Condition monitoring and fault detection in roller bearing used in rolling mill by acoustic emission and vibration analysis

Abstract

Bearings for rolling elements are essential components of rotating devices and bearing failure can lead to machine failure. As a result, early identification of such defects, as well as the seriousness of damage when the bearing is in use, will aid in the prevention of machine failure and breakdown. Vibration is caused by defective bearings, and these vibration signals may be used to test the bearings. This paper defines the use of acoustic emission to predict fault detection in rolling mill roller bearings in relation to the gradual rise in defect size. The signals acquired by the transducers from the bearings are distorted by other faults and mechanical noise from the machinery, so vibration-based approaches are commonly used in bearing fault diagnosis. A rolling mill machine's condition monitoring involves keeping a close eye on the performance of the roller bearings and detecting bearing faults until they cause any problems. Acoustic emission is a non-destructive testing tool used in structural health control that is gaining interest in the diagnosis of bearing defects. Vibration signals provide a wealth of information about a bearing's operating state. A bearing test rig was designed and developed to investigate various defects in rolling element bearings in a real-world environment. Researchers had previously been unable to create a connection between defect sizes and vibration amplitude despite studying various types of seeded defects with random sizes. The experimental research discussed in this paper focuses on a seeded defect of the same kind that grows in size over time on the outer race of radially loaded cylindrical roller bearings, as well as running the defective bearing at different speeds and loads. Data was collected simultaneously using Acoustic Emission (AE) and vibration probes for better diagnosis. The use of acoustic emission (sound) obtained from the near field region of bearings in good and simulated faulty conditions for fault diagnosis is presented, and it is concluded that the Acoustic Emission (AE) approach is superior to identify faults in roller bearings used in rolling mills over a range of speed and load conditions at gradual increase of defect size.