Dynamic modeling of double cage induction machines for diagnosis of rotor faults

Double cage induction motors are typically employed in applications that require high starting torque for frequent, loaded starts. This makes the starting (outer) cage susceptible to fatigue failure since it must withstand the large thermo-mechanical stress associated with the high starting current and long acceleration time. Conventional FFT-based spectrum analysis techniques are insensitive to outer cage faults due to the small outer cage current under steady state operation, and the fault indicators and thresholds applied to single cage rotors cannot be used. Therefore, there is a strong industrial need for research and development of analysis techniques and advanced fault detection algorithms for double cage machines. However, there are only a few publications that investigate double cage rotor faults. In this paper, a dynamic model for rotor cage faults in double cage motors, which is the first step towards double cage rotor fault research, is derived. A 7.5 Hp separate end ring fabricated copper double cage rotor is built for experimental verification of the proposed fault model (copper inner cage/brass outer cage). The tests performed under startup and steady state operation show that the proposed fault model can provide sufficient accuracy and can be used as the basis for interpretation of FFT results and development of advanced fault detection algorithms.