Multivariate research in areas of phosphorus cast-iron brake shoes manufacturing using the statistical analysis and the multiple regression equations

• Published in: IOP conference series. Materials Science and Engineering Vol. 200; no. 1; pp. 12022 - 12030
• Main Authors: Kiss, I, Cioată, V G, Alexa, V, Raţiu, S A
• Format: Journal Article
• Language: English
• Published: Bristol IOP Publishing 01.05.2017
• Subjects: Brake shoes; Cast iron; Chemical composition; Correlation coefficients; Footwear; Freight cars; Graphical representations; Hardness; Iron and steel making; Manufacturing; Multivariate analysis; Phosphorus; Railroad wheels; Regression coefficients; Regression models; Statistical analysis; Subsystems; Variables
• ISSN: 1757-8981; 1757-899X
• DOI: 10.1088/1757-899X/200/1/012022

The braking system is one of the most important and complex subsystems of railway vehicles, especially when it comes for safety. Therefore, installing efficient safe brakes on the modern railway vehicles is essential. Nowadays is devoted attention to solving problems connected with using high performance brake materials and its impact on thermal and mechanical loading of railway wheels. The main factor that influences the selection of a friction material for railway applications is the performance criterion, due to the interaction between the brake block and the wheel produce complex thermos-mechanical phenomena. In this work, the investigated subjects are the cast-iron brake shoes, which are still widely used on freight wagons. Therefore, the cast-iron brake shoes - with lamellar graphite and with a high content of phosphorus (0.8-1.1%) - need a special investigation. In order to establish the optimal condition for the cast-iron brake shoes we proposed a mathematical modelling study by using the statistical analysis and multiple regression equations. Multivariate research is important in areas of cast-iron brake shoes manufacturing, because many variables interact with each other simultaneously. Multivariate visualization comes to the fore when researchers have difficulties in comprehending many dimensions at one time. Technological data (hardness and chemical composition) obtained from cast-iron brake shoes were used for this purpose. In order to settle the multiple correlation between the hardness of the cast-iron brake shoes, and the chemical compositions elements several model of regression equation types has been proposed. Because a three-dimensional surface with variables on three axes is a common way to illustrate multivariate data, in which the maximum and minimum values are easily highlighted, we plotted graphical representation of the regression equations in order to explain interaction of the variables and locate the optimal level of each variable for maximal response. For the calculation of the regression coefficients, dispersion and correlation coefficients, the software Matlab was used.