Slice spectrum and time-frequency analysis of vibration acceleration signal of the elastic support of the high-pressure rotor

The structure of an aero-engine high-pressure rotor is complex, and the frequency components of vibration signals are rich. The traditional spectrum analysis method can not identify its vibration characteristic frequency well, so it can not use frequency characteristics to identify its faults. In order to master the frequency components and time-frequency distribution of the vibration signal of the elastic support of the high-pressure rotor under different conditions, firstly, build a double rotor test bench, carry out a vibration test, and obtain the shaft and elastic support vibration signals under normal working conditions and typical faults; Secondly, the local mean decomposition (LMD) combined with the slice spectrum analysis method is used to adaptively decompose and analyze the frequency components of the obtained vibration signals, based on the frequency characteristics of each fault type, the identification frequency of the vibration fault signals based on the elastic support is given; Then, for the product function (PF) component that can represent the fault frequency component, the Hilbert spectrum is used to conduct time-frequency analysis on it, and the time-frequency distribution characteristics of each typical faults are mastered. The collected signals are further analyzed by using PF energy as an indicator. The results show that when the high-pressure rotor fails, the PF energy representing the fault information in the elastic support signal has a significant increase trend; Finally, the universality of the results and the effectiveness of the method are verified by taking the connection looseness fault signal as an example. The methods and conclusions presented in this paper can provide some reference for aero-engine rotor vibration feature extraction and fault identification.