Adaptive Delamination by S-method A methodology aimed at addressing computational complexity of analyzing delamination in large structural components made of laminated composites is proposed. The classical ply-by-ply discretization of individual layers may increase the size of the problem by an order of magnitude in comparison to the laminated shell or plate element meshes. The manuscript features delamination indicators that pinpoint the onset and propagation of delamination fronts with striking accuracy. Once the location of delamination has been identified, the discrete solution space of the classical laminated plate/shell element is hierarchically enriched by a combination of weak and strong discontinuities to adaptively track the evolution of delamination fronts. The so-called adaptive s-method proposed herein is equivalent in terms of approximation space to the extended finite element method, but offers sparser matrices and added flexibility in transitioning from weak to strong discontinuities. Numerical examples suggest that despite an overhead that comes with adaptivity, the adaptive s-method is computationally advantageous over the classical ply-by-ply discretization especially as the problem size increases.
A methodology aimed at addressing computational complexity of analyzing delamination in large structural components made of laminated composites is proposed. The classical ply-by-ply discretization of individual layers may increase the size of the problem by an order of magnitude in comparison to the laminated shell or plate element meshes. The manuscript features delamination indicators that pinpoint the onset and propagation of delamination fronts with striking accuracy. Once the location of delamination has been identified, the discrete solution space of the classical laminated plate/shell element is hierarchically enriched by a combination of weak and strong discontinuities to adaptively track the evolution of delamination fronts. The so-called adaptive s-method proposed herein is equivalent in terms of approximation space to the extended finite element method, but offers sparser matrices and added flexibility in transitioning from weak to strong discontinuities. Numerical examples suggest that despite an overhead that comes with adaptivity, the adaptive s-method is computationally advantageous over the classical ply-by-ply discretization especially as the problem size increases.
