[macchesney]

Hello,

I have a plot here test1.PNG that shows the modelled turbulent viscosity ratio in the IDDES formulation in a backwards facing step geometry. As you can see once the flow enters the separated region, the turbulent viscosity ceases to be modelled and its effects are instead captured by the turbulent fluctuations in the flow with LES. Here you can see the turbulent intensity profiles generated by calculating v' / vref. test2.PNG.

My question is: is it possible to estimate the turbulent viscosity (ie. the ratio of apparent viscous forces due to turbulence with molecular viscosity) given the turbulence intensity?

Thanks for you interest!

[FMDenaro]

Originally Posted by macchesney

Hello,

I have a plot here Attachment 58611 that shows the modelled turbulent viscosity ratio in the IDDES formulation in a backwards facing step geometry. As you can see once the flow enters the separated region, the turbulent viscosity ceases to be modelled and its effects are instead captured by the turbulent fluctuations in the flow with LES. Here you can see the turbulent intensity profiles generated by calculating v' / vref. Attachment 58612.

My question is: is it possible to estimate the turbulent viscosity (ie. the ratio of apparent viscous forces due to turbulence with molecular viscosity) given the turbulence intensity?

Thanks for you interest!

If you are using an LES with the (static or dynamic) Smagorinsky model, in your code is computed the eddy viscosity field. You just need to access the data. Turbulent intensity alone cannot provied a full field.

[macchesney]

Hi there.. from my somewhat limited understanding, the idea behind the turbulent viscosity parameter is that turbulent fluids behave as if their viscosity has increased.. so in these models we substitute a region of turbulence with a fluid with an apparent viscosity. This is only done for the Smagorinsky RANS model however.. within DES we have RANS regions and LES regions.. therefore when I show the turbulent viscosity parameter in a plot, I can only see the turbulent viscosity parameter in the rans regions (inlet boundary layer). After the step LES takes over and the modelled turbulent viscosity disappears.. what I am actually interested in is what is the apparent viscosity of the entire turbulent region.. I am able to obtain the turbulence intensity from the LES region.. and was wondering if I can use this to estimate the apparent viscosity of the the turbulent region.. that ms for your reply.

[FMDenaro]

Originally Posted by macchesney

Hi there.. from my somewhat limited understanding, the idea behind the turbulent viscosity parameter is that turbulent fluids behave as if their viscosity has increased.. so in these models we substitute a region of turbulence with a fluid with an apparent viscosity. This is only done for the Smagorinsky RANS model however.. within DES we have RANS regions and LES regions.. therefore when I show the turbulent viscosity parameter in a plot, I can only see the turbulent viscosity parameter in the rans regions (inlet boundary layer). After the step LES takes over and the modelled turbulent viscosity disappears.. what I am actually interested in is what is the apparent viscosity of the entire turbulent region.. I am able to obtain the turbulence intensity from the LES region.. and was wondering if I can use this to estimate the apparent viscosity of the the turbulent region.. that ms for your reply.

No. In LES the SGS model can be based on the Bousinnesq as same as happens for the RANS formulation. If you have an LES with a Smagorinsky model you have a local eddy viscosity variable. What you have in the LES with the dynamic Smagorinsky model is that the eddy viscosity vanishes (correctly) in regions of laminar flow condition, for example close a wall for y+ =O(1).