Bulletin of the American Physical Society
69th Annual Meeting of the APS Division of Fluid Dynamics
Volume 61, Number 20
Sunday–Tuesday, November 20–22, 2016; Portland, Oregon
Session D2: Wind Turbines: Simulations |
Hide Abstracts |
Chair: Dennice Gayme, Johns Hopkins University Room: A106 |
Sunday, November 20, 2016 2:57PM - 3:10PM |
D2.00001: A simple dynamic wake model for time dependent wind turbine yaw Carl Shapiro, Charles Meneveau, Dennice Gayme This work develops a time dependent wake model for wind farms that better captures the spanwise and streamwise propagation of fluctuations generated by changes in turbine thrust and yaw angle. The model builds on classic wake models by incorporating time dependence and turbine yawing. These extensions enable us to capture the spanwise skewness in the yawed turbine wake as well as the dynamic advection of the wake downstream. This model is then compared to large eddy simulations of a wind farm with upstream rows of wind turbines dynamically yawing their rotors. An important advantage of the model is it allows us to take advantage of predictions of dynamic flow phenomena to coordinate the action of individual wind turbines for farm level control. We use the model to further explore the potential of wind farms to use wind turbine yaw to provide important services to the power grid through power tracking. [Preview Abstract] |
Sunday, November 20, 2016 3:10PM - 3:23PM |
D2.00002: An LES study of vertical-axis wind turbine wakes aerodynamics Mahdi Abkar, John O. Dabiri In this study, large-eddy simulation (LES) combined with a turbine model is used to investigate the structure of the wake behind a vertical-axis wind turbine (VAWT). In the simulations, a recently developed minimum dissipation model is used to parameterize the subgrid-scale stress tensor, while the turbine-induced forces are modeled with an actuator-line technique. The LES framework is first tested in the simulation of the wake behind a model straight-bladed VAWT placed in the water channel, and then used to study the wake structure downwind of a full-scale VAWT sited in the atmospheric boundary layer. In particular, the self-similarity of the wake is examined, and it is found that the wake velocity deficit is well characterized by a two-dimensional elliptical Gaussian distribution. By assuming a self-similar Gaussian distribution of the velocity deficit, and applying mass and momentum conservation, an analytical model is developed and tested to predict the maximum velocity deficit downwind of the turbine. [Preview Abstract] |
Sunday, November 20, 2016 3:23PM - 3:36PM |
D2.00003: An actuator line model simulation with optimal body force projection length scales Luis Martinez-Tossas, Matthew J. Churchfield, Charles Meneveau In recent work (Mart\'inez-Tossas et al.~``Optimal smoothing length scale for actuator line models of wind turbine blades'', preprint), an optimal body force projection length-scale for an actuator line model has been obtained. This optimization is based on 2-D aerodynamics and is done by comparing an analytical solution of inviscid linearized flow over a Gaussian body force to the potential flow solution of flow over a Joukowski airfoil. The optimization provides a non-dimensional optimal scale $\epsilon/c$ for different Joukowski airfoils, where $\epsilon$ is the width of the Gaussian kernel and $c$ is the chord. A Gaussian kernel with different widths in the chord and thickness directions can further reduce the error. The 2-D theory developed is extended by simulating a full scale rotor using the optimal body force projection length scales. Using these values, the tip losses are captured by the LES and thus, no additional explicit tip-loss correction is needed for the actuator line model. The simulation with the optimal values provides excellent agreement with Blade Element Momentum Theory. [Preview Abstract] |
Sunday, November 20, 2016 3:36PM - 3:49PM |
D2.00004: Assessing the Impacts of Low Level Jets' Negative Wind Shear over Wind Turbines. Walter Gutierrez, Arquimedes Ruiz-Columbie, Murat Tutkun, Luciano Castillo Nocturnal Low Level Jets (LLJs) are defined as relative maxima in the vertical profile of the horizontal wind speed at the top of the stable boundary layer. Such peaks constitute major power resources, since they are observed at altitudes within the heights of commercial-size wind turbines. However, a wind speed maximum implies a transition from a positive wind shear below the maximum height to a negative one above. The effect that such transition inflicts on wind turbines has not been thoroughly studied. Here we focused on the impacts that the LLJ negative wind shears have over commercial size wind turbines. Using actual atmospheric LLJ data of high frequency as input for the NREL aeroelastic simulator FAST, different scenarios were created varying the LLJ maximum height with respect to the wind turbine hub height. We found only slight changes in the deflection and load averages for those scenarios, whereas the corresponding variances appear to decrease when a larger portion of the wind turbine sweeping area is affected by the negative shear. The exception was observed in the junction between the tower top and the nacelle, where a deflection maximum was detected that might reveal a critical structural point. [Preview Abstract] |
Sunday, November 20, 2016 3:49PM - 4:02PM |
D2.00005: Modeling velocity space-time correlations in wind farms Laura J. Lukassen, Richard J.A.M. Stevens, Charles Meneveau, Michael Wilczek Turbulent fluctuations of wind velocities cause power-output fluctuations in wind farms. The statistics of velocity fluctuations can be described by velocity space-time correlations in the atmospheric boundary layer. In this context, it is important to derive simple physics-based models. The so-called Tennekes-Kraichnan random sweeping hypothesis states that small-scale velocity fluctuations are passively advected by large-scale velocity perturbations in a random fashion. In the present work, this hypothesis is used with an additional mean wind velocity to derive a model for the spatial and temporal decorrelation of velocities in wind farms. It turns out that in the framework of this model, space-time correlations are a convolution of the spatial correlation function with a temporal decorrelation kernel. In this presentation, first results on the comparison to large eddy simulations will be presented and the potential of the approach to characterize power output fluctuations of wind farms will be discussed. [Preview Abstract] |
Sunday, November 20, 2016 4:02PM - 4:15PM |
