A novel detection method for diagnosis of rotor eccentricity in three-phase induction motor

Abstract

In this paper, a novel method to diagnose eccentricity fault in rotor of three-phase induction motor based on the current signal, is proposed along with experimental validation. It utilizes three-phase current in time series into Clarks two phase α-β 90 • apart vectors assimilating into 0 • measuring reference frame starting point with the aid of key-phasor averaging method. α-β sinusoidal waves are applied to an interpolation method which gives an improved Fast Fourier transform of amplitude and phase angle of α-β components. These transformed two-phase current signals are again measured to get amplitude and phase. Finally, the amplitude and phase of α-β are used to generate holo-spectrum with an ellipse of the first order. Since unbalance fault occur at peak fundamental 1X harmonic frequency component, so its amplitude and phase angle are utilized for orbit construction. A comparison is made on rotor orbit and its eccentricity values, constructed from vibration signals and three-phase current signals, both measured simultaneously. This leads to a direct relationship in terms of orbit eccentricity and its longitudinal major axis inclination. The results show that at low frequency the eccentricity values of vibration and current signals orbits coincide hence the unbalance fault is detectable in induction motor by only utilizing the current signals. This method represents rotor non-linear dynamic behavior and is immune to unsymmetrical and anisotropic surface properties of the rotor and can measure rotor condition representing an accurate severity level of air gap eccentricity or, in mechanical terms, rotor unbalance fault. The new technique is a progressive development in induction motor air gap eccentricity fault detection and is applicable to harsh running conditions due to its simplified and non-invasive approach.

Maximal overlap discrete wavelet packet transforms and variants of neutrosophic cubic cross-entropy-based identification of rotor defects

“…In this work, MODWPT is applied at three levels to generate eight sub-frequency bands. The energy E of the vibration signal at the eight sub-frequency bands is computed and represented as E = e 0 3 , e 1 3 , e 2 3 , e 3 3 , e 4 3 , e 5 3 , e 6 3 , e 7 3 .…”

“…The rotor is an influential component of any rotating machinery [1][2][3][4], the catastrophic failure of which could lead to the machine's sustainability and thus impede the total working of the machinery. In the current scenario, many entropy and signal processing technique-based rotor defect identification methodologies [5][6][7][8] have been deployed for observing non-linear and non-stationary characteristics of the vibration signals of rotors.…”

Special feature on rotating machinery condition monitoring by connecting physics-based and data-driven methods

“…Focusing on turbines, Khalique et al [14] proposed a method to diagnose eccentricity faults in the rotor of a three-phase induction motor based on the current signal. They utilized three-phase induction motor current vectors 120 • apart and transformed them into 2D signals by applying Clark's transformation α and β apart at 90 • .…”