Force Feedback Control of Tool Deflection in Miniature Ball End Milling.

Abstract

Previous research at North Carolina State University focused on open-loop compensation of machining errors associated with tool deflection in miniature ball end milling. These methods utilized tool force models to predict deflections an precompensate tool paths off-line to achieve dimensional tolerance and accuracy in finished parts. Accuracy depended on the tool force model, its cutting parameters, and workpiece alignment and dimensional accuracy. Real-time force feedback has the potential to further improve the accuracy of profiles created during machining. This paper demonstrates that force feedback can be used to predict tool deflection and compensate for deflection during the milling operation, reducing susceptibility to uncertainties in model parameters and workpiece alignment. Two specific force feedback approaches are presented here: cutting depth prediction (based on a non-dynamic cutting force model) and tool deflection prediction (using a non-dynamic model of tool stiffness). Real-time control algorithms incorporating both methods were implemented and evaluated on a high-speed air bearing spindle. A non-dynamic tool force model developed previously at the Precision Engineering Center used measured forces to predict depth of cut. A separate tool stiffness model was developed to predict tool deflections (axial and radial) based on measured forces. Experiments involving machined grooves in hard steel workpieces, including simple slotting cuts and three-dimensional finishing operations, were conducted at various tool tilt angles to evaluate the effectiveness of force feedback control. Results indicate that profile errors can be reduced up to 80% compared to non-compensated cases. These results confirm that real-time force feedback control can significantly improve the dimensional tolerance and accuracy of injection molds created using miniature ball end mills.