On-Line Preload Monitoring for Anti-Friction Spindle Beatings of High-Speed Machine Tools

Abstract

Catastrophic and premature bearing failure caused by excessive thermally-induced bearing preload is a major design problem for spindle bearings in high-speed machine tools. Due to a lack of low cost and easy to maintain on-line preload measuring techniques, the traditional solution is to limit the maximum spindle speed and the initial bearing preload. This solution is incompatible with the trend of high-speed machining, which requires increasing both spindle speed and spindle stiffness. Therefore, it would be valuable if thermally-induced preload can be monitored on-line for regulating bearing thermal behavior at high speeds. This paper proposes using a dynamic state observer based on a preload model to estimate the spindle bearing preload via low cost thermocouples attached to the bearing outer ring and the spindle housing. The observer is based on a state-space model capable of describing the transient preload behavior of the spindle bearing. The temperatures of the outer ring and housing are used as the feedback signals for the preload observer. The observer gains are determined systematically to account for modeling errors, unknown parameters, nonlinearities, and measurement noise. In particular, the modeling errors due to unexpected factors such as bearing skidding, wear, and lubricant deterioration are compensated by a Modeling Error Compensator (MEC). By using the MEC, the error dynamics of the observer can be converted into a form suitable for applying existing observer techniques such as the Extended Kalman Filter (EKF). This preload observer has been successfully validated on two different bearing configurations operated at different speeds. The results show that the model-based monitoring technique, which combines the measurement of outer ring and housing temperature and a robust state observer, can be an effective and low-cost solution for preload monitoring in high-speed machine tools.