Application of Artificial Neural Network for Prediction of Heat Treated Sintered Steels Properties
Heat treatment is an important method for improving the mechanical properties of industrial parts that are made through the powder metallurgy. Most PM steels are subjected to hardening and tempering, and it is due to this treatment that tempered martensite is formed. After heat treatment, these steel’s mechanical properties are affected by the heat treatment parameters and the initial density. In this paper, in order to make an evaluation of the effect of the above parameters, FN-0205 PM steel with various densities is heat treated in different austenite conditions and tempering time. Their mechanical properties are then evaluated and recorded. Afterwards, this data obtained by experimental procedure are predicted for various conditions. The method employed here is the well-known feedforward Artificial Neural Network (ANN) with the Back Propagation (BP) learning algorithm. Comparison between predicted values and experimental data, in the present study, indicate that the predicted results from this model are in good agreement with the experimental values.
Andreas Öchsner and Graeme E. Murch
H. Khorsand et al., "Application of Artificial Neural Network for Prediction of Heat Treated Sintered Steels Properties", Defect and Diffusion Forum, Vols. 273-276, pp. 323-328, 2008