Computational Method for Heat Partition at the Rail-Armature Interface Based on Least Squares Regression

Abstract

The sliding electrical contact interface will generate a huge total heat source including Joule heat and frictional heat when an armature slides along a pair of rails driven by Lorentz forces. The total heat is partitioned at the contact interface flowing into armature and rail respectively. The heat partition at the rail-armature interface and the resulting temperature rise have important effects on the melt wear characteristics of the contact material. A computational method was proposed for determining heat partition at the rail-armature interface based on least squares regression. Temperatures equations of rail and armature were developed in terms of an unknown heat partition function. Solution equations of heat partition function were derived by matching the average temperature of rail and armature at the contact surfaces. The heat partition function was assumed to be a polynomial form and an estimate for the heat partition function was obtained by minimizing the error term in the least squares sense. Then, effects of polynomial order on the heat partition function are analyzed. Furthermore, the heat partition solution from the least squares regression and previous analytical method are compared. Variation characteristics of heat partition coefficient were also analyzed. In the end, the variation mechanism and influencing factors of heat partition coefficient were discussed.