Bulletin of the American Physical Society
71st Annual Meeting of the APS Division of Fluid Dynamics
Volume 63, Number 13
Sunday–Tuesday, November 18–20, 2018; Atlanta, Georgia
Session L14: Wind Energy: Experiments |
Hide Abstracts |
Chair: Sean Bailey, University of Kentucky Room: Georgia World Congress Center B301 |
Monday, November 19, 2018 4:05PM - 4:18PM |
L14.00001: LiDAR measurements and modeling of onshore wind farms on flat and complex terrains Stefano Letizia, Lu Zhan, Giacomo Valerio Iungo In order to understand and model interactions between the atmospheric boundary layer and multi-MW onshore wind farms, two experimental campaigns were carried out: the first one for a relatively flat terrain in Panhandle, TX, for which wind farm performance is mainly driven by the atmospheric stability, and a second site in Colorado characterized by a complex topography and low-level jets. Wind velocity measurements were performed through a scanning wind LiDAR, which are coupled with SCADA and meteorological data. LiDAR measurements of wind-turbine wakes are post-processed, clustered and ensemble-averaged based on atmospheric stability regime and operative conditions of the wind turbines. The resulting statistics of the wake velocity fields are used for tuning of a parabolic RANS model. Specifically, LiDAR data are leveraged for tuning the RANS turbulence closure for different atmospheric stability regimes and estimating thrust force over the blade span by coupling the LiDAR data and results of the respective RANS simulations. The RANS model is assessed against power production of individual turbines recorded through the SCADA showing a satisfactory agreement. |
Monday, November 19, 2018 4:18PM - 4:31PM |
L14.00002: Investigation of near wake modulation induced by utility-scale turbine operation Aliza Abraham, Jiarong Hong Super-large-scale flow visualization using natural snowfall has been implemented on a 2.5 MW horizontal-axis wind turbine (EOLOS turbine) to extract qualitative and quantitative information about the utility-scale turbine wake at unprecedented resolution, revealing the influence of incoming flow and turbine operation on near-wake behavior. However, past visualization studies were only conducted on streamwise-wallnormal planes, limiting analysis of spanwise wake behavior. In this study, we present a 3-hour video dataset that captures the full cross section of the EOLOS turbine wake at 0.2 rotor diameters downstream of the turbine, with the light sheet oriented parallel to the rotor plane. Based on Taylor’s frozen hypothesis, these plan-view images are stacked over time to visualize the temporal variation of coherent flow structures in the near wake. Specifically, the wake envelope marked by the positions of helical tip vortex structures allows us to quantify the deflection as well as the expansion/contraction of the wake. These wake behaviors are then correlated with turbine operational parameters such as yaw error, blade pitch, etc. to provide insight into the wake modulation associated with flow-turbine interaction. |
Monday, November 19, 2018 4:31PM - 4:44PM |
L14.00003: Wake meandering in a fully developed experimental scale wind turbine array John J Turner V, Martin Wosnik It is proposed that large scale atmospheric eddies push the wake of single turbines around. This atmospheric wake interaction, a dynamic shift in the wake over time, is a phenomenon known as wake meandering. Unlike a bluff body shedding frequency, the meandering phenomenon for a turbine is not characterized by a well-pronounced peak in the frequency domain, but rather by a bumb spread over a larger low-frequencies range. Single turbines and an array of 19 rows and 5 columns of turbines were positioned within a simulated atmospheric boundary layer (ABL) in a large boundary layer wind tunnel. High resolution velocity time series was obtained at high enough frequencies to resolve Strouhal number trends behind individual and coalescing wakes of turbines in ABL flow. It is proven that the downstream continuation of the meandering is provided via the intrinsic single wind-turbine ABL coupling behavior. It is proven that this meandering presents itself in large experimental arrays and that far downstream the peak meandering frequency is dominated by the turbine spacing. |
