High-fidelity modeling of yarn nesting in plain-woven composites using virtual fiber and voxel methods

Compaction-induced yarn nesting in multilayered plain-woven composites leads to complex mesostructures, which complicate accurate mechanical property prediction. This study develops an integrated virtual fiber method (VFM) and voxel-based finite element framework for nested multilayer plain-woven composites. The compacted yarn geometry is first obtained through virtual-fiber compaction simulation. The deformed virtual-fiber information is then converted into voxel-level material fields. In this virtual fiber-to-voxel conversion, the local fiber volume fraction and local fiber orientation are calculated for each voxel and used to assign spatially varying yarn properties and material orientations. The framework is further combined with progressive damage analysis to study the effect of yarn nesting on stiffness, damage evolution, and tensile strength. An automated software platform is developed to connect virtual fiber generation, compaction simulation, virtual fiber-to-voxel conversion, material assignment, and finite element model generation. A parametric study of 64 nesting configurations shows that nesting patterns affect both elastic modulus and tensile strength, with a stronger effect on tensile strength. The random nesting simulations using 4-layer periodic models and a full-thickness 11-layer model show good agreement with tensile experiments, with the 11-layer model giving relative errors of 1.2% for modulus and 3.9% for strength.