Use of artificial neural network for prediction of ion nitrided case depth in Fe–Cr alloys

In this work, a simple artificial neural network (ANN) model using back-propagation training algorithm for ion nitriding behaviour of Fe–Cr alloys was established. The case depth data were extracted from experimental data and used in the formation of training sets of ANN in order to predict case depth of ion nitrided Fe–Cr alloys, 2.5% Cr intervals for 5–20% Cr. The modelling results confirm the feasibility of this approach and show good agreement with experimental data by Alves et al. (Mater Sci Eng 2002; 279A: 10–15) with high accuracy. A contour diagram as a function of Cr (wt.%) and ion nitriding time for Fe–Cr alloy was constructed for industrial application. It is concluded that a considerable saving in terms of cost and time could be obtained from using the trained ANN model and, it provides more useful data from relatively small experimental databases.