Influence of loading rate on the delamination response of untufted and tufted carbon epoxy non-crimp fabric composites/Mode II

Abstract

This research investigates dynamic delamination Mode II of carbon Non-Crimp Fabric (NCF) epoxy composites. Tufted and untufted configurations are considered. In order to achieve dynamic Mode II loading a modified end loaded split apparatus (ELS) is mounted on a drop weight impact tower. The high speed events are monitored through load recording and high speed image acquisition. Optical analysis of crack length and displacement combined with specimen geometry information and flexural properties allows the calculation of the delamination force and dynamic strain energy release rate in Mode II (G_IIC). The untufted specimens show unstable failure once damage is initiated. Tufts are beneficial on the dynamic Mode II response as they increased energy absorption considerably and reduced the crack speed significantly. G_IIC for the tufted composites is slightly positively dependent on the ELS specimen loading velocity. Optical analysis proves to be a reliable means to achieve high speed Mode II characterisation, whereas load signals are not sufficiently accurate to provide enough information for possible rate effects detection.

Introduction

Composite shell structures reinforced by spars and ribs are in use in the aerospace, motorsport and marine industry. The structures can reach high speed and can sometimes collide with unexpected objects. Out of plane impact response is dictated by the resistance to delamination in Mode II for low speed impacts. Loading rate effects on the crack propagation in Mode II would therefore influence the response to impact of the laminate and the damage produced. It is consequently of particular interest to predict and understand the dynamic delamination in Mode II of composite laminates.

Textile composites offer reduced manufacturing time over pre-impregnated techniques by the use of resin infusion methods. The use of non-crimp fabrics consisting of two unidirectional (UD) plies stitched together increases the speed of draping and reduces the manufacturing time while maintaining an excellent in-plane response due to the reduced crimp compared to woven materials. However, the response to out of plane loading can be poor as the material is likely to delaminate, which can reduce the structural strength dramatically. Through the thickness reinforcement enhances the resistance to delamination considerably and is under investigation in this work.

Different experimental set ups can provide delamination Mode II measurements but none of them has been standardised. End notch flexural (ENF) specimen [1], end loaded split (ELS) [2] or short length specimens are possible test configurations to quantify delamination in Mode II. Dynamic experiments can be carried out via hydraulic test machines, drop towers or Hopkinson pressure bars. Smiley and Pipes [1] investigated the dependence of Mode II fracture toughness on strain rate via a screw-driven and a hydraulic testing machine (up to 0.092 m/s) with ENF type specimens. It was found that crack growth was changing from ductile to brittle mode as the rate of loading increased. This change resulted in a considerable drop in G_IIC. Nwosu et al. [3] investigated Mode II delamination in graphite fibre epoxy specimens at high strain rate with a Split Hopkinson pressure bar and ENF and centre notch flexure CNF specimen configurations. Results showed that delamination resistance and energy absorbed in fracture increased with impact energy. Wu and Dzenis [4] used a three-point-bending and compact shearing set ups with a Hopkinson pressure bar and found no significant rate effects in Mode II. Cairn [5] used a modified ENF set up in a drop tower and registered superior fracture initiation and higher resistance to propagation under dynamic conditions. Tsai et al. [6] used a modified ENF tests in a drop weight tower in conjunction with a special insert film that delayed crack initiation allowing high crack speed to be reached. Using finite element analysis and experimental results it was concluded that no significant rate effects were achieved for crack speeds up to 1100 m/s. Using a similar set up Cantwell [7] detected positive rate sensitivity. It was also noticed that offsetting the interface by a few degrees would increase the rate sensitivity during testing. Using CNF and FE analysis Maikuma et al. [8] and Caimmi et al. [9] observed decreasing fracture energy for increasing loading rate on carbon composites with screw-driven or drop tower test devices. Caimmi et al. [9] utilised a new specimen set up called CENF (compact edge-notched shear specimen) and found increasing fracture energy with loading rates. With the use of ENF and drop weight loading on glass-fibre/vinyl-ester specimen Compston et al. [10] showed no significant rate effects. In general the results reported in the literature indicate some discrepancy due to the difference of delamination test set up, loading apparatus, and materials tested.

Through the thickness reinforcements by means of stitches were investigated in Mode II, static delamination in carbon epoxy laminates by Wood et al. [11]. It was found that the stitch effects could be measured but did not significantly improved the delamination resistance.

This research investigates the dynamic delamination behaviour of standard and tufted carbon NCF composites in Mode II. Results based on load cell readings and image recordings are presented and compared. The optical technique is utilised to investigate strain rate effects and to understand better the crack loading.