Answers are grounded in this paper and cite their sources.
The Virtual Element Method (VEM) has proven to be an effective framework for geometrically exact beam analysis. However, existing formulations are restricted to straight beam elements, requiring initially curved geometries to be approximated by polylines. This approximation significantly affects the convergence properties of the method, particularly when coarse meshes are used. This work presents a Virtual Element formulation for two-dimensional geometrically exact Reissner beams with arbitrary initial curvature. The proposed approach embeds the description of the beam axis directly into the element formulation, allowing the actual initial geometry to be directly incorporated at the element level. The virtual approximation spaces, projection operators, and numerical integration are consistently reformulated in the curved reference geometry, while the overall structure of the standard VEM formulation is preserved. Since all geometry-dependent quantities are evaluated only during the preprocessing phase, the additional implementation effort and computational overhead with respect to straight-axis formulations remain limited. The formulation is validated through several nonlinear benchmark problems involving initially curved beams subjected to large displacements and rotations. A comparison with the corresponding straight-axis formulation shows that directly accounting for the initial curvature significantly improves the convergence behavior, particularly for coarse discretizations. The proposed formulation accurately reproduces the structural response and recovers the optimal convergence rates. The influence of the numerical quadrature rule is additionally investigated, revealing a curvature-induced locking mechanism and showing that optimal convergence is obtained through an appropriate reduced integration, with the number of Gauss points matching the polynomial degree adopted for the displacement projection.
No takes yet. Share an insight, caveat, or question.
Chirianni et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: