Experimental determination of the carbon solubility limit in ferritic steels

Despite the existence of a number of published results, the data on the solubility of carbon in alpha iron are still inaccurate. An analysis of published experimental results shows that available values vary greatly (between 50 and 100 ppm by wt, for example, at 600 °C). These discrepancies make it difficult to optimize the metallurgical processes of low-carbon or ultralow-carbon alloys. An experimental methodology, using the measurement of the thermo-electric power (TEP) of the alloy, was set up. This enabled us to deduce the quantity of free interstitials in the matrix by measuring the amount of interstitials which segregate on dislocations after a deformation of the sample. This technique was used in the case of an Al-killed steel containing 0.2 pct Mn. The limit of solubility of carbon was determined with a precision of ±2 ppm between 550 °C and 730 °C. This limit of solubility can be analytically described by the relation C_{(wt pct)}=6.63 exp (−11.8_kcal·mol ⁻¹/RT), which is shown to be valid only for temperatures above 400 °C. We show experimentally that the residual concentration of carbon at low temperature is much greater than the value predicted by the extrapolation of this relation. Complementary studies on steels with various C and Mn contents allow us to verify the validity of the proposed methodology.