Elastic properties were predicted for AS4, IM7, T300, and T650-35 graphite fibers. An inverse method was employed using lamina and epoxy properties taken from literature. Fiber properties predicted using finite element analysis based on a hexagonal microstructure, finite element analysis based on a random microstructure, and the Mori-Tanaka averaging scheme were compared. It was observed that the Mori-Tanaka averaging scheme and finite element analysis based on a hexagonal microstructure predicted nearly identical fiber properties. In contrast, randomness in the arrangement of fibers results in significantly different predictions for the longitudinal shear modulus of the fiber. The three microstructural models were also used to predict properties for isotropic E-glass 21xK43. It was observed that the three models predicted nearly the same properties for the glass fibers.

No abstract

The effect of the free-edge on the interlaminar stresses that develop in a thin-ply [0°/90°]s laminated composite under uniaxial tension was explored using finite element models that directly modeled the random heterogeneous microstructure and a model that treated the plies as homogeneous, orthotropic materials. The deformed cross-sections were compared for the two cases, showing that the homogeneous model generally captured the displacement field well but the heterogenous model exhibited local perturbations due to the microstructure. The interlaminar normal stress distributions along the ply interfaces were very different for two models. The heterogeneous model exhibited a complex pattern of stresses that were sensitive to nearby fibers. A comparison of the interlaminar shear stress distributions for the two models showed better agreement, though the heterogeneous model greatly differed from the homogeneous model when fibers or matrix pockets were close to the ply interface. After separating the effect of the fibers and the free-edge effect, the stresses that develop due to the free-edge effect matched well between the two models, except very close to the free-edge. It was shown that the Poisson’s ratios of the fibers and matrix significantly affect the stress distributions along the ply interface, and a matrix Poisson’s ratio could be selected that reduces the effect of the microstructure on the interlaminar normal stress. The study highlighted that a full 3D analysis can provide new insights for classical problems, and more optimal design of composites will require consideration of microstructural effects.

A non-idealized finite element model of a plain orthogonally woven textile composite was subjected to tension along the warp direction, and the predicted stress state was investigated. The effect of refining the geometry and mesh on the volume average stresses and the percentage of each constituent at different stress levels was explored. For the particular textile architecture considered, which consisted of large reinforcement tows and complex tow cross sections, it was shown that the typical mesh refinement in the literature might suffice for volume average stresses, but a higher mesh refinement is needed to accurately capture stress concentrations. The locations of stress concentrations within each constituent were identified. For the three types of tows, [Formula: see text], transverse normal stress in the local coordinate system, in the wefts was predicted to be the most severe component of stress. For the layers of wefts that are crossed over or under by a binder, stress concentrations developed where the warps were the most distorted. Whereas, for the interior layer of wefts, stress concentrations developed where a binder came closest to the weft. In the matrix, [Formula: see text], the normal stress in the direction of the load, concentrations developed where a binder came close to a warp or weft. The locations of peak cross-sectionally averaged stresses along the tow paths were shown to match the locations of local stress concentrations. However, it was observed that many of the stress concentrations might be sensitive to the method used to create the finite element model, boundary conditions, or accounting for the variation of local fiber-volume fraction that results from a variation of cross-sectional area.

The onset and growth of damage in fiber/matrix composites under transverse loads were modelled using cohesive elements and representative volume elements of randomly arranged fibers. Switching between iterative schemes, using an appropriate tolerance and load increment size, and using an extrapolated solution as an initial guess for load increments led to over an order of magnitude reduction in the solution time. The effect of several model parameters on the failure properties for the next larger scale was studied. The crack path did exhibit a dependence on the mesh, but the RVE strength and amount of dissipated energy in the representative volume element did not vary more than 4% for any of the mesh refinements considered. Periodic boundary conditions minimally interfered with the localization of damage when the localized band of damage did not extend across the entire RVE or when the damage naturally localized parallel to a boundary or diagonal of the representative volume element. A local method for quantifying the energy dissipated within the representative volume element was proposed, which provides an improved accuracy and flexibility. An approach to precisely define the dominant crack was given, which allowed the energy dissipate diffusely and along the dominant crack to be separated. It was shown that the predicted critical strain energy release rate for the representative volume element was sensitive to the representative volume element unless the diffusely dissipated energy was accounted for separately. The proposed technique for calculating failure properties within a multiscale framework has the potential to be applied to other damage models.