Use of isothermal representation for infrared-based fault detection in induction motors

In electrical engineering, the utilization of infrared thermography has been traditionally confined to the periodic inspection of electric installations, distribution lines and power transformers, among other applications. Over recent years, this technique is being increasingly incorporated in electric motors predictive maintenance programs. This is due to the fact that it has been reported its suitability for detecting certain types of problems, such as deficient bearing lubrication, cooling system problems, stator winding asymmetries or problems in the transmission system, among others. This paper presents an infrared-based methodology for detecting bearing anomalies in induction motors. It relies on the use of isothermal representations that enable to visualize the temperature gradient and, hence, to facilitate the localization of the anomaly source. At the same time, it enables an approximated computation of the motor losses, as it is carried out in the paper. The method is applied to both laboratory motors and industrial motors operating in a petrochemical plant. The results prove the potential of the presented infrared-based methodology to reliably detect the considered anomalies.