Use of air-gap torque spectra for squirrel cage rotor fault identification

Identification of rotor faults in squirrel cage induction motors by estimation of air-gap torque spectra has gained considerable attention recently. This has the advantage that the signature spectrum generated by faults is in the low frequency region. But the torque computation needs a machine model. The standard healthy machine model has been proposed in the literature to estimate the signatures but the correlation between the actual torque and the estimated torque is not clear. This paper presents a simulation study of an induction motor with faulted rotor bars and presents a comparison with torque spectra predicted by healthy machine dq model excited with faulted machine currents. It is shown that the spectral information produced by the latter is able to reproduce not only the frequency but also the amplitude of the torque spectrum to a fair degree of accuracy.