PSI - Issue 57

Philipp Ulrich Haselbach et al. / Procedia Structural Integrity 57 (2024) 169–178 P. U. Haselbach and P. Berring / Structural Integrity Procedia 00 (2023) 000–000

171

3

Wind and Energy Systems, as part of the ReliaBlade (2023) project granted by the Danish Energy Agency through the EUDP grant 640180068, a blade full-scale wind turbine blade embedded with artificial damages was designed and manufactured. It will be tested under static and fatigue load scenarios according to design loads from aeroelastic simulations.

2. Method

For this purpose, the baseline of the DTU 12.6m reference wind turbine blade (Haselbach, P. U. et al. (2020), Haselbach, P. U. et al. (2020)) is modified and artificial defects embedded, which are supposed to grow under cycling loading. One of the implemented failure types is an area, where debonding should take place initiated / triggered by an unbonded area and is supposed to grow steadily into regions with only partly bonds of around 30 % and 50 % of the contact surfaces area, respectively. It is aimed on to investigate whether existing methods can monitor the damage growth rates and sizes and also allow to predict numerically the expected growth rates and sizes as for di ff erent load scenarios in order to establish a reliable PHM systems. 2.1. DTU 12.6m wind turbine rotor blade The DTU 12.6m wind turbine rotor blade is designed as a demonstrator and reference platform. Nevertheless, the blade design refers to a real blade that has been optimized for maximizing annual energy production subject to loads, deflection and mass constraints as replacement alternative for older 150kW turbines and can be operated as pitch controlled or as stall-regulated blade. In an iterative design loop between aerodynamic design, aeroelastic design and structural design a balanced design was reached based on fatigue and ultimate strength design load cases. The detailed design process is described in Haselbach, P. U. et al. (2020). 2.2. Finite element model of the DTU 12.6m blade The original structural blade model was generated using DTU’s in-house software Blade Modeling Tool (BMT), generating a 3D volumetric model for the commercial finite element pre / post-processing software Abaqus CAE 2022 (Abaqus (2022)). The blade geometry is discretized with 208,744 eight-node general purpose linear brick elements with reduced integration (Abaqus type C3D8I) with a characteristic element length between 0.010 m and 0.015 m. The volumetric geometry and its composite layup are specified with layered-elements, where only one element through the thickness is used to discretize the airfoil. The entire blade geometry is modelled based on input data of 99 blade cross sections generated by DTU Wind HawtOpt2 optimizing workflow. These cross sections describe the outer geometry (airfoil) of the blade, where 59 di ff erent regions per cross section, as e.g. radial position R = 0.00 m – 0.120 m, are used to assign the layup in the specific regions. The thickness is specified via the volumetric height of the element (element thickness in 3-direction) and the layup / ply thickness is relative to the entire stack of plies. Finally, the individual cross sections are connected by spline curves and interpolation surfaces to obtain a volume representation of the blade. The plybook of the composite layup varies depending on the structural requirements and consists of 6 to 54 plies though the thickness. Only the adhesive bonds in the trailing edge region as well as between the shear webs and caps are described individually. The composite properties, with its material, layup and ply orientations are assigned to the layered continuum solid elements. The layup of the DTU 12.6 m wind turbine blade was designed to experience a maximum of around 4000 µϵ in ϵ 11 along the blade span. For the wind turbine blade exposed to flapwise suction to pressure side static loading including PSF (1.35 PSF for the aeroelastic load calculations and 1.1 PSF for experimental loading), the highest strains in the spar cap region occur between 3.8 m and 9 m radial position along the wind turbine blade span as shown in Figure 1. Initiation shear web and cap debonding in this region could be attractive due to the high strain levels. However, beside shear web and cap debonding, also other failure types, like wrinkles and trailing edge debonding, are placed in the same blade. Thus, each failure type got predefined regions assigned in order to avoid failure accumulation in the region of highest stresses and strains and also to allow the placing of load saddles on the blade for the experiments. Hence, it was decided to place the shear web and cap debonding on the upwind side of the blade between the blade root (r = 0 m) and a radial position along the blade span of r = 4m.