Dynamic behavior of large faults toward severity estimation in bearings

Abstract

Estimation of fault severity throughout bearing life is required for bearing prognostics and remaining useful life estimation. The useful life limit varies based on the function and criticality of the bearing and the machine. Most research has focused on the initial stages when faults are small. This paper presents novel severity categories of a spall-like fault located on a deep groove bearing outer race by investigating the dynamic behavior. Different stages of fault severity, classified as small, medium, and large faults, are characterized from fault initiation to faults extending to angles corresponding to several balls interacting simultaneously with the fault. This research combines dynamic modeling and experiments. The dynamic model is used to understand dynamic behavior, and the experiments are used to validate the results. The dynamic behavior is used to analyze and explain new unique patterns in the vibration signatures of different fault severities. This study enhances the capabilities for estimating fault severity in ball bearings, providing valuable insights that contribute to the development of more accurate and effective maintenance strategies. This is particularly relevant in scenarios where the system can operate efficiently even with large faults exceeding the “small” category.