ABSTRACT
The shape of hot yield stress curves (surfaces) is complicated because of the microstructural changes and phase transformations that occur during the hot working process. Such dependencies are difficult to describe mathematically. In this article, modeling of hot yield stress curves for carbon silicon steel (electric steel, 1.87%Si, 0.05–0.08%C) by genetic programming was carried out. Strain rate, strain, and temperature were the parameters monitored experimentally. On the basis of the experimental data, a model for prediction of hot yield stress curves was obtained by genetic programming. The study showed that genetic programming is a powerful prediction tool.