Neural network prediction of flow stress of Ti–15–3 alloy under hot compression

Hot compression experiments are conducted on Ti–15–3 alloy specimens using a Gleeble-1500 Thermal Simulator. These tests have been focused to obtain flow stress data under varying conditions of strain, strain rate and temperature. High temperature flow characteristics are studied for Ti–15–3 alloy in terms of stress–strain curves and a predicting model for the calculation of flow stress has been established with an artificial neural network method. Results show that the neural network can correctly reproduce the flow stress in the sampled data and it also can predict well the non-sampled data. These studies are significant for determining the hot-forging processing parameters of Ti–15–3 alloy.