A new empirical formula for the bainite upper temperature limit of steel

Abstract

The definition of the practical upper temperature limit of the bainite reaction in steels is discussed. Because the theoretical upper temperature limit of bainite reaction, B 0, can neither be obtained directly from experimental measurements, nor from calculations, then, different models related to the practical upper temperature limit of bainite reaction, B S, are reviewed and analyzed first in order to define the B 0 temperature. A new physical significance of the B S and B 0 temperatures in steels is proposed and analyzed. It is found that the B 0 temperature of the bainite reaction in steels can be defined by the point of intersection between the thermodynamic equilibrium curve for the austenite→ferrite transformation by coherent growth (curve Z $$\gamma \to \overrightarrow \alpha $$ ) and the extrapolated thermodynamic equilibrium curve for the austenite→cementite transformation (curve ES in the Fe-C phase diagram). The B S temperature for the bainite reaction is about 50–55 °C lower than the B 0 temperature in steels. Using this method, the B 0 and B S temperatures for plain carbon steels were found to be 680 °C and 630 °C, respectively. The bainite reaction can only be observed below 500 °C because it is obscured by the pearlite reaction which occurs prior to the bainite reaction in plain carbon steels. A new formula, B S(°C) =, 630-45Mn-40V-35Si-30Cr-25Mo-20Ni-15W, is proposed to predict the B S temperature of steel. The effect of steel composition on the B S temperature is discussed. It is shown that B S is mainly affected by alloying elements other than carbon, which had been found in previous investigations. The new formula gives a better agreement with experimental results than for 3 other empirical formulae when data from 82 low alloy steels from were examined. For more than 70% of these low alloy steels, the B S temperatures can be predicted by this new formula within ±25°C. It is believed that the new equation will have more extensive applicability than existing equations since it is based on data for a wide range of steel compositions (7 alloying elements).