Monday, November 19, 2018 4:44PM - 4:57PM |
L14.00004: How High is High Enough? Horizontal Axis Wind Turbine Aerodynamics at Full-Scale Reynolds Numbers Using Laboratory Models Mark A. Miller, Janik Kiefer, Carsten Westergaard, Marcus Hultmark Achieving the high Reynolds numbers and matched tip speeds of full-scale horizontal axis wind turbines in a laboratory setting remains a major challenge for experimentalists. As a result, large-scale experiments and field tests are typically required to achieve dynamic similarity with the flow of interest. This forces most small-scale experiments on these types of systems to be performed at reduced Reynolds numbers. The underlying assumption being that lower, laboratory Reynolds numbers are still large enough to safely assume inertia effects dominate viscous ones. However, the experimental evidence for this is limited and has not been extensively confirmed. Current work at Princeton University utilizes a specialized, high-pressure wind tunnel facility to adjust the Reynolds number independent of the tip speed ratio. Results are available at Reynolds numbers based on free-stream conditions and diameter up to 14 million, a value typically encountered by field turbines of 25 meters in diameter operating at 8-10 m/s. Trends with Reynolds number are discussed and support presented for Reynolds number invariance occurring at values much larger than previously found. |
Monday, November 19, 2018 4:57PM - 5:10PM |
L14.00005: Abstract Withdrawn
|
Monday, November 19, 2018 5:10PM - 5:23PM |
L14.00006: Defining the Characteristic Velocity for the Reynolds Number in Turbine Experiments Hannah Ross, Brian Polagye The Reynolds number can significantly affect the performance of lift-based wind and water turbines, particularly at laboratory-scale. It is, therefore, important to use a consistent and physically-meaningful definition for the length and velocity scales. In these applications, the most meaningful length scale is the blade chord. For simplicity, common definitions of the Reynolds number take the characteristic velocity as either the tangential speed of the blade or the free-stream velocity. However, the flow over the blade can vary with rotational position and depends on the flow induction, free-stream velocity, and tangential blade speed. To explore the most meaningful choice of the characteristic velocity, this study tests a cross-flow current turbine over a range of Reynolds numbers. The Reynolds number is varied first by changing the temperature of the water, which alters the viscosity. The turbine is tested again at constant temperature and varying free-stream velocity. The implications of the choice of characteristic velocity scale are explored by determining the conditions under which performance is invariant between the two experiments. |
Monday, November 19, 2018 5:23PM - 5:36PM |
L14.00007: Turbulent wakes generated by 1 m diameter wind turbinemodels Martin Wosnik, Gregory G Taylor-Power The turbulent wakes generated by two 1 m diameter wake generator models were studied in the Flow Physics Facility (FPF) at the University of New Hampshire. The FPF has a cross section of 6.0m (W) x 2.7m (H), resulting in a blockage ratio of less than 5%. Horizontal and vertical velocity profiles were measured downstream of a 1 m porous disk up to 50 m downstream. The disk wake elongates in the vertical direction as it evolves far downstream, due to influence of wall boundary layers, and shifts downward due to the presence of the wind turbine model tower. Streamwise and azimuthal velocity profiles were measured in the wake of a 1 m diameter scale model wind turbine to 20 diameters downstream. The mean swirl appears to decouple from the mean velocity deficit as the wake evolves downstream; however, the mean azimuthal velocity W still exhibits a W∼x-1∼Uo3/2 scaling at intermediate downstream locations, where Uo is the wake centerline velocity deficit. Both disk and wind turbine wakes exhibit the classical high-Reynolds number scaling for axisymmetric turbulent wakes up to 20 diameters downstream. Axial and angular momentum integrals demonstrate that the mean pressure gradient is non-negligible in the near wake.
