%0 Journal Article
%T Computational Actuator Disc Models for Wind and Tidal Applications
%A B. Johnson
%A J. Francis
%A J. Howe
%A J. Whitty
%J Journal of Renewable Energy
%D 2014
%I Hindawi Publishing Corporation
%R 10.1155/2014/172461
%X This paper details a computational fluid dynamic (CFD) study of a constantly loaded actuator disc model featuring different boundary conditions; these boundary conditions were defined to represent a channel and a duct flow. The simulations were carried out using the commercially available CFD software ANSYS-CFX. The data produced were compared to the one-dimensional (1D) momentum equation as well as previous numerical and experimental studies featuring porous discs in a channel flow. The actuator disc was modelled as a momentum loss using a resistance coefficient related to the thrust coefficient (). The model showed good agreement with the 1D momentum theory in terms of the velocity and pressure profiles. Less agreement was demonstrated when compared to previous numerical and empirical data in terms of velocity and turbulence characteristics in the far field. These models predicted a far larger velocity deficit and a turbulence peak further downstream. This study therefore demonstrates the usefulness of the duct boundary condition (for computational ease) for representing open channel flow when simulating far field effects as well as the importance of turbulence definition at the inlet. 1. Introduction The actuator disc method has been used together, with the Reynolds averaged Navier-Stokes (RANS) equations, for many years and for many applications including helicopter rotors [1], horizontal axis wind turbines [2], and horizontal axis tidal turbines [3, 4] alike. The actuator disc method represents a turbine as a simple disc of similar dimensions to the rotor and is used to approximate the forces applied to the flow. The forces are implemented as body loads or as negative momentum source terms on the flow as it passes through the disc. The actuator disc approximation has a number of benefits over modelling the full rotor geometry. The most significant benefit amongst these is the reduction in computational expense especially for multiple rotor simulations. Full rotor simulations require a fine mesh to capture the boundary layer and separation along the blade surface, as well as the solution of the unsteady compressible Navier-Stokes equations. A full transient rotor simulation is needed, allowing the rotor blades to rotate in order to capture the wake. The actuator disc method allows for coarser meshes to be used and the incompressible Navier-Stokes equations to be solved as long as the Mach number is below 0.3, for this study the mach number is below 0.00021. Additionally, steady-state solutions can be obtained, vastly reducing the computational
%U http://www.hindawi.com/journals/jre/2014/172461/