The development of time-dependent compensation model for roller CVC profile generation in precision grinding

Abstract

In the process of CVC (continuous variable crown) profile generation for large rollers (D≥600mm, L≥600mm), the profile accuracy reduces dramatically due to the undercut and overcut errors, which are primarily owing to wheel width and time-dependent wheel wear. The conventional method to improve the profile accuracy in industry relies on “trial and error.” In this study, a time-dependent compensation model for multi-pass roller grinding is established, which is capable to predict the roller profile accurately with consideration of the relative wheel-roller motion and wheel wear. Based on the prediction results, the undercut and overcut errors are minimized through using continuous feed and modified roller profile curve. The experiment of model verification shows that the mean error and root mean squared error are only 5.9μm and 7.7μm, respectively, which verify the model prediction accuracy. The results of compensation experiment show that the roller profile accuracy is improved effectively by using continuous feed or modified roller profile curve and reaches the best when using both two ways. After full compensation, the total and average roller profile deviations reduce by 31.37% and 44.87%, respectively, which verify the effectiveness of the proposed time-dependent compensation model.