Neural Network Model for Isothermal Pearlite Transformation. Part I: Interlamellar Spacing

The present paper is the first of a two-part paper which deals with a neural network model to describe the isothermal pearlite formation. The isothermal austenite-to-pearlite transformation has been analyzed using a neural network technique within a Bayesian framework. In this framework, the pearlite interlamellar spacing and growth rate of pearlite can be represented as a general empirical function of variables such as Mn, Cr, Ni, Si and Mo alloying contents and temperature which are of great importance for the pearlite growth mechanisms. The method has limitations owing to its empirical character, but it has been demonstrated that it can be used in such way that the predicted trends make metallurgical sense. In this first part paper, the method has been used to examine the relative importance of the alloying elements on pearlite interlamellar spacing.