Abstract: The bending performance of fabrics is one of the key factors determining their deformation ability. Figuring out the bending mechanical properties of fabrics is crucial for designers and engineers to make accurate predictions and optimizations at the early stages of product development. By establishing simulation models to mimic and analyze the impact of different design parameters on fabric bending performance, the structure, materials, and processes of fabrics can be optimized, thereby enhancing the product's overall performance and quality. Furthermore, finite element simulation of fabric properties can significantly shorten the research and development cycle and reduce experimental costs. Therefore, establishing predictable finite element models for fabric bending at the mesoscopic level to study the bending performance of fabrics can better meet the needs of modern product development. High-performance fabrics like Kevlar® and carbon fiber fabrics, celebrated for their strength and lightweight attributes, are extensively used in multifunctional products such as protective gear and sports equipment. A deeper understanding of the bending performance of these high-performance materials can further elevate the design and manufacturing standards of multifunctional products to meet complex application requirements.
 Taking the high-performance Kevlar® plain woven fabric as the research object, this paper proposes a mesoscopic finite element modeling method based on the sag bending of the fabric under its self-weight to simulate the bending mechanical response at the yarn level. The preprocessing module of the ABAQUS finite element software is utilized to model the buckling of the yarns and assemble them into plain woven fabrics. In this study, a linear elastic orthotropic material model is utilized, and linear reduced integration solid elements are meshed, followed by a mesh sensitivity analysis. A mesh size of 0.25 mm × 0.25 mm is ultimately selected. The fabric model is subjected to fixed boundary conditions and a global gravity load to simulate the bending of the fabric under gravity. Due to the complex contact nonlinearities between the yarns in the mesoscopic model, an explicit solver is used for the calculations. However, to mitigate numerical oscillations caused by the stability limitations of the explicit algorithm, damping is added as a control measure in this study. Through the aforementioned method, the bending of a 7 cm Kevlar® fabric under its self-weight is successfully simulated, and the droop displacement and deformation angle at the pendant end are extracted and compared with experimental results. The differences are found to be within 6%, and the bending shapes from the simulation and experiments are basically consistent, demonstrating good simulation results. To validate the model's generalizability, an 11 cm Kevlar® fabric model and a 9 cm glass fiber fabric model are also created. The droop displacement and deformation angles of these fabric models fall within the experimental error range, and the bending shapes of the models closely match the experimental results after contour similarity matching, indicating that the established fabric bending model possesses good generalizability.
 The mesoscopic model of fabric self-weight bending established in this paper provides a modeling approach for studying fabric bending performance at the yarn level, which helps to minimize the consumption of experimental fabrics and shorten the product development cycle.

Key words: self-weight bending, mesoscopic simulation, finite element, Kevlar? fabric, damping

摘要： 为深入理解织物弯曲性能，优化织物设计，建立了基于纱线级别的织物自重弯曲细观有限元模型。采用弹性正交各向异性材料模型，划分线性缩减积分实体单元，并对织物模型施加全局重力载荷和固定边界条件；通过添加阻尼控制显式求解出现的数值振荡，对比了有无阻尼对织物弯曲模拟的影响，模拟了7 cm长度的Kevlar®织物自重弯曲，发现有限元模型下落位移和变形角度与实验平均值差异小于6%；并通过轮廓相似度对比，显示织物有限元模型与试验弯曲形态基本一致；进一步建立了11 cm长度的Kevlar®织物和9 cm长度玻璃纤维织物弯曲模型，验证了模型的泛化能力，为从细观尺度研究织物弯曲性能提供了建模思路。

关键词: 自重弯曲, 细观模拟, 有限元, Kevlar?织物, 阻尼