Adhesion in the wheel–rail contact under contaminated conditions

Railway vehicles require a certain level of adhesion between wheel and rail to operate efficiently, reliably, and economically. Different levels of adhesion are needed depending on the vehicle running conditions. In the wheel tread–railhead contact, the dominant problem is low adhesion, as low adhesion on the railhead negatively affects railway operation: on one hand, the vehicle will lose traction resulting in delay when driving on low-adhesion tracks; on the other hand, low adhesion during deceleration will extend the braking distance, which is a safety issue. This thesis examines the influence of several contaminants, i.e., water, oil, and leaves, on the adhesion in the wheel tread–railhead contact. This study will improve our knowledge of the low-adhesion mechanism and of how various contaminants influence adhesion. The thesis consists of a summary overview of the topic and three appended papers (A– C). Papers A and B focus mainly on water and oil contamination examined using two methods, numerical simulation and lab testing. In paper A, real measured wheel and rail surfaces, lowand high-roughness surfaces, along with generated smooth surfaces are used as input to the numerical model for predicting the adhesion coefficient. Waterlubricated, oil-lubricated, and dry contacts are simulated in the model. In the research reported in paper B, scaled testing using a mini traction machine (MTM) was carried out to simulate the wheel–rail contact under lubricated conditions. Two types of disc surfaces of different roughnesses were run at different contact pressures and temperatures. A stylus machine and atomic force microscopy (AFM) were used to measure the surface topography. A study of leaf contamination on the railhead surface, based on field testing, is presented in paper C. Railhead surface samples were cut and the friction coefficient was measured on five occasions over the course of a year. Electron spectroscopy for chemical analysis (ESCA) and glow discharge optical emission spectrometry (GD-OES) were used to detect the chemical composition of the leaf-contamination layer on the