Statistical analysis of heats with targeted overheating realised in the EAF at Třinec steelworks

Abstract

Data sets were obtained from series of heats of tool and bearing steels produced in electric arc furnace (EAF) with subsequent treatment in VOD unit and targeted change (setting) of higher overheating of steel above the liquidus by chemical heating. The data can be divided to input, process and output quality parameters, which is typical division for so called qualimetrics. The paper gives a short overview with description of use of methods of multi-dimensional statistical analysis (exploratory analysis, analysis of time series, correlation and regression, multi-dimensional methods, PLS – Partial Least Squares) for analysis of parameters and links between quality indices for produced steel and input and set process parameters for production of steel in the EAF.