A mapping-based method capturing the mesoscopic morphological characteristics of 3D woven fabric torsion structures

Upon torsional molding, 3D woven fabrics accommodate a complex mesoscopic morphology that significantly governs the mechanical properties of composite structures. This paper is aimed to propose a macro-mesoscopic mapping method for characterizing the mesoscopic morphology of 3D woven fabric torsion structures. To this end, hyperelastic constitutive molding simulations of fabrics are performed to determine the macroscopic deformation profile within the structure in first. Then the mesoscopic details are resolved with the unit cell deformed accordingly at each digital element. A macro-mesoscopic mapping relationship is thus established by constructing a multiscale topological description function so as to tune the mesoscopic unit cells in consistency with the macroscopic manifold. Finally, the collection of mapping data is utilized for geometric reconstruction, establishing a mesoscale geometrically refined descriptive model about the torsion structure. The mesoscopic morphology of the reconstructed model is shown to agree well with experimental results. Moreover, the mapping method offers a significant time-saving strategy as the calculation time is only within a few minutes on a personal computer, in comparison with existing treatments necessitating 2-day computation on a workstation against similar tasks. The applicability of the present method to resolve other multiscale manifolds is also demonstrated.