Research And Design Of Measurement System For Turbine Cavitation Noise

Cavitation and erosion have severely impacted to the hydro-turbine operations, which reduce the machine's efficiency and cause accidents. Therefore, it's necessary to develop a suitable system which can get and manage a quantity of data, and can analyse the characteristics of hydro-turbine cavitation.Based on the characteristic descriptions and analysis methods of the cavitation noise, a construction of cavitation noise detection system is presented by focusing on the following aspects: (1) design of data acquisition card for cavition noise based on consecutive sample, (2) development of signal sample and saving software, (3) development of cavitation anaysis software.A data acquisition card, which is premised with no aliasing distortion in the valid frequence range, is designed for collecting the cavitation noise. The card has the following features: wide frequence range, high sampling frequece, consecutive sample, and gain adjustability. In order to realize the consecutive sample, which is the difficulty for the design, the three following technologies are presented: (1) digital signal decimate algorithm, which can cut the redundant frequence band brought by the over-samplings, and can reduce the data transfer through PCI bus, but won't bring the aliasing distortion between 1Hz and 1MHz; (2) storage method with two alternative SRARs based on FPGA, which can prevent the data after signal decimate from being overlaid by the new data, and therefore ensures that the data can transfer through PCI bus consecutively; (3) Scatter/Gather DMA transfer and the structure of two DMA memory buffers, which can improve the efficiency of the data transfer through PCI bus.To control the data acquisition card and get the signal with PC computer, software for signal sample and saving is designed. The software also has the function of saving and managing data. Based on the data acquisition system, software for cavitation anaysis is designed. The software provides three following sorts of analysis methods: the general siginal processes, the allied signal analysises, and the special cavitation analysises. The identification algorithm for cavitation state based on cavitation noise is designed and integrated to judge the state of cavitation. And the calculation of cavitation erosion is designed to evaluate the loss of the erosion.