David C. Swanson
Karl M. Reichard
Michael B. Van Dyke
Appl. Res. Lab., Penn State Univ., P. O. Box 30, State College, PA 16804
Rotational elements of machinery, such as bearings and gears, have been shown to exhibit specific vibration signal spectral patterns over various stages of failure, associated with certain defect frequencies characteristic of the elements and their components. It has been demonstrated that during advancing stages of wear, spectral peaks at defect frequencies are accompanied by increasing numbers of sidebands of other defect frequencies. In this paper, the use of wavelet template analysis to characterize vibration signals of defective rotational elements is presented as a method for identifying and quantifying element wear. The premise of this approach is that the dynamics of a worn component are characterized by an increasing degree of variation around its rotational frequency as the component degrades at its points of contact with other components, resulting in a vibration signal analogous to a frequency-modulated signal.