CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   fixed pressure (http://www.cfd-online.com/Forums/main/1125-fixed-pressure.html)

john August 4, 1999 02:38

fixed pressure
 
I am working on incomressible fluid flow problem with the following specifications. Inlet B.C. :fixed velocity boundary condition at 6 m/s with turbulence intensity of 2% OUTLET B.C : fixed pressure boundary condition at 0 pressure with the same turbulence. Problem: At the outlet there is INFLOW. Any Explanation. Geometry : Like Heat Exchanger Baffles are placed in the flow domain.

John C. Chien August 4, 1999 04:07

Re: fixed pressure
 
(1). You need to find out the real flow pattern first. (2). The flow pattern is determined by your problem specification, which is a strong function of the problem geometry. (3). It is likely that the outlet is not placed far downstream and there is a flow separation there. (4). So, move the outlet condition further downstream to avoid the complication first. (5). In the pressure-based solver, the solution can also run into reverse flow during iteration. But it should disappear by the time the solution converges, if the exit boundary condition is placed properly further downstream.

Jin Wook LEE August 4, 1999 08:27

Re: fixed pressure
 
It is best to place the outlet far downstream, but sometimes it is impossible. In that case, you should check the possibility of reverse flow in the real flow system. For special cases, inflow through the boundary, including outlet, is possible correct solution. In that case, outlet is not a real outlet but a mixed inflow/outflow boundary.

Hoping that it might be helpful to you.

Sincerely, Jin Wook LEE

Biao Zhou August 4, 1999 10:21

Re: fixed pressure
 
The real flow pattern is dependent on the specific geometry and sometime reverse flow from outlet is expected for some engineering application, for example, some burner.

You also can check if your boundary is set correctly with the following step:

(1) calculate a short case; (2) calculate a long case; (3) compare the presure and velocity fields between the two cases for some specified cross-sections. If the boundary is right, then the results should be same or very closed.


clifford bradford August 6, 1999 00:34

Re: fixed pressure
 
for your inlet you need 2 bc other than the turbulence intensity. also as Jin said the reversed flow may be physically realistic in which case your code needs to check whether the boundary is inflow or outflow if it is outflow your one bc is adequate if it is inflow you must then specify another bc(?).another option may be to use nonreflecting bc at the exit which may allow the reversed flow.


All times are GMT -4. The time now is 06:59.