3D Incompressible turbulent flow validation case
 

Dear all. I'm looking for a couple of three dimensional steady incompressble turbulent flow cases. My solver is based on density based but extent to the incompressible region using a preconditioning method. So far, For an incompressible laminar flow the solver has been validated. Now, I am about to validate the incompressible turbulent flow. Can you all give any advices ? Thanks in advance.

Yu

Re: 3D Incompressible turbulent flow validation ca
 

Dear Yu,

There are several incompressible test cases that can be simulated. The easier options are to use two-dimensional test cases themselves and extrude in the z-direction. Some typical problems are the flow past a Backward facing step and 3D turbulent lid-driven cavity. As a suggestion, you can check for the following points with respect to your code.

1. Ability to run with preconditioning on at higher Mach such as at M=2.

2. Mach independent convergence for M<0.1 for steady state problems.

3. Delta(p) ~ O(M^2)

4. Use of reference pressure to avoid round off errors at very low M (1e-5 and less). This could be a serious problem in 3D, as I have personally encountered the same.

The check (1) is more of a consistency check, showing a truly unified code while checks (2) and (3) are necessary to validate the preconditioned solver. You could refer to INRIA report on preconditioning (you can google and get it) for more details. (4) is more of a numerical trick and practical fix. These checks can done even for laminar flow.

Out of interest since I also have a preconditioned code, 1) What preconditioner do you employ ? 2) What is the lower bound on the Mach that your code goes for ? 3) What are the practical applications you are looking forward to employ the code for ?

Hope this helps

Regards,

Ganesh

Re: 3D Incompressible turbulent flow validation ca
 

Dear ganesh. Thanks for your reply. It really helps me and i will consider those points you suggested.

And my answers about your question are

1)My preconditioning code is applied using the preconditioner which Weiss and Smith suggested. If you want to know more detail, find this paper AIAA Journal, Vol. 33, No. 11,1995,pp 2050~2057

2)I don't have any specific region for low-speed so far. It depends on the practical applications. I've done the research on the flow-field around NACA0012 airfoil in Ma=0.00001 for checking the order of convergence.

3)I haven't decide yet what the applications would be proper. There would be lots of discussion on that part.

Regards.

Yu.

P.s: Do you have any suggestions about full three-dimensional validation cases for RANS model, i mean not a case two dimensional phenomena is dominant.