Spindle Radial Accuracy Evaluation and Fault Diagnosis Based on a Normal Distribution

A spindle as a key part of the machine tools have a significant effect on the machining precision and surface roughness. In this paper, spindle radial accuracy evaluation and fault diagnosis based on a normal distribution are investigated. Based on the least-square circle, the radial error model is developed. A least-square circle approximate algorithm is proposed to calculate the coordinates of the least-square circle center. Then on the basis of the normal distribution, a method to evaluate the synchronous radial error and the asynchronous radial error of the spindle is proposed. Furthermore, a fault diagnosis method is also given to detect the condition of the spindle based on the Kolmogorov-Smimov test. At last, the evaluation method and fault diagnosis method are verified experimentally. This work is useful for spindle radial accuracy tests and degradation tests of reliability studies.