A comparison of cyclostationary and envelope analysis in the diagnostics of rolling element bearings

Some machine signals, while being almost periodic, are not exactly phase-locked to shaft speeds. Typical examples are the impulsive signals from faults in rolling element bearings, which because of slip, are cyclostationary of second order. Thus the spectral correlation diagram has discrete cyclic frequencies and integrating the spectral correlation diagram along the frequency axis produces a discrete frequency spectrum, which has thus been suggested as a simpler way of presenting the data. However, as shown in this paper, this gives the same result as a Fourier transform of the squared signal, which is much more easily produced directly. The result is then very closely related to "envelope analysis" which has long been used in the diagnostics of rolling element bearings. This paper demonstrates that the optimum results are normally obtained by analyzing the squared envelope of an analytic signal (with bandpass filtered one-sided spectrum).