Predictive modeling of roughness change in multistep machining

Following sustainability in manufacturing, the machining chain can be optimized by either reducing the time and energy consumption of each operation or eliminating the unnecessary operations subjected to keeping the quality of the final product as consistent. However, the roadblock in designing an optimum machining chain is lack of prediction tool to interact between the included operations. In this paper, an integrated algorithm is developed to simulate the surface roughness generation and following modification caused by milling and burnishing, respectively. Predict the surface roughness generation by milling process and its alternation after burnishing. The algorithm works on the basis of clouds of points which were generated in the engagement region of tool and workpiece and their transformation from tool to workpiece coordinate systems. Moreover, some mechanical attributes of the process regarding effect of surface work hardening and elastic rebound were added to the algorithm to enhance the accuracy of simulation. To verify the results, a series of burnishing experiments with multi-roller rotary tool have been carried out on the surface of the finish-milled samples and the surface roughness change was taken into investigation. The obtained results showed that by applying the work hardening and springback effect to predictive algorithm the prediction accuracy of roughness at submicron level enhances up to 50%. It was also found that the most influential parameters influencing the surface roughness after milling-burnishing sequence are milled surface roughness, burnishing force and pass number. In addition, results showed that applying burnishing after rough machining consumes lots of energy to achieve nanoscale surface finish. Accordingly, the sequence of rough-milling, finish-milling and burnishing results in achieving sound surface finish within significantly shorter period of time and applied force.