Abstract
	Program ENERGY 4.0, developed for calculation of parameters describing dynamic recrystallization (DRX), includes now two ways of methodology of calculation. Input data coming from the torsion plastometer SETARAM - VÍTKOVICE is first processed by a conventional method of successive linear regressions. In this stage the exclusion of temperature levels is carried out, in which a state featured by one phase of tested steels is not maintained. The following automatic computation using the simplex algorithm of non-linear regression can make particular parameters for a description of DRX more precise. Those calculations are entirely objective and yield in the least statistical errors but their results can not always match up the physical-based idea. In total 22 steels of different chemical composition - from simple structural steels, through microalloyed steels of various types, high alloyed tool steels up to corrosion-resisting (stainless) austenitic steels - were included in the analyzed data set. A key (decisive) criterion for including in an evaluated group of steels was maintaining of mono-phase structural state in the whole range of evaluated temperatures of testing. One can compare influence of both methods on calculated parameters, as well as accuracy of both methods in determination of values of the maximum stress and peak to strain.