Development of bearing fault detection system based on vibration signal

Bearing is a type of rolling element in which are vastly used in industry or even for home appliances. The functions may be differ with each other, but most importantly it allow the movement of any shaft to rotate smoothly. The position and movement of this rolling element are varied for the required functions. The movement of this rolling element could sometimes disturbed by the bearing problems which account at about 40% of machines failure. Therefore, this research are conducted to emphasize bearing fault detection. Bearing conditions are monitored for 8 hours in order to receive a smooth vibration signal for the defected bearing. In that time duration, the vibration signal received may not be clear as this is due to outside noise vibration in which interrupts the vibration signal collection. But with skills developed in handling error, we can minimize the unwanted vibration signal. 5 sets of bearing are used for this research which includes 3 defected bearings in which we tested those bearings for analysis by using Fast Fourier Transform (FFT). The initial vibration signal received in which was conducted for the healthy bearing, is used as a benchmark to compare with the defected bearing’s vibration signal. The initial data collection was conducted using Dasylab. The collected data saved for analysis in Matlab. The initiations towards comparing the vibration signal can be determined from the bearing’ s feature frequency itself, outer race, inner race, the ball spin and also the bearing cage frequency. Nevertheless, FFT proved to be an effective method on monitoring bearing defect. For future research, a method of detecting the type of defect should be emphasize as this may reduce time consumption before the whole machine could damage