Vibration-based condition monitoring of rotating machines using a machine composite spectrum

Abstract

Vibration-based condition monitoring (VCM) requires vibration measurement on each bearing pedestal using a number of vibration transducers and then signals processing for all the measured vibration data to identify fault(s), if any, in a rotating machine. Such a large vibration data set makes the diagnosis process complex generally for a large rotating machine supported through a number of bearing pedestals. Hence a new method is used to construct a single composite spectrum using all the measured vibration data set. This composite spectrum is expected to represent the dynamics of the complete machine assembly and can make fault diagnosis process relatively easier and more straightforward. The paper presents the concept of the proposed composite spectrum which was applied to a laboratory test rig with different simulated faults; healthy and three faulty cases named misalignment, crack shaft, and shaft rub. A comparison between the composite spectrum with and without the coherence has been investigated for the simulated faults in the rig. It has been observed that the coherent composite spectrum provides much better diagnosis compared to the non-coherent composite spectrum.