Abstract:

An approach based on model was employed to identify the rotor bending deformation degree. The original vibration signals of the rotor system were recorded, the difference between the online detected vibration signals and the original signals were computed. Combining with the rotor system dynamics model in normal state, the equivalent loads on each node were obtained, and the online bending degree of the rotor was identified by analyzing the equivalent loads. In view of a lot of noise existing in the signals, a denoise approach based on the empirical mode decomposition was proposed and employed. The results of the numerical examples show that the proposed approach is effective in identification of the rotor bending deformation degree.

Key words: mechanics, rotor system, identification of rotor, bending defotmation