Analyzing periodic signals in rotating pin-on-disc tribotest measurements using discrete fourier transform algorithm

In rotating pin-on-disc tribometer testing, friction coefficient data often contains various periodic components. The periodic oscillation in the measured signal can be caused by unleveled sample mounting or inhomogeneous sample surfaces along the measurement track. In any case, the periodic components need to be separated from the constant or stochastic friction signals in order to observe either the fine features of the friction coefficient or the evolution of wear characteristics. For this purpose, the discrete Fourier transform (DFT) algorithm can be used to process the data and separate the periodic components from the raw friction signal. This study demonstrates how the DFT method can enhance the analysis of pin-on-disc friction coefficient data to eliminate the artifact due to uneven sample mounting, detect the sample inhomogeneity, and follow the wear of the surface. The DFT method works much better in separating and removing the periodic noises than the commonly used box smoothing method.