A novel modeling method for the mechanical behavior of 3D woven fabrics considering yarn distortion

Mengran Li,Kai Liu,Jingran Ge,Junbo Xie,Zengfei Liu,Binbin Zhang,Jian Huang,Jun Liang
DOI: https://doi.org/10.1016/j.compscitech.2022.109691
IF: 9.1
2022-11-10
Composites Science and Technology
Abstract:Three-dimensional woven fabrics feature complex meso-architectures, such as yarn torsion and cross-section variation. This paper aims to developing a novel numerical modeling method to simulate the mechanical behaviors of reinforcements considering realistic yarn geometries. The initial as-woven unit cell is first simulated by the digital element analysis. A reconstruction algorithm is then used to generate mesoscale high-fidelity solid geometric models from digital chains. Fiber directional orientation within torsional yarns of the high-fidelity model is described by a digital-chain tracing method and analyzed in detail. Furthermore, based on the high-fidelity model, a hyper-elastic constitutive model is employed for the numerical analysis of the mechanical properties of reinforcements. The finite element analysis (FEA) of transverse compression, tensile deformation and in-plane shearing deformation combined with in-situ Micro-CT experiments are carried out to verify the method. The proposed method agrees well with experimental tests. Defining fiber orientations of torsional yarns using the digital-chain tracing method leads to more accurate simulation results. The developed numerical method can be extended for the composites design and structure optimization due to its high-fidelity simulation capabilities.
materials science, composites
What problem does this paper attempt to address?
This paper aims to address the challenges faced in the mechanical behavior simulation of 3D braided fabrics. Specifically, the paper attempts to solve the following problems: 1. **Complex microstructure**: 3D braided fabrics have complex meso - structures, such as yarn twisting and cross - sectional changes. Traditional idealized unit cell models cannot accurately capture these complex structures, resulting in reduced accuracy of mechanical property prediction. 2. **High - fidelity modeling of yarn geometry**: Existing geometric modeling methods (such as automatic reconstruction based on Micro - CT images) can capture the real yarn structures of dry - state 3D braided fabrics, but are not suitable for pre - design analysis of new fabrics and are costly. Therefore, a method that can accurately simulate the yarn geometry at the design stage is required. 3. **Accurate description of fiber orientation**: For complex 3D braided fabrics, yarn twisting can cause the fiber orientation to be unevenly distributed within each yarn segment. If the fiber orientation is defined only according to the yarn path, there will be a deviation from the actual situation, thus affecting the prediction accuracy of the fabric's mechanical response. To solve the above problems, the paper proposes a new numerical high - fidelity modeling method. The specific steps are as follows: - **Digital element analysis of the initial braided geometry**: Use a Python script to establish an initial loose digital element unit cell model of the 3D braided fabric, and obtain the actual braided geometry through digital element analysis. - **Reconstruction of a solid geometric model from digital element chains**: Based on the data of digital element chains, generate a high - fidelity solid geometric model. Project nodes onto a plane and use the alpha - shapes algorithm to generate cross - sectional contours, and finally construct six - node prism elements. - **Digital chain - tracing method for fiber orientation**: Through the digital chain - tracing method, accurately describe and analyze the fiber orientation within twisted yarns to ensure the accuracy of the material orientation. - **Numerical analysis of the hyperelastic constitutive model**: Based on the high - fidelity model, use the hyperelastic constitutive model for numerical analysis to simulate the mechanical behavior of the fabric. The effectiveness of this method has been verified through lateral compression, tensile deformation, and plane shear deformation experiments. In conclusion, the method proposed in this paper can not only more accurately simulate the mechanical behavior of 3D braided fabrics but also provides strong support for the design and structural optimization of composite materials.