Collision detection for virtual machine tools and virtual robot arms using the Shared Triangles Extended Octrees method

Abstract

The performance of virtual machine tools and virtual robot arms relies on the use of efficient and precise collision detection methods. This study proposes an octree-based collision detection method, called the ‘shared triangles extended octrees’, applied on virtual machine tools and virtual robot arms. The proposed scheme combines the high computational efficiency of the octree test and the high numerical accuracy of an analytical surface boundaries intersection test. In the proposed algorithm, the overlapping voxels between neighbouring geometries in the virtual mechanisms are identified using octrees, and the intersections of the triangles within these voxels are then checked. It works efficiently, with rapid generation of deformed geometry by shared triangles without using decomposition.The proposed collision detection scheme is implemented on virtual machine tools and the virtual robot arms in a virtual manufacturing cell. The virtual machine tools and virtual robot arms are composed of component trees, which describe the kinematic relation between components. The necessity to check for collisions between each component is indicated by a Boolean matrix. The pairs of components selected by the Boolean matrix are checked using the proposed shared triangles extended octree method.The proposed collision detection method is an efficient tool for verifying the manipulation of CNC machine tools and robot arms and is of great help to CAD/CAM engineers and manufacturing engineers and operators of machine tools and robot arms.