Wear and airborne noise interdependency at asperitical level: analytical modelling and experimental validation.

Generation of wear and airborne sound is inevitable during friction processes. Most correlation between the wear and the sound generated during a sliding process have been experimental. Analytical models do exist, but they remain scarce and do not fully account for the wear and the airborne sound generation especially at asperitical level. The model developed in this research attempts to fill the gap by providing a quantifiable relationship between the wear generated and the sound emitted in a simple pin-on-disc setup. It provides a relationship between the wear and the sound from an asperitical level. This is done by examining the conditions at which wear would occur on an asperity distribution. The asperity distribution is considered to be exponential, although a Gaussian distribution was also considered. Impact forces are calculated on a per-asperity basis and the wear and vibrational displacement is calculated as a result. This leads to the quantification of wear and acoustic noise. The model is validated using a pin-on disc setup for three varied materials (iron (4% carbon content), mild steel (0.18% carbon content) and aluminium T351) under two loads (10 N and 20 N) at 300 RPM. The loads and speeds were chosen so as to observe a range of wear behaviour while remaining within the constraints of the lab limitations and the safety of the force sensors on the tribometer. Temperatures are also examined, and a second set of validation experiment is performed at temperatures of 40 °C and 60 °C. The model computes the predicted wear and sound pressure, and it is compared with the experimental sound pressure measured by the microphone and the wear measured by the tribometer sensors. Sound pressure is chosen as a measure over frequency because it is easier to analyse and compare. The theoretical model agrees with the experimental results with a varying error of 10 to 15 % error in iron and aluminium. However, a larger error is observed in the case of mild steel. The model could be refined to improve the accuracy as it assumes point impacts on the asperities where a distributed impact would be more suitable. Furthermore, the pin is assumed a single asperity to simplify the model at the expense of accuracy. Overall, the experimental results are in fair correlation with the theoretical results and this model provides the first step in quantifying wear using only the recorded sound pressure.