The response of stiffened aluminium panels subjected to extreme pressure loading was studied experimentally and numerically. An accurate description of these load conditions and how they interact with structures is critical in the design of protective buildings and in many other applications. This thesis is part of an ongoing research project at SIMLab, NTNU, that investigates the effective action of extreme surface pressures on plated structures.
The SIMLab Shock Tube Facility was used to generate shock waves similar to those from far field explosive detonations. A recently designed rig defined the spacial settings and conditions for the test specimen - a welded panel with three stiffeners, all made of the aluminium alloy AA6082-T6. The panels were were exposed to load intensities of various magnitudes resulting in different responses and levels of deformation. The loading was obtained by using similar tests on a rigid plate equipped with pressure sensors. The displacements of the panels was experimentally measured by use of high-speed cameras and digital image correlation methods. A new design was successfully implemented for measuring reaction forces in the supports using load cells.
Numerical models of the experiments were established using an uncoupled approach following the Eulerian-Lagrangian method. The simulated pressure loads were generated in EUROPLEXUS and prescribed to the Lagrangian model of the stiffened panel in ABAQUS/Explicit. A material model was adopted from previous work on the panels, using a Modified Johnson Cook constitutive equation with Voce hardening rule. Material parameters for the aluminium alloy were identified from tensile tests.
The tests covered a range of responses, from purely elastic response to inelastic deformation and failure of the panel. Comparison of the experimental and numerical results showed a a general good agreement of the data, where the model was able to predict most of the observed events in the experiments. The numerical results had slightly lower displacement in some tests, and not all modes of failure was replicated. The reaction forces had oscillations that seemed to match in frequency with the displacement results. The total reaction forces in the supports were especially comparable with the measurement results.
The pressure loads generated by the Eulerian simulations in EUROPLEXUS experience somewhat lower magnitudes in pressure than those obtained with experiments on rigid plates. Suggested sources of error were the method of applying the load to the Lagrangian model, and an insufficient mesh discretization for the fluid subdomain. A parameter study on the effects of load, mesh, element type, and material parameters was performed. A marked improvement in both displacement and reaction force was found when the simulated load was increased to match the measured.