A spatially nonlinear generalised actuator disk model for the simulation of horizontal axis wind and tidal turbines. (1st March 2020)

• Record Type: Journal Article
• Authors: Edmunds, Matt; Williams, Alison J.; Masters, Ian; Banerjee, Arindam; VanZwieten, James H.
• Is Part Of: Energy. Volume 194(2020)
• Publication Date: 2020-03-01
• Subjects: horizontal axis turbine -- Finite volume -- Hydrodynamics -- Aerodynamics -- Incompressible flow; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2019.116803
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: Efficient numerical simulation of renewable energy wind and tidal turbines is important for the layout of devices in farms. Computational Fluid Dynamics (CFD) approaches using blade geometry resolved models are computationally expensive. Therefore, most array models use source term representations of rotors, normally actuator disk, actuator line or blade element disk. Unfortunately, these methods rarely capture enough physics to accurately predict power and at the same time correctly characterise the wake velocity field and turbulent structures. This study describes a new Generalised Actuator Disk CFD model (GAD-CFD), that achieves the required accuracy for the simulation of horizontal axis wind and tidal turbines and their wakes. This new method combines a finite volume CFD code with additional source terms representing the rotor, including: correct consideration of losses along the foil by modification of the distribution of downwash; a concise downwash distribution computation; recognition that foil cross section varies along the length; dynamically changing Reynolds numbers and the application of a tip radius correction. Also reported are foil lift and drag coefficients and their variation with thickness, surface roughness and Reynolds number, which is necessary for the proper characterisation the whole rotor. The effectiveness of this approach is investigated and validated against two experiments, and demonstrates improvements over traditional source term methods, in particular the correct CFD approach to tip losses and consequent downstream wake prediction. This study provides confidence in application to both small scale flume studies and large scale array deployments in both the marine and wind environments.