Evaluation of Tube Formability in Hydroforming by Artificial Neural Network

During tube hydroforming, formability of stainless steel tube is often obtained by experiment or FEM simulation. In this paper, a back-propagation artificial neural network (BP-ANN) model is built to evaluate material formability in tube hydroforming. The comparison of experiment results and evaluation results indicates that the proposed ANN can accurately evaluate material formability. In the post optimization for hydroforming parameter, this proposed can be used to replace FEM simulation, and a lot of time should be saved in the search for the optimal solution. This method is also applied to predict formability of other material and different type part.