Progressive segmentation for MRR-based feed-rate optimization in CNC machining

Keeping a constant cutting force in CNC machining is very important for obtaining better stability of cutting operation and improving topography, texture and geometry of the machined surface. This paper presents a feed-rate optimization approach based on Material Removal Rate (MRR). Given a tool-path with predefined feed-rates, the geometry of raw material, and the shape of cutter, the histogram of MRR in very fine resolution can be efficiently computed by using a GPU-based geometric modeling kernel. Starting from the evaluation given on the finest histogram of MRR, error-controlled subdivision algorithms are developed to progressively segment the tool-path into user-specified number of sub-regions. Different feed-rates are assigned to different sub-regions so that nearly constant MRR can be achieved while keeping the shape of the given tool-path unchanged. Experimental tests taken on real examples verify the effectiveness of this method.