Modulation Signal Bispectrum Analysis of Acoustic Signals for the Impeller Wear Detection of Centrifugal Pumps

Alsadak Daraz, Samir Alabied, Fengshou Gu, Andrew D. Ball

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The impeller is one of the main mechanical rotating components in a centrifugal pump which contains vanes at the middle between two faces often called as a closed impeller. Basically, the plain vane leads to the edge square to the back shroud. The impeller in the pump needs more attention as the mechanical wear can cause severe damage reducing the lifetime and efficiency of the pump significantly. In this paper impeller wear of centrifugal pump is investigated based on the acoustic mechanism under different flow rates. A little wear that appears in the area of impeller inlet rapidly increases with the rise in peripheral velocity. The area where the impact velocity is below the critical value is known as the uniform corrosive wear area and it grows with the increase of the impeller velocity. In fact, the wear mechanisms are different from one area to another; it is larger when the tangential component to the impact velocity is high which is responsible for the failure mode in the impeller. Based on this comprehension, the modulated signal bispectrum (MSB) analysis has been used for extracting the incipient fault signatures with its unique capability of noise reduction along with the modulated the signals. The result from the experimental process shows the diagnostic features developed by modulated signal bispectrum process can diagnose the impeller wear in an early stage. Moreover, the MSB analysis of acoustic signals and their spectrum provides a good attestation of full separation between healthy and faulty conditions under different flow rates.

Conference

Conference25th IEEE International Conference on Automation and Computing
Abbreviated titleICAC 2019
CountryUnited Kingdom
CityLancaster
Period5/09/197/09/19
Internet address

Fingerprint Dive into the research topics of 'Modulation Signal Bispectrum Analysis of Acoustic Signals for the Impeller Wear Detection of Centrifugal Pumps'. Together they form a unique fingerprint.

Cite this