Back to Search Start Over

Enhancing five-axis CNC toolpath smoothing: Overlap elimination with asymmetrical B-splines.

Authors :
Hu, Yifei
Jiang, Xin
Huo, Guanying
Su, Cheng
Li, Hexiong
Shen, Li-Yong
Zheng, Zhiming
Source :
CIRP: Journal of Manufacturing Science & Technology; Sep2024, Vol. 52, p36-57, 22p
Publication Year :
2024

Abstract

In the realm of five-axis Computer Numerical Control (CNC) machining, the challenge of minimizing velocity, acceleration, and jerk fluctuations has prompted the development of various local corner smoothing methods. However, existing techniques often rely on symmetrical spline curves or impose pre-set transition length constraints, limiting their effectiveness in reducing curvature and maintaining velocity at critical corners. To comprehensively address these limitations comprehensively, a novel approach for local smoothing of five-axis linear toolpaths is presented in this paper. The proposed method introduces two asymmetrical B-splines at corners, effectively smoothing both the tool-tip position in the workpiece coordinate system and the tool orientation in the machine coordinate system. To fine-tune transition curve scales and minimize velocity disparities between adjacent corners, an overlap elimination scheme is employed. Furthermore, the two-step strategy emphasizes the synchronization of tool-tip position and tool orientation while considering maximal approximation error. The outcome is a blended five-axis toolpath that significantly reduces curvature extremes in the smoothed path, achieving reductions ranging from 36.28 % to 45.51 %. Additionally, the proposed method streamlines the interpolation process, resulting in time savings ranging from 5.68 % to 8.78 %, all the while adhering to the same geometric and kinematic constraints. [ABSTRACT FROM AUTHOR]

Details

Language :
English
ISSN :
17555817
Volume :
52
Database :
Supplemental Index
Journal :
CIRP: Journal of Manufacturing Science & Technology
Publication Type :
Academic Journal
Accession number :
178400634
Full Text :
https://doi.org/10.1016/j.cirpj.2024.05.013