Fault Signature Analysis of Industrial Machines

The condition monitoring and fault diagnosis is an important function of the intelligent mechanics. Understanding equipment’s condition helps engineers avoiding inevitable catastrophe that could cause unplanned downtime and affect the system reliability and safety. In addition, alleviation of the machine’s incipient faults helps the machine running efficiently and reduces energy consumption. This paper presents field case studies of industrial induction motors including pumps and fans from several facility factories, and uses electrical data not only to analyze machine’s condition and but also to detect the faults. Current spectrum analysis is mainly employed to detect the anomaly. In addition, we propose a ratio of current divided by voltage (C/V) spectrum to represent the steady gain property of the system. This ratio is expected to mitigate the effect of Variable-frequency drive’s noise and disturbance on current spectrum and to reduce the false alarm caused by Variable-frequency drive (VFD). The combination of current spectrum and spectrum ratio is beneficial in the conditional analysis of industrial motors.