Condition monitoring on bearing faults of electric motor – Based on sound signal

Abstract

The condition monitoring of rotating shaft is important in terms of system maintenance and process automation, especially for the bearing which is most common component in machine or apparatuses. In this project, a general model of faulty bearing based on sound signal will be developed. The microphone is used to detect the frequency signals of sound from new bearing, used bearing and bearing with artificial damage. The graph of frequency response for different bearing conditions is plotted using Matlab Programming. This graph will be used to diagnose fault in bearings on the other electrical motor. The maximum noise value for each type of bearing condition was taking as a range to determine the condition of bearing if tested at motor that have same model and speed. Experiment result shows, value of noise produce for new bearing are between 47.07 dB until 65.57 dB. While, value of noise produce for used bearing are between 49.50 dB until 66.58 dB and value of noise produce by damage bearing are between 55.33 dB until 70.07 dB.