CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Profile Boundary conditions (http://www.cfd-online.com/Forums/cfx/111854-profile-boundary-conditions.html)

diamondx January 16, 2013 15:34

Profile Boundary conditions
 
Dear cfd-onliners,
I'm about to start my master degree, in some cases i might divide my simulation into two cases, and use an outlet region as an inlet condition in a new analysis. i have always been curious about this method since i'm not sure it can work correctly because:

When you perform a simulation with the first case, the OUTLET used, geometrically speaking, should be extended a little (right ?), otherwise problems with convergence will occur, and error like the reversed flow warning, or just because i often hear "OUTLET should be big"

If we extend the outlet, how should it be extended ? the first thing my mind come up with is to include a similar and simplified second case. why ? so we can respect and well anticipate the flow behavior...

The other way is to just add a big domain as outlet... if we do that, i'm not sure our variable in the outlet (velocity,pressure,temperature) will be correct. Why ?
In my aerodynamic class, i always learned that the flow follow a certain path-line, when doing external aerodynamics study of car for example, my teacher used to tell me the particle pushes the one behind it and it continues like that, hence the flow deviate around the car. SO what's next really implies change to the upstream...

i hope i was enough clear, really sorry for this long explanation. hoping i can get some help.

Thanks a lot in advance,

ghorrocks January 16, 2013 17:48

The outlet should not be big. In fact it is often easier when the outlet is small. But the outlet should be far enough downstream to not affect the results of the area being modelled, and not have any recirculations going through it.

The best way to do this extension of the outlet is by modelling whatever is actually downstream in the system you are modelling. If it is a long pipe then model more of the pipe. If exhausts to atmosphere, model a big atmosphere region.


All times are GMT -4. The time now is 11:59.