Natural Convection BC
I'm doing my final year project on Natural Convection in fin array.. I investigate the Natural Convection of the air domain. I created 3D volume of air which partial of the face of that air volume is fin wall and base which temp constant, i set th BC as below:
Fin wall Wall (constant temp 400K) Fin base Wall (constant temp 400K) Other surface = Pressure inlet and outlet randomly (both oso "0" presure) And Operation Condition = 101325 pascal, 300K temp and tick gravity(y= 9.81) Material = air with ideal gas Solution Initialization = all zero value Picture uploaded: http://img210.imageshack.us/img210/357/fin2b.jpg http://img402.imageshack.us/img402/5941/finmeshed.jpg But i got error ms as:  reversed flow in 650 faces on pressureoutlet 3. reversed flow in 1248 faces on pressureinlet 4. reversed flow in 5000 faces on pressureoutlet 5. 10 2.6736e01 2.8976e05 5.2343e02 2.4865e02 3.7876e08 0:00:00 0 No error handler available Error: FLUENT received a fatal signal (SEGMENTATION VIOLATION). Error Object: () Error message from graphics function Close_Segment: Nothing open  can't Close a 'segment'  Anyone tell me wat to do? I set the error bc? thanks... I 'm a beginer here, hope ur help... 
No one reply me?

natural convection BC
Quote:
i think you should use boussinesq model instead of ideal gas in the material selection panel of for air and set the appropriate thermal expansion coefficient.... 
Thanks rahul.. But izit ideal gas more suitable to which has large temp diff?
thermal expansion coefficient what value i should set? I jus set default value.. 
Hi, yipseng
Let us see how are the fins oriented , Are they horizontal of vertical W.R.T. gravity. Once this is clear. You can find out whether the natural convection flow is laminar or turbulent. In many cases where the flow is turbulent in natural convection cases it is tough to converge. In that case you can reduce you gravity to 100 times run it in stage to reach the value of 9.81 m^2/sec. Let me know the exact errors you encountered. Uday 
Picture uploaded.. Anyone noe what the problem??

I've got a problem with natural convection, too. The suggestion to check if flow is turbolent or laminar is surely good! In case of turbolent flow i suggest to use Keps models.
Hi! 
any 1 know how to solve? Thanks

i know but it is you project, so you have to done it yourself.
best regards. 
Dear Yapiseng, I happen to simulate the same model with you. But I set the fin base as heat flux and I consider the airfin coupled domain. And I also set BC as symmetric and pressureout. I did have solutions. I wonder if you have succeeded with a solution and what's the error. How much was that?
Maybe we could help each other. 
All times are GMT 4. The time now is 15:30. 