• Energy Research

Abstract The present review paper provides a comprehensive overview of the research activities on wind turbine aeroacoustics at DTU over the last 20 years, as well as it gives the state-of-the-art of noise prediction models for wind turbines under complex inflow conditions. Various noise generation models developed at DTU are described and analyzed, including models based on the acoustic analogy, flow-acoustics splitting techniques, Amiet's model, and various engineering models. Some of the models are coupled to existing aero-elastic software and computational fluid mechanics models developed at DTU, and implemented in the simulation platform WindSTAR (Wind turbine Simulation Tool for AeRodynamic noise). This simulation platform consists of WindSTAR-Gen, dealing with models for generation of noise and design of low-noise wind turbines, and WindSTAR-Pro, which is developed to handle the modeling of long distance acoustic propagation. As specific features of the WindSTAR-Pro package, the rotation of the noise sources is modeled, the propagation simulations combine the so-called Parabolic Equations (PE) propagation model with numerical flow simulations to take into account effects from wind turbine wakes, atmospheric turbulence and wind shear.

AbstractThe noise emission of wind turbines and farms can be an important and limiting factor for future cost reductions and growth of wind energy. Closing scientific and technological gaps on wind turbine noise is thus directly supporting the further development of renewable energy while reducing adverse reactions toward wind farms. The present article is providing guidance on the most relevant research directions from an engineering perspective, namely: simulation methods, wind tunnel testing, and wind turbine design. Each topic is addressed separately and specific scientific challenges are identified. Future research directions that may improve our physical understanding of wind turbine noise, as well as facilitate the deployment of wind energy, are outlined. It is concluded that future scientific research on the topic of wind turbine noise should be conducted in a multidisciplinary context to maximize its impact. The suggested topics shall be seen as a collection of what is seen as the most relevant topics across research and product development but shall not be seen as exclusive or interlinked with specific development plans.This article is categorized under: Sustainable Energy > Wind Energy Human and Social Dimensions > Social Acceptance

Abstract This work is concerned with the experimental study of airfoil stall and the modelling of stall noise. Using pressure taps and high-frequency surface pressure microphones flush-mounted on airfoils measured in wind tunnels and on an operating wind turbine blade, the characteristics of stall are analyzed. This study shows that the main quantities of interest, namely convection velocity, spatial correlation and surface pressure spectra, can be scaled highlighting the universal nature of stall independently of airfoil shapes and flow conditions, although within a certain range of experimental conditions. Two main regimes for the scaling of the correlation lengths and the surface pressure spectra, depending on the Reynolds number of the flow, can be distinguished. These results are used to develop a model for the surface pressure spectra within the detached flow region valid for Reynolds numbers ranging from 1 × 10 6 to 6 × 10 6 . Subsequently, this model is used to derive a model for stall noise. Modelled noise spectra are compared with experimental data measured in anechoic wind tunnels with reasonably satisfactory agreement.

AbstractThe aim of this work is to improve aeroelastic simulation codes by accounting for the unsteady aerodynamic forces that a blade experiences in static stall. A model based on a spectral representation of the aerodynamic lift force is defined. The drag and pitching moment are derived using a conditional simulation technique for stochastic processes. The input data for the model can be collected either from measurements or from numerical results from a Computational Fluid Dynamics code for airfoil sections at constant angles of attack. An analysis of such data is provided, which helps to determine the characteristics of stall. The model is applied to wind turbine rotor cases, including the stand still condition, and results are compared to experimental data. Copyright © 2009 John Wiley & Sons, Ltd.

Abstract The vortex-induced forces on an extruded cylinder with a span of two diameters representative of a finite segment of a non-tapered wind turbine tower at a very high Reynolds number (Re = 8.0×106) are numerically investigated using an incompressible Navier-Stokes flow solver with an Improved Delayed Detached Eddy Simulation (IDDES) turbulence model and correlation-based boundary layer transition modelling. The solution shows spanwise correlated structured vortex shedding with the Strouhal number St = 0.48. The boundary layer transition is found to occur at θ transition = 70 ◦ , and boundary layer separation is found to occur at θ separation = 120 ◦ . Results from the grid dependency study strongly imply that when using IDDES, the Strouhal number converges to higher values than previously reported by the literature as the grid is refined, with results ranging from St ∼ 0.44 using a grid with 4.2 × 106 cells, to St = 0.48 for the finest considered grid with 33 × 106 cells. This behaviour is not seen for URANS, where St = 0.33 for the finest grid.

The objective of this paper is an improved understanding of the physics of the aeroelastic motion of wind turbine blades in order to improve the numerical models used for their design. Two- and three-dimensional Navier–Stokes calculations of the flow around a wind turbine airfoil using the k−ω SST and Detached Eddy Simulation (DES) turbulence models, as well as an engineering semiempirical dynamic stall model, are conducted. The computational results are compared to the experimental results that are available for both the static airfoil and the pitching airfoil. It is shown that the Navier–Stokes simulations can reproduce the main characteristic features of the flow. The DES model seems to be able to reproduce most of the details of the unsteady aerodynamics. Aerodynamic work computations indicate that a plunging motion of the airfoil can become unstable.

ABSTRACTThis paper describes an extensive assessment and a step by step validation of different turbulent boundary‐layer trailing‐edge noise prediction schemes developed within the European Union funded wind energy project UpWind. To validate prediction models, measurements of turbulent boundary‐layer properties such as two‐point turbulent velocity correlations, the spectra of the associated wall pressure fluctuations and the emitted trailing‐edge far‐field noise were performed in the laminar wind tunnel of the Institute of Aerodynamics and Gas Dynamics, University of Stuttgart. The measurements were carried out for a NACA 643‐418 airfoil, at Re = 2.5 ×106, angle of attack of −6° to 6°. Numerical results of different prediction schemes are extensively validated and discussed elaborately. The investigations on the TNO‐Blake noise prediction model show that the numerical wall pressure fluctuation and far‐field radiated noise models capture well the measured peak amplitude level as well as the peak position if the turbulence noise source parameters are estimated properly including turbulence anisotropy effects. Large eddy simulation based computational aeroacoustic computations show good agreements with measurements in the frequency region higher than 1 kHz, whereas they over‐predict the sound pressure level in the low‐frequency region. Copyright © 2011 John Wiley & Sons, Ltd.