Razbin Milad, Salehian Mortaza, Gharehaghaji Ali Akbar
Department of Textile Engineering, Amirkabir University of Technology, Tehran, Iran.
School of Aerospace, Mechanical & Mechatronic Engineering, The University of Sydney, Sydney, NSW, 2006, Australia.
Sci Rep. 2024 Oct 11;14(1):23800. doi: 10.1038/s41598-024-74602-2.
Different forms of close-packed yarns can be produced by varying the number of monofilaments in the core region, ranging from one to five. Numerous efforts have been made to model or simulate the mechanical response of close-packed yarns; however, previous studies have predominantly focused on one or two monofilaments in the core. In this study, we propose an analytical approach that combines a geometrical model with an artificial neural network (ANN) to predict the tensile behavior of close-packed yarns containing 2 to 5 monofilaments in the core region. The novelty of this hybrid model lies not only in accounting for more than two monofilaments in the core but also in extending the prediction range from elastic to viscoelastic-plastic behavior. Validation of the proposed method showed excellent agreement between experimental and theoretical results. Numerical simulations further confirmed that the results align with theoretical predictions, demonstrating the model's accuracy in predicting the tensile behavior of close-packed yarns. This modeling approach has the potential to significantly improve the understanding and modeling of textile structures.
通过改变芯区单丝的数量（范围从1到5），可以生产出不同形式的紧密排列纱线。人们已经做出了许多努力来对紧密排列纱线的力学响应进行建模或模拟；然而，先前的研究主要集中在芯区的一到两根单丝上。在本研究中，我们提出了一种将几何模型与人工神经网络（ANN）相结合的分析方法，以预测芯区含有2至5根单丝的紧密排列纱线的拉伸行为。这种混合模型的新颖之处不仅在于考虑了芯区中超过两根的单丝，还在于将预测范围从弹性行为扩展到粘弹塑性行为。对所提出方法的验证表明，实验结果与理论结果之间具有良好的一致性。数值模拟进一步证实，结果与理论预测相符，证明了该模型在预测紧密排列纱线拉伸行为方面的准确性。这种建模方法有可能显著提高对纺织结构的理解和建模。
