Time domain signal analysis to detect bearing faults using motor current signature analysis

Abstract

Bearings are critical components in the rotating machinery. The need for an easy and effective fault diagnosis technique has led to increase the use of motor current signature analysis (MCSA). In this research, a fault detection system for bearings was developed and then different faults were simulated and investigated in the test rig. MCSA is utilized since it represents a reliable approach for fault recognition in rotating machinery, time domain signals analysis technique was utilized to extract some indicative features, such as such as root mean square, kurtosis and skewness. However, in addition to the machine healthy condition two fault types, which are inner race fault and outer race fault, were introduced in the test rig. Three current sensors, type SCT013, were interfaced to Arduino MEGA 2560 microcontroller and utilized together for the purpose of data acquisition, to record the motor current signals. Then, the captured signals were analyzed and different time domain features were extracted. The results showed the effectiveness of using MCSA based time domain signal analysis in detection and diagnosis different bearings faults.