Diagnostic of rotor windings of induction motors in rundown mode using an external magnetic field

Broken rotor bars in auxiliary motor rotors are a common cause of failures, leading to significant economic losses at power plants. One of the most promising methods for their timely detection is spectral analysis of the external magnetic field. However, the use of this method in rundown mode, which has several undeniable advantages, has received little attention. Thus, the purpose of this study is to detect special signs of broken rotor bars in the external magnetic field signal spectra of asynchronous motors in rundown mode, taking into account residual magnetization. The study has been conducted using simulation models of two induction motors developed in the Ansys software package. Spectral analysis of the external magnetic field induction has been used for monitoring, based on determining the harmonic amplitudes of a fictitious rotor winding. For processing, a short-time Fourier transform method is used applying the Flattop window function for more accurate determination of harmonic amplitudes. Frequency-time spectra have been analyzed for both motor models under varying loads. It has been shown that both in the presence of rotor winding currents and after their attenuation due to residual magnetization, the amplitudes of harmonics associated with the fictitious rotor winding significantly increase in the spectra of induction motors with broken rotor bars. Moreover, the higher the motor load in the pre-rundown mode, the greater the amplitudes of harmonics associated with the fictitious rotor winding. The results of the research have confirmed the feasibility to detect rotor winding damage based on the increase in harmonic amplitudes associated with the fictitious rotor winding, existing in the external magnetic field of both high-voltage and low-voltage motors.