|
Monday, November 19, 2018 5:36PM - 5:49PM |
L14.00008: Several near-wake phenomena of a utility-scale wind turbine Teja Dasari, Aliza Abraham, Jiarong Hong Flow visualization and PIV via natural snowfall serves as an effective tool to probe the flow field around utility-scale wind turbines (Hong et al. Nature Comm. 2014). Using this tool, we have collected several datasets over 2015-2018 in the near-wake (<D) of the 2.5 MW turbine at the EOLOS station. These data provide 2D velocity field on the streamwise-wallnormal planes over the entire vertical span of the wake at multiple spanwise locations, as well as quantitative information of all the key coherent structures generated from the turbine, including blade tip/root vortices, trailing vortex sheets, hub vortices and turbine tower structures, etc. The follow-up analysis reveals several interesting near-wake phenomena, such as intermittent wake contraction, flow acceleration around the nacelle, and oscillation of nacelle wake, etc. Particularly, the movement of the nacelle-induced flow structures is shown to cause significant change of mean flow and turbulent kinetic energy profiles of the near wake. With conditional sampling, the interconnection among the near-wake velocity field, coherent structure behavior and inflow/turbine operational conditions is also investigated. |
Monday, November 19, 2018 5:49PM - 6:02PM |
L14.00009: Abstract Withdrawn
|
Monday, November 19, 2018 6:02PM - 6:15PM |
L14.00010: PIV of the aerodynamics of an airfoil subjected to icing Magnus Kyrkjebø Vinnes, Leon Li, R. Jason Hearst Wind turbines located in cold climates face several issues due to ice accretion. These include power losses, mechanical failure and changes in the aerodynamic behaviour. While the effect on symmetric airfoils and airfoils intended for aviation has been covered in the literature, airfoils designed for wind turbines have received limited attention. Therefore, particle image velocimetry experiments have been conducted to investigate the flow field over a NREL S826 airfoil, with different ice accretions at the leading edge. The three accretions represent different icing conditions, namely rime, glaze and mixed conditions. The mixed icing has a horn geometry, while the others have a streamlined geometry. Experiments were also conducted on the clean airfoil for comparison. Different angles of attack were used, ranging from -4deg to 16deg for all geometries. From the measurements, mean velocity, turbulent kinetic energy, mean vorticity and instantaneous swirl have been calculated. The results demonstrate how the separation bubble changes with angle of attack for the horn accretion. A separation bubble over the streamlined accretions at high angle of attack was also found. The shear layer between the recirculating region and the freestream are also described for all experimental cases. |
Monday, November 19, 2018 6:15PM - 6:28PM |
L14.00011: Reducing Dynamic Stall Effects and Load Fluctuations on Wind Turbine Blades using Trailing Edge Flap Farid Samara, David Johnson Dynamic stall on wind turbine blades often leads to severe fatigue and high load fluctuations that tend to decrease the lifespan of the blades. In this study the influence of trailing edge flap (TEF) on dynamic stall effects are investigated on an oscillating S833 airfoil with a chord length of 178 mm at a Reynolds number of 1.8×105 and a reduced frequency of k= 0.06 and 0.1. Surface pressure measurements along with strain at the blade support are collected for all cases inside a 0.61 m square wind tunnel . The static lift and moment coefficients will be presented for different pitch and flap angles. The lift and moments curves are then presented for the dynamic stall cases for a mean pitch angle of 0° and 10° to represent stall onset and deep stall cases. The flap is also oscillating at the same pitch frequency but with different phase lags to study the influence of flap motion. The curves clearly show the leading edge vortex formation and convection and how it is influenced by the TEF. The coefficient of pressure for different cases are also presented in a contour plot to illustrate how the pressure along the airfoil chord changes for the pitching cycle. Finally a conclusion is presented on how the TEF can control the load fluctuation experienced by the blades. |
Monday, November 19, 2018 6:28PM - 6:41PM |
L14.00012: Improving the efficiency and the wake recovery rate of vertical-axis turbines using detached end-plates Thierry Villeneuve, Matthieu Boudreau, Guy Dumas Aiming to improve the performance of vertical-axis turbines, this work is devoted to the development of detached end-plates, i.e., stationary end-plates that are not in contact with the turbine blades. The motivation is that the use of traditional attached end-plates has already been proven to be rather ineffective to increase the power coefficient of vertical-axis turbines because of the resistive torque (additional drag) generated by these wingtip devices moving with the blades (Gosselin et al., 2016). In this study, DDES simulations are performed using an unconfined H-Darrieus turbine with and without detached end-plates. It is found that the power coefficient is significantly increased when properly designed detached end-plates are used while the overall turbine drag coefficient remains almost unchanged. In addition to an improved power coefficient, we show that the use of detached end-plates can also lead to a faster recovery of the mean streamwise velocity in the turbine wake. An analysis of the streamwise momentum transport mechanisms allows us to conclude that the dominant contribution to the velocity recovery remains related to the transport term involving the mean spanwise velocity, as Boudreau & Dumas (2017) showed it for a vertical-axis turbine without end-plate. |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700