A combined experimental-numerical approach for generating statistically equivalent fibre distributions for high strength laminated composite materials

Abstract

A technique is presented where actual experimental distributions, measured from a high strength carbon fibre composite, are considered in the development of a novel method to generate statistically equivalent fibre distributions for high volume fraction composites. The approach uses an adjusted measure of nearest neighbour distribution functions to define inter-fibre distances. The statistical distributions, characterising the resulting fibre arrangements, were found to be equivalent to those in the actual microstructure. Finite element models were generated and used to determine the effective elastic properties of the composite and excellent agreement was obtained. The algorithm developed is simple, robust, highly efficient and capable of reproducing actual fibre distributions for high strength laminated composite materials. It does not require further heuristic steps, such as those seen in fibre stirring/shaking algorithms, in order to achieve high volume fraction microstructures and provides a useful alternative to both microstructure reproduction and random numerical models.