CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Outlet Boundary Condition (http://www.cfd-online.com/Forums/cfx/109682-outlet-boundary-condition.html)

aledexe November 23, 2012 10:02

Outlet Boundary Condition
 
Hello everybody,

I am trying to estimate a pressure loss along a "pipe", which is actually an internal flow path with high pressure fluid. The geometry consists of a channel with a series of bends and restrictions, which at a certain point comes into a small cavity and then again proceeds with a almost straight channel (only one inlet/one outlet). I know both static pressure and mass flow rate at the inlet, but I have no information about the outlet; what are the possible combinations of boundary conditions I should use generally in this case? I noticed that imposing a flow rate in the outlet (as I have no sources in the system, it should be the same as in the inlet..) the solution definitely is not going anywhere...

Thanks in advance

vitulaaak November 23, 2012 16:23

Hello,
Use total pressure and temperature at inlet. You can use either mass flow or static pressure at the outlet. You can roughly estimate the static pressure by estimating the mass flow target and guess the loss of total pressure.

aledexe November 24, 2012 06:36

Do you think it's necessary to fix an inlet temperature? Why? Also, about the total pressure to impose, in this case should I calculate it based on the average velocity at inlet (from the known flow rate)?

ghorrocks November 25, 2012 05:13

I would model this with a mass flow rate at the inlet and a zero pressure at the outlet. If you do not know the total pressure then just use static pressure - make sure the difference is not significant before deciding this. Then the pressure loss along the gizmo will come from the inlet pressure.

Only model temperature if it is required. If temperature does not do anything interesting then leave it out and you will simplify the simulation marginally (ie no heat equation).

aledexe November 25, 2012 07:09

Thanks to you both for the answers. Anyway (for ghorrocks), I also thought to use a flow rate inlet condition combined with zero pressure at the oulet (I know this is a robust option..), but isn't this kind of setting affecting the pressure loss calculation, as the outlet pressure is not actually zero?
Also, when you talk about static or total pressure, you mean at the outlet? I can estimate the velocity at the outlet is not negligible in determining the total pressure anyway...

ghorrocks November 25, 2012 17:34

Quote:

but isn't this kind of setting affecting the pressure loss calculation, as the outlet pressure is not actually zero?
Do you have any physics which depends on pressure (eg compressible flow, cavitation)? If no, then it is only the relative pressure which is important and you can set it to any value you like at any point you like - but zero will minimise numerical round off errors.

As for whether it should be total or static pressure: The pressure gradient is essentially perpendicular to a wall for high Re flows. So this means for pipe-like flows the static pressure is essentially the same across the flow. So this means a static pressure outlet is usually the best choice. The total pressure one is more suited to more open flows where part of the flow has a high velocity jet. So this suggests a mass flow rate inlet and a static pressure outlet - and fortuntely this is the most robust BC configuration for most flows.

Also note you cannot specify a total pressure outlet - this is described in the documentation.

aledexe November 25, 2012 17:40

Thank you very much for the good explanation Ghorrocks!

Bytheway, you are right I cannot specify any total pressure at the outlet as explained in the modeling guide


All times are GMT -4. The time now is 06:40.