Reliable Flux-Based Detection of Induction Motor Rotor Faults From the Fifth Rotor Rotational Frequency Sideband

Abstract

Motor current signature analysis (MCSA) is widespread in the field as a means of providing remote, online monitoring of industrial induction motor rotor faults. However, numerous cases of false indications with MCSA produced due to asymmetries in the rotor structure or load leading to unnecessary inspection or forced outages have been reported. Although alternative test methods under motor standstill or starting transient immune to the false indications have been proposed, a reliable method capable of "online" monitoring is desirable. In this article, a new fault indicator in the airgap or stray flux capable of reliable online detection is proposed. It is shown that the fifth sideband of the rotor rotational speed frequency can provide detection of rotor faults immune to most of the false MCSA indications. Experimental results under the most common false indications are provided to verify the claims made for cases where online MCSA produces false indications. The proposed fault indicator is also shown to be immune to eccentricity and common load defects such as load unbalance and misalignment.