[Mojtabam]

Hello, I am trying to simulate a horizontally homogeneous neutral atmospheric boundary layer, using the inflow conditions for the wind speed profile, turbulent kinetic energy and the turbulent dissipation rate, shown in the attachment - inlet boundary conditions. The results related to these simulations are shown in the attachments - Energy and wind speed (the three simulations are very similar). As can be seen, except for the wind speed profile, horizontal inhomogeneity is present in the turbulent kinetic energy. In my opinion, this is because, while the consistency is guaranteed directly between the upstream wind speed profile and the roughness characteristics at the ground, no direct treatment of the turbulence quantities at the ground is carried out based on their upstream values. Therefore, I believe that one way to improve the horizontal homogeneity of the turbulence quantities close to the ground surface is specifying directly the values of these quantities at the ground using their values from the upstream boundary. Considering this, I would like to know if there is any way to specify directly the values of turbulence quantities (I am using the quadratic pressure-strain Reynolds stress model) at the ground?

[ping]

i am aware from others that there have historically issues with accurate a.b.l. modelling but the help now has some detailed information - have you applied all the recommendations from this extensive topic: Simulating Physics > Guidelines for Applying Simcenter STAR-CCM+ > Atmospheric Boundary Layers

i would also have a good look on the steve support site and failing that contact your support engineer.

[Mojtabam]

Thank you Ping for your reply.



As you mentioned correctly, the simulation of a neutral ABL has long been a problem in CFD community. Additionally, several solutions have already been proposed to resolve this issue. However, these solutions are applicable to the two-equation based turbulence models, not the more complex Reynolds-Stress turbulence models, where there are six transport equations for each Reynolds stress term.

Regarding the recommendations provided by STAR-CCM+, I have applied all of them. These recommendations, again, are applicable mostly to two-equation based turbulence models, and also to the cases where the turbulent kinetic energy profile is uniform.

I believe that one way to improve the results is specifying directly the values of turbulent kinetic energy and its dissipation rate using their values from the upstream boundary. However, when a wall boundary condition is activated there is no direct way to do this... Maybe I have to add some source terms into the turbulent transport equations.