Artificial Neural Network for Modeling the Tensile Properties of Ferrite-Pearlite Steels: Relative Importance of Alloying Elements and Microstructural Factors

Abstract

An artificial neural network (ANN) model was developed to predict the tensile properties as a function of alloying element and microstructural factor of ferrite-pearlite steels. The input parameters of the model were composed of alloying elements (Mn, Si, Al, Nb, Ti, and V) and microstructural factors (pearlite fraction, ferrite grain size, interlamellar spacing, and cementite thickness), while the output parameters of the model were yield strength and tensile strength. Although the ferrite-pearlite steels have complex relationships among the alloying elements, microstructural factors, and tensile properties, the ANN model predictions were found to be more accurate with experimental results than the existing equation model. In the present study the individual effect of input parameters on the tensile properties was quantitatively estimated with the help of the average index of the relative importance for alloying elements as well as microstructural factors. The ANN model attempted from the metallurgical points of view is expected to be useful for designing new steels having required mechanical properties.