Search

Dynamic inflow describes the effect of load overshooting due to a fast change in rotor loading. This effect results from the inertia of the global flow field which only adapts slowly to a fast variation in pitch angle or local wind velocity. This effect is still not fully understood and one of the main approaches to enhance the understanding are wind tunnel experiments with scaled wind turbines. The objective here is to present and discuss the recent experiments on dynamic inflow due to gusts in the turbulent wind tunnel in the WindLab of the University of Oldenburg.

To enable further increases in wind turbine dimensions, and thus to lower the levelized cost of energy, active and passive concepts for load reduction have to be further developed. One of such concepts is individual pitch control, which has been thoroughly analysed and validated in simulations and experiments (e.g., see [1] and references therein), and is currently used by some modern multi-megawatt wind turbines. On the other hand, although well known from aeronautics, concepts based on rotor blades with trailing edge flaps or leading edge slats have still not found any commercial application in wind energy. One obstacle is seen in the lack of a thorough and convincing validation of such concepts. In this work, we present our wind tunnel setup for experimental validation of advanced concepts for wind turbine load reduction, and analyse the interaction between rigid slats and the wind turbine control system in different inflow conditions. The wind tunnel at the University of Oldenburg has a nozzle of 3m by 3m and an active grid, which can impress a wide range of tailored turbulent flow conditions and coherent gusts on the wind flow [2]. The fully controllable and aerodynamically scaled version of a 5MW reference turbine MoWiTO (Model Wind Turbine Oldenburg) with a rotor diameter of 1.8m features the measurement of flapwise blade root and tower bending moments, rotor speed, and torque[3]. MoWiTO can be operated with two sets of blades – a set with rigid leading edge slats[4], and a reference set without slats. Additionally, the real-time hardware enables implementation of different control algorithms, but for the purpose of this work, two control algorithms will be used: a baseline controller for torque and collective pitch control, and an individual pitch controller designed to reduce once-per-revolution blade bending moments. Using the active grid, the capability of the rigid slats in combination with individual pitch control will be analysed in different inflow conditions, including turbulent and sheared flows, and wind gusts. [1] Bossanyi et al., IEEE Trans. on Control System Technology 21, 1067 (2013) [2] Kröger et al., Journal of Physics: Conference Series 1037, 052002 (2018) [3] Berger et al., Journal of Physics: Conference Series 1104, 012026 (2018) [4] Manso Jaume et al., Deutsche Windenergie Konferenz (2015)

To test concepts and controls of wind turbines under atmospheric-like inflow conditions at lab-scale, a method is developed in the turbulent wind tunnel of the University of Oldenburg. Commonly, wind turbine models are investigated in wind tunnels under stationary conditions with constant mean velocities. In the atmospheric boundary layer, the flow is unsteady and more complex situations occur. Normally, new concepts cannot be tested under such conditions in the wind tunnel. The herein presented method allows for generating atmospheric-like inflow situations in the wind tunnel and thereby capturing wind turbine power curves and dynamics at lab-scale. In this study, the 3 x 3m² active grid in the Oldenburg wind tunnel [1] and the model wind turbine with 1.8 m diameter (MoWiTO 1.8 [2]) are used. The wind velocity is changed by continuously varying the fan speed. To be close to the velocity distribution in the atmospheric boundary layer, the generated wind velocity by the fans is Weibull distributed. Additionally, the active grid generates a certain statistic of the flow (TI, intermittency). The turbulence intensity is calculated based on 10s sections and found to be constant for the most probable velocities. For higher velocities the turbulence intensity is slightly decreased. The generation and hence the velocity time series is completely reproduceable. This tailored continuous variation of the mean wind velocity in combination with a constant statistic (turbulence intensity) allows to generate atmospheric-like turbulence in the wind tunnel. A 30 minutes velocity time series (corresponding to 25h in free field) generated like this in the wind tunnel can be used to measure the full power curve of a model wind turbine. In addition to the common estimated power curves based on 10-minute averages, also dynamic effects, e.g. of turbine controls, can be studied. To do so, the Langevin power curve analysis is used [3]. The presented method therefore allows for an easy test procedure to investigate the general behavior of a wind turbine under realistic atmospheric-like conditions. In combination with the Langevin power curve analysis this is giving insight into dynamic effects. This allows for extensive investigations of different concepts of control and design, such as leading edges slats on the rotor blades (part of the SmartBlades2 project), in the wind tunnel under realistic reproduceable atmospheric-like conditions. [1] Kröger et al., J. Phys.: Conf. Ser. 1037, 052002, (2018). [2] Berger et al., J. Phys.: Conf. Ser. 1104, 012026, (2018). [3] Wächter et al., J. Wind Energy 14, (2011)., {"references":["Kröger et al., J. Phys.: Conf. Ser. 1037, 052002, (2018).","Berger et al., J. Phys.: Conf. Ser. 1104, 012026, (2018)","Wächter et al., J. Wind Energy 14, (2011)."]}