The present thesis investigates the performance of current aerospace structural materials such as Glare, a fibre metal laminate, to the catastrophic consequence of sabotage-induced blast loadings on commercial aircraft. The aim is to quantify the effects of these blast events and establish if remedial action can in some manner increase the chances of aircraft survivability.
Within the EU funded VULCAN consortium, a coordinated effort has been devised to determine the dynamic deformation and fracture behaviour of structural materials subject to blast loadings using both experimental and numerical techniques. Test data from small-scale experimental blast trials have been verified and validated by the author using robust and efficient finite element models. Numerical studies have shown that Glare has potential to be a strong candidate for blast attenuating structures, exhibiting superior blast resistance compared to monolithic aluminium plates. Furthermore, a blast vulnerability and survivability analysis was devised to illustrate various failure scenarios in scaled fuselage structures.
To address the macroscopic crack propagation in large-scale shell structures to blast loadings, well-controlled dynamic fracture experiments have been performed. This configuration, which consists of closed-end pressurised barrels with a through-thickness crack, is designed to capture the underlying dynamic phenomena under investigation whilst keeping the computational effort manageable. Quantitative fracture metrics obtained from high speed imaging systems have shown that Glare exhibits much lower average crack velocities than Aluminium 2024-T3 and CFRP.
Experimental boundary and loading conditions served as well-defined input parameters to large-scale finite element models using cohesive elements. It has been shown that rate-independent cohesive models, initially verified using quasi-static fracture toughness tests, are insufficient to capture the dynamic crack growth rates. Alterative rate-dependent models have been discussed and implemented which take into account the influence of loading rate on the cohesive traction and energy dissipation. An inverse problem of cohesive zone modelling is performed to obtain mode-I cohesive zone laws. The comparison shows that both the experiments and the numerical simulations result in very similar crack initiation times and produce crack tip velocities of acceptable agreement.
