Al Thobiani, Faisal (2011) The Non-intrusive Detection of Incipient Cavitation in Centrifugal Pumps. Doctoral thesis, University of Huddersfield.

This thesis investigates methods for the detection of incipient cavitation in centrifugal pumps. The thesis begins by describing the working of the centrifugal pump which makes this type of pump particularly prone to cavitation. The basic mechanisms of cavitation are described, which explain why this phenomenon is so damaging.

The thesis reports the results of experiments to predict the onset cavitation using a range of statistical parameters derived from: the vibration signal obtained from an accelerometer on the pump casing, the airborne acoustic signal from a microphone close to the outlet of the pump and the waterborne acoustic signal from a hydrophone in the outlet pipe close to the pump. An assessment of the relative merits of the three methods for the detection of incipient cavitation is given based on a systematic investigation of a range of statistical parameters from time and frequency domain analysis of the signals.

It is shown that is the trends in the features extracted are more than their absolute values in detecting the onset of cavitation. A number of recommendations are made as to which features are most useful, and how future work incorporating these suggestions could give a powerful method for detecting incipient cavitation.

A major contribution of this research programme is the development of a novel capacitive method for the detection of cavitation. Some basic theory is presented to show the principles of the device and then the details of its construction and placement in the test rig built for the purpose. The data for the tests using the capacitive sensor are given and we can say definitely that it has been confirmed as a method of detecting cavitation in a pipe system, and that it is a promising method for the detection of the onset of cavitation.

Faisal_Al_Thobiani_-_Final_Thesis.pdf - Accepted Version
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (6MB) | Preview


Downloads per month over past year

Add to AnyAdd to TwitterAdd to FacebookAdd to LinkedinAdd to PinterestAdd to Email