Crack fault diagnosis of rotor systems using wavelet transforms

As the traditional Fourier transformation method cannot process the non-linear and non-stationary vibration signal of the cracking rotor system, the 3-D waterfall spectrum in combination with reassigned wavelet scalogram method is presented to analyze the temporal frequency characteristics of the crack fault. Three different crack depths of rotor crack faults were simulated on a cracking rotor test-bed. The experiment results show that the present method is effective in determining the frequency structure of the signal, especially the twofold frequency component, which can be measured at half of the critical speed. The present method is an effective tool to diagnose cracks in rotor system and the frequency characteristic can be used as references in the diagnosis.