CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   orifice type boundary condition (http://www.cfd-online.com/Forums/cfx/99370-orifice-type-boundary-condition.html)

antonio April 2, 2012 06:01

orifice type boundary condition
 
Dear all,
I am trying to simulate a rectangular channel with clean water and constant height at time t=0s in which is injected a mixture of water and sediments. I want to preserve a constant water height during the entire simulation, thereby by mass conservation law the flow rate that is injected must be equal to a flow rate that leaves the domain. I am thinking of doing this by inserting a hollow in my outlet boundary (that is in most of its extension a simple wall) and to prescribe there a flow rate equal to the inlet flow rate. Does anyone have tried a similar approach?Do you think that this can be done as I have described it?
Best regards

ghorrocks April 4, 2012 07:29

You cannot define a flow with a flow rate in and out, it is not well posed. You need to prescribe pressure somewhere.

Have you considered doing a single phase simulation with the free surface replaced with a pressure boundary?

antonio April 6, 2012 07:22

Dear Glenn/All
at the present moment I am trying, following your recommendation, the option of replacing the free surface by a pressure boundary. I am using an opening boundary condition (entrainment option) with relative pressure = 0. In my specific case, I do not have turbulence at this specific boundary. Which way in cfx do you consider the best to specify zero level of turbulence ?I think zero gradient could be an option...Any opinions?Many thanks

ghorrocks April 15, 2012 18:03

The turbulence BC is probably not very important. You should do a sensitivity analysis to check but you will probably find it does not matter.

antonio April 20, 2012 04:45

Dear Glenn, I am having some problems in my simulation. At the end of 8 seconds of simulation walls are erected at an "outlet" (in the totality of the corresponding area). Basically I am using the following boundary conditions:
inlet-> velocity
top of domain-> i am replacing my free surface with an opening boundary with the option entrainment (with the opening pressure option selected). I am prescribing a relative pressure =0
lateral walls and bottom->no slip smooth wall
downstream section --> the vast majority of this section is a a no-slip smooth wall however in order to mantain the initial water level constant i have incribed in this section an "orifice" at atmospheric pressure (i have tried the option static pressure and average static pressure=0) .
Do you have any ideia why walls are being erected?
Regards.

antonio April 20, 2012 05:55

Could this be due to the fact that in my downstream section I have an outlet in which the flow will be tangent to it (please see attached by outlet boundary) ? I read something about this here:
http://www.cfd-online.com/Forums/cfx...yant-flow.html

http://img521.imageshack.us/img521/6...ion.jpghttp://

antonio April 20, 2012 05:57

http://img521.imageshack.us/img521/6...eamsection.jpg

ghorrocks April 20, 2012 07:47

I do not understand your question.

But the walls at the outlet are generated to stop back flow. So something is generating the backflow. It could be a convergence problem, or it could be real.

antonio April 20, 2012 08:35

Thanks Glenn.
Sorry, I will try to be more specific. The draw that you see above is my downstream section. The vast majority of this section it is in fact a wall, however the orificio section it is in contact with atmosphere (imagine that you have a wall with an oriffice). Hence I have specified the region "orificio" as an outlet with pressure=0. At the moment I am receiving a message saying something like "A wall has been placed in portion(s) of an OUTLET". I do not see any reason for this...the flow in the "orificio" region should "point" out of the domain...I do not think itīs pysical this "feedback effect" at "orificio". This articial wall is affecting the global quality of my results...Any suggestion/idea?

ghorrocks April 21, 2012 07:29

Use the post processor to find what the flow is doing. Zoom into the boundary to see the back flow which has been stopped. That should give hints as to what the problem is.


All times are GMT -4. The time now is 01:33.