Application of the Empirical Mode Decomposition to condition monitoring of damper bars in synchronous motors

Abstract

Damper bars play an important role in synchronous motors. Among other functions, they enable the direct-on-line startup of these machines. Moreover, they allow enhancing their stability during load transients. However, the large currents and stresses to which those elements are subjected during their operation intervals increase the probability of winding deterioration and of eventual breakage. This paper presents a logical extension of a transient-based diagnosis approach, whose application was successfully assessed in conventional induction motors. The method is applied here to diagnose the condition of damper bars in synchronous motors. It is based on the capture of the stator startup current of the machine and its further analysis in order to track the characteristic transient evolutions of a particular fault-related harmonic in the time-frequency map. For this purpose, the Hilbert-Huang Transform (HHT), based on the Empirical Mode Decomposition (EMD) method, is used here as proposed signal processing tool, even though other tools may be also viable. Simulations as well as experimental results on a laboratory synchronous machine prove the validity of the approach for condition monitoring of such elements.