The Prediction of Geometrical Configuration and Ductile Fracture Using the Artificial Neural network for a Cold Forged Product

신경망을 이용한 냉간 단조품의 기하학적 형상 및 연성파괴 예측

This paper suggests the scheme to simultaneously accomplish prediction of fracture initiation and geomeytical configuration of deformation in metal forming processes using the artificial neural network. A three-layer neural network is used and a back propagation algorithm is adapted to train the network. The Cookcroft-Lathjam criterion is used to estimate whether fracture occurs during the deformation process. The geometrical configuration and the value of ductile fracture are measured by finite element method. The predictions of neural network and numerical results of simple upsetting are compared. The proposed scheme has successfully predicted the geometrical configuration and fracture initiation.