ANN constitutive model for high strain-rate deformation of Al 7075-T6

Abstract

An artificial neural network (ANN) constitutive model was developed for Al 7075-T6 based on flow data found in the literature and orthogonal machining tests. The use of orthogonal machining data allowed the ANN network to be trained and tested at high strain-rates of deformation common in machining operations. A new ANN method of network construction (training and validation) was successfully applied to the sparse high strain-rate regime. The method of training and validation, 0.632e stop training method, requires less experimentation to determine network parameters and makes the most efficient use of scarce data. The ANN predictions at high strain-rates where compared with and shown to be superior to a parametric constitutive model.