In-situ monitoring of friction surfaces and friction modelling by surface pattern analysis

Abstract

In-situ monitoring of wear and friction surfaces is difficult to achieve because friction occurs between solid surfaces, often even in the presence of a lubricant. To understand a tribological system, it is important to determine how the wear surfaces change over time. In many methods, the surfaces are observed only after tribological tests are completed and the test piece is removed from the tribotester. To overcome these limitations, we have developed a laser-strobe technique that records surface photographs of the same region of the wear surface during each cycle. The recorded photographs were analyzed by pattern matching methods and correlations between the images were determined. Characteristic patterns corresponding to friction track sizes and shapes were identified and used to perform the correlation calculations. A model was also proposed to understand the results of the friction and wear test. The model assumed that the surface consists of small domains and that the friction patterns are generated by the interaction of these domains. Preliminary results indicate that the model effectively describes surface modification due to wear and friction.