RANS Reynolds-stress model

WeFarm 2.0

Xiaodong Zhang's picture
Submitted by Xiaodong Zhang on May 5, 2015 - 10:41am
Main hypothesis

Almost the same as WeFarm 1.0, they both are general purpose CFD solvers modified for atmospheric boundary layer flow. This version uses OpenFoam 2.0.7. solve. Other differences with respect to WeFarm 1.0:

1. In the k-ε turbulence model, the model constant = 0.036 based on wind farm measurements.

2. Wall functions based on roughness height Kw are adopted.

WeFarm 1.0

Xiaodong Zhang's picture
Submitted by Xiaodong Zhang on May 5, 2015 - 10:39am
Main hypothesis

1. Steady state Atmospheric Boundary Layer flow. Incoming wind velocity profile is based on logarithmic distribution with stratification amendments for non-neutrual conditions. Different from surface layer, the mixing length L is defined as: 1/L = 1/z + 1/Lm + 1 / (zi - z), where z is evelation, zi is thickness of ABL and Lm is a middle length.  

2. No gravity and vertical pressure gradient for neutrual stratification. Potential temperature is adopted for non-neutrual conditions, and a reverse cap is simulated for the convective boundary layer.  

3. A logrithmic profile is used for turbulent flow wall function.