A Voxel Based Automatic Tool Path Planning Approach Using Scanned Data as the Stock

Abstract

This paper presents a novel automatic tool path planning approach which employs the scanned data as the stock. Using the scan data as the stock allows a new geometrical approach to measure the difference between the stock model and desired final geometry. This approach employs a voxel model as the underlying geometric representation, which enables an efficient parallel tool path planning algorithm based on a general-purpose graphics processing unit (GPGPU) approach. The approach is composed of two stages: first, to identify the excess material to be cut and reduce the required GPU memory, a novel voxelization algorithm for machining was developed to voxelize the scanned data and the model of the part in a single 3D grid. Second, slice-by-slice contour parallel tool paths for roughing and 3D contour parallel tool paths for finishing were generated directly based on the voxel model. The proposed approach has been successfully implemented to machine a part representative of the welded blade tip and a part representative of a cast workpiece with a freeform surface into parts with desired final geometry. The experiment results show that the proposed approach is efficient and feasible in generating tool paths when scanned data was used as the stock.