Deteksi Cacat Bantalan Bola Pada Pompa Sentrifugal Menggunakan Spektrum Getaran

Abstract

One of the common fault in the centrifugal pump is faulty bearing. Bearings play a very important role for smooth rotation of a shaft. A bearing condition must be constantly monitored to ensure top performance of a pump. Therefore, a method is needed to detect an early defect in the bearings. One of the most widely used methods for bearing faults detection is based on the vibration analysis. Vibration analysis can detect a defect in a bearing without having to disassemble the machine. Furthermore, and it is fast and easy to be implemented. This study aims to develop a fault detection method on the ball bearing using spectrum analysis by applying envelope analysis. This research uses experimental method with three bearings conditions i.e. normal (no fault), outer race fault, and inner race fault. The type of ball bearings used are self aligning double row bearings. The vibration signal from each of bearing condition is taken from the centrifugal pump vibration test rig and measured using accelerometer sensor which is acquired directly with DAQ and then processed into Matlab. The analysis gives the result of frequency spectrum and envelope spectrum. This study concludes that the high amplitude on the frequency that coincide with the frequency of Ball Pass Frequency Outer Race (BPFO) and Ball Pass Frequency Inner Race (BPFI) make an indication of damage to the bearing on the outer and inner race respectively. The envelope spectrum gives better results as compared to the result of the frequency spectrum. This is because the high amplitude of low frequency generated from other components is blocked and removed using a high-pass filter. Consequently, it becomes easier to detect a low amplitude of high frequency vibration signal from a faulty bearing.