This paper presents the work on six single-stringer specimens manufactured using the card-sliding technique with non-crimp fabrics and adopting a Double-Double (DD) stacking sequence. These specimens, representative of sub-structure level components, are used to investigate post-buckling and failure in aerospace structures. Two specimens maintain a constant thickness cross-section, while four are tapered, two of which incorporate a Teflon insert in the stringer flange. All specimens are tested under compression loading conditions, inducing skin buckling, skin-stringer separation, and eventual collapse. Numerical simulations are validated by experimental results and serve to analyze the specimens behavior and the failure mode. The load versus displacement curves of both experimental tests and Finite Element Method (FEM) analyses are compared, along with the out-of-plane displacement field. Subsequently, the observed failure modes are discussed, focusing on the various mechanisms that occurred and considering the impact of flanges and stiffener tapering. Both the FEM simulations and experimental tests demonstrate good agreement, with the flanges tapering revealing notable results. This offers promising evidence of a viable solution to optimize aeronautical structures and enhance resistance to skin-stringer separation.

In this research, conduction welded C-struts, part of a thermoplastic composite fuselage designed and manufactured in the framework of the Clean Sky 2 STUNNING project, are investigated. Five specimens made of two C-section profiles are manufactured and welded using conduction welding in three different configurations with variations in the direction and distance of the two welded joints. Preliminary numerical analysis using the virtual crack closure technique are conducted to obtain an initial evaluation of the specimens behavior, in preparation of the tests. Experiments are performed under quasi-static loading conditions to measure the strength of the welds. Comparisons with the preliminary numerical analyses show a good agreement in terms of the predicted maximum load, while a clear difference is observed in the initial stiffness, due to the compliance of the support structure. The numerical model is updated, leading to results that closely match the experimental behavior. For all the analyzed specimens, the separation occurs suddenly and no signs of propagation are observed. Experimental and numerical data show no relevant difference in the joint strength among the different conduction welding configurations.

This research proposes a computationally efficient methodology using a Constrained Variational Asymptotic Method (C-VAM) for non-linear buckling analysis on a hat-stringer panel with delamination defects. Starting with the geometrically non-linear kinematics, the VAM procedure reduces the three-dimensional (3-D) strain energy functional to an analogous 2-D plate model and evaluates the closed form warping solutions. Utilising the resulting warping solutions and recovery relations for the skin and the stringer, displacement continuity at the three-dimensional level is enforced between the stringer and the skin based on the pristine and delaminated interface regions. Consequently, the constrained matrices obtained from C-VAM is incorporated into an in-house developed non-linear finite element framework. Using the developed formulation, a stiffened panel with delamination of 40 mm between the stringer and the skin is analysed under compression. The results have been validated locally and globally, employing experimental data and 3-D finite element analysis (FEA). Experiments are carried out on the co-cured panel by applying quasi-static loading with displacement-controlled conditions, and 3-D FEA is carried out in Abaqus. Load-response plots have been obtained to validate the results at the global level, and they are in excellent agreement with experiments and 3-D FEA. Subsequently, out-of-plane displacement contour plots are obtained; the number of half waves and wave intensity in 3-D FEA and C-VAM are comparable, although there are minor differences compared to the experimental findings. The proposed framework is shown to be computationally efficient by over 55% as compared to 3-D FEA for performing non-linear buckling analysis on the stiffened composite structure considered in the current work.

This study aims at better understanding the damage tolerance of stiffened composite panels subjected to fatigue loads in the post-buckling regime. Ten single-stringer hat-stiffened specimens with an initial delamination between the skin and the stringer foot were manufactured, and then tested under quasi-static and fatigue loads in post-buckling conditions, with different load levels and load ratios. The tests were monitored with digital image correlation and an ultrasonic system, providing data on the displacements, strains, and extension of the delamination length. The quasi-static results showed that the delamination onset, when the initial delamination begins propagating, occurred at loads over twice the buckling load, while collapse occurred for values almost 20% higher than the delamination onset. During fatigue testing at load levels below the delamination onset, the specimens were able to sustain 150000 cycles and then, when tested statically after fatigue, the average load at collapse was reduced by less than 10% with respect to the quasi-static benchmark. When the maximum load during fatigue was increased to 5% over the delamination onset load, the specimens still withstood between 8000 and 16500 cycles before collapse, depending on the load ratio. It was also seen that for tests at the same load level, the specimens with high load ratio had a slower damage propagation.