The objective of this study is a development of a robust method for estimation of the size of a spall located on the bearing raceway by analysis of the measured vibration signal. Rolling element bearings are an essential part of rotating machinery. Therefore, preventing bearing failure is a subject of great interest. Effective failure preventing requires maintenance decision that is based on both diagnosis of bearing faults and estimation of the fault severity. Spalling is one of the most common bearing faults. Therefore, estimation of the spall size is the focus of the study.
During each interaction of a rolling element with a spall located on a bearing raceway, two events take place. The first event is the entrance of the rolling element into the spall, which is followed by a system response to a step function. The second event is the exit of the rolling element from the spall, which is followed by system response to an impulse. The distance between those events can be used for estimation of the spall size. The objective of the presented research is to locate those events in the vibration signal measured next to a faulty bearing.
Several studies were published on this subject. Those studies present algorithms that are based on sophisticated signal-processing techniques, which might be inapplicable for practical noisy systems. Development of a more robust method for separation of those events might give a better solution for practical rotating machinery. The proposed method is a simpler algorithm based on insights from an analysis of the mechanical properties of the system and its numerical simulations.
