Analysis Of Shaft Vibration Feature Of Large Steam Turbine Unit Under Fast Changed Load

The operating safety and stability of a steam power generation unit with large capacity that participates in the load regulation is highlighted in our country but this aspect is lack of further study. The 600MW steam power units was studied. Shaft dynamics and stability, vibration performance in fault condition and fault feature extraction methods are studied with the support of science and technology project of nation and enterprise for providing reference basis for operation and fault diagnosis of units. Research results have practical implications in technology and engineering for improving the safety and reliability of units in peaking operation.The main research contents and achievement are as follows:(1) Calculation and analysis of dynamical characteristics were carried up for the shaft of a 600MW subcritical thermal power unit under rapid changes of load by using the transfer matrix method. The results show that dangerous shaft section of 600MW shaft is located at joint part between each rotor. When the load increases rapidly, shafting torsional response possesses damped oscillation property, which indicates that the shaft is in a stable state when the load increases rapidly.(2) Shaft vibration characteristics of the steam turbine by transient process and under abnormal running conditions were analyzed. Abnormal vibration response characteristics are investigated and signal feature extraction methods are studied for shaft fault diagnosis. Vibration characteristics of slow hot deformation phenomenon of a 600M W unit rotor in high-intermediate pressure are researched and the cause of the rotor thermal deformation is confirmed.(3) The research methods of monitoring and diagnosis for steam turbine rubbing fault. When the slight rubbing fault occurs, fault information is very weak in the relative shaft vibration response signal. Aiming at this characteristic, a new fault monitoring and diagnosis method is proposed that combined structure vibration monitoring and signal processing. The information of high-frequency impact vibration energy is used in this method and is suitable for abnormal vibration of the rubbing fault that contains transient impact. As relative shaft vibration monitoring is insensitive for impact vibration response caused by rubbing, a rubbing fault detection method by using wavelet singular value is proposed. The rubbing transient impact signal at specific scales is obtained by the multi-resolution analysis of wavelet. The amplitude envelope of transient impact signal is extracted. Then wavelet singular value detection method is used for detecting the saltation information caused by rubbing fault. It improves the effect of extracting features for weak rubbing fault obviously.