A Novel Approach to Rapid Toolpath Computation Using GPU Hardware

Current graphics processors (GPUs) provide tremendous processing power that outperforms higher-end central processing units. Although GPUs were originally targeted for computer graphics applications, their programmability has opened up possibilities for other applications. This research aims at developing methodologies and algorithms for the rapid generation of CNC surface machining toolpaths using GPU hardware. It will focus on the specific toolpath generation technique developed recently by the principal investigator. A major challenge of this research is the re-formulation of the time-critical tasks in toolpath computation into graphical problems so that the parallel processing power of the GPU can be fully utilized. The outcome will not only improve the efficiency of the current CAM process but also open the possibility of developing a fully automatic CAM system whereby a large set of machining strategies and parameter settings can be attempted and compared systemically to find the optimal toolpath.