Towards accurate and efficient single fibre characterization to better assess failure strength distribution

Abstract : Failure in unidirectional organic composites, loaded longitudinally, usually originates within fibres, which makes their failure kinetics a key parameter for failure modelling. Correctly capturing fibre strength statistics and failure kinetics is crucial to the success of composite stochastic strength models. The characterization of fibres is challenging, as the fibre diameter can be just a few microns. Nevertheless mechanical testing of single fibres is the only unambiguous means of characterising fibres or exploring their morphologies. Despite great progress in characterisation techniques, many obstacles remain to obtain accurate fibre strengths. Most studies have acknowledged the difficulty in obtaining fibre strength data which can accurately represent the entire fibre population in a given application. In this study an automated single fibre tensile testing machine has been used which overcomes most obstacles usually faced during fibre strength determination. Advanced statistical data processing techniques have been implemented to determine the reliability of the obtained results. Such analysis permits the calculation of confidence/accuracy associated with the generated fibre strength data. This may be used to extract more useful information from the limited resources available. The inadequacy of the popular statistical distributions for representing fibre strength has been highlighted. It is shown that inclusion of a trunca-tion parameter in the Weibull distribution could help in estimating the strength behaviour of a given fibre population.