Designing of the chemical composition of steels basing on the hardenability of constructional steels

The paper presents the original method of modelling of the relationships between chemical composition of alloy constructional steel and its hardenability, employing neural networks. Basing on the experimental results of the hardenability investigations, which employed Jominy method, the model of the neural networks was developed and fully verified experimentally. The model makes it possible to obtain Jominy hardenability curves basing on the steel chemical composition. The model of neural networks, making it possible to design the steel chemical composition, basing on the known Jominy hardenability curve shape, was developed also and fully verified numerically. (author)