The mesoscale model with realistic yarn geometries could improve the accuracy of property prediction of woven composite structures. This paper aims to propose a novel mesoscale modeling method for the woven composite twisted structures. Firstly, the torsional deformation of the yarns in the unit cell model is simulated with high-fidelity using digital element fibers embedded into the matrix solid meshes. Then, the solid geometry model of the twisted composite structures is reconstructed from the digital element deformation model by combining the affine transforms. Finally, the elastic responses of the twisted composite structure under cantilever loading are predicted by assigning the element information of the established mesoscale model to the macroscale model. It is shown that the geometric morphology and predicted mechanical responses of the high-fidelity model are consistent with the experimental results. Furthermore, the effect of twist angles on the stiffness properties of the twisted composite structures is analyzed using the presented numerical method. The proposed digital modeling method could provide a guide for the early design of woven composite structures.

中文翻译：

---

结合数字元素嵌入模型和仿射变换的编织复合扭曲结构的介观建模

具有真实纱线几何形状的介观模型可以提高机织复合材料结构性能预测的准确性。本文旨在提出一种新型的编织复合材料扭曲结构的介观建模方法。首先，使用嵌入矩阵实体网格的数字元素纤维对晶胞模型中纱线的扭转变形进行高保真度模拟。然后，结合仿射变换，根据数字单元变形模型重建扭曲复合结构的实体几何模型。最后，通过将建立的细观模型的单元信息分配给宏观模型，预测悬臂载荷下扭转复合材料结构的弹性响应。结果表明，高保真模型的几何形貌和预测的力学响应与实验结果一致。此外，利用所提出的数值方法分析了扭转角对扭转复合材料结构刚度性能的影响。所提出的数字建模方法可以为编织复合材料结构的早期设计提供指导。