Fluid Structure Interactions for Wind Turbines

FSI-WT project built upon the knowledge and softwares libraries developed in the research funded project ICADA and NOWITECH. An open source code named IFEM was further developed in the FSI-WT project. Since, IFEM is based on an emerging technology and was at a low TRL, in parallel we also extended / optimized / validated the open source code OpenFoam to equip the industry partners with a toolkit having high TRL. These tools were validated using the data generated by the participating partners both within the projects and also in other parallel running projects. 1) IFEM - An object oriented Isogeometric finite element library We have in the project enhanced the open source object oriented library IFEM to enable aerodynamic simulation of flow around wind turbine blades. We have achieved results for lift and drag that compares well with experiments for 2D and 3D section analysis with realistic high Reynolds flow. Furthermore, we have demonstrated proof of concept for fully coupled fluid-structure interaction simulation of (very) flexible structures in low Reynolds flow. In addition we have developed divergence conforming CFD-solvers (for pointwise fulfilment of the incompressibility constraints) that shows promising results related to reduced order modelling (ROM). 2) WTFoam validated towards wind tunnel tests Based on the open source library OpenFoam, we have developed WTFoam that may be utilised for 2D as well as 3D forced fluid-structure interaction simulations of flow around wind turbines at realistic wind velocities. WTFoam has been validated towards wind tunnel tests performed at NTNU, and we obtained equal or better accuracy compared to all other groups that have addressed the same tests. This study demonstrated that WTFoam using three dimensional forced fluid-structure interaction simulation for handling the rotation of the turbine, are capable of accurate prediction (at least at laboratory scale) of the conversion of kinetic energy from the wind into mechanical energy for production of electricity. Furthermore, the tests show that WTFOam is able to predict reasonable accurate wake vortecities. 3) WTFoam used to investigate simulation of full scale wind turbines (NREL 5MW) A complete aerodynamics analysis of NREL 5MW wind turbine at full scale was conducted with an incremental level of geometric approximations (from simple 2D to complete 3D) using WTFoam. This investigation revealed that if computational resources are limited, and the full-scale 3D solution cannot be adopted, a 2.5D spatial dimension (i.e. 3D slices around represntative cross sections) should be preferred rather then performing a simplistic 2D analysis. Cyclic periodicity (120°) can become a legitimate alternative if turbine?s overall torque calculations are under consideration. Whereas, if the purpose is to evaluate the horizontal wake deficit profiles in the downstream direction, (360°) rotor sector can produce plausible results. However, for vertical profiles, nacelle and tower should be employed to get a realistic picture of the wake in the downstream direction. 4) Multiscale modelling of wind turbine in a turbulent atmospheric boundary layer SINTEF brought together with Meteorological Institute a multiscale simulation tool (HARMONIE-SIMRA) for micro-scale wind simulation into the project. The coupled model have been tested by doing real time forecast of the wind and turbulence intensity in the vicinity of the onshore Bessaker wind farm. The tool has been extended with Kalman filters to correct the numerical prediction using the observation data from the wind turbines (production data have been transformed to wind velocities). Furthermore, to enable accurate simulation of offshore wind turbines the atmospheric code HARMONIE and the wave modelling code WAM were coupled uni- and bi-directionally with each other. Our studies show that bi-directional coupling is more accurate for high significant waves. We have also performed high-fidelity simulation of turbulent air flow over propagating water waves (air-sea interaction) by means Large Eddy Simulation (LES) techniques. We have produced a data set of the wave growth process, where a flat water surface is forced by the turbulent flow to produce fully developed nonlinearly interaction water waves . This dataset may be of use in further development of wind forcing modules of operational wave forecasting models such as WAM. The multiscale system (HARMONIE-SIMRA-WAM) is capable of handling atmospheric stratification and to provide turbulent atmospheric inflow conditions for aerodynamic flow simulation around onshore and offshore wind turbines. To enable the needs of the industry to have tools that are light, easy to use and computationally efficient so that they could be run in real time on a desktop pc or laptop. To this end we started working on a Reduced Order Modeling tool-kit which will use the high fidelity simulation results generated in the FSI-WT project.

The deep sea offshore wind potential is huge, but will only be realised provided that costs are reduced to a competitive level. This project addresses exactly this, presenting results on technical issues related to the core of wind energy i.e. the convers ion of kinetic energy from the wind into mechanical energy that is used to produce electricity. By enabling more understanding of the physics behind the wind turbine interaction through detailed fluid-strucutre interction simulation we may increase the co mpetence in developing more optimized wind turbines as well as more yielding wind farm layouts. Numerical simulation tools are an invaluable way of gaining new insights into these issues with the possibility of integrating many physical models influencin g the performance of the wind turbine. Up to now the integrated numerical simulation tools used by the wind energy industry apply simplified empirical or parametrized models to compute the aerodynamical forces on the turbine blades. Simplified numerical m odels are computationally efficient, but important details of the flow field and nonlinearities in the interaction of the air flow and rotor blades are not resolved. Coupled fluid-structure interaction (FSI) simulations are needed for accurate modelling of wind turbines, and also to provide input parameters, to verify and to improve parametrized models. Detailed FSI analysis will be even more important in the design of offshore wind turbines because of the extreme wind conditions at sea. New methods fo r accurate, robust and efficient simulation of FSI for wind turbines will be developed in the project. The methods will be implemented to form a state-of-the-art numerical simulation tool for coupled FSI simulations of a full scale rotor utilizing the com putational power of modern parallel hardware architectures. Furthermore, the atmospheric interaction is important and will be taken into consideration by establishing representative input profiles.