Study of the significance of differences of average chemical composition of the melt and its temperature in terms of the LD-process

Abstract

Development of modern metallurgical production and engineer requires the scalene study of process of thermodynamics and kinetics of process of the oxygen converting. Creation of the determined model of the oxygen-converter melting is not possible, a statistical method was therefore used. The aim of the real research was determination of meaningfulness of distinction of mean values of chemical composition and temperature of metallic fusion of converter redistribution with the use of t - criterion for independent selections. For realization of statistical analysis the programmatic package of STATISTICA 8 was used. As a method of verification of statistical hypothesis about equality of middle two selections, the criteria of self-reactance and non-parametric statistics were used. Arrays of data, that analyses, were presented 88th melting from the first converter and 93th melting got on the second converter. It is set a calculation, that diagrams of scope for variables (Т, °С), [C], [S], [P] answer distribution of variables under the law of normal distribution. At the same time distribution of variables [Mn] and a_[0], appraised by the same method, it is fallen short of the law of normal distribution. It is set that for variables (Т, °С) and [C] a hypothesis about equality of dispersions cannot be accepted. Calculation of comparison middle conducted on methodology of values of t - criterion with the separate estimations of dispersions. The results of calculation show that mean values of variable [C] meaningfully differ for fusions of both converters: 0,044 and 0,037% accordingly. At the same time mean values of variable (Т, °С) meaningfully not distinguished. It is shown that hypothesis of equality of dispersions for variables [P] and [S] confirmed: the values of variables belong to the same general aggregates. Values middle for these variables: 0,0079 and 0,0085% for a variable [P], and for a variable [S] 0,0134 and 0,0128% accordingly for the first and second converters and does not differentiate meaningfully. The use of criterion Mann-Whitney for the case of the use of method of non-parametric statistics specifies on absence of meaningful distinctions in values middle for a variable a_[0]. Nevertheless, distinctions middle for a variable [Mn] appeared meaningful