Characterisation and modelling of the variability of mechanical properties of natural fibre composites.

Composite materials offer excellent opportunities for modern and environmentally friendly design of structural components. The combination of very good to superior mechanical properties and low mass density with a wide range of reinforcement architectures and processing and forming technologies gives the designer attractive options for highly tailored and optimised ecient structures.

A wide range of material families are available forfibre reinforcement, and natural fibres provide particular advantages with respect to the environment. Much research and development have already been done on the characterisation of different types of natural bres (ax, hemp, jute, sisal, bamboo, ldots), with remarkably high levels for brefistifness and strength. Specific strength is in the same range as E-glass and specic stifness is even better. However, when compared to the commonly used man-made fibres like glass and carbon, mechanical properties exhibit a very high degree of scatter, with intervals where the upper bound is sometimes twice as high as the lower bound, even for bres of the same type. This is a significant problem for the quality of the final product, as variability may become unacceptably high

 The objective of the proposed project is to quantify variability at different levels of the full cycle of product design, starting with the extraction of the technical fibre, over the composite processing technology, to the macroscopic properties of the product, using experimental identification and numerical analysis. The envisaged model is a quantified random field of the macroscopic component stifness tensor, based on multi-scale modelling. Vibration characteristics of a composite plate will be used as indicators to tune plate stifiness properties.