Detection of developing bearing failures by means of vibration analysis.

The application of vibration, as the sensing medium for measurements in predictive and preventive maintenance, is steadily increasing on many types of machinery and production systems. One of its earliest applications is the monitoring of the condition of rolling element bearings. Although the procedure has been used in industry for over two decades, many uncertainties still remain and complete consensus on its effectiveness does not yet exist. The growing interest in the application of knowledge based expert systems for the control of larger installations emphasizes further the need for additional studies in this area. This study presents the results of an experimental investigation of the basic relationships between developing failures and the resulting changes in vibration response on small size, high speed, ball bearings. These bearings are commonly used in high speed spindles installed in automated metal cutting systems. It has been found that acceleration, in comparison with displacement and velocity, has the best capability of detecting the development of ball bearing faults. Among the acceleration sensing parameters, the conventional acceleration frequency range shows almost the same effectiveness as the high frequency range descriptors, such as spike energy and acoustic emission. In the study of the statistical descriptors, kurtosis shows measuring values close to 3 at the beginning and the end of the bearing life. Higher values are shown in between the two. By comparison of measuring effectiveness in low and high frequency ranges, it has been concluded that for the majority of industrial applications, the best results are obtained by monitoring vibration acceleration in the frequency band of approximately 6 kHz to 20 kHz.Dept. of Mechanical, Automotive, and Materials Engineering. Paper copy at Leddy Library: Theses & Major Papers - Basement, West Bldg. / Call Number: Thesis1991 .L355. Source: Dissertation Abstracts International, Volume: 53-01, Section: B, page: 0516. Chairperson: L. K. Smedick. Thesis (Ph.D.)--University of Windsor (Canada), 1990