U-shaped Part Springback Forecasting Based on the Finite Element Method and the Neural Network

In the paper, finite element simulation parameters settings of u-shaped part bending were expounded and Dynaform was employed as simulating tool in study. Given the material parameters, orthogonal experiments in which the springback angle was the evaluating target were used, and different results under the conditions of different Blank Holding Force, Friction Coefficient, the Gap between Moulds and Radius of Moulds were given. Then an artificial neural network which can be used to train and later to predict the springback was build. The results of research shown, U-shaped part springback forecasting method basing on the neural network would be a new way of plate stamping process and mould design, and it was also having great importance to shorten mould design cycle, optimize the molding process, reduce the cost and improve the quality of stamping.