Modelling of Multi-Contact Sliding Wear using Frictional Noise

Abstract

Friction-induced vibrations in sliding contacts generate wear, a phenomenon where relative motion between surfaces cause material degradation. This degradation is primarily due to asperities on contacting surfaces that collide and deform, leading to the wearing out of the surfaces. The contact of surfaces generate both wear and the emission of frictional noise. This noise, often characterized by its frequency and amplitude, serves as an indicator of the wear severity and the dynamic behavior of the contact surfaces. Thus, establishing a direct dependence on the two factors. The occurrence of frictional noise due to wear is also an indicative of stick-slip condition in surface contact scenario. Stick-slip occurs when two colliding surfaces alternately stick and slide over each other. In the context of predictive maintenance, wear and friction sound generated at the contact can lead to interesting revelations about the type of wear, and its intensity, thus leading to comprehensive analysis of potential component failure. This research proposes a simplistic approach towards estimating incremental wear in a multi-contact scenario using a vibrational analysis approach and in turn goes a step forward to model its associated sound. The proposed model is based on Hertzian contact model and Archard wear principle, in order to predict wear depth and its associated frictional noise. Furthermore, experimental validation is performed to confirm the results obtained from the analytical model. Predicted wear depth and frictional sound are compared to the experimental values obtained using a standardized pin-on-disc tribometer setup affixed with a free-field microphone to capture airborne noise. By iii estimating incremental wear and associated noise, the research aids in predictive maintenance strategies, allowing for early detection of potential component failures in machines based on the analysis of friction-induced vibrations and noise. From the results, it can be concluded that the predictive model is a good estimator of wear and frictional noise for hard materials whereas an increasing error is observed in terms of softer materials, like Aluminum. The results show good conformity between the proposed analytical model values and the standardized experiments, hence ensuing that within certain limitations, the proposed model and the intended approach can effectively be used as a good estimator of wear and its sound in a multi-contact scenario.