Gearbox diagnosis using a modified Fourier series

The vibration signals measured from a gearbox are complex multi-component signals, which are often non-stationary and time varying. When conducting gearbox failure detection, sometimes it is not simple to interpret the results coming from traditional Fourier analysis and other methods such as wavelet analysis, especially when the gear has an incipient failure. In this paper, a modified Fourier series (MFS) is proposed. It is based on an improvement of the traditional Fourier series analysis and comes from the analysis of the gear meshing vibration signal model. Its principle is firstly introduced, and then an algorithm for calculating its coefficients is presented. In the case of truck gearbox vibration diagnosis, the proposed method is shown to be satiating.