IEEE TRANSACTIONS ON MAGNETICS, VOL. 44, NO. 12, DECEMBER 2008 4673 A New Pattern for Detecting Broken Rotor Bars in Induction Motors During Start-Up Jawad Faiz and Bashir-Mahdi Ebrahimi Department of Electrical and Computer Engineering, School of Electrical and Computer Engineering, Center of Excellence on Applied Electromagnetic Systems, University of Tehran, Tehran, Iran This paper introduces a novel frequency pattern for diagnosis of broken rotor bars in induction motors during the start-up period. This is important in the case of start-up of large induction motors with long starting time and also motors with frequent start-up. The amplitude of the side-band components of the introduced frequency pattern is a suitable index for diagnosing the broken bars and de- termining the number of faulty bars precisely during start-up. The paper investigates the performance of a faulty motor. Because the number of broken rotor bars and the load are two major parameters affecting the performance of a faulty motor, the frequency spec- trum of torque and speed in a motor with different numbers of broken rotor bars and different loads is analyzed. The time-stepping ﬁnite-element method (TSFEM) is used to model the motor. This model takes into account the spatial distribution of the stator windings, nonuniform magnetic permeance of the air gap, nonlinear characteristic of the cores, and continuous rotation of the rotor. The paper compares simulation results in steady-state mode with experimental results and shows that the results agree well. Index Terms—Broken rotor bars, induction motor fault diagnosis, starting transients, time-stepping-ﬁnite element method. I. INTRODUCTION S QUIRREL-CAGE induction motors are widely used in small and large industries, such as electric power stations, oil reﬁners, and factories. Product lines will be stopped if these motors fail to operate. In addition to the disturbance of their operation, induction motor faults shorten their life-time. Therefore, diagnosis of the faults is important from perfor- mance improvement and longer life-time points of view. A broken rotor bar is a common fault in induction motors. All methods reported until 1987 have put limits on the type and degree of fault, and there is no general and comprehensive method for faults diagnosis (for different numbers of broken bars and poles). Symmetrical components are among the ﬁrst applied methods for broken bar diagnosis [1]. In [2], the Norton theorem has been used for analysis of broken rotor bars. It shows that it is impossible to diagnose the disturbance due to the fault in the rotor using torque-speed characteristic or stator current. Therefore, a search coil has been suggested to do the job. In [3], a double-cage induction motor has been analyzed in steady-state mode for broken rotor bars assuming a current passing the outer cage. However, there are considerable differ- ences between the predicated and measured values which is due to neglecting the saturation and inter-bar currents. Alternative fault diagnosis methods are estimation and measurement of the apparent resistance in broken rotor bars [4]; however, in this method, variation of the rotor resistance with tempera- ture must be taken into account. Variations of the rotor and stator resistances have been taken into account ofﬂine, which decreases the accuracy of the suggested method. In [5], the presence of sidebands in speciﬁc harmonics on both sides of the fundamental component of the ﬂux in faulty machine has been detected using a search coil. But this fault diagnosis method is not usable for all types of machines with different structures. Digital Object Identiﬁer 10.1109/TMAG.2008.2002903 Color versions of one or more of the ﬁgures in this paper are available online at http://ieeexplore.ieee.org. Also, location of the detecting coil has a considerable effect on the fault diagnosis. The relation between the magnetomotive force (MMF) distribution of an induction motor and varieties of broken rotor bars has been discussed in [6], but saturation and inﬂuence of the spatial harmonics have not been taken into account. Magnetic ﬁeld distribution in a healthy machine and a ma- chine with ﬁve broken bars have been obtained by using the ﬁ- nite-element method (FEM), and a few search coils have been then ﬁxed inside and outside of the machine in order to diagnose the broken rotor bars [7]. However, the broken bars current have been assumed equal to zero, due to the presence of the inter-bar currents which is not quite true. It is shown in [8] and [9] that both the level of load and the fault inﬂuence the stator current spectrum. If the stator current is visualized in the synchronous reference frame, an existing magnetic asymmetry inﬂuences both and components of the current, while the load ﬂuctuation affects only the com- ponent. In [10], a simulation method has been presented based on the rotating ﬁeld theory for an induction motor with nonsinu- soidal air gap ﬂux distribution, in the presence of different load types. However, this method is useful only for a particular type of fault and nonideal operating conditions. In [11]–[13], appli- cation of intelligent techniques has been proposed for diagnosis of broken rotor bars. In [14] and [15], it has been assumed that there are three symmetrical windings on the stator and loops on the rotor, and with equation , steady-state analysis of the motor has been carried out, and the presence of components in the stator current and components in the torque and speed have been shown. In [16], and compo- nents have been calculated using Park’s transformation, and the attempt has been made to diagnose the fault by comparing the Park’s current patterns in the healthy and faulty motor. In [17], the winding function method has been used to model the faulty induction motor under broken rotor bars. In Section II of the paper, the fault diagnosis algorithm is introduced and its different parts are discussed. In Section III, modeling of the motor with broken rotor bars using time-step- ping ﬁnite-element method (TSFEM) is presented. The experi- mental setup is described in Section IV and the basic principles 0018-9464/$25.00 © 2008 IEEE Authorized licensed use limited to: Tehran University Trial User. Downloaded on January 22, 2009 at 06:51 from IEEE Xplore. Restrictions apply.