RANS eddy viscosity


Daniel Cabezon's picture
Submitted by Daniel Cabezon on May 5, 2015 - 9:59am
Main hypothesis

The model derives from a previous elliptic model and it is inspired on the parabolic technique of other models such as UPMPARK and Windfarmer but using the actuator disk technique to represent the wind turbine instead of wind speed deficit. 

The wind turbine is represented as an actuator disk uniformly loaded. This means that the wind turbine acts as a sink of momentum, associated to the drag force exerted over the incoming flow. The reference wind speed for each disk is initially calculated from the wind speed at the position of the disk and corrected through the method proposed by Calaf et.al.

The solution algorithm consists of a decomposition of the domain into a finite number of adjacent subdomains that are solved sequentially in the axial direction, using the output of each subdomain as input for the next one. This is done until the end of the domain is reached. This way the computational time becomes significantly lower in comparison to the solution of a single domain by means of a purely elliptic approach.

VestasFOAM 1.1.0 - Steady

Submitted by Yavor Hristov on May 5, 2015 - 12:00am
Main hypothesis

The basis of VestasFOAM is built upon the simpleFoam solver distributed with the publically available OpenFOAM release and suitable for steady, incompressible, turbulent and isothermal flows. This solver and associated k-epsilon turbulence model [2] have been expanded to include appropriate boundary conditions for ABL flow [3], Coriolis force, Durbin realizability constraint [4], extension to stratified flows as wells as canopy [5] and buoyancy source terms. Meshes are automatically generated with Pointwise [6] using structured hyperbolic extrusion ensuring the highest possible quality. In order to reduce the mesh size, the horizontal mesh goes through a step-wise reduction in resolution with height once sufficiently away from the terrain. This creates a hybrid mesh with "hanging-node" type architecture maximizing efficiency without sacrificing mesh quality near the terrain and turbines.


Submitted by Carlos Peralta on May 4, 2015 - 6:02pm
Main hypothesis

Steady state solver based on OpenFOAM's simpleFoam (version 2.1.1). Isotropic eddy-viscosity turbulence using the Boussinesq approximation, homogeneous forest canopy and actuator disk solver.

EllipSys3D ABL

Tilman Koblitz's picture
Submitted by Tilman Koblitz on May 4, 2015 - 5:49pm
Main hypothesis

The EllipSys3D code is a multiblock finite volume discretization of the incompressible Reynolds Averaged Navier-Stokes (RANS) equations in general curvilinear coordinates.  The code uses a collocated variable arrangement, and Rhie/Chow interpolation [iv] is used to avoid odd/even pressure decoupling. As the code solves the incompressible flow equations, no equation of state exists for the pressure, and the SIMPLE algorithm of [v] is used to enforce the pressure/velocity coupling. The EllipSys3D code is parallelized with MPI for executions on distributed memory machines, using a non-overlapping domain decomposition technique.

The solution is advanced in time using a 2nd order iterative time-stepping (or dual time-stepping) method.  In each global time-step the equations are solved in an iterative manner, using under-relaxation. First, the momentum equations are used as a predictor to advance the solution in time.  At this point in the computation the flowfield will not fulfil the continuity equation. The rewritten continuity equation (the so called pressure correction equation) is used as a corrector making the predicted flowfield satisfy the continuity constraint.  This two step procedure corresponds to a single sub-iteration, and the process is repeated until a convergent solution is obtained for the timestep. When a convergent solution is obtained, the variables are updated, and we continue with the next timestep.

The three momentum equations are solved decoupled using a red/black Gauss-Seidel point solver. The solution of the Poisson system arising from the pressure correction equation is  accelerated using a multigrid method. In order to accelerate the overall algorithm, a three level grid sequence and local time stepping are used.

CRES-flow NS

John Prospathopoulos's picture
Submitted by John Prospathopoulos on May 4, 2015 - 5:22pm
Main hypothesis

CRES-flow NS is an in-house RANS solver using the k-ω turbulence model for closure and the actuator disk theory for the simulation of the embedded wind turbines. The momentum equations are numerically integrated introducing a matrix-free pressure correction algorithm which maintains the compatibility of the velocity and pressure field corrections. Discretization is performed with a finite volume technique using a body-fitted coordinate transformation on a structured curvilinear mesh. Convection terns are handled by a second order upwind scheme bounded through a limiter, whereas centred second order schemes are employed for the diffusion terms. Velocity-pressure decoupling is prevented by a linear fourth order dissipation term added into the continuity equation. The k-ω turbulence model has been suitably modified for atmospheric conditions. Stratification is considered through an additional production term added to each one of the k and ω transport equations to account for the buoyancy effect.