Neural networks-based prediction of hardenability of high performance carburizing steels for automotive applications

The new quenching processes for automotive applications, which follow the cementation stage, include the application of pressurized gas for cooling during quenching. Therefore, it is of utmost importance to have an accurate estimate of the hardenability behavior of carburizing steels, which show a higher Carbon content with respect to traditional materials. These new cooling processes also require properly designed new steels in terms of alloying contents, which ensure a proper response to heat treatment. In the present paper a neural network-based approach to the prediction of the hardenability profile is proposed, which can be applied both for the design of the steel chemistry and for assessing the suitability of the steel at the steel shop level, in order to suitable adjusting the cooling process after quenching