Voxel-Based Adaptive Toolpath Planning Using Graphics Processing Unit for Freeform Surface Machining

Today, freeform surfaces are widely used on products in automobile, aerospace, and die/molds industries, which are generally manufactured using multi-axis computer numerical control (CNC) machines. Frequent changes in the design of products necessitate creating CNC part programs that need fast and accurate toolpath generation methods. Traditional toolpath generation methods involve complex computations and are unable to consider multiple surface patches together. The voxel-based computer-aided design (CAD) model provides the ability to represent the multi-patch surfaces in a discretized manner, which can be processed using an advanced parallel computing framework for accurate tool path planning. This paper presents a new method to generate an adaptive iso-planar toolpath for a three-axis CNC machining of freeform surfaces that are projectable onto a plane without overlap using the voxel-based part model. The algorithm is designed to work on a graphics processing unit (GPU) that allows parallel processing for faster toolpath generation. The proposed approach consists of two main steps, an algorithm to generate gouge-free cutter location points from the voxel-based CAD model and an algorithm to find out the sidestep and forward step from the cutter location points to create the final CNC tool path. A new image-processing technique has been proposed to identify gouge by detecting the shadow surface voxels and their intersection with the cutting tool. The developed system was extensively tested and compared with various reported toolpath planning strategies for machining complex freeform surface parts. The results show that the developed method is computationally efficient, robust, and accurate in generating adaptive planar toolpath.