Damage and strain rate optical characterisation of standard and tufted non crimp fabric carbon composites for Meso-scale impact models

As global warming is a prime concern to the wellbeing of the planet, lighter planes are a requirement to reduce CO2 emissions. Light structures made of carbon epoxy composite materials are of particular interest but are sensitive to impacts such as hail or bird strikes. Static and dynamic testing of composite specimens and structures (from aeronautical standard) through novel testing methods was the first aim of this PhD research. Subsequently it led to novel material characterisation and material parameters calibration for which numerical simulations of impact responses could be developed. In this research further investigation under static, dynamic and impact loading of two certified aeronautical materials occurred. Carbon non crimp fabric epoxy (tufted and untufted) response was investigated. Novel tests and testing methods were developed for in-plane and for delamination focussing on the use of optical analysis using digital image correlation (DIC) and high speed cameras. A novel damage detection method was proposed using DIC. The experimental data set was used to calibrate a damage model with imposed strain rate laws and added delamination Mode I and II interface prior to a punch validation study. A novel compression apparatus designed for DIC usage worked well in static and dynamic. A novel intermediate strain rate tensile test worked better on bias direction lay-up than on axial one. Dynamic DIC method proved of interest to record strains up to strain rate achieved with a Split Hopkinson pressure bar apparatus. In quasi static tufting reduced axial properties considerably but had little effect in shear loading, in addition it increased significantly the resistance to delamination and reduced the crack speed in dynamic. The damage fields generated allowed for the detection of damage progression for various load cases. More damage occurred in compression and shear than in tension as the tufted laminate showed more pronounced damage than the untufted one. The dynamic effect of tufting on in-plane and impact response was reduced as it increased considerably delamination resistance in Mode I and II. For both tufted and untufted NCF composites, strong strain rate effects were detected from a low speed on the in–plane strength and failure strain as no or little effects were recorded on the material stiffness. Novel dynamic delamination Mode I and II tests combined with optical analysis provided possibilities to detect rate effects and crack speed propagation while loaded in pure mode I and II. No strain rate effects were recorded in delamination Mode I apart from a slight effect during crack initiation which was stronger for the tufted material. In Mode II a slight rate effect was detected for the tufted interface during crack propagation. During out-of-plane impact loading at intermediate speed, a minor negative loading rate effect was detected.The model calibrated in damage, delamination and strain rate prove useful for dynamic DCB representation and assessment of possible mix mode crack loading. Modelling tufts as P-link was of interest but requires further investigation. Damage and strain rate was well modelled in tension, compression and bias direction loading, even if the strain rate shear law would require some modification. A Meso-scale model was validated successfully by means that the model responses would follow the experimental trends in quasi static loading but with the modification of 4 parameters among ±50. Further research could extend its use for impact modelling. This research showed the complexity of the failure mechanism occurring in composite materials, modelling them at high speed in the plane and in out of plane impact remains a difficult challenge. Carbon composites damage sensitivity is significant and invisible to the naked eye for some load cases and lay-ups necessitating regular non destructive testing on aging airframe.