CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   How to model an airflow over a water? (

Luk January 9, 2008 05:27

How to model an airflow over a water?
Hi all, I want to model very simple situation. The hot air blows over an cold very dense and viscous other liquid flow and heat them. The problem is that I dont want to model momentum transfer between air and the other fluid (just to avoid convergence problems since I know that the influence of air on the second dense fluid can be neglected). Because of that I dont want to model it in single domain like in "Flow over a bump tutorial". Time ago Glen from this forum suggested to build a model with two separated domains (one for air and second for this other fluid) and make interface. However my problem is that it is impossible to include one fluid for first of domains and another for second one. If I set 2 fluid in one domain automatically these fluid occures for second one and it starts all the problems with momentum transfer etc. Convergence is very slow. Can anyone explain how to set different fluid for different domains as there will be only one fluid in each of domains?

I would be very grateful, Luk

Glenn Horrocks January 9, 2008 18:03

Re: How to model an airflow over a water?

Is the fluid in the liquid domain moving? If it is not moving you could assume it to be a solid and do a standard CHT simulation.

Didn't somebody mention how to put different fluids in separate domains? If not ask you support office how to do this, I have not done it I just know it can be done.

Glenn Horrocks

CycLone January 9, 2008 18:11

Re: How to model an airflow over a water?
Hi Glenn,

You can put them in different domains, but there won't be any heat transfer between them. Although I wonder if you could put a GGI between them and remove the "conservative interface flux" condition for mass and momentum (by editing the CCL) and leave only the condition for energy?


Luk January 10, 2008 04:39

Re: How to model an airflow over a water?
Hi all, Unfortunately the second fluid is not that dense that I can treat it as solid body. My problem is that if I assume two domains with interface between them there are huge problems with convergence. My potential solution is similar to that proposed by CycLone - to assume no momentum transfer between them only energy transfer. The second solution would be to make 2 domains but only a single substance on each but I dont know how to do it. If I add a substance to one of the domains it is automaticaly added to second.


All times are GMT -4. The time now is 01:55.