Methodology for ball screw support bearing fault analysis with screw nut vibration signal

Ball screw support bearing degradation is a critical factor of machining precision of machine tools. In order to ensure machine efficiency and position precision, this paper proposes a method of monitoring the condition of support bearing of ball screw system through the Empirical Mode Decomposition (EMD) with envelop spectrum analysis. The proposed method can extract the diagnosis feature of support bearing from the vibration signal of ball screw nut. Vibration of ball screw nut is a typical Amplitude Modulation Frequency Modulation (AM-FM) signal. To handle multi-scale dynamic AM-FM vibration signal, Empirical Mode Decomposition decomposes raw signal into different modes with their respective center frequencies. And the specific discrete frequency signal and random noise can be separated by soft thresholding denoising method. At last, the envelop spectrum of reconstruction signal is calculated to diagnosis the fault characteristic. The vibration signals collected from ball crew nut are used to demonstrate the performance of the proposed method to identify the support bearing fault. The results verify the effectiveness of the method in extracting fault characteristics and diagnosing faults of support bearing.