Estimation of flow curve and friction coefficient by means of a one-step ring test using a neural network coupled with FE simulations

This paper is concerned with application of artificial neural network (ANN) to the ring compression test for simultaneous determination of the flow curve of the material and the friction factor. The developed ANN model was trained using data from 700 finite element (FE) simulations of the ring test. The load curve of this test and the final internal diameter of the sample are the inputs for this ANN model and the outputs are the strength coefficient, strain hardening exponent and the friction factor. It was found that the outputs of the developed ANN model were in good agreement with the experimental results