Adhesive Joints in Wind Turbine Blades

The industrial goal of this PhD project is to enable manufacturing of larger wind turbine blades by improving the existing design methods for adhesive joints. This should improve the present joint design such that more efficient wind turbine blades can be produced. The main scientific goal of the project is to develop new- and to improve the existing design rules for adhesive joints in wind turbine blades. The first scientific studies of adhesive joints were based on stress analysis, which requires that the bond-line is free of defects, but this is rarely the case for a wind turbine blade. Instead linear-elastic fracture mechanics are used in this project since it is appropriate to assume that a crack can initiate and propagate from a pre-existing defect. The project was divided into three sub-projects. In the first sub-project, the effect of different parameters (e.g. laminate thickness, post curing and test temperatures) on the formation of transverse cracks in the adhesive were tested experimentally. It was assumed that the transverse cracks evolved due to a combination of mechanical- and residual stresses in the adhesive. A new approach was developed that allows the residual stress to be determined in several different ways. The accuracy of different ways of measuring residual stresses in the adhesive was tested by applying five different methods on a single sandwich test specimen (laminate/adhesive/laminate) that was instrumented with strain gauges and fiber Bragg gratings. Quasi-static tensile tests of sandwich specimens showed that higher post curing temperature and lower test temperature had a negative effect on the formation of transverse cracks in the adhesive i.e. transverse cracks initiated at lower applied mechanical loadings. The effect of increased laminate thickness was minimal under both static and cyclic loading. In the second sub-project, tunneling cracks in adhesive joints were analyzed numerically and experimentally. Simulations with a new tri-material finite element model showed that the energy release rate of the tunneling crack could be reduced by embedding a so-called buffer-layer with a well-chosen stiffness and -thickness. However, it was found for adhesive joints in wind turbine blades that the laminates were already sufficiently stiff. Thus, the effect of a stiffer buffer-layer was small in comparison with the effect of reducing the thickness of the adhesive layer. A new approach was in combination with a generic tunneling crack tool used to predict the cyclic crack growth rate for tunneling cracks in the adhesive joint of a full scale wind turbine blade. Model predictions were tested on a full scale wind turbine blade that was loaded excessively in an edgewise fatigue test in a laboratory. It was demonstrated that the model predictions were in agreement with measurements on the full scale test blade.
In the third sub-project crack deflection at interfaces in adhesive joints was investigated experimentally. Therefore, it was necessary to design a test specimen, where a crack could propagate stable and orthogonal towards a bi-material interface. A four-point single-edge-notch-beam (SENB) test specimen loaded in displacement control (fixed grip) was designed and manufactured for the purpose. In order to design the test specimen, new models were established to ensure stable crack growth and thus enable that crack deflection could be observed during loading (in-situ). A new analytical model of the four-point SENB specimen was derived, and together with numerical models it was found that the test specimen should be short and thick and the start-crack length relatively deep for the crack to propagate in a stable manner. Using the design from the developed models, crack deflection at interfaces for different material systems was tested successfully. For test specimens in selected test series it was observed that a new crack initiated at the interface before the main crack propagated and reached the interface. This cracking mechanism was used to develop a novel approach to determine the cohesive strength of the interface. The novel approach was applied to determine the cohesive strength of different material systems including an adhesive/laminate interface. It was found that the cohesive strength of the interfaces was small in comparison with the macroscopic strength of the adhesive.