Design of an Experimental Test Set-up for Buckling of a Trailing Edge Composite Bonded Joint Specimen Based on Semi-analytical and Numerical Models

Abstract

The trailing edge of wind turbine blades are commonly manufactured as an adhesive joint of the pressure-side and suction-side composite panels of the blade. Under some conditions, a lead-to-trailing (LTT) edgewise bending moment can induce buckling at the trailing edge adhesive joint, which may lead to early failure of the blade due to delamination. As a structural instability, buckling in wind turbines has been the focus of much research especially in full-scale tests and more recently at the sub-component level. These higher-level tests, however, are done on pre-manufactured wind turbine blades and require extensive preparation in order to adapt the testing rig to each blade section, as well as incurring into elevated costs.
An additional test level has been suggested for elements and details of wind turbine blades. It has been suggested that this level can fulfill many purposes: New concepts, modifications, material combinations and orientations can be tested, partial safety factors of larger scale tests can be reduced or even certifying minor details of the blade can be done at the element and detail level. As such, the focus of this project is to develop a testing method for a simplified trailing edge bonded joint with a custom designed hinged clamping system upon which a compressive moment can be imposed to induce buckling.
The design of this test will initially be based on a semi-analytical buckling plate model, where in-plane and out-of-plane displacements are coupled through the Von Karman strain-displacement relations. This semi-analytical tool is employed to quickly estimate the buckling loads for plates of varying dimensions. Strain-free imperfections can be included in the model for twisted/pre-bent plates in order to estimate their effect on reducing the load-bearing capacity of the structure. The semi-analytical tool is complemented with FE models for all the design parameters.
The semi-analytical and numerical results are compared to demonstrate the agreement of both approaches aimed to provide a sturdy base for the research. Next, the experimental buckling loads and force-displacement curves are shown against the predictions from the previous approaches with good agreement. Nevertheless, the observable discrepancies between the experimental and numerical results showed that the desired joint-fixity at the boundaries was not fully realized, therefore leading to a slightly different post-buckling behavior. In the end, suggestions are given to improve on the experimental clamping system in order to improve and expand the scope of this research.