Prediction of Surface Roughness of End Milling for Cycloidal Gears Based on Orthogonal Tests

End milling method is applied to machining of cycloidal gears to improve the cutting quality and efficiency. The influence of milling parameters on the surface roughness is investigated based upon orthogonal tests with the four factors and four levels, as well as analysis of range and variance. A model to predict the surface roughness is built up on basis of the probability statistics and multivariate nonlinear regression analysis. Significance tests are conducted on the prediction model, and the interactive effect of these parameters on the surface roughness is figured out so as to propose optimization schemes. The results show that the shaft inclination angle has the biggest impact on the surface roughness, followed by the feed per tooth, the radial feed and the spindle speed. The prediction model of surface roughness is proved to have high prediction accuracy. This study aims to provide references for the improvement of machining quality of cycloidal gears and optimization of milling parameters.