I'm doing LES calculations for backward-facing step. Smagorinsky Subgrid model is being used.

In terms of wall treatment, I have choices between scalable and automatic. I do understand what scalable wall function is. But I wonder what automatic subgrid scale model do in LES. Please advise me if anyone has an idea.

Ideally, I would like to turn off wall functions and resolve detailed flows including laminar sublayer. Is there any way to turn off wall functions?

Regards, Sungho

Hi Sungho

If you want to resolve detailed flow including laminar sublayer then you should ideally have YPlus equal to 1 and use the automatic option for wall functions. I think if you use automatic option, you can consider the wall functions turned off. However, this is important only for heat transfer problems. It should not make much difference for other cases.

Can I ask what are you using to validate your results. Is this an article or experimental data. Which paper or experimental data are you using and what are your boundary conditions. I am also interested in setting up this case. Thanks

Reagrds

Thank you.

I do have all the results with the scalable wall function (with yplus<1). I'm doing calculations with automatic option. I will compare the results to see whether it affects a lot.

By the way, my case if from

- S. Jovic, and D. M. Driver, Backward-facing step Measurements at low reynolds number, Reh=5000, NASA technical memorandum 108807

- Hung Le, Direct Numerical Simulation of Turbulent Flow over a Backward-facing step, PhD thesis, Stanford University, 1995.

Hi Sungho

Thank you for your answer. Did you use periodic boundary between inlet and outlet to generate turbulence. I do not have access to thesis, can you send me a copy please. As I said I also want to set up this case. I hope you can help.

Regards

I do have the thesis and papers but not in electronic version.

If you can't access the thesis, then alternatively, you can check the following paper

H. Le, P. Moin, and J. Kim "Direct Numerical Simulation of Turbulent Flow over a Backward-Facing step", 1997, J. Fluid Mech., vol. 330, pp 349-374.

At the moment, I didn't generate turbulent at the inlet. However, it seems flow becomes turbulent spontaneously just after the step.

Hi Sungho

Did you use normal inlet and outlet? How many flow through (residence times) did you run your case for. My feeling is that we can not have a LES solution unless we generate proper turbulence in our domain. One way of generating turbulence would be, ofcourse, to use periodic boundary at the inlet and outlet.

How did you set up your case? Did you get good results for your rms plots? Thanks in advance.

Regards

Hi,

Yes. I used normal inlet and outlet conditions. I ran 4 flow through times and gathered statistics from the last 2 flow through times.

The result was very good without generating turbulence at the inlet partly because, as I said, the backward facing step, with the given Reynolds number, ensures with that flow becomes turbulent spontaneously after the step.

I'm running more flow through times.

And I'm also interested in generating inlet turbulence. I know that stanford guys did auxilary flat plate calculations to generate inlet turbulence.

I'm not sure whether you can use the inlet/exit periodic conditions because the heights (h1 and h2) are different between the inlet and exit.

|-----------------------------| h1 | |

|------| | h2

| |

|______________________|

I also think it would be a good idea to communicated by personal emails if we want to discuss further in detail.

Regards, Sungho

The figure of my backward facing step became distorted. What I meant was inlet height is different from the exit height.