open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,

I am a bit confused about the Smagorinski model described in the user manual and would be grateful if one could clarify.

From my understanding from the manual, choosing Smagorinski model as the turbulence model implies that eddy viscosity is calculated (modelled) through the formulation/equation which is then being added to the molecular viscosity to form the effective viscosity. This effective viscosity is being put in the diffusion term of the RANS equations, irrespective of whether the eddies are larger or smaller than the mesh size.

So essentially this is different from Smagorinski model in LES with low-pass filter where eddies larger than mesh size are computed and eddies smaller than mesh size are modelled.

Am I right about this?

Besides, are there any requirements of y+ value for using Smagorinski model as the horizontal turbulence model (mixing length for vertical) with Nikuradse law of bottom friction and no-slip bottom boundary condition?

Thanks in advance!

www.opentelemac.org/index.php/assistance...turbulence-modelling

I just realised that there is a similar post in the forum. But it's not answered yet.

Hello,

This turbulence model was introduced a very long time ago in TELEMAC-2D and TELEMAC-3D before I arrived, I am not a specialist of it.

Anyway, I think you are right, it is a turbulence model for RANS and not the filter used in LES.

I am not sure there are any requirements for y+ using this Smagorinsky turbulence model in TELEMAC-3D over horizontal directions (or TELEMAC-2D). It seems to be often used with quite big elements so that large domains can be modeled.

Hope this helps,

Chi-Tuan