Abstract:

Objective This article aims to improve the needling quality and production efficiency of three-dimensional (3-D) composite preforms with irregular shapes. Based on the optimization of 3-D needling path planning algorithms and robotic collaborative needling, a dual-robot needling path planning technique was proposed. This method established an improved needling path planning approach for multi-curvature complex preforms by constructing needling points and auxiliary points, and calculating spatial attitudes. The complex preforms were divided into two zones, and each zone was individually needled by the dual-robot system, further reducing the production cycle.

Method This study employed two 6-axis 3-D needling robotic devices to prepare special-shaped preforms using quartz cloth/felt laminates. Rapid output of needling points and auxiliary points was achieved based on CATIA software. The normal vectors of the needling points were obtained by constructing four co-vertex triangular mesh planes. Leveraging spatial vectors, rotation matrices, and Euler angle calculations, the developed partitioned needling path planning computer aided manufacturing (CAM) software successfully generated highly precise and executable robot programs. The rationality and feasibility of the technique were verified through offline robot simulation and physical experiments.

Results The needling path planning method was validated through experiments, and the experimental results showed that by dividing the complex preform into two needling zones, and by reasonably allocating the needling trajectory using CATIA software and post-processing CAM software, the robot's end pose could be accurately calculated. The two needling robots could simultaneously needle the preform, without any collision, doubling the production efficiency during the entire needling process.

By comparing to the results from the double robot partitioned needling experiment, it could be seen that the needling trajectory simulated by AUTOCAD was highly consistent with the needling trajectory results on the surface of the preform in the experiment. This indicated that by constructing a triangular mesh plane with four common vertices to obtain the normal vector of the needle puncture point, the problem of inaccurate calculation of the normal vector in areas with large curvature changes such as corners of the preform could be effectively avoided, and the overall uniformity of the preform could be significantly improved.

In addition, based on Python and QT Desinger, combined with the calculation method of robot end pose and Kawasaki AS programming language, the CAM software for partitioned needling trajectory planning of needling robots had been developed. It could quickly process the position information of needling points and auxiliary points exported by CATIA, generate robot executable programs, and prove the reliability of the algorithm through simulation and experimental verification. The software has a simple interface and is easy to use. This trajectory planning method is also applicable to the tufting and I fiber implantation process in the weaving and forming of preform with variable curvature surfaces.

Conclusion The method for outputting needing points and auxiliary points based on CATIA had successfully achieved parametric output of the coordinates of these points. Leveraging the established calculation methods for spatial vectors, rotation matrices, and Euler angles, the partitioned needling path planning CAM software had been successfully developed, enabling the efficient and precise output of executable robot programs. The offline simulation process for dual-robot partitioned needling was collision-free, with precise positioning of the end-effectors, and the CAM post-processing software algorithm was reliable. The needling marks on the rotary preform's surface were highly consistent with the computer-simulated trajectories. The resulting preform exhibited a smooth surface and good uniformity. The dual-robot system efficiently achieved the needling of the quasi-rotary preform, doubling the production efficiency compared to a single robot. This method is suitable for mass production weaving of quasi-rotary needling preforms. Furthermore, this path planning approach provides a reference for robot tufting and robot I-fiber sewing processes in fabricating 3-D preforms.

Key words: needling robot, partitioned needling, path planning, special-shaped preform, composite