| Centrifugal pump is one of the widely used fluid mechanical machines,and it plays an important role in the departments of electric power,petrochemical industry, agriculture irrigation and textile industry. Cavitation phenomenon is a common fault happened in pumps. With regard to centrifugal pump, the appearance and growth of cavitation can change the velocity distribution of the fluid passageway, and cause the decline of efficiency and the lift, and lead to vibration and noise and impeller damage.Fault diagnosis has become an independent and cross-subject synthetical information-processing technology.It is one of the basic measures to ensure safe operation of equipments. It can predict the early stage faults and judge the reasons of these faults so as to avoid or reduce accidents of the equipments. Finding suitable methods of diagnosis and identification, or fast and exact fault removal ,and preventing damage and enhancing efficiency and service life are of great importance.Firstly, this text simulated different cavitation conditions of the centrifugal pump working in specified operating mode on the closed experimental device and collected the pressure fluctuation signal of the centrifugal pump inlet, and then researched the cavitation fault diagnosis of the centrifugal pump according to the characteristic of the signals. Lastly, analyzed the signals using different methods.This text used four methods as'Autocorrelation analysis combined with Short Time Fourier Transform(STFT)'and'correlative dimension analysis'and'energy entropy analysis of EMD'and'wavelet multi-scales statistical features analysis'. This text also compared four methods and pointed their limitations and characterstics. This text preliminarily explored the way of cavitation fault diagnosis of the centrifugal pump using pressure fluctuation method,and the experimental and analytical results can be regarded as the reference to the engineering application.
|
PDF Full Text Request