Examination of a technique for the early detection of failure in gears by signal processing of the time domain average of the meshing vibration

Abstract

An existing technique for the signal processing of the time domain average of the tooth meshing vibration of gears is examined with application to the early detection of failure. It is shown that the “regular” signal extracted from the time domain average of the gear vibration by the elimination of all except the fundamental and harmonics of the tooth meshing frequency forms the time domain average of the vibration of a single tooth. The “residual” signal which is obtained by subtracting the regular signal from the original time domain average gives the departures of the vibration from the average. Numerical and practical examples are given.