Return to search

3D imaging of the tensile failure mechanisms of carbon fibre composites

Synchrotron radiation computed tomography (SRCT) has been used to analyse the tensile failure mechanisms in carbon fibre/epoxy composites. Two specimen types were analysed – in situ loaded coupons and filament wound samples, taken from incrementally loaded cylinders and scanned “post mortem”. The effects of fibre, matrix and interfacial properties on the initiation and accumulation of fibre breaks have been analysed. Breaks accumulated on a power law curve as a function of fibre stress; however the fibre and matrix moduli had little effect on accumulation. Initial analysis of the fibre Weibull moduli showed little correlation between Weibull modulus and break accumulation. Singlets initiated in low fibre volume fraction areas; however a full investigation into the effects of varying fibre volume fraction has not been possible. Attention was focused on the formation of interacting groups of broken fibres (clusters), as they are believed to be the strength-defining failure event. The in situ coupons had much larger maximum cluster sizes than the filament wound counterparts (14 vs. 9), and a correlation between high break density and low cluster percentage is proposed. No simple correlations were found between fibre/matrix moduli and the clustering parameters. Clusters formed in one load step, and did not grow from singlets or smaller clusters, which suggests a dynamic process. The interface is suggested to be key to damage initiation and propagation. The work provides links between experimental studies and simulation tools by informing and validating a micromechanical tensile failure model. Comparisons between experimental and modelled results found that the model accurately predicted the composite failure strain but not the complex damage accumulation processes. The model under-predicted both cluster size and the proportion of interacting breaks; this is attributed to the inaccurate modelling of the stress transfer process. Both experimentally and analytically the dominant parameter controlling clustering was the overall stress concentration factor. This has been infrequently analysed in work published in the literature, and is the recommended focus of the future work.

Identiferoai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:628703
Date January 2014
CreatorsMorton, Hannah
ContributorsReed, Philippa
PublisherUniversity of Southampton
Source SetsEthos UK
Detected LanguageEnglish
TypeElectronic Thesis or Dissertation
Sourcehttps://eprints.soton.ac.uk/366506/

Page generated in 0.0103 seconds