Statistical Approach for Preload Monitoring of Ball Screw Drives

Preload loss is one of the most frequent causes of ball screw drive failures. To avoid a reduction in production quality or sudden machine breakdowns, condition monitoring systems are of interest. One method to observe the preload conditions is the use of strain gauge sensors embedded in the nut. Evaluating the strain signal, we show that a loss of preload is associated with irregular motion of the balls. Based on that, a novel statistical method to characterize the smoothness of the balls' motion is introduced. The approach results in a health value, which is directly associated with the preload condition of the ball screw.