Bulletin of the American Physical Society
65th Annual Meeting of the APS Division of Fluid Dynamics
Volume 57, Number 17
Sunday–Tuesday, November 18–20, 2012; San Diego, California
Session G31: Wind Energy I |
Hide Abstracts |
Chair: Luciano Castillo, Texas Tech University Room: 33B |
Monday, November 19, 2012 8:00AM - 8:13AM |
G31.00001: Grid Sensitivity Analysis of Simulations of a Flow around a Single Rotating Wind Turbine Blade Bryan E. Kaiser, Michael A. Snider, Svetlana V. Poroseva, Rob O. Hovsapian Design of a wind farm layout with the purpose of optimizing the power outcome requires accurate and reliable simulations of a flow around and behind wind turbines. Such computations are expensive even for a single turbine. To find an optimal set of simulation parameters that satisfies both requirements in simulation accuracy and cost in an acceptable degree, a sensitivity study on how the parameters' variation influences results of simulations should be conducted at the early stage of computations. In the current study, the impact of a grid refinement, grid stretching, and cell shape on simulation results is analyzed in a flow around a single rotating blade utilized in a mid-sized Rim Driven Wind Turbine design (U.S. Patent \#7399162) developed by Keuka Energy LLC, and in its near wake. Simulation results obtained with structured and unstructured grids are compared. Industry relies on commercial software for conducting fluid flow simulations. Therefore, STAR-CCM+ software was used in our study. A choice of a turbulence model was made based on our previous sensitivity study of flow simulations over a rotating disk (see M. A. Snider, S. V. Poroseva, AIAA-2012-3146). [Preview Abstract] |
Monday, November 19, 2012 8:13AM - 8:26AM |
G31.00002: Coupling meso- and micro-scale fluid dynamics codes for wind-energy computing Ignas Satkauskas, Michael Sprague, Matt Churchfield Enabled by peta-scale supercomputing, the next generation of computer models for wind energy will simulate a vast range of scales and physics, spanning from wind-turbine structural dynamics and blade-scale turbulence to meso-scale atmospheric flow. This work focuses on new mathematical interface conditions and computational algorithms for coupling meso-scale numerical-weather-prediction codes with micro-scale turbine-vicinity fluid-dynamics codes. Here, an inherent challenge exists when the weather code is based on the compressible Euler equations while the turbine-vicinity flow is modeled by the incompressible Navier-Stokes equations. We propose several one- and two-way code-interaction approaches. These approaches are implemented in a two-dimensional testing platform composed of two in-house codes: (1) a finite-difference code that mimics the weather research and forecasting (WRF) solver and (2) an embedded-domain code based on a common finite-volume approach. [Preview Abstract] |
Monday, November 19, 2012 8:26AM - 8:39AM |
G31.00003: LES of turbulent flow past axial flow turbines and turbine arrays: Model development and validation Fotis Sotiropoulos, Seokkoo Kang, Xiaolei Yang, Leonardo Chamorro, Craig Hill We present recent progress towards the numerical simulation of turbulent flows past axial-flow wind and hydrokinetic turbines and farms. For simulating multi-turbine arrays, we combine turbine parameterization approaches (actuator disk and actuator line models) with our curvilinear-immersed boundary (CURVIB) LES model. Simulations are carried out both for aligned and staggered wind farms and the computed results are compared with wind tunnel experiments carried out at the St. Anthony Falls Laboratory (SAFL) atmospheric boundary layer wind tunnel. Turbine geometry resolving simulations also employ the CURVIB-LES solver with a wall model and very fine computational grids. Simulations are reported for a complete model marine turbine mounted at the bottom of a straight open channel and the computed results are compared with laboratory experiments obtained in the SAFL Main Channel. The simulated flowfields are analyzed to elucidate the structure of the turbine wake, identify large-scale instabilities, and quantify the mechanisms of turbulence production in the near and far wakes. [Preview Abstract] |
Monday, November 19, 2012 8:39AM - 8:52AM |
G31.00004: The Penn State ``Cyber Wind Facility'' James Brasseur, Ganesh Vijayakumar, Adam Lavely, Tarak Nandi, Balaji Jayaraman, Pankaj Jha, Alex Dunbar, Javier Motta-Mena, Sue Haupt, Brent Craven, Robert Campbell, Sven Schmitz, Eric Paterson We describe development and results from a first generation Penn State ``Cyber Wind Facility'' (CWF). The aim of the CWF program is to develop and validate a computational ``facility'' that, in the most powerful HPC environments, will be basis for the design and implementation of cyber ``experiments'' at a level of complexity, fidelity and resolution to be treated similarly to field experiments on wind turbines operating in true atmospheric environments. We see cyber experiments as complimentary to field experiments in the sense that, whereas field data can record over ranges of events not representable in the cyber environment, with sufficient resolution, numerical accuracy, and HPC power, it is theoretically possible to collect cyber data from more true, albeit canonical, atmospheric environments can produce data from extraordinary numbers of sensors impossible to obtain in the field. I will describe our first generation CWF, from which we have quantified and analyzed useful details of the interactions between atmospheric turbulence and wind turbine loadings for an infinitely stiff commercial-scale turbine rotor in a canonical convective daytime atmospheric boundary layer over horizontally homogeneous rough flat terrain. Supported by the DOE Offshore Initiative and the National Science Foundation. [Preview Abstract] |
Monday, November 19, 2012 8:52AM - 9:05AM |
G31.00005: Proper Orthogonal Decomposition analysis of kinetic energy entrainment in large wind farms Claire Verhulst, Charles Meneveau Vertical entrainment of kinetic energy is thought to play an important role in the dynamics of very large wind farms (Calaf et al., Phys Fluids 2010; and Cal et al. J. Ren. Sust. Energy 2010). To elucidate dominant mechanisms and flow physics of this vertical transfer of kinetic energy, we use Proper Orthogonal Decomposition (POD) to extract dominant flow structures from snapshots of velocity fields generated using Large Eddy Simulation of flow in an infinite turbine array in the atmospheric boundary layer. The POD analysis shows that the dominant modes are large streamwise counter-rotating vortices located above the turbines. The contribution of each POD mode to kinetic energy entrainment at the turbine level is then quantified and the modes are ordered by this contribution. Interestingly, the number of POD modes needed to represent dominant portions of the kinetic energy entrainment is less that the number needed to represent similar portions of the kinetic energy in the turbulent field. This suggests that understanding and controlling only a small number of flow structures may be relevant to the design of very large wind farms. In addition, to understand how the array layout affects the POD modes, several turbine orientations (aligned, staggered, etc) will be discussed. [Preview Abstract] |
Monday, November 19, 2012 9:05AM - 9:18AM |
G31.00006: An improved effective roughness height model for optimization of wind farm layout Xiaolei Yang, Fotis Sotiropoulos An improved effective roughness height model is developed to account for the different effects of streamwise turbine spacing and spanwise turbine spacing, which cannot be well captured by the classic model when the ratio of streamwise spacing to spanwise spacing is not equal to 1. The central idea of the present model is approximating the nominal incoming velocity by a kinematic model, which is the time- and horizontal-averaged velocity at hub height in the classic model. The prediction capability of the present model is validated by comparing with the results from large-eddy simulation of infinite large aligned wind farms. The present model is also tested for finite-size wind farms and staggered wind farms. [Preview Abstract] |
Monday, November 19, 2012 9:18AM - 9:31AM |
G31.00007: Wind Turbine Wakes with Actuator Line Aerodynamics Yulia Peet Actuator line aerodynamics (AL) model is becoming increasingly popular for characterization of the flow field and the turbulent wake created by the rotated turbines. AL model does not require boundary layer resolution and is thus significantly more efficient than the fully-resolved computations. Thus, simulation of multiple wind turbines and characterization of turbulent wakes in such multiple-turbine configurations is possible with the current model. In this talk, we investigate the properties of wind turbine wakes calculated by Large Eddy Simulations with the actuator line model, as a function of several parameters, including Reynolds number, tip speed ratio and distance between the turbines. Spectral element fluid dynamics code Nek5000 is used for the simulations. [Preview Abstract] |
Monday, November 19, 2012 9:31AM - 9:44AM |
G31.00008: Computational investigation of hydrokinetic turbine arrays in an open channel using an actuator disk-LES model Seokkoo Kang, Xiaolei Yang, Fotis Sotiropoulos While a considerable amount of work has focused on studying the effects and performance of wind farms, very little is known about the performance of hydrokinetic turbine arrays in open channels. Unlike large wind farms, where the vertical fluxes of momentum and energy from the atmospheric boundary layer comprise the main transport mechanisms, the presence of free surface in hydrokinetic turbine arrays inhibits vertical transport. To explore this fundamental difference between wind and hydrokinetic turbine arrays, we carry out LES with the actuator disk model to systematically investigate various layouts of hydrokinetic turbine arrays mounted on the bed of a straight open channel with fully-developed turbulent flow fed at the channel inlet. Mean flow quantities and turbulence statistics within and downstream of the arrays will be analyzed and the effect of the turbine arrays as means for increasing the effective roughness of the channel bed will be extensively discussed. [Preview Abstract] |
Monday, November 19, 2012 9:44AM - 9:57AM |
G31.00009: Vertical Mean Kinetic Energy Entrainment in a Scaled Wind Turbine Array Andrew Newman, Don Drew, Luciano Castillo A 2D model of the Mean Kinetic Energy (MKE) of a scaled wind turbine array was analyzed to understand how turbulent transport brings MKE into arrays from the Turbulent Boundary Layer above. This was done by applying a Proper Orthogonal Decomposition to particle image velocimetry data and constructing modal expansions for the Reynolds stress terms which appear in the transport equation along a horizontal surface above the array. These terms have been shown to be of the same order of magnitude as the power extracted from the turbines. It was also found that 75{\%} of the total Reynolds shear stress was carried by the first 13 modes. A strong correlation between a mode's Reynolds shear stress and its contribution to the MKE entrainment was demonstrated. Thus, a small number of modes are responsible for a large quantity of the MKE entrainment for the array. Modal streamwise length scales were determined; it was found that modal length decreases with increasing mode number. By considering sums of modes the largest scales observable in the experiment (13 rotor diameters) were shown to contribute over 50{\%} of the MKE entrainment. [Preview Abstract] |
Monday, November 19, 2012 9:57AM - 10:10AM |
G31.00010: Measurements in an axisymmetric turbulent wake with rotation downstream of a model wind turbine Nathaniel Dufresne, Martin Wosnik Energy production data from several of the existing offshore wind farms indicate that turbine arrays may enter a stall condition which can cause an overall energy production shortfall(which can exceed 10\%). This deep array stall is (presumably) due to the wakes generated by turbines upstream interacting with turbine rotors downstream. It is hypothesized that there is a critical array spacing at which this stall occurs, but that this spacing is dependent on rotor thrust $c_T$ (which is determined by tip-speed ratio $\lambda$ and power coefficient $c_P$ of the rotor), Reynolds number, upstream conditions, and possibly wall roughness. An experimental investigation of the axial and azimuthal velocity field measurements in the wake of a single 3-bladed wind turbine with rotor diameter of 0.91m was conducted. The turbine was positioned in the free stream, near the entrance of the 6m x 2.5m test section of the UNH FPF, which can achieve test section velocities of up to 15 m/s and Reynolds numbers $\delta^+ = \delta u_\tau/\nu \approx 30,000$. Hot-wire anemometry was used to obtain velocity field measurements. The data obtained will be used to examine similarity scaling functions for velocity, wake growth, and turbulence derived from an equilibrium similarity analysis of the far wake. [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. |
© 2019 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
1 Research Road, Ridge, NY 11961-2701
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700