Influence on rotor broken bar fault diagnosis of mechanical torque pulsations by means of FFT

Induction machines are connected to a wide variety of mechanical elements. A failure in drive or load will affect the performance of the whole system. Therefore, it will cause malfunction of the “healthy element” and it may lead to it she destruction with time. Nowadays it is possible to monitor the state of drive and load by analysing the electrical machine signatures applying MCSA such as FFT and statistical indicators like neural networks fuzzy logic algorithms. In order to be effective it has to be defined with features of the signals which are going to be studied, as well as the magnitudes which imply a fault. This paper attempts to point out the effects over the typical signatures used for fault diagnosis of rotor cage and mechanical faults when the fault frequency of both load and drive coincides by means of FFT applied to the stator phase currents, supported with simulation and experimental results.