Nanoscale digital image correlation at elementary fibre/matrix level in polymer–based composites

Klavzer, Nathan [UCL] Gayot, Sarah [UCL] Coulombier, Michaël [UCL] Nysten, Bernard [UCL] Pardoen, Thomas [UCL]

Multiscale mechanical modelling aims at predicting the failure of composites from the fibre/matrix level up to the component scale. Existing frameworks are limited by the lack of reliable experimental data and by an incomplete understanding of the submicron deformation and failure mechanisms. A novel digital image correlation (DIC) method has been developed for the characterisation of the nanoscale mechanical response in composites, based on latest advances in surface patterning. Indium has been deposited on unidirectional composites leading to a dense, homogeneous speckle with particle diameter around 15 nm. The specimens were subjected to transverse compression in a scanning electron microscope, while minimising distortion effects. Strain concentration areas, like submicron shear bands and fibre–matrix interphases, were successfully captured for two systems: a carbon fibre-reinforced thermoset and a glass fibre-reinforced thermoplastic. DIC results were compared with alternative experimental data, obtained by atomic force microscopy, and with finite element simulations based on a conventional elastoplastic model.