Chaotic Characteristics Analysis of Acoustic Emission Signals from Cavitation in Francis Turbine

In order to realize the diagnosis of the cavitation state of a Francis turbine, the analysis of acoustic emission signals (AE) was carried out for the fault diagnosis of the cavitation state of a Francis turbine under different load conditions in this paper. Sensors were arranged at the guide vane arms and the tailwater inlet gate, and the AE signals were collected under ten load conditions ranging from no load to full load of 140MW. The Largest Lyapunov exponent (LLE) was calculated, which is combined with the time-domain characterization and phase space reconstruction techniques to analyze the chaotic characteristics and its changing rules. The experimental results show that the cavitation activity has a significant load interval characteristic. The active cavitation interval at the guide vane arm is 30 ~ 75MW, and the one at the tailwater inlet gate is 45 ~ 75MW. The AE signals under the above conditions exhibit highly nonlinear and chaotic characteristics, while those tend to be stable under low load and full load conditions. The cavitation diagnosis method based on the chaotic characteristics of AE signals can effectively identify the load interval and activity intensity of hydraulic turbine cavitation, which is of engineering application value to ensure the safe and stable operation of the hydraulic turbine.