[vavnoon]

Dear friends,
I have a basic question about validation. If I want to simulate a laminar flow in a tank, will I use a turbulent case for validation of my simulation procedure? Exactly I would like to check my scheme, sliding mesh, for fluid motion.

I would be so thankful if you guide me in this regard.

[aerosayan]

Originally Posted by vavnoon

If I want to simulate a laminar flow in a tank, will I use a turbulent case for validation of my simulation procedure.

Depends on what you're trying to study really.
If there is a very high difference between the results of the laminar and turbulent cases, you probably can't use a laminar simulation to validate a transient simulation's result, but you most probably could validate your laminar result from a previously validated transient result.


But in some cases you can get away with it.
Say you're studying the pressure drop in a pipe which has a sudden change in diameter.
There will be a lot of turbulence in this model, so you decided to solve it with RANS or LES. We know that the LES simulation will be able to resolve the turbulent losses better than the RANS simulation. However you know from previous literature that the turbulent losses are negligible, so you can validate your RANS simulation with another author's published LES simulation (Since we only use published papers or experiments to validate our data).


Try to plot the data you want to analyze (let's say CL/CD) along with the published data. If both plots agree with each other, then you've validated your data.
If the plots don't agree, then the turbulence is pretty significant to properly simulate the case.
If the plots agree, then the turbulence probably has very little effect on the final result.



Regards

[vavnoon]

Thank you for your comment. I think it is a good idea.