Prediction Research Of Surface Roughness In High Speed Milling Of Hardened Steel

As one of the most advanced manufacturing technologies, High speed machining is developing rapidly in recent years. It has a serious advantages such as high machining efficiency, good surface quality. And It can realize green machining and difficult-to-process materials machining. High speed machining has been widely used in the mold processing, automobile manufacturing, and aerospace. However, the surface quality, which has a significant influence on performance and service life of parts, become a key factors of impeding the development and popularization of high speed machining. Therefore, a systematic research on the mechanism and prediction of surface roughness in high speed machining, have great theoretical significance and practical value.Based on the high speed milling experiments of45steel, the influence of milling parameters and workpiece hardness on the surface roughness is studied in this paper. And the surface roughness prediction models are established based on the experiment data. These works not only lay a theoretical basis for the cutting mechanism research and milling parameters optimization, but also provide technical support for the popularization of high speed machining.Firstly, the influence of workpiece hardness on the surface roughness and milling force is studied through signal factor milling experiments in this paper. And the results show that, with the increase of workpiece hardness, the surface roughness increase first and then decrease; but the milling force decrease first and then increase.Then, Based on orthogonal experiments principle, the high speed milling experiments of45steel are conducted. The surface roughness is considered as experiments index and the workpiece hardness, milling speed, cutting depth and feed rate are considered as influence factors in the experiments. The experiments results is analyzed by the means of rang analysis and influence of each factors on the surface roughness is studied. The results indicate that, with the increase of workpieces hardness, surface roughness increase first and then decrease; and with the increase of milling speed, surface roughness decrease. It also show a trend of increase with the increase of feed rate and milling depth. And the range analysis shows that the feed rate make the largest influence on surface roughness, followed by workpiece hardness, milling speed and milling depth in turn.Finally, based on the methods of regression, least square support vector and particle swarm optimization-least square support vector, the surface roughness prediction models of high speed machining45steel are established respectively by using the orthogonal experiments data. Through experimental verification and comparison, It shows that, the prediction error of regression model is large; and the prediction error of LSSVM model is7.61%, which is improved compared with regression. However, The prediction error of PSO-LSSVM model is only3.632%, which is just50%of LSSVM and it avoids the problems of parameters selection by using the PSO method to optimize the parameters. The surface roughness prediction model of PSO-LSSVM has a high prediction accuracy and generalization ability which can predict the surface roughness precisely.