3D simulation of double-needle bar warp-knitted clustered pile fabrics on DFS

The shortcomings of traditional pile fabric design and simulation in simulating clustered structures are addressed in this study by proposing a new three-dimensional (3D) simulation method. To accurately capture the details of clustered pile fabrics, a fine geometric model of a single pile loop is first established, and a novel geometric transformation algorithm, combined with rotational transformations and trigonometric calculations, is employed to achieve the natural aggregation of pile fibers, resulting in a highly realistic clustered structure. Additionally, to enhance design flexibility and accuracy, a pile design pattern is introduced, and for the first time, a Depth-First Search (DFS) algorithm is utilized for precise pattern recognition and determination of the directions of pile fibers, effectively solving the challenges of clustered area identification. To enhance the visual realism of the simulation, the Phong illumination model is adopted to simulate the lighting effects on the fabric surface, and blurring processes are applied to eliminate sharp edges and noise, resulting in a softer and more natural simulation output. Based on the recognition results from the DFS algorithm, the directions of the pile fibers are finely adjusted to achieve a realistic simulation of clustered pile fabrics. Experimental validation demonstrates that the proposed method produces 3D simulation results that closely resemble actual pile fabrics, demonstrating significant practical value and innovation.