A GPU-Accelerated Approach for Collision Detection and Tool Posture Modification in Multi-Axis Machining

作者：Wang, Jing; Luo, Ming^*; Zhang, Dinghua

来源：IEEE Access, 2018, 6: 35132-35142.

DOI：10.1109/ACCESS.2018.2848938

• CAM
• collision avoidance
• high performance computing
• GPU
• machining
• manufacturing
• tool posture modification

Collision detection and avoidance between solid bodies are one of the most important problems in path planning for robotics machining or multi-axis machining. While planning toolpath for multi-axis milling, accurate collision detection is usually time-consuming among complex solid bodies in the computer environment. Furthermore, how to avoid the collision automatically within limited space in the path planning stage still needs lots of experience. To this end, this paper presents a general collision detection and tool posture automatic adjustment approach for the multi-axis milling process. First, by analyzing the contact state of the tool-workpiece, the calculation model of the interference quantity is determined. A unified tool constraint mathematical model based on the interference quantity and the interference type is established. Second, three types of tool adjustment strategies are constructed, the sequential quadratic programming method is used to solve the model and the graphics processing unit-based high-performance computing technology is employed to accelerate the solution process. Finally, the developed method is validated for automatic collision and tool posture adjustment in the five-axis milling of a blisk. The presented method can be integrated into commercial CAD/CAM software for rapid tool collision detection and tool orientation modification.

• 出版日期2018
• 单位西北工业大学