D2.00006: Using Reconstructed POD Modes as Turbulent Inflow for LES Wind Turbine Simulations Jordan Nielson, Kiran Bhaganagar, Vejapong Juttijudata, Sirod Sirisup Currently, in order to get realistic atmospheric effects of turbulence, wind turbine LES simulations require computationally expensive precursor simulations. At times, the precursor simulation is more computationally expensive than the wind turbine simulation. The precursor simulations are important because they capture turbulence in the atmosphere and as stated above, turbulence impacts the power production estimation. On the other hand, POD analysis has been shown to be capable of capturing turbulent structures. The current study was performed to determine the plausibility of using lower dimension models from POD analysis of LES simulations as turbulent inflow to wind turbine LES simulations. The study will aid the wind energy community by lowering the computational cost of full scale wind turbine LES simulations, while maintaining a high level of turbulent information and being able to quickly apply the turbulent inflow to multi turbine wind farms. This will be done by comparing a pure LES precursor wind turbine simulation with simulations that use reduced POD mod inflow conditions. The study shows the feasibility of using lower dimension models as turbulent inflow of LES wind turbine simulations. Overall the power production estimation and velocity field of the wind turbine wake are well captured with small errors. [Preview Abstract] |
Sunday, November 20, 2016 4:15PM - 4:28PM |
D2.00007: Proper orthogonal decomposition of a large eddy simulation during a diurnal cycle for very large wind farms Naseem Ali, Gerard Cortina, Nicholas Hamilton, Marc Calaf, Raúl Cal The structure of the turbulent flow within large wind farms under different atmospheric flow stratification (stable, unstable and neutral) and compares it to the case when there are no turbines present. Spectral analysis is further applied to the corresponding proper orthogonal modes to identify the characteristic wavenumber and be able to relate it to the actual wind farm structure and wake-to-wake interactions. The variation in the number of needed modes between the different cases decreases with increasing value of cumulative energy, which confirms that the major difference between the different study cases resides at the largest turbulent kinetic energy containing scales of each case. The POD modes show the stratification impact of the flow structure and distinguish the flow layers. The spectral analysis displays the domain size and the distance between the rotors as distinctive scales within the wind farm. [Preview Abstract] |
Sunday, November 20, 2016 4:28PM - 4:41PM |
D2.00008: Anisotropy stress invariants of a large eddy simulation during a diurnal cycle for very large wind farms Raúl Cal, Naseem Ali, Nicholas Hamilton, Gerard Cortina, Marc Calaf Reynolds stress invariants of the turbulent flow within large wind farms under different atmospheric flow stratification (stable, unstable and neutral); cases without turbines is also evaluated. Lumley triangle and barycentric map are used to quantify the anisotropic stress tensor within the wind farm and atmospheric boundary layer. Dependent on the thermal stratification, the unstable and neutral cases of the wind farm and no wind farm display the minimum second invariant in contrast to the stable case that shows the maximum invariants. Scaled color is used to present the invariant as a function of domain height. The unstable stratification approaches the isotropy limit at high layers of the domain and the stable stratification leads the turbulence flow to be one component flow. The principle eigenvalues are also shown, where they show its effects on the vicinity of the swept area of the rotor. Finally, spheroid visualization is pursued to understand and interpret the realizable turbulence flow within the wind farm and atmospheric boundary layer. [Preview Abstract] |
Sunday, November 20, 2016 4:41PM - 4:54PM |
D2.00009: Focused-based multifractal analysis of the wake in a wind turbine array utilizing proper orthogonal decomposition Hawwa Kadum, Naseem Ali, Raúl Cal Hot-wire anemometry measurements have been performed on a 3 x 3 wind turbine array to study the multifractality of the turbulent kinetic energy dissipations. A multifractal spectrum and Hurst exponents are determined at nine locations downstream of the hub height, and bottom and top tips. Higher multifractality is found at 0.5D and 1D downstream of the bottom tip and hub height. The second order of the Hurst exponent and combination factor show an ability to predict the flow state in terms of its development. Snapshot proper orthogonal decomposition is used to identify the coherent and incoherent structures and to reconstruct the stochastic velocity using a specific number of the POD eigenfunctions. The accumulation of the turbulent kinetic energy in top tip location exhibits fast convergence compared to the bottom tip and hub height locations. The dissipation of the large and small scales are determined using the reconstructed stochastic velocities. The higher multifractality is shown in the dissipation of the large scale compared to small-scale dissipation showing consistency with the behavior of the original signals. [Preview Abstract] |
Sunday, November 20, 2016 4:54PM - 5:07PM |
D2.00010: Investigation of wake characteristics in wind farm varying turbulent inflow condition Jisung Na, Eunmo Koo, Munoz-Esparza Domingo, Emilia Kyung Jin, Rodman Linn, Joon Sang Lee In this study, we investigate the wake characteristics in wind farm varying turbulent property at inlet condition. To solve the flow with wind turbines and its wake, we use large eddy simulation (LES) technique with actuator line method (ALM). The wake characteristics in wind farm is important mainly in performance of wind farm because non-fully recovered wake induced by upstream wind turbines interferes power generation at downstream wind turbines. Turbulent inflow which contains the information of turbulence in atmospheric boundary layer is one of the key factors for describing the wake in wind farm accurately. We perform the quantitative analysis of velocity deficit and turbulent intensity in whole cases. In the comparison between cases with and without turbulent inflow, we observe that wake in case with turbulent inflow is more diffused to span-wise direction. And we analyze the coherent structures behind wind turbines at each row. Through above-analysis, we reveal how the wake is interacted with performance of wind farm. [Preview Abstract] |
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