Simulative damage prediction based on material models at micro level for non-crimp fabric laminates (DamageDict)

For the best possible utilisation of the lightweight design potential of fiber-reinforced plastic (FRP) composites, the anisotropic mechanical properties have to be experimentally determined or known precisely. Beneath the anisotropy, the complex multiscale FRP structure leads to mayor challenges. The structure of a FRP component ranges from the single fiber (micro level), the roving, the single layer (meso level) up to the laminate at component level (macro level). For an exact and reliable design of a component, not only the knowledge of fiber-matrix adhesion, fiber orientation and mechanical properties but also the knowledge of the damage and failure behavior are important. Therefore, many different failure criteria exist for different approaches. Easy to use criteria like the maximum strength-/strain criterion are sometimes too imprecise and do not reflect the different failure modes of the composites. Puck's action plane strength criterion, on the other hand, allows the failure of FRP structures to be predicted accurately, but the experimental effort required in advance is enormous to get all necessary material data.

In DamageDict project, a simulation tool is being developed which keeps the computational and experimental effort for dimensioning FRP components low. A multi-scale approach should lead to this by dividing the calculation of the micromechanical properties of the laminate into smaller micro models (s. Fig. 1). These results are taken into account for the computation of the single layer, the results of which are in turn used for component simulations. Starting  at micro level, the material behavior of fibers and matrix is simulated and the crack initiation is predicted in the digital Material-Lab “GeoDict”. By means of tension-compression-torsion tests of unimpaired and specifically damaged specimens (contribution of a definite imperfection), the mechanical properties are characterised. These data are compared with micro sections and in-situ X-Ray tensile tests (s. Fig. 2), and subsequently the fiber structure  is simulated. To optimize the micro models, the properties of the fibers, matrix and fiber-matrix adhesion are additionally determined experimentally. These data are taken into account in the finite element model at meso level (ply level) (s. Fig. 3) which is used to predict the failure of the fiber-plastic composite. Here, a stitching as a local weakness point is simulated where damage preferentially occurs and can be detected. With the micro models and the finite-element model of the laminate, a precise component design can be realised, taking into account the material properties down to the fiber level

The project ”DamageDict – Simulative damage prediction based on material models at micro level for scrim laminates“ is funded by the federal ministry of economic affairs and climate action within the framework of the “central innovation program for SMEs” (funding reference ZF4052328LF9).

Contact:
Christian Becker, M.Sc.
Design of Composite Structures
E-Mail: christian.becker@ivw.uni-kl.de
Tel. 0631 2017 206