An Experimental Study on Vibration Signatures for Detecting Incipient Cavitation in Centrifugal Pumps Based on Envelope Spectrum Analysis

Authors

1 Department of Mechanical Engineering, Faculty of Engineering, Mustansiriyah University, Baghdad, Iraq

2 Institute of Sound and Vibration Research (ISVR), Digital Signal Processing and Control Group, University of Southampton, SO17 1BJ, UK

Abstract

Vibration signatures have been studied for monitoring the condition of centrifugal pumps by many researchers, however, there is limited published information on the application of vibration analysis to incipient pump cavitation. The paper will review the state of the art in the field and develops an effective signal processing approach based on envelope spectral analysis to close this gap. A purpose-built test rig was employed for recording vibration signals from a centrifugal pump at a wide range of operating conditions. The collected data was then processed using time domain and frequency domain analysis methods. The study has shown that the vibration energy concentrated mainly in the frequency range between 8-15 kHz. At the flow rates less than 300(l/min), i.e. in design flow rate range, the vibration amplitudes remain constant and does not show a notable change by the flow rate increase. However, a notable increase in the vibration level is evident when the flow rate exceeds 300 (l/min). Analysis the result of filtered vibration signatures has revealed that vibration signal parameters: peak value, root mean squared (RMS), crest factor along with vibration spectrum allow development of cavitation (for the flow rates higher than 300 (l/min)) to be diagnosed reliably. However, conventional signal processing methods may not produce a clear separation of the incipient cavitation from the healthy baseline. Therefore, envelope spectrum analysis has been carried out on recorded vibration signatures to detect the onset of cavitation from the baseline and satisfactory results have been perceived.

Keywords

Journal of Applied Fluid Mechanics
Volume 12, Issue 6
November 2019
Pages 2057-2067

Files

Cited by

Share

How to cite

Statistics