COMPLAS 2025

Multiscale Modeling of Textile Yarns and Nets via Particle-based Equilibrium and Finite Element Analysis

  • Gomez, Javier (ADVANCED MATERIAL SIMULATION)
  • Oroya, Jesús (ADVANCED MATERIAL SIMULATION)

Please login to view abstract download link

Contact‐rich textile structures pose significant challenges for finite element (FE) simulations, primarily due to the multitude of contact interfaces and the need to accurately capture unknown equilibrium positions of numerous filaments. 3D continuum approaches become computationally prohibitive when each filament must be discretized and contact constraints tracked from uncertain initial configurations. To address these limitations, one‐dimensional finite element formulations have been extensively used [1, 2]. This paper proposes an alternative hybrid methodology that integrates discrete particle simulations, using packages such as LAMMPS, with continuum‐based FE analysis. In the first stage, each filament within a braided yarn or net is represented by a chain of discrete particles. This particle‐based model accounts for axial and bending stiffness, as well as frictional contact, allowing localized rather than global contact detection. Consequently, the system converges rapidly to an equilibrium arrangement that accurately reflects real-world contact interactions, yet with markedly reduced computational overhead compared to large-scale continuum formulations. Once an equilibrium is reached, the geometry of the filaments, along with any residual stress state, is exported into a Finite Element framework. Here, a two-steps simulation is conducted. First, a preliminary tying analysis lightly constrains the geometry to preserve the equilibrium positions, minimizing large and unpredictable displacements. Second, a fully coupled FE analysis applies user-defined mechanical loads to evaluate structural behaviour in a more traditional continuum setting. By commencing from a realistic configuration, the number of contact interactions in the FE solver remains tractable and more robust. Numerical examples on braided yarns and fishing nets demonstrate that this approach not only predicts equilibrium shapes more efficiently than conventional continuum methods but also delivers reliable mechanical assessments.