Statistical analysis using the multiple regression research in areas of the indefinite chilled cast–iron rolls manufacturing

Abstract

To analyze the metallurgical processes is used, mainly, the statistical fundamental methods that permit to draw conclusions, from the observed values, about the repartition of the frequencies of various parameters, about their interaction, about verification validity of certain premises, and about the research of the dependencies among different parameters. In this sense, the realization of optimum chemical compositions of the cast–iron can constitute a technical efficient way to assure the exploitation properties, the material from which the rolling mills rolls are manufactured having an important role in this sense. This paper reviews key aspects of roll material properties and presents an analysis of the influences of chemical composition upon the mechanical properties of the indefinite cast iron rolls. Now, using the multivariate research, we present some mathematical correlations and graphical interpretations between the hardness and the chemical composition. Using the double and triple correlations variation boundaries for the chemical composition, in view the obtaining the optimal values of the hardness of indefinite cast iron rolls, are obtained. The partial results and evidence obtained by actual experiments are presented. For the multiple regression equations, correlation coefficients and graphical representations the software MATLAB was used.