Direct numerical simulations of the Navier-Stokes equations are performed to achieve a better understanding of the behaviour of wakes generated by wind turbines. The simulations are performed by combining the in-house developed computer code EllipSys3D with the actuator-line methodology. In the actuator-line method, the blades are represented by lines along which body forces representing the loading are introduced. The body forces are determined by computing local angles of attack and using tabulated aerofoil coefficients. The advantage of using the actuator-line technique is that it is not needed to resolve blade boundary layers and instead the computational resources are devoted to simulating the dynamics of the flow structures. In the present study, approximately 5 million mesh points are used to resolve the wake structure in a 120-degree domain behind the turbine. The results from the computational fluid dynamics (CFD) simulations are evaluated and the downstream evolution of the velocity field is depicted. Special interest is given to the structure and position of the tip vortices. Further, the circulation from the wake flow field is computed and compared to the distribution of circulation on the blades.

The aim of the present project is to obtain a better understanding of thestability properties of wakes generated by wind turbine rotors. To accomplishthis a numerical study on the stability of the tip vortices of the Tjaereborgwind turbine has been carried out. The numerical model is based on large eddysimulations of the Navier-Stokes equations using the actuator line method togenerate the wake and the tip vortices. To determine critical frequencies theflow is disturbed by inserting harmonic perturbations.The results show that the instability is dispersive and that growth arisesonly for some specific frequencies and type of modes. The study also providesevidence of a relationship between the turbulence intensity and the length ofthe near wake. The relationship however needs to be calibrated against measurements.

Large eddy simulations of the Navier-Stokes equations are performed to achieve a better under-standing of the behaviour of wakes generated by wind turbines. The simulations are performed bycombining the in-house developed computer code EllipSys3D with the actuator-disc methodology.In the actuator-disc method the blades are represented by a disc at which body forces representingthe loading are introduced. The body forces are determined by computing local angles of attack andusing tabulated aerofoil coefficients. The advantage of using the actuator-disc technique is that itis not needed to resolve blade boundary layers and instead the computational resources are devotedto simulating the dynamics of the flow structures. In the present study approximately six millionsmesh points are used to resolve the wake structure in a park containing up to 9 turbines. Theresults from the CFD simulations are evaluated and the downstream evolution of the velocity fieldis depicted. Special interest is given to what extent the production is dependent of inflow angle.The result clearly indicates that turbulent inflow has a strong impact on the result and leads to amore complex flow structure.

Large eddy simulations of the Navier-Stokes equations are performed tosimulate the Horns Rev off shore wind farm 15 km outside the Danish westcoast. The aim is to achieve a better understanding of the wake interactioninside the farm. The simulations are performed by combining the in-housedeveloped computer code EllipSys3D with the actuator-disc methodology. Inthe actuator-disc method the blades are represented by a disc at which bodyforces representing the aerodynamic loading are introduced. The body forcesare determined by computing local angles of attack and tabulated aerofoil coefficients.The advantage of using the actuator-disc technique is that it is notnecessary to resolve blade boundary layers since the computational resourcesare devoted to simulating the dynamics of the flow structures.In the present study approximately 13.6 million mesh points are used toresolve the wake structure in the park. The park contains 80 wind turbinesdistributed over an area of about 20km2. Since it is not possible to simulate allturbines, the 2 central columns of turbines have been simulated with periodicboundary conditions. This corresponds to an infinitely wide farm with 10turbines in downstream direction. Simulations were performed within plusminus 15 degrees of the turbine alignment, making the wide farm approximationreasonable.The results from the CFD simulations are evaluated and the downstreamevolution of the velocity field is depicted. Special interest is given to whatextent the production is dependent on the inflow angle and turbulence level.The study shows that the applied method captures the main productionvariation within the wind farm. The result further demonstrates that levelsof production correlate well with measurements. However, in some cases thevariation of the measurement data is caused by variation of measurement conditionswith inflow angles.

Numerical simulations of the Navier-Stokes equations are performed toachieve a better understanding of the behaviour of wakes generated by windturbines. The simulations are performed by combining the in-house developedcomputer code EllipSys3D with the actuator line and actuator disc methodologies.In the actuator methods the blades are represented by lines or a discat which body forces representing the loading are introduced. The body forcesare determined by computing local angles of attack and using tabulated aerofoilcoefficients. The advantage of using the actuator techniques is that theyare not needed to resolve blade boundary layers. Instead the computationalresources are devoted to simulating the dynamics of the flow structures. Inthe present study we run simulations using both the actuator line and theactuator disc methodologies. The influence from changing a number of parametersis evaluated. The study serves as a validation and sensitivity study toboth the actuator line and the actuator disc methods. The result shows thatboth methods produce realistic results. There is however a sensitivity to someparameters.