Novel non-invasive flux-based airgap eccentricity detection for induction machines and its practical validation

The airgap between the stator and rotor of the rotating machines plays a significant role in determining the performance and efficiency of the machine. Any non-uniformity in the airgap introduced either at commissioning or during operation must be identified at the earliest. This is usually achieved with the help of flux sensors installed in the airgap between stator and rotor or by special constructional modifications. This paper presents a method for detecting airgap eccentricity in rotating machines using non-intrusive flux sensors which sense external leakage flux, along with information about type of eccentricity (static, dynamic, or mixed) and its severity. Detecting the type of eccentricity is as important as detecting the eccentricity well in time. The former helps in deciding the appropriate repair actions while the latter helps in avoiding severe damage and catastrophic failures. A finite element method-based modelling was first attempted to understand the characteristics and signatures of the different eccentricity conditions followed by the development of the detection method. The proposed method is then validated experimentally by implementing the detection method in a protection relay and testing on a real machine. Depending on severity of the condition, either alarms or trip signals are generated by the relay.