Multi-factor regression analysis of the process of rails contact welding on K1000 machine

20th International Scientific and Research Conference "Metallurgy: Technologies, Innovation, Quality. Metallurgy-2017"

Shevchenko, R.A.^1 ; Kozyrev, N.A.^1 ; Kryukov, R.E.^1 ; Patrushev, A.O.^1 ; Usoltsev, A.A.^1

Siberian State Industrial University, 42 Kirova street, Novokuznetsk

654007, Russia^1

关键词: Functional dependence;    Multiple regression analysis;    Production test;    Railroad rails;    Regression model;    Technological parameters;    Transverse bending;    Welding parameters;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/411/1/012088/pdf
DOI  :  10.1088/1757-899X/411/1/012088

来源: IOP

【 摘 要 】

The article describes the technology of welding railroad rails with an electrocontact method by a pulsating reflow on a K1000 machine and production tests for a static three-point transverse bending. The analysis of the technology of welding rail rails on K1000 machine made it possible to determine the parameters of the welding mode and to reveal the essential (determining) characteristics of the technology for welding rails. The evaluation of the influence of the welding parameters on the value of the flexural force Pbend (kN) and the value of the deflection fdef (mm) was carried out by means of multiple regression analysis, which makes it possible to study the regularities of the change in the resultant index depending on the behavior of various factors. Based on the available data, a mathematical model of the process of contact welding of rails was created. The functional dependence of the influence of technological parameters on the force Pbend. Using the regression model, it is proposed to predict the quality of the welded seam and control the technological parameters of the process of contact welding of rails.