mets2016_poster_SaleAliseda
1 like49 views
Experiments and numerical simulations were conducted to study the flow field of an array of axial-flow marine hydrokinetic turbines operating in a flume. Particle image velocimetry and rotor torque measurements were collected from the experiments and compared to computational fluid dynamics simulations using Reynolds-averaged Navier-Stokes and large eddy simulation models. The simulations captured some wake characteristics accurately but differed from particle image velocimetry in their representation of the near wake and nacelle effects. Adding turbulent inflow to the large eddy simulation improved the realism of the wake behavior observed in experiments and other studies.
![Fig. 1 (top) solid model of the re-designed ¡°Reference Model 1¡± turbine as built, (bottom)
configuration of the turbines in the Bamfield Marine Science Centre flume, and measured
efficiency of the turbine array comparing to a variety of RANS and LES models
Experiments and numerical simulations are conducted for an array of axial-flow Marine
Hydro-Kinetic (MHK) turbines operating in a flume. This study aims to understand the
influence of coherent structures in high Reynolds number wakes on energy extraction and
dynamical rotor control processes. In the flume experiments, rotor torque and rotational
position measurements are collected, and the flow field is characterized by Acoustic Doppler
Velocimetry (ADV) and simultaneous imaging with Particle Image Velocimetry (PIV). The
performance of 3 turbines are characterized under varying stream-wise spacing, cross-
stream offsets, and varying yaw angles to the inflow. To study the effect of unsteady
hydrodynamics, the turbines are outfitted with either open-loop or closed-loop feedback
rotor-speed controllers, designed to maintain target tip-speed-ratios. The experimental
results were compared against Computational Fluid Dynamics (CFD) simulations using
different models to represent the effect of the rotor and for different strategies to simulate the
turbulent flow field: both Large Eddy Simulation (LES) and Reynolds-Averaged Navier-Stokes
(RANS). In the LES and RANS simulations, turbine models within the class of "actuator
methods" are included which impose body forces upon the flow field in form of disks or lines.
The laboratory scale MHK turbines (designed, manufactured and flume tested) demonstrated
high efficiency and wake characteristics similar to the full scale reference model turbine. A
marked improvement in efficiency was seen, although contributed by several factors such as:
rotor-speed controller, Reynolds scaling, and blockage effects. Differences between the PIV,
RANS, and LES confirm that the turbine models have certain strengths and shortcomings.
Namely, the overset mesh method performs well to capture the wake of nacelle, while the LES
does not include the nacelle effect and therefore predicts the wake inaccurately. Adding a
turbulence inflow to the LES greatly improves the realism of the model, showing near wake
breakdown and wake meandering, which was observed here and has been observed in other
MHK and wind turbine studies.
THE FLOW FIELD OF A TWO-BLADED HORIZONTAL AXIS TURBINE VIA
COMPARISON OF LES AND RANS SIMULATIONS AGAINST
EXPERIMENTAL PIV FLUME MEASUREMENTS
Danny Sale and Alberto Aliseda (dsale@uw.edu, aaliseda@uw.edu)
University of Washington, Mechanical Engineering, Seattle, WA, U.S.A.
Proceedings of the 4th Marine Energy Technology Symposium, Washington, D.C., April 25-27
ACKNOWLEDGEMENTS
Thanks to Ben Strom, Rob Cavagnaro, Dom Forbush, John Bates, Michelle Hickner and Brian
Polagye for their help setting up and conducting the experiments. This work was financially
supported by the National Science Foundation under the Sustainable Tidal Energy project, and the
Naval Facilities Engineering Command through the Northwest National Marine Renewable Energy
Center. Help from the Bamfield Marine Sciences Centre and specially Eric Clelland is greatly
appreciated. This work was facilitated through the use of infrastructure provided by the Hyak
supercomputer system at the University of Washington.
[1] V. Neary, C. Hill, L. Chamorro, B. Gunawan, F. Sotiropoulos, ¡°Experimental Test Plan ¨C DOE Tidal and
River Reference Turbines¡± Technical Report, Oak Ridge National Laboratory, 2012.
