An extensive dataset with mixed-mode fracture resistance experimental data (R-curves) and fracture process parameters is presented. The fracture resistance values are extracted from double cantilever beam specimens subject to un-even bending moments. The tested specimens are made of a unidirectional composite which undergoes large-scale fibre bridging during fracture. The data set contains both raw data (forces from two load-cells, time, acoustic emission signals, and opening displacement measurements) and processed data (J-integral, end-opening displacements, fracture process parameters) of each test. In the repository, MATLAB scripts are also provided to facilitate recreating the processed data from the raw data.

A multiscale computational framework is presented to assess the effect of voids that may arise from a scarf repair and its influence on the post-repair lifetime of wind turbine blades. The approach links a full scale blade model with a detailed repair region model and a microscale representation of polymer adhesives containing voids. Boundary conditions from the global blade model are transferred to the scarf repair model, which subsequently provides input to a microscale representative volume element (RVE) of the adhesive containing voids using the submodelling technique. This RVE is combined with a continuum damage mechanics formulation to simulate high-cycle fatigue and estimate the lifetime for different void contents. The effect of void content resulting from scarf repair is evaluated under both quasi-static and high-cycle fatigue loading, enabling lifetime predictions. In the simulations, it was demonstrated that the lifetime of repaired blade is 4 times lower for the repair with 4% void content as compared with the repair with 1% void content.