Prevent a phase to leave from a certain outlet among multiple outlets
 

Hi, I have a eulerian two-phase, 3d, laminar, pipe flow with separate inlets and outlets for both phases. In my set up, streamlines show that both phases attempt to leave from both the oulets. How to make them not leave from other outlet.
I suppose one way could be If phases can consider the other's outlets as walls. How can it be done? Using UDF or without UDF.

In this image, larger portion of the left side of the pipe is gas inlet and smaller bottom portion represents other denser phase outlet. Some of the coming in gas streamlines, returns in hurry from other oulet to out of domain. How do I prevent this. I am using velocity inlets and pressure outlets along with no slip at wall. Any comments are appreciated.

why shouldn't they escape for that outlet? it is very physical and natural for this to happen. what is your application? why dont you want it to escape from that outlet.

imagine two doors in a room, if you open both air will rush in/out from both, you cant expect one door to not having an flow through it.

Thanks for your reply. It is a counter current flow of two phases (gas and solids) in a rotating pipe. Gas has a certain axial velocity at its inlet. Solids too have a certain velocity from other side of pipe. Lower portion of the section in figure is for the solid particles as they need some outlet in that section. Similar structure is made on the other side of pipe. Physically also, entering gas with a certain velocity at inlet, why should it move backward unless velocity is too low. If lower portion in figure was a wall, it won't move backward in Fluent except few recirculations. I think in this set-up, gas looks to the nearest outlet there is, and moves out. I need to distinguish outlets for both phases. Its like, inlet for one phase is outlet for other phase.

One door on each of two opposite sides of a room, one takes air in and sand out. Other takes Sand in and air out. It can be made possible if room is inclined and also rotating.:cool:

Then you need to make the simulation inclined and rotating it cant predict by itself what you want it to do.

As for the reason that only a small portion of the air should go from that outlet. It is because of the initial conditions. Initialize by giving the simulation a headstart.

For example there will be a pressure difference between the two outlets which will cause one outlet to draw out much more air. there are many ways to do this. a few which i can tell from the back of my head are: patch the region with the air velocities as the initialization rather than just specifying the velocity at the inlet.

Or you can also calculate the mass flow rate of the air from the inlet, this mass flow can be specified at the target mass flow rate in the pressure outlet boundary condition options. Note that this will cause a pressure difference though between the two outlets. the outlet where you want the air to go towards will be at a lower pressure as it is sucking the air out.

Hope this helps.

Thanks for your precious time and explanation.
I will initialize with patching the regions with phase velocities. For rotation of pipe, I am using single moving reference frame. It is inclined (just 1 degree). According to user guide, target mass flow rate option is not available for multiphase flow. It is actually a rotary kiln problem. I will be updating my results.

Yes but you can specify a pressure, do you know what the pressure drops are from experimental results? putting them in will help a lot. or from the mass flows calculate the velocity at the inlet and the use a velocity inlet condition instead. so you can specify the target mass flow at the outlet.

Dear hwet,

I have inlet mass flow rates (or velocities) for both phases. I am using inlet velocities for both phases. Expected or target mass flow rates at outlets are also known. You mentioned that I can still specify target mass flow rates at outlets. But as it is multiphase flow, I cant see that option in pressure
outlet. Do you mean I should use velocity inlet instead of pressure outlet and put a negative velocity there.

I have not found the pressure information in experimental works I am referring. Pacthing axial velocitis for both phases gave better results in terms of convergence. Still phases seem to move towards opposite outlets.

regards,

not very sure what you are trying to do. if there are solids escaping from that outlet there will surely be gas escaping from it as well. maybe you will have to write a udf. i am not sure