Study of three phase induction motors with incipient rotor cage faults under different supply conditions

Abstract

The majority of rotor related faults in three-phase induction motors are due to broken bars and end rings. These faults occur primarily due to the thermal, magnetic, mechanical, environmental stresses that the rotor has to undergo during the routine usage. Faults involving several broken bars cause excessive vibration, noise and sparking during motor starting. Fabricated type rotors have more incidents of rotor bar and end ring breakage than cast rotors. On the other hand, cast rotors are more difficult to repair once they fail. Once a bar breaks; the condition of the neighboring bars also deteriorates progressively due to increased stresses. To prevent such a cumulative destructive process, the problem should be detected early, that is, when the bars are beginning to crack. This condition can be visualized as continuous increase in rotor bar resistance which increases from its nominal value to infinity when the bar is fully broken. Any experimental study to diagnose broken bar faults is costly as it causes irreversible damage to the rotor. Thus, a model based approach to simulate broken bar related faults at various degrees of severity is indeed essential. The present paper evaluates through simulation the line current spectrum of an induction motor at the incipient stage of bar breakage. The model can also be extended to multiple, full blown broken bar case. The speed and torque ripples caused by broken bars can also be studied. The rules and laws generated through such simulations can then be used in neural network based diagnostic tools. Results in case of complete broken bars are validated by finite element calculations. Experimental results with up to four bars partially broken with machine operating from balanced sinusoidal and inverter fed supply are presented. Simulation results showing that certain abnormal power supply conditions can produce broken bar like spectrum have also been included.