Determination of a coefficient of thermal expansion by machine learning

Objective of this work is to model the thermal expansion coefficients of selected steel grade and compare results with those measured by TMA method. Coefficient of thermal expansion is described as a function of steel composition (C, Mn, P, S, Si, Cr, Ni, Mo) and temperature.Experimental values are described and compared with model. Correlation analysis of these data sets is done. Presented model is based on using artificial neural network and represents a preliminary test of method capability to be used for such problems class – for predicting of thermophysical properties depending on composition and temperatre.