Although the Foufier Transform has been used with great success on periodic signals arising in mechanical and electromechanical devices, there is no conventional method of equivalent power that is available for the detection and analysis of transients. Yet, in many cases, transients carry the bulk of the critical information. Examples are transients arising from hinge pin wear in check valves and rubbing of rotating parts in machinery. These low-level impulsive signals are very hard to detect with conventional techniques, yet awareness of the degrading condition can prevent catastrophic failure. Researchers are examining the use of the Wavelet Transform. In the last decade, it has been placed on a firm mathematical foundation and has recently been applied with promise in various nonmechanical fields. Its attractive feature is its ability to analyze a signal based on local rather than global features. The project evaluates the use of wavelet transform technology as a basis for analyzing transients. Its initial application is in the analysis of mechanical and electromechanical transients. The research focuses on fundamental issues, namely repeatability of wavelet coefficients in the presence of arbitrary time shifts, and the ability to individually identify transients.The potential commercial application as described by the awardee: Research will result in the detection, identification, and analysis of shaft rub, motor arcing, gear anomalies, check valve hinge pin wear, and any fault-indicating transient associated with mechanical or electromechanical devices. The research will prove to be useful in utilities, petrochemical plants, and manufacturing facilities.
