CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   Main CFD Forum (https://www.cfd-online.com/Forums/main/)
-   -   CFD-ACE compressible flow convergence problem (https://www.cfd-online.com/Forums/main/69106-cfd-ace-compressible-flow-convergence-problem.html)

DelphineL October 12, 2009 11:48

CFD-ACE compressible flow convergence problem
 
Hello everybody!

I would like to simulate a reservoir of water & air, heated from outside and observe the natural convection due to density change with temperature.

I use for that CFD-ACE

But I have convergence problem:

If I set the air & water density as constant, it converges, but I will never see natural convection
If I set the air compressible (ideal gas law) & the water constant, it converges. But the water is just a first step before a simulation with propane, so I would like to make it compressible
If I set the air with constant density & the water with piecewire linear with T, it does not converge.
==> all the other parameters are kept similar...

So the problem is obviously coming from the compressibility of the water...

Did anyone already try to simulate natural convection with 1 liquid and 1 gas on top of each other?

Thanks a lot in advance!

DelphineL October 21, 2009 06:50

Hello!

I know that nobody repplied, but, just in case someone could be interested by this question:
the problem is that CFD-ACE does not support a solution with 2 phases with changing density, especially because of a parameter called "reference density".
Therefore, the best is to set the reference density to the one of the liquid, and for the gas, add a flow source as a subroutine, that is linked to convection: like g*(rho - rho_ref)*V

I still need to implement this subroutine because I never did it, so if one of you already implemented a USOURCE, please let me know!


All times are GMT -4. The time now is 07:10.