1) The distinction appears from your choice of grid refinement close to the wall. Of course, then a proper SGS model have to be used ( Enhanced Wall Functions in https://www.sharcnet.ca/Software/Fluent6/html/ug/node514.htm).

2) see 1)

3) yes, any CFD code that solves the unsteady 3D form of the NS equations can perform a DNS.

Thank you Filippo.

Hi Joemon,

I can answer your last question.

Applying a very fine grid does not necessarily guarantees that you can use Fluent or any other CFD solver to solve a specific flow near the wall, and especially with an extremely highly computationally demanding turbulence modelling approach/model such as the direct numerical simulation (DNS).

Actually, DNS was deemed by some researchers very impractical for some simulating flow in some applications, such as indoor ventilation and airflow, which is my field.

So, my short answer is: maybe this approach can work for resolving wall bounded models DEPENDING ON YOUR CASE STUDY, MODEL SIZE, and RESULTS SOUGHT. But what I'm sure of that the runs will be extremely lengthy in terms of time (expect months maybe!)

You can also refer for the last table (Table 3) in my recent review article for more information about different turbulence models:

http://www.tandfonline.com/eprint/69yQsAit7nFHJ27dTScj/full

Regards,

Hossam

Hello Hossam

I too have the same view on Ansys Fluent and DNS capabilities. LES with SGS models are available and no direct mentioning of DNS in Fluent options.

@Filippo, did you mean the no SGS model LES with a Kolmogrov scale mesh is equivalent to DNS in Ansys Fluent.

Yes, it MUST be a realizable simulation (apart for the available computational resources) for any bug-free written CFD code and consistent numerical discretization of the NS equations.

In Fluent you can simply use the laminar flow assumption, however I suggest setting always a second order time and centred space discretization.