Analysis Of Centrifugal Pump Cavitation Based On Underwater Noise Test

Occurrence of cavitation in centrifugal pump, Will produce cavitation in the low pressure area, and collapse in high pressure area. The cavitation bubble collapse will generate signals, and these signals include the characteristics of cavitation. Therefore, research is needed on the cavitation noise signal. Cavitation signal is submerged in the background noise, so that it is difficult to extract and analysis. Although there are many signal processing methods used in the extraction and analysis of signal. Because the cavitation signal belongs to the weak signal areas, therefore, the traditional signal processing methods exist many disadvantages. This paper collected the cavitation noise signals of a centrifugal pump cavitation, to find a new approach to the analysis of the underwater noise signal, extraction of signal characteristics of cavitation.The main task of this paper is to study the cavitation of centrifugal pump, at present, the research on cavitation has become mature by the numerical simulation method,obtained the performance of Centrifugal pump. Then, for the same type of centrifugal pump cavitation, the noise signal acquisition of centrifugal pump cavitation, The power spectral analysis method and empirical mode decomposition algorithm were used for analysis and extraction of the cavitation noise signals, To have found out the characteristics of the cavitation signal and the law of variation.Through the analysis of the power spectrum of cavitation noise signal can be concluded, the characteristics of cavitation signals in the frequency domain mainly concentrated in low frequency, While in middle and high frequency band has no obvious characteristics. Using this method can well reflect the development process of cavitation of centrifugal pump, and the definition of two power threshold, which can quantitatively determine the degree of development of cavitation of centrifugal pump.By adopting the empirical mode decomposition method for the analysis of cavitation noise signal, the decomposition component of signals have existed mutation pulse component obviously. This shows that the method can be used to effectively detecte mutation pulse signals in the strong background noise. The number of pulse signal of decomposition components in the statistics can be found, Only the number of pulse signal of 1 decomposed component with degree of cavitation gradually increased, This is consistent with the analysis results of the mechanism. Using this method can well reflect the development process of the cavitation of centrifugal pump cavitation of centrifugal pump, the judgment of the stage, at the same time, this method can also be used to estimate incipient cavitation. Finally, the results with the results obtained by the numerical simulation were compared, the two results are consistent, which shows that the numerical simulation results are correct.