the predicted tearing locations were then obtained and directly corresponded to the number of locations the specific blank holder forces. These maps, as established, could be directly used in a closed-loop control system enabling tearing defects to be prevented or eliminated.
Wrinkling Prediction in Deep Drawing by Using Response Surface Methodology and Artificial Neural Network