Model for predicting Bearing Performance from Profiles of Ball Races

Abstract

AN analytical technique for predicting precision bearing performance from race profilometry has been developed in this laboratory. This development is an outcome of research directed towards improving the torque ripple characteristics of angular contact bearings operating at very slow speeds. Race profiles are frequently used during the manufacturing process to measure the effects of finishing operations and as a final quality control check on individual races. They are obtained by monitoring the radial displacement of a stylus as it traces the ball path on a bearing race. Sensitivities as low as 5 µinches are easily obtained with most standard equipment and the resulting data are recorded on polar charts which are easy to read. Determination of the torque ripple characteristics of an assembled bearing, however, has in the past required torque testing in preload conditions. Poritsky et al.¹ developed a method for determining the steady state frictional torque in a pivot bearing of ideal geometry based on the slip mechanism in the contact areas between ball and race. Local deviations from ideal geometry and the resulting change in ball-race contact forces give rise to the variations in steady state frictional torque referred to here as torque ripple.