A Harmonic Wavelets Approach for Extracting Transient Patterns From Measured Rotor Vibration Data

Abstract

Vibration analysis is a powerful diagnostic tool for rotating machinery problems. Traditional approaches to vibration signature analysis have focused on the Fourier transform, which tends to average out transient effects. Recent work in the area of wavelets has allowed for the characterization of signals in frequency and in time, which, if properly interpreted, can provide substantial insight, particularly with regard to transient behaviors. There are many different types of wavelets, but the harmonic wavelet was developed specifically for vibration analysis. It uses an algorithm based upon the fast Fourier transform (FFT), which makes it particularly attractive to many in the vibration analysis community. This paper considers the harmonic wavelet as a tool for extracting transient patterns from measured vibration data. A method for characterizing transient behaviors using the harmonic wavelet is described and illustrated with simulation and experimental results.