The current trend of wind turbine upscaling has led to the use of long and slender blades prone to large structural deformations. In the present study, the impact of the aeroelastic effects is assessed for the NREL-5MW wind turbine in a turbulent wind. To this end, an actuator curve method coupled to a one-dimensional finite-element structural solver is implemented in a fourth-order finite difference code that can perform large eddy simulation (LES) of realistic winds. The approach is computationally affordable compared to blade-resolved simulations and hence long time series can be computed. This, combined to the ability of the LES to capture the relevant scales of the unsteadiness wind, leads to a better estimation of the fluctuating loads and power of the turbine. The results are here shown for one and two wind turbines operating in a neutrally stable atmospheric boundary layer. It appears that the blade mostly deforms according to its first bending mode. The rotation of the blade in the sheared atmospheric flow is responsible for large amplitude deformations, but the turbulence also plays a role in causing deformations at higher frequencies with a smaller amplitude. The azimuthal distributions of loads and of power are significantly affected when the aeroelasticity is considered.

The present study aims at assessing the Actuator Disk (AD) method supplemented with an Individual Pitch Control (IPC) strategy, at a resolution appropriate for the Large Eddy Simulation of large wind farms. The IPC scheme is based on a state-of-the art individual pitch control, generalized to be applied to an AD approach. This procedure also requires an accurate recovery of the flapwise bending moment on each blade, which is not trivial for a disk-type model. In order to compute flapwise moments on each blade, blade trajectories are reproduced through the disk and the AD aerodynamic forces are interpolated onto these virtual blades at each time step. We verify the AD model with IPC in simulations of an isolated wind turbine, for different wind speeds and turbulence intensities, and in a configuration with two rotors. We compare the AD statistics with those obtained using an Actuator Line (AL) method. The comparison done in terms of equivalent moment shows that the AD and AL simulations produce very similar results.

The advance in wind turbine technology has led to a tendency to gather the turbines into wind farms to benefit from economies of scale. However, the farm configuration often leads to a drastic power loss due to the interactions between the turbines. To mitigate this issue, the turbines are typically misaligned with the dominant wind direction to deviate their wake. In this paper, a tilt-induced wake deviation strategy is studied in a farm configuration using large eddy simulation of an atmospheric boundary layer and an actuator line model for the blade. The results show an effective wake deviation and a significant increase of the total power production of the farm. The evolution of the time-averaged and unsteady loads is also assessed in the wind farm, and the resulting damage equivalent load is shown to be reasonably modified when the turbines are tilted. The tilted configuration is also compared with a wind farm with yawed turbines and is found to lead to a larger power output for a reduced damage equivalent load.

Wind energy is usually harnessed using wind turbines but it is not the only way of collecting this energy. In this paper, Airborne Wind Energy Systems (AWES) are considered. The understanding of the wake shed by such devices is of primary importance when considering farms, yet it has not been much studied so far, and even less so when considering kites with soft curved wing. In this work, LES is used to analyse the wake of AWES. Two types of wings are considered: a straight wing, modelling the reference rigid wing AWES and a curved wing accounting for leading edge inflatable kites. Both types of wings are modelled using actuator lines. The wake analysis reveals that the wake of curved wings tends to be stronger and to recover faster. For the case of the tilted trajectory, asymmetries in the wake are also noted. The difference is explained by the azimuthal variation of the circulation distribution on the wing. It also relates to some wakes characteristics for tilted wind turbines. The wake of the curved wing is also more sensitive to the variation of the loading when it flies tilted trajectory, due to the varying loading on its tips.

This work presents an investigation of the aeroelastic effects occurring on the IEA 15-MW reference wind turbine in a turbulent atmospheric boundary layer. Large Eddy simulation is carried out using an actuator line method coupled to the finite element beam solver BeamDyn. The results show that the blades experience significant displacements, leading to an important variation of the loads along the blade span. The turbulence also interacts with the blades natural frequencies, resulting in a variation of the spectra of the various loads. The impact of the flexibility is also evaluated on the wake by comparing the undeformed configuration with a statically and dynamically deformed rotor. It appears that the mean deformation plays an important role in the wake development due to changes in the load distribution, but that the dynamic effects have minimal impact on the wake behavior.