A Study on the Ball Screw Preload Identification Method Using Experimental Natural Frequency

The ball screw feed drive system (BFDS) is widely used in most machining machines because of its accuracy and reduced error caused by vibrations. After long-term operation, the ball screw component will lose stiffness, so its health condition needs to be regularly monitored. In this study, a procedure for the health condition monitoring of ball screws was established using an experimental vibration-based method. The reduction of ball screw stiffness is assessed through the preload level loss and is directly reflected in the vibration responses of the BFDS. First, a dynamic model for BFDS is used to investigate the vibration modes of the system, including axial and torsional vibrations. In this case, the axial stiffness directly determines the axial vibration response of the ball screw, which can be directly identified or measured from the experimental model. The dynamic parameters of BFDS are identified, and then the natural vibration frequency results are calculated from the dynamic equations by using the eigenvalue problem. The vibration frequency of the BFDS is determined based on the vibration measurement signals recorded directly from the accelerometers mounted at the ball nut through a real-time signal processing system. Finally, the calculated results are compared with the experimental vibration measurement results to verify the correctness of this study. From there, the experiment preload level of the ball screw is quickly and efficiently identified for BFDS condition monitoring.