| Rub-impact is one of the common faults in rotating machinery such as steam turbine unit. The rub-impact fault is a direct reason for rotor complex vibration and instability. When rub-impact occurs, elastic deformation causes stress wave emission which mean the phenomenon of rub-impact acoustic emission. Therefore, the acoustic emission signals with rub-impact information could be used to identify and diagnose faults, thus it is important to ensure the safe operation of the rotating machinery.In order to better understand the acoustic emission signals which generated in rotor rub-impact. A finite element propagation model of acoustic emission signal was established in ANSYS/LS-DYNA software by using the theory of elastic wave propagation in solid. The acoustic emission signal in steel plate mode propagation path, dynamic response in time and frequency domain and the attenuation of signal energy were studied firstly.In order to use method of array signal processing to find out the rub-impact fault location, then the far field model and near field model based on array signals were developed in this thesis. In this study, the simulation method was used to compare the conventional beam-forming method and the MUSIC algorithm, and the effects of different parameters in the array of choices on resolution performance of MUSIC algorithm were also analyzed.In order to make the acoustic emission signal source localization process by using simple operation, the acoustic emission signal acquisition system was established based on LabVIEW software platform. The developed system integrates data collection, file management, signal analysis and source localization module. The signal analysis module includes time domain analysis, frequency domain analysis and time-frequency domain analysis. The source localization module includes the near-field and far-field acoustic emission source location which based on array signal processing, and the experiment on steel sheet verify the reliability of the acquisition system.The rub-impact test was carried out on the simulation of 300MW steam turbo-generator rotor test rig. The experimental results show that a section of frequency information could be extracted from the acoustic emission signal by using band-pass filter method. The rub-impact position between the static and dynamic components could be effectively diagnosed by utilizing MUSIC algorithm. All of these studies could provide an effective and feasible method for the monitoring of rub-impact fault. |
PDF Full Text Request