Incipient fault diagnosis in squirrel-cage induction motors

Behrooz Mirafzal, Marquette University


In this dissertation, a new diagnostic method for detecting rotor broken-bars and inter-turn short-circuit faults in induction motors is introduced based on electromagnetic field laws governing the operation and performance of induction machines. The basis of this new method is to monitor the orientation of the axis of the induction machine's magnetic field, and hence compute (estimate) the range of its oscillation due to the above mentioned types of faults in a synchronously rotating frame of reference. Differences in the pendulous oscillations due to broken-bars and inter-turn faults required different signal processing approaches (or algorithms) for detecting the range of the pendulous oscillations, the so-called swing angle indices. Moreover, these algorithms for obtaining the swing angle indices have been developed as practical fault indicators which diagnose effectively these faults even in the presence of some system and measurement noises. Furthermore, these algorithms and techniques have been applied, experimentally tested and investigated for a wide range of operating conditions, for two quite distinct and different induction motors, whose designs are of two distinct natures, under different load levels and different excitation systems. Using these swing angle indices for rotor broken-bars and stator inter-turn short circuit faults, namely, Δδ1 and Δδ sc , respectively, the severity of any such fault was effectively evaluated in a much more reliable manner than in case of using more conventional fault indicators such as frequency spectral side bands or negative sequence current/impedance components, etc. Specifically, this characteristic regarding the evaluation of the fault severity is also evaluated by making comparisons between the swing angle indices and the well-known side band component techniques as well as negative sequence current component for the cases of broken bars and inter-turn faults, respectively. Thus, in this dissertation, all three main steps of a diagnostic procedure are addressed concerning the so-called swing angle indices, while these three steps are (1) introducing the fault signature (or index), (2) developing the fault identification technique and (3) ascertaining the strength or robustness of the fault index for purposes of fault severity evaluations. (Abstract shortened by UMI.)

Recommended Citation

Mirafzal, Behrooz, "Incipient fault diagnosis in squirrel-cage induction motors" (2005). Dissertations (1962 - 2010) Access via Proquest Digital Dissertations. AAI3194798.