A global-local superposition-based beam finite element for predicting mode I delamination
摘要
This study proposes a quasi-2D composite beam finite element based on a global–local superposition, combined with an interlaminar damage model, with the aim of accurately assessing, predicting, and evaluating delamination in composite laminates, as well as structural health, under both environmental and operational conditions. The changes in the mechanical properties of the laminated material induced by delamination are computed as a function of the interfacial displacements The main advantage of this formulation lies in its computational efficiency, ensured by the global-local superposition of displacement fields, which not only captures the expected zig-zag behavior of stresses and displacements, but also makes the total number of degrees of freedom (DOFs) independent of the number of layers. This is achieved because the local DOFs, whose quantity is related to the interpolation functions in each layer, are efficiently eliminated through the application of boundary conditions and the CZ0continuity criteria, allowing the displacements and stresses to be determined exclusively by the DOFs used in beam elements. In this context, the finite element (FE) formulation is applied only to mode I delamination (opening) with a bilinear loading–unloading curve, considering cases of different interlaminar thicknesses, loading conditions, and environmental situations. It is worth noting, however, that the new proposal enables the inclusion of other delamination modes as well as formulations based on different forms of stress–displacement curves for loading and unloading. Finally, the proposed beam element provides a concise, detailed, and efficient analysis of laminated structures of varying complexity, delivering relevant and consistent results regarding structural damage behavior. Moreover, it enhances damage predictability, contributes to structural health monitoring, and improves design efficiency in composite structures, since the algorithm is simple to implement, requires low computational cost, and offers results comparable to commercial software for two-dimensional analyses, making it an excellent economic and effective option.