The detection of shaft misalignments using motor current signals from a sensorless variable speed drive

Shaft misalignments are common problems in rotating machines which cause additional dynamic and static loads, and vibrations in the system, leading to early damages and energy loss. It has been shown previously that it is possible to use motor current signature analysis to detect and diagnose this fault in motor drives. However, with a variable speed drive (VSD) system, it becomes difficult to detect faults as the drive compensates for the small changes from fault effects and increased noise in the measured data. In this paper, motor current signatures including dynamic and static data have been investigated for misalignment diagnosis in a VSD system. The study has made a systemic comparison of different control parameters between two common operation modes: open loop and sensorless control. Results show that fault detection features on the motor current from the sensorless mode can be the same as those of the open loop mode, however, the detection and diagnosis is significantly more difficult. In contrast, because of the additional frictional load, features from static data show results of early detection and diagnosis of different degrees of misalignment is as good as that from conventional vibration methods.