Experimental study on the optimum time for conducting bearing maintenance

Bearings are critical components in the products of many industries, and their failure can result in long downtime and costly maintenance. Prolonging a bearing's service in a safe manner is vital for operators of equipment, and condition monitoring is regarded as one of the best approaches to achieve this. In general practice, condition monitoring does successfully indicate the presence and growth of bearing faults. However, standard condition monitoring techniques do not usually have a proven method to determine the best time to conduct maintenance of a defective bearing. Traditionally, maintenance will be conducted when the measured vibration of a bearing is found to exceed an acceptable vibration level, as defined by an industry standard. However, the industry standard is only a general guideline to the design and operation of the kind of machines/components under certain conditions. The actual dynamic response of a bearing is reliant on a variety of factors, such as lubrication, loading, temperature, and operational and environmental conditions. As a consequence, although the empirical vibration levels suggested by the standard are helpful in ensuring the safe operation of a bearing, they cannot guarantee full utilization of the residual life of a defective bearing. The purpose of this paper is to try and find a feasible solution for this issue, and moreover validate it through a number of experiments conducted under various loading and operational conditions. In the research, four dimensionless condition monitoring criteria: normalized information entropy; J-Divergence; Kurtosis; and a composite criterion based on them, are employed to assess the actual health condition of the test bearings with different types and severity levels of failures. Experimental results have shown that in comparison of the industry standard, the proposed method provides an effective and feasible approach for predicting the optimum time to conduct bearing maintenance. It is deemed that the achievements of this work will help operators in further improving their management of assets.