Abstract
This study proposes a new simulation method of workpiece shape with voxel representation for multi axis controlled machining process. Usually, voxel representation requires enormous number of computational complexity and computer memory to estimate accurate shape of machined workpiece. In this paper, a new simulation algorithm based on octree collision detection with multi level voxel representation is proposed. The proposed algorithm is designed to adapt an characteristic of GPU hardware architecture for parallel processing. In the proposed algorithm, as attribute of each voxel, block number of NC program from the moment when the voxel is machined with endmill is described in computer memory. Then, by referring the attributes, the voxeles locate near the cutting tool can be identified within a short time. By applying the proposed algorithm, the prototype simulation system is developed. The developed system can estimate shape of the workpiece in machining process within a several micro seconds at an arbitrary moments.
