Open Access

ARTICLE

Paralleling Collision Detection on Five-Axis Machining

Cheng-Yan Siao¹, Jhe-Wei Lin¹, Ting-Hsuan Chien^2,*, Rong-Guey Chang¹

1 Department of Computer Science and Information Engineering, Advanced Institute of Manufacturing with High-Tech Innovations, National Chung Cheng University, Chiayi, 621005, Taiwan
2 Department of Computer Science and Information Management, Providence University, Taichung, 433303, Taiwan

* Corresponding Author: Ting-Hsuan Chien. Email:

(This article belongs to the Special Issue: Machine Learning and Deep Learning for Transportation)

Intelligent Automation & Soft Computing 2021, 29(2), 559-569. https://doi.org/10.32604/iasc.2021.018252

Received 02 March 2021; Accepted 05 April 2021; Issue published 16 June 2021

Abstract

With the rapid growth of the Fourth Industrial Revolution (or Industry 4.0), five-axis machining has played an important role nowadays. Due to the expensive cost of five-axis machining, how to solve the collision detection for five-axis machining in real-time is very critical. In this paper, we present a parallel method to detect collision for five-axis machining. Moreover, we apply the bounding volume hierarchy technique with two-level bounding volume represent the surface or solid of the object to reduce triangle meshes inside each axis of the five-axis machine tool, and then matching the operating range limit of the five-axis machine tool itself, delete the no colliding triangle mesh. Additionally, we also propose some optimization with loop unrolling and prefetching techniques to improve performance of collision detection. Our approach can reduce the execution time significantly by computing six separating axes in plan and eleven separating axis in non-plan between two triangle meshes based on the characteristic of GPUs (Graphics Processing Units) for program acceleration. Our proposed work consists of kinematic analysis and interpolation for axes to save the numerous collision detection for five-axis machining computations. In this experiment, the result shows that using the proposed approach above can achieve approximately 37.1 times speedup than that of CPU.

---

Keywords

---