International audienceThis paper investigates a computational strategy for studying the interactions between multiple through-the-width delaminations and global or local buckling in composite laminates taking into account possible contact between the delaminated surfaces. In order to achieve an accurate prediction of the quasi-static response, a very refined discretization of the structure is required, leading to the resolution of very large and highly nonlinear numerical problems. In this paper, a nonlinear finite element formulation along with a parallel iterative scheme based on a multiscale domain decomposition are used for the computation of 3D mesoscale models. Previous works by the authors already dealt with the simulation of multiscale delamination assuming small perturbations. This paper presents the formulation used to include geometric nonlinearities into this existing multiscale framework and discusses the adaptations that need to be made to the iterative process in order to ensure the rapid convergence and the scalability of the method in the presence of buckling and delamination. These various adaptations are illustrated by simulations involving large numbers of DOFs

No abstract

This paper is devoted to the study of a micro-macro LaTIn-based Domain Decomposition Method for which the partitioning, the geometry and the boundary conditions play a major role in the number of iterations to convergence and in the scalability. To confront these obstacles, an analysis of the macroscopic space and of the search direction -two parameters of the strategy -is proposed for traction, bending and buckling examples. Then, we propose a new search direction which takes into account the global stiffness of the structure and that limits the need to enrich the macroscopic space. This choice leads to a minimal number of iterations, a reduced computation time and the scalability of the strategy. The enhanced parameter is then applied to the simulation of combined buckling and delamination of a composite 3D plate.

No abstract

In this work, a numerical-experimental study of the interlaminar zone for an unidirectional glass fiber reinforced epoxy composite is carried out in order to predict the load-displacement curves of a double cantilever beam test. First, an experimental mechanical characterization of the laminated composite was made through quasi-static in-plane tensile and bending tests and out-of-plane delamination tests (i.e. double cantilever beam tests or opening mode I). The main results have shown that the elastic module in the fiber direction is E 1 ¼ 32.1 GPa and the fracture process is characterized by a critical energy release rate G IC ¼ 1466 N/m. In order to be able to predict delamination using finite element analysis, a bilinear cohesive zone model is adopted. This law has two parameters-the initial stiffnessK and the critical traction t n0-to be fit for reducing the distance between numerical and experimental double cantilever beam load-displacement curves. That means an optimization problem, which is here solved by proposing a very simple and cheap three-step procedure, avoiding expensive three-dimensional simulations: (i) the numerical double cantilever beam test is coded in the OCTAVE software using one-dimensional beam elements and the bilinear cohesive zone model; (ii) several onedimensional simulations are performed varying at the cohesive law's parameters in order to build the objective function; and (iii) a genetic algorithm from the Scilab optimization toolbox is then used to determine the interface's parameters minimizing the objective function. This method has proposedK ¼ 2.74 Â 10 14 N/m 3 and t n0 ¼ 1.99 MPa as the best parameters fitting the curves. Finally, the identified bilinear cohesive law's parameters are used to carry out a full three-dimensional finite element analysis simulation in ANSYS, showing the same response than the simplified beam code, but with the possibility to obtain more accurate information about the fracture process (e.g. the shape of the process zone) or to use more complicated geometries or boundary conditions. In this work, we show the ability to predict the material behavior through different methods, using the best characteristics of each one.