An investigation into current and vibration signatures of three phase induction motors

Ashnibha, Rmdan Abdussalm (2012) An investigation into current and vibration signatures of three phase induction motors. Doctoral thesis (PhD), Manchester Metropolitan University.

Preview

Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (5MB) | Preview

Abstract

This research aimed at investigating the relationship between three phase induction motors vibration (MVS) and current signatures (MCS). This is essential due to the cost of vibration measuring equipment and in cases where vibration of interest point is not accessible; such as electrical submersible pumps (ESP) used in oil industry. A mathematical model was developed to understand the effects of two types of induction motors common faults; rotor bar imperfections and phase imbalance on the motor vibration and current signatures. An automated test facility was developed in which 1.1 kW three phase motor could be tested under varying shaft rotation speeds and loads for validating the developed model. Time and frequency domains statistical parameters of the measured signals were calculated for fault detection and assessing its severity. The measured signals were also processed using the short time Fourier transform (STFT), the Wigner-Ville distribution (WVD), the continuous wavelet transform (CWT) and discrete wavelet transform (DWT) and wavelet multi-resolution analysis (MRA). The non-stationary components, representing faults within induction motor measured vibration and current signals, were successfully detected using wavelet decomposition technique. An effective alternative to direct vibration measurement scheme, based on radial basis function networks, was developed to the reconstruction of motor vibration using measurements of one phase of the motor current. It was found that this method captured the features of induction motor faults with reasonable degrees of accuracy. Another method was also developed for the early detection and diagnosis of faults using an enhanced power factor method. Experimental results confirmed that the power factor can be used successfully for induction motor fault diagnosis and is also promising in assessing fault severity. The suggested two methods offer inexpensive, reliable and non-intrusive condition monitoring tools that suits real-time applications. Directions for further work were also outlined.