Research On Cutting Parameters Optimization In Milling Based On Tool Wear Monitoring

In the process of NC machining, the selection of suitable cutting parameters is prerequisite to improve the machining efficiency and control the surface quality of workpiece. The traditional method to select the cutting parameters, which depended on the experience of operators or the off-line parameters optimization strategy, tended to ignore the effect of tool wear condition on the surface quality. However, the cutting performance and effect of tool are changing continuously along with the change of tool wear state, especially when the tool is on the condition of intense wear which has a significant impact on the surface quality of workpiece. In addition, the cutting force signal is very sensitive to the subtle changes in the cutting process, which can accurately and effectively reflect the state of cutting tool, while a kind of new force-measure tool holder integrated with the sensors of force and torque is designed by our laboratory to settle the installation deficiencies of traditional force-measure instruments. Therefore, the technology of cutting parameters optimization in milling based on tool wear monitoring is studied in this paper,in which the cutting parameters are adjusted along with the state of tool monitored online through the force signal to make the optimization objectives on the best status.Firstly, the overall struct of the online parameters optimization system was designed, while the strategies of monitoring the tool wear condition and optimizing the cutting parameters online were developed according to the practical optimization problems. In order to establish the tool wear state model,the milling experiment scheme and the method to extract the tool wear characteristics were designed. The parameters optimization model for milling process with the objectives of material removal rate and surface roughness deviation was established in the accordance with the requirement of cutting efficiency and quality.Then, the experimental data were analyzed to extract the characteristics of the cutting force signals,from which the tool wear characteristics were selected by the method of correlation analysis. The models of tool wear state and surface roughness deviation were established by using the neutral network optimized by genetic algorithm, and the inputs of first model are the tool wear characteristics and cutting parameters whlie the second one are the length of tool wear and cutting parameters. In addition, a linear regression model of surface roughness without considering the impact of tool wear on the cutting quality was established, in order to do the off-line optimization.Finally, the on-line parameters optimization system was developed by using the combined programming with MATLAB and Lab VIEW. In order to make the comparison between the optimization effect of on-line strategy and off-line strategy, the off-line parameter optimization model was optimized by genetic algorithm and the experiment was conducted with the result to validated the effect. The on-line parameter optimization experiment was conducted then. After every cutting path,the tool wear characteristics extracted from the cutting force signals collected from the force-measure tool holder as well as the cutting parameters were used to calculate the length of tool wear through the tool wear model. At last, the parameter optimization model on this condition of tool wear could be obtained, which was optimized by genetic algorithm to get the parameters for the next cutting path. The above operations were repeated until the tool was worn out. It was can be concluded that the on-line parameter optimization strategy had a better optimization effect than the off-line one according to the comparison of the exprimental results.