Modelling the Effect of Alloy Elements

The information about the quantitative effects of alloy elements on the hot strength of steels is scarce and contradictory. Its effects can change according to alloy content and specific conditions of the hot forming process, as they can activate different hardening or softening mechanisms. Three empirical approaches were initially adopted in this work to rank and model hot strength in steels in function of their chemical composition: activation energy of deformation Q, mean hot flow stress and parameters of the Hajduk equation. Both alloy elements effect and steel ranking regarding to hot strength varied according to the selected parameter. Besides that, no simple and precise model could be developed from these parameters. This fact led to the development of a neural network with this purpose; its input parameters were C, Mn, Si, Al, Cr, Cu, Nb, V, Ti and N contents, as well strain temperature, degree and strain rate. This new approach showed better precision than the former trials, making easy to forecast the hot strength behavior of new steel alloys. This solution proved to be simple and very efficient, minimizing the development time of a mathematical model for this purpose.