BC - Wall Roughness High

Rahul_P1

Hi community 

once again, I've to ask you and please your advice!

Defining the BC for a WALL:
The 'Turbulence wall type' is chosen as 'Wall function'. Now I've got the opportunity to determine the 'Roughness height' and here we've got the blur!
Which kind of roughness is the correct one? Is it the surface quality (in german R_a, R_z ...) or is it the roughness for fluid mechanics, for example for pipe flow? In German it's called 'Sandrauheit k_s' - sorry couldn't find a translation for it.... 

Hope you understood what I mean 

Thanks for your help! 

Rahul_P1

The roughness height is determined for the surfaces (element faces) bounding a turbulent boundary layer, so it is the first one in your query,

}
TURBULENCE_WALL( 'no-slip wall' ) {
type = wall_function
roughness_height = 0.001
active_type = all
}
specifies that two surfaces of the element set are on a no-slip boundary, that a turbulence wall function with an average wall roughness height of 0.001 is to be used to capture the turbulent boundary layer, and that both surfaces are active.

I presume you are doing a pipe flow analysis, I however do not see much change in the results with this type of analysis, with or without the roughness height specified,

You can explore the other turbulence wall modelling options,

Turbulence wall specifies a turbulent boundary layer wall and the types of Type of the turbulence wall modelling in Acusolve
Low Reynolds number – Low Reynolds number damping function
Wall function – Turbulence wall function
Running avg wall function – Turbulence wall function based on running average field

Rahul_P1

Hi 

thanks for help!

No I ain't got a pipe flow analysis  - I've got a closed volume filled with air. The inlets are 5 circular holes. And there are 5 circular outlets. In- and outlets are located at the lower section of the volume. I analyse the flow streaming inside the volume, crossing the bottom plate and leaving through the outlets. The bottom plate owns boundry layer and it should be tubulent. Aside from in- and outlets the walls own a boundry layer. It's also interesting if some fluid streams up to the upper section of the volume.

This is why I ask for wall roughness, because the bottom plate also own some edging.

You wrote: 'that a turbulence wall function with an average wall roughness height of 0.001 is to be used to capture the turbulent boundary layer'.
Do I understand it right?: Using HyperWorks for meshing, the boundry layer is defined e.g. 5mm = 0.005m. Is it necessary to use a wall roughness high of 0.005??? Or is it unimportant?
another option would be to define the roughness high as the surface quality of polished steel (the bottom plate). Would be this maybe right?

Thanks again for your help 

Rahul_P1

DG_Student_TUHH

The 0.001 was in reference to the example deck portion, in flower brackets in my first post.

The general structure of turbulent boundary layer has laminar region near the wall (sub layer), then transition region and fully turbulent core flow. So your first grid point will decide which regime you have to resolve. If y+ is < 5 then your turbulence model will resolve the sub layer and if y+> 50 then it will resolve the turbulent core.

The roughness height is necessary if you are interested in frictional forces due to unsmooth wall surface. If you are having a smooth surface like polished stainless steel then there really may be no need to specify the height.

The 1st layer height and turbulence model will take care of your Turbulent boundary layer, so there is no need to define the wall roughness height (0.005, in your case).

Rahul_P1

Thank You!
Honestly, you are fantastic!  Thank you also for your patience! 

Please correct me a last time, if I'm wrong:
Because I'm using HyperMesh for meshing a 3D-CFD-Tetramesh (the core + the boundry layer), where I define the 1st layer height, it is not necessary to define the wall roughness height. Defining the height in AcuSolve would be the same as done twice the same thing?

Once again thank You for your patience!!!

Rahul_P1

DG_Student_TUHH

Yes, you really may not need to specify the roughness height in acusolve for your case,

Regarding the first layer height, if you are creating the volume meshing in HyperMesh, there is no need to define it in Acuconsole.
If suppose you are using Acuconsole for the volume mesh, try using the boundary layer options in meshing parameters, that would help you control the first element height.