[2] M.J. Churchfield, Y. Li, P.J. Moriarty. ¡°A large-eddy simulation study of wake propagation and power
production in an array of tidal-current turbines¡±. Phil. Trans. R. Soc. A 28 February 2013 vol. 371 no. 1985
[3] Kornev, N. & Hassel, E. (2007). ¡°Method of random spots for generation of synthetic inhomogeneous
turbulent fields with prescribed autocorrelation functions¡±. Communications in Numerical Methods
Engineering, Vol. 23, Issue 1, pp. 35-43.
[4] T Javaherchi, S Antheaume, A Aliseda, ¡°Hierarchical methodology for the numerical simulation of the flow
field around and in the wake of horizontal axis wind turbines: Rotating reference frame, blade element
method and actuator disk model¡±, Wind Engineering, 2014
[5] T. Javaherchi, J. Seydel, N. Stelzenmuller, and A. Aliseda ¡°Experimental and Numerical Analysis of a
Scale-Model Horizontal Axis Hydrokinetic Turbine¡± Marine Energy Technology Symposium, 2014.
[6] V. Neary et al (2011) ¡°Field Measurements at River and Tidal Current Sites for Hydrokinetic Energy
Development: Best Practices Manual¡± Oak Ridge National Laboratory, ORNL/TM-2011/419
[7] R.J. Adrian and J. Westerweel. Particle Image Velocimetry. Cambridge: Cambridge University Press, 2011.
[8] Computer-Aided-Engineering tools for wind, wave, & tidal energy research & development. Northwest
National Marine Renewable Energy Center. https://github.com/nnmrec
Fig. 2 Visualization of the blade tip-vortex interaction via injecting bubbles into the flow.
Fig 3. Mean velocity and mean vorticity fields measured via Particle Image Velocimetry (PIV)
in the near wake of an axial-flow marine hydrokinetic turbine (the PIV fields of view are
shown to scale). The PIV identifies key features of the wake, including the blade tip-vortex
interaction, flow separation off the nacelle, and diffusion of the turbulent wake.
Fig 4. Comparison of the velocity fields measured via Particle Image Velocimetry (PIV), the
¡°1D-Momemtum¡± RANS model, and the ¡°actuator line¡± LES model.](https://image.slidesharecdn.com/95bf5a75-9537-43b8-82ae-ff070ae31813-170110050249/85/mets2016_poster_SaleAliseda-1-320.jpg)
More Related Content
What's hot (20)
Viewers also liked (17)
Similar to mets2016_poster_SaleAliseda (20)
mets2016_poster_SaleAliseda
- 1. Fig. 1 (top) solid model of the re-designed ¡°Reference Model 1¡± turbine as built, (bottom) configuration of the turbines in the Bamfield Marine Science Centre flume, and measured efficiency of the turbine array comparing to a variety of RANS and LES models Experiments and numerical simulations are conducted for an array of axial-flow Marine Hydro-Kinetic (MHK) turbines operating in a flume. This study aims to understand the influence of coherent structures in high Reynolds number wakes on energy extraction and dynamical rotor control processes. In the flume experiments, rotor torque and rotational position measurements are collected, and the flow field is characterized by Acoustic Doppler Velocimetry (ADV) and simultaneous imaging with Particle Image Velocimetry (PIV). The performance of 3 turbines are characterized under varying stream-wise spacing, cross- stream offsets, and varying yaw angles to the inflow. To study the effect of unsteady hydrodynamics, the turbines are outfitted with either open-loop or closed-loop feedback rotor-speed controllers, designed to maintain target tip-speed-ratios. The experimental results were compared against Computational Fluid Dynamics (CFD) simulations using different models to represent the effect of the rotor and for different strategies to simulate the turbulent flow field: both Large Eddy Simulation (LES) and Reynolds-Averaged Navier-Stokes (RANS). In the LES and RANS simulations, turbine models within the class of "actuator methods" are included which impose body forces upon the flow field in form of disks or lines. The laboratory scale MHK turbines (designed, manufactured and flume tested) demonstrated high efficiency and wake characteristics similar to the full scale reference model turbine. A marked improvement in efficiency was seen, although contributed by several factors such as: rotor-speed controller, Reynolds scaling, and blockage effects. Differences between the PIV, RANS, and LES confirm that the turbine models have certain strengths and shortcomings. Namely, the overset mesh method performs well to capture the wake of nacelle, while the LES does not include the nacelle effect and therefore predicts the wake inaccurately. Adding a turbulence inflow to the LES greatly improves the realism of the model, showing near wake breakdown and wake meandering, which was observed here and has been observed in other MHK and wind turbine studies. THE FLOW FIELD OF A TWO-BLADED HORIZONTAL AXIS TURBINE VIA COMPARISON OF LES AND RANS SIMULATIONS AGAINST EXPERIMENTAL PIV FLUME MEASUREMENTS Danny Sale and Alberto Aliseda (dsale@uw.edu, aaliseda@uw.edu) University of Washington, Mechanical Engineering, Seattle, WA, U.S.A. Proceedings of the 4th Marine Energy Technology Symposium, Washington, D.C., April 25-27 ACKNOWLEDGEMENTS Thanks to Ben Strom, Rob Cavagnaro, Dom Forbush, John Bates, Michelle Hickner and Brian Polagye for their help setting up and conducting the experiments. This work was financially supported by the National Science Foundation under the Sustainable Tidal Energy project, and the Naval Facilities Engineering Command through the Northwest National Marine Renewable Energy Center. Help from the Bamfield Marine Sciences Centre and specially Eric Clelland is greatly appreciated. This work was facilitated through the use of infrastructure provided by the Hyak supercomputer system at the University of Washington. [1] V. Neary, C. Hill, L. Chamorro, B. Gunawan, F. Sotiropoulos, ¡°Experimental Test Plan ¨C DOE Tidal and River Reference Turbines¡± Technical Report, Oak Ridge National Laboratory, 2012. [2] M.J. Churchfield, Y. Li, P.J. Moriarty. ¡°A large-eddy simulation study of wake propagation and power production in an array of tidal-current turbines¡±. Phil. Trans. R. Soc. A 28 February 2013 vol. 371 no. 1985 [3] Kornev, N. & Hassel, E. (2007). ¡°Method of random spots for generation of synthetic inhomogeneous turbulent fields with prescribed autocorrelation functions¡±. Communications in Numerical Methods Engineering, Vol. 23, Issue 1, pp. 35-43. [4] T Javaherchi, S Antheaume, A Aliseda, ¡°Hierarchical methodology for the numerical simulation of the flow field around and in the wake of horizontal axis wind turbines: Rotating reference frame, blade element method and actuator disk model¡±, Wind Engineering, 2014 [5] T. Javaherchi, J. Seydel, N. Stelzenmuller, and A. Aliseda ¡°Experimental and Numerical Analysis of a Scale-Model Horizontal Axis Hydrokinetic Turbine¡± Marine Energy Technology Symposium, 2014. [6] V. Neary et al (2011) ¡°Field Measurements at River and Tidal Current Sites for Hydrokinetic Energy Development: Best Practices Manual¡± Oak Ridge National Laboratory, ORNL/TM-2011/419 [7] R.J. Adrian and J. Westerweel. Particle Image Velocimetry. Cambridge: Cambridge University Press, 2011. [8] Computer-Aided-Engineering tools for wind, wave, & tidal energy research & development. Northwest National Marine Renewable Energy Center. https://github.com/nnmrec Fig. 2 Visualization of the blade tip-vortex interaction via injecting bubbles into the flow. Fig 3. Mean velocity and mean vorticity fields measured via Particle Image Velocimetry (PIV) in the near wake of an axial-flow marine hydrokinetic turbine (the PIV fields of view are shown to scale). The PIV identifies key features of the wake, including the blade tip-vortex interaction, flow separation off the nacelle, and diffusion of the turbulent wake. Fig 4. Comparison of the velocity fields measured via Particle Image Velocimetry (PIV), the ¡°1D-Momemtum¡± RANS model, and the ¡°actuator line¡± LES model.