Investigations on Non-Invasive Fault Diagnostic Techniques for Three-Phase Induction Motor with Mixed Eccentricity

Abstract

Induction motors are extensively used in several industries such as oil, steel, petrochemical to name a few, and their failure may lead to the plant shut down resulting in the heavy financial losses. The motor operation management units are usually supported with protective gear which has the ability to detect the induction motor faults and stall the operation on detection. However, as a routine practice, it is essential to diagnose the defect in the motor by prognostic studies. There are various signature analysis techniques widely practiced in industries to detect nature and causes of failures in induction motor parts. Out of known causes of failures in these motors, 10 to 15% is attributed to the air gap eccentricity faults, thus demanding detection at the earliest. The major objective of this research work is to investigate the robust signatures to characterise the air gap eccentricity in the induction motor. To meet this objective, efficacy of various signatures that mark the presence of eccentricity are explored, and as an outcome it is found that d-q components of stator current can be used as the most suitable detector of machine failure due to mixed air gap eccentricity. Power signatures, Torque signatures and Power factor signatures were used for mixed eccentricity detection in the past by various researchers. These techniques use characteristic harmonics produced at the frequency signifying disturbances produced due to an air gap mixed eccentricity and used as signatures in air gap eccentricity diagnosis. In the present work it is shown that, the extracted d-q components of the stator currents contain characteristic harmonics signifying the presence of eccentricity and proposed as a better technique for performing the signature analysis. This technique is more advantageous and cost effective when compared to torque and power signatures analysis methods, since they are commonly used in controllers which are handling motor control. The scope of research work is modularized into following four phases: Phase 1: At this phase, mathematical expressions are derived for d-q components of stator currents in synchronous reference frame for the induction motor suffering from mixed air gap eccentricity. From the derived expressions, it is shown that they contain eccentricity specific harmonics in them. Phase 2: Here, a dynamic model of squirrel cage induction motor suffering from air gap mixed eccentricity fault is developed. And this model is used in simulatingii various mixed air gap eccentricity conditions and stored as a test data set. By systematic frequency signature analysis performed on the extracted d-q components of stator currents, it is shown that simulation results validate the mathematical expressions derived earlier. Phase3: In this phase, observations made at the phases 1 and 2 are experimentally validated by conducting experiments on 3HP three phase induction motor suffering from mixed eccentricity. In addition a motor mounting frame having special provision for introducing eccentricity is specially fabricated to introduce various degrees of eccentricity. Phase 4: To understand implications of supply voltage imbalances on eccentricity, detection technique is further investigated at this phase. The d-q components of stator currents in synchronous reference frame are often used in controller as control variables, as these are found to be dc signals. But as a part of the research investigation it is also shown that d-q components of stator currents extracted from stator currents of an air gap eccentric machine in synchronous reference frame will contain DC component superimposed with ripple oscillations. This ripple is due to eccentricity and it can be characterized by harmonics frequency even when the machine is fed with 3 phase sinusoidal voltages and running under constant load. As an outcome of all above work finally an integrated air gap eccentric condition monitoring unit and a controller unit for an induction motor has been proposed.