Vibration event detection: a monitoring method for slow speed bearings

Traditional signal analysis methods appear to fail in their ability to provide consistent meaningful information when presented with data from large slow moving slew bearings. A number of reasons for this are presented. Statistics obtained from vibration data collected from a large Coal Reclaimer and an experimental test-rig is discussed. The Coal Reclaimer rotates at 4.3 rpm about two vertically mounted, large, slew bearings. The experimental test-rig rotates at 1 rpm in the horizontal plane. These statistics are compared to the results obtained using a simple event detection algorithm. The event detection algorithm is detailed and its strengths discussed relative to other methods. It is found that the event detection method provides a consistent statistical view of the condition of the slew bearing but not necessarily better than simple statistical measures. The event detection algorithm is now being used as a condition monitoring tool on the test-rig designed to specifically condition monitor horizontally mounted slow speed (1 rpm) bearings to failure