No abstract

Modelling of the onset and propagation of matrix cracks in fibre-reinforced composites on the micro scale requires adequate representation of the material microstructure. In the present work we compare simulated and real fibre arrangements found in unidirectional composites, using statistical descriptors. The comparison is done for geometrical and mechanical parameters such as distributions of the fibre positions and the stress fields. The real fibre arrangements are extracted from microscopy images of a 3D non-crimp woven carbon/epoxy composite with fibre volume fractions of 58-68%. The modelled fibre arrangements are generated using a heuristic random microstructure generation algorithm. The stress fields are compared for the case of transverse tension. A good correlation between statistical parameters of the real and simulated fibre arrangements is obtained.

No abstract

Steel fibers, with their high stiffness and high ductility, have a potential to provide a new range of properties in polymer composites, in comparison with carbon and glass fiber composites. However, the high stiffness contrast between the steel fiber and the polymer matrix plus the fiber’s non-circular cross-section are likely to generate high stress concentrations in a composite under transverse loading. In the present study, these stress concentrations are analyzed using finite element modeling and compared with the case of carbon and glass fiber composites. The study is performed for an isolated fiber and multiple fibers in hexagonal and random packings with 40% and 60% of fiber volume fractions. According to the results, in spite of a high contrast between the stiffness values of steel and glass fibers, no significant difference between the transverse stress concentrations was observed for steel and glass fibers in the hexagonal packing due to the difference in material properties. Differences in stress concentrations were noted for the case of randomly packed fibers. The polygonal cross-section of steel fibers was found to introduce extreme stress concentrations.

No abstract