Abstract
Recently published experimental data on the variation of the highest temperature at which Widmanstiittenferrite can be seen to form at a detectable rate, as a function of steel chemistry, are analysed theoretically. It is found that the data can be predicted to a fair accuracy if it is assumed that the nucleation of Widmanstiitten ferrite occurs by a mechanism similar to that of martensitic nucleation, but with the diffusion of carbon during nucleation, and if it is additionally assumed that the growth of Widmanstiitten ferrite can only be sustained when the chemical driving force exceeds a specific stored energy term.
MST/1210