Proposal of a Prediction Model for Product Shape in Sequential Press Forming and Its Experimental Validation

Abstract

Sequential press forming is a method used to form a workpiece while changing its location and to manufacture large structures, such as storage tanks. To optimize the sequential press forming conditions, accurate and efficient predictions of product shapes using numerical simulations are required. However, such predictions are currently difficult because the computational cost is large when sequential press forming processes are simulated. In this study, a new prediction model for product shape in sequential press forming was proposed. The detailed ideas are as follows. First, a machine learning model that predicts a product shape formed by a press forming process was developed. The model represents the product shape using a radial basis function network. Then, the predicted results were used as the input data for the next press forming process, and this procedure was repeated to predict the entire sequential press forming process. The predicted results were compared with experimental results, and it was confirmed that the predicted results had good accuracy. Because this procedure allows the prediction of sequential processes by using the machine learning model for a press forming process, these results showed that the proposed model enables efficient predictions of product shape in sequential press forming.