Predicting milling force variation in time and space domain for multi-toothed face milling

Variations of milling force in time domain and space domain are the common behaviors of face milling for large-scale component with discontinuous multi-holed surface, which specifically generate surface variations along feed direction and circumferential direction. An accurate prediction model of local milling force is a key issue to reliable simulation of the milling force variations and the resultant machined surface variations. However, milling force variation modeling is still a significant challenge due to changing of instantaneous chip thickness in angular domain and the non-constant cutting coefficients under different milling conditions. Therefore, a general milling force model with cutting coefficients derived from average forces of milling experiments may loss the robustness when applied to different milling conditions. To address this issue, this paper attempts to provide a methodology to identify the unknown cutting coefficients in dual-mechanism local force model for face milling through the periodic averages integrated with multiple angular local forces which are measured at specific sampling angular positions in each single test. And each of the dual-mechanism cutting coefficients is modeled as multi-quadratic regression equation individually based on response surface methodology (RSM) via the design of experiments. The influence of feed per tooth, spindle speed, and cutting depth on the cutting coefficients are studied in order to build a model which can predict reliably the local milling force for different process parameter combinations. The proposed prediction model of milling force variations in time and space domains is verified by experimental